<%BANNER%>

Impact of Surface Chemistry on Adsorption

Permanent Link: http://ufdc.ufl.edu/UFE0021222/00001

Material Information

Title: Impact of Surface Chemistry on Adsorption Tailoring of Activated Carbon
Physical Description: 1 online resource (132 p.)
Language: english
Creator: Bach, Morgana T
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: activated, activation, carbon, chemistry, dissolved, oxygen, ph, surface
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The present work summarizes research into the impact of the surface chemistry of activated carbon on the water treatment phenomena of pH excursions, the tailoring of activated carbon through manipulation of the dissolved oxygen (DO) content of the water used to create steam for activation, and, most importantly, the impact of these changes in surface chemistry on adsorption. The increase in total surface acidity and some individual acidic functional groups associated with air treatments of reactivated carbons were reaffirmed as the mechanism by which pH excursions are overcome with air treatments. The proposal that carbon dioxide (CO_2) treatments overcome pH excursions via physisorption, and subsequent desorption, of CO_2 forming carbonic acid upon water immersion was supported through an analysis of mass gained during CO_2 treatment. Both air and CO_2 treatments were not found to suppress the calcium chemistry mechanism responsible for pH excursions. In a comparison of the adsorption performance of the treated carbons for 2-methylisoborneol (MIB), a common taste- and odor-causing compound, the untreated carbon removed less MIB than the air-treated carbon in water containing natural organic matter (NOM) due to decreased adsorption of the competitive NOM with the air-treated carbon related to its increased surface acidity. CO_2 treatment was proven effective, as shown by its increased removal of MIB, as it deterred water adsorption, which is deleterious to MIB adsorption. In a comparison of adsorption performance for phenol, both the untreated and CO_2-treated carbons performed similarly while the air-treated carbon did not perform as well due to its increased surface acidity. A novel approach to tailoring activated carbon via varying levels of DO in the water used to create steam for the activation of a wood-based precursor was evaluated using the Box-Behnken design of experiments. The statistical analysis confirmed the impact of changes in DO on the performance of the resulting carbon for adsorption of MIB and supported variation in DO as a possible means of tailoring activated carbon.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Morgana T Bach.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021222:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021222/00001

Material Information

Title: Impact of Surface Chemistry on Adsorption Tailoring of Activated Carbon
Physical Description: 1 online resource (132 p.)
Language: english
Creator: Bach, Morgana T
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2007

Subjects

Subjects / Keywords: activated, activation, carbon, chemistry, dissolved, oxygen, ph, surface
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The present work summarizes research into the impact of the surface chemistry of activated carbon on the water treatment phenomena of pH excursions, the tailoring of activated carbon through manipulation of the dissolved oxygen (DO) content of the water used to create steam for activation, and, most importantly, the impact of these changes in surface chemistry on adsorption. The increase in total surface acidity and some individual acidic functional groups associated with air treatments of reactivated carbons were reaffirmed as the mechanism by which pH excursions are overcome with air treatments. The proposal that carbon dioxide (CO_2) treatments overcome pH excursions via physisorption, and subsequent desorption, of CO_2 forming carbonic acid upon water immersion was supported through an analysis of mass gained during CO_2 treatment. Both air and CO_2 treatments were not found to suppress the calcium chemistry mechanism responsible for pH excursions. In a comparison of the adsorption performance of the treated carbons for 2-methylisoborneol (MIB), a common taste- and odor-causing compound, the untreated carbon removed less MIB than the air-treated carbon in water containing natural organic matter (NOM) due to decreased adsorption of the competitive NOM with the air-treated carbon related to its increased surface acidity. CO_2 treatment was proven effective, as shown by its increased removal of MIB, as it deterred water adsorption, which is deleterious to MIB adsorption. In a comparison of adsorption performance for phenol, both the untreated and CO_2-treated carbons performed similarly while the air-treated carbon did not perform as well due to its increased surface acidity. A novel approach to tailoring activated carbon via varying levels of DO in the water used to create steam for the activation of a wood-based precursor was evaluated using the Box-Behnken design of experiments. The statistical analysis confirmed the impact of changes in DO on the performance of the resulting carbon for adsorption of MIB and supported variation in DO as a possible means of tailoring activated carbon.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Morgana T Bach.
Thesis: Thesis (Ph.D.)--University of Florida, 2007.
Local: Adviser: Mazyck, David W.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2007
System ID: UFE0021222:00001


This item has the following downloads:


Full Text





IMPACT OF SURFACE CHEMISTRY ON ADSORPTION: TAILORING OF ACTIVATED
CARBON




















By

MORGANA T. BACH


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

2007

































Copyright 2007 Morgana Bach


































To my mother and sister









ACKNOWLEDGMENTS

I thank my mom, Teresa Bach, and my sister, Rhiannon Bach, for their love and support

over the many years I have pursued my academic career. I would also like to thank Link

Charlott for his support and respect of my accomplishments in the academic world.

I thank David Mazyck for his support and advice for the last five years. His tendency to

think "outside of the box" and his insight into the academic and business world is admirable. I

consider myself lucky to have had him as an advisor and I am surely a better person for it.












TABLE OF CONTENTS


page

ACKNOWLEDGMENT S .............. ...............4.....


LIST OF TABLES ................ ...............8............ ....


LI ST OF FIGURE S .............. ...............9.....


AB S TRAC T ............._. .......... ..............._ 10...


CHAPTER


1 INTRODUCTION ................. ...............12......... .....


2 LITERATURE REVIEW ..........._..._ ...............13.......__......


Activation of Activated Carbon ................ ............... ..............................13
G asification................ .... ...........1
Creation of Functional Groups ................. ...............15.......... ....
Impact of Activation Parameters ................. ......... ...............17. ...
Type of Precursor ................. ...............17................
Type of Oxidant ......................... ... .............2
Time, Temperature, and Oxidant Flow ................. ...............23........... ...
Ad sorption .............. .. ...............25...
Physical Adsorption............... ....... .. ........2
Effect of Surface Chemistry on Adsorption .............. ...............27....
Current Uses of Activated Carbon. ................ ................... .........................29
Reactivation of Activated Carbon ................ ......... ........ ......... ................30

pH Excursions in Water Treatment ................ .. ..... ...............3
pH Excursion Mechanisms for Activated Carbons .............. ...............33....
Methods to Overcome pH Excursions ................. ............ ... ............... 36....
Influence of Dissolved Oxygen in Water used to Create Steam ................. .....................40

3 HYPOTHESIS AND OBJECTIVES ................. ...............44................


pH Excursions ................. ...............44.................
Hypothesis .............. ...............44....
Obj ectives .............. ..... .... ..._.. .... .. ........ ......... .......4
Tailoring of Activated Carbon Using Dissolved Oxygen (DO) ................ ......................45
Hypothesis .............. ...............45....
Obj ectives ................. ...............45....... ......

4 EXPERIMENTAL METHODS .............. ...............46....


Reactivated Carbon Samples for Overcoming pH Excursions ......___............. ...... .........46
Carbon Precursor for Activations .............. ...............46....











Air Treatments and CO2 Treatments of Reactivated Carbons ........__. ........ _.._.............47
Activations with Varying DO Content in Water .............. ...............47....
W ater Contact pH .................. ...............48..
Calcium Solution Concentration............... .............4
Adsorption Experiments (Batch) .................. ...............49..
Rapid Small-Scale Column Tests (RSSCTs)............... ...............50
BET Surface Area Measurements .............. ...............51....
Pore Size Distribution Measurement ................. ...............52........ .....
pH at the Point of Zero Charge ........._._.. ........... ...............52...
Boehm Titration............ .............. ...............5
Statistical Design of Experiments: Box Behnken ....__. ................. .. ....._...........5


5 STRATEGIES FOR OVERCOMING pH EXCURSIONS FOR REACTIVATED
GRANULAR ACTIVATED CARBON: AIR AND CARBON DIOXIDE
TREAT MENT S ................. ...............58.......... ......


Creation of a pH Stable Carbon............... ...............58.
Surface Acidity of Air-Treated Carbons. ................ ....................... ................59
Surface Acidity of CO2 Treated Carbons .............. ...............60....
Verification of Mechanisms .............. ...............61....
Summary and Conclusions .............. ...............63....

6 METHODOLOGIES FOR OVERCOMING PH EXCURSIONS FOR REACTIVATED
GRANULAR ACTIVATED CARBON: EFFECT ON ADSORPTION
PERF ORMANCE ................. ...............67.................


Impact of Wetting Time on MIB Adsorption for CO2 Treated Carbon .............. ... ........._...70
MIB Adsorption from Natural Water .............. ...............71....
Phenol Adsorption from Deionized Water .............. ...............71....
Conclusions............... ..............7

7 IMPACT OF DISSOLVED OXYGEN ON ACTIVATION OF POWDERED
ACTIVATED CARBON............... ...............77.

Dissolved Gases in Solution .............. ...............78....
Box-Behnken Approach and Mass Loss................ ...............81.
M IB Removal .............. ...............84....
Physical Properties .............. ...............86....
Surface Chemistry .............. ...... .. .. .. ... .. .......8
Chemical versus Physical Methods of Removing DO ................. .............................93
Phenol Removal ................. ...............95.......... ......
Summary and Conclusions .............. ...............96....

8 CONTRIBUTIONS TO SCIENCE ................. ......... ...............105 ....


9 FUTURE WORK ................. ...............106................











APPENDIX

A SAMPLE CALCULATIONS FOR DISSOLVED GASES .............. ......................0

B PRELIMINARY STUDIES FOR DETERMINATION OF LEVELS FOR BOX
BEHNKEN ................. ...............110......... ......

C BOX-BEHNKEN RE SULT S ................. ...............114...............

LIST OF REFERENCES ................ ...............123................

BIOGRAPHICAL SKETCH ................. ...............132......... ......











LIST OF TABLES


Table page


4-1. RSSCT and full-scale design parameters .............. ...............56....

4-2. 22 full factorial .............. ...............57....

4-3. Box-Behnken 33 fractional factorial .............. ...............57....

5-1. Calcium in solution with treatments ................. ...............64...........

6-1. Characteristics of sample carbons .............. ...............58....

6-2. DOC remaining in solution after contact with 20 ppm carbon............... ...............74.

7-1. Concentration of dissolved gases in water at 273 K............... ...............98...

7-2. Levels of four factors employed in the Box-Behnken design .............. .....................9

7-3. Box-Behnken statistical results for mass loss. ................ ......... ........ ................98

7-4. Surface area versus treatment time, treatment temperature, and SC .................. ...............99

7-5. Box-Behnken statistical results for MIB removal .............. ...............99....

7-6. BET surface area and total pore volume of four carbons ......___ .......__ .............99

7-7. Box-Behnken statistical results for pHpze.....__ ....._____ ........___ ..........10

7-8. Activation conditions of four carbons and their removal of MIB in dichloromethane .......100

7-9. BET surface area and total pore volume carbons treated with sodium sulfite ....................100

7-10. Box-Behnken statistical results for phenol removal ................. ................ ......... .100











LIST OF FIGURES


Fiare page

2-1. Proposed mechanism for the formation of COOH on an activated carbon surface ..............42

2-2. Full-scale water treatment plant effluent pH and calcium concentration data after
reactivated GAC was returned to service............... ...............42

2-3. Comparison of performance in the removal of MIB for reactivated carbons created with
varying DO content in the water used to create steam. ......_.__ ........__ ........._....43

4-1. Box-Behnken design of 33............. ...............57...

5-1. Effect of air exposure time and temperature on water contact pH ................. ............._..64

5-2. Effect of treatment time and gas flow on resulting water contact pH of CO2 treated
carbon ................. ...............65.................

5-3. Effect of air treatments after 60-minute contact time on individual acidic functional
group s ................ ...............65.................

5-4. Effect of CO2 treatments on individual functional groups. ............. .....................6

6-1. Percent removal of MIB in DI water versus dose of carbon (24-hr contact time) ........._......74

6-2. Percent removal of MIB in raw water versus dose of carbon (24-hr contact time). .............75

6-3. Percent MIB removal versus wetting time in DI water with a CO2 and untreated
carbons .............. ...............75....

6-4. Effluent phenol concentration in DI water versus BV treated with a CO2 treated, air
treated, and an untreated carbon. ............. ...............76.....

7-1. Mass loss versus time and temperature .............. ...............101....

7-2. Mass loss versus SC and temperature............... ..............10

7-3. MIB removal versus time and temperature ..........._......_ ....__ ..............102

7-4. MIB removal versus SC and DO ..........._...... ._ ...............102

7-5. Pore size distributions of sample carbons............... ...............103

7-6. pHpzc versus SC and temperature............... ..............10

7-7. PSDs of carbons treated with sodium sulfite ....__ ......_____ .......___ .........10

7-8. Phenol removal versus time and temperature ....__ ......_____ .......___ ...........0









Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IMPACT OF SURFACE CHEMISTRY ON ADSORPTION: TAILORING OF ACTIVATED
CARBON

By

Morgana T. Bach

August 2007

Chair: David W. Mazyck
Major: Environmental Engineering Sciences

The present work summarizes research into the impact of the surface chemistry of

activated carbon on the water treatment phenomena of pH excursions, the tailoring of activated

carbon through manipulation of the dissolved oxygen (DO) content of the water used to create

steam for activation, and, most importantly, the impact of these changes in surface chemistry on

adsorption.

The increase in total surface acidity and some individual acidic functional groups

associated with air treatments of reactivated carbons were reaffirmed as the mechanism by which

pH excursions are overcome with air treatments. The proposal that carbon dioxide (CO2)

treatments overcome pH excursions via physisorption, and subsequent desorption, of CO2

forming carbonic acid upon water immersion was supported through an analysis of mass gained

during CO2 treatment. Both air and CO2 treatments were not found to suppress the calcium

chemistry mechanism responsible for pH excursions. In a comparison of the adsorption

performance of the treated carbons for 2-methylisoborneol (MIB), a common taste- and odor-

causing compound, the untreated carbon removed less MIB than the air-treated carbon in water

containing natural organic matter (NOM) due to decreased adsorption of the competitive NOM

with the air-treated carbon related to its increased surface acidity. CO2 treatment was proven









effective, as shown by its increased removal of MIB, as it deterred water adsorption, which is

deleterious to MIB adsorption. In a comparison of adsorption performance for phenol, both the

untreated and CO2-treated carbons performed similarly while the air-treated carbon did not

perform as well due to its increased surface acidity.

A novel approach to tailoring activated carbon via varying levels of DO in the water used

to create steam for the activation of a wood-based precursor was evaluated using the Box-

Behnken design of experiments. The statistical analysis confirmed the impact of changes in DO

on the performance of the resulting carbon for adsorption of MIB and supported variation in DO

as a possible means of tailoring activated carbon.









CHAPTER 1
INTRODUCTION

Adsorbents have been commonly used to solve environmental issues and researchers

have delved into the mechanisms responsible for the success of a particular adsorbent. In the

case of activated carbon, although the surface chemistry of activated carbon is as important as its

physical properties in water treatment, only the optimization of the latter is well researched. In

other words, it is known what physical properties are desired for a contaminant' s removal and

how to create these physical properties. On the other hand, while there are many methods to

alter the surface chemistry (i.e., functional groups) of activated carbon, there are few guidelines

available for its optimization, as it is not known what type of functional groups are best for a

particular contaminant. Furthermore, it is not clear how to control the quantity, type, and

location of these functional groups.


To elucidate the role of surface chemistry in adsorption, the present work summarizes

advanced research into the impact of surface chemistry on the water treatment phenomena of pH

excursions, the tailoring of the surface chemistry of activated carbon through manipulation of the

dissolved oxygen content of the water used to create steam for activation, and, most importantly,

the impact of these changes in surface chemistry on adsorption. By investigating not only how

changes in surface chemistry are instilled into activated carbon but also the specific groups

formed and any effect on adsorption, this research may be a step towards bridging the gap

between potential optimization of the physical properties and surface chemistry of adsorbents.









CHAPTER 2
LITERATURE REVIEW

Activation of Activated Carbon

Activated carbon is created when a carbonaceous precursor, generally wood, peat, lignite-

coal, bituminous-coal, or other material possessing relatively high carbon content, undergoes a

thermal process involving pyrolysis and oxidation to increase the surface area and develop a

range of pore sizes within the precursor. Pyrolysis is carried out in an inert environment, or

oxygen starved, to ensure that the raw material does not gasify but rather is transformed into

char. The oxidation process, or simply activation, is performed in the presence of an oxidizing

gas, such as air, steam, carbon dioxide (CO2), Of a COmbination of these, at temperatures between

973-1 173 K in order to oxidize the surface of the carbon pores and further develop the internal

pore structure via gasification. In physical activation, the pyrolysis and oxidation steps are

separate, while in chemical activation the two processes are combined (Snoeyink and Summers,

1999).

Gasification

Gasification by oxygen proceeds according to the reactions shown in Equations 2-1 to 2-

6 (Walker et al., 1959). A free active carbon site is designated by Cc and a surface complex is

designated by C(n), where n is the type of complex. The preference of the oxidation reaction for

free active sites, associated with high energy edge sites, has been shown by Smith and Polley

(1956) and Laine et al. (1963).

2Cf + 02(g) + 2C(0) (2-1)

C(0) + CO(g) (2-2)

Cr + 02(g) + C(O2) (2-3)

C(O2) + CO2(g) (2-4)









CO(g) + C(0) + CO2 8)+ f (2-5)

Cr + CO2 (g) + 2CO(g) (2-6)

In Equations 2-1 to 2-4, gasifieation occurs either through evolution of carbon monoxide

(CO) or carbon dioxide (CO2) from the carbon surface. These reaction products may then react

with the carbon surface (Equations 2-5, 2-6). The rate of gasifieation is inhibited by reactions

with CO (Equation 2-5), as this causes the concentration of surface oxides available to

participate in Equation 2-2 to decrease (Walker et al., 1959). The CO2 produced may gasify the

carbon surface (Equation 2-6). Indeed, CO2 has been used as an activating agent though it does

require a relatively pure quantity of this gas, which may be difficult to obtain in some areas,

while air and steam are ubiquitous.

Unlike air activation, steam activation is an endothermic process that is easier to control

and better suited for carbons with high surface activity (Hassler, 1974; Kirk-Othmer, 1977).

Typical reaction mechanisms for steam activation are found in Equations 2-7 to 2-9 (Walker et

al., 1959; Puri, 1970; Lussier et al., 1998).

Cr + H120 + C(0) + H12 (2-7)

2Cf + H2 + 2C(H) (2-8)

CO(g) + H120 + CO2(g) + H2 (2-9)

The initial reaction of the steam with the carbon surface where the oxygen is exchanged

from the water molecule to the carbon surface to create a surface oxide (Equation 2-7), which

may devolve as CO, produces hydrogen gas (Walker et al., 1959; Puri, 1970; Lussier et al.,

1998). As discussed with air oxidation, the CO produced may react to decrease the rate of

gasifieation by reacting with a surface oxide to produce CO2. The hydrogen produced (Equation

2-7) may also react with the carbon surface to yield a surface hydrogen complex (Equation 2-8),









which inhibits the gasifieation reaction as this site will not react further as may a C(0) site

(Walker et al., 1959; Puri, 1970; Menendez et al., 1996; Lussier et al., 1998; Bansal et al., 2002).

The reverse of Equation 2-7 may also lead to inhibition by hydrogen removing a C(0). The

water-gas shift reaction (Equation 2-9) is where the water vapor is broken down to CO2 and

hydrogen gas, the products of which may either activate the surface (Equation 2-6) or inhibit

gasifieation (Equation 2-8).

Creation of Functional Groups

Functional groups (designated here as C(n)) form on the surface of activated carbon due

to the presence of heteroatoms, any atom present in the carbon structure that is not carbon or

hydrogen, such as oxygen, nitrogen and sulfur. Functional groups are formed from these

heteroatoms due to the activation and reactivation scheme used and exposure to the atmosphere

and as such these groups are created at a variety of temperatures though there are temperatures

for which groups experience optimal formation. Therefore, activation affects both the physical

structure of the carbon, via gasifieation, and also the chemical properties of the carbon through

formation of these functional groups.

The thermal stability of oxygen-containing functional groups is such that the functional

groups that devolve as CO (e.g., phenolic and quinone groups) are thermally unstable at 973 K

and functional groups that devolve as CO2 (e.g., carboxylic and lactone groups) are thermally

unstable at 773 K (Puri, 1970; Otake and Jenkins, 1993). This relative thermal stability is often

used as a measurement of the concentration of these functional groups on the carbon surface

under a process called temperature programmed desorption (TPD). During TPD, where the

temperature is slowly raised to 1223 K in an inert environment, the quantities of CO and CO2

devolved from the surface are measured and this quantity, and the knowledge of the chemistry

behind the creation of the devolved gas (e.g., carboxylic groups devolve as CO2), yields a










qualitative indicator of the functional groups on the carbon surface (Leon y Leon et al., 1992;

Salame and Bandosz, 2001; Bansal et al., 2002). Fourier transform infrared spectroscopy

(FTIR, which measures the absorption spectra of the carbon surface to determine individual

groups, is also an available "dry" technique of evaluating functional groups. "Wet" methods are

also available, such as the Boehm titration method where specific bases or acids are contacted

with the carbon's surface then titrated to determine the quantity of base or acid that has reacted

with the acidic or basic functional groups on the carbon surface (El-Sayed and Bandosz, 2003;

Ania et al., 2004). In a study by Salame and Bandosz (2001) comparing the Boehm titration

technique, TPD, and FTIR, it was determined that Boehm titration allows detection of groups of

certain strength while omitting other acids and bases such as ketones, esters, and ethers. TPD,

found to detect all functional groups present on the carbon surface, was limited to the broad

category of CO-producing and CO2-producing functional groups and thus this method lacks

information on individual functional groups. Finally, though the FTIR method provides

information on individual functional groups, they found it produced complicated information

where mathematical treatments, requiring several approximations, were necessary to clarify the

data.

By measuring the oxygen-containing functional groups that desorbed from carbon using

TPD it was determined that the optimum temperature for chemisorption of oxygen, using air as

the oxidizing agent, occurs around 673 K (Wiegand, 1937; Puri, 1970). Interestingly, the

maximum capacity for a carbon to adsorb bases, or its acidic nature, has also been shown to lie

close to 673 K (Puri, 1970). The fact that the maximum chemisorption of oxygen and the

maximum capacity to adsorb bases occur at the same temperature has led many researchers to









the conclusion that the acidic nature of carbon is closely related to its oxygen content (Puri,

1970; Otake and Jenkins, 1993; Barton, 1997; Ania et al., 2004).

The carboxyl group (COOH) has been suggested as the oxygen-containing functional

group responsible for the acidic nature of carbon by several researchers (Kruyt and de Kadt,

1931; Puri, 1970). The temperature at which these carboxyl groups were developed was 673 K

in either oxygen or air. More recently, Otake and Jenkins (1993) researched air treated carbons

and found a positive linear relationship between the chemisorbed oxygen concentration desorbed

as a CO2 COmplex, which is indicative of the presence of carboxyl groups on the carbon surface,

and total acidity of the carbon. Furthermore, these authors proposed that the oxygen functional

groups existing on the air-oxidized char were carboxyl groups and were formed following the

mechanism of Figure 2-1.

The importance of these oxygen-containing functional groups on adsorption and their

influence on the surface chemistry of activated carbon will be discussed in later sections.

Impact of Activation Parameters

The impact of the activation process on the physical and chemical properties of the

carbon has been discussed in general terms but not the effect of the individual activation

parameters on these properties. Activation parameters include, but are not limited to, precursor

choice, type and flow of oxidant, time of activation, and temperature of activation. The

following is a discussion of how and why variation of these parameters, as they are the easiest to

control, can affect the resulting carbon's physical and chemical properties.

Type of Precursor

The two main groups of raw material for activated carbon are coal and biomass. Types

of coals employed include anthracite, lignite, and bituminous coals (Streat et al., 1995; Zou and

Han, 2001). By far the most common precursor for activated carbon used in research is coal;









subsequently, many of the theories on physical and chemical adsorption onto activated carbon

are based on coal-based activated carbon and may not hold with biomass-based carbons due to

the differing properties of these carbons, as discussed further in this section. Examples of

biomass used as raw material for activated carbon include sewage sludge, coconut shells,

coconut coir dust, fruits stones (e.g., peaches, plums, and cherries), straw, and nut shells

(Gergova et al., 1994; Heschel and Klose, 1995; Streat et al., 1995; Albers et al., 2003; Rio et al.,

2005; Souza Macedo et al., 2006). Various types of wood, including pine, oak, and beech have

also been investigated (Heschel and Klose, 1995; Streat et al., 1995; Albers et al., 2003; Garcia-

Garcia et al., 2003).

The drive to investigate different types of biomass as a precursor for activated carbon

does not seem to lie so much in an effort to improve carbon for the removal of a specific

compound or water matrix as much as it is an economic or environmental consideration. For

example, a study pertaining to the use of pine as a source of activated carbon was prompted by

the use of pine as an energy source in the study area, due to the lack of other available energy

sources such as coal (Garcia-Garcia et al., 2003). The large supply of char remaining after

burning of the pine prompted the investigation into use of this raw material for activated carbon,

which could then be sold for an economic benefit. The use of a waste product to create an

economically and environmentally useful product is what has prompted the study of sewage

sludge, coconut shells (due to local abundance in tropical countries), fly ash from power plants,

straw, and used tires, as these wastes would either be incinerated or placed in a land fill in most

circumstances (Streat et al., 1995; Rio et al., 2005; Sarkar et al., 2006; Souza Macedo et al.,

2006).









A successful carbon in these studies is one that performs as well as traditional carbons in

removal of a target compound. These studies employ a number of raw materials and evaluate

them based on either adsorption performance or a common method of indicating performance,

such as pore size distribution, and they may also analyze the effect of varying activation

parameters on the properties of the resulting carbon (Streat et al., 1995; Garcia-Garcia et al.,

2003; Rio et al., 2005). For example, Heshel and Klose (1995) employed several agricultural by-

products (e.g., fruit stones, nut shells, and wood) to determine the type of raw material that

produced the highest quality activated carbon. High quality was defined as a carbon that was

abrasion resistant, having low macroporosity (i.e., few large pores), and having high carbon

yield, where the latter relates to the quantity of carbon lost during activation. The authors

concluded that the highest quality carbon from agricultural by-products was produced from

coconut shells, followed by peach stones, plum stones, hazelnut shells, walnut shells, and cherry

stones. The authors indicate that a large pore size in the raw material leads to large pores in the

carbon. The wood based carbons had higher porosity than either the nut shells or the fruit stones

and this led to decreased abrasion resistance.

Heshel and Klose (1995) found that the yield of carbon correlated with the O/C atomic

ratio in the carbon. The authors concluded that the morphology of the raw materials determines

several important properties of carbon due to the cross-linking cellulose and lignin molecules. A

study by Albers et al. (2003) using wood and coconut shells as the raw material demonstrated

from scanning electron microscope (SEM) pictures that the inherent structure of the raw material

is present even after pyrolysis. Gergova et al. (1994) and Khezami et al. (2005) observed the

same phenomenon of a woody structure that is preserved and becomes visible after the high

temperatures employed in pyrolysis. Even at the atomic level, differences in the graphite









structure of activated carbons are apparent based on the raw material employed (Albers et al.,

2003).

Gergova et al. (1994) also note that lignin and cellulose in the raw materials of

agricultural waste may affect porosity development which may relate to the finding of Hershel

and Klose (1995) that the O/C atomic ratio plays a role in char yield. Heschel and Klose (1995)

and Khezami et al. (2005) note that lignin as a source of activated carbon produces the highest

char yield, which may relate to porosity and the fact that lignin has a lower oxygen content than

cellulose. These studies highlight the importance of the choice of raw material when creating

activated carbon.

Type of Oxidant

Examples of oxidants that may be used in the thermal activation process include air (C-

02), steam (C-H20), carbon dioxide (C-CO2), Of a COmbination of these. While use of either

oxidant to oxidize the surface of carbon, and thereby improve its surface area, will produce an

acceptable activated carbon, there are advantages and disadvantages to each oxidant.

Walker et al. (1959) indicates that the activation energy for C-O2 is leSs than C-H20

which is in turn less than C-CO2. This is due to the activation energy being determined by either

the adsorption (e.g., creation of C(0)) or desorption (e.g., evolution of CO(g)) step. The C-O2

adsorption step is much more exothermic than the C-H20 adsorption step and due to this

exothermic reaction the lifetime of the functional groups on the C-O2 reacted carbon may be

short. Given that the desorption of the surface oxide created by the C-O2 reaction occurs

relatively quickly, the activation energy for C-O2 is determined by the adsorption step. The C-

H20 and C-CO2 adsorption step, having a lower exothermic nature, may have a longer lifetime

for its adsorbed group so that desorption is the rate limiting step. However, while C-O2 may

require lower activation energy, energy is released during the reaction of air with the carbon such









that control of the process may be difficult and generally is better suited for low activity carbons

(Walker et al., 1959; Kirk-Othmer, 1977).

The literature is rife with comparisons of C-CO2 VeTSus C-H20 activations. Tomk6w et

al. (1977) investigated the porous structures of carbons resulting from C-O2, C-H20, and C-CO2

and concluded that C-H20 created a higher distribution of pores of different sizes due to the

progressive widening of pores once they are formed. C-CO2 activations yielded mainly

micropores (i.e., pore openings smaller than 2 nm) suggesting a higher occurrence of pore

creation rather than pore widening. These findings were supported by Ryu et al. (1993).

However, Wigmans (1989) found the opposite trend to Tomk6w et al. (1977) where C-CO2

produced widening of micropores to mesopores (i.e., pore openings between 2-50 nm). Alcaniz-

Monge et al. (1994) and Rodriguez-Reinoso et al. (1995) worked to clarify this apparent

contradiction by investigating the difference in the gases' diffusion and accessibility to the pores.

Interestingly, at low burnoffs (e.g., low levels of gasification) the results of Wigmans (1989)

were confirmed by Alcaniz-Monge et al. (1994), where the C-H20 showed a higher production

of micropores, though at higher burnoffs the C-CO2 had a higher concentration of micropores

confirming the results of Tomk6w et al. (1977). Overall, however, the authors surmised that

Tomk6w et al.'s (1977) view of pore formation dominated and this was attributed to the

difference in diffusion of the CO2 and steam molecules.

Walker et al. (1954) found that the order of reaction rates were such that the rate of C-

H20 progressed at approximately three times the rate of C-CO2 and the rate of C-O2 progressed

at several orders of magnitude faster than both C-H20 and C-CO2. Later studies further

confirmed that C-CO2 gasification is slower than that of steam gasification likely due to the

faster diffusion rate of the H20 molecule (Ryu et al., 1993; Alcaniz-Monge et al., 1994; Walker









et al., 1996). Rodriguez-Reinoso et al. (1995) used the relative reaction rates of C-H20 and C-

CO2 to clarify the impact of burnoff on the development of pores with steam and carbon dioxide.

Carbon dioxide reacts first to create and widen narrow micropores so that at low burnoff the

micropores will dominate. Given further burnoff(i.e., further opportunity for the CO2 to react

with the carbon), the narrow micropores will become mesopores and macropores (i.e., pores

greater than 50 nm) which yields a more diverse pore structure. Steam, on the other hand, reacts

to continually create and then expand micropores so that micropores do not remain intact very

long due to the higher reaction rate of C-H20.

Another possible explanation proposed by Rodriguez-Reinoso et al. (1995) was that C-

CO2 created a higher number of oxygen containing functional groups, located on the openings of

the pores, and that these groups, being more thermally stable than those produced by steam

activation, would close the pores to further diffusion of carbon dioxide which would have

allowed for pore widening. Later work by these authors (Molina-Sabio et al., 1996) supported

this hypothesis. Walker et al. (1996) also suggests that hydrogen is a greater inhibitor in the

steam reaction (Equation 2-7) which limits the creation of pores and would explain the decrease

in micropore formation. Regardless of the explanation, the conclusion is that steam activation

yielded a wider pore size distribution than activation with carbon dioxide (Tomk6w et al., 1977;

Alcaniz-Monge et al., 1994; Rodriguez-Reinoso et al., 1995; Molina-Sabio et al., 1996; Walker

et al., 1996). While micropores are preferred for the adsorption of some small molecules, it is

generally considered advantageous to have a wide pore size distribution as the larger pore sizes

act as "roadways" by which the target molecule can travel to its final adsorption site (Tennant

and Mazyck, 2003). To this end, steam activation is often employed.









In terms of activation with oxygen, Tomk6w et al. (1977) determined that these carbons

have a high development of pores but only at low burnoff and this was related to the blocking of

the pores due to creation of functional groups at the edge of the pore openings. All carbons

activated with oxygen in this study were found to have a higher concentration of polar centers

than either the C-H20 or C-CO2 carbons, which is related to the influence of oxygen-containing

functional groups on polarity. In addition, at higher temperatures, where these oxygen-

functional groups would be expected to devolve (i.e., be removed from the carbon surface), the

pore structure resulting from C-O2 is improved.

Time, Temperature, and Oxidant Flow

Gergova et al. (1993) evaluated the effect of activation temperature and activation time

on the functional groups created during steam activation of several different types of agricultural

precursors. The trends indicate an increase in the quantity of oxygen-containing functional

groups with treatment time, from 1 to 3 hrs, at lower temperatures (873 K) and a decrease in

these groups with an increase in treatment time at higher temperatures (973 K), though this trend

was not true of all precursors. This relates to the degradation of functional groups at higher

temperatures and the limited creation of new functional groups to replace the ones lost, a finding

supported by Arriagada et al. (1997).

All carbons in Gergova et al.'s (1993) study showed a decrease in solid yield, the mass of

carbon remaining after activation, with an increase in activation time and temperature. Given the

nature of the gasification reaction, it is anticipated that with higher temperature, which would

speed gasification reactions, and longer treatment time, which would allow more gasification to

occur, that the solid yield would decrease and this is supported in many studies (Gergova et al.,

1993; Gonzalez et al., 1994; Rodriguez-Reinoso et al., 1995; Gergova et al., 1996; Martin-

Gullon et al., 1996; Tennant and Mazyck, 2003). As the degree of gasification increases with an









increase in temperature and treatment time, there is also generally an increase in the volume of

pores and also the surface area within the carbon due to the removal of carbon atoms from the

carbon surface to create new pores and higher surface area (Gergova et al., 1994; Centeno and

Stoeckli, 1995; Rodriguez-Reinoso et al., 1995; Gergova et al., 1996; Mazyck and Cannon,

2002; Zhang et al., 2003). However, overactivation can occur at high temperatures and or

treatment time as external mass loss exceeds internal mass loss, where pores are developed,

and/or the walls of pores begin to collapse (Gergova et al., 1994; Tennant and Mazyck, 2003;

Mackenzie et al., 2005).

Gonzalez et al. (1994) found that the rate of steam flow did not significantly affect the

resulting micropore and mesopore formation in activated carbon and that low temperatures

favored the widening of narrow micropores. Mazyck and Cannon (2002) also observed that

steam flow had no impact on pore size distribution. Gonzalez et al. (1994) indicate that, though

an increase in gasification rate has been observed with an increase in the flow of carbon dioxide

due to the presence of a concentration gradient which compels the oxidant to enter the pores at a

faster rate (i.e., bulk diffusion), the increase in steam flow yields more water vapor on the

particles' outer surface resulting in mass loss and/or the creation of macropores. Martin-Gullon

et al. (1996) further this theory by suggesting that the inhibiting factors produced by gasification

(CO and H2) (Equation 2-5 and 2-7, respectively) would dilute the steam on the internal portions

of the carbon thus leaving the highest concentrations of pure steam on the external surface of the

particle which would lead to external gasification rather than internal. The dilution theory was

supported by Arriagada et al. (1997).

Rodriguez-Reinoso et al. (1995), however, found that at lower temperatures the steam

flow had no effect while at higher temperatures the rate of steam flow yielded an increase in










gasification. An increase in the volume of mesopores with an increase in steam flow was also

observed by Martin-Gullon et al. (1996). Rodriguez-Reinoso et al. (1995) clarified this

contradiction by noting that at higher temperatures the inhibiting effect of the hydrogen formed

during gasification (Equation 2-7) is less effective. As steam flow is increased one would

anticipate that more oxidant will react with the carbon; however, at lower temperatures, the

increase in steam flow yields a corresponding increase in the amount of hydrogen produced by

the reaction of Equation 2-7 which would inhibit reactions so that similar quantities of

gasification would occur regardless of the increase in oxidant. At higher temperatures, where the

gas-water shift is less efficient, an increase in oxidant would indeed yield an increase in

oxidation without as significant an affect of hydrogen inhibition.

Adsorption

Physical Adsorption

Adsorption of a substance onto activated carbon in water occurs when the substance is

concentrated more on the surface of the activated carbon than it is in the bulk solution. The

substance that accumulates is called the adsorbate and the activated carbon is termed the

adsorbent. Physisorption, or physical adsorption, occurs when the accumulation results from

dispersion forces while chemisorption, or chemical adsorption, results from an exchange or

sharing of electrons between the adsorbate and the surface of the activated carbon. The

properties of activated carbon that make it ideal as an adsorbent are a significant surface area to

which the adsorbate may accumulate and an extensive internal pore structure in which the

adsorbate may become trapped, thus encouraging its accumulation.

The rate of removal, or adsorption kinetics, is governed by the rates of the four steps

composing physical adsorption. The first step, bulk transport, where the adsorbate moves from

the bulk solution to the boundary layer of the adsorbate, is encouraged by the turbulent flow









surrounding the activated carbon. Film resistance to transport follows bulk transport and

involves the transport of the adsorbate through the stationary layer of water surrounding the

adsorbent. Film resistance may also be minimized by increased flow rate past the adsorbent.

The third step is internal pore transfer where the adsorbate travels through the adsorbent' s pores

via surface or pore diffusion to existing adsorption sites. Internal pore transfer is often the rate-

limiting step as the final step, physical adsorption, occurs almost instantaneously, so it has little

effect on the rate of the overall reaction (Snoeyink and Summers, 1999). The rate of adsorption,

therefore, is influenced by the same variables affecting diffusion rate, including the concentration

gradient and the temperature of the system (Bansal et al., 1988).

The relative hydrophobicity of the adsorbate can be a driving force for its accumulation

on the adsorbent. Thus, the adsorption of a substance generally decreases with increased

contaminant solubility in water (Snoeyink and Summers, 1999). The pore size of the activated

carbon also plays a maj or role in adsorption as, for example, a large natural organic matter

(NOM) molecule will be unable to fit into small pores (i.e, micropores). An activated carbon

with a large volume of mesopores and macropores would be better suited for NOM removal

(Newcombe, 1999). Conversely, a larger microporous volume is generally preferred when the

target compound is a small molecule. However, an increase in mesopores can be associated with

an increase in adsorption of small molecules (Tennant and Mazyck, 2003, Nowack et al., 2004;

Rangel-Mendez and Cannon, 2005) though the mesopores may not be the final adsorption sites

for the compound (for example 2-methylisoborneol (MIB)). The mesopores simply serve to aid

in diffusion of the target compound through the pores to their final adsorption sites, likely

micropores. It is common knowledge that mesopores aid in diffusion particularly in fine

powders and that the adsorption of small molecules is most effective in micropores due to the










greater number of contact points between the molecule and the adsorbent (Suzuki, 1990;

Pelekani and Snoeyink, 1999; Quinlivin et al., 2005; Rangel-Mendez and Cannon, 2005).

Effect of Surface Chemistry on Adsorption

While general trends related to strictly physical adsorption have been discussed, the

impacts of surface chemistry on adsorption must also be discussed, as there are cases where it is

the chemical properties of the activated carbon that dominate adsorption rather than its physical

properties.

The addition of oxygen to the surface of an activated carbon has been shown to alter its

wettability, catalytic and electronic properties, as well its adsorption capabilities (Puri, 1970;

Bansal et al., 2002; Szymaniski, 2002). Acidic oxygen-containing functional groups are mainly

present on the outer surface or edge of the basal plane due to limited diffusion into the

micropores (Puri, 1970; Menendez et al., 1996; Ania et al., 2004). As these outer sites constitute

the maj ority of the adsorption surface, the concentration of oxygen on the surface has a great

impact on the adsorption capabilities of the carbon (Puri, 1970).


The presence of oxygen imparts a polar nature to the activated carbon, which results in a

preference for removal of the more polar component of a solution, e.g., H20 (Kipling and

Gasser, 1960). El-Sayed et al. (2003) found that the adsorption of valeric acid onto activated

carbon decreased as the density of acid groups on the surface increased. These authors suggested

that the polar nature of the activated carbon surface, associated with its high surface oxygen

group content, resulted in the preferential adsorption of water molecules. In addition, the

presence of carboxylic groups acted as physical obstacles to the interaction of the valeric acid

and the activated carbon surface. Bansal et al. (2002) also attributed the decrease in phenol

adsorption with increased oxidation of activated carbon to the increasing hydrophilic character of









the oxidized carbons. Ania et al. (2004) found a decrease in the adsorption of salicylic acid as

the oxygen-containing functional groups increased at low adsorbate concentrations. At high

levels of salicylic acid, however, the study found that the adsorption capacity of the carbon was

similar to an activated carbon that had not been oxidized and these authors proposed that the

higher concentration of valeric acid allowed the water molecules adsorbed onto the hydrophilic

carbon sites to be displaced.

Pendleton et al. (1997) researched the effects of surface chemistry on the adsorption of

MIB and linked decreased adsorption with an increase in the hydrophilic nature of the oxidized

carbon. They also found a strong correlation (R2=0.99) between oxygen content and the number

of hydrophilic sites. A subsequent study by the same research team (Considine et al., 2001)

further supported the link between oxygen content and decreased adsorption of MIB. In this

study, the adsorption of MIB in a dichloromethane solvent, a non-polar solvent, was investigated

as opposed to previous experiments in water. The results showed that adsorption of MIB was

independent of oxygen content in this solvent, further supporting the contention that adsorption

of water as a result of surface oxidation impedes MIB adsorption.

Karanfil and Kitis (1999) found that surface oxidation, and the resulting density of strong

acid functionalities, decreased the adsorption of dissolved organic matter (DOC) (the dissolved

portion of NOM). Water clusters forming around the polar hydrophilicc) functional groups on

the surface reduced the ability of the DOC to reach the smaller pore sizes needed for its

adsorption.

The chemisorption of oxygen and the resulting functional groups have also been found to

destroy delocalized xn-electrons due to electron localization (Leon y Leon et al., 1992; Ania et al.,

2004). The adsorption of phenol has been attributed to the electron-donor acceptor complexes









formed between the delocalized electrons on the basal planes on the surface of the carbon and the

aromatic ring of the phenol (Bansal et al., 2002; Ania et al., 2004). A decrease in phenol

adsorption on oxidized samples has been shown by a number of researchers (Bansal et al., 2002;

Ania et al., 2004). As the bonding sites are destroyed through chemisorption of oxygen, the

adsorption of phenol decreases (Ania et al., 2004). Pereira et al. (2003) performed a study on an

acidic anionic dye, where the presumed adsorption mechanism involved the delocalized xn-

electrons of the carbon surface and the free electrons of the dye molecule, and found that the

adsorption of the anionic dye was hindered by the presence of acidic oxygen-containing

functional groups on the surface of the activated carbon. It should be noted that, in addition to

both the increased hydrophilicity and the destruction of xn-bonds, the formation of acidic-oxygen

containing functional groups create a weakening in the donor-acceptor (acid/base) adsorption

mechanism (Ania et al., 2004).

Current Uses of Activated Carbon

Following activation, the carbon may be distributed as either powdered activated carbon

(PAC) or granular activated carbon (GAC). PAC is activated carbon ground in various types of

mills so that 65-90% passes through a number 325 mesh (45Cpm) sieve while GAC consists of

any larger activated carbon particles and is the portion of activated carbon used in the packed bed

contact basins of water treatment plants. GAC can be used as a substitute for typical granular

filter media, in that it removes suspended matter, though through adsorption GAC is also capable

of removing organic compounds such as synthetic organic chemicals (SOCs), NOM, and taste

and odor causing compounds (Gerber, 1968; Frederick et al., 2001a; Bansal et al., 2002).

SOCs, such as most pesticides, may be contaminants in both ground and surface water.

SOCs at chronic exposure levels may cause neurological and kidney effects and, for some

pesticides, cancer (Hammer and Hammer, 2001). One common organic compound that is










targeted for removal by activated carbon is phenol which can be both manufactured and naturally

occurring and is relatively common in the environment due to discharge from industries such as

those producing plastics (Jung et al., 2001; Bansal et al. 2002).

NOM, a mixture of fulvic, humic acids, hydrophilic acids, and carbohydrates, can impart

color to water and may also react with chlorine, added for disinfection purposes, to create

disinfection by-products (DBPs). DBPs, such as chloroform, a trihalomethane, are considered

carcinogenic (Bull et al., 1982). It is, therefore, important to reduce the NOM concentration in

the water before disinfection with chlorine occurs. Adsorption ofNOM onto GAC is enhanced if

calcium complexes with the NOM prior to adsorption by activated carbon (Frederick et al.,

2001a).


Taste and odor causing compounds, the removal of which constitutes the primary reason

water treatment utilities use activated carbon, may result from biological growth or industrial

activities. An example of a taste and odor causing compound resulting from biological growth is

2-methylisoborneol (MIB), which can be produced by cyano-bacteria, or blue-green algae

(Gerber, 1968). MIB, described as having an earthy-musty odor, enters a water treatment plant

when the source water is surface water experiencing a bloom in the growth of the MIB producing

organi sm. If occurrences of MIB are relatively short in duration PAC may be used as an

additive during water treatment rather than a permanent GAC filter.

Reactivation of Activated Carbon

During normal use, activated carbon will eventually become saturated with adsorbates so

that the treated water exceeds the desired level of adsorbate removal, or reaches breakthrough.

At this point, the spent activated carbon can be reactivated to restore the adsorption capacity of

the GAC so that it can be reused rather than simply disposed of and replaced with virgin GAC.









The time to breakthrough for a GAC bed can range from about 6 months to 5 years, depending

on the type of adsorbate, the influent concentration, and the desired treatment level (Kawamura,

2000).

Thermal or chemical regeneration may be employed, though thermal reactivation is the

most commonly employed scheme. Thermal reactivation consists of four steps: drying,

desorption, pyrolysis, and gasification. The drying process removes water and some highly

volatile adsorbates. At higher temperatures, thermal desorption occurs with vaporization of

volatile adsorbates and decomposition of unstable adsorbates to more volatile components. A

pyrolytic process follows using high temperatures, between 923-1123 K, in an inert environment

that converts heavy or non-volatile adsorbates to char. The final step, gasification at

temperatures above 973 K in steam, CO2, Of a COmbination of both, involves desorption of the

vapors and gaseous products of char and their exit from the pores of the reactivated carbon

(Clark and Lykins, 1989; Cannon et al., 1993; Snoeyink and Summers, 1999).

Reactivation has several effects on the activated carbon, aside from the intended effect of,

ideally, restored adsorption capacity. Mass loss during the activation process may range from

10-15% of the original GAC and is the result of loss through the transfer of GAC from the

treatment site to the regeneration site and/or burning of the GAC during reactivation (Clark and

Lykins, 1989). The loss is compensated for by the addition of virgin GAC before the reactivated

GAC is returned to service. The surface chemistry of the activated carbon is also affected by

reactivation as most oxygen-containing functional groups are stripped during reactivation at

temperatures greater than 973 K and in an inert environment (Menendez et al., 1996; Pereira et

al., 2003).










Inorganic molecules, such as calcium, may be adsorbed during normal operation of a

GAC filter and are not removed during the activation or reactivation processes. The inorganic

molecules remaining on the surface of the reactivated carbon prevent these adsorption sites from

being freed for further adsorption and thus compromise the ability of the reactivation process to

restore the adsorption capacity of the carbon (Clark and Lykins, 1989). One well-documented

effect of calcium presence in the reactivation process is calcium catalysis (Knappe et al., 1992;

Cannon et al., 1993; Mazyck and Cannon, 2000, 2002). Calcium catalysis increases the number

of active sites on the surface of the carbon rather than, in the traditional sense of a catalyst,

reducing the activation energy. Calcium catalyzes the gasification step of reactivation and is

characterized by increased mass losses during the reaction process and pore enlargement

(Knappe et al., 1992). As discussed previously, the pore size of the activated carbon can affect

its ability to adsorb certain compounds, and an increase in pore size may make it unsuitable for

certain applications. There is also cost associated with the addition of virgin GAC to replace the

increased mass lost during the reactivation process.

Another result of reactivation is the incidence of pH excursions, an increase in the pH of

treated water when reactivated carbon is returned to service (Farmer et al., 1996; Farmer et al.,

1998). The duration and extent of the pH excursion will determine its impact on the system.

pH Excursions in Water Treatment

Many water and wastewater utilities have found that when thermally reactivated GAC is

returned to service, the pH of the subsequently treated water is elevated (e.g., pH > 9) for days or

even weeks (Mazyck et al., 2005). Water with high pH levels is unfit for distribution to

customers, as it exceeds the United States Environmental Protection Agency secondary standards

of the National Secondary Drinking Water Regulations (6.5 < pH < 8.5) and could also bring

about calcium carbonate (CaCO3) precipitation problems in distribution systems.









A water utility in the Virginia-American Water Company, under the umbrella of the

American Water Works Company, measured increases in pH in their backwash water when a

reactivated GAC was placed back in service (Figure 2-2). These data are representative of a full-

scale GAC filter bed adsorber. As shown, the reactivated GAC posed an adverse effect on water

quality in that it elevated the filter bed adsorber' s effluent pH to greater than 11. Even after 6

days, the pH was still above 8.5. Note that the pH of natural water is approximately 7.

In addition to the pH, the water utility also measured the concentration of calcium in the

effluent for two days. As shown in Figure 2-2, as the calcium concentration decreased, the water

pH likewise decreased. The influent calcium concentration in their backwash water was

relatively constant over the two days. No other operational changes could have contributed to

the higher Ca2+ and OH- levels; these results are fairly typical, as reported by American Water

Works Company personnel. These results led to the hypothesis of a calcium chemistry

mechanism proposed by Mazyck et al. (2005) and discussed further below.

pH Excursion Mechanisms for Activated Carbons

A mechanism involving anion exchange with hydroxide ions following carbon

protonation has been proposed by Farmer and coworkers to explain pH excursions (Carr and

Farmer, 1995; Dussert et al., 1995; Farmer et al., 1996). These authors suggested that pH

excursions are a function of the carbon surface that is altered during high temperature activation

or reactivation. Activated carbons exposed to high temperatures in a reducing atmosphere during

manufacturing and reactivation tend to adsorb strong acids in water. Garten and Weiss (1957)

classified these as HI-type carbons. In contrast, L-type carbons are those produced by surface

oxidation; these are known to adsorb strong bases in water. Farmer et al. (1996) argued that this

acid adsorption during water treatment might involve the protonation of pyrone-type surface

groups or other similar functionalities on the carbon surface, in agreement with Leon y Leon et









al. (1992). In addition, they suggested that loading the carbon with sulfate, chloride, or other

anions that are present in water could neutralize the positive charge on the carbon surface.

Furthermore, they proposed that this charge neutralization could also occur through an anion

exchange process involving sulfates and hydroxides on the carbon surface. In other words, they

surmised that the carbon could adsorb sulfates or other anions from water while releasing

hydroxyls, and that this ion exchange process could cause the pH to rise. Farmer et al. (1996)

did acknowledge that this anion exchange mechanism requires additional testing and

confirmation. To date, however, there has been no further evaluation by these authors presented

in the literature that would either prove or disprove their hypothesis.

Dussert et al. (1995) supported this anion exchange mechanism because they found no

significant pH excursion when virgin GAC was immersed in Milli-Q (anion-free) water, and they

concurred that anions were required to trigger a pH effect. These authors came to their

conclusions after studying virgin GACs that removed 2 to 9 mg of sulfate/g GAC. In Milli-Q

water spiked with 80 mg sulfate/L, they measured the water pH and sulfate capacity for several

virgin carbons. Not only did they find that pH excursions only occurred when anions such as

sulfate were present, but they also reported that, as the sulfate capacity increased, pH also

increased. In their study, the carbons that did not exhibit pH excursions also did not significantly

remove sulfate ions from solution (< 0.7 mg/g). They also found that sulfate (SO42-) TemOVal

was greater than nitrate (NO3-) or chloride (Cl-) removal. They did not measure calcium

concentrations for their carbons; however, they did find that the ash content of the carbon could

not be correlated with pH excursions.

To discern if an anion exchange mechanism was also valid in describing pH excursions

with reactivated GAC, Mazyck et al. (2005) tested reactivated carbons under conditions similar









to those used by Dussert et al. (1995) to support an anion exchange mechanism. According to

the anion exchange mechanism supported by Dussert et al. (1995), the sulfate uptake should be

highest when there is a high pH excursion. The data of Mazyck et al. (2005) demonstrated an

opposite trend where, as the pH excursions of the reactivated GAC increased, the sulfate uptake

decreased. For example, at a water contact pH, a predictor of pH excursions, of 8.75 the sulfate

uptake was 1.4 mg sulfate/g GAC while at a water contact pH of 10.6 there was no measurable

sulfate uptake. In addition, pH excursions were found to occur even in Milli-Q water, where

sulfate ions were not present suggesting that anions are not responsible for the observed pH

excursions with reactivated GAC. In light of these results, Mazyck et al. (2005) concluded that

the uptake of sulfate and the subsequent release of hydroxyl ions (i.e., an anion exchange

mechanism) was not responsible for the observed pH excursions with reactivated GACs.

The observation of a relationship between calcium and pH excursions, shown in Figure 2-

2, led to a hypothesis by Mazyck et al. (2005) of a calcium chemistry mechanism behind pH

excursions. During potable water treatment, water-soluble calcium can adsorb onto GAC as a

complex with NOM (Nowak and Cannon, 1997; Nowak et al., 1999; Frederick and Cannon,

2001; Frederick et al., 2001a, 2001b). Moreover, calcium is abundant in many coal-based

carbons. Thermal reactivation of spent GAC at temperatures above 1123 K decomposes the

oxygenated functionality of the NOM that complexes with calcium. This process results in the

formation of CaO, which is thermodynamically the most stable species at these high

temperatures (Mazyck and Cannon, 2000). When the reactivated carbon is returned to service,

the calcined CaO can react with water to form Ca(OH)2 (Equation 2-10). Furthermore, Ca(OH)2

can dissolve and subsequently dissociate as Ca2+ and 20H- (Equation 2-11). The release of OH-

would elevate the pH of the subsequently treated water.










CaO + H20 4 Ca(OH)2 (2-10)

Ca(OH)2 8 C2+ + 20H- (2-11)

Calcium chemistry and thermodynamics as related to GAC and thermal reactivation have

been previously discussed in the literature at length (Cannon et al., 1994; Mazyck and Cannon,

2000, 2002). Therefore, it was proposed that the manner by which the reactivated GAC was

contributing both calcium and hydroxyl ions to the water was via the mechanisms described in

Equations 2-10 and 2-11 above. Furthermore, it was rationalized that the deviation from a [OH-]

(molar concentration of hydroxyl ions) to [Ca2+] (mOlar concentration of calcium ions) ratio of 2,

is caused by the OH- that originated from Ca(OH)2 being consumed to some extent by the natural

buffering capacity of the water. Thus, the data from the Virginia-American Water Company

(Figure 2-2) could be taken as a semi-quantitative example of the mechanisms described in

Equations 2-10 and 2-11.

In further support of the calcium chemistry mechanism, a significant correlation was

found between pH and calcium present in solution and altering the reactivation process to avoid

the creation of calcium species that will participate in the proposed calcium chemistry, using a

reactivation scheme proposed by Mazyck and Cannon (2000, 2002), was found successful in

overcoming pH excursions (Mazyck et al., 2005). Both of these finding supported the

hypothesis of the calcium chemistry mechanism dominating pH excursions with reactivated

carbons.

Methods to Overcome pH Excursions

The three broad categories of methods to overcome pH excursions are on-site treatments,

tailoring of the reactivation process, and post-reactivation treatments performed before the

reactivated GAC is returned to service.









One on-site method that would occur prior to reactivation is the reduction of the calcium

adsorbed onto the activated carbon during water treatment. To achieve this, either the NOM

complexes with a stronger cation, thus out-competing calcium, or the calcium-NOM complex is

removed before it comes into contact with the GAC. The use of an iron coagulant would be

ideal as it could facilitate both of the above goals. Iron (Fe3+) is a Stronger cation than calcium

and would displace the calcium from the NOM though competition with NOM by hydroxyl

species may hinder the process near a neutral pH. Nowack et al. (1999) proposed that, at near

neutral pH, soluble iron hydroxide would precipitate depleting the amount of iron available for

further reactions with NOM. The iron would also coagulate the calcium-NOM complexes

already formed and this floc would be allowed to settle out before reaching the GAC. Nowack et

al. (1999) and Frederick et al. (2001) have shown that the addition of iron coagulant to a system

containing both calcium and NOM yielded a decrease in the amount of calcium-NOM complexes

adsorbed to the carbon as well as a relatively level iron loading onto the GAC. These results

indicate that, not only is iron addition successful in decreasing calcium loading, but that the

primary mechanism responsible for this is coagulation of the iron-NOM and/or calcium-NOM

complexes. The accumulation of iron is not detrimental to the reactivation process as iron

catalysis is suppressed by the presence of sulfur, which is ubiquitous in coal-based carbons

(Nowack et al., 1999). A near neutral pH was indeed found to hinder iron-NOM formation as

the hydroxide species out-competed the NOM for complexation with the iron ions. Therefore,

Nowack et al. (1999) and Frederick et al. (2001) concluded that a combination of pH adjustment

to approximately 6.0 pH and iron addition, on the order of 5 mg/L, are the most effective in

reducing calcium loading. Water treatment facilities that operate above this range of pH are

expected to accumulate higher levels of calcium onto their GAC.









Other on-site treatments to overcome pH excursions after the reactivated carbon is

returned to service include backwashing, recycling of the high pH water, and pH adjustments

before the treated water is discharged. Both backwashing and recycling involve running water

through the GAC until the pH reaches acceptable levels, which reduces the yield of treated water

through the system (Farmer et al., 1998). In addition, backwashing may require a holding tank,

as the high pH water cannot be discharged, which will restrict the allowable backwashing time

(Carr and Farmer, 1995). Acid treatment of the high pH water prior to discharge to stabilize the

pH can be time-consuming and expensive, due to additional monitoring equipment.

Acid-washing a traditionally reactivated carbon prior to use, aside from being time-

consuming and expensive, may not even produce a pH stable carbon (Dussert et al., 1995). In

addition, oxygen surface complexes are introduced to the surface of the carbon if an oxidizing

acid is used and this can impact the adsorption capabilities of the reactivated carbon, as discussed

previously. Depending on the acid, the micropore volume may also decrease as a result of

treatment.

A reactivation protocol proposed by Mazyck et al. (2005) was discussed that can

overcome calcium catalysis and the resulting pH excursions by reactivating the carbons at

temperatures below that at which CaO is formed, 1123 K. It is the CaO form that participates in

the calcium chemistry mechanism. A decrease in the amount of CaO available to participate in

the calcium chemistry reactions will result in a decreased pH excursion.

Oxidizing the reactivated carbon, outlined in a patent by Dussert et al. (1995), or treating

the reactivated carbon with carbon dioxide (CO2), outlined in a patent by Farmer et al. (1998),

have been offered as means to create a pH stable carbon. Both patents propose that oxidation or

CO2 treatment, depending on the patent, neutralizes the surface of the reactivated carbon, thus









inhibiting any anion-exchange during water treatment. As the focus of a previous study (Mazyck

et al., 2005) was the role of calcium in pH excursions and the result was that the anion-exchange

mechanism does not fully explain the excursions found in reactivated carbons, there must be an

additional explanation for why these treatments are able to create pH stable carbons.

Previous work by this author (Bach, 2004) determined that air treatments of reactivated

carbons create pH stable carbons (i.e., pH < 8.5) through the formation of acidic oxygen-

containing functional groups on the carbon surface which balance out the hydroxyl ions added by

the calcium chemistry mechanism. This was supported by an evaluation of the total surface

acidity of the air-treated versus untreated carbon. However, given the discussion of the role of

individual functional groups on surface acidity, specifically the carboxyl group, an evaluation of

individual functional groups would be required to further substantiate this proposal. CO2

treatment was also found to create a pH stable carbon and it was proposed that CO2 treatments

form pH stable reactivated carbons via physisorption, and subsequent desorption, of CO2

forming H2CO3 and then HCO3- upon water immersion which also balances out the hydroxyl

addition by the calcium chemistry mechanism. This hypothesis was supported by similar total

surface acidity found between an untreated and a CO2 treated carbon, though an analysis of

individual functional groups rather than total acidity would further support this theory. The

literature also demonstrates that there would be no chemisorption of CO2 onto CaO to form

CaCO3 which would not participate in the proposed calcium chemistry mechanism of Equations

2-10 and 2-11 (Mazyck and Cannon, 2000).

Following the treatments to overcome pH excursions and after evaluating their success,

an air treated, a CO2 treated and an untreated reactivated carbon were evaluated for their

performance in the removal of a common taste and odor causing compound, MIB (Bach, 2004).









The results demonstrated that the untreated reactivated GAC removed more MIB than the air

treated reactivated GAC in organic free water, due to an increase in the effects of surface acidity

with the air-treated carbon, while this trend was reversed when NOM was present. The

hypothesis behind the improved performance of the air treated carbon in water containing NOM

was that the air treated carbon's surface charge, related to its surface acidity, made it less

attractive to the negatively charged NOM leaving more adsorption sites available for MIB. In

both waters, the CO2 treated reactivated GAC removed the greatest quantity of MIB. The CO2

treated reactivated GAC was shown to adsorb slightly less water in a study of water vapor

removal from air, suggesting that the CO2 treated carbon deterred water adsorption, which is

deleterious to MIB adsorption. These findings highlight the importance of investigating changes

in adsorption performance following deliberate changes in surface chemistry, as was done in

order to overcome pH excursions.

Influence of Dissolved Oxygen in Water used to Create Steam

To this point, the impacts of surface chemistry on activated carbon, including its role in

adsorption, as well as methods to tailor this surface chemistry for a specific goal (e.g.,

overcoming pH excursions) have been discussed. Rather than tailor surface chemistry following

reactivation, a method to tailor reactivated carbon's surface chemistry during the reactivation

process has also been explored, namely altering the dissolved oxygen (DO) content of the water

used to create steam for steam reactivation (Chestnutt et al., 2007).

Water may be converted to steam for several industrial applications including laundry

services, locomotion (i.e., steam engines), food manufacturing, wood pulp products, and

activated carbon creation. When water is heated at sufficient temperatures to convert it to steam,

any DO present in the system will devolve as gaseous oxygen (Ol). The devolution of oxygen in

boiler systems has been documented as creating corrosion and pitting within the piping systems









employed in these facilities (Bizot and Bailey, 1997). First, the metal iron in the pipe dissolves

to iron (II) in an anodic reaction followed by water' s reaction with hydroxide ions to form

ferrous hydroxide (Fe(OH)2) (Equation 2-12). The ferrous hydroxide then reacts with oxygen to

form iron (III) hydroxide (Equation 2-13), the dehydration of which forms iron (III) oxide more

commonly known as rust (Bizot and Bailey, 1997).

Fe2+ + 20H- + 4 Fe(OH)2 (s) (2-12)

4Fe(OH)2 + 02 + 2H20 + 4Fe(OH)3(s) (2-13)

This reaction demonstrates the impact that the DO present in water used to create steam

may have on materials with which it comes into contact. With the impact of DO in mind,

Chestnutt et al. (2007) studied the effect of DO in the water used to create steam on the

reactivation of carbon. While several reactivation scenarios (i.e., different time, temperature

regiments, etc.) were evaluated, Figure 2-3 shows an analysis of the performance of three

reactivated carbons exposed to the same reactivation scenario only with differing DO values in

the water used to create the steam. The performance in this study was evaluated using a column

study where the effluent concentration of the target compound, MIB, was evaluated versus the

volume of water treated, or bed volumes (BV). As demonstrated by Figure 2-3, the change in the

DO content of the water used to create steam yielded a difference in the performance of the

resulting reactivated carbon. Namely, the reactivated carbon exposed to steam resulting from

water with high DO treated fewer bed volumes to breakthrough than a reactivated carbon

exposed to steam created from low DO water. Here, breakthrough is defined as the number of

bed volumes treated before the effluent exceeds the odor threshold concentration (OTC) of MIB

(i.e., the concentration of MIB above which it is generally noted by the human palate) of 10 ng/L

(Rashash et al., 1997). The impact of DO on the adsorption performance of the resulting carbons










was related to an increase in surface acidity of the carbons as the level of DO in the water used to

make steam increased. The impact of surface acidity on MIB adsorption has been discussed at

length.

While the mechanism behind the creation of increased acidity with an increase in the DO

content of the water was not investigated in the work of Chestnutt et al. (2007), the results of this

work do indicate that altering the DO of the water used to create steam for reactivation of carbon

is a plausible means by which to alter surface chemistry and thereby influence the adsorption

properties of reactivated carbon.


C C



Formed during air
tarpt ment


O OH
\\C


O C OH


H20


Carbon surface

Figure 2-1. Proposed mechanism for the formation of COOH on an activated carbon surface.


11.5




S10.5

10.0






8.5


100


SE




10A


0 2 4 6
Time (Days)

Figure 2-2. Full-scale water treatment plant effluent pH and calcium concentration data after
reactivated GAC was returned to service (Mazyck et al., 2005).













35 n-I-~MidDO

30 H -A-High DO

-1 25

S20

S15



100


0 2000 4000 6000 8000 10000
BV

Figure 2-3. Comparison of performance in the removal of MIB for reactivated carbons created
with varying DO content in the water used to create steam (Chestnutt et al., 2007).









CHAPTER 3
HYPOTHESIS AND OBJECTIVES

The obj ectives of the current work can be divided into two main categories: research

related to overcoming pH excursions and research related to tailoring activated carbon using the

DO content of the water used to create steam.


pH Excursions

Hypothesis

* Air treatments overcome pH excursions via increased surface acidity, specifically the
carboxyl group and not via suppression of the calcium chemistry mechanism.

* CO2 treatments overcome pH excursions via physisorption of CO2 to the carbon surface
and subsequent desorption to form carbonic acid and not via suppression of the calcium
chemistry mechanism.

* Performance of a CO2 treated carbon for removal of MIB is higher than both air treated
and untreated carbon due to a decreased wetting time associated with the CO2 treated
carbon.

* Performance of an air treated carbon for MIB removal in water containing NOM is better
than that of an untreated carbon due to the preference of the untreated carbon for NOM
adsorption.

* Methods to overcome pH excursions, namely air and CO2 treatments, will adversely affect
the removal of phenol using a granular activated carbon column as increases in surface
acidity are known to be deleterious to the adsorption of phenol and the surface chemistry
of carbon can affect adsorption in both batch and column systems.

Obj ectives

* Investigate individual acidic functional groups on the air-treated, CO2-treated, and
untreated carbons versus the ability of these treatments to overcome pH excursions.

* Investigate mass change of carbon, indicative of CO2 adsorption, during CO2 treatment.

* Evaluate calcium in solution after contact with air- and CO2- treated carbons to determine
relative progression of the calcium chemistry mechanism as compared to an untreated
carbon.

* Determine relative adsorption of MIB for a CO2-treated and an untreated carbon after
contact with only water for various lengths of time to confirm a difference in wetting time.










* Measure concentration of NOM remaining in solution after contact with MIB and an air-
treated, CO2-treated, and an untreated carbon to support hypothesis of competitive
adsorption between MIB and NOM.

* Compare the performance of an air-treated, CO2-treated, and untreated carbon for the
removal of phenol in a column study.

Tailoring of Activated Carbon Using Dissolved Oxygen (DO)

Hypothesis

Altering the DO content of the water used to create steam during steam activation of a wood-
based precursor will alter the surface chemistry/physical structure of the activated carbon
affecting adsorption performance.

Obj ectives

* Evaluate the impacts of four different activation parameters (time, temperature, steam flow,
and DO content of the water used to created steam) at three different levels (low, medium,
and high) on the adsorption performance, physical properties, and chemical properties of
the resulting activated carbon.

* Evaluate different methods of altering DO in the water used to create steam for activation
and their impact on removal of MIB.









CHAPTER 4
EXPERIMENTAL IVETHODS

Reactivated Carbon Samples for Overcoming pH Excursions

Ten coal-based reactivated GAC samples, designed for adsorption from an aqueous

solution, were received from the NORIT Americas, Inc. plant in Pryor, OK. These carbons had

been used in potable water treatment, had become spent, and had been subsequently reactivated

using traditional methods in Pryor, OK. A composite of three of these samples (i.e., these

samples were commingled) having the highest calcium content (5.6-7.8 g Ca/kg carbon) and also

exhibiting the highest pH excursion was used throughout the research into pH excursions. The

BET surface area of the composite was 896 m2/g with a micropore (< 2 nm) volume of 0.3 5

cm3/g and a mesopore/macropore (>2 nm) volume of 0. 13 cm3/g

Carbon Precursor for Activations

A wood-based fly ash was used as the carbonaceous precursor in this study. Fly ash from

wood-combustion has been investigated as an addition to composts to aid in adsorbing odors,

controlling pH and moisture content, and nutrient addition (Campbell et al., 1997; Rosenfeld and

Henry, 2000; Das et al., 2003). Wood-based fly ash also outperformed a coal fly ash in the

catalytic oxidation of H2S and methanethiol, due to the higher surface area of the wood-based fly

ash, in a study of sulfur compound removal from gaseous waste streams (Kastner et al., 2003).

The fly ash used in this study was a side product of combustion taking place at a wood

processing facility in Florida. By utilizing a wood-based fly ash, the activation uses a renewable

resource as well as recycling a waste stream which enhances the environmental impetus to use

such a material.









Air Treatments and CO2 Treatments of Reactivated Carbons

Air treatments were conducted in ambient air using a muffle furnace. Three grams of

carbon were placed in crucibles at the desired temperature inside a muffle furnace where they

were continuously exposed to 52.5 mmol 02/g GAC for 15, 30, or 60-minutes. The air treatment

temperatures used were 623 K, 673 K, and 723 K based on Dussert et al.'s (1995) preferred

temperature range for the air treatments of between 623 K and 773K. After treatment, the

samples were allowed to cool in a dessicator and were stored until further use.

A clamshell furnace similar to that used by Mazyck and Cannon (2000, 2002) held a

quartz-fluidized bed wherein CO2 treatments were carried out at ambient temperatures. A porous

plate near the center of the quartz-fluidized bed allowed a flow of 0.9 L/min of CO2 to contact

three grams of sample for 20 minutes for an exposure of 281 mmol CO2/g carbon. The treatment

is similar to that employed by Farmer et al. (1996) with the exception that dry GAC was exposed

to CO2 instead of wet GAC.

Activations with Varying DO Content in Water

Using the same clamshell furnace used for CO2 treatments, activations involved exposing

five grams of sample, resting on a porous plate near the center of the quartz-fluidized bed, to a

steam flow under the desired conditions. Steam flow is reported as g H20/g carbon which is the

mass of water, determined by the flow of steam as measured with a flow gauge multiplied by the

given period of activation, divided by the mass of carbon with which the steam comes into

contact. Activations were preceded by a pyrolysis step where the five grams of sample were

exposed to the activation temperature in a flow of pure nitrogen gas for a time equal to the steam

exposure (i.e., activation time).

Henry's law was used to calculate the theoretical DO level in the deionized water used to

create steam under varying experimental conditions, as well as the concentration of other









dissolved gases, in order to compare these theoretical values to the measured DO concentrations.

Henry's law (Equation 4-1) relates the concentration of gas in solution to the partial pressure of

the gas above the solution. Partial pressure (Px), according to Dalton's Law, is defined as the

pressure exerted by a single gas in a gaseous mixture where the total pressure equals the sum of

the partial pressures of each gas in the mixture (Petrucci and Harwood, 1997). C equals the

concentration of gas X in solution (mol/L); kH is Henry's constant for gas X (mol/L/atm); and Px

is the partial pressure of gas X (atm).

Cx = kH Px (4-1)

The DO content of the water was manipulated based on the gas (e.g., pure nitrogen, air,

or pure oxygen) used to pressurize a water vessel set at 10 psi (gauge), which then delivered the

water to a preheater set at approximately 723 K which then sent steam to the fluidized bed

reactor. Nitrogen gas pressurization yielded a DO concentration of less than 4 mg/L, air

pressurization yielded a DO of approximately 7.5 mg/L, and pure oxygen pressurization created

a DO concentration of greater than 10 mg/L. The DO concentration was measured inline using a

Mettler Toledo inline DO probe placed just prior to a water gauge which measured water flow

into the preheater. Following activation, the carbon samples were cooled to ambient

temperatures in the furnace under a nitrogen flow to prevent interactions with the atmosphere at

high temperatures.

Water Contact pH

The water contact pH of the carbons was determined by mixing approximately 2 grams of

as-received GAC with 80 mL of Milli-Q water for 30 minutes, then measuring the pH. This ratio

of GAC to solution was the same used by Dussert et al. (1995) and Farmer et al. (1996). In

initial exploratory experiments, pH measurements were taken every five minutes, and it was

observed that after 20 or 30 minutes the pH reached a pseudo-equilibrium level (less than 0.1









change in pH over 5 minutes). Dussert et al. (1995) used a similar technique and concluded that

this protocol provides a good prediction of the pH that would be measured during water

treatment service. For the experiments described herein, each reactivated carbon was sampled

twice and the pH for these two samples was within f 0.4 pH units.

Calcium Solution Concentration

Following the water contact pH experiments, the GAC was separated from the solution

utilizing a vacuum filtration apparatus (0.45 Clm Whatman filter paper). The calcium levels

leached out of the GAC during the 30-minute water contact pH experiment were analyzed by

inductively coupled plasma (ICP) emission spectroscopy (Perkin Elmer, Optima 3200RL). The

measured concentrations of the replicas were within f 3.5% of one another.


Adsorption Experiments (Batch)

For batch equilibrium studies, PAC was created by grinding the carbons in a crucible and

sieving the resulting particles through a 325 mesh (45 Cpm). Then, a 0.5 L stock solution of 153

ng radiolabeled C14-MIB/L was created by blending 95.7 CLL of an 800,000 ng MIB/L-methanol

stock solution with 500 mL of either deionized (DI) water or natural water influent to the

Manatee County Water Treatment Plant (Bradenton, Florida) (Total organic carbon=10.0 mg/L;

Specific UV Absorbance=4.78 L/mg-m; pH=7). Note, MIB (1 ,2,7,7-tetramethyl-exo-

bicyclo(2.21 )heptan-2-ol) is a moderately hydrophobic compound of low molecular weight with

a molecular size of approximately 6 A+ (Newcombe et al., 2002; Rangel-Mendez and Cannon,

2005). From the 153 ng radiolabeled C14-MIB/L solution, 50 mL allotments were added to

separate gas tight syringes followed by addition of 10 ppm carbon. The syringes were then

sealed, permitting a small amount of headspace, and allowed to mix on a rotisserie style mixer.

After 24 hours, a 0.45 Clm luer-lock nylon filter, attached to each syringe, was used to separate










the carbon from 9 mL of the solution, which was then tested according to previous studies using

radiolabeled C14-MIB (Gillogly et al., 1998; Tennant and Mazyck, 2003). Each sample was

measured twice for MIB and the mean of these two runs was used for the final value. Typical

reproducibility was within +1%. The entire process, starting with 50 mL samples of either DI or

natural water, was repeated for each carbon. Each dose experiment was run in replicate and the

error bars shown in the figures represent one standard deviation from the mean of these replicate

runs.

Select samples were analyzed for NOM remaining in solution using a Tekmar Dohrmann

Apollo HS 9000 with an autosampler to measure nonpurgeable dissolved organic carbon (DOC).

Error is represented as plus or minus one standard deviation from the mean of triplicate analyses.

For batch studies with phenol, 30 mL of 5 mg/L phenol where mixed in vials containing

50 ppm carbon for 24 hours. The PAC was then separated from the solution utilizing a vacuum

filtration apparatus (0.45 Clm Whatman filter paper). The samples were stored at 277 K under

low light conditions until they were ready to be measured at which point dilutions were

employed to ensure detection on a gas chromatography-mass spectroscopy instrument (Varian,

Saturn 2100P). The averages of replicate experiments are reported with error bars representing

one standard deviation from the mean.

Rapid Small-Scale Column Tests (RSSCTs)

For column studies with phenol as the target compound, a commercially available

activated carbon (F400) was exposed to the same air and CO2 treatments used to overcome pH

excursions and then these carbons were evaluated using RSSCTs. Details pertaining to the

RSSCT design are listed in Table 4-1. The advantages of RSSCTs are that: (1) they can provide

a breakthrough profile in a fraction of the time compared to full- or pilot-scale studies; (2) a









smaller volume of water is required for testing; and (3) extensive isotherm studies are not

required to predict full-scale performance (Crittenden et al., 1986; Crittenden et al., 1991).

Although the RSSCT can significantly reduce the time and cost of the study, the limitation of the

RSSCT is that it can over-predict full-scale GAC performance (Summers et al., 1989; Crittenden

et al., 1991); nevertheless, it is not expected that the relative comparison between carbons would

change.

The RSSCTs employed in this study were designed using full-scale design parameters

and scaling equations available in the literature (Crittenden et al., 1991). The RSSCT column

was manufactured from acrylic (1.5 cm x 0.41 cm) (0.59 in. x 0.16 in.) with Teflon inserts to

prevent the sorption of compounds. Similarly, a Masterfiex L/S rigid PTFE pump head, PTFE

tubing, and PTFE fittings were used throughout the system to avoid sorption of compounds.

Throughout the entire experiment, a flow meter monitored water throughput (4.4 mL/min), and

all experiments were performed in duplicate.

The various reactivated carbons were evaluated using the RSSCT by spiking aliquots of

deinoized water with a phenol concentration of 5 mg/L. At the time of sampling, a 20 mL vial

was placed at the discharge tube where 13 mL of sample were collected over 3 minutes. The

samples were stored at 277 K under low light conditions until they were ready to be measured at

which point dilutions were employed to ensure detection on a gas chromatography-mass

spectroscopy instrument. The averages of replicate experiments are reported with error bars

representing one standard deviation from the mean.

BET Surface Area Measurements

To determine surface area, the carbon samples were first outgassed at 383 K for 24 hours.

Then, an N2 adsorption isotherm was carried out at relative pressures between 0.05 and 0.35 in a

NOVA 1200 (Quantachrome) gas sorption analyzer. The resulting data was then analyzed using









the BET model to determine the surface area of each sample. Duplicate samples were analyzed

with average error less than 5%.

Pore Size Distribution Measurement

Pore size distributions (PSDs) were performed using a Quantachrome Autosorb I. Each

carbon sample was first outgassed at 383 K for 24 hours. The sample cell was then placed in a

liquid nitrogen bath which created an analysis temperature of approximately 77 K with nitrogen

gas used as the adsorbate. The density functional theory (DFT) was used to develop the pore

size distribution data in the micropores (< 2 nm) from the nitrogen isotherm with a relative

pressure range from 10-6 to 1, while the Barrett, Joyner, Halenda (BJH) model was used for all

pores greater than 2 nm.

pH at the Point of Zero Charge

To determine the pH at the point of zero charge (pHpze), where the point of zero charge is

defined as the condition where the charge on the surface of the carbon is neutral or zero, an

aqueous system where the carbon sample dominates the pH of the system was evaluated. To this

end, 0.5 g of carbon sample was mixed with 5 mL of nanopure water which was previously

bubbled with nitrogen gas for 20 minutes to remove any carbon dioxide from the system that

may interfere with the influence of the carbon sample on the water pH. The headspace above the

solution was also purged with nitrogen gas. The vial was then covered with parafilm to

eliminate mixing with the atmosphere and the sample was allowed to mix for 24 hours in order

for the carbon to reach equilibrium with the solution. After 24 hours, the pH of the sample was

determined using a standard pH electrode and this pH measurement was recorded as the pHpze.

Each pHpze measurement was performed in duplicate with each experiment performed in

replicate. The replicates were within +1%.









Boehm Titration

The concentration of acidic oxygen complexes created on the surface of the carbon

samples was determined according to the Boehm titration method (Boehm, 1966). An allotment

of 0.5 g of carbon and 25 mL of either 0.05 N NaOH, 0.25 N NaOH, 0.05 N NaHCO3, or 0.05 N

Na2CO3 WeTO Shaken in sealed vials for three days. The suspensions were then filtered using a

0.45 Cpm filter and 20 mL of the filtrate was titrated with 0. 1 N H2SO4 to a pH of 4.5. The

amount of base consumed by each sample was determined by comparing each titration to that of

a blank. Any base consumed by the sample results from the neutralization of acidic functional

groups and is dependant on the type of base used. Carboxyl groups react with NaHCO3; lactone

groups are measured with the difference between the reaction with Na2CO2 and NaHCO3,

phenolic groups are measured with the difference between reactions with 0.05 NaOH and

Na2CO2; the difference between 0.25 NaOH and 0.05 NaOH measures carbonyl groups; total

acidity is measured with 0.05 NaOH. The error in the measure of functional groups represents

the highest and lowest values found in repeated experiments.

Statistical Design of Experiments: Box Behnken

Traditionally, the evaluation of the effect of several variables on a result is made using a

full factorial, one-factor-at-a-time, or the fractional factorial approach to obtaining data. The full

factorial approach involves the collection of data relating to all possible combinations of the

factors (variables). With four factors under investigation in this research and a proposed three

levels (settings) for each factor, the full factorial method corresponds to 81 experiments. The

benefit of such an approach is that the choice of optimum conditions will be clear when the data

are analyzed; the optimum experimental conditions will be run as one of the experiments.

However, performing at least 81 experiments, as this number does not take into account replicate

experiments, can be expensive in terms of both cost and time.









Given the drawbacks of a full factorial approach, many researchers revert to the one-

factor-at-a-time approach, which evaluates the effect of varying one factor while holding all

other factors constant. Once the level that produces the best result has been determined for this

factor, the level is fixed while another factor is evaluated with the remaining factors constant.

The one-factor-at-a-time approach used in this research would yield nine experiments. Clearly,

this approach would be tempting to a researcher when compared with the 81 runs required for a

full factorial approach. However, the one-factor-at-a-time approach has significant drawbacks

when considering the validity of the final optimum conditions, as it does not take into account

interactions. An interaction between factors occurs when two or more factors acting together

have an effect on the result greater than the factors acting alone.

The fractional factorial approach mediates the cost and result validity issue by utilizing

the principles of the design of experiments (DOE). DOE is defined as the evaluation of the

simultaneous effect of changing variables on the final product in order to gain the maximum

amount of definitive information with the minimal amount of expended resources. The

fractional factorial approach uses only a predetermined set of experiments chosen from the full

factorial. Unlike the one-factor-at-a-time approach, there is a logical plan behind the choice of

experiments run from the full factorial and statistical analysis yields a final result with a specific

associated value of certainty, without the requirement to run each combination in the full

factorial. Therefore, fractional factorials, specifically a Box-Behnken design (Box and Behnken,

1960), will be used in this research with 29 runs required.

The proposed Box-Behnken design is formed by combining a 2k factorial design, where 2

is the number of levels and there are k variables being studied, with incomplete block designs to









produce a 3k fractional factorial (Montgomery, 1991). A 22 full factorial design is shown in

Table 4-2.

In a Box-Behnken design, each pair of factors being studied is linked to a 22 factorial

where the two factors are varied at two levels (high and low) while the third factor is fixed at its

center level (middle). The combination of factorials in this fashion for a Box-Behnken 33 design

is shown in Table 4-3.

A visual representation of the Box-Behnken design of 33 is Shown in Figure 4-1, where

each dot represents experiments and the vertices represent the three factors. It should be noted

from the 33 design that, although 12 experiments are required to complete the combination of

factorials, there is also an additional point at the center.

A 34 design, which is the suggested Box-Behnken design for this experiment, requires

four axis for which the visual representation is less simple but the same theory applies. This

design would contain 24 design points plus a certain number of runs at the center point. It is

from this replication of center points that analysis of error can be made. Myers (1995)

recommends 3-5 replicates of the center points and the upper limit of this will be used in this

Box-Behnken design for a total of 29 runs (i.e., 24 runs plus five runs at the center point).

The center point is replicated to obtain the pure error associated with the experiment.

Pure error in an experiment accounts for any random error in experimentation that is not

accounted for by the lack of the data' s fit to the proposed model. By utilizing the five replicate

experiments at the center point, pure error has four degrees of freedom (five replicate

experiments at the center point minus one) and there are 10 degrees of freedom for lack of fit. In

this way, the Box-Behnken is able to incorporate a level of significance in the final result without

the need to repeat every experiment. To ensure that the error is random, the required 29 runs are










performed in a computer generated random order. The pure error is then used to analyze the lack

of fit, while the total residual error (i.e., pure error plus lack of fit) is used to determine the

significance of each factor evaluated according to the common evaluation method known as the

F-value. Equation 4-1 shows how to calculate the F-value to determine the significance of the

model's lack of fit. It is desired that the models lack of fit be insignificant (i.e., it is desired that

the generated model to predict results fits well with the data).

Sum of Squares Inck of fit/degree of freedom lack of fit
F-value = (4-1)
Sum of Squares Pure Err-or/degr ee of freedom pure error

To determine whether an F-value is significant to a 0.05 level of significance, or a 95%

confidence interval, as used in this study, the P-value (Prob > F) was employed. The P-value

gives the probability of obtaining an F-value at least that high due solely to random error, where

a P-value less than 0.05 would indicate a significant parameter such that there is less than a 5%

chance that an F-value that large could be due solely to random error. Equation 4-2 shows how

to calculate the F-value to determine the significance of a factor.

Sum of Squares Factor/degree of freedom factor
F-value = (4-2)
Sum of Squares Residual/degree of freedom residual

For this study, the Box-Benken approach was evaluated using Design-Expert software,

version 6.0.5, distributed by Stat-Ease, Inc. in 2001.

Table 4-1. RSSCT and full-scale design parameters
Parameter Full-scale RSSCT*
Grain Size (U.S. mesh) 12 x 40 170 x 200
EBCT (min) 7.5 0.045
Hydraulic loading rate (mm/min) 167.5 335.1
Time to process 1000 BV 5.2 d 0.74 hr
*Constant diffusivity conditions based on Speitel et al. (2001)









Table 4-2. 22 full factorial
Experiment Variable A Variable B
1 Low Low
2 Low High
3 High Low
4 High High

Table 4-3. Box-Behnken 33 fractional factorial
Experiment Variable A Variable B Variable C
1 Low Low Middle
2 Low High Middle
3 High Low Middle
4 High High Middle
5 Low Middle Low
6 Low Middle High
7 High Middle Low
8 High Middle High
9 Middle Low Low
10 Middle Low High
11 Middle High Low
12 Middle High High
Center run Middle Middle Middle









*--~~--




Figure 4-1. Box-Behnken design of 33









CHAPTER 5
STRATEGIES FOR OVERCOMING pH EXCURSIONS FOR REACTIVATED GRANULAR
ACTIVATED CARBON: AIR AND CARBON DIOXIDE TREATMENTS

Previous research into the mechanisms behind pH excursions have demonstrated the

plausibility of a calcium chemistry mechanism behind pH excursions experienced with

reactivated carbons (Mazyck et al., 2005). Further research has suggested that mechanisms to

overcome pH excursions, namely air and carbon dioxide (CO2) treatments, are effective in

overcoming pH excursions due to addition of acidity to solution either by increased total surface

acidity, as with air treatments, or physisorption of carbon dioxide and subsequent desorption to

form carbonic acid, as with CO2 treatments (Bach, 2004). However, additional research needs to

be done to support these mechanisms and, therefore, the intent of the current work is to bolster

the hypothesis made in the previous work via additional experiments. Due to the fact that the

current work builds directly from previous work, certain figures have been reproduced for the

reader' s benefit and are marked as "(Bach, 2004)" to differentiate the previous work from new

work done in this study.


Creation of a pH Stable Carbon

Air treatments of a coal-based reactivated carbon were conducted according to the patent

of Farmer et al. (1998) to investigate the creation of a pH stable reactivated GAC which has a

water contact pH less than 8.5 pH units, as this pH falls between the NPDES permitted discharge

pH of 6 to 9 while still allowing for a small safety factor (Bach, 2004). A pH stable carbon is not

expected to exhibit a pH excursion upon return to service (Dussert et al., 1995; Farmer et al.,

1998; Mazyck et al., 2005). The resulting water contact pH of the air treated carbons, treated at

three different temperatures and treatment times, measured in deionized (DI) water, is shown in

Figure 5-1. Dashed lines at a water contact pH of 8.5 and 6.5 highlight the cutoff between an










acceptable and unacceptable water contact pH incorporating a safety factor. The first data bar, at

time zero, represents the water contact pH of the untreated carbon (i.e., reactivated GAC).

While a pH stable carbon is not created at any of the three treatment times when the

temperatures is 623 K, after 60-minutes of air treatment at 673 K and after only 30-minutes of

treatment at 723 K a pH stable carbon is created. These results indicate that treatments to

overcome pH excursions with virgin carbons are also successful in overcoming pH excursions

with reactivated carbon in DI water.

The CO2 treatment scenario, wherein the composite sample was exposed to a CO2 flOW Of

0.9 L/min for 20 minutes (281 mmol CO2/g) at ambient temperature, also created a pH stable

carbon (pH=8.3) (Bach, 2004). The CO2 flow rate and contact time were chosen based on

preliminary experiments demonstrating that this time and flow rate were sufficient to create a pH

stable carbon (Figure 5-2). At 20 minutes, a pH stable carbon was created at the lowest carbon

dioxide flow (i.e., 0.9 L/min) and, therefore, this treatment time and flow were used in all

subsequent analyses. Again, the proposed method to overcome pH excursions experienced with

virgin carbons in water containing anions is also successful in overcoming pH excursions with

reactivated GAC in DI water.

Surface Acidity of Air-Treated Carbons

The air treatments tested herein bound the ideal temperature for production of acidic

functional groups, specifically carboxyl groups, which leads to the hypothesis that air treatments

overcome pH excursions by creating acidic functional groups on the GAC surface which balance

out the hydroxyl ions imparted to the water via the calcium chemistry mechanism. The quantity

of individual functional groups on the surface of air treated carbons with a treatment time of 60-

minutes and the untreated carbon is shown in Figure 5-3.









As expected, the air treatments that created pH stable carbons (i.e., treatments at 673 K

and 723 K) show an increase in total acidity compared to the untreated carbon. The 623 K

treatment, which did not create a pH stable carbon, had a similar quantity of oxygen-containing

functional groups as the untreated carbon except for lactone groups, which were slightly more

prevalent in the air treated carbon. Interestingly, though 60-minute air treatments at 673 and 723

K yielded pH stable carbons, only the treatment at 723 K showed an increase in the quantity of

carboxyl groups compared to the untreated carbon. This result suggests that several acidic

functional groups are responsible for creating pH stable carbons, as the 673 K treatment yielded

a pH stable carbon without a significant increase in carboxyl groups. The 723 K air treated

carbon also showed an increase in the lactone and phenolic groups compared to all other

samples. The carbonyl functional group was found to form to a significant extent in the 673 K

sample compared to all other samples and, like the 723 K sample, it showed an increase in the

lactone and phenolic groups compared to the untreated sample. Given this information, it seems

likely that the lactone and phenolic groups play a role in creating a pH stable carbon at these

temperatures.

Surface Acidity of CO2 Treated Carbons

The CO2 treated and the untreated carbon yielded similar quantities of acidic functional

groups on the carbon surfaces (Figure 5-4), indicating that addition of surface acidity is not

responsible for overcoming pH excursions with CO2 treatment.


Here, the mechanism that creates a pH stable carbon is likely physisorption of CO2 to the

carbon surface. Upon immersion into water, the CO2 desorbs and reacts with water to form

carbonic acid (H2CO3) and then bicarbonate (HCO3-) after deprotonation at the pH typically

observed in water treatment. Based on calculations predicting the pH of a system where only the









dissolution of CO2 from the atmosphere and the calcium chemistry mechanism influences pH,

the theoretical equilibrium pH is approximately 8.6 compared to the measured water contact pH

of 8.3 with the CO2 treated carbon. The theoretical equilibrium pH for the untreated carbon is

also approximately 8.5 though the measured water contact pH is 10.7. It is possible that the

additional CO2, which is expected to dissolve upon contact with water, on the CO2 treated

carbon's surface allows the system to reach its equilibrium pH at a faster rate than the untreated

carbon. An analysis of the weight gain following CO2 treatment showed a slight increase in

weight which, assuming the mass gain is due solely to CO2 physisorption, corresponds to an

addition of 1.1*10"5 M CO2. The equilibrium concentration of CO2 dissolved into solution from

the atmosphere at room temperature is 0.9*10"5 M CO2/L, which may support the hypothesis that

CO2 treatment increases the rate at which the solution reaches equilibrium with the atmosphere.

For example calculations of CO2 in SOlution, see Appendix A.

As with air treatment, it is proposed that the CO2 treatment balances out the basic effects

of the calcium chemistry mechanism, for the bicarbonate buffers the hydroxyl ions released from

the calcium hydrolysis, thus depressing the magnitude of the pH excursion. It is not anticipated

that CO2 treatment forms CaCO3 to any great extent, which would not participate in the pH

excursion mechanism, because the optimum chemisorption temperature of CO2 to CaO, to form

CaCO3, is near 573 K and the CO2 treatments used herein occurred at ambient temperatures (295

K) (Mazyck and Cannon, 2000).

Verification of Mechanisms

To support the hypothesis that neither air treatments nor CO2 treatments inhibit the

calcium chemistry mechanism from progressing, an examination of the calcium that dissolved

from the carbons shown in Figure 5-3 and Figure 5-4 was performed (Table 5-1).









An analysis of the calcium in solution after CO2 treatment showed that the untreated

carbon yielded 34 mg/L Ca2+ and the CO2 treated carbon created approximately 30 mg/L Ca2+

These findings suggest that CO2 treatment results in a slight decrease in dissolved calcium,

though it is not significant enough to fully account for the creation of a pH stable carbon as this

change of 4 mg/L Ca2+ WOuld only be expected to decrease the pH by 0.7 versus the observed

decrease in pH of 2.1i.

Table 5-1 also shows an unexpected increase in dissolved calcium present in solution at

air treatment temperatures of 673 and 723 K (60 and 69 mg/L, respectively). An increase in the

concentration of calcium in solution suggests that there is more calcium involved in the proposed

calcium chemistry mechanism responsible for the pH excursion. A plausible explanation is that

not all of the calcium adsorbed to the activated carbon during service reacted to form CaO during

reactivation. During the subsequent air treatment to create a pH stable carbon, the calcium not

already present as CaO (e.g., CaCO3) reacted to form CaO, the thermodynamically stable species

at the temperatures and partial pressure of oxygen used during air treatments herein (Mazyck and

Cannon, 2000). Conversely, a portion of the original adsorbed calcium-NOM complex may have

remained through reactivation and was subsequently converted to CaO by the air treatments

employed herein. To determine if air treatments act to create more CaO on the carbon surface,

which then participate in the proposed calcium chemistry mechanism, treatments at the same

temperature but in a nitrogen environment were performed and the resulting calcium in solution

evaluated. The nitrogen treatments resulted in approximately half the concentration of calcium

in solution (32 and 34 mg/L at 673 and 723 K, respectively) that was obtained with air treatments

at the same temperature, supporting the hypothesis presented above. This suggests interplay

between the addition of increased basicity (as demonstrated by an increase in dissolved calcium










and thus more hydroxyl ions) and increased acidity (as demonstrated by Figure 5-3) with air

treatments .

Summary and Conclusions

Air treatments and CO2 treatments designed to create pH stable carbons from virgin

carbons in anion-containing water were successful at creating pH stable reactivated carbons in

deionized (ion-free) water. These findings support the work of Mazyck et al. (2005) which

questioned the plausibility of an anion-exchange mechanism and the suppression thereof as an

explanation for the creation of a pH stable carbon. In other words, the treatments created pH

stable carbons even in anion-free water.

Air treatments of reactivated carbons at 673 K and 60-minutes or 723 K for 30- or 60-

minutes were found to create pH stable carbons through the formation of acidic oxygen-

containing functional groups on the carbon surface. It is proposed that CO2 treatments form pH

stable reactivated carbons via physisorption, and subsequent desorption, of CO2 forming H2CO3

and subsequently HCO3- upon water immersion. While an increase in the mass of the carbon due

to CO2 physisorption was observed, additional experiments would need to be performed to

confirm this hypothesized mechanism. However, an analysis of the functional groups on the

CO2 treated carbons' surface and an analysis of the calcium in solution following contact with

the treated carbons ruled out both increased acidity and calcium chemistry suppression as

plausible mechanisms behind the creation of a pH stable carbon with CO2 treatment.









Table 5-1. Calcium in solution with treatments
Temperature (K) Calcium in solution (+ Img/L)
Ambient (Untreated) 34
Ambient (CO2) 30
623 (Air) 37
673 (Air) 60
723 (Air) 69
673 (Nitrogen) 32
723 (Nitrogen) 34


O Untreated
O 623 K
O3 673 K
S723 K


-1- ~










15


30
ime (min)


Figure 5-1. Effect of air exposure time and temperature on water contact pH of reactivated
carbon. Dashed lines represent a pH of 8.5 and 6.5, respectively (Bach, 2004).























64


60



























2 '' "' "
5 10 20
Time (min.)

Figure 5-2. Effect of treatment time and gas flow on resulting water contact pH of CO2 treated
carbon.


600


? 500

400


S300


S200

S100
L.


Untreated


623 K 673 K


723 K


Figure 5-3. Effect of air treatments after 60-minute contact time at varying temperature on
individual acidic functional groups.











250
oCarboxyl
cna Lactone
'0 2 00
I Phenolic
u, CCarbonyl
0- 150
o 0Total Acidity

"a 100


50



Untreated Carbon Dioxide


Figure 5-4. Effect of CO2 treatments on individual functional groups.









CHAPTER 6
METHODOLOGIES FOR OVERCOMING pH EXCURSIONS FOR REACTIVATED
GRANULAR ACTIVATED CARBON: EFFECT ON ADSORPTION PERFORMANCE

The work discussed in Chapter 5 supported the hypothesis that air treatments of

reactivated carbons create pH stable carbons (i.e., pH < 8.5) through the formation of acidic

oxygen-containing functional groups on the carbon surface which balance out hydroxyl ions

added by the calcium chemistry mechanism. The hypothesis that CO2 treatments form pH stable

reactivated carbons via physisorption, and subsequent desorption, of CO2 forming H2CO3 and

then HCO3- upon water immersion which again balances out the hydroxyl addition by the

calcium chemistry mechanism was also supported. The addition of oxygen to the surface of an

activated carbon and its effect on adsorption is well-known and therefore a follow-up study by

Bach (2004) evaluated the removal of the target compound MIB using the treated carbons.

To determine the effect that the treatments designed to create pH stable carbons have on

adsorption, a pH stable air treated carbon (673 K, 60-minute treatment), a CO2 treated carbon,

and an untreated carbon were evaluated for their removal of MIB in both deionized (DI) water

(i.e., organic free) and natural water (containing 10 mg/L natural organic matter (NOM)

measured as total organic carbon (TOC)). The characteristics of the pH stable carbons and an

untreated carbon are shown for comparison in Table 6-1. Note that the BET surface area of the

carbons can be considered similar. The percent removal of MIB in DI water versus the

powdered carbon dose of a CO2 treated, air treated, or an untreated carbon evaluated in the work

of Bach (2004) is reproduced in Figure 6-1.


Figure 6-1 indicates that the CO2 treated carbon had a higher adsorption capacity for MIB

than either the air treated or untreated carbon and the air treated carbon had less of an adsorption









capacity for MIB compared to the untreated carbon. Note that MIB removal is not affected by

pH change and MIB does not affect pH (Herzing et al., 1977).

The studies of Pendleton et al. (1997) showed the effects of activated carbon surface

chemistry on the adsorption of MIB and linked decreased adsorption with an increase in the

hydrophilic nature of an oxidized carbon. As MIB will adsorb to both hydrophobic and

hydrophilic adsorption sites, direct competition for adsorption sites between water and MIB, the

mechanism suggested by Pendleton and coworkers, would not explain the decreased adsorption

of MB with oxygenation of the carbon's surface. Therefore, given the increased acidity of the

air treated carbons (Table 6-1) and previous research showing the impact of an oxidized surface

on adsorption, the air treated carbon in the study of Bach (2004) did not remove as much MIB as

the untreated carbon due to the presence of water clusters on the oxygenated surface of the air

treated carbon which impeded MIB adsorption.

While the impact of air treatment and the associated oxidation of the carbon surface on

adsorption are well understood, an increase in removal with CO2 treatment was unexpected given

the similar surface acidity of the CO2 treated carbon compared to the untreated carbon. Bach

(2004) proposed that the CO2 treated carbon delays adsorption of water as the CO2 is desorbed

from the surface thereby leaving the adsorption sites free for MIB adsorption. A slight increase

in the adsorption of water vapor following CO2 treatment, while indicative of a change in the

way carbon behaves with water molecules, may not translate to the water based system used in

the adsorption study and therefore additional work needs to be done to confirm this hypothesis.

NOM is ubiquitous in surface waters and the concentration of NOM can be 100,000

times that of the trace contaminant MIB and, furthermore, the adsorption of NOM has been

shown to decrease the removal of MIB via competitive adsorption (Newcombe et al., 2002;









Tennant, 2004). Therefore, it was of interest to compare the performance of the air treated, CO2

treated, and untreated carbons for MIB removal in natural water which was done in the work of

Bach (2004) and is reproduced for the reader's convenience in Figure 6-2.

Unlike the DI water scenario, the air treated carbon removed more MIB in natural water

than did the untreated carbon. NOM can adsorb through electrostatic forces (as well as non-

electrostatic forces) wherein the negative charge of the NOM is attracted to a positive charge on

the carbon surface (Newcombe, 1999). The pHpze,the pH at the point of zero charge where the

surface carries a negligible charge, can be used to determine the charge on a surface

(Ravichandran, 1998; Tennant, 2004). In a solution with a pH above the pHpze, the carbon

surface will carry a negative charge and a solution with a pH below the pHpze will yield a

positive surface charge. The untreated carbon with a pHpze of 10.4 (Table 6-1) in contact with a

natural water pH of approximately 7 will have a positive surface charge. Given this difference in

surface charge, it was hypothesized that the untreated carbon would adsorb more NOM and, as

there are fewer adsorption sites available after NOM adsorption, less MIB adsorption would

occur (i.e., competitive adsorption). Bach (2004) also hypothesized that the air treated carbon,

with a pHpze of 8 in contact with a natural water of pH 7, will have less of a positive surface

charge than the untreated carbon and will not adsorb as much NOM, leaving more adsorption

sites available for MIB. However, the NOM remaining in solution was not evaluated to support

the hypothesis of competitive adsorption.

Therefore, the intent of the present work is to bolster the hypotheses made in the work of

Bach (2004) concerning differences in adsorption performance of the treated carbons. An

additional goal is to evaluate the removal of a compound other than MIB, one that will have

different chemical properties, to determine if the pH excursion treatments impact the adsorption









of other compounds. To this end, the removal of phenol by the treated carbons in a column study

was also be evaluated.

Impact of Wetting Time on MIB Adsorption for CO2 Treated Carbon

The CO2 treated carbon was found to remove more MIB (10%) than the untreated carbon

at the lower carbon doses (Figure 6-1), despite the similarity in their total acidity (Table 6-1). A

possible explanation for this occurrence is a disparity between the time it takes the CO2 treated

carbon to wet compared to the untreated carbon. To determine if this proposed mechanism is

plausible, the adsorption experiments were repeated for the CO2 treated carbon and the untreated

carbon with DI water except that the respective carbons were mixed with DI water for a period

of time (i.e., 0, 5, 60, and 150 minutes; wetting time) before the MIB was added, instead of

simultaneous water/MIB addition. The results of this experiment are shown in Figure 6-3 with

error bars representing one standard deviation of replicate experiments.

As the wetting time increased (i.e., the longer the carbon was in contact with only DI

water before addition of MIB) the lower the adsorption capabilities of both the CO2 and the

untreated carbon. This is expected, as the carbons will adsorb water which blocks the adsorption

sites available for MIB when it is added after 5, 60, or 150 minutes. The observation that the

adsorption capacities of the CO2 and untreated carbon are equal after 150 minutes wetting

indicates that the adsorption advantage of the CO2 treated carbon is likely related to its

temporarily decreased water adsorption. This phenomenon is expected to hold true even in

natural waters, where the CO2 treated carbon also outperformed the air treated and untreated

carbons (Figure 6-2).

In separate experiments, CO2 bubbling of an MIB solution containing the untreated

carbon demonstrated no effect on MIB removal when the volatilization of MIB from solution

was taken into account. Volatilization under these experimental conditions (i.e., bubbling with










CO2) yielded an MIB removal of approximately 50%. Therefore, simply the presence of CO2 in

water and any effect it has on pH did not affect MIB adsorption, a finding supported by Herzing

et al. (1977).

MIB Adsorption from Natural Water

An increase in removal of MIB with the air treated carbon at low carbon doses compared

to the untreated carbon, as shown in Figure 6-2, was hypothesized to relate to competitive

adsorption with NOM. The phenomenon of decreased MIB removal with an increase in NOM

removal over a range of pH values is supported in the work of Tennant (2004). To substantiate

this hypothesis in the work of Bach (2004), Table 6-2 shows the results of the evaluation of the

concentration of NOM remaining in solution, represented by dissolved organic carbon (DOC),

after 24-hr of contact with a 20 ppm dose of carbon and 153 ng/L MIB. The concentration of

NOM remaining in solution is lowest for the untreated carbon and highest with the air treated

and CO2 treated carbon, as anticipated by the theory that NOM is attracted to the positive charge

on the untreated carbon. Given that fewer adsorption sites are available for MIB once NOM has

adsorbed to the surface, it is anticipated that the removal of MIB will decrease as the removal of

NOM increases, which is the case when comparing Figure 6-2 to Table 6-2. The CO2 treated

carbon outperforms the air treated carbon, despite the similar removal of NOM, due to the CO2

treated carbon's increased wetting time.

Phenol Adsorption from Deionized Water

To determine whether the impact on adsorption performance with treatments to overcome

pH excursions is dependant on MIB being the target compound and on batch studies, the

performance of an air treated, CO2 treated, and an untreated commercially available carbon

(F400) was evaluated for the removal of phenol in DI water via a column study. The results of









this study are shown in Figure 6-4 where the effluent phenol concentration from the column is

shown versus bed volumes (BV) treated.

As with the batch studies, the treatments to overcome pH excursions, performed on a

commercially available activated carbon in this case, showed a difference in adsorption

performance depending on the treatment used. Not surprisingly, given the dependence of phenol

adsorption on the oxygenation of the surface (Bansal et al., 2002; Salame and Bandosz, 2003),

the air treated carbon results in fewer bed volumes treated before a measurable concentration of

phenol is produced in the effluent. Moreover, the phenol concentration is higher in the effluent

of the air treated carbon' s column at similar bed volumes treated by the CO2 and untreated

carbons. Again, the impact of the increased surface oxygenation of the air-treated carbon,

negatively impacts the performance of the air treated carbon for phenol removal due to an

increase in the carbon's surface oxygenation.

Interestingly, Figure 6-4 does not show an increase in the performance of the CO2

treatment as compared to the untreated carbon as found with MIB removal in batch studies. It is

possible that the difference in wetting, which played a role in the batch studies, does not manifest

in the column studies due to the difference in adsorption kinetics (e.g., batch versus flow-through

systems). More likely, the slight impact on adsorption that results from a difference in wetting

time (e.g., approximately 10% improved removal of MIB) in DI water batch studies containing

ng/L concentrations of MIB does not manifest in a solution containing mg/L concentrations of

the target compound (phenol). In other words, the high concentration of phenol drives the

compound onto the carbon's surface via a concentration gradient that overcomes the slight

influence of wetting time. Under these conditions, the dominating factor may be the relative










oxygenation of the carbon' s surface and, as has been previously discussed, the carbon dioxide

and untreated carbon have similar surface acidity.

Conclusions

The adsorption capacity for MIB of two GACs that have undergone two separate

treatments to overcome pH excursions, namely air and CO2 treatments, were compared in a study

by Bach (2004). The untreated carbon removed more MIB than the air treated carbon while this

trend was reversed when NOM was present. Increase in surface acidity with the air treated

carbon created preferential adsorption sites for water, which out-competed MIB in water not

containing NOM. When NOM was present, the negatively charged NOM was attracted to the

positively charged surface of the untreated carbon via electrostatic forces, thus out-competing

MIB for adsorption sites. In addition, the hypothesis that in both water scenarios the CO2 treated

carbon removed the most MIB due to a longer wetting time was confirmed by analysis ofMIB

removal versus wetting time. Finally, an analysis of the removal of phenol following air and

CO2 treatments on a commercially available carbon showed a decrease in performance with air

treatment, as found in batch studies with MIB. However, similar performance of the CO2 and

untreated carbons for removal of phenol was found as compared to the improved performance of

the CO2 carbon versus the untreated carbon in MIB removal. These results reaffirm that

treatments to create pH stable reactivated carbons will change the carbon's adsorption

capabilities and, thus, use of these treatments should be based on more than their ability to create

pH stable carbons.










Table 6-1. Characteristics of sample carbons
Treatment pHpze BET surface area (+ 5%) Total acidity (peq/g)
Carbon dioxide 8.3 885 170
Air 8.0 927 400
Untreated 10.4 896 160

Table 6-2. DOC remaining in solution after contact with 20 ppm carbon
Treatment DOC remaining in solution (mg/L)
Carbon dioxide 36.8 & 0.7
Air 34.8 & 1.0
Untreated 31.0 &1.5


o
E 60


E 40


20


0


1 2 6
Carbon Dose (mg/L)

of MIB in DI water versus dose of carbon (24-hr contact time)


Figure 6-1. Percent removal
(Bach, 2004)










100


80



4 0



20


0


0 Carbon Dioxide Treated
O Air Treated =
. o Untreated


20 40
Carbon Dose (mg/L)


Figure 6-2. Percent removal of MIB in raw water (TOC = 10 mg/L) versus dose of carbon (24-hr
contact time) (Bach, 2004).


60

50


$ 40

p~30

20

10

0


0 Carbon Dioxide Treated
0 Untreated


Wetting Time (min.)

Figure 6-3. Percent MIB removal versus wetting time in DI water with a CO2 and untreated
carbons (carbon dose = 2ppm, contact time = 24 hrs).














~tUntreated F400











0 5000 10000 15000
BV

Figure 6-4. Effluent phenol concentration in DI water versus BV treated with a CO2 treated, air
treated, and an untreated carbon.









CHAPTER 7
IMPACT OF DISSOLVED OXYGEN ON ACTIVATION OF POWDERED ACTIVATED
CARBON

Despite the millenniums over which activated carbon has been used and the centuries of

research into the mechanisms behind activated carbon's success, there is currently minimal

research effort being put into tailoring the carbon surface to enhance carbon's capacity for

specific contaminants. The lack of tailoring means that activated carbon is often not utilized to

its full potential. For example, perchlorate, a chemical found in rocket fuel and increasingly

discovered in drinking water supplies, had minimal affinity for activated carbon in preliminary

studies (Chen et al., 2005a). However, subsequent studies using an ammonia-tailoring process

found that the tailored activated carbon removed four times as much perchlorate as a standard

activated carbon, indicating that many studies using standard activated carbon may not represent

the full capabilities of activated carbon for this particular pollutant (Chen et al., 2005a, 2005b).

As new contaminants are found and existing regulations made more stringent, the ability

to efficiently tailor activated carbons for removal of specific compounds, by manipulation of

pore size and/or surface chemistry, is crucial. Therefore, the obj ective of this work is to tailor

the thermal activation of a wood-based carbon by changing the dissolved oxygen (DO) content

of the water used to create the steam used for activation. Research by Chestnutt et al. (2007) has

determined that altering the DO in the water used to create steam for reactivation of carbon can

alter the resulting reactivated carbon but it could be possible to modify the activated carbon from

its inception using this method rather than during the reactivation process. By developing a

method of activation that requires only a shift in DO content to create activated carbons with

differing properties, capable of adsorbing a variety of contaminants, activated carbon is thus

made adaptable to the changing types of and standards for contaminant removal. To this end, the

ability of the DO content of the water used to create steam for activation of a wood-based fly ash









to impact adsorption was evaluated along with the impact on the carbon's chemical and physical

properties. In addition, the interaction of the DO and typical tailoring parameters of temperature,

time, and oxidant flow rate was evaluated.

Dissolved Gases in Solution

Alteration of the dissolved gases in solution will occur as a result of attempts to change

the DO and, therefore, the quantity and possible effects of the change in these dissolved gases

should be evaluated. The dissolved gases in solution chosen for evaluation at the individual

experimental conditions were those naturally found in the atmosphere at the four highest

concentrations, namely, nitrogen, oxygen, argon, and carbon dioxide. The system employed in

this work includes a pressure vessel set at 10 psi (gauge) for a total pressure on the system of

24.7 psi (absolute) where the pressuring gas was air, oxygen, or nitrogen. For comparison, the

theoretical concentration of dissolved gases in a system at atmospheric pressure was also

calculated. Complete purity of the gas above the systems pressurized in nitrogen and oxygen is

likely not a realistic condition but, for the purpose of this exercise, it is assumed that any air

remaining in the system will contribute only negligibly to the final dissolved gas concentrations.

The theoretical dissolved gas concentrations are shown in Table 7-1 along with the measured

values for the DO. An example calculation using Henry's law can be found in Appendix A.

A theoretical DO value of 9. 1 mg/L at atmospheric pressure fit within expected values,

while DO levels under pressurization of the system, even in air, were higher than the measured

values. A brief review of industrial techniques shows that utilizing a pressure vessel to increase

the DO level is not unique. The purpose of dissolved air flotation (DAF) is to infuse water with

a high level of DO so that, when the DO devolves from solution, the oxygen bubbles will lift

microscopic particles to the surface which are then removed (Appleton, 2006). DO levels used

in DAF technology may be as high as 40 mg/L and may be attained through pressurization. A









patent by Barnett (1999) employs a pressure system to imbue water with DO at five times

atmospheric pressure, which is indicated as approximately 50 mg/L, to aid in biological

treatment of a wastewater. These findings indicate that the theoretical DO levels found under

pressurization are within reason.

However, there is a significant discrepancy between the theoretical DO level, predicted

using Henry's law, and the measured DO level. The reason for the failure of Henry' s law to fit

the experimental data obtained could be one of several. While Henry's law is expected to fail at

high pressures (Petrucci and Harwood, 1997), research has shown that the Henry' s constant of

nitrogen from 14.4 psi to approximately 573 psi only varied by 10% and only 8% in a study of

nitrogen with oil at pressures of 14.7 psi to 735 psi (Gerth, 1985). Given that the experiments

herein employ pressure not even twice that of atmospheric, it is not considered likely that the

system deviates significantly from Henry's law due solely to an increase in pressure.

Henry's law also applies only in systems at equilibrium and, moreover, gas/liquid

equilibrium is a very slow process compared to many other reactions involving dissolved

species. Equilibrium in gas/liquid systems depends on the interfacial area available for gas

transfer, which explains why bubbling a liquid with gas increases the rate of equilibration as the

small gas bubbles create higher surface area for water/gas contact compared to gas above a layer

of water (Benjamin, 2002). The system employed herein did utilize some form of bubbling,

more in the nitrogen system than in others; the system with oxygen easily exceeded the limit of

detection of the DO probe (i.e., greater than 20 mg/L) with even seconds of oxygen bubbling.

Given the ease with which the DO level can be changed by bubbling the water with oxygen for a

short amount of time, it is likely this latter phenomenon is creating the discrepancy between the

experimental DO levels and that predicted by Henry's law; namely, the system is not at










equilibrium. The portion of the reaction that progresses over the duration of the experiment

leads to the experimentally determined values which, as seen in Table 7-1, are at most half as

high as the theoretical values. Of most interest is the difference between theoretical and

experimental values for the oxygen pressurization scenario. For the purposes of this

experiment' s goals ( i.e., determination of the effect of a change in oxygen), the gas/liquid

system does not need to be at equilibrium as long as the DO level is relatively steady throughout

the experiment and uniform between experiments. The DO level throughout experiments was

easily held within +/- 1 mg/L.

At high temperatures, as the water turns from liquid to vapor, the dissolved gases will

devolve and will be present at higher than atmospheric concentrations inside the furnace. The

use of high temperatures in inert environments, such as nitrogen and argon, as a means to remove

oxygen-containing functional groups and create a more basic carbon is common (Leon y Leon et

al., 1992; Menendez et al., 1996; Barton et al., 1997; Considine et al., 2001; Lillo-Rodenas et al.,

2005). However, it is not expected that the small concentration of inert gas (i.e., nitrogen and

argon) (e.g., 8*10-6 mol at the highest) versus oxidizing gas (i.e., water vapor and oxygen) (e.g.,

0.886 mol H20) is high enough to qualify as an inert environment where the chemisorption of

oxygen would be deterred (See Appendix A for example calculation of moles of dissolved gas

versus moles of water vapor). Aside from the possibility of creating an inert environment during

the oxidation step, the adsorption of these inert gases during the oxidation step would be

negligible as the preceding pyrolysis step would saturate the carbon with as much nitrogen as it

could adsorb at that temperature (Puri, 1970).

Given that carbon dioxide in solution can change the pH of the system and its associated

alkalinity, it is of interest to calculate the theoretical shift in pH with an increase in carbon










dioxide in solution. Whereas water exposed to the air at atmospheric pressure would be expected

to have a pH of 5.7, based solely on carbon dioxide dissolution in water, water exposed to air at

1.68 times atmospheric pressure would be expected to have a pH of 5.59, only slightly less than

that at atmospheric pressure (See Appendix A for calculations). It is not expected that this slight

shift in pH will cause any significant change in the activation of carbon with this water used as

the basis to make steam.

Carbon dioxide is also used as an oxidant in the activation of carbon according to the

reactions shown in Chapter 2. However, given that the concentration of carbon dioxide is an

order of magnitude less than that of oxygen, which can react with twice as many reactive sites as

can carbon dioxide (see Chapter 2), the contribution of carbon dioxide activation in all scenarios

seems negligible. Given the discussion of the relative impact of nitrogen, argon, and carbon

dioxide, it is clear why changes in dissolved oxygen are taken as the dominant influencing factor

in these activations.

Box-Behnken Approach and Mass Loss

In order to evaluate the impact of DO and other activation parameters on the resulting

activated carbon, the Box-Behnken approach, an example of a statistical design of experiments,

was employed. The first step in the Box-Behnken design process is to determine the levels of

each of the four factors to be evaluated (i.e., activation time, activation temperature, steam to

carbon ratio (i.e., steam flow), and DO level). Preliminary experiments indicated what levels of

the four factors would allow for even spacing of levels, as required by the Box-Behnken

approach, be practical to employ, and encompass the broadest range of impact on the carbon. A

more detailed discussion of the selection of the levels of these parameters, including results of

preliminary tests, can be found in Appendix B. The results of these studies yielded the three

levels of the four factors shown in Table 7-2.










Next, 29 separate activated carbons were created according to the Box-Behnken design of

experiments. Of these carbons, 25 were unique based on the setting of the activation time,

activation temperature, SC, and DO level while four samples were replicates of the center point.

A complete list of the activation parameters for each of the 29 carbons can be found in Appendix



The initial mass of carbon before activation and the final mass after activation were

recorded for each carbon in order to determine the mass lost during activation. The mass loss is

a measure of the gasification of the carbon (i.e., the degree of activation) and was the first

response evaluated. The mass loss data were input into the Design-Expert software which

analyzed the significance of each factor in mass loss and produced a model based on these data

to predict the mass loss that would result over the entire range of levels and factors studied. For

a complete list of mass loss for each carbon and the entire output from Design-Expert, see

Appendix C. Pertinent statistical information and results from the Box-Behnken analysis have

been summarized in Table 7-3. Note that a Prob > F (P-value) greater than 0.05 for lack of fit

(i.e., insignificant lack of fit) and a low standard deviation, which includes both the lack of fit

and random error, are preferred.

The statistically significant factors (Prob > F of less than 0.05) in mass loss were found to

be temperature and time with significant interactions between temperature and time and also

temperature and SC. The final column of Table 7-3 represents the predicted weighting of each

parameter where the summation of the activation parameters, multiplied by their weighting

factor, and the intercept would yield the predicted mass loss. Note that a negative weighting

factor indicates a detrimental impact of this factor on mass loss. Figure 7-1 shows a 3-D model

produced by Design-Expert of the predicted mass loss, in percent of initial mass, versus time and










temperature of activation, both significant factors. The levels of the remaining two factors are

set at their medium value.

The model resulting from the Box-Behnken analysis predicts that as the time and

temperature of activation increases so will the mass loss of the carbon. The slight curves in the

graph are indicative of an interaction where the two factors acting together have an influence

greater than the factor acting alone. The statistical significance of the interaction between time

and temperature is affirmed by the results of Design-Expert as shown in Table 7-3. Indeed, the

predicted model is a quadratic model due to the presence of an interaction between temperature

and time and also temperature and SC.

The influence of temperature and time on mass loss during activation, namely an increase

in mass loss with an increase in time and temperature, is an expected result as this phenomenon

is well-supported in the literature (Gergova et al., 1993; Gonzalez et al., 1994; Rodriguez-

Reinoso et al., 1995; Gergova et al., 1996; Martin-Gullon et al., 1996; Tennant and Mazyck,

2003). An increase in temperature will speed gasification reactions, which remove carbon atoms

and therefore mass, while an increase in activation time will allow for more gasification. At

623K, the interaction between time and temperature of activation is evident as the increase in

mass loss with time is negligible indicating that at this low temperature there is minimal

gasification that will occur such that longer treatment times do not impact mass loss. Indeed, an

activation temperature of 623 K is low compared to that normally employed in activation (e.g.,

1073K) (Kawamura, 2000). The influence of treatment time and temperature on surface area is

confirmed in Table 7-4, where there is an increase in surface area with time and temperature

except at 623 K where surface area is unaffected by an increase in treatment time.









The interaction between temperature and SC in mass loss can be seen in Figure 7-2 in the

curved sides of this 3-D graph. Note, however, that SC alone is not a statistically significant

factor influencing mass loss; SC is only significant when interacting with temperature.

An interaction between temperature and SC has been shown in the literature where only

at high temperatures does the steam flow influence the gasification of the resulting carbon.

Hydrogn is produced as a reaction of the water molecule with the carbon surface and also the

water-gas shift reaction (See Chapter 2). The hydrogen may then react with an active site on the

carbon surface which prohibits further reaction of the steam with that active site, thus inhibiting

gasification (Walker et al., 1959; Puri, 1970; Menendez et al., 1996; Lussier et al., 1998; Bansal

et al., 2002). It has been postulated by Rodriguez-Reinoso et al. (1995) that at higher steam flow

and higher temperatures the inhibiting reactant, hydrogen, is less effective. Therefore, an

increase in steam (i.e., oxidant) will produce greater gasification and mass loss but only at higher

temperatures where the creation of the inhibiting agent hydrogen, which would increase with an

increase in steam flow, does not influence gasification. The influence of SC and temperature on

surface area is supported in Table 7-4, where there is an increase in surface area with SC except

at low temperature.

The Design-Expert results for mass loss, indicating a positive relationship between time

and temperature and also an interaction with time and temperature and with SC and temperature,

are an affirmation of the reliability of this approach as these results conform to phenomenon

found in the literature concerning activation with steam. DO was not found to be a significant

parameter in mass loss during activation.

MIB Removal

Following creation of the 29 activated carbons, these carbons were then powdered to

create a PAC slurry for use in subsequent batch adsorption studies. The first performance










response of interest was the performance of the activated carbons for removal of the common

taste- and odor-causing compound MIB. Preliminary experiments demonstrated that a PAC dose

of 10 mg/L in contact with 153 ng/L of MIB for 24 hours would yield a suitable range of

removal. The MIB removal data were input into the Design-Expert software which analyzed the

significance of each factor in the removal of MIB and produced a quadratic model to predict the

performance of the carbons over the entire range of levels studied. For a complete list of the

MIB removal of each carbon and the statistical results of Design-Expert, see Appendix C.

Pertinent statistical information for the MIB removal response is summarized in Table 7-5.

All parameters were found to be statistically significant factors (Prob > F of less than

0.05) influencing MIB removal, whether acting independently or as an interaction, as expected

given that the selection of these factors was based on their influence on adsorption of MIB.

While the final column of Table 7-5 contains the weighting factor of each parameter' s influence

on the predicted MIB removal, Figure 7-3 shows a 3-D model of the predicted removal of MIB

versus both time and temperature of activation. As the time and temperature of activation

increases, so does the removal of MB. Note that the shape of the mass loss and MIB removal

curves versus time and temperature are similar at higher temperatures. Figure 7-4 shows MIB

removal versus SC and DO, the interaction of which was found to be significant, as well as SC

alone, in the removal of MIB.

SC acting alone has a significant impact on removal of MIB but the interaction between

DO and SC is also significant as demonstrated in Figure 7-4 where the trend in removal changes

depending on the level of DO. At low SC, an increase in DO is detrimental while at high SC a

change in DO yields negligible change in MIB removal.









The trends in MIB removal with changes in temperature, activation time, SC, and DO

relate to chemical and/or physical changes in the carbon caused by these activation parameters.

As such, the following sections explore the chemical and physical properties of the activated

carbons in order to elucidate the role of each of the four factors on these properties and

sub sequent adsorption of MIB.

Physical Properties

The mass loss, which can be taken as a general indicator of the gasification of the

carbons, was shown to be impacted by temperature and time. It is well known that an increase in

mass loss generally correlates to an increase in both surface area and pore volume as gasification

removes carbon atoms from the internal surface of the carbon (Gergova et al., 1994; Centeno et

al., 1995; Rodriguez-Reinoso et al., 1995; Gergova et al., 1996; Mazyck and Cannon, 2002;

Zhang et al., 2003). Therefore, it is expected that the increase in MIB removal with time and

temperature, predicted in Figure 7-3, is due in part to the increase in the surface area of the

carbons related to temperature and time as found in Table 7-4. An improvement in the removal

of MIB with time at low temperature, where the physical structure of the carbon remains similar,

may be related to changes in surface chemistry as discussed further in later sections.

In order to evaluate the influence of the extreme levels of SC and DO on the physical

properties of the carbon, and to determine if physical properties dominate the observed

phenomenon of Figure 7-4, four carbons representing the four corners of Figure 7-4 were

evaluated for their BET surface area and pore size distributions. Descriptions of the conditions

of activation for these four carbons as well as their BET surface area and their pore volumes are

shown in Table 7-6. These carbons were labeled based on a code where SC is designated S with

levels at low (L) or high (H) and the same applies for the DO (D) level.









Within standard error for BET surface area measurements, only SHDL and SLDH have

different surface area. A similar BET surface area but an increase in mass loss as predicted with

an increase in steam flow (Figure 7-2) indicates that the changes in steam flow yield different

pore size distributions making up this surface area. A summary of the pore volume is presented

in Table 7-6 with the PSD plots shown in Figure 7-5.

Evaluation of the PSDs of these four carbons shows a predominance of micropores with a

small volume of mesopores and macropores. In a study by Tennant and Mazyck (2003) for MIB

removal using PAC, carbons having a large volume of large micropores and small mesopores

(1.2-10 nm) were found to correlate to the highest removal of MIB. Therefore, this range is also

presented in Table 7-6 as this gives the range of pore sizes that are expected to most significantly

affect MIB removal.

From Table 7-6, it is apparent that an increase in steam flow (SHDL versus SLDL or

SHDH versus SLDH) will produce a higher volume of pores in the range of 1.2-10 nm. An

increase in the PSD with an increase in steam flow is anticipated based on the correlation

between increase in steam flow and mass loss, where a higher concentration of oxidant yields

more gasification. However, the similarity in SHDH and SLDL indicates that a decrease in

steam flow can still produce a similar pore volume to a higher steam flow if only the DO content

were decreased which would have advantages in terms of cost savings of having to pump a lower

flow of steam to get the same result. SLDH, having a low steam flow and also high DO, has the

lowest volume of pores in the 1.2-10 nm range.

The influence of DO on the pore size distribution, namely the occurrence of a higher

volume of 1.2-10 nm with a lower DO may relate to a study of the activation of carbon with

oxygen by Tomkow et al. (1977). These authors suggested that only limited micropores are









formed during activation with oxygen due to the formation of oxygen-containing functional

groups at the entrance of the micropores which cuts off the micropores from further activation as

the oxidant cannot fit into the narrowed entrance created by the functional groups. In the current

research, where the DO in the water is expected to devolve upon conversion of the water into

steam, the small quantities of free oxygen may react with the carbon surface to create oxygen-

containing functional groups which would impede further activation by steam. The influence of

DO on the surface chemistry of the carbon will be discussed in the following section.

As the volume of 1.2-10 nm pores are expected to correlate to MIB adsorption, it is

anticipated that SLDH, with the lowest volume in the ideal range of pore sizes, would remove

the least MIB. In the predicted MIB removal of Figure 7-4 this theory is confirmed in that

SLDH is the only carbon to have a significantly different removal of MB. The similarity in the

ideal range volume of SLDL and SHDH explains their similar removal of MIB but one may

expect that SHDL would have higher removal as it has a higher volume of ideal pores compared

to the other carbons. It is most likely that the small increase in this ideal pore range volume from

SLDL and SHDL (i.e., 13%) compared to the large increase in this pore range from SLDH and

SHDH (i.e., 39%) is not significant enough to manifest differences in performance outside of the

standard deviation.

It is clear that the DO content of the water used to create steam for activation of these

carbons influenced the physical structures of these carbons but the influence of the DO on the

chemical properties of the carbon must also be investigated.

Surface Chemistry

Following creation of the 29 activated carbons, the samples were then analyzed for their

pH at the point of zero charge (pHpze). The pHpze is often used as a measure of the carbon' s

surface acidity where a lower pH,ze indicates a more acidic surface (El-Sayed and Bandosz,









2003; Ania et al., 2004). Furthermore, activated carbon acidity can be related to the content of

the oxygen-containing functional groups on the carbon surface which are often acidic (Puri,

1970; Otake and Jenkins, 1993; Barton, 1997; Ania et al., 2004). The pH,ze data was input into

the Design-Expert software which analyzed the significance of each factor in the resulting pHpze

of the carbons and produced a linear model to predict the pHpze of the carbons over the entire

range of levels studied. For a complete list of the pHpze values of each carbon and the statistical

results of the Design-Expert process, see Appendix C. Pertinent statistical information for the

pHpze response is shown in Table 7-7.

A linear model best fit the data, which means that no interactions were statistically

significant and only temperature and SC were found to be statistically significant factors

impacting the pHpze. Figure 7-6 shows a 3-D model of the predicted pHpze versus both SC and

temperature of activation.

As demonstrated in Figure 7-6, as both the SC and temperature of activation increase so

does the pHpze (i.e., the activated carbon becomes more basic/less acidic). The impact of

temperature on pHpze relates to the decomposition of acidic oxygen-containing functional groups

starting at 773K. Removal of oxygen containing-functional groups to create a more basic carbon

through the use of high temperatures in inert environments is a common practice (Leon y Leon et

al., 1992; Menendez et al., 1996; Barton et al., 1997; Considine et al., 2001; Lillo-Rodenas et al.,

2005). While inert conditions at high temperatures were not deliberately employed for this

purpose in the current research, following activation the reactor was allowed to cool to room

temperature in a flow of nitrogen gas before the carbon sample was removed. During the

cooling of the sample from as high as 973K to less than 373K, the carbon is briefly exposed to

high temperatures in an inert environment which may strip the surface of the functional groups









formed during the activation process. A less acidic carbon with increasing activation

temperature supports the finding of Figure 7-3, where improved MIB removal was predicted as

temperature of activation increases; a less acidic surface would attract fewer water molecules

leavings the adsorption sites free to remove MIB (Pendleton et al., 1997; Considine et al., 2001).

Though the pHpze is predicted to be constant regardless of activation time, as activation time was

not found to be a significant factor in pHpze, there may be changes in the quantity of basic groups

with time that are not evident in this study of surface chemistry but manifest in improved

removal of MIB with activation time at even the lowest temperature employed (Figure 7-3).

The occurrence of an increase in steam flow relating to an increase in pHpze (i.e., an

increase in basicity) may be explained by an associated increase in the formation of hydrogen as

the steam flow is increased. Recall that hydrogen is created as a result of the water molecule' s

reaction with the carbon's surface and/or the water-gas shift reaction. The free hydrogen

molecule would then react with active sites to form hydrogen-containing functional groups

(Walker et al., 1996). These hydrogen-containing functional groups would occur on sites that

could have contained an oxygen-containing functional group, as Bansal et al. (1974) determined

that both oxygen and hydrogen adsorb to the same reactive sites, so that the formation of an

acidic-oxygen containing functional group is avoided. Aside from possible direct competition

of the hydrogen and oxidant for adsorption sites, the hydrogen also blocks active sites from

further reactions with atmospheric oxygen to create functional groups following activation. The

process of preventing oxygen chemisorption by preadsorption with hydrogen has been proposed

by Bansal et al. (1974) and confirmed by Mendendez et al. (1996) where hydrogen exposure was

used to create a stable basic carbon.









As a less acidic surface is favorable for the removal of MIB, an increase in steam flow,

the creation of hydrogen, and the corresponding increase in pH,ze would be expected to

correspond to improved MIB removal. In support of the link between steam, hydrogen, and

MIB removal, Nowack et al. (2004) observed similar performance for removal of MIB in a

steam treated carbon compared to a carbon treated in hydrogen at a similar temperature. These

authors took the similar performance of the steam and hydrogen treated carbons as support of the

influence of the water-gas shift' s production of hydrogen and subsequent reaction of this

hydrogen with the carbon surface on improved removal of MIB. The influence of steam flow on

surface chemistry supports the results of Figure 7-4 which shows that at high DO there is an

improved removal of MIB with increasing SC.

The expected increase in performance as SC is increased is only significant at high DO,

at low DO the performance of the carbons remains the same regardless of steam flow. It is

possible that at high DO, the competition provided by hydrogen is more significant as there are

more oxidants to compete with (i.e., both steam and Ol). Without this competition from

hydrogen, the higher oxygen in contact with the carbon, as the DO devolves from the water upon

creation of steam, results in the creation of more acidic functional groups and therefore a more

acidic carbon. At low DO, hydrogen is competing with a smaller quantity of oxygen and the

resulting impact of competition with hydrogen is less significant and produces negligible results

on total acidity. However, the DO level was not found to significantly impact the total acidity of

the carbon, which could indicate that, while the impact on the total number of oxygen-containing

functional groups remains the same regardless of DO, the type of individual oxygen-containing

functional groups changes with the DO level. A change in the dominant type of oxygen-

containing functional group would not necessarily change the total acidity.









Activation with oxygen would produce active sites with different energy levels than those

produced by steam reactions and these active sites may react to form different functional groups

when exposed to the atmosphere. Puri (1970) notes that the activity of the surface may

determine the type of group formed, with more active surfaces producing lactones while less

active sites produce carbonyls. Otake and Jenkins (1993) and Molina-Sabio et al. (1996) also

make note of differences in the thermal stability of surface oxides depending on the method of

oxidation (air versus HNO3 activation and steam versus CO2 activation, respectively). The

findings of these studies, along with the knowledge of the relative stability of functional groups

created by oxygen and steam reactions, where the exothermic nature of the C-O2 adsorption step

makes the resulting functional groups short-lived, support a hypothesis that different oxidants

can yield different types of functional groups. Therefore, by passifying the surface of the carbon

in low DO versus high DO, the formation of different functional groups may be avoided though

the total acidity remains unchanged. In other words, if oxygen reacts with the surface in the high

DO conditions, different types of individual functional groups may result than if mostly steam

reacted with the surface, as in the low DO conditions.

A preliminary analysis of the precursor and activated carbons demonstrated that these

carbons are highly basic, as indicated in Figure 7-6, which makes their analysis via the Boehm

titration method, a common wet method of evaluating individual acidic functional groups,

difficult. In a study by El-Sayed and Bandosz (2003) of individual functional groups using the

Boehm titration method, when the total basicity was almost 3 times the measured total acidity the

Boehm titration method measured negligible quantities of individual acidic functional groups.

Therefore, to evaluate whether surface chemistry is a factor in the observed changes in Figure 7-

4, the same four carbons representing the four corners of Figure 7-4 and described in Table 7-6









were evaluated for their removal of MIB in dichloromethane, a nonaqueous solvent. The goal of

this experiment is to determine whether MIB removal changes depending on the activation

conditions even in a scenario where the acidic surface's interaction with the polar water molecule

has been removed. The MIB removal of these four carbons in the dichloromethane solvent,

where the carbon dose was increased to 1100 mg/L, is shown in Table 7-8.

The removal of MIB in dichloromethane is low compared to the removal of MIB in

water. Considine et al. (2003) also evaluated the removal of MIB from dichloromethane using

activated carbon and the decrease in removal of MIB in dichloromethane was explained by the

difference in the polarity of the solvents and the way the solvent interacts with the carbon

surface. Within standard error, the removals of MIB in dichloromethane with the four corner

point carbons are similar. The similar removal supports the theory that a difference in surface

chemistry is partly responsible for the difference in removal of the SLDH carbon, as when the

water molecules and the hypothesized interaction of the water molecules with the acidic

functional groups are removed from the solution SLDH performs just as well as the other three

carbons.

Chemical versus Physical Methods of Removing DO

Removal of DO can be accomplished either by physical means (e.g., bubbling with

nitrogen gas or passing the water through membranes) or chemical means (e.g., addition of an

oxygen scavenger). For use in boiler streams, it has been found that physical methods do not

allow for removal of all of the dissolved oxygen and therefore the streams are often polished

with an oxygen scavenger (Bizot and Bailey, 1997). Sodium sulfite is a common oxygen

scavenger employed in the boiler industry because it is relatively cheap, safe, and effective at

removing DO. As this study has so far focused on the removal of oxygen through physical

means, it was also of interest to evaluate a chemical means of removing dissolved oxygen to









determine if the results are similar. To this end, the SHDL and SLDL activation scenarios

described in Table 7-6 were repeated but with sodium sulfite used to decrease the DO. The

carbons which mimic the DO levels used in the Box-Behnken approach (i.e., 3 mg/L DO) are

labeled SHDLs and SLDLs, where the subscript s is used to denote that these conditions were

created as a result of sodium sulfite addition. In addition, two carbons were created under the

same conditions as SHDL and SLDL except that the DO was reduced to 0 mg/L, which is easily

done with sodium sulfite compared to physical removal of DO which cannot remove all DO.

The 0 mg/L DO carbons are labeled SHDLo and SLDLo. The MIB removal and the chemical

and physical properties of these four activated carbons are shown in Table 7-9.

Both sets of carbon, comparing low and high SC at a given DO concentration, have

similar BET surface area and similar trends in pHpze, namely a decrease in pHpze at low steam

flow. The removal of MIB with the SHDLs and SLDLs also showed similar trends to the

removal of MB with SHDL and SLDL (approximately 62% MIB removal) with no significant

change in removal with a change in SC. The MIB removal with SHDLo and SLDLo, however,

showed a decrease in removal with SHDLo compared to SLDLo. The difference in removal with

the 0 mg/L DO activations may relate to the difference in their ideal pore size volume (1.2-10

nm). The PSD plots for these carbons are shown in Figure 7-7.

Of the four carbons created with sodium sulfite as an oxygen scavenger, only SHDLo had

a significantly different pore volume in the ideal range of pores than the other three carbons. The

decrease in the volume of this ideal pore range explains the decrease in MIB removal with this

carbon. It may be that at this lower DO, where more sodium sulfite was added to reach 0 mg/L

of DO, and a higher SC, where a higher volume of water vapor containing sodium sulfite










contacted the carbon, more sodium sulfite interacts with the carbon surface, thus impacting

adsorption.

Phenol Removal

The same PAC slurries created from the 29 activated carbons of the Box Behnken where

used in a phenol batch adsorption study. The performance of the carbons for removal of phenol

is of interest because of the different adsorption mechanism for the adsorption of phenol (i.e, xn- xn

bonding) versus MIB adsorption. In addition, phenol is a common toxic organic compound in

surface waters for which activated carbon is used. Preliminary experiments demonstrated that a

PAC dose of 50 mg/L in contact with 5 mg/L of phenol for 24 hours would yield a range of

removal that was easily detectable. The phenol removal data was input into Design-Expert

software which analyzed the significance of each factor in the removal of phenol and produced a

linear model, indicating a lack of statistically significant interactions between factors, to predict

the performance of the carbons over the entire range of levels studied. For a complete list of the

phenol removal of each carbon and the statistical results of the Design-Expert process, see

Appendix C. Pertinent statistical information for the phenol removal response is shown in Table

7-10.

Only temperature and time were found to be statistically significant factors influencing

phenol removal. Figure 7-8 shows a 3-D model of the predicted removal of phenol versus both

time and temperature of activation.

As with MIB removal, where both time and temperature were found to be significant

factors, it is likely that increases in surface area and/or pore size distribution associated with

increased temperature of activation and treatment time simply yielded more adsorption sites for

phenol removal. In addition, the decrease in total surface acidity found with increasing

temperature may also play a role in the increased removal of phenol with temperature as an









increase in surface acidity has also been found to be detrimental to phenol adsorption (Bansal et

al., 2002; Ania et al., 2004).

Neither steam flow nor DO was found to be a significant factor in phenol removal. It is

possible that the influence of surface chemistry and physical changes brought about by DO were

overcome by the higher concentrations of phenol employed herein such that the concentration

gradient drove the phenol onto the carbon surface despite the obstacles. Furthermore, as it was

suggested that the different functional groups may have been created with different levels of DO

and this is what impacts MIB removal, the particular acidic functional groups formed in the low

SC and high DO carbons may not impact phenol adsorption as it impacts MIB adsorption. For

example, in a study by Salame and Bandosz (2003) the decreased removal of phenol by wood-

based activated carbons was shown to have a stronger correlation to an increase in the quantity of

carboxylic groups than to total surface acidity. Another possibility is that the wood-based

activated carbon used herein responds differently than coal-based carbons, as the difference in

coal and wood-based carbons has been discussed. Specifically, one adsorption mechanism for

phenol onto activated carbon is xn- xn bonding where the aromatic phenol and the aromatic xn-

electrons on the carbon surface interact. However, wood and wood-based precursors are seen to

have higher H/C elemental ratios, which suggest a lower aromatic structure, than their coal-based

counterparts (Buczek et al., 1993; Gergove et al., 1993; Pendleton et al., 1997; El-Sayed and

Bandosz, 2003; Fernades et al., 2003). Therefore, the removal of phenol with the wood-based

activated carbon used herein may not be impacted by changes in surface chemistry as it may

have a low occurrence of xn- xn bonding in its original form.

Summary and Conclusions

The Box-Behnken approach was employed to evaluate the influence of four factors on

the physical and chemical properties and performance of the resulting activated carbons. An









initial evaluation of mass loss following activation using the Box-Behnken design confirmed the

validity of this approach as an increase in mass loss with activation time and activation

temperatures as well as an interaction between activation temperature and SC are phenomenon

supported in the literature.

The Box-Behnken approach also predicted an increase in MIB and phenol removal with

activation time and temperature. An increase in the physical properties of the carbon with

activation time and temperature, supported by an observed increase in mass loss under these

conditions, as well as a decrease in surface acidity with an increase in activation temperature

were deemed responsible for an increase in removal of these compounds.

An interaction between SC and DO was also seen with the removal of MIB where only at

high DO and low SC did these factors impact MIB removal in the form of decreased removal.

Neither DO nor SC was found to be significant in the removal of phenol. The physical

properties of the carbon, namely a decrease in the PSDs of the carbons, explained a decrease in

removal at the highest DO and lowest SC level. However, a low SC level was also found to

correspond to a more acidic carbon which can impact MIB removal. While the total acidity of

the carbons treated at low SC remained unchanged regardless of the DO level, the high DO

activated carbon had decreased removal of MB. It was proposed that different groups are

formed at high DO, where the free oxygen devolved from the water upon creation of steam,

while the total acidity remains unchanged. A lack of difference in removal found in

dichloromethane, a nonpolar solvent, supports that the difference in removal found with low SC

and high DO may be a combination of both physical and surface chemistry changes, though this

theory would need to be confirmed with new techniques to evaluate surface chemistry on the

heterogeneous and basic activated carbons used herein.









An additional study of DO removal with sodium sulfite indicated that high levels of

sodium sulfite may be detrimental to MIB removal, though the amount required to decrease DO

to 3 mg/L did not impact performance.

While the impact of DO on surface chemistry may require further study, it is clear that

the DO level in the water used to create steam for activation of carbon does influence the

resulting carbon's ability to adsorb certain compounds.

Table 7-1. Concentration of dissolved gases in water at 273 K
Gas (mg/L) Air (14.7 psi) Air (24.7 psi) Oxygen (24.7 psi) Nitrogen (24.7 psi)
Nitrogen 14.8 25.2 0 32.3
Oxygen 9.1 15.5 73.6 0
Oxygen (measured) 9.0 7-8 12-13 2-3
Argon 0.6 0.9 0 0
Carbon Dioxide 0.5 0.9 0 0

Table 7-2. Levels of four factors employed in the Box-Behnken design
Factor Low Medium High
Temperature (K) 673 823 973
Time (min.) 10 30 50
Steam to carbon ratio (SC) (g H20/g carbon) 0.4 1.4 2.4
DO (mg/L) 3 7.5 12

Table 7-3. Box-Behnken statistical results for mass loss
Parameter F-value Prob > F-value Coefficients of Value
Equation
Lack of fit (F-value) 2.90 0.156
DO (mg/L) 0.82 0.377 0.91
Temperature (K) 185.0 < 0.0001 -0.14
Time (min.) 7.54 0.0125 -0.31
SC (g H20/ g carbon) 3.78 0.0662 -7.39
Temperature2 39.6 < 0.0001 0.00013
DO *Temperature 3.63 0.0713 -0.0018
Temperature Time 9.53 0.0058 0.00065
Temperature SC 12.1 0.0023 0.015
Intercept 45.79
Standard deviation 1.27
R-squared 0.93










Table 7-4. Surface area versus treatment time, treatment temperature, and SC (DO = 7.5 mg/L)
Temperature (K) Time (min.) SC (g H20/g carbon) BET Surface Area (m2/g) (A 5%)
973 10 1.4 541
973 50 1.4 615
673 10 1.4 470
673 50 1.4 498
673 30 0.4 468
673 30 2.4 496
973 30 0.4 537
973 30 2.4 606

Table 7-5. Box-Behnken statistical results for MIB removal
Parameter F-value Prob > F-value Coefficients of Value
Equation
Lack of fit (F-value) 2.44 0.2009
DO (mg/L) 4.12 0.0559 -4.76
Temperature (K) 73.6 < 0.0001 -0.038
Time (min.) 39.2 < 0.0001 1.12
SC (g H20/ g carbon) 5.16 0.0344 -10.33
DO2 2.25 0.1491 0.2
Temperature2 14.5 0.0011 0.00046
DO *Time 2.70 0.1158 -0.064
DO *SC 6.47 0.0193 1.99
Intercept 126.7
Standard deviation 7.0
R-squared 0.88

Table 7-6. BET surface area and total pore volume of four carbons
Activated SC DO BET Micro- Meso- Macro- Pore
carbon (g H20/ (mg/L) surface pore pore pore volume
label g carbon) area volume volume volume (1.2-10
(A 5%) (< 2 nm) (2-50 nm) (>50 nm) nm)
(cc/g) (cc/g) (cc/g) (cc/g)
SHDL 2.4 3 501 0.232 0.021 0.003 0.0371
SLDL 0.4 3 461 0.217 0.021 0.006 0.0327
SHDH 2.4 12 478 0.221 0.022 0.006 0.0356
SLDH 0.4 12 430 0.202 0.020 0.006 0.0257









Table 7-7. Box-Behnken statistical results for pHpze
Prob > F-value Coefficients of Value
Parameter F-value
Equation
Lack of fit (F-value) 0.34 0.9544
DO (mg/L) 0.83 0.3699 -0.017
Temperature (K) 143.5 < 0.0001 0.0066
Time (min.) 2.32 0.1409 0.0063
SC (g H20/ g carbon) 13.4 0.0013 0.30
Intercept 7.0
Standard deviation 0.28
R-squared 0.87

Table 7-8. Activation conditions of four carbons and their removal of MIB in dichloromethane
(time = 30 min., temperature = 873K)
Activated carbon label MIB removal (%) (A 1.6%)
SHDL 5.4
SLDL 5.8
SHDH 6.2
SLDH 5.8

Table 7-9. BET surface area and total pore volume carbons treated with sodium sulaite
Activated MIB pHpze BET Micro- Meso- Macro- Pore volume
carbon removal (A 0.1) surface pore pore pore (1.2-10 nm)
label (%) area volume volume volume (cc/g)
(A 2.5%) (m2/g) (< 2 nm) (2-50 nm) (>50 nm)
(' 5%) (cc/g) (cc/g) (cc/g)
SHDLs 64 11.67 506 0.227 0.025 0.006 0.0381
SLDLs 59 11.31 484 0.230 0.021 0.006 0.0361
SHDLo 49 11.63 469 0.225 0.019 0.006 0.0305
SLDLo 57 11.37 494 0.236 0.019 0.006 0.0355

Table 7-10. Box-Behnken statistical results for phenol removal
Prob > F-value Coefficients of Value
Parameter F-value
Equation
Lack of fit (F-value) 2.64 0.1791
DO (mg/L) 0.21 0.6517 -0.10
Temperature (K) 47.8 < 0.0001 0.046
Time (min.) 4.92 0.0362 0.11
SC (g H20/ g carbon) 1.47 0.2368 1.22
Intercept 29.8
Standard deviation 3.47
R-squared 0.69



























'10 ~673 Temperature (K)
Figure 7-1. Mass loss versus time and temperature (SC = 1.4 g H20/g carbon, DO = 7.5 mg/L)





22
18
15
11




2.4
1.9 973
1 898
SC 823
(gH2/ crbn0.9 78 Temperature (K)
0.4 673

Figure 7-2. Mass loss versus SC and temperature (time = 30 min., DO = 7 mg/L)













6 79
"ii 64
o49
e34



50
40` 973
30 898
Time (min.) 20` 823
748 Temperature (K)
10 673
Figure 7-3. MIB removal versus time and temperature (SC = 1.4 g H20/g carbon, DO = 7.5
mg/L)


67
~-60 -


E 47
P~40


(g H20/g carbon) DO (mg/L)
Figure 7-4. MIB removal versus SC and DO (time = 30 min., temperature = 873K)












0.3




S0.2
o

O






0
1 10
Pore Size

Figure 7-5. Pore size distributions of sample carbons


100
e (A)


1000


12.4
11.7
11.1
10.5
9.8


Figure 7-6. pHpze versus SC and temperature (time


30 min., DO = 7 mg/L)
















;*0.
O

0.1







0
1 10 100
Pore Size (A)

Figure 7-7. PSDs of carbons treated with sodium sulfite




S69
S64
g 60
E 55
5
O s


1000


Figure 7-8. Phenol removal versus time and temperature (SC = 1.4 g H20/g carbon, DO = 7
mg/L)









CHAPTER 8
CONTRIBUTIONS TO SCIENCE

* First to determine that activated carbon treated with CO2 yields improved removal of MIB
due to a delay in the wetting of the carbon surface;


* Supported the importance of investigating individual acidic functional groups on the surface
of activated carbon as not all of the individual acidic functional groups measured were
altered following air treatment and yet the water contact pH and the air-treated carbon's
performance in the removal of MIB and NOM was altered following air-treatment;

* First to demonstrate that variations in the DO content of water used to create steam for
activation of a wood-based carbon alters the subsequent performance of the activated carbon
for removal of MIB through a combination of physical and chemical changes in the carbon
surface;


* Highlighted the necessity to tailor activated carbon for a specific contaminant as the
performance of the activated carbons for removal of phenol, unlike MIB, was unaffected by
alterations in the DO content of the water used to create steam for activation.









CHAPTER 9
FUTURE WORK

* Explore the impact of wetting time of activated carbon, as it pertains to the slurries by which
PAC and GAC are transported, on the resulting adsorption performance of the activated
carbon;

* Investigate the influence of the DO content of water used to create steam for activation of
carbon on the subsequent adsorption performance of the activated carbon for removal of MIB
and phenol in natural water (e.g., water containing NOM);

* Determine a method to either evaluate the quantity and type of individual acidic oxygen-
containing functional groups on the surface of highly basic, heterogeneous carbons or
evaluate the influence of the DO content of water used to create steam on the surface
chemistry of a more homogenous surface (e.g., silica gel).















Nitrogen 0.01559 7.80E-01
Oxygen 0.03109 2. 10E-0 1
Argon 0.03412 9.00E-03
Carbon Dioxide 0.8981"f 3.15E-04
*Weiss (1970) and "fTaiz and Zeiger (2006)


0.223 torrO2 / tOTT


APPENDIX A
SAMPLE CALCULATIONS FOR DISSOLVED GASES

Table A-1. Henry's constants (at 273 K) and partial pressure of atmospheric gases


Bunsen coefficient*


Air (14.7 psi)


Sample calculation of dissolved gas concentration in solution

Scenario: Air, 24.7 psi = 1277 torr

Henry's law expressed using a volumetric solubility coefficient (i.e, the Bunsen Method):
Po
Co o2 o, Where PO2 = partial pressure, BO2 = Bunsen coefficient, and K
02 (760 / 1000 / K)2
molecular weight/molecular volume

BO2 = 0.03109


0.21 O,
PO2 = (1277torr 17.52 torr) 2
torr

264 .5torrO2 /torr
o21 760


264.5 torr Ol/ torr



-*0.03109 = 15.5 mg/L


1000 mg / g (32 g/ mol / 22.4L / mol)

Sample calculation of change in pH with change in carbon dioxide in solution

Scenario: Atmospheric pressure


Pco
760 /1000 /K 0


Bunsen Method: C ,


BCO2 = 0.8981

PO2 = (760torr -


0.0003 CO,
17.52 torr) 2
torr


0.223 torr Ol/ torr


C O


* 0.8981 = 0.376 mg/L


760 / 1000 mg / g /(44 g / mol / 22 .4L / mol)


Cco12 = 0.376mg/L mol /44000mgCO2, = 9*"10-6 mol/LI


































0.377 torrO2 / torr


[H' ]*[HCO3 ]
k=
[ H2CO3 i


[H' ]*[HCO3
>[H[H+ ]*[HCO3 ] = 9*"10-6 *10-6.35 = 4.02*10-12
CO2


Charge Balance: [H ] = [HCO3-] + [CO32-] + [OH ], assume [CO32-] and [OH-] are negligible in
this scenario


4.02*10-12
[H ]= ~[H ]2
[H ]


4.02*10-1 [H ]= 44.2*10-12


2.0*"10-6mol/ L


pH = -log(2.0 10-6) = 5.70

Scenario: 1.68 times atmospheric pressure


Pco
760 /1000 /K 0


Bunsen Method: C ,


BCo2 = 0.8981 mg/L/torr


0.0003 CO,
PO2 = (1277torr 17.52 torr) 0.377 torr Ol/ torr
torr


C O


*0.8981 = 0.6381 mg/L


760 / 1000 mg / g /(44 g / mol / 22 .4L / mol)


CIco2 = 0.6381Img/L "mol/ 44000mgCO2~ = 15 10 mol /LI


[H' ]*[HCO3 ]
k=
[H2CO3 i


[H' ]*[HCO, ]
3 --> H ']*[HCO3 ]
"'1


15*10-6*10-6.35


6.48*10-12


Charge Balance: [H ] = [HCO3-] + 2[CO32-] + [OH ], assume [CO32-] and [OH-] are negligible in
this scenario

6.48*10-12 [' 64 0 [' Ji811- 0 o
[H ]=[H ]=6.8 0-2 [ ]= 64*01=3.*06 l/
[H ]


pH = -log(3.0*"10-6)= 5.59

Scenario: Atmospheric pressure plus calcium chemistry


P,,
760 / 1000 / K2


Bunsen Method: Cco 2


BCO2 = 0.8981













0.223 torrO, / torr


Charge Balance: [H ] + 2[Ca2+] [HCO3-] + 2[CO32-] + [OH ], assume [CO32-] and [OH-] are
negligible in this scenario and the calcium in solution (as measured) is 2.5 mg/L or 8*10-4 M.

4.02*10'
[H ]= -2*'(8*"1 0-4) ~[H ] = 2.5 1*'10-92
[H ]

pH = log(2.51*10 9) = 8.60

Moles of dissolved gases versus moles of steam and sample calculation

Table A-2. Moles of dissolved gases and water vapor versus treatment scenario
Level of 02 (mO1 02)
SC (g H20/ g carbon) mol H20 3 mg/L 7.5 mg/L 12 mg/L
0.4 0.111 1.88*10- 4.69*10- 7.50*10-
1.4 0.389 6.56*10-7 1.64*10-6 2.63*10-6
2.4 0.667 1.13*10-6 2.81*10-' 4.50*10-6


PO2 = (760torr 17.52 torr) .03 O 0.223 torr Ol/ torr
torr


C CO'


* 0.8981 = 0.376 mg/L


760 / 1000 mg / g /(44 g / mol / 22 .4L / mol)


Ci.o) = 0.376mg/L mol /44000mgCO2, = 9*"10-6 mol/LI


[H' ]*[HCO, ]
k =
[ H2CO, ]


[H' ]*[HCO, ]
>[H' ]* [HCO, ]
COo,


9*"10-6 *10-6.35 = 4.02*10-12


Scenario: 1.4 g H20/g carbon and 3 mg/L oxygen

1.4 g H20/ g carbon 5 g carbon = 7g H20
7 g H20 / (18 g/mol) (molecular weight of H20) = 0.3 89 moles of H20

7 g H20 / (1000 g/L) (density of water) = 0.007 L H20
3 mg/L Oxygen 0.007 L / (32000 mg/mol) = 6.56*10-7 moles of oxygen









APPENDIX B
PRELIMINARY STUDIES FOR DETERMINATION OF LEVELS FOR BOX BEHNKEN

To begin any experiment, one must first determine the levels (or settings) of the variables

to be evaluated. In this study, the variables of activation time, activation temperature, steam to

carbon ratio (SC), and the dissolved oxygen (DO) content in the water used to create steam were

evaluated to determine the impact of these parameters on the adsorption performance of the

resulting carbons for MIB and also the impact of these parameters on both the chemical and

physical properties of the resulting carbons. The following section discusses how the levels were

determined for the Box-Behnken approach. Note that no attempt is made to explain observed

phenomenon as that is outside the scope of this section.

For the latter variable, DO content, it was decided to use the same demarcations used in a

previous study on the impact of DO content on reactivation of carbon (Chesnutt et al., 2007). In

a study by Chestnutt et al. (2007), the DO contents were characterized as being low (less than 4

mg/L), medium (between 7-8 mg/L), and high (greater than 10 mg/L). In order to use the Box-

Behnken approach, there must be three levels for each variable and these levels must be evenly

spaced. Therefore, the DO content levels in this study were set at 3 mg/L, 7.5 mg/L, and 12

mg/L and where achieved by pressuring the water vessel with either nitrogen, air, or oxygen gas.

Given that activations are typically carried out at high temperatures (1173 K),

preliminary experiments were performed at 1023 K and also a lower temperature, as low

temperatures are known to manifest differences in surface chemistry, of 823 K. The activation

time, 30 min., and SC, 0.68 g H20/g carbon, for these studies were based on steam activations

used in a previous study of steam activation by Tennant and Mazyck (2003). The results of the

first experiment varying both temperature and DO and evaluating MIB removal in batch studies

are shown in Figure B-1.









Activations at 1023 K did not manifest a difference in MIB removal but the activations at

823K demonstrated that higher DO in the water used to create steam for activation does alter the

carbons performance for MIB removal. It was also noted that in the study of Chesnutt et al.

(2007) a very low temperature (648 K) was employed which manifested significant difference in

MIB removal with altered DO. Therefore, in order to encompass the range of results where DO

impacts MIB and conditions where it does not, the temperatures employed for the Box-Behnken

approach were 673 K, 823 K, and 973 K.

Next, to shorten activation time and thereby cut experiment time in half, an activation

time of 15 minutes was evaluated at the same steam flow as Figure B-1 (0.68 SC) and an

activation temperature of 823 K. The performance of these carbons for removal of MIB is

shown in Figure B-2.

From Figure B-2, it is apparent that there is a change in MIB removal depending on the

activation time employed. As 15 and 30 minutes of treatment create different removals of MIB,

and in order to create evenly spaces treatment times that represent extreme conditions, treatment

times of 10, 30, and 50 minutes were employed in the Box-Behnken study.

Finally, an analysis of flow was performed where the steam flow was raised to the

highest possible in the system (2.84 g H20/g carbon) and compared to the previously employed

flow based on Tennant and Mazyck (2003). The performance of these carbons for removal of

MIB is shown in Figure B-3.

There is a change in MIB removal depending on the SC employed. As 0.68 g H20/g

carbon and 2.84 g H20/g carbon create different removals of MIB, and in order to create evenly

spaces treatment times that represent extreme conditions, SC of 0.4 g H20/g carbon, 1.4 g H20/g

carbon, and 2.4 g H20/g carbon were employed in the Box-Behnken study.









These preliminary studies highlight the importance of studying interactions between

factors, as clearly DO is not an influencing factor under all conditions and this underlines why

the Box-Behnken statistical design of experiments was employed in this study.


100
90
80
70
60
50
40
30
20
10
0


I Fly ash; 10 mg/L
I Fly ash; 5 mg/L
m823K; 10 mg/L
HO023K; 5 mg/L


F lyash


Low DO Mid DO High DO


Figure B-1. MIB removal versus activation temperature and DO (SC = 0.68 g H20/g carbon,
activation time = 30 min.).


O'
Flyash 15 min.

Figure B-2. MIB Removal versus time and DO (temperature
carbon).


30 min.

823 K, SC = 0.68 g H20/g













MLow DO
M~igh D


90

80



40

2 30

20
10
30


0.68 g H2C
carbon


2.84 g H20/g
carbon


Figure B-3. MIB removal versus steam flow and DO (temperature = 823 K, time = 30 min.).




















Dissolved Carbon MIB Phenol
oxygen Temperature Time (g H20/g removal Mass loss removal
Std. (mg/L) (K) (min.) carbon) (%) (%) pzc (pH) (%)
1 3 673 30 1.4 56.5 8 9.9 56.8
2 12 673 30 1.4 42.4 9 10.0 49.6
3 3 973 30 1.4 94.7 21 11.8 65.9
4 12 973 30 1.4 66.9 17 12.1 67.4
5 7.5 823 10 0.4 37.1 11 10.5 52.6
6 7.5 823 50 0.4 65.7 10 11.0 59.4
7 7.5 823 10 2.4 42.4 10 11.0 54.8
8 7.5 823 50 2.4 62.9 12 11.4 62.7
9 3 823 30 0.4 72.6 11 11.0 63.2
10 12 823 30 0.4 49.2 11 10.4 61.6
11 3 823 30 2.4 56.8 11 11.5 62.4
12 12 823 30 2.4 69.3 11 11.7 63.0
13 7.5 673 10 1.4 30.5 10 9.9 45.9
14 7.5 973 10 1.4 75.4 15 12.0 68.1
15 7.5 673 50 1.4 58.7 10 10.1 54.4
16 7.5 973 50 1.4 95.9 23 11.9 68.8
17 3 823 10 1.4 37.3 10 11.2 53.6
18 12 823 10 1.4 50.5 10 10.9 55.6
19 3 823 50 1.4 76.5 12 11.6 56.4
20 12 823 50 1.4 66.5 11 11.0 55.6
21 7.5 673 30 0.4 38.5 8 9.7 50.7
22 7.5 973 30 0.4 71.1 14 12.0 64.8
23 7.5 673 30 2.4 63.0 8 10.5 59.2
24 7.5 973 30 2.4 95.3 22 12.1 64.8
25 7.5 823 30 1.4 55.3 9 11.2 63.2
26 7.5 823 30 1.4 57.0 11 10.7 59.0
27 7.5 823 30 1.4 56.2 11 11.2 59.0
28 7.5 823 30 1.4 45.2 11 11.8 63.6
29 7.5 823 30 1.4 52.4 11 11.6 62.3


APPENDIX C
BOX-BEHNKEN RESULTS

Table C-1. Box-Behnken carbons activation conditions, MIB removal, mass loss, pHpze, phenol
removal
Factor Response Response Response Response


1 2 3 4


Factor A Factor B C


Factor D
Steam/










Table C-2. Analysis of variance (ANOVA) results for mass loss (Prob > F of less than 0.05 are
significant factors)
Source Sum of squares DF Mean square F-value Prob > F
Model 422.83 8 52.85 32.80 < 0.0001 significant
A 1.32 1 1.32 0.82 0.3769
B 298.75 1 298.75 185.42 < 0.0001 significant
C 12.15 1 12.15 7.54 0.0125 significant
D 6.08 1 6.08 3.78 0.0662
B2 63.80 1 63.80 39.60 < 0.0001 significant
AB 5.84 1 5.84 3.63 0.0713
BC 15.35 1 15.35 9.53 0.0058 significant
BD 19.55 1 19.55 12.13 0.0023 significant
Residual 32.22 20 1.61
Lack of fit 29.66 16 1.85 2.90 0.1565 not significant
Pure error 2.56 4 0.64
Cor total 455.05 28

Table C-3. Statistical data for mass loss
Parameter Values
Std. dev. 1.26
Mean 12.0
C.V. 10.5
PRESS 99.5
R-squared 0.929
Adj R-squared 0.900
Pred R-squared 0.781
Adeq precision (ratio greater than 4 desired) 20.4

Table C-4. Data for coefficient of parameters for mass loss
Factor Coefficient DF Standard 95% CI 95% CI VIF Actual
estimate error low high factors
Intercept 10.82 10.31 10.17 11.46 45.79
A-DO -0.33 1 0.37 -1.10 0.43 1 0.91
B-Temperature 4.99 1 0.37 4.23 5.75 1 -0.14
C-Time 1.01 1 0.37 0.24 1.77 1 -0.31
D-Steam to carbon ratio 0.71 1 0.37 -0.05 1.48 1 -7.39
B2 3.01 1 0.48 2.01 4.01 1 0.00013
AB -1.21 1 0.63 -2.53 0.12 1 -0.0018
BC 1.96 1 0.63 0.63 3.28 1 0.00065
BD 2.21 1 0.63 0.89 3.53 1 0.015











Table C-5. Diagnostic case studies for mass loss


Standard
order
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29


Actual
value
8
9
21
17
11
10
10
12
11
11
11
11
10
15
10
23
10
10
12
11
8
14
8
22
9
11
11
11
11


Predicted
value
7.96
9.72
20.36
17.28
9.10
11.11
10.52
12.53
10.44
9.77
11.86
11.20
9.79
15.85
7.89
21.78
10.14
9.48
12.15
11.49
10.34
15.90
7.34
21.74
10.82
10.82
10.82
10.82
10.82


Student
residual
-0.19
-1.28
0.99
-0.10
1.63
-0.71
-0.25
-0.77
0.51
1.46
-0.68
-0.54
0.06
-0.92
2.62
1.63
0.15
0.47
-0.49
-0.14
-2.20
-2.39
0.99
0.79
-1.27
0.24
0.19
0.21
0.05


Cook's
distance
0.00
0.18
0.11
0.00
0.09
0.02
0.00
0.02
0.01
0.07
0.01
0.01
0.00
0.09
0.76
0.30
0.00
0.01
0.01
0.00
0.54
0.63
0.11
0.07
0.01
0.00
0.00
0.00
0.00


Outlier
t
-0.18
-1.30
0.99
-0.10
1.70
-0.70
-0.24
-0.76
0.50
1.51
-0.67
-0.53
0.06
-0.92
3.15
1.71
0.15
0.46
-0.48
-0.14
-2.46
-2.75
0.99
0.79
-1.29
0.23
0.18
0.21
0.05


Residual
-0.17
-1.15
0.89
-0.09
1.82
-0.79
-0.27
-0.86
0.56
1.63
-0.75
-0.60
0.06
-0.83
2.35
1.47
0.17
0.52
-0.54
-0.16
-1.97
-2.14
0.89
0.71
-1.57
0.29
0.23
0.26
0.06


Leverage
0.50
0.50
0.50
0.50
0.23
0.23
0.23
0.23
0.23
0.23
0.23
0.23
0.50
0.50
0.50
0.50
0.23
0.23
0.23
0.23
0.50
0.50
0.50
0.50
0.06
0.06
0.06
0.06
0.06










































Table C-8. Data for coefficient of parameters for MIB removal
Coefficient Standard 95% CI 95% CI Actual
Factor estimate DF error low high VIF factors
Intercept 54.16 1 2.127 49.72 58.59 126.7
A-DO -4.13 1 2.037 -8.38 0.12 1-4.76
B-Temperature 17.48 1 2.037 13.23 21.72 1-0.038
C-Time 12.75 1 2.037 8.50 17.00 1 1.12
D-Steam to carbon ratio 4.63 1 2.037 0.38 8.87 1 -10.33
A2 4.03 12.685 -1.57 9.63 1.019 0.2
B2 10.24 1 2.685 4.64 15.84 1.019 0.00046
AC -5.80 1 3.528 -13.16 1.56 1-0.064
AD 8.98 1 3.528 1.62 16.33 1 1.99


Table C-6. ANOVA results for MIB
Source Sum of squares DF
Model 7306 8
A 205 1
B 3665 1
C 1951 1
D 257 1
A2 112 1
B2 724 1
AC 135 1
AD 322 1
Residual 996 20
Lack of fit 903 16
Pure error 92 4
Cor total 8302 28


removal
Mean square
913
205
3665
1951
257
112
724
135
322
50
56
23


F-value
18.35
4.12
73.62
39.19
5.16
2.25
14.54
2.70
6.47

2.44


Prob> F
< 0.0001
0.0559
< 0.0001
< 0.0001
0.0344
0.1491
0.0011
0.1158
0.0193


significant

significant
significant
significant

significant

significant


0.2009 not significant


Table C-7. Statistical data for MIB removal
Parameter Values
Std. dev. 7.05
Mean 60.0
C.V. 11.7
PRESS 2380.4
R-squared 0.880
Adj R-squared 0.832
Pred R-squared 0.713
Adeq precision (ratio greater than 4 desired) 15.3




















































Table C-10. ANOVA results for pHpze
Sum of
Source Squares DF Mean square F-value Prob > F
Model 12.94 4 3.23 40.01 < 0.0001 significant
A 0.07 1 0.07 0.83 0.3699
B 11.60 1 11.60 143.53 < 0.0001 significant
C 0.19 1 0.19 2.32 0.1409
D 1.08 1 1.08 13.36 0.0013 significant
Residual 1.94 24 0.08
Lack of fit 1.22 20 0.06 0.34 0.9544 not significant
Pure error 0.72 4 0.18
Cor total 14.88 28


Table C-9. Diagnostic case studies for MIB removal


Standard
order
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29


Actual
value
56.5
42.4
94.7
66.9
37.1
65.7
42.4
62.9
72.6
49.2
56.8
69.3
30.5
75.4
58.7
95.9
37.3
50.5
76.5
66.5
38.5
71.1
63.0
95.3
55.3
57.0
56.2
45.2
52.4


Predicted
value
55.09
46.82
90.04
81.77
36.78
62.28
46.03
71.53
66.67
40.45
57.97
67.65
34.17
69.12
59.67
94.62
43.77
47.10
80.87
61.00
42.30
77.25
51.55
86.50
54.16
54.16
54.16
54.16
54.16


Student
residual
0.24
-0.76
0.80
-2.55
0.05
0.56
-0.60
-1.42
1.21
1.78
-0.24
0.34
-0.61
1.04
-0.16
0.21
-1.32
0.69
-0.89
1.12
-0.63
-1.02
1.89
1.46
0.17
0.42
0.30
-1.33
-0.26


Cook's
distance
0.00
0.03
0.03
0.33
0.00
0.01
0.01
0.08
0.17
0.38
0.01
0.01
0.01
0.04
0.00
0.00
0.21
0.06
0.09
0.15
0.02
0.04
0.14
0.09
0.00
0.00
0.00
0.02
0.00


Outlier
t
0.24
-0.75
0.79
-3.02
0.05
0.55
-0.59
-1.46
1.22
1.89
-0.23
0.33
-0.60
1.04
-0.16
0.21
-1.34
0.68
-0.89
1.13
-0.62
-1.02
2.04
1.50
0.17
0.41
0.30
-1.36
-0.25


Residual
1.41
-4.42
4.66
-14.87
0.32
3.42
-3.63
-8.63
5.93
8.75
-1.17
1.65
-3.67
6.28
-0.97
1.28
-6.47
3.40
-4.37
5.50
-3.80
-6.15
11.45
8.80
1.14
2.84
2.04
-8.96
-1.76


Leverage
0.31
0.31
0.31
0.31
0.26
0.26
0.26
0.26
0.52
0.52
0.52
0.52
0.27
0.27
0.27
0.27
0.52
0.52
0.52
0.52
0.27
0.27
0.27
0.27
0.09
0.09
0.09
0.09
0.09
























Table C-12. Data for coefficient of parameters for pHpze
Coefficient Standard 95% CI 95% CI Actual
Factor estimate DF error low high VIF factors
Intercept 11.09 1 0.053 10.984 11.202 7.01
A-DO -0.08 1 0.082 -0.244 0.094 1 -0.017
B-Temperature 0.98 1 0.082 0.814 1.153 1 0.0066
C-Time 0.13 1 0.082 -0.044 0.294 1 0.0063
D-Steam to carbon ratio 0.30 1 0.082 0.131 0.469 1 0.30


Table C-11. Statistical data for pHpze
Parameter Values
Std. dev. 0.284
Mean 11.0
C.V. 2.56
PRESS 2.57
R-squared 0.869
Adj R-squared 0.847
Pred R-squared 0.826
Adeq precision (ratio greater than 4 desired) 21.7



















































Table C-14. ANOVA results for phenol removal
Source Sum of squares DF Mean square F-value Prob > F
Model 656.55 4 164.14 13.60 < 0.0001 significant
A 2.52 1 2.52 0.21 0.6517
B 576.85 1576.85 47.80 < 0.0001 significant
C 59.41 1 59.41 4.92 0.0362 significant
D 17.76 117.76 1.47 0.2368
Residual 289.60 24 12.07
Lack of fit 269.20 20 13.46 2.64 0.1791 not significant
Pure error 20.41 4 5.10
Cor total 946.15 28


Table C-13. Diagnostic case studies for pHaze


Standard
order
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29


Actual
value
8
9
21
17
11
10
10
12
11
11
11
11
10
15
10
23
10
10
12
11
8
14
8
22
9
11
11
11
11


Predicted
value
7.96
9.72
20.36
17.28
9.10
11.11
10.52
12.53
10.44
9.77
11.86
11.20
9.79
15.85
7.89
21.78
10.14
9.48
12.15
11.49
10.34
15.90
7.34
21.74
10.82
10.82
10.82
10.82
10.82


Student
vesidual
-0.19
-1.28
0.99
-0.10
1.63
-0.71
-0.25
-0.77
0.51
1.46
-0.68
-0.54
0.06
-0.92
2.62
1.63
0.15
0.47
-0.49
-0.14
-2.20
-2.39
0.99
0.79
-1.27
0.24
0.19
0.21
0.05


Cook's
distance
0.00
0.18
0.11
0.00
0.09
0.02
0.00
0.02
0.01
0.07
0.01
0.01
0.00
0.09
0.76
0.30
0.00
0.01
0.01
0.00
0.54
0.63
0.11
0.07
0.01
0.00
0.00
0.00
0.00


Outlier
t
-0.18
-1.30
0.99
-0.10
1.70
-0.70
-0.24
-0.76
0.50
1.51
-0.67
-0.53
0.06
-0.92
3.15
1.71
0.15
0.46
-0.48
-0.14
-2.46
-2.75
0.99
0.79
-1.29
0.23
0.18
0.21
0.05


Residual
-0.17
-1.15
0.89
-0.09
1.82
-0.79
-0.27
-0.86
0.56
1.63
-0.75
-0.60
0.06
-0.83
2.35
1.47
0.17
0.52
-0.54
-0.16
-1.97
-2.14
0.89
0.71
-1.57
0.29
0.23
0.26
0.06


Leverage
0.50
0.50
0.50
0.50
0.23
0.23
0.23
0.23
0.23
0.23
0.23
0.23
0.50
0.50
0.50
0.50
0.23
0.23
0.23
0.23
0.50
0.50
0.50
0.50
0.06
0.06
0.06
0.06
0.06
























Table C-16. Data for coefficient of parameters for phenol removal
Coefficient Standard 95% CI 95% CI Actual
Factor estimate DF error low high VIF factors
Intercept 59.46 1 0.65 58.13 60.79 29.76
A-DO -0.46 1 1.00 -2.53 1.61 1 -0.10
B-Temperature 6.93 1 1.00 4.86 9.00 1 0.046
C-Time 2.23 1 1.00 0.16 4.29 1 0.11
D-Steam to carbon ratio 1.22 1 1.00 -0.85 3.29 1 1.22


Table C-15. Statistical data for phenol removal
Parameter Values
Std. dev. 3.47
Mean 59.4
C.V. 5.84
PRESS 434.2
R-squared 0.693
Adj R-squared 0.642
Pred R-squared 0.541
Adeq precision (ratio greater than 4 desired) 12.6











Table C-17. Diagnostic case studies for phenol removal


Standard
order
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29


Actual
value
56.8
49.6
65.9
67.4
52.6
59.4
54.8
62.7
63.2
61.6
62.4
63.0
45.9
68.1
54.4
68.8
53.6
55.6
56.4
55.6
50.7
64.8
59.2
64.8
63.2
59.0
59.0
63.6
62.3


Predicted
value
52.99
52.07
66.85
65.94
56.02
60.47
58.45
62.90
58.70
57.79
61.14
60.22
50.30
64.17
54.75
68.62
57.70
56.78
62.15
61.23
51.31
65.18
53.75
67.61
59.46
59.46
59.46
59.46
59.46


Student
residual
1.23
-0.80
-0.31
0.47
-1.10
-0.34
-1.18
-0.07
1.45
1.23
0.41
0.90
-1.42
1.27
-0.11
0.06
-1.32
-0.38
-1.85
-1.81
-0.20
-0.12
1.76
-0.91
1.10
-0.14
-0.14
1.21
0.83


Cook's
distance
0.08
0.03
0.00
0.01
0.06
0.01
0.07
0.00
0.11
0.08
0.01
0.04
0.10
0.08
0.00
0.00
0.09
0.01
0.17
0.17
0.00
0.00
0.16
0.04
0.01
0.00
0.00
0.01
0.00


Outlier
t
1.24
-0.79
-0.30
0.46
-1.11
-0.34
-1.19
-0.06
1.48
1.24
0.40
0.89
-1.45
1.28
-0.11
0.06
-1.34
-0.37
-1.96
-1.91
-0.19
-0.12
1.84
-0.90
1.10
-0.13
-0.13
1.22
0.83


Residual
3.81
-2.47
-0.95
1.46
-3.42
-1.07
-3.65
-0.20
4.50
3.81
1.26
2.78
-4.40
3.93
-0.35
0.18
-4.10
-1.18
-5.75
-5.63
-0.61
-0.38
5.45
-2.81
3.74
-0.46
-0.46
4.14
2.84


Leverage
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.20
0.03
0.03
0.03
0.03
0.03










LIST OF REFERENCES

Albers, P.W., Pietsch J., Krauter, J., Parker, S.F., 2003. Investigations of activated carbon
catalyst supports from different natural sources. Phys. Chem. Chem. Phys. 5, 1941-1949.

Alcaniz-Monge, J., Cazorla-Amoros, D., Linares-solano, A., 1994. Effect of the activating gas
on tensile strength and pore structure of pitch-based carbon fibres. Carbon 32 (7), 1277-1283.

Ania, C.O., Parra, J.B., Pis, J.J., 2004. Oxygen-induced decrease in the equilibrium adsorptive
capacities of activated carbons. Adsorption Science and Technology 22 (4), 337-351.

Appleton, Wisconsin, 2006. Wastewater treatment facility: Treatment plant tour. Accessed: July
9. .

Arriagada, R., Garcia, R., MolinaSabio, M., RodriguezReinoso, F., 1997. Effect of steam
activation on the porosity and chemical nature of activated carbons from Eucalyptus globulus
and peach stones. Microporous Materials 8 (3-4), 123-130.

Bach, M.T., 2004. The role of calcium in and methodologies for overcoming pH excursions for
reactivated granular activated carbon. Thesis (M.E.). University of Florida.

Benjamin, M.M., 2002. Water Chemistry. McGraw-Hill; New York.

Bansal, R.C., Aggarwall, D., Goyal, M., Kaistha, B.C., 2002. Influence of carbon-oxygen surface
groups on the adsorption of phenol by activated carbons. Ind. Journal of Chem. Tech. 9, 290.

Bansal, R.C., Donnet, J.B., Stoeckli, F., 1988. Active Carbon. Marcel Dekker, Inc.; New York.

Bansal, R.C., Vastola, F.J., Walker, P.L., 1974. Influence of hydrogen chemisorption on the
subsequent chemisorption of oxygen on activated graphon. Carbon 12, 355.

Barnett, K.E., 1999. Wastewater treatment method and plant. United States Patent Number
5,961,830.

Barton, S.S., Evans, M.J.B, Halliop, E., MacDonald, J.A.F., 1997. Acidic and basic sites on the
surface of porous carbon. Carbon 35 (9), 1361-1366.

Bizot, P.M., Bailey, B.R., 1997. Sodium sulfoxylate formaldehyde as a boiler additive for
oxygen scavenging. United States Patent Number 5,660,736.

Boehm, H., 1966. Functional groups on the surfaces of solids. Angewandte Chemie Inernational
Edition 5, 533-522.

Box, G.E.P., Behnken, D.W., 1960. Some new three level designs for the study of quantitative
variables. Technometrics 2 (4), 455-475.










Buczek. B., 1993. Development of properties within particles of active carbons obtained by a
steam activation process. Langmuir 9, 2509-2512.

Bull, R.J., Robinson, M., Meier, J.R., Stober, J., 1982. Use of biological assay systems to assess
the relative carcinogenic hazards of disinfection by-products. Environmental Health
Perspectives 46, 215-227.

Campbell, A.G., Folk, R.L., Tripepi, R.R., 1997. Wood ash as an amendment in municipal
sludge and yard waste composting processes. Compost Science & Utilization 5 (1), 62-73.

Cannon, F.S., Snoeyink, V.L., Lee, R.G., Dagois, G., DeWolfe, J.R., 1993. Effect of calcium in
field-spent GACs on pore development during regeneration. JAWWA 85 (3), 76-89.

Cannon, F.S., Snoeyink, V.L., Lee, R.G., Dagois, G., 1994. Reaction mechanism of calcium-
catalyzed thermal regeneration of spent granular activated carbon. Carbon 32 (7), 1285-1301.

Carr, S., Farmer, R.W., 1995. React-pHTAI improves pH in carbon effluent. The National
Environmental Joumnal 5 (2), 54.

Centeno, T.A., Stoeckli, F., 1995. The oxidation of an asturian bituminous coal in air and its
influence on subsequent activation by steam. Carbon 33 (5), 581-586.

Chen ,W.F., Cannon, F.S., Rangel-Mendez, J.R., 2005a. Ammonia-tailoring of GAC to enhance
perchlorate removal. I: Characterization of NH3 thermally tailored GACs. Carbon 43 (3), 573-
580.

Chen, W.F., Cannon, F.S., Rangel-Mendez, J.R., 2005b. Ammonia-tailoring of GAC to enhance
perchlorate removal. II: Perchlorate adsorption. Carbon 43 (3), 581-590.

Chestnutt, T.E., Bach, M.T., Mazyck, D.W., 2007. Improvement of thermal reactivation of
activated carbon for the removal of 2-methylisoborneol. Water Res. 41, 79-86.

Clark, R.M., Lykins, B.W., 1989. Granular Activated: Design, Operation and Cost. Lewis
Publishers; Chelsea, MI.

Considine, R., Denoyel, R., Pendleton, P., Schumann, R., Wong, S.H., 2001. The influence of
surface chemistry on activated carbon adsorption of 2-methylisobomeol from aqueous solution.
Colloids and Surfaces A: Physicochemical and Engineering Aspects 179 (2-3), 271-280.

Crittenden, J.C., Berrigan, J.K., Hand, D.W., 1986. Design of rapid small-scale adsorption tests
for a constant diffusivity. Journal of Water Pollution Control Federation 58 (4), 3 12-3 19.

Crittenden, J.C., Reddy, P.S., Arora, H., Trynoski, J., Hand, D.W., Perram, D.L., Summers, RS.,
1991. Predicting GAC performance with rapid small-scale column tests. JAWWA 83 (1), 77-87.










Das, K.C., Melear, N.D., Kastner, J.R., Buquoi, J.Q., 2003. Influence of ash amendment on odor
emissions and aerobic biodegradation of biosolids mixes. Transactions of the ASAE 46 (4),
1185-1191.

Dussert, B.W., Farmer, R.W., Hayden, R.A., 1995. Oxidized activated carbon for the control of
pH and alkalinity in water treatment applications. United States Patent 5,466,378.

El-Sayed, Y., Bandosz, T.J., 2003. Effect of increased basicity of activated carbon surface on
valeric acid adsorption from aqueous solution activated carbon. Phys. Chem. Chem. Phys 5 (21),
4892-4898.

Farmer, R.W., Dussert, B.W., Kovacic, S.L., 1996. Improved granular activated carbon for the
stabilization of wastewater pH. American Chemical Society, Division of Fuel Chemistry 41 (1),
456-460.

Farmer, R.W., Kovacic, L., Matviya, M., Wadhwa, N.P., 1998. Activated carbon treated by
carbon dioxide for the stabilization of treated water pH. United States Patent Number 5,714,433.

Femnandes, M.B., Skjemstad, J.O., Johnson, B.B., Wells, J.D., Brooks, P., 2003.
Characterization of carbonaceous combustion residues. I. Morphological, elemental, and
spectroscopic features. Chemosphere 51, 785-795.

Frederick, H.T., Cannon, F.S., Dempsey, B.A., 2001a. Calcium and TOC loading: Effect of
hydroxyl and carboxyl substituents. Colloids and Surfaces A: Physicochemical and Engineering
Aspects 191 (1-2), 161-177.

Frederick, H.T., Cannon, F.S., Dempsey, B.A., 2001b. Calcium loading onto granular activated
carbon with salicylate or phthalate. Colloids and Surfaces A: Physicochemical and Engineering
Aspects 177 (2-3), 157-168.

Frederick, H.T., Cannon, F.S., 2001. Calcium-NOM loading onto GAC. JAWWA 93(12), 77-89.

Garcia-Garcia, A., Greg6rio, A., Franco, C., Pinto, F., Boavida, D., Gulyurtlu, I., 2003.
Unconverted chars obtained during biomass gasification on a pilot-scale gasifier as a source of
activated carbon production. Bioresource Technology 88, 27-32.

Garten, V.A., Weiss, V.E., 1957. The ion and electron exchange properties of activated carbon in
relation to its behavior as a catalyst and adsorbent. Rev Pure Appl Chem 7, 69-122.

Gerber, N.N., 1968. A volatile metabolite of actinomycetes, 2-methylisobomeol. Journal of
Antibiotics 22 (10), 508.

Gergova, K., Eser. S., 1996. Effects of activation method on the pore structure of activated
carbons from apricot stones. Carbon 34 (7), 879-888.










Gergova, K., Petrov, N., Eser, S., 1994. Adsorption properties and microstructure of activated
carbons produced from agricultural by-products by steam pyrolysis. Carbon 32 (4), 693-702

Gergova, K., Petrov, N., Butuzova, L., Minkova, V., Isaeva, L., 1993. Evolution of the active
surface of carbons produced from various raw-materials by steam pyrolysis activation. Journal of
chemical technology and biotechnology 58 (4), 321-330.

Gerth, W.A., 1985. Applicability of Henry's law to hydrogen, helium, and nitrogen solubilities in
water and olive oil at 37 OC and pressures up to 3 00 atmospheres. Archives of Biochemistry and
Biophysics 241 (1), 187-199.

Gillogly, T., Sneoyink, V., Elarde, J., Wilson, C., Royale, E., 1998. 14C-MIB adsorption on PAC
in natural water. JAWWA 90(1), 98-108.

Gonzalez, M.T., Molina-Sabio, M., Rodriguez-Reinoso, F., 1994. Steam activation of olive
stone chars, development of porosity. Carbon 32 (8), 1407-1413.

Hammer, M.J., Hammer, M.J., Jr., 2001. Water and Wastewater Technology. Prentice Hall;
Upper Saddle River, NJ.

Hassler, J.W., 1974. Purifieation with activated carbon: industrial, commercial, environmental.
Chemical Pub. Co.; New York.

Herzing, D., Snoeyink, V., Wood, N., 1977. Activated carbon adsorption of the odorous
compounds 2-methylisoborneol and geosmin. JAWWA 69, 223.

Heschel, W., Klose, E., 1995. On the suitability of agricultural by-products for the manufacture
of granular activated carbon. Fuel 74 (12), 1786-1791.

Jung, M.W., Ahn, K.H., Lee, Y., Kim, K.P., Rhee, J.S., Park, J.T., Paeng, K.J., 2001. Adsorption
characteristics of phenol and chlorophenols on granular activated carbons (GAC).
Microchemical Journal 70 (2), 123-131.

Karanfil, T., Kitis, M., 1999. Role of granular activated carbon surface chemistry on the
adsorption of organic compounds. 2. Natural Organic Matter. Environ. Science and
Technology 33 (18), 3225-3233.

Kastner, J.R., Das, K.C., Buquoi, Q., Melear, N,D., 2003. Low temperature catalytic oxidation of
hydrogen sulfide and methanethiol using wood and coal fly ash. Environ. Sci. Technol. 37 (11),
2568-2574.

Kawamura, S., 2000. Integrated design and operation of water treatment facilities. John Wiley &
Sons; New York.










Khezami. L., Chetouani, A., Taouk, B., Capart, R., 2005. Production and characterisation of
activated carbon from wood components in powder: Cellulose, lignin, xylan. Powder
Technology 157, 48-56.

Kipling, J.J., Gassser, C.G., 1960. Adsorption from liquid mixtures at solid surfaces. J.
Phys.Chem 64 (6), 710.

Kirk-Othmer, 1977. Encyclopaedia of Chemical Technology, Third Edition, Wiley; New York.

Knappe, D.R.U., Snoeyink, V.L., Dagois, G., DeWolfe, J.R., 1992. Effect of calcium on thermal
regeneration of GAC. JAWWA 84 (8), 73-80.

Kruyt, H.R., De Kadt, G.S., 1929. Die ladung der kohle. Kolloid Chem. Beihefte 47, 44.

Laine, N.R., Vastola, F.J., Walker, P.L., 1963. Importance of active surface area in carbon-
oxygen reaction. Journal of Physical Chemistry 67 (10), 2030.

Leon y Leon, C.A., Solar, J.M., Calemma, V., Radovic, L.R., 1992. Evidence for the protonation
of basal plane sites on carbon. Carbon 30 (5), 797-81 1.

Lillo-Rodenas, M.A., Cazorla-Amoros, D., Linares-Solano, A., 2005. Behaviour of activated
carbons with different pore size distributions and surface oxygen groups for benzene and toluene
adsorption at low concentrations. Carbon 43 (8), 1758-1767.

Lussier, M.G., Zhang, Z., Miller, D.J., 1998. Characterizing rate inhibition in steam/hydrogen
gasification via analysis of adsorbed hydrogen. Carbon 36 (9), 1361-1369.

MacKenzie, J.A., Tennant, M.F., Mazyck, D.W., 2005. Tailored GAC for the effective control of
2-methylisoborneol. JAWWA 97 (6), 76-87.

Martin-Gullon, I., Asensio, M., Font, R., Marcilla ,A., 1996. Steam-activated carbons from a
bituminous coal in a continuous multistage fluidized bed pilot plant. Carbon 34 (12), 1515-1520.

Mazyck, D.W., Cannon, F.S., 2000. Overcoming calcium catalysis during the thermal
reactivation of granular activated carbon: Part I. Steam-curing plus ramped-temperature N2
treatment. Carbon 38 (13), 1785-1799.

Mazyck, D.W., Cannon. F.S., 2002. Overcoming calcium catalysis during the thermal
reactivation of granular activated carbon. Part II. Variation of steam-curing reactivation
parameters. Carbon 40, 241-252.

Mazyck, D.W., Cannon, F.S., Bach, M.T., Radovic, L.R., 2005. The role of calcium in high pH
excursions with reactivated GAC. Carbon 43, 511-518.










Menendez, J.A., Phillips, J., Xia, B., Radovic, L.R., 1996. On the modification and
characterization of chemical surface properties of activated carbon: In the search of carbons with
stable basic properties. Langmuir 12 (18), 4404-4410.

MolinaSabio, M., Gonzalez, M.T., RodriguezReinoso, F., Sepulveda-Escribano, A., 1996. Effect
of steam and carbon dioxide activation in the micropore size distribution of AC. Carbon 34 (4),
505-509.

Montgomery, D.C., 1991. Design and analysis of experiments. Wiley; New York.

Myers, M.H., 1995. Response surface methodology. Wiley; New York.

Newcombe, G., Morrison, J., Hepplewhite, C., Knappe, D.R.U., 2002. Simultaneous adsorption
of MB and NOM onto activated carbon. II. Competitive effects. Carbon 40, 2147.

Newcombe, G., 1999. Charge vs. porosity-some influences on the adsorption of natural organic
matter (NOM) by activated carbon. Water Science & Technology 40 (9): 191-198.

Nowack, K.O., Cannon, F.S., Arora, H., 1999. Ferric chloride plus GAC for removing TOC. J
AWWA 91 (2), 65-78.

Nowack, K.O., Cannon, F.S., Mazyck, D.W., 2004. Enhancing activated carbon adsorption of 2-
methylisoborneol: Methane and steam treatments. Environ. Sci. Technol. 38, 276-284.

Nowack, K.O., Cannon, F.S., 1997. Control of calcium buildup in GAC: Effect of iron
coagulation. Carbon 35 (9), 1223-1237.

Otake, Y., Jenkins, R.G., 1993. Characterization of oxygen-containing surface complexes created
on a microporous carbon by air and nitric-acid treatment. Carbon 31 (1), 109-121.

Pelekani, C., Snoeyink, V.L., 1999. Competitive adsorption in natural water: Role of activated
carbon pore size. Water Research 33 (5), 1209-1219.

Pendleton, P., Wong, S.H., Schumann, R., Levay, G., Denoyel, R., Rouquerol, J., 1997.
Properties of activated carbon controlling 2-methylisoborneol adsorption. Carbon 35, 1 141.

Petrucci, R.H., Harwood, W.S., 1997. General chemistry: principles and modern applications.
Prentice Hall; New Jersey.

Pereira, M.F.R., Soares, S.F., Orfhao, J.J.M., Figueiredo, J.L., 2003. Adsorption of dyes on
activated carbons: influence of surface chemical groups. Carbon 41 (4), 811-821.

Puri, B.R, 1970. Surface Complexes on Carbon in Chemistry and Physics of Carbon, Vol. 6, ed.
P.L. Walker, Jr., Marcel Dekker; New York.










Quirilivan, P.A., Li, L., Knappe, D.R.U., 2005. Effects of activated carbon characteristics on the
simultaneous adsorption of aqueous organic micropollutants and natural organic matter.
Water Research 39 (8), 1663-1673.

Rangel-Mendez, J.R., Cannon, F.S., 2005. Improved activated carbon by thermal treatment in
methane and steam: Physicochemical influences on MIB sorption capacity. Carbon 43, 467.

Rashash, D.M.C., Dietrich, A.M., Hoehn, R.C., 1997. FPA of selected odorous compounds.
JAWWA 89 (4), 131-141.

Ravichandran, M., Aiken, G.R., Reddy, M.M., Ryan, J.N., 1998. Enhanced dissolution of
cinnabar mercuricc sulfide) by dissolved organic matter isolated from the Florida Everglades.
Eny. Sci.& Tech. 32, 3305.

Rio, S., Faur-Brasquet, C., Le Coq, L., Le Cloirec, P., 2005. Production and characterization of
adsorbent materials from an industrial waste. Adsorption 11, 793-798.

Rodriguez-Reinoso, F., Molina-Sabio, M., Gonzalez, M.T., 1995. The use of steam and CO2 aS
activating agents in the preparation of activated carbons. Carbon 33 (1), 15-23.

Rosenfeld, P.E., Henry, C.L., 2000. Wood ash control of odor from biosolids application.
Journal of Environmental Quality 29 (5), 1662-1668.

Ryu, S.K., Jin, H., Gondy, D., Pusset, N., Ehrburger, P., 1993. Activation of carbon-fibers by
steam and carbon-dioxide. Carbon 31 (5), 841-842.

Salame, I.I., Bandosz, T.J., 2001. Surface chemistry of activated carbons: Combining the results
of temperature-programmed desorption, Boehm, and potentiometric titrations. J. Colloid and
Interface Sciences 240, 252-258.

Salame, I.I., Bandosz, T.J., 2003. Role of surface chemistry in adsorption of phenol on activated
carbons. Journal of Colloid and Interface Science 264 (2), 307-312.

Sarkar, M., Acharya, P.K., 2006. Use of fly ash for the removal of phenol and its analogues from
contaminated water. Waste Management 26,559-570.

Smith, W.R., Polley, M.H., 1956. The oxidation of graphitized carbon black. Journal of Physical
Chemistry 60 (5), 689-691.

Snoeyink, V., Summers, R.S., 1999 "Chapter 13. Adsorption of organic compounds." Water
Quality and Treatment, AWWA; Denver, CO.

Souza-Macedo, J., Costa-Junior, N.B., Almeida, L.E., Silva-Vieira, E.F., Cestari, A.R., Fatima-
Gimenez, I., Carrefio, N.L.V., Barreto, L.S., 2006. Kinetic and calorimetric study of the
adsorption of dyes on mesoporous activated carbon prepared from coconut coir dust. Journal of
Colloid and Interface Science 298, 515-522.










Speitel, G.E., Dvorak, B.I,. Morley, M.C., 2001. Rapid small-scale column test. AEESP
Environmental Engineering Processes Laboratory Manual. Association of Environmental
Engineering and Science Professors; Champaign, Illinois.

Streat, M., Patrick, J.W., Perez, M.J.C., 1995. Sorption of phenol and para-chlorophenol from
water using conventional and novel activated carbons. Wat. Res. 29 (2), 467-472.

Summers, R.S., Haist, B., Koehler, J., Ritz, J., Zimmer, G., Sontheimer, H., 1989. The influence
of background organic matter on GAC adsorption. JAWWA 81 (5), 66-74.

Suzuki, M., 1990. Adsorption engineering. Elsevier; New York.

Szymaniski, G.S., Karpiniski, Z., Biniak, S., Swistkowski, A., 2002. The effect of the gradual
thermal decomposition of surface oxygen species on the chemical and catalytic properties of
oxidized activated carbon. Carbon 40 (14), 2627-2639.

Taiz, L., Zeiger, E., 2006. Plant Physiology, Fourth Edition. Sinauer Associates, Inc.;
Sunderland, MA.

Tennant, M.F., Mazyck, D.W., 2003. Steam-pyrolysis activation of wood char for superior
odorant removal. Carbon 41, 2195.

Tennant, M.F., 2004. Activation and use of powdered activated carbon for removing 2-
methylisoborneol in water utilities. Dissertation (Ph.D.). University of Florida.

Tomk6w, K., Siemieniewska, T., Czechowski, F., Jankowska, A., 1977. Formation of porous
structures in activated brown-coal chars using 02, CO2, and H20 as activating agents. Fuel 56
(2), 121-124.

Walker, P.L., Rusinko, F., Austin, L.G., 1959. Gas reactions of carbon. Advan. Catal. 11, 133-
221.

Walker, P.L., 1996. Production of activated carbons: Use of CO2 VeTSus H20 as activating agent.
Carbon 34 (10), 1297-1299.

Weiss, R.F., 1970. The solubility of nitrogen, oxygen, and argon in water and seawater. Deep
Sea Research 17, 721-735.

Wiegand, W.B., 1937. pH properties of colloidal carbon. Ind. Eng. Chem. 29, 953.

Wigmans, T., 1989. Industrial-aspects of production and use of activated carbons. Carbon 27
(1), 13-22.

Zhang, Y.Z., Lu, Z., Maroto-Valer, M.M., Andresen, J.M., Schobert, H.H., 2003. Comparison of
high-unburned-carbon fly ashes from different combustor types and their steam activated
products. Energy & Fuels 17 (2), 369-377.










Zou, Y., Han, B.X., 2001. High-surface area activated carbon from Chinese coal. Energy &
Fuels 15, 1383-1386.









BIOGRAPHICAL SKETCH

Morgana Bach was born in St. Petersburg, Florida, where she was raised by her mother.

She graduated from Gibbs High School in June 1998. From there, she traveled to Gainesville,

Florida, to attend the University of Florida where I received she received her bachelor of science

degree with highest honors and her Master of Engineering in Environmental Engineering

Sciences.

She stayed on at the University of Florida to work on her doctor of philosophy degree

with Dr. David W. Mazyck in the field of environmental engineering sciences with a focus on

water treatment. She looks forward to working towards the betterment of mankind through

research into the technologies which provide clean water and air.





PAGE 1

1 IMPACT OF SURFACE CHEMISTRY ON ADSORPTION: TAILORING OF ACTIVATED CARBON By MORGANA T. BACH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGRE E OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007

PAGE 2

2 Copyright 2007 Morgana Bach

PAGE 3

3 To my mother and sister

PAGE 4

4 ACKNOWLEDGMENTS I thank my mom, Teresa Bach, and my sister, Rhiannon Bach, for their love and support over the many years I have p ursued my academic career. I would also like to thank Link Charlott for his support and respect of my accomplishments in the academic world. I thank David Mazyck for his support and advice for the last five years. His tendency to consider myself lucky to have had him as an advisor and I am surely a better person for it.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............... 4 LIST OF TABLES ................................ ................................ ................................ ........................... 8 LIST OF FIGURES ................................ ................................ ................................ ......................... 9 ABSTRACT ................................ ................................ ................................ ................................ ... 10 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .................. 12 2 LITERATURE REVIEW ................................ ................................ ................................ ....... 13 Activation of Activated Carbon ................................ ................................ .............................. 13 Gasification ................................ ................................ ................................ ...................... 13 Creation of Functional Groups ................................ ................................ ........................ 15 Impact of Activation Pa rameters ................................ ................................ ..................... 17 Type of Precursor ................................ ................................ ................................ ..... 17 Type of Oxidant ................................ ................................ ................................ ....... 20 Time, Temper ature, and Oxidant Flow ................................ ................................ .... 23 Adsorption ................................ ................................ ................................ .............................. 25 Physical Adsorption ................................ ................................ ................................ ......... 25 Effect of Surface Chemistry on Adsorption ................................ ................................ .... 27 Current Uses of Activated Carbon ................................ ................................ .......................... 29 Reactivation of Activated Carbon ................................ ................................ .......................... 30 pH Excursions in Water Treatment ................................ ................................ ........................ 32 pH Excursion Mechanisms for Activated Carbons ................................ ......................... 33 Methods to Overcome pH Excursions ................................ ................................ ............. 36 Influence of Dissolved Oxygen in Water used to Create Steam ................................ ............ 40 3 HYPOTHESIS AND O BJECTIVES ................................ ................................ ...................... 44 pH Excursions ................................ ................................ ................................ ......................... 44 Hypothesis ................................ ................................ ................................ ....................... 44 Objectives ................................ ................................ ................................ ........................ 44 Tailoring of Activated Carbon Using Dissolved Oxygen (DO) ................................ ............. 45 Hypothesis ................................ ................................ ................................ ....................... 45 Objectives ................................ ................................ ................................ ........................ 45 4 EXPERIMENTAL METHODS ................................ ................................ ............................. 46 Reactivated Carbon Samples for Overcoming pH Excursions ................................ ............... 46 Carbon Precursor for Activations ................................ ................................ ........................... 46

PAGE 6

6 Air Treatments and CO 2 Treatments of Reactivated Carbons ................................ ................ 47 Activations with Varying DO Content in Water ................................ ................................ .... 47 Water Contact pH ................................ ................................ ................................ ................... 48 Calcium Solution Concentration ................................ ................................ ............................. 49 Adsorption Experiments (Batch) ................................ ................................ ............................ 49 Rapid Small Scale Column Tests (RSSCTs) ................................ ................................ .......... 50 BET S urface Area Measurements ................................ ................................ .......................... 51 Pore Size Distribution Measurement ................................ ................................ ...................... 52 pH at the Point of Zero Charge ................................ ................................ ............................... 52 Boehm Titration ................................ ................................ ................................ ...................... 53 Statistical Design of Experiments: Box Behnken ................................ ................................ ... 53 5 S TRATEGIES FOR O VERCO MING p H EXCURSIONS FOR REACTIVATED GRANULAR ACTIVATED CARBON: AIR AND CARBON DIOXIDE TREATMENTS ................................ ................................ ................................ ...................... 58 Creation of a pH Stable Carbon ................................ ................................ .............................. 58 Surface Acidity of Air Treated Carbons ................................ ................................ ................. 59 Surface Acidity of CO 2 Treated Carbons ................................ ................................ ............... 60 Verification of Mechanisms ................................ ................................ ................................ ... 61 Summary and Conclusions ................................ ................................ ................................ ..... 63 6 METHODOLOGIES FOR OVERCOMING PH EXCURSIONS FOR REACTIVATED GRANULAR ACTIVATED CARBON: EFFECT ON ADSORPTION P ERFORMANCE ................................ ................................ ................................ ................... 67 Impact of Wetting Time on MIB Adsorption for CO 2 Treated Carbon ................................ 70 MIB Adsorption from Natural Water ................................ ................................ ..................... 71 Phenol Adsorption from Deionized Water ................................ ................................ ............. 71 Conclusions ................................ ................................ ................................ ............................. 73 7 IMPACT OF DISSO LVED OXYGEN ON ACTIVATION OF POWDERED ACTIVATED CARBON ................................ ................................ ................................ ........ 77 Dissolved Gases in Solution ................................ ................................ ................................ ... 78 Box Behnken Approach and Mass Loss ................................ ................................ ................. 81 MIB Removal ................................ ................................ ................................ ......................... 84 Physical Properties ................................ ................................ ................................ .......... 86 Surface Chemistry ................................ ................................ ................................ ........... 88 Chemical versus Physical Methods of Removing DO ................................ ........................... 93 Phenol Removal ................................ ................................ ................................ ...................... 95 S ummary and Conclusions ................................ ................................ ................................ ..... 96 8 CONTRIBUTIONS TO SCIENCE ................................ ................................ ...................... 105 9 FUTURE WORK ................................ ................................ ................................ .................. 106

PAGE 7

7 APPENDIX A SAMPLE CALCULATIONS FOR DISSOLVED GASES ................................ ................. 107 B PRELIMINARY STUDIES FOR DETERMINATION OF LEVELS FOR BOX BEHNKEN ................................ ................................ ................................ ........................... 110 C BOX BEHNKEN RESULTS ................................ ................................ ............................... 114 LIST OF REFERENCES ................................ ................................ ................................ ............. 123 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ....... 132

PAGE 8

8 LIST OF TA BLES Table page 4 1. RSSCT and full scale design parameters ................................ ................................ .............. 56 4 2 2 2 full factorial ................................ ................................ ................................ ....................... 57 4 3 Box Behnken 3 3 fractional factorial ................................ ................................ ...................... 57 5 1 Calcium in solution with treatments ................................ ................................ ...................... 64 6 1 Characteristics of sample carbons ................................ ................................ ......................... 5 8 6 2 DOC remaining in solution after contact with 20 ppm carbon ................................ .............. 74 7 1 Concentration of dissolved gases in water at 273 K ................................ .............................. 98 7 2 Levels of four factors employed in the Box Behnken design ................................ ............... 98 7 3 Box Behnken st atistical results for mass loss ................................ ................................ ........ 98 7 4 Surface area versus treatment time, treatment temperature, and SC ................................ ..... 99 7 5 Box Behnken statistical results for MIB removal ................................ ................................ 99 7 6 BET surface area and total pore volume of four carbons ................................ ...................... 99 7 7 Box Behnken statist ical results for pH pzc ................................ ................................ ............ 100 7 8 Activation conditions of four carbons and their removal of MIB in dichloromethane ....... 100 7 9 BET surface area and total pore volume carbons treated with sodium sulfite .................... 100 7 10 Box Behnken statistical results for phenol removal ................................ .......................... 100

PAGE 9

9 LIST OF FIGURES Fig ure page 2 1. Proposed mechanism for the formation of COOH on an activated carbon surface .............. 42 2 2. Full scale water t reatment plant effluent pH and calcium concentration data after reactivated GAC was returned to service. ................................ ................................ .......... 42 2 3. Comparison of performance in the removal of MIB for reactivated carbons create d with varying DO content in the water used to create steam ................................ ....................... 43 4 1 Box Behnken design of 3 3 ................................ ................................ ................................ ..... 57 5 1. Effect of air exposure ti me and temperature on water contact pH ................................ ........ 64 5 2 Effect of treatment time and gas flow on resulting water contact pH of CO 2 treated carbon. ................................ ................................ ................................ ................................ 65 5 3 Effect of air treatments after 60 minute contact time on individual acidic functional groups. ................................ ................................ ................................ ................................ 65 5 4 Effect of CO 2 treatments on individual functional groups. ................................ ................... 66 6 1 Percent removal of MIB in DI water versus dose of carbon (24 hr contact time) ................ 74 6 2 Percent removal of MIB in raw water versus dose of carbon (24 hr contact time). ............. 75 6 3 Percent MIB removal versus wetting time in DI water with a CO 2 and untreated carbons ................................ ................................ ................................ ............................... 75 6 4 Effluent phenol concentration in DI water versus BV treated with a CO2 treated, air treated, and an untreated carbon. ................................ ................................ ....................... 76 7 1 Mass loss versus time and temperature ................................ ................................ ............... 101 7 2 Mass loss versus SC and temperature ................................ ................................ .................. 101 7 3 MIB removal versus time and temperature ................................ ................................ ......... 102 7 4 MIB removal versus SC and DO ................................ ................................ ......................... 102 7 5 Pore size distributions of sample carbons ................................ ................................ ............ 103 7 6 pHpzc vers us SC and temperature ................................ ................................ ....................... 103 7 7 PSDs of carbons treated with sodium sulfite ................................ ................................ ....... 104 7 8 Phenol removal versus time and temperature ................................ ................................ ...... 104

PAGE 10

10 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IMPACT OF SURFACE CHEMISTRY ON A DSORPTION: TAILORING OF ACTIVATED CARBON By Morgana T. Bach August 2007 Chair: David W. Mazyck Major: Environmental Engineering Sciences The present work summarizes research into the impact of the surface chemistry of activated carbon on the water trea tment phenomena of pH excursions, the tailoring of activated carbon through manipulation of the dissolved oxygen (DO) content of the water used to create steam for activation, and, most importantly, the impact of these changes in surface chemistry on adsor ption. The increase in total surface acidity and some individual acidic functional groups associated with air treatments of reactivated carbons were reaffirmed as the mechanism by which pH excursions are overcome with air treatments. The proposal that carbon dioxide (CO 2 ) treatments overcome pH excursions via physisorption, and subsequent desorption, of CO 2 forming carbonic acid upon water immersion was supported through an analysis of mass gained during CO 2 treatment. Both air and CO 2 treatments were not found to suppress the calcium chemistry mechanism responsible for pH excursions. In a comparison of the adsorption performance of the treated carbons for 2 methylisoborneol (MIB), a common taste and odor causing compound, the untreated carbon removed less MIB than the air treated carbon in water containing natural organic matter (NOM) due to decreased adsorption of the competitive NOM with the air treated carbon related to its increased surface acidity. CO 2 treatment was proven

PAGE 11

11 effective as shown by its increased removal of MIB, as it deterred water adsorption, which is deleterious to MIB adsorption. In a comparison of adsorption performance for phenol, both the untreated and CO 2 treated carbons performed similarly while the air treated carbon did n ot perform as well due to its increased surface acidity. A novel approach to tailoring activated carbon via varying levels of DO in the water used to create steam for the activation of a wood based precursor was evaluated using the Box Behnken design of ex periments. The statistical analysis confirmed the impact of changes in DO on the performance of the resulting carbon for adsorption of MIB and supported variation in DO as a possible means of tailoring activated carbon.

PAGE 12

12 CHAPTER 1 INTRODUCTION Adsorb ents have been commonly used to solve environmental issues and researchers have delved into the mechanisms responsible for the success of a particular adsorbent. In the case of activated carbon, although the surface chemistry of activated carbon is as imp ortant as its physical properties in water treatment, only the optimization of the latter is well researched. In how to create these physical properties. On the other hand, while there are many methods to alter the surface chemistry (i.e., functional groups) of activated carbon, there are few guidelines available for its optimization, as it is not known what type of functional groups are best for a particular con taminant. Furthermore, it is not clear how to control the quantity, type, and location of these functional groups. To elucidate the role of surface chemistry in adsorption, the present work summarizes advanced research into the impact of surface chemistr y on the water treatment phenomena of pH excursions, the tailoring of the surface chemistry of activated carbon through manipulation of the dissolved oxygen content of the water used to create steam for activation, and, most importantly, the impact of thes e changes in surface chemistry on adsorption. By investigating not only how changes in surface chemistry are instilled into activated carbon but also the specific groups formed and any effect on adsorption, this research may be a step towards bridging the gap between potential optimization of the physical properties and surface chemistry of adsorbents.

PAGE 13

13 CHAPTER 2 LITERATURE REVIEW Activation of Activated Carbon Activated carbon is created when a carbonaceous precursor, generally wood, peat, lignite coal, bituminous coal, or other material possessing relatively high carbon content, undergoes a thermal process involving pyrolysis and oxidation to increase the surface area and develop a range of pore sizes within the precursor. Pyrolysis is carried out in a n inert environment, or oxygen starved, to ensure that the raw material does not gasify but rather is transformed into char. The oxidation process, or simply activation, is performed in the presence of an oxidizing gas, such as air, steam, carbon dioxide (CO 2 ), or a combination of these, at temperatures between 973 1173 K in order to oxidize the surface of the carbon pores and further develop the internal pore structure via gasification In physical activation, the pyrolysis and oxidation steps are separa te, while in chemical activation the two processes are combined (Snoeyink and Summers 1999). Gasification Gasification by oxygen proceeds according to the reactions shown in Equations 2 1 to 2 6 (Walker et al. 1959). A free active carbon site is desig nated by C f and a surface complex is designated by C(n), where n is the type of complex. The preference of the oxidation reaction for free active sites, associated with high energy edge sites, has been shown by Smith and Polley (1956) and Laine et al. (19 63). 2C f + O 2 (g) 2C(O) (2 1) C(O) CO (g) (2 2) C f + O 2 (g) C(O 2 ) (2 3) C(O 2 ) CO 2 (g) (2 4)

PAGE 14

14 CO(g) + C(O) CO 2 (g) + C f (2 5) C f + CO 2 (g) 2CO(g) (2 6) In Equations 2 1 to 2 4 gasification occurs either through evolution of carbon monoxide (CO) or carbon dioxide (CO 2 ) from the carbon surface. These reaction products may then react with the carbon surface ( Equations 2 5, 2 6) The rate of gasification is inhibited by reactions with CO ( Equation 2 5) as this causes the concentration of surface oxides available to participate in Equation 2 2 to decrease (Walker et al. 1959). The CO 2 produced may gasify the carbon surface ( Equation 2 6) Indeed, CO 2 has been used as an activating agent though it does require a relatively pure quantity of this gas, which may be difficult to obt ain in some areas, while air and steam are ubiquitous. Unlike air activation, steam activation is an endothermic process that is easier to control and better suited for carbons with high surface activity (Hassler 1974; Kirk Othmer 1977). Typical reactio n mechanisms for steam activation are found in Equations 2 7 to 2 9 (Walker et al. 1959; Puri 1970; Lussier et al. 1998). C f + H 2 O C(O) + H 2 (2 7) 2C f + H 2 2C(H) (2 8) CO(g) + H 2 O CO 2 (g) + H 2 (2 9) T he initial reaction of the steam with the carbon surface where the oxygen is exchanged from the water molecule to the carbon surface to create a surface oxide (Equation 2 7) which may devolve as CO, produces hydrogen gas (Walker et al. 1959; Puri 1970 ; Lussier et al., 1998). As discussed with air oxidation, the CO produced may react to decrease the rate of gasification by reacting with a surface oxide to produce CO 2 The hydrogen produced (Equation 2 7) may also react with the carbon surface to yield a surface hydrogen complex (Equation 2 8 ),

PAGE 15

15 which inhibits the gasification reaction as this site will not react further as may a C(O) sit e (Walker et al. 1959; Puri 1970; Menndez et al., 1996; Lussier et al., 1998; Bansal et al., 2002). The reverse of Equation 2 7 may also lead to inhibition by hydrogen removing a C(O). T he water gas shift reaction ( Equation 2 9) is where the water vap or is broken down to CO 2 and hydrogen gas, the products of which may either activate the surface ( Equation 2 6) or inhibit gasification ( Equation 2 8) Creation of Functional Groups Functional groups (designated here as C(n)) form on the surface of activ ated carbon due to the presence of heteroatoms, any atom present in the carbon structure that is not carbon or hydrogen, such as oxygen, nitrogen and sulfur. Functional groups are formed from these heteroatoms due to the activation and reactivation scheme used and exposure to the atmosphere and as such these groups are created at a variety of temperatures though there are temperatures for which groups experience optimal formation. Therefore, activation affect s both the physical structure of the carbon, vi a gasification, and also the chemical properties of the carbon through formation of these functional groups. The thermal stability of oxygen containing functional groups is such that the functional groups that devolve as CO (e.g., phenolic and quinone gr oups) are thermally unstable at 973 K and functional groups that devolve as CO 2 (e.g., carboxylic and lactone groups) are thermally unstable at 773 K (Puri, 1970; Otake and Jenkins, 1993 ). This relative thermal stability is often used as a measurement of the concentration of these functional groups on the carbon surface under a process called temperature programmed desorption (TPD). During TPD, where the temperature is slowly raised to 1223 K in an inert environment the quantities of CO and CO 2 devolved from the surface are measured and this quantity, and the knowledge of the chemistry behind the creation of the devolved gas ( e g ., carboxylic groups devolve as CO 2 ), yield s a

PAGE 16

16 qualitative indicator of the functional groups on the carbon surface (Leon y Leon et al., 1992; Salame and Bandosz, 2001; Bansal et al., 2002). Fourier transform infrared spectroscopy (FTIR), which measures the absorption spectra of the carbon surface to determine individual also available, such as the Boehm titration method where specific bases or acids are contacted with the acidic or basic f unctional groups on the carbon surface (El Sayed and Bandosz, 2003; Ania et al., 2004). In a study by Salame and Bandosz (2001) comparing the Boehm titration technique, TPD, and FTIR, it was determined that Boehm titration allows detection of groups of c ertain strength while omitting other acids and bases such as ketones, ester s and ethers. TPD, found to detect all functional groups present on the carbon surface, was limited to the broad category of CO producing and CO 2 producing functional groups and t hus this method lacks information on individual functional groups. Finally though the FTIR method provide s information on individual functional groups, they found it produce d complicated information where mathematical treatments, requiring several approx imations, were necessary to clarify the data. B y measuring the oxygen containing functional groups that desorbed from carbon using TPD it was determined that the optimum temperature for chemisorption of oxygen, using air as the oxidizing agent, occurs aro und 673 K (Wiegand, 1937; Puri, 1970). Interestingly t he maximum capacity for a carbon to adsorb bases, or its acidic nature, has also been shown to lie close to 673 K (Puri 1970). The fact that the maximum chemisorption of oxygen and the maximum capac ity to adsorb bases occur at the same temperature has led many researchers to

PAGE 17

17 the conclusion that the acidic nature of carbon is closely related to its oxygen content ( Puri 1970 ; Otake and Jenkins 1993 ; Barton 1997 ; Ania et al. 2004). The carboxyl gr oup (COOH) has been suggested as the oxygen containing functional group responsible for the acidic nature of carbon by several researchers (Kru y t and de Kadt 1931 ; Puri 1970). The temperature at which these carboxyl groups were developed was 673 K in ei ther oxygen or air. More re cently, Otake and Jenkins (1993) researched air treated carbons and found a positive linear relationship between the chemisorbed oxygen concentration desorbed as a CO 2 complex which is indicative of the presence of carboxyl gro ups on the carbon surface, and total acidity of the carbon. Furthermore, these authors proposed that the oxygen functional groups existing on the air oxidized char were carboxyl groups and were formed following the mechanism of Figure 2 1. The importance of these oxygen containing functional groups on adsorption and their influence on the surface chemistry of activated carbon will be discussed in later sections. Impact of Activation Parameters The impact of the activation process on the physical and chemic al properties of the carbon has been discussed in general terms but not the e ffect of the individual activation parameters on these properties. Activation parameters include, but are not limited to, precursor choice, type and flow of oxidant, time of act ivation, and temperature of activation. The following is a discussion of how and why variation of these parameters, as they are the easiest to control, can affect the resulting Type of Precursor The two main grou ps of raw material for activated carbon are coal and biomass. Types of coals employed include anthracite, lignite, and bituminous coals (Streat et al., 1995; Zou and Han, 2001). By far the most common precursor for activated carbon used in research is co al;

PAGE 18

18 subsequently, many of the theories on physical and chemical adsorption onto activated carbon are based on coal based activated carbon and may not hold with biomass based carbons due to the differing properties of these carbons, as discussed further in this section. Examples of biomass used as raw material for activated carbon include sewage sludge, coconut shells, coconut coir dust, fruits stones (e.g., peaches, plums, and cherries), straw, and nut shells (Gergova et al., 1994; Heschel and Klose, 1995; Streat et al., 1995; Albers et al., 2003; Rio et al., 2005; Souza Macedo et al., 2006). Various types of wood, including pine, oak, and beech have also been investigated (Heschel and Klose, 1995; Streat et al., 1995; Albers et al., 2003; Garca Garca e t al., 2003). The drive to investigate different types of biomass as a precursor for activated carbon does not seem to lie so much in an effort to improve carbon for the removal of a specific compound or water matrix as much as it is an economic or envi ronmental consideration. For example, a study pertaining to the use of pine as a source of activated carbon was prompted by the use of pine as an energy source in the study area, due to the lack of other available energy sources such as coal (Garca Garc a et al., 2003). The large supply of char remaining after burning of the pine prompted the investigation into use of this raw material for activated carbon, which could then be sold for an economic benefit. The use of a waste product to create an economi cally and environmentally useful product is what has prompted the study of sewage sludge, coconut shells (due to local abundance in tropical countries), fly ash from power plants, straw, and used tires, as these wastes would either be incinerated or placed in a land fill in most circumstances (Streat et al., 1995; Rio et al., 2005; Sarkar et al., 2006; Souza Macedo et al., 2006).

PAGE 19

19 A successful carbon in these studies is one that performs as well as traditional carbons in removal of a target compound. These studies employ a number of raw materials and evaluate them based on either adsorption performance or a common method of indicating performance, such as pore size distribution, and they may also analy ze the effect of varying activation parameters on the pr operties of the resulting carbon ( Streat et al., 1995; Garca Garca et al., 2003; Rio et al., 2005 ). For example, Heshel and Klose (1995) employed several agricultural by products (e.g., fruit stones, nut shells, and wood) to determine the type of raw ma terial that produced the highest quality activated carbon. High quality was defined as a carbon that was abrasion resistant, having low macroporosity (i.e., few large pores), and having high carbon yield, where the latter relates to the quantity of carbon lost during activation. The authors concluded that the highest quality carbon from agricultural by products was produced from coconut shells, followed by peach stones plum stones hazelnut shells walnut shells and cherry stones. The authors indicate that a large pore size in the raw material leads to large pores in the carbon. The wood based carbons had higher porosity than either the nut shells or the fruit stones and this led to decreased abrasion resistance. Heshel and Klose (1995) found that t he yield of carbon correlated with the O/C atomic ratio in the carbon. The authors concluded that the morphology of the raw materials determines several important properties of carbon due to the cross linking cellulose and lignin molecules. A study by Albe rs et al. (2003) using wood and coconut shells as the raw material demonstrated from scanning electron microscope (SEM) pictures that the inherent structure of the raw material is present even after pyrolysis. Gergova et al. (1994) and Khezami et al. (200 5) observed the same phenomenon of a woody structure that is preserved and becomes visible after the high temperatures employed in pyrolysis. Even at the atomic level, differences in the graphite

PAGE 20

20 structure of activated carbons are apparent based on the ra w material employed (Albers et al., 2003). Gergova et al. (1994) also note that lignin and cellulose in the raw materials of agricultural waste may affect porosity development which may relate to the finding of Hershel and Klose (1995) that the O/C atomi c ratio plays a role in char yield. Heschel and Klose (1995) and Khezami et al. (2005) note that lignin as a source of activated carbon produces the highest char yield, which may relate to porosity and the fact that lignin has a lower oxygen content than cellulose. These studies highlight the importance of the choice of raw material when creating activated carbon. Type of Oxidant Examples of oxidants that may be used in the thermal activation process include air (C O 2 ), steam (C H 2 O), carbon dioxide (C C O 2 ), or a combination of these. While use of either oxidant to oxidize the surface of carbon, and thereby improve its surface area, will produce an acceptable activated carbon, there are advantages and disadvantages to each oxidant. Walker et al. (1959) indicates that the activation energy for C O 2 is less than C H 2 O which is in turn less than C CO 2 This is due to the activation energy being determined by either the adsorption (e.g., creation of C(O)) or desorption (e.g., evolution of CO (g) ) step. The C O 2 adsorption step is much more exothermic than the C H 2 O adsorption step and due to this exothermic reaction the lifetime of the functional groups on the C O 2 reacted carbon may be short. Given that the desorption of the surface oxide created by the C O 2 reaction occurs relatively quickly, the activation energy for C O 2 is determined by the adsorption step. The C H 2 O and C CO 2 adsorption step, having a lower exothermic nature, may have a longer lifetime for its adsorbed group so that desorption is the rate limiting step. However, while C O 2 may require lower activation energy, energy is released during the reaction of air with the carbon such

PAGE 21

21 that control of the process may be difficult and generally is better suited for low activity carbons (Walker e t al. 1959; Kirk Othmer 1977). The literature is rife with comparisons of C CO 2 versus C H 2 O activations. Tomkw et al. (1977) investigated the porous structures of carbons resulting from C O 2 C H 2 O and C CO 2 and concluded that C H 2 O created a high er distribution of pores of different sizes due to the progressive widening of pores once they are formed. C CO 2 activations yielded mainly micropores (i.e., pore openings smaller than 2 nm) suggesting a higher occurrence of pore creation rather than pore widening. These findings were supported by Ryu et al. (1993). However, Wigmans (1989) found the opposite trend to Tomkw et al. (1977) where C CO 2 produced widening of micropores to mesopores (i.e., pore openings between 2 50 nm). Alcaniz Monge et al. (1994) and Rodrguez Reinoso et al. (1995) worked to clarify this apparent Interestingly, at low burnoffs (e.g., low levels of gasification) the results of Wigmans (1989) were confirmed by Alcaniz Monge et al. (1994), where the C H 2 O showed a higher production of micropores, though at higher burnoffs the C CO 2 had a higher concentration of micropores confirming the results of Tomkw et al. (1977). Overall however, the authors surmised that difference in diffusion of the CO 2 and steam molecules. Walker et al. (1954) found that the order of reaction rates were such that the rate of C H 2 O progressed at approximately three times the rate of C CO 2 and the rate of C O 2 progressed at several orders of magnitude faster than both C H 2 O and C CO 2 Later studies further confirmed that C CO 2 gasification is slower than that of st eam gasification likely due to the faster diffusion rate of the H 2 O molecule (Ryu et al., 1993; Alcaniz Monge et al., 1994; Walker

PAGE 22

22 et al., 1996). Rodrguez Reinoso et al. (1995) used the relative reaction rates of C H 2 O and C CO 2 to clarify the impact of burnoff on the development of pores with steam and carbon dioxide. Carbon dioxide reacts first to create and widen narrow micropores so that at low burnoff the micropores will dominate. Given further burnoff (i.e., further opportunity for the CO 2 to reac t with the carbon), the narrow micropores will become mesopores and macropores (i.e., pores greater than 50 nm) which yields a more diverse pore structure. Steam, on the other hand, reacts to continually create and then expand micropores so that micropore s do not remain intact very long due to the higher reaction rate of C H 2 O. Another possible explanation proposed by Rodrguez Reinoso et al. (1995) was that C CO 2 created a higher number of oxygen containing functional groups, located on the openings of the pores, and that these groups, being more thermally stable than those produced by steam activation, would close the pores to further diffusion of carbon dioxide which would have allowed for pore widening. Later work by these authors ( Molina Sabio et al 1996) supported this hypothesis Walker et al. (1996) also suggests that hydrogen is a greater inhibitor in the steam reaction (Equation 2 7) which limits the creation of pores and would explain the decrease in micropore formation. Regardless of the e xplanation the conclusion is that steam activation yielded a wider pore size distribution than activation with carbon dioxide (Tomkw et al., 1977; Alcaniz Monge et al., 1994; Rodrguez Reinoso et al., 1995; Molina Sabio et al., 1996; Walker et al., 1996) While micropores are preferred for the adsorption of some small molecules, it is generally considered advantageous to have a wide pore size distribution as the larger pore sizes act as roadways by which the target molecule can travel to its final adso rption site (Tennant and Mazyck, 2003) To this end, steam activation is often employed.

PAGE 23

23 In terms of activation with oxygen, Tomkw et al. (1977) determined that these carbons have a high development of pores but only at low burnoff and this was related to the blocking of the pores due to creation of functional groups at the edge of the pore openings. All carbons activated with oxygen in this study were found to have a higher concentration of polar centers than either the C H 2 O or C CO 2 carbons, which is related to the influence of oxygen containing functional groups on polarity. In addition, at higher temperatures, where these oxygen functional groups would be expected to devolve (i.e., be removed from the carbon surface), the pore structure resultin g from C O 2 is improved. Time, Temperature, and Oxidant Flow Gergova et al. (1993) evaluated the effect of activation temperature and activation time on the functional groups created during steam activation of several different types of agricultural prec ursors. The trends indicate an increase in the quantity of oxygen containing functional groups with treatment time, from 1 to 3 hrs, at lower temperatures (873 K) and a decrease in these groups with an increase in treatment time at higher temperatures (97 3 K), though this trend was not true of all precursors. This relates to the degradation of functional groups at higher temperatures and the limited creation of new functional groups to replace the ones lost, a finding supported by Arriagada et al. (1997). carbon remaining after activation, with an increase in activation time and temperature. Given the nature of the gasification reaction, it is anticipated that wit h higher temperature, which would speed gasification reactions, and longer treatment time, which would allow more gasification to occur, that the solid yield would decrease and this is supported in many studies (Gergova et al., 1993; Gonzalez et al., 1994; Rodrguez Reinoso et al., 1995; Gergova et al., 1996; Martin Gullon et al., 1996; Tennant and Mazyck, 2003). As the degree of gasification increases with an

PAGE 24

24 increase in temperature and treatment time, there is also generally an increase in the volume of pores and also the surface area within the carbon due to the removal of carbon atoms from the carbon surface to create new pores and higher surface area (Gergova et al., 1994; Centeno and Stoeckli, 1995; Rodrguez Reinoso et al., 1995; Gergova et al., 1996 ; Mazyck and Cannon, 2002; Zhang et al., 2003). However, overactivation can occur at high temperatures and or treatment time as external mass loss exceeds internal mass loss, where pores are developed, and/or the walls of pores begin to collapse (Gergova et al., 1994; Tennant and Mazyck, 2003; Mackenzie et al., 2005). Gonzalez et al. (1994) found that the rate of steam flow did not significantly affect the resulting micropore and mesopore formation in activated carbon and that low temperatures favored the widening of narrow micropores. Mazyck and Cannon (2002) also observed that steam flow had no impact on pore size distribution. Gonzalez et al. (1994) indicate that, though an increase in gasification rate has been observed with an increase in the flow of carbon dioxide due to the presence of a concentration gradient which compels the oxidant to enter the pores at a faster rate (i.e., bulk diffusion), the increase in steam flow yields more water vapor on the and/or the creation of macropores. Martin Gullon et al. (1996) further this theory by suggesting that the inhibiting factors produced by gasification (CO and H 2 ) (Equation 2 5 and 2 7, respectively) would dilute the steam on the internal portions of the c arbon thus leaving the highest concentrations of pure steam on the external surface of the particle which would lead to external gasification rather than internal. The dilution theory was supported by Arriagada et al. (1997). Rodrguez Reinoso et al. (199 5), however, found that at lower temperatures the steam flow had no effect while at higher temperatures the rate of steam flow yielded an increase in

PAGE 25

25 gasification. An increase in the volume of mesopores with an increase in steam flow was also observed by Martin Gullon et al. (1996). Rodrguez Reinoso et al. (1995) clarified this contradiction by noting that at higher temperatures the inhibiting effect of the hydrogen formed during gasification (Equation 2 7) is less effective. As steam flow is increased one would anticipate that more oxidant will react with the carbon; however, at lower temperatures, the increase in steam flow yields a corresponding increase in the amount of hydrogen produced by the reaction of Equation 2 7 which would inhibit reactions so that similar quantities of gasification would occur regardless of the increase in oxidant. At higher temperatures, where the gas water shift is less efficient, an increase in oxidant would indeed yield an increase in oxidation without as significant an affect of hydrogen inhibition. Adsorption Physical Adsorption Adsorption of a substance onto activated carbon in water occurs when the substance is concentrated more on the surface of the activated carbon than it is in the bulk solution. The substance that accumulates is called the adsorbate and the activated carbon is termed the adsorbent. Physisorption, or physical adsorption, occurs when the accumulation results from dispersion forces while chemisorption, or chemical adsorption, results from an exc hange or sharing of electrons between the adsorbate and the surface of the activated carbon. The properties of activated carbon that make it ideal as an adsorbent are a significant surface area to which the adsorbate may accumulate and an extensive intern al pore structure in which the adsorbate may become trapped, thus encouraging its accumulation. The rate of removal, or adsorption kinetics, is governed by the rates of the four steps composing physical adsorption. The first step, bulk transport, where the adsorbate moves from the bulk solution to the boundary layer of the adsorbate, is encouraged by the turbulent flow

PAGE 26

26 surrounding the activated carbon Film resistance to transport follows bulk transport and involves the transport of the adsorbate throug h the stationary layer of water surrounding the adsorbent. Film resistance may also be minimized by increased flow rate past the adsorbent. via surface or pore diffusion to existing adsorption sites. I nternal pore transfer is often the rate limiting step as the final step, physical adsorption, occurs almost instantaneously, so it has little effect on the rate of the overall reaction (Snoeyink and Summers 1 999). The rate of adsorption, therefore, is influenced by the same variables affecting diffusion rate, including the concentration gradient and the temperature of the system (Bansal et al. 1988). The relative hydrophobicity of the adsorbate can be a dri ving force for its accumulation on the adsorbent. Thus, the adsorption of a substance generally decreases with increased contaminant solubility in water (Snoeyink and Summers 1999). The pore size of the activated carbon also plays a major role in adsorp tion as, for example, a large natural organic matter (NOM) molecule will be unable to fit into small pores (i.e, micropores ) An activated carbon with a large volume of mesopores and macropores would be better suited for NOM removal (Newcombe, 1999) Con versely, a larger microporous volume is generally preferred when the target compound is a small molecule. However, a n increase in mesopores can be associated with an increase in adsorption of small molecules (Tennant and Mazyck, 2003, Nowack et al., 2004; Rangel Mendez and Cannon, 2005 ) though the mesopores may not be the final adsorption sites for the compound (for example 2 methylisoborneol (MIB)). The mesopores simply serve to aid in diffusion of the target compound through the pores to their final ads orption sites, likely micropores. It is common knowledge that mesopores aid in diffusion particularly in fine powders and that the adsorption of small molecules is most effective in micropores due to the

PAGE 27

27 greater number of contact points between the molecu le and the adsorbent (Suzuki, 1990; Pelekani and Snoeyink, 1999; Quinlivin et al., 2005; Rangel Mendez and Cannon, 2005 ). Effect of Surface Chemistry on Adsorption While general trends related to strictly physical adsorption have been discussed, the impa cts of surface chemistry on adsorption must also be discussed, as there are cases where it is the chemical properties of the activated carbon that dominate adsorption rather than its physical properties. The addition of oxygen to the surface of an activa ted carbon has been shown to alter its wettability, catalytic and electronic properties, as well its adsorption capabilities (Puri 1970 ; Bansal et al. 2002 ; 2002). Acidic oxygen containing functional groups are mainly present on the outer surface or edge of the basal plane due to limited diffusion into the micropores (Puri 1970 ; Menendez et al. 1996 ; Ania et al. 2004). As these outer sites con stitute the majority of the adsorption surface, the concentration of oxygen on the surface has a great impact on the adsorption capabilities of the carbon (Puri 1970). The presence of oxygen imparts a polar nature to the activated carbon, which results in a preference for removal of the more polar component of a solution, e.g., H 2 O (Kipling and Gasser 1960). El Sayed et al. (2003) found that the adsorption of valeric acid onto activated carbon decreased as the density of acid groups on the surface incr eased. These authors suggested that the polar nature of the activated carbon surface, associated with its high surface oxygen group content, resulted in the preferential adsorption of water molecules. In addition, the presence of carboxylic groups acted as physical obstacles to the interaction of the valeric acid and the activated carbon surface. Bansal et al. (2002) also attributed the decrease in phenol adsorption with increased oxidation of activated carbon to the increasing hydrophilic character of

PAGE 28

28 t he oxidized carbons. Ania et al. (2004) found a decrease in the adsorption of salicylic acid as the oxygen containing functional groups increased at low adsorbate concentrations. At high levels of salicylic acid, however, the study found that the adsorpt ion capacity of the carbon was similar to an activated carbon that had not been oxidized and these authors proposed that the higher concentration of valeric acid allowed the water molecules adsorbed onto the hydrophilic carbon sites to be displaced. Pend leton et al. (1997) researched the effects of surface chemistry on the adsorption of MIB and linked decreased adsorption with an increase in the hydrophilic nature of the oxidized carbon. They also found a strong correlation (R 2 =0.99) between oxygen conte nt and the number of hydrophilic sites. A subsequent study by the same research team (Considine et al. 2001) further supported the link between oxygen content and decreased adsorption of MIB. In this study, the adsorption of MIB in a dichloromethane sol vent, a non polar solvent, was investigat ed as opposed to previous experiments in water. The results showed that adsorption of MIB was independent of oxygen content in this solvent further supporting the contention that adsorption of water as a result of surface oxidation impedes MIB adsorption. Karanfil and Kitis (1999) found that surface oxidation, and the resulting density of strong acid functionalities decreased the adsorption of dissolved organic matter (DOC) (the dissolved portion of NOM ). Water c lusters forming around the polar (hydrophilic) functional groups on the surface reduced the ability of the DOC to reach the smaller pore sizes needed for its adsorption. The chemisorption of oxygen and the resulting functional groups have also been found to destroy electrons due to electron localization (Leon y Leon et al., 1992; Ania et al., 2004). The adsorption of phenol has been attributed to the electron donor acceptor complexes

PAGE 29

29 formed between the delocalized electrons on the basal planes on the s urface of the carbon and the aromatic ring of the phenol (Bansal et al., 2002; Ania et al., 2004). A decrease in phenol adsorption on oxidized samples has been shown by a number of researchers (Bansal et al., 2002; Ania et al., 2004). As the bonding site s are destroyed through chemisorption of oxygen, the adsorption of phenol decreases (Ania et al., 2004). Pereira et al. (2003) performed a study on an acidic anionic dye where the presumed adsorption mechanism involved the de electrons of the carbon surface and the free electrons of the dye molecule, and found that the adsorption of the anionic dye was hindered by the presence of acidic oxygen containing functional groups on the surface of the activated carbon. It should be noted that, in addition to bonds, the formation of acidic oxygen containing functional groups creat e a weakening in the donor acceptor (acid/base) adsorption mechanism (Ania et al. 2004) Current Uses of Activated Carbon Following activation, the carbon may be distributed as either p owdered activated carbon (PAC) or g ranular activated carbon (GAC) PAC is activated carbon ground in various types of mills so that 65 90% passes through a num consists of any larger activated carbon particles and is the portion of activated carbon used in the packed bed contact basins of water treatment plants GAC can be used as a substitute for typical granular filter media in that it removes suspended matter, though through adsorption GAC is also capable of removing organic compounds such as synthetic organic chemicals (SOCs) NOM and taste and odor causing compounds (Gerber, 1968; Frederick et al. 2001a ; Bansal et al., 2002) SOCs, such as most pesticides, may be contaminants in both ground and surface water. SOCs at chronic exposure levels may cause neurological and kidney effects and, for some pesticides, cancer (Hammer and Hammer 2001 ). One common organic compoun d that is

PAGE 30

30 targeted for removal by activated carbon is phenol which can be both manufactured and naturally occurring and is relatively common in the environment due to discharge from industries such as those producing plastics (Jung et al., 2001; Bansal et al. 2002 ). NOM, a mixture of fulvic humic acids, hydrophilic acids, and carbohydrates, can impart color to water and may also react with chlorine, added for disinfection purposes, to create disinfection by products (DBPs). DBPs, such as chloroform, a t rihalomethane, are considered carcinogenic ( Bull et al., 1982 ). It is, therefore, important to reduce the NOM concentration in the water before disinfection with chlorine occurs. A dsorption of NOM onto GAC is enhanced if calcium complexes with the NOM pr ior to adsorption by activated carbon (Frederick et al. 2001a). Taste and odor causing compounds, the removal of which constitutes the primary reason water treatment utilities use activated carbon, may result from biological growth or industrial activ ities. An example of a taste and odor causing compound resulting from biological growth is 2 methylisoborneol (MIB), which can be produced by cyano bacteria, or blue green algae (Gerber, 1968) MIB, described as having an earthy musty odor, enters a wate r treatment plant when the source water is surface water experiencing a bloom in the growth of the MIB producing organism. If occurrences of MIB are relatively short in duration PAC may be used as an additive during water treatment rather than a permanen t GAC filter. Reactivation of Activated Carbon During normal use, activated carbon will eventually become saturated with adsorbates so that the treated water exceeds the desired level of adsorbate removal, or reaches breakthrough. At this point, the spent activated carbon can be reactivated to restore the adsorption capacity of the GAC so that it can be reused rather than simply disposed of and replaced with virgin GAC.

PAGE 31

31 The time to breakthrough for a GAC bed can range from about 6 months to 5 years, depen ding on the type of adsorbate, the influent concentration, and the desired treatment level (Kawamura, 2000) Thermal or chemical regeneration may be employed, though thermal reactivation is the most commonly employed scheme. Thermal reactivation consis ts of four steps: drying, desorption, pyrolysis, and gasification. The drying process removes water and some highly volatile adsorbates. At higher temperatures, thermal desorption occurs with vaporization of volatile adsorbates and decomposition of unsta ble adsorbates to more volatile components. A pyrolytic process follows using high temperatures, between 923 1123 K in an inert environment that converts heavy or non volatile adsorbates to char. The final step, gasification at temperatures above 973 K in steam, CO 2 or a combination of both, involves desorption of the vapors and gaseous products of char and their exit from the pores of the reactivated carbon (Clark and Lykins 1989 ; Cannon et al. 1993 ; Snoeyink and Summers 1999). Reactivation has sev eral effects on the activated carbon, aside from the intended effect of, ideally, restored adsorption capacity. Mass loss during the activation process may range from 10 15% of the original GAC and is the result of loss through the transfer of GAC from th e treatment site to the regeneration site and/or burning of the GAC during reactivation (Clark and Lykins 1989). The loss is compensated for by the addition of virgin GAC before the reactivated GAC is returned to service. The surface chemistry of the ac tivated carbon is also affected by reactivation as most oxygen containing functional groups are stripped during reactivation at temperatures greater than 973 K and in an inert environment (Menendez et al. 1996 ; Pereira et al. 2003).

PAGE 32

32 Inorganic molecules such as calcium, may be adsorbed during normal operation of a GAC filter and are not removed during the activation or reactivation processes. The inorganic molecules remaining on the surface of the reactivated carbon prevent these adsorption sites from being freed for further adsorption and thus compromise the ability of the reactivation process to restore the adsorption capacity of the carbon (Clark and Lykins 1989). One well documented effect of calcium presence in the reactivation process is calcium catalysis (Knappe et al. 1992 ; Cannon et al. 1993 ; Mazyck and Cannon 2000 2002). Calcium catalysis increases the number of active sites on the surface of the carbon rather than, in the traditional sense of a catalyst, reducing the activation energy. Calcium catalyzes the gasification step of reactivation and is characterized by increased mass losses during the reaction process and pore enlargement (Knappe et al. 1992). As discussed previously the pore size of the activated carbon can affect its ab ility to adsorb certain compounds, and an increase in pore size may make it unsuitable for certain applications. There is also cost associated with the addition of virgin GAC to replace the increased mass lost during the reactivation process. Another re sult of reactivation is the incidence of pH excursions, an increase in the pH of treated water when reactivated carbon is returned to service (Farmer et al. 1996 ; Farmer et al. 1998). The duration and extent of the pH excursion will determine its impact on the system. pH Excursions in Water Treatment Many water and wastewater utilities have found that when thermally reactivated GAC is returned to service, the pH of the subsequently treated water is elevated (e.g., pH > 9) for days or even weeks (Mazyck et al. 2005). Water with high pH levels is unfit for distribution to customers, as it exceeds the United States Environmental Protection Agency secondary standards of the National Secondary Drinking Water Regulations (6.5 < pH < 8.5) and could also brin g about calcium carbonate (CaCO 3 ) precipitation problems in distribution systems.

PAGE 33

33 A water utility in the Virginia American Water Company, under the umbrella of the American Water Works Company, measured increases in pH in their backwash water when a reac tivated GAC was placed back in service (Figure 2 2 ). These data are representative of a full scale GAC filter bed adsorber. As shown, the reactivated GAC posed an adverse effect on water o greater than 11. Even after 6 days, the pH was still above 8.5. Note that the pH of natural water is approximately 7. In addition to the pH, the water utility also measured the concentration of calcium in the effluent for two days. As shown in Figure 2 2 as the calcium concentration decreased, the water pH likewise decreased. The influent calcium concentration in their backwash water was relatively constant over the two days. No other operational changes could have contributed to the higher Ca 2+ and OH levels; these results are fairly typical, as reported by American Water Works Company personnel. These results led to the hypothesis of a calcium chemistry mechanism proposed by Mazyck et al. (2005) and discussed further below. pH Excursion Mechanism s for Activated Carbons A mechanism involving anion exchange with hydroxide ions following carbon protonation has been proposed by Farmer and coworkers to explain pH excursions (Carr and Farmer 1995 ; Dussert et al. 1995 ; Farmer et al ., 1996). These auth ors suggested that pH excursions are a function of the carbon surface that is altered during high temperature activation or reactivation. Activated carbons exposed to high temperatures in a reducing atmosphere during manufacturing and reactivation tend to adsorb strong acids in water. Garten and Weiss (1957) classified these as H type carbons. In contrast, L type carbons are those produced by surface oxidation; these are known to adsorb strong bases in water. Farmer et al. (1996) argued that this acid a dsorption during water treatment might involve the protonation of pyrone type surface groups or other similar functionalities on the carbon surface, in agreement with Leon y Leon et

PAGE 34

34 al. (1992). In addition, they suggested that loading the carbon with sulf ate, chloride, or other anions that are present in water could neutralize the positive charge on the carbon surface. Furthermore, they proposed that this charge neutralization could also occur through an anion exchange process involving sulfates and hydro xides on the carbon surface. In other words, they surmised that the carbon could adsorb sulfates or other anions from water while releasing hydroxyls, and that this ion exchange process could cause the pH to rise. Farmer et al. (1996) did acknowledge tha t this anion exchange mechanism requires additional testing and confirmation. To date, however, there has been no further evaluation by these authors presented in the literature that would either prove or disprove their hypothesis. Dussert et al. (1995 ) supported this anion exchange mechanism because they found no significant pH excursion when virgin GAC was immersed in Milli Q (anion free) water, and they concurred that anions were required to trigger a pH effect. These authors came to their conclusio ns after studying virgin GACs that removed 2 to 9 mg of sulfate/g GAC. In Milli Q water spiked with 80 mg sulfate/L, they measured the water pH and sulfate capacity for several virgin carbons. Not only did they find that pH excursions only occurred when anions such as sulfate were present, but they also reported that, as the sulfate capacity increased, pH also increased. In their study, the carbons that did not exhibit pH excursions also did not significantly remove sulfate ions from solution (< 0.7 mg/g ). They also found that sulfate (SO 4 2 ) removal was greater than nitrate (NO 3 ) or chloride (Cl ) removal. They did not measure calcium concentrations for their carbons; however, they did find that the ash content of the carbon could not be correlated wi th pH excursio ns. To discern if an anion exchange mechanism was also valid in describing pH excursions with reactivated GAC, Mazyck et al. (2005) tested reactivated carbons under conditions similar

PAGE 35

35 to those used by Dussert et al. (1995) to support an anio n exchange mechanism. According to the anion exchange mechanism supported by Dussert et al. (1995), the sulfate uptake should be highest when there is a high pH excursion. The data of Mazyck et al. (2005) demonstrated an opposite trend where, as the pH e xcursions of the reactivated GAC increased, the sulfate uptake decreased. For example, at a water contact pH, a predictor of pH excursions, of 8.75 the sulfate uptake was 1.4 mg sulfate/g GAC while at a water contact pH of 10.6 there was no measurable sul fate uptake. In addition, pH excursions were found to occur even in Milli Q water, where sulfate ions were not present suggesting that anions are not responsible for the observed pH excursions with reactivated GAC. In light of these results, Mazyck et al (2005) concluded that the uptake of sulfate and the subsequent release of hydroxyl ions (i.e., an anion exchange mechanism) was not responsible for the observed pH excursions with reactivated GACs. The observation of a relationship between calcium and pH excursions, shown in Figure 2 2, led to a hypothesis by Mazyck et al. (2005) of a calcium chemistry mechanism behind pH excursions. During potable water treatment, water soluble calcium can adsorb onto GAC as a complex with NOM (Nowak and Cannon 1997 ; Nowak et al., 1999 ; Frederick and Cannon 2001 ; Frederick et al. 2001a 2001b). Moreover, calcium is abundant in many coal based carbons. Thermal reactivation of spent GAC at temperatures above 1123 K decomposes the oxygenated functionality of the NOM that complexes with calcium This process results in the formation of CaO, which is thermodynamically the most stable species at these high temperatures (Mazyck and Cannon 2000). When the reactivated carbon is returned to service, the calcined CaO can r eact with water to form Ca(OH) 2 ( Equation 2 10 ). Furthermore, Ca(OH) 2 can dissolve and subsequently dissociate as Ca 2+ and 2OH ( E quation 2 11 ). The release of OH would elevate the pH of the subsequently treated water.

PAGE 36

36 CaO + H 2 O Ca(OH) 2 (2 10) Ca(O H) 2 Ca 2+ + 2OH (2 11) Calcium chemistry and thermodynamics as related to GAC and thermal reactivation have been previously discussed in the literature at length (Cannon et al. 1994 ; Mazyck and Cannon 2000 2002). Therefore, it was proposed that the manner by which the reactivated GAC was contributing both calcium and hydroxyl ions to the water was via the mechanisms described in E quations 2 10 and 2 11 above. Furthermore, it was rationalized that the deviation from a [OH ] (molar concentration of hy droxyl ions) to [Ca 2+ ] (molar concentration of calcium ions) ratio of 2, is caused by the OH that originated from Ca(OH) 2 being consumed to some extent by the natural buffering capacity of the water. Thus, the data from the Virginia American Water Compan y (Figure 2 2) could be taken as a semi quantitative example of the mechanisms described in E quations 2 1 0 and 2 11 In further support of the calcium chemistry mechanism, a significant correlation was found between pH and calcium present in solution and altering the reactivation process to avoid the creation of calcium species that will participate in the proposed calcium chemistry, using a reactivation scheme proposed by Mazyck and Cannon (2000, 2002), was found successful in overcoming pH excursions (Ma zyck et al., 2005). Both of these finding supported the hypothesis of the calcium chemistry mechanism dominating pH excursions with reactivated carbons. Methods to Overcome pH Excursions The three broad categories of methods to overcome pH excursions are on site treatments, tailoring of the reactivation process, and post reactivation treatments performed before the reactivated GAC is returned to service.

PAGE 37

37 One on site method that would occur prior to reactivation is the reduction of the calcium adsorbed o nto the activated carbon during water treatment. To achieve this, either the NOM complexes with a stronger cation, thus out competing calcium, or the calcium NOM complex is removed before it comes into contact with the GAC. The use of an iron coagulant w ould be ideal as it could facilitate both of the above goals. Iron (Fe 3+ ) is a stronger cation than calcium and would displace the calcium from the NOM though competition with NOM by hydroxyl species may hinder the process near a neutral pH. Nowack et al (1999) proposed that, at near neutral pH, soluble iron hydroxide would precipitate depleting the amount of iron available for further reactions with NOM. T he iron would also coagulate the calcium NOM complexes already formed and this floc would be allow ed to settle out before reaching the GAC. Nowack et al. (1999) and Frederick et al. (2001) have shown that the addition of iron coagulant to a system containing both calcium and NOM yielded a decrease in the amount of calcium NOM complexes adsorbed to the carbon as well as a relatively level iron loading onto the GAC. These results indicate that, not only is iron addition successful in decreasing calcium loading, but that the primary mechanism responsible for this is coagulation of the iron NOM and/or cal cium NOM complexes. The accumulation of iron is not detrimental to the reactivation process as iron catalysis is suppressed by the presence of sulfur, which is ubiquitous in coal based carbons (Nowack et al. 1999). A near neutral pH was indeed found to hinder iron NOM formation as the hydroxide species out competed the NOM for complexation with the iron ions. Therefore, Nowack et al. (1999) and Frederick et al. (2001) concluded that a combination of pH adjustment to approximately 6.0 pH and iron additio n, on the order of 5 mg/L, are the most effective in reducing calcium loading. Water treatment facilities that operate above this range of pH are expected to accumulate higher levels of calcium onto their GAC.

PAGE 38

38 Other on site treatments to overcome pH exc ursions after the reactivated carbon is returned to service include backwashing, recycling of the high pH water, and pH adjustments before the treated water is discharged. Both backwashing and recycling involve running water through the GAC until the pH r eaches acceptable levels, which reduces the yield of treated water through the system (Farmer et al. 1998). In addition, backwashing may require a holding tank, as the high pH water cannot be discharged, which will restrict the allowable backwashing time (Carr and Farmer 1995). Acid treatment of the high pH water prior to discharge to stabilize the pH can be time consuming and expensive, due to additional monitoring equipment Acid washing a traditionally reactivated carbon prior to use, aside from bein g time consuming and expensive, may not even produce a pH stable carbon (Dussert et al. 1995). In addition, oxygen surface complexes are introduced to the surface of the carbon if an oxidizing acid is used and this can impact the adsorption capabilities of the reactivated carbon, as discussed previously Depending on the acid, the micropore volume may also decrease as a result of treatment. A reactivation protocol proposed by Mazyck et al. (2005) was discussed that can overcome calcium catalysis and the resulting pH excursions by reactivating the carbons at temperatures below that at which CaO is formed, 1123 K It is the CaO form that participates in the calcium chemistry mechanism. A decrease in the amount of CaO available to participate in the calciu m chemistry reactions will result in a decreased pH excursion. Oxidizing the reactivated carbon, outlined in a patent by Dussert et al. (1995), or treating the reactivated carbon with carbon dioxide (CO 2 ) outlined in a patent by Farmer et al. (1998), hav e been offered as means to create a pH stable carbon. Both patents propose that oxidation or CO 2 treatment, depending on the patent, neutralizes the surface of the reactivated carbon, thus

PAGE 39

39 inhibiting any anion exchange during water treatment. As the focu s of a previous study (Mazyck et al., 2005) was the role of calcium in pH excursions and the result was that the anion exchange mechanism does not fully explain the excursions found in reactivated carbons, there must be an additional explanation for why th ese treatments are able to create pH stable carbons Previous work by this author (Bach, 2004) determined that air treatments of reactivated carbons create pH stable carbons (i.e., pH < 8.5) through the formation of acidic oxygen containing functional gro ups on the carbon surface which balance out the hydroxyl ions added by the calcium chemistry mechanism. This was supported by an evaluation of the total surface acidity of the air treated versus untreated carbon. However, given the discussion of the role of individual functional groups on surface acidity, specifically the carboxyl group, an evaluation of individual functional groups would be required to further substantiate this proposal. CO 2 treatment was also found to create a pH stable carbon and i t w as proposed that CO 2 treatments form pH stable reactivated carbons via physisorption, and subsequent desorption, of CO 2 forming H 2 CO 3 and then HCO 3 upon water immersion which also balances out the hydroxyl addition by the calcium chemistry mechanism. Th i s hypothesis was supported by similar total surface acidity found between an untreated and a CO 2 treated carbon, though an analysis of individual functional groups rather than total acidity would further support this theory. The literature also demonstrat es that there would be no chemisorption of CO 2 onto CaO to form CaCO 3 which would not participate in the proposed calcium chemistry mechanism of Equation s 2 10 and 2 11 (Mazyck and Cannon, 2000) Following the treatments to overcome pH excursions and aft er evaluating their success, an air treated, a CO 2 treated and an untreated reactivated carbon were evaluated for their performance in the removal of a common taste and odor causing compound, MIB (Bach, 2004).

PAGE 40

40 The results demonstrated that the untreated r eactivated GAC removed more MIB than the air treated reactivated GAC in organic free water due to an increase in the effects of surface acidity with the air treated carbon, while this trend was reversed when NOM was present. The hypothesis behind the imp roved performance of the air treated carbon in water containing NOM attractive to the negatively charged NOM leaving more adsorption sites available for MIB. In both waters, the CO 2 treated reactivated GAC removed the greatest quantity of MIB. The CO 2 treated reactivated GAC was shown to adsorb slightly less water in a study of water vapor removal from air, suggesting that the CO 2 treated carbon deterred water a dsorption, which is deleterious to MIB adsorption. These findings highlight the importance of investigating changes in adsorption performance following deliberate changes in surface chemistry, as was done in order to overcome pH excursions. Influence of Dissolved Oxygen in Water used to Create Steam To this point, the impacts of surface chemistry on activated carbon, including its role in adsorption, as well as methods to tailor this surface chemistry for a specific goal (e.g., overcoming pH excursions) h ave been discussed. Rather than tailor surface chemistry following process has also been explored, namely altering the dissolved oxygen (DO) content of the wat er used to create steam for steam reactivation (Chestnutt et al., 2007). Water may be converted to steam for several industrial applications including laundry services, locomotion (i.e., steam engines), food manufacturing, wood pulp products, and activated carbon creation. When water is heated at sufficient temperatures to convert it to steam, any DO present in the system will devolve as gaseous oxygen (O 2 ). The devolution of oxygen in boiler systems has been documented as creating corrosion and pitting w ithin the piping systems

PAGE 41

41 employed in these facilities ( Bizot and Bailey, 1997 ). First, the metal iron in the pipe dissolves ferrous hydroxide (Fe(OH) 2 ) (Eq uation 2 12). The ferrous hydroxide then reacts with oxygen to form iron (III) hydroxide (Eq uation 2 1 3), the dehydration of which forms iron (III) oxide more commonly known as rust ( Bizot and Bailey, 1997 ). Fe 2+ + 2OH 4 Fe(OH) 2 (s) (2 12) 4Fe(OH) 2 + O 2 + 2H 2 O 4Fe(OH) 3 (s) (2 13) This reaction demonstrates the impact that the DO present in water used to create steam may have on materials with which it comes in to contact. With the impact of DO in mind, Chestnutt et al. (2007) studied the effect of DO in the water used to create steam on the reactivation of carbon. While several reactivation scenarios (i.e., different time, temperature regiments, etc.) were ev aluated, Figure 2 3 shows an analysis of the performance of three reactivated carbons exposed to the same reactivation scenario only with differing DO values in the water used to create the steam. The performance in this study was evaluated using a column study where the effluent concentration of the target compound, MIB, was evaluated versus the volume of water treated, or bed volumes (BV). As demonstrated by Figure 2 3, the change in the DO content of the water used to create steam yielded a difference in the performance of the resulting reactivated carbon. Namely, the reactivated carbon exposed to steam resulting from water with high DO treated fewer bed volumes to breakthrough than a reactivated carbon exposed to steam created from low DO water. Here breakthrough is defined as the number of bed volumes treated before the effluent exceeds the odor threshold concentration (OTC) of MIB (i.e., the concentration of MIB above which it is generally noted by the human palate) of 10 ng/L (Rashash et al., 1997 ). The impact of DO on the adsorption performance of the resulting carbons

PAGE 42

42 was related to an increase in surface acidity of the carbons as the level of DO in the water used to make steam increased. The impact of surface acidity on MIB adsorption has been discussed at length. While the mechanism behind the creation of increased acidity with an increase in the DO content of the water was not investigated in the work of Chestnutt et al. (2007), the results of this work do indicate that altering the DO of t he water used to create steam for reactivation of carbon is a plausible means by which to alter surface chemistry and thereby influence the adsorption properties of reactivated carbon. Figure 2 1. Proposed mechanism for the fo rmation of COOH on an activated carbon surface Figure 2 2. Full scale water treatment plant effluent pH and calcium concentration data after reactivated GAC was returned to service (Mazyck et al., 2005). + H 2 O Formed during air treatment C C O O O C C O OH O OH Carbon surface

PAGE 43

43 Figure 2 3. Comparison of performance in the removal of MIB for reactivated carbons created with varying DO content in the water used to create steam (Chestnutt et al., 2007).

PAGE 44

44 CHAPTER 3 HYPOTHESIS AND OBJEC TIVES The objectives of the current work can be divided into two main categories: research related to overcoming pH excursions and research related to tailoring activated carbon using the DO content of the water used to create steam. pH Excursions Hypothesis Air treatments overcome pH excursions via increased surface acidity, specifically the c arboxyl group and not via suppression of the calcium chemistry mechanism. CO 2 treatments overcome pH excursions via physisorption of CO 2 to the carbon surface and subsequent desorption to form carbonic acid and not via suppression of the calcium chemistry mechanism. Performance of a CO 2 treated carbon for removal of MIB is higher than both air treated and untreated carbon due to a decreased wetting time associated with the CO 2 treated carbon. Performance of an air treated carbon for MIB removal in water co ntaining NOM is better than that of an untreated carbon due to the preference of the untreated carbon for NOM adsorption Methods to overcome pH excursions, namely air and CO 2 treatments, will adversely affect the removal of phenol using a granular activat ed carbon column as increases in surface acidity are known to be deleterious to the adsorption of phenol and the surface chemistry of carbon can affect adsorption in both batch and column systems Objectives Investigate individual acidic functional groups on the air treated, CO 2 treated, and untreated carbons versus the ability of these treatments to overcome pH excursions. Investigate mass change of carbon, indicative of CO 2 adsorption, during CO 2 treatment. Evaluate calcium in solution after contact wit h air and CO 2 treated carbons to determine relative progression of the calcium chemistry mechanism as compared to an untreated carbon. Determine relative adsorption of MIB for a CO 2 treated and an untreated carbon after contact with only water for vario us lengths of time to confirm a difference in wetting time.

PAGE 45

45 Measure concentration of NOM remaining in solution after contact with MIB and an air treated, CO 2 treated, and an untreated carbon to support hypothesis of competitive adsorption between MIB and N OM. Compare the performance of an air treated, CO 2 treated, and untreated carbon for the removal of phenol in a column study. Tailoring of Activated Carbon Using Dissolved Oxygen (DO) Hypothesis A ltering the DO content of the water used to create steam du ring steam activation of a wood based precursor will alter the surface chemistry/physical structure of the activated carbon affecting adsorption performance Objectives Evaluate the impacts of four different activation parameters (time, temperature, steam flow, and DO content of the water used to created steam) at three different levels (low, medium, and high) on the adsorption performance, physical properties, and chemical properties of the resulting activated carbon. Evaluate different methods of alteri ng DO in the water used to create steam for activation and their impact on removal of MIB.

PAGE 46

46 CHAPTER 4 EXPERIMENTAL METHODS Reactivated Carbon Samples for Overcoming pH Excursions Ten coal based reactivated GAC samples, designed for adsorption from an aqu eous solution, were received from the NORIT Americas, Inc. plant in Pryor, OK. These carbons had been used in potable water treatment, had become spent, and had been subsequently reactivated using traditional methods in Pryor, OK. A composite of three of these samples (i.e., these samples were commingled) having the highest calcium content (5.6 7.8 g Ca/kg carbon) and also exhibiting the highest pH excursion was used throughout th e research into pH excursions The BET surface area of the composite was 89 6 m 2 /g with a micropore (< 2 nm ) volume of 0.35 cm 3 /g and a mesopore/macropore (>2 nm ) volume of 0.13 cm 3 /g. Carbon Precursor for Activations A wood based fly ash was used as the carbonaceous precursor in this study. Fly ash from wood combustion has been investigated as an addition to composts to aid in adsorbing odors, controlling pH and moisture content, and nutrient addition ( Campbell et al., 1997; Rosenfeld and Henry, 2000; Das et al., 2003) Wood based fly ash also outperformed a coal fly ash in the catalytic oxidation of H 2 S and methanethiol, due to the higher surface area of the wood based fly ash, in a study of sulfur compound removal from gaseous waste streams (Kastner et al., 2003). The fly ash used in this study was a side product of combustion taking place at a wood processing facility in Florida. By utilizing a wood based fly ash, the activation uses a renewable resource as well as recycling a waste stream which enhances the environmental impetus to use such a material.

PAGE 47

47 Air Treatments and C O 2 Treatments of Reactivated Carbons Air treatments were conducted in ambient air using a muffle furnace. Three grams of carbon were placed in crucibles at the desired temperature inside a muffle furnace where they were continuously exposed to 52.5 mmol O 2 /g GAC for 15, 30, or 60 minutes. The air treatment temperature range for the air treatments of between 623 K and 773K. After treatment, the samples were allowed t o cool in a dessicator and were stored until further use. A clamshell furnace similar to that used by Mazyck and Cannon (2000 2002) held a quartz fluidized bed wherein CO 2 treatments were carried out at ambient temperatures. A porous plate near the cent er of the quartz fluidized bed allowed a flow of 0.9 L/min of CO 2 to contact three grams of sample for 20 minutes for an exposure of 281 mmol CO 2 /g carbon. The treatment is similar to that employed by Farmer et al. (1996) with the exception that dry GAC w as exposed to CO 2 instead of wet GAC. Activations with Varying DO Content in Water Using the same clamsh ell furnace used for CO 2 treatments, a ctivations involved exposing five grams of sample, resting on a porous plate near the center of the quartz flu idiz ed bed, to a steam flow under the desired conditions Steam flow is reported as g H 2 O/g carbon which is the mass of water, determined by the flow of steam as measured with a flow gauge multiplied by the given period of activation, divided by the mass of carbon with which the steam comes into contact. A ctivations were preceded by a pyrolysis step where the five grams of sample w ere exposed to the activation temperature in a flow of pure nitrogen gas for a time equal to the steam exposure (i.e., activation time) level in the deionized water used to create steam under varying experimental conditions, as well as the concentration of other

PAGE 48

48 dissolved gases, in order to compare these theoretical values to t he measured DO concentrations. (Equation 4 1) relates the concentration of gas in solution to the partial pressure of the gas above the solution. Partial pressure (P x ) pressure exerted by a single gas in a gaseous mixture where the total pressure equals the sum of the partial pressures of each gas in the mixture (Petrucci and Harwood, 1997 ). C equals the concentration of gas X in solution (mol/L); k H P x is the partial pressure of gas X (atm). C x = k H P x ( 4 1 ) The DO content of the water was manipulated based on the gas (e.g., pure nitrogen air, o r pure oxygen) used to pressurize a water vessel set at 10 psi (gauge) which then delivered the water t o a preheater set at approximately 723 K which then sent steam to the fluidized bed reactor. Nitrogen gas pressurization yielded a DO concentration of less than 4 mg/L air pressurization yielded a DO of approximately 7.5 mg/L, and pure oxygen pressurizat ion created a DO concentration of greater than 10 mg/L. The DO concentration was measured inline using a Mettler Toledo inline DO probe placed just prior to a water gauge which measured water flow into the preheater. Following activation, the carbon samp les were cooled to ambient temperatures in the furnace under a nitrogen flow to prevent interactions with the atmosphere at high temperatures. Water Contact pH The water contact pH of the carbons was determined by mixing approximately 2 grams of as receiv ed GAC with 80 mL of Milli Q water for 30 minutes, then measuring the pH. This ratio of GAC to solution was the same used by Dussert et al. (1995) and Farmer et al. (1996). In initial exploratory experiments, pH measurements were taken every five minutes and it was observed that after 20 or 30 minutes the pH reached a pseudo equilibrium level (less than 0.1

PAGE 49

49 change in pH over 5 minutes). Dussert et al. (1995) used a similar technique and concluded that this protocol provides a good prediction of the pH t hat would be measured during water treatment service. For the experiments described herein, each reactivated carbon was sampled twice and the pH for these two samples was within 0.4 pH units. Calcium Solution Concentration Following the water contact pH experiment s the GAC was separated from the solution utilizing a vacuum filtration apparatus (0.45 m Whatman filter paper) The c alcium levels leached out of the GAC during the 30 minute water contact pH experiment were analyzed by inductively couple d plasma (ICP) emission spectroscopy (Perkin Elmer, Optima 3200RL) The measured concentrations of the replicas were within 3.5% of one another. Adsorption Experiments (Batch) For batch equilibrium studies, PAC was created by g rinding the carbons in a c rucible and Then a 0.5 L stock solution of 153 ng radiolabeled C 14 methanol stock solution with 500 mL of either deionized (DI) water or natural water influent to the Manatee County Water Treatment Plant (Bradenton, Florida) (Total o rganic c arbon=10.0 mg/L; Specific UV Absorbance=4.78 L/mg m; pH=7). Note, MIB (1,2,7,7 tetramethyl exo bicyclo(2.21)heptan 2 ol) is a moderately hydrophobi c compound of low molecular weight with a molecular size of approximately 6 (Newcombe et al., 2002 ; Rangel Mendez and Cannon 2005). From the 153 ng radiolabeled C 14 MIB/L solution, 50 mL allotments were added to separate gas tight syringes followed by addition of 10 ppm carbon The syringes were then sealed, permitting a small amount of headspace, and allowed to mix on a rotisserie style mixer. After 24 hours, a 0.45 m luer lock nylon filter, attached to each syringe, was used to separate

PAGE 50

50 the carbon from 9 mL of the solution, which was then tested according to previous studies using radiolabeled C 14 MIB (Gillogly et al. 1998 ; Tennant and Mazyck 2003). Each sample was measured twice for MIB and the mean of these two runs was used for the final value Typical reproducibility was within 1%. The entire process, starting with 50 mL samples of either DI or natural water, was repeated for each carbon. Each dose experiment was run in replicate and the error bars shown in the figures represent one standa rd deviation from the mean of these re plicate runs. Select samples were analyzed for NOM remaining in solution using a Tekmar Dohrmann Apollo HS 9000 with an autosampler to measure nonpurgeable dissolved organic carbon (DOC) Error is represented as plus or minus one standard deviation from the mean of triplicate analyses. For batch studies with phenol, 30 mL of 5 mg/L phenol where mixed in vials containing 50 ppm carbon for 24 hours. The PAC was then separated from the solution utilizing a vacuum filtra tion apparatus (0.45 m Whatman filter paper) The samples were stored at 277 K under low light conditions until they were ready to be measured at which point dilutions were employed to ensure detection on a gas chromatography mass spectroscopy instrument (Varian, Saturn 2100P). The averages of replicate experiments are reported with error bars representing one standard deviation from the mean. Rapid Small Scale Column Tests (RSSCTs) For column studies with phenol as the target compound, a commercially available activated carbon (F400) was exposed to the same air and CO 2 treatments used to overcome pH excursions and then these carbons were evaluated using RSSCTs. Details pertaining to the RSSCT design are listed in Table 4 1 The advantages of RSSCTs a re that: (1) they can provide a breakthrough profile in a fraction of the time compared to full or pilot scale studies; (2) a

PAGE 51

51 smaller volume of water is required for testing; and (3) extensive isotherm studies are not required to predict full scale perfor mance (Crittenden et al., 1986; Crittenden et al., 1991). Although the RSSCT can significantly reduce the time and cost of the study, the limitation of the RSSCT is that it can over predict full scale GAC performance (Summers et al., 1989; Crittenden et a l., 1991); nevertheless, it is not expected that the relative comparison between carbons would change. The RSSCTs employed in this study were designed using full scale design parameters and scaling equations available in the literature (Crittenden et al. 1991). The RSSCT column was manufactured from acrylic (1.5 cm x 0.41 cm) (0.59 in. x 0.16 in.) with Teflon inserts to prevent the sorption of compounds Similarly, a Masterflex L/S rigid PTFE pump head, PTFE tubing, and PTFE fittings were used througho ut the system to avoid sorption of compounds Throughout the entire experiment, a flow meter monitored water throughput (4.4 mL/min), and all experiments were performed in duplicate The various reactivated carbons were evaluated using the RSSCT by s piking aliquots of deinoized water with a phenol concentration of 5 m g/L. At the time of sampling, a 20 mL vial was placed at the discharge tube where 13 mL of sample were collected over 3 minutes. The samples were stored at 277 K under low light conditi ons until they were ready to be measured at which point dilutions were employed to ensure detection on a gas chromatography mass spectroscopy instrument. The averages of replicate experiments are reported with error bars representing one standard deviatio n from the mean. BET Surface Area Measurements To determine surface area, the carbon samples were first outgassed at 383 K for 24 hours. Then, an N 2 adsorption isotherm was carried out at relative pressures between 0.05 and 0.35 in a NOVA 1200 (Quantach rome) gas sorption analyzer. The resulting data was then analyzed using

PAGE 52

52 the BET model to determine the surface area of each sample. Duplicate samples were analyzed with average error less than 5%. Pore Size Distribution Measurement Pore size distribution s (PSDs) were performed using a Quantachrome Autosorb I. Each carbon sample was first outgassed at 383 K for 24 hours. The sample cell was then placed in a liquid nitrogen bath which created an analysis temperature of approximately 77 K with nitrogen gas used as the adsorbate. The density functional theory (DFT) was used to develop the pore size distribution data in the micropores (< 2 nm ) from the nitrogen isotherm with a relative pressure range from 10 6 to 1 while the Barrett, Joyner, Halenda ( BJH ) m odel was used for all pores greater than 2 nm pH at the Point of Zero Charge To determine the pH at the point of zero charge (pH pzc ), where the point of zero charge is defined as the condition where the charge on the surface of the carbon is neutral or zero, an aqueous system where the carbon sample dominates the pH of the system was evaluated. To this end, 0.5 g of carbon sample was mixed with 5 mL of nanopure water which was previously bubbled with nitrogen gas for 20 minutes to remove any carbon diox ide from the system that may interfere with the influence of the carbon sample on the water pH. The headspace above the solution was also purged with nitrogen gas. The vial was then covered with parafilm to eliminate mixing with the atmosphere and the sa mple was allowed to mix for 24 hours in order for the carbon to reach equilibrium with the solution. After 24 hours, the pH of the sample was determined using a standard pH electrode and this pH measurement was recorded as the pH pzc Each pH pzc measureme nt was performed in duplicate with each experiment performed in replicate. The replicates were within 1%.

PAGE 53

53 Boehm Titration The concentration of acidic oxygen complexes created on the surface of the carbon samples was determined according to the Boehm titr ation method (Boehm 1966). An allotment of 0.5 g of carbon and 25 mL of either 0.05 N NaOH, 0.25 N NaOH, 0.05 N NaHCO 3 or 0.05 N Na 2 CO 3 were shaken in sealed vials for three days. The suspensions were then filtered using a he filtrate was titrated with 0.1 N H 2 SO 4 to a pH of 4.5. The amount of base consumed by each sample was determined by comparing each titration to that of a blank. Any base consumed by the sample results from the neutralization of acidic functional group s and is dependant on the type of base used Carboxyl groups react with NaHCO 3 ; lactone groups are measured with the difference between the reaction with Na 2 CO 2 and NaHCO 3 ; phenolic groups are measured with the difference between reactions with 0.05 NaOH and Na 2 CO 2 ; the difference between 0.25 NaOH and 0.05 NaOH measures carbonyl groups; total acidity is measured with 0.05 NaOH The error in the measure of functional groups represents the highest and lowest values found in repeated experiments. Statistica l Design of Experiments: Box Behnken Traditionally, the evaluation of the e ffect of several variables on a result is made using a full factorial, one factor at a time, or the fractional factorial approach to obtaining data. The full factorial approach inv olves the collection of data relating to all possible combinations of the factors (variables). With four factors under investigation in this research and a proposed three levels (settings) for each factor, the full factorial method corresponds to 81 exper iments. The benefit of such an approach is that the choice of optimum conditions will be clear when the data are analyzed; the optimum experimental conditions will be run as one of the experiments. However, performing at least 81 experiments, as this num ber does not take into account replicate experiments, can be expensive in terms of both cost and time.

PAGE 54

54 Given the drawbacks of a full factorial approach, many researchers revert to the one factor at a time approach, which evaluates the effect of varying o ne factor while holding all other factors constant. Once the level that produces the best result has been determined for this factor, the level is fixed while another factor is evaluated with the remaining factors constant. The one factor at a time appro ach used in this research would yield nine experiments. Clearly, this approach would be tempting to a researcher when compared with the 81 runs required for a full factorial approach. However, the one factor at a time approach has significant drawbacks w hen considering the validity of the final optimum conditions, as it does not take into account interactions. An interaction between factors occurs when two or more factors acting together have an e ffect on the result greater than the factors acting alone. The fractional factorial approach mediates the cost and result validity issue by utilizing the principles of the design of experiments (DOE). DOE is defined as the evaluation of the simultaneous e ffect of changing variables on the final product in orde r to gain the maximum amount of definitive information with the minimal amount of expended resources. The fractional factorial approach uses only a predetermined set of experiments chosen from the full factorial. Unlike the one factor at a time approach, there is a logical plan behind the choice of experiments run from the full factorial and statistical analysis yields a final result with a specific associated value of certainty, without the requirement to run each combination in the full factorial. Ther efore, fractional factorials, specifically a Box Behnken design (Box and Behnken, 1960) will be used in this research with 29 runs required. The proposed Box Behnken design is formed by comb in ing a 2 k factorial design where 2 is the number of levels and there are k variables being studied with incomplete block designs to

PAGE 55

55 produce a 3 k fractional factorial (Montgomery 1991). A 2 2 full factorial design is shown in Table 4 2. In a Box Behnken design each pair of factors being studied is linked to a 2 2 f actorial where the two factors are varied at two levels (high and low) while the third factor is fixed at its center level (middle). The combination of factorials in this fashion for a Box Behnken 3 3 design is shown in Table 4 3. A visual representation of the Box Behnken design of 3 3 is shown in Figure 4 1, where each dot represents experiments and the vertices represent the three factors. It should be noted from the 3 3 design that, although 12 experiments are required to complete the combination of fa ctorials, there is also an additional point at the center. A 3 4 design, which is the suggested Box Behnken design for this experiment, requires four axis for which the visual representation is less simple but the same theory applies. This design would con tain 24 design points plus a certain number of runs at the center point. It is from this replication of center points that analysis of error can be made. Myers (1995) recommends 3 5 replicates of the center points and the upper limit of this will be used in this Box Behnken design for a total of 29 runs (i.e., 24 runs plus five runs at the center point). The center point is replicated to obtain the pure error associated with the experiment Pure error in an experiment accounts for any random error in experimentation that is not experiments at the center point, pure error has four degrees of freedom (five replicate experiments at the center point minus o ne) and there are 10 degrees of freedom for lack of fit. In this way, the Box Behnken is able to incorporate a level of significance in the final result without the need to repeat every experiment. To ensure that the error is random, the required 29 runs are

PAGE 56

56 performed in a computer generated random order. The pure error is then used to analyze the lack of fit, while the total residual error (i.e., pure error plus lack of fit) is used to determine the significance of each factor evaluated according to the common evaluation method known as the F value. Equation 4 1 shows how to calculate the F value to determine the significance of the the generated model to predict results fits well with the data). F value = (4 1) To determine whether an F value is significant to a 0.05 level of significance, or a 95% confidence interval, as used in this study, the P value (Prob > F) was employed. The P value gives the probability of obtaining an F value at least that high due solely to random error, where a P value less than 0.05 would indicate a significant parameter such that there is less than a 5% chance that an F value that large co uld be due solely to random error. Equation 4 2 shows how to calculate the F value to determine the significance of a factor. F value = (4 2) For this study, the Box Benken approach was evaluated using Design Expert soft ware, version 6.0.5, distributed by Stat Ease, Inc. in 2001. Table 4 1. RSSCT and full scale design parameters Parameter Full s cale RSSCT* Grain Size (U.S. mesh) 12 x 40 170 x 200 EBCT (min) 7.5 0.045 Hydraulic l oading r ate (mm/min) 167.5 335.1 Time to p rocess 1000 BV 5.2 d 0.74 hr *Constant diffusivity conditions based on Speitel et al. ( 2001 )

PAGE 57

57 Table 4 2 2 2 full factorial Experiment Variable A Variable B 1 Low Low 2 Low High 3 High Low 4 High High Table 4 3. Box Behnken 3 3 fractional factorial Experiment Variable A Variable B Variable C 1 Low Low Middle 2 Low High Middle 3 High Low Middle 4 High High Middle 5 Low Middle Low 6 Low Middle High 7 High Middle Low 8 High Middle High 9 Middle Low Low 10 Middle Low Hig h 11 Middle High Low 12 Middle High High Center run Middle Middle Middle Figure 4 1. Box Behnken design of 3 3

PAGE 58

58 CHAPTER 5 STRATEGIES FOR OVERC OMING p H EXCURSIONS FOR REA CTIVATED GRANULAR ACTIVATED CARBON: AI R AND CARBON DIOXIDE TREATMENTS Previou s research into the mechanisms behind pH excursions have demonstrated the plausibility of a calcium chemistry mechanism behind pH excursions experienced with reactivated carbons (Mazyck et al., 2005). Further research has suggested that mechanisms to ove rcome pH excursions, namely air and carbon dioxide (CO 2 ) treatments, are effective in overcoming pH excursions due to addition of acidity to solution either by increased total surface acidity, as with air treatments, or physisorption of carbon dioxide and subsequent desorption to form carbonic acid, as with CO 2 treatments (Bach, 2004). However, additional research needs to be done to support these mechanisms and, therefore, the intent of the current work is to bolster the hypothesis made in the previous wo rk via additional experiments. Due to the fact that the current work builds directly from previous work, certain figures have been reproduced for the work done i n this study. Creation of a pH Stable Carbon Air treatments of a coal based reactivated carbon were conducted according to the patent of Farmer et al. (1998) to investigate the creation of a pH stable reactivated GAC which has a water contact pH less than 8.5 pH units, as this pH falls between the NPDES permitted discharge pH of 6 to 9 while still allowing for a small safety factor (Bach, 2004) A pH stable carbon is not expected to exhibit a pH excursion upon return to service (Dussert et al. 1995; Farm er et al. 1998; Mazyck et al. 2005). The resulting water contact pH of the air treated carbons treated at three different temperatures and treatment times measured in deionized (DI) water, is shown in Figure 5 1 Dashed lines at a water contact pH of 8.5 and 6.5 highlight the cutoff between an

PAGE 59

59 acceptable and unacceptable water contact pH incorporating a safety factor. The first data bar, at time zero, represents the water contact pH of the untreated carbon (i.e., reactivated GAC). While a pH stable c arbon is not created at any of the three treatment times when the temperatures is 623 K, after 60 minutes of air treatment at 673 K and after only 30 minutes of treatment at 723 K a pH stable carbon is created. These results indicate that treatments to ov ercome pH excursions with virgin carbons are also successful in overcoming pH excursions with reactivated carbon in DI water. The CO 2 treatment scenario, wherein the composite sample was exposed to a CO 2 flow of 0.9 L/min for 20 minutes (281 mmol CO 2 /g) at ambient temperature, also created a pH stable carbon (pH=8.3) (Bach, 2004) The CO 2 flow rate and contact time were chosen based on preliminary experiments demonstrating that this time and flow rate were sufficient to create a pH stable carbon (Figure 5 2 ). At 20 minutes, a pH stable carbon wa s created at the lowest carbon dioxide flow (i.e., 0.9 L/min) and, therefore, this treatment time and flow w ere used in all subsequent analyses. Again, the proposed method to overcome pH excursions experienced with virgin carbons in water containing anions is also successful in overcoming pH excursions with reactivated GAC in DI water. Surface Acidity of Air Treated Carbons The air treatments tested herein bound th e ideal temperature for production of acidic functio nal groups, specifically carboxyl groups, which leads to the hypothesis that air treatments overcome pH excursions by creating acidic functional groups on the GAC surface which balance out the hydroxyl ions imparted to the water via the calcium chemistry m echanism. The quantity of individual functional groups on the surface of air treated carbons with a treatment time of 60 minutes and the untreated carbon is shown in Figure 5 3

PAGE 60

60 As expected, the air treatments that created pH stable carbons (i.e., treatme nts at 673 K and 723 K) show an increase in total acidity compared to the untreated carbon. The 623 K treatment, which did not create a pH stable carbon, had a similar quantity of oxygen containing functional groups as the untreated carbon except for lact one groups, which were slightly more prevalent in the air treated carbon. Interestingly, though 60 minute air treatments at 673 and 723 K yielded pH stable carbons, only the treatment at 723 K showed an increase in the quantity of carboxyl groups compared to the untreated carbon. This result suggests that several acidic functional groups are responsible for creating pH stable carbons, as the 673 K treatment yielded a pH stable carbon without a significant increase in carboxyl groups. The 723 K air treate d carbon also showed an increase in the lactone and phenolic groups compared to all other samples. The carbonyl functional group was found to form to a significant extent in the 673 K sample compared to all other samples and, like the 723 K sample, it sho wed an increase in the lactone and phenolic groups compared to the untreated sample. Given this information, it seems likely that the lactone and phenolic groups play a role in creating a pH stable carbon at these temperatures. Surface Acidity of CO 2 Tr eated Carbons The CO 2 treated and the untreated carbon yielded similar quantities of acidic functional groups on the carbon surfaces (Figure 5 4 ), indicating that addition of surface acidity is not responsible for overcoming pH excursions with CO 2 treatmen t. Here, the mechanism that creates a pH stable carbon is likely physisorption of CO 2 to the carbon surface. Upon immersion into water, the CO 2 desorbs and reacts with water to form carbonic acid (H 2 CO 3 ) and then bicarbonate (HCO 3 ) after deprotonation at the pH typically observed in water treatment. Based on calculations predicting the pH of a system where only the

PAGE 61

61 dissolution of CO 2 from the atmosphere and the calcium chemistry mechanism influences pH, the theoretical equilibrium pH is approximately 8 6 compared to the measured water contact pH of 8.3 with the CO 2 treated carbon. T he theoretical equilibrium pH for the untreated carbon is also approximately 8.5 though the measured water contact pH is 10.7. It is possible that the additional CO 2 which is expected to dissolve upon contact with water, on the CO 2 treated carbon. An analysis of the weight gain following CO 2 treatment showed a slight increase in weight which, assuming the mass gain is due solely to CO 2 physisorption, corresponds to an addition of 1.1* 10 5 M CO 2 T he equilibrium concentration of CO 2 dissolved into solution from the atmosphere at room temperature is 0.9 10 5 M CO 2 /L, which may support the hypothesis that CO 2 treatment increases the rate at which the solution reaches equilibrium with the atmosphere. For example calculations of CO 2 in solution, see Appendix A. As with air treatment, it is proposed that the CO 2 treatment balances out the basic e ffects of the calcium chemistry mechanism, for the bicarbonate buffers the hydroxyl ions released from the calcium hydrolysis, thus depressing the magnitude of the pH excursion. It is not anticipated that CO 2 treatment forms CaCO 3 to any gr eat extent, which would not participate in the pH excursion mechanism, because the optimum chemisorption temperature of CO 2 to CaO, to form CaCO 3 is near 573 K and the CO 2 treatments used herein occurred at ambient temperatures (295 K) (Mazyck and Cannon 2000). Verification of Mechanisms To support the hypothesis that neither air treatments nor CO 2 treatments inhibit the calcium chemistry mechanism from progressing, an examination of the calcium that dissolved from the carbons shown in Figure 5 3 and F igure 5 4 was performed (Table 5 1).

PAGE 62

62 An analysis of the calcium in solution after CO 2 treatment showed that the untreated carbon yielded 34 mg/L Ca 2+ and the CO 2 treated carbon created approximately 30 mg/L Ca 2+ These findings suggest that CO 2 treatment results in a slight decrease in dissolved calcium, though it is not significant enough to fully account for the creation of a pH stable carbon as this change of 4 mg/L Ca 2+ would only be expected to decrease the pH by 0.7 versus the observed decrease in pH of 2.1. Table 5 1 also shows an unexpected increase in dissolved calcium present in solution at air treatment temperatures of 673 and 723 K (60 and 69 mg/L, respectively). An increase in the concentration of calcium in solution suggests that there is more calcium involved in the proposed calcium chemistry mechanism responsible for the pH excursion. A plausible explanation is that not all of the calcium adsorbed to the activated carbon during service reacted to form CaO during reactivation. During th e subsequent air treatment to create a pH stable carbon, the calcium not already present as CaO (e.g., CaCO 3 ) reacted to form CaO, the thermodynamically stable species at the temperatures and partial pressure of oxygen used during air treatments herein (Ma zyck and Cannon 2000). Conversely, a portion of the original adsorbed calcium NOM complex may have remained through reactivation and was subsequently converted to CaO by the air treatments employed herein. To determine if air treatments act to create mo re CaO on the carbon surface, which then participate in the proposed calcium chemistry mechanism, treatments at the same temperature but in a nitrogen environment were performed and the resulting calcium in solution evaluated. The nitrogen treatments resu lted in approximately half the concentration of calcium in solution (32 and 34 mg/L at 673 and 723 K, respectively) that was obtained with air treatments at the same temperature, supporting the hypothesis presented above. This suggests interplay between t he addition of increased basicity (as demonstrated by an increase in dissolved calcium

PAGE 63

63 and thus more hydroxyl ions) and increased acidity (as demonstrated by Figure 5 3 ) with air treatments. Summary a nd Conclusions Air treatments and CO 2 treatments desig ned to create pH stable carbons from virgin carbons in anion containing water were successful at creating pH stable reactivated carbons in deionized (ion free) water. These findings support the work of Mazyck et al. (2005) which question ed the plausibilit y of an anion exchange mechanism and the suppression thereof as an explanation for the creation of a pH stable carbon. In other words, the treatments created pH stable carbons even in anion free water. Air treatments of reactivated carbons at 673 K and 60 minutes or 723 K for 30 or 60 minutes were found to create pH stable carbons through the formation of acidic oxygen containing functional groups on the carbon surface. It is proposed that CO 2 treatments form pH stable reactivated carbons via physisorpt ion, and subsequent desorption, of CO 2 forming H 2 CO 3 and subsequently HCO 3 upon water immersion. While an increase in the mass of the carbon due to CO 2 physisorption was observed, additional experiments would need to be performed to confirm this hypothes ized mechanism. However, an analysis of the functional groups on the CO 2 the treated carbons ruled out both increased acidity and calcium chemistry suppression as p lausible mechanisms behind the creation of a pH stable carbon with CO 2 treatment

PAGE 64

64 Table 5 1. Calcium in solution with treatments Temperature (K) Calcium in solution ( 1mg/L) Ambient (Untreated) 34 Ambient (CO 2 ) 30 623 (Air) 37 673 (Air) 60 723 (Air) 69 673 (Nitrogen) 32 723 (Nitrogen) 34 Figure 5 1. Effect of air exposure time and temperature on water contact pH of reactivated carbon Dashed lines represent a pH of 8.5 and 6.5, respectively (Bach, 2004).

PAGE 65

65 Figure 5 2 Effect of treatmen t time and gas flow on resulting water contact pH of CO 2 treat ed carbon. Figure 5 3. Effect of air treatments after 60 minute contact time at varying temperature on individual acidic functional groups.

PAGE 66

66 Figure 5 4 Effect of CO 2 treatments on individual functional groups.

PAGE 67

67 CHAPTER 6 METHODOLOGIES FOR OV ERCOMING p H EXCURSIONS FOR REA CTIVATED GRANULAR ACTIVATED C ARBON: EFFECT ON ADS ORPTION PERFORMANCE The work discussed in Chapter 5 supported the hypothesis that air treatments of reactivated carbons cre ate pH stable carbons (i.e., pH < 8.5) through the formation of acidic oxygen containing functional groups on the carbon surface which balance out hydroxyl ions added by the calcium chemistry mechanism. The hypothesis that CO 2 treatments form pH stable re activated carbons via physisorption, and subsequent desorption, of CO 2 forming H 2 CO 3 and then HCO 3 upon water immersion which again balances out the hydroxyl addition by the calcium chemistry mechanism was also supported The addition of oxygen to the su rface of an activated carbon and its effect on adsorption is well known and therefore a follow up study by Bach (2004) evaluated the removal of the target compound MIB using the treated carbons. To determine the effect that the treatments designed to cre ate pH stable carbons have on adsorption, a pH stable air treated carbon (673 K, 60 minute treatment), a CO 2 treated carbon, and an untreated carbon were evaluated for their removal of MIB in both deionized ( DI ) water (i.e., organic free) and natural water (containing 10 mg/L natural organic matter (NOM) measured as total organic carbon (TOC) ). The characteristics of the pH stable carbons and an untreated carbon are shown for comparison in Table 6 1 Note that the BET surface area of the ca rbons can be co nsidered similar The percent removal of MIB in DI water versus the powdered carbon dose of a CO 2 treated, air treated, or an untreated carbon evaluated in the work of Bach (2004) is reproduced in Figure 6 1 Figure 6 1 indicates that the CO 2 treated car bon had a higher adsorption capacity for MIB than either the air treated or untreated carbon and the air treated carbon had less of an adsorption

PAGE 68

68 capacity for MIB compared to the un treated carbon. Note that MIB removal is not affected by pH change and MIB does not affect pH (Herzing et al. 1977). The studies of Pendleton et al. (1997) showed t he effects of activated carbon surface chemistry on the adsorption of MIB and linked decreased adsorption with an increase in the hydrophilic nature of an oxidized carbon. As MIB will adsorb to both hydrophobic and hydrophilic adsorption sites, direct competition for adsorption sites between water and MIB, the mechanism suggested by Pendleton and coworkers, would not explain the decreased adsorption of MIB with oxy air treated carbons (Table 6 1 ) and previous research showing the impact of an oxidized surface on adsorption, t he air treated carbon in th e study of Bach (2004) did not remov e as much MIB as the untreated carbon due to the presence of water clusters on the oxygenated surface of the air tr e ated carbon which impeded MIB adsorption. While the impact of air treatment and the associated oxidation of the carbon surface on adsorptio n are well understood, an increase in removal with CO 2 treatment was unexpected given the similar surface acidity of the CO 2 treated carbon compared to the untreated carbon. Bach (2004) proposed that the CO 2 treated carbon delays adsorption of water as th e CO 2 is desorbed from the surface thereby leaving the adsorption sites free for MIB adsorption. A slight increase in the adsorption of water vapor following CO 2 treatment, while indicative of a change in the way carbon behaves with water molecules, may n ot translate to the water based system used in the adsorption study and therefore additional work needs to be done to confirm this hypothesis. NOM is ubiquitous in surface waters and the concentration of NOM can be 100,000 times th at of the trace contamina nt MIB and, furthermore, t he adsorption of NOM has been shown to decrease the removal of MIB via competitive adsorption (Newcombe et al. 2002 ;

PAGE 69

69 Tennant 2004). Therefore, it was of interest to compare the performance of the air treated, CO 2 treated, and u ntreated carbons for MIB removal in natural water which was done in the work of 2 Unlike the DI water scenario, the air treated carbon removed more MIB in natural water than did the untreated carbon. NOM can adsorb through electrostatic forces (as well as non electrostatic forces) wherein the negative charge of the NOM is attracted to a positive charge on the carbon surface (Newcombe 1999). The pH pzc ,the pH at the point of zero cha rge where the surface carries a negligible charge, can be used to determine the charge on a surface (Ravichandran 1998; Tennant 2004). In a solution with a pH above the pH pzc the carbon surface will carry a negative charge and a solution with a pH belo w the pH pzc will yield a positive surface charge. The untreated carbon with a pH pzc of 10.4 (Table 6 1) in contact with a natural water pH of approximately 7 will have a positive surface charge. Given this difference in surface charge, it was hypothesize d that t he untreated carbon would adsorb more NOM and, as there are fewer adsorption sites available after NOM adsorption, less MIB adsorption would occur (i.e., competitive adsorption). Bach (2004) also hypothesized that t he air treated carbon, with a pH pzc of 8 in contact with a natural water of pH 7, will have less of a positive surface charge than the untreated carbon and will not adsorb as much NOM, leaving more adsorption sites available for MIB. However, the NOM remaining in solution was not evalua ted to support the hypothesis of competitive adsorption. Therefore, the intent of the present work is to bolster the hypotheses made in the work of Bach (2004) concerning differences in adsorption performance of the treated carbons. An additional goal is to evaluate the removal of a compound other than MIB, one that will have different chemical properties, to determine if the pH excursion treatments impact the adsorption

PAGE 70

70 of other compounds. To this end, the removal of phenol by the treated carbons in a co lumn study was also be evaluated. Impact of Wetting Time o n MIB Adsorption f or CO 2 Treated Carbon The CO 2 treated carbon was found to remove more MIB (10%) than the untreated carbon at the lower carbon doses (Figure 6 1) despite the similarity in their to tal acidity (Table 6 1 ). A possible explanation for this occurrence is a disparity between the time it takes the CO 2 treated carbon to wet compared to the untreated carbon. To determine if this proposed mechanism is plausible, the adsorption experiments were repeated for the CO 2 treated carbon and the untreated carbon with DI water except that the respective carbons were mixed with DI water for a period of time (i.e., 0, 5, 60, and 150 minutes; wetting time) before the MIB was added, instead of simultaneo us water/MIB addition. The results of this experiment are shown in Figure 6 3 with error bars representing one standard deviation of replicate experiments. As the wetting time increased (i.e., the longer the carbon was in contact with only DI water befor e addition of MIB) the lower the adsorption capabilities of both the CO 2 and the untreated carbon. This is expected, as the carbons will adsorb water which blocks the adsorption sites available for MIB when it is added after 5, 60, or 150 minutes. The ob servation that the adsorption capacities of the CO 2 and untreated carbon are equal after 150 minutes wetting indicates that the adsorption advantage of the CO 2 treated carbon is likely related to its temporarily decreased water adsorption. This phenomenon is expected to hold true even in natural waters, where the CO 2 treated carbon also outperformed the air treated and untreated carbons (Figure 6 2). In separate experiments, CO 2 bubbling of an MIB solution containing the untreated carbon demonstrated no e f fect on MIB removal when the volatilization of MIB from solution was taken into account. Volatilization under these experimental conditions (i.e., bubbling with

PAGE 71

71 CO 2 ) yielded an MIB removal of approximately 50%. Therefore, simply the presence of CO 2 in wa ter and any effect it has on pH did not affect MIB adsorption, a finding supported by Herzing et al. (1977). MIB Adsorption f rom Natural Water An increase in removal of MIB with the air treated carbon at low carbon doses compared to the untreated carbon, as shown in Figure 6 2, was hypothesized to relate to competitive adsorption with NOM. The phenomenon of decreased MIB removal with an increase in NOM removal over a range of pH values is supported in the work of Tennant (2004). To substantiate this hypo thesis in the work of Bach (2004), Table 6 2 shows the results of the evaluation of the concentration of NOM remaining in solution, represented by dissolved organic carbon (DOC), after 24 hr of contact with a 20 ppm dose of carbon and 153 ng/L MIB. The co ncentration of NOM remaining in solution is lowest for the untreated carbon and highest with the air treated and CO2 treated carbon, as anticipated by the theory that NOM is attracted to the positive charge on the untreated carbon. Given that fewer adsorp tion sites are available for MIB once NOM has adsorbed to the surface, it is anticipated that the removal of MIB will decrease as the removal of NOM increases, which is the case when comparing Figure 6 2 to Table 6 2. The CO2 treated carbon outperforms th e air treated carbon, despite the similar removal of NOM, due to the CO2 Phenol Adsorption f rom Deionized Water To determine whether the impact on adsorption performance with treatments to overcome pH excursions is dependant on MIB being the target compound and on batch studies, the performance of an air treated, CO 2 treated, and an untreated commercially available carbon (F400) was evaluated for the removal of phenol in DI water via a column study. The results of

PAGE 72

72 this study are shown in Figure 6 4 where the effluent phenol concentration from the column is shown versus bed volumes (BV) treated. As with the batch studies, the treatments to overcome pH excursions, performed on a commercially available activated carb on in this case, showed a difference in adsorption performance depending on the treatment used. Not surprisingly, given the dependence of phenol adsorption on the oxygenation of the surface (Bansal et al., 2002; Salame and Bandosz 2003 ), the air treated carbon results in fewer bed volumes treated before a measurable concentration of phenol is produced in the effluent. Moreover, the phenol concentration is higher in the effluent 2 and untreated carbons. Again, the impact of the increased surface oxygenation of the air treated carbon, negatively impacts the performance of the air treated carbon for phenol removal due to an Interestingly, Figure 6 4 does not show an increase in the performance of the CO 2 treatment as compared to the untreated carbon as found with MIB removal in batch studies. It is possible that the difference in wetting, which played a role in the batch studies, does not manifest in the column studies due to the difference in adsorption kinetics (e.g., batch versus flow through systems). More likely, the slight impact on adsorption that results from a difference in wetting time (e.g., approximately 10% improved removal o f MIB) in DI water batch studies containing ng/L concentrations of MIB does not manifest in a solution containing mg/L concentrations of the target compound (phenol). In other words, the high concentration of phenol drives the urface via a concentration gradient that overcomes the slight influence of wetting time. Under these conditions, the dominating factor may be the relative

PAGE 73

73 and un treated carbon have similar surface acidity. Conclusions The adsorption capacity for MIB of two GACs that have undergone two separate treatments to overcome pH excursions, namely air and CO 2 treatments, were compared in a study by Bach (2004) The untreat ed carbon removed more MIB than the air treated carbon while this trend was reversed when NOM was present. Increase in surface acidity with the air treated carbon created preferential adsorption sites for water, which out competed MIB in water not contain ing NOM. W hen NOM was present, the negatively charged NOM was attracted to the positively charged surface of the untreated carbon via electrostatic forces, thus out competing MIB for adsorption sites. In addition, the hypothesis that in both water scenar ios the CO 2 treated carbon removed the most MIB due to a longer wetting time was confirmed by analysis of MIB removal versus wetting time Finally, an analysis of the removal of phenol following air and CO 2 treatments on a commercially available carbon sh owed a decrease in performance with air treatment, as found in batch studies with MIB. However, similar performance of the CO 2 and untreated carbons for removal of phenol was found as compared to the improved performance of the CO 2 carbon versus the untre ated carbon in MIB removal. The se results reaffirm that capabilities and, thus, use of these treatments should be based on more than their ability to create pH stable c arbons.

PAGE 74

74 Table 6 1. Characteristics of sample carbons Treatment pH pzc BET surface area ( 5%) Total acidity ( eq/g) Carbon dioxide 8.3 885 170 Air 8.0 927 400 Untreated 10.4 896 160 Table 6 2. DOC remaining in solution af ter contact with 20 ppm carbon Treatment DOC remaining in solution (mg/L) Carbon dioxide 36.8 0.7 Air 34.8 1.0 Untreated 31.0 1.5 Figure 6 1 Percent removal of MIB in DI water versus dose of carbon (24 hr contact time) (Bach, 2004)

PAGE 75

75 Figure 6 2 Percent removal of MIB in raw water (TOC = 10 mg/L) versus dose of carbon (24 hr contact time) (Bach, 2004). Figure 6 3 Percent MIB removal versus wetting time in DI water with a CO 2 and untreated carbons (carbon dose = 2ppm, contact time = 24 hr s).

PAGE 76

76 Figure 6 4 Effluent phenol concentration in DI water versus BV treated with a CO2 treated, air treated, and an untreated carbon.

PAGE 77

77 CHAPTER 7 IMPACT OF DISSOLVED OXYGEN ON ACTIVATION OF POWDERED ACTIVATE D CARBON Despite the millenniums over which a ctivated carbon has been used and the centuries of specific contaminants The lack of tailoring means that activated carbon is often not utilized to its full potential. For example, perchlorate, a chemical found in rocket fuel and increasingly discovered in drinking water supplies, had minimal affinity for activated carbon i n preliminary studies (Chen et al., 2005a). However, subsequent studies using an ammonia tailoring process found that the tailored activated carbon removed four times as much perchlorate as a standard activated carbon, indicating that many studies using s tandard activated carbon may not represent the full capabilities of activated carbon for this particular pollutant (Chen et al., 2005a, 2005b). As new contaminants are found and existing regulations made more stringent, the ability to efficiently tailor activated carbons for removal of specific compounds, by manipulation of pore size and/or surface chemistry, is crucial. Therefore, the objective of this work is to tailor the thermal activation of a wood based carbon by changing the dissolved oxygen (DO) content of the water used to create the steam used for activation. Research by Chestnutt et al. (2007) has determined that altering the DO in the water used to create steam for reactivation of carbon can alter the resulting reactivated carbon but it could be possible to modify the activated carbon from its inception using this method rather than during the reactivation process. By developing a method of activation that requires only a shift in DO content to create activated carbons with differing properti es, capable of adsorbing a variety of contaminants, activated carbon is thus made adaptable to the changing types of and standards for contaminant removal. To this end, the ability of the DO content of the water used to create steam for activation of a wo od based fly ash

PAGE 78

7 8 properties. In addition, the interaction of the DO and typical tailoring parameters of temperature, time, and oxidant flow rate was evaluated. Dissolved Gases in Solution Alteration of the dissolved gases in solution will occur as a result of attempts to change the DO and, therefore, the quantity and possible e ffects of the change in these dissolved gases should be evaluated. The dissolved gas es in solution chosen for evaluation at the individual experimental conditions were those naturally found in the atmosphere at the four highest concentrations, namely, nitrogen, oxygen, argon, and carbon dioxide. The system employed in this work includes a pressure vessel set at 10 psi (gauge) for a total pressure on the system of 24.7 psi (absolute) where the pressuring gas was air, oxygen, or nitrogen. For comparison, the theoretical concentration of dissolved gases in a system at atmospheric pressure w as also calculated. Complete purity of the gas above the systems pressurized in nitrogen and oxygen is likely not a realistic condition but, for the purpose of this exercise, it is assumed that any air remaining in the system will contribute only negligib ly to the final dissolved gas concentrations. The theoretical dissolved gas concentrations are shown in Table 7 1 along with the measured A theoretical DO value of 9. 1 mg/L at atmospheric pressure fit within expected values, while DO levels under pressurization of the system, even in air, were higher than the measured values. A brief review of industrial techniques shows that utilizing a pressure vessel to increase th e DO level is not unique. The purpose of dissolved air flotation (DAF) is to infuse water with a high level of DO so that, when the DO devolves from solution, the oxygen bubbles will lift microscopic particles to the surface which are then removed (Applet on, 2006). DO levels used in DAF technology may be as high as 40 mg/L and may be attained through pressurization A

PAGE 79

79 patent by Barnett (1999) employs a pressure system to imbue water with DO at five times atmospheric pressure, which is indicated as approx imately 50 mg/L, to aid in biological treatment of a wastewater. These findings indicate that the theoretical DO levels found under pressurization are within reason. However, there is a significant discrepancy between the theoretical DO level, predicted u high pressures (Petrucci and Harwood, 1997), research has shown nitrogen from 14.4 psi to approximately 573 psi only varied by 10% and only 8% in a study of nitrogen with oil at pressures of 14.7 psi to 735 psi (Gerth, 1985) Given that the experiments herein employ pressure not even twic e that of atmospheric, it is not considered likely that the equilibrium is a very slow process compared to many other reactions involving dissolved species. Equilibrium in gas/liquid systems depends on the interfacial area available for gas transfer, which explains why bubbling a liquid with gas increases the rate of equilibration as the s mall gas bubbles create higher surface area for water/gas contact compared to gas above a layer of water (Benjamin, 2002). The system employed herein did utilize some form of bubbling, more in the nitrogen system than in others; the system with oxygen eas ily exceeded the limit of detection of the DO probe (i.e., greater than 20 mg/L) with even seconds of oxygen bubbling. Given the ease with which the DO level can be changed by bubbling the water with oxygen for a short amount of time, it is likely this la tter phenomenon is creating the discrepancy between the

PAGE 80

80 equilibrium. The portion of the reaction that progresses over the duration of the experiment leads to the experi mentally determined values which, as seen in Table 7 1, are at most half as high as the theoretical values. Of most interest is the difference between theoretical and experimental values for the oxygen pressurization scenario. For the purposes of this ex system does not need to be at equilibrium as long as the DO level is relatively steady throughout the experiment and uniform between experiments. The DO level thro ughout experiments was easily held within +/ 1 mg/L. At high temperatures, as the water turns from liquid to vapor, the dissolved gases will devolve and will be present at higher than atmospheric concentrations inside the furnace. The use of high tempera tures in inert environments, such as nitrogen and argon, as a means to remove oxygen containing functional groups and create a more basic carbon is common (Leon y Leon et al. 1992; Menndez et al. 1996; Barton et al. 1997; Considine et al. 2001; Lillo Rodenas et al. 2005). However, it is not expected that the small concentration of inert gas (i.e., nitrogen and argon) (e.g., 8*10 6 mol at the highest) versus oxidizing gas (i.e., water vapor and oxygen) (e.g., 0.886 mol H 2 O) is high enough to qualify a s an inert environment where the chemisorption of oxygen would be deterred (See Appendix A for example calculation of moles of dissolved gas versus moles of water vapor). Aside from the possibility of creating an inert environment during the oxidation ste p, the a dsorption of these inert gases during the oxidation step would be negligible as the preceding pyrolysis step would saturate the carbon with as much nitrogen as it could adsorb at that temperature ( Puri, 1970 ). Given that carbon dioxide in solutio n can change the pH of the system and its associated alkalinity, it is of interest to calculate the theoretical shift in pH with an increase in carbon

PAGE 81

81 dioxide in solution. Whereas water exposed to the air at atmospheric pressure would be expected to have a pH of 5.7, based solely on carbon dioxide dissolution in water, water exposed to air at 1.68 times atmospheric pressure would be expected to have a pH of 5.59, only slightly less than that at atmospheric pressure (See Appendix A for calculations). It is not expected that this slight shift in pH will cause any significant change in the activation of carbon with this water used as the basis to make steam. Carbon dioxide is also used as an oxidant in the activation of carbon according to the reactions shown in Chapter 2. However, given that the concentration of carbon dioxide is an order of magnitude less than that of oxygen, which can react with twice as many reactive sites as can carbon dioxide (see Chapter 2), the contribution of carbon dioxide activatio n in all scenario s seems negligible. Given the discussion of the relative impact of nitrogen, argon, and carbon dioxide, it is clear why changes in dissolved oxygen are taken as the dominant influencing factor in these activations. Box Behnken Approach and Mass Loss In order to evaluate the impact of DO and other activation parameters on the resulting activated carbon, the Box Behnken approach, an example of a statistical design of experiments, was employed. The first step in the Box Behnken design pro cess is to determine the levels of each of the four factors to be evaluated (i.e., activation time, activation temperature, steam to carbon ratio (i.e., steam flow), and DO level). Preliminary experiments indicated what levels of the four factors would al low for even spacing of levels, as required by the Box Behnken approach, be practical to employ, and encompass the broadest range of impact on the carbon. A more detailed discussion of the selection of the levels of these parameters, including results of preliminary tests, can be found in Appendix B. The results of these studies yielded the three levels of the four factors shown in Table 7 2.

PAGE 82

82 Next, 29 separate activated carbons were created according to the Box Behnken design of experiments. Of these c arbons, 25 were unique based on the setting of the activation time, activation temperature, SC, and DO level while four samples were replicates of the center point. A complete list of the activation parameters for each of the 29 carbons can be found in Ap pendix C. The initial mass of carbon before activation and the final mass after activation were recorded for each carbon in order to determine the mass lost during activation. The mass loss is a measure of the gasification of the carbon (i.e., the degre e of activation) and was the first response evaluated. The mass loss data were input into the Design Expert software which analyzed the significance of each factor in mass loss and produced a model based on these data to predict the mass loss that would r esult over the entire range of levels and factors studied. For a complete list of mass loss for each carbon and the entire output from Design Expert, see Appendix C. Pertinent statistical information and results from the Box Behnken analysis have been su mmarized in Table 7 3. Note that a Prob > F (P value) greater than 0.05 for lack of fit (i.e., insignificant lack of fit) and a low standard deviation, which includes both the lack of fit and random error, are preferred. The statistically significant fact ors (Prob > F of less than 0.05) in mass loss were found to be temperature and time with significant interactions between temperature and time and also temperature and SC. The final column of Table 7 3 represents the predicted weighting of each parameter where the summation of the activation parameters, multiplied by their weighting factor, and the intercept would yield the predicted mass loss. Note that a negative weighting factor indicates a detrimental impact of this factor on mass loss. Figure 7 1 sh ows a 3 D model produced by Design Expert of the predicted mass loss, in percent of initial mass, versus time and

PAGE 83

83 temperature of activation, both significant factors. The levels of the remaining two factors are set at their medium value The model resul ting from the Box Behnken analysis predicts that as the time and temperature of activation increases so will the mass loss of the carbon. The slight curves in the graph are indicative of an interaction where the two factors acting together have an influen ce greater than the factor acting alone. The statistical significance of the interaction between time and temperature is affirmed by the results of Design Expert as shown in Table 7 3. Indeed, the predicted model is a quadratic model due to the presence of an interaction between temperature and time and also temperature and SC. The influence of temperature and time on mass loss during activation, namely an increase in mass loss with an increase in time and temperature, is an expected result as this phen omenon is well supported in the literature (Gergova et al., 1993; Gonzalez et al., 1994; Rodrguez Reinoso et al., 1995; Gergova et al., 1996; Martin Gullon et al., 1996; Tennant and Mazyck, 2003). An increase in temperature will speed gasification reacti ons, which remove carbon atoms and therefore mass, while an increase in activation time will allow for more gasification. At 623K, the interaction between time and temperature of activation is evident as the increase in mass loss with time is negligible indicating that at this low temperature there is minimal gasification that will occur such that longer treatment times do not impact mass loss. Indeed, an activation temperature of 623 K is low compared to that normally employed in activation (e.g., 1073K ) (Kawamura, 2000). The influence of treatment time and temperature on surface area is confirmed in Table 7 4, where there is an increase in surface area with time and temperature except at 623 K where surface area is unaffected by an increase in treatmen t time.

PAGE 84

84 The interaction between temperature and SC in mass loss can be seen in Figure 7 2 in the curved sides of this 3 D graph. Note, however, that SC alone is not a statistically significant factor influencing mass loss; SC is only significant when inte racting with temperature. An interaction between temperature and SC has been shown in the literature where only at high temperatures does the steam flow influence the gasification of the resulting carbon. Hydrogn is produced as a reaction of the water mo lecule with the carbon surface and also the water gas shift reaction (See Chapter 2) The hydrogen may then react with an active site on the carbon surface which prohibits further reaction of the steam with that active site, thus inhibiting gasification ( Walker et al. 1959; Puri 1970; Menndez et al. 1996; Lussier et al. 1998; Bansal et al. 2002 ) It has been postulated by Rodrguez Reinoso et al. (1995) that at higher steam flow and higher temperatures the inhibiting reactant, hydrogen is less effe ctive Therefore, an increase in steam (i.e., oxidant) will produce greater gasification and mass loss but only at higher temperatures where the creation of the inhibiting agent hydrogen which would increase with an increase in steam flow, does not influ ence gasification. The influence of SC and temperature on surface area is supported in Table 7 4, where there is an increase in surface area with SC except at low temperature. The Design Expert results for mass loss, indicating a positive relationship bet ween time and temperature and also an interaction with time and temperature and with SC and temperature, are an affirmation of the reliability of this approach as these results conform to phenomenon found in the literature concerning activation with steam. DO was not found to be a significant parameter in mass loss during activation. MIB Removal Following creation of the 29 activated carbons, these carbons were then powdered to create a PAC slurry for use in subsequent batch adsorption studies. The first performance

PAGE 85

85 response of interest was the performance of the activated carbons for removal of the common taste and odor causing compound MIB. Preliminary experiments demonstrated that a PAC dose of 10 mg/L in contact with 153 ng/L of MIB for 24 hours woul d yield a suitable range of removal. The MIB removal data were input into the Design Expert software which analyzed the significance of each factor in the removal of MIB and produced a quadratic model to predict the performance of the carbons over the ent ire range of levels studied. For a complete list of the MIB removal of each carbon and the statistical results of Design Expert, see Appendix C. Pertinent statistical information for the MIB removal response is summarized in Table 7 5. All parameters w ere found to be statistically significant factors (Prob > F of less than 0.05) influencing MIB removal, whether acting independently or as an interaction, as expected given that the selection of these factors was based on their influence on adsorption of MIB. While the final column of Table 7 on the predicted MIB removal, Figure 7 3 shows a 3 D model of the predicted removal of MIB versus both time and temperature of activation. As the time an d temperature of activation increases, so does the removal of MIB. Note that the shape of the mass loss and MIB removal curves versus time and temperature are similar at higher temperatures. Figure 7 4 shows MIB removal versus SC and DO, the interaction of which was found to be significant, as well as SC alone, in the removal of MIB. SC acting alone has a significant impact on removal of MIB but the interaction between DO and SC is also significant as demonstrated in Figure 7 4 where the trend in removal changes depending on the level of DO. At low SC, an increase in DO is detrimental while at high SC a change in DO yields negligible change in MIB removal.

PAGE 86

86 The trends in MIB removal with changes in temperature, activation time, SC, and DO relate to chem ical and/or physical changes in the carbon caused by these activation parameters. As such, the following sections explore the chemical and physical properties of the activated carbons in order to elucidate the role of each of the four factors on these pr operties and subsequent adsorption of MIB. Physical Properties The mass loss, which can be taken as a general indicator of the gasification of the carbons, was shown to be impacted by temperature and time. It is well known that an increase in mass loss generally correlates to an increase in both surface area and pore volume as gasification removes carbon atoms from the internal surface of the carbon (Gergova et al., 1994; Centeno et al., 1995; Rodrguez Reinoso et al., 1995; Gergova et al., 1996; Mazyck and Cannon, 2002; Zhang et al., 2003). Therefore, it is expected that the increase in MIB removal with time and temperature, predicted in Figure 7 3, is due in part to the increase in the surface area of the carbons related to temperature and time as foun d in Table 7 4 An improvement in the removal of MIB with time at low temperature, where the physical structure of the carbon remains similar, may be related to changes in surface chemistry as discussed further in later sections. In order to evaluate t he influence of the extreme levels of SC and DO on the physical properties of the carbon, and to determine if physical properties dominate the observed phenomenon of Figure 7 4, four carbons representing the four corners of Figure 7 4 were evaluated for th eir BET surface area and pore size distributions. Descriptions of the conditions of activation for these four carbons as well as their BET surface area and their pore volumes are shown in Table 7 6. These carbons were labeled based on a code where SC is designated S with levels at low (L) or high (H) and the same applies for the DO (D) level.

PAGE 87

87 Within standard error for BET surface area measurements, only SHDL and SLDH have different surface area. A similar BET surface area but an increase in mass loss as predicted with an increase in steam flow (Figure 7 2) indicates that the changes in steam flow yield different pore size distributions making up this surface area. A summary of the pore volume is presented in Table 7 6 with the PSD plots shown in Figure 7 5. Evaluation of the PSDs of these four carbons shows a predominance of micropores with a small volume of mesopores and macropores. In a study by Tennant and Mazyck (2003) for MIB removal using PAC, carbons having a large volume of large micropores and small mesopores (1.2 10 nm) were found to correlate to the highest removal of MIB. Therefore, this range is also presented in Table 7 6 as this gives the range of pore sizes that are expected to most significantly affect MIB removal. From Table 7 6, it is apparent that an increase in steam flow (SHDL versus SLDL or SHDH versus SLDH) will produce a higher volume of pores in the range of 1.2 10 nm. An increase in the PSD with an increase in steam flow is anticipated based on the correlation between increa se in steam flow and mass loss, where a higher concentration of oxidant yields more gasification. However, the similarity in SHDH and SLDL indicates that a decrease in steam flow can still produce a similar pore volume to a higher steam flow if only the D O content were decreased which would have advantages in terms of cost savings of having to pump a lower flow of steam to get the same result. SLDH, having a low steam flow and also high DO, has the lowest volume of pores in the 1.2 10 nm range. The in fluence of DO on the pore size distribution, namely the occurrence of a higher volume of 1 2 10 nm with a lower DO may relate to a study of the activation of carbon with oxygen by Tomko w et al. (1977). These authors suggested that only limited micropores are

PAGE 88

88 formed during activation with oxygen due to the formation of oxygen containing functional groups at the entrance of the micropores which cuts off the micropores from further activation as the oxidant cannot fit into the narrowed entrance created by the functional groups. In the current research, where the DO in the water is expected to devolve upon conversion of the water into steam, the small quantities of free oxygen may react with the carbon surface to create oxygen containing functional groups whic h would impede further activation by steam. The influence of DO on the surface chemistry of the carbon will be discussed in the following section. As the volume of 1.2 10 nm pores are expected to correlate to MIB adsorption, it is anticipated that SLDH, with the lowest volume in the ideal range of pore sizes, would remove the least MIB. In the predicted MIB removal of Figure 7 4 this theory is confirmed in that SLDH is the only carbon to have a significantly different removal of MIB. The similarity in the ideal range volume of SLDL and SHDH explains their similar removal of MIB but one may expect that SHDL would have higher removal as it has a higher volume of ideal pores compared to the other carbons. It is most likely that the small increase in this ideal pore range volume from SLDL and SHDL (i.e., 13%) compared to the large increase in this pore range from SLDH and SHDH (i.e., 39%) is not significant enough to manifest differences in performance outside of the standard deviation. It is clear that the DO content of the water used to create steam for activation of these carbons influenced the physical structures of these carbons but the influence of the DO on the chemical properties of the carbon must also be investigated. Surface Chemistry Following creation of the 29 activated carbons, the samples were then analyzed for their pH at the point of zero charge (pH pzc ). The pH pzc surface acidity where a lower pH pzc indicates a more acidic surface (El Sayed and B andosz,

PAGE 89

89 2003; Ania et al., 2004). Furthermore, activated carbon acidity can be related to the content of the oxygen containing functional groups on the carbon surface which are often acidic ( Puri 1970 ; Otake and Jenkins 1993 ; Barton 1997 ; Ania et al. 2004 ) The pH pzc data was input into the Design Expert software which analyzed the significance of each factor in the resulting pH pzc of the carbons and produced a linear model to predict the pH pzc of the carbons over the entire range of levels studied. For a complete list of the pH pzc values of each carbon and the statistical results of the Design Expert process, see Appendix C. Pertinent statistical information for the pH pzc response is shown in Table 7 7. A linear model best fit the data, which mean s that no interactions were statistically significant and only temperature and SC were found to be statistically significant factors impacting the pH pzc Figure 7 6 shows a 3 D model of the predicted pH pzc versus both SC and temperature of activation. As demonstrated in Figure 7 6, as both the SC and temperature of activation increase so does the pH pzc (i.e., the activated carbon becomes more basic/less acidic). The impact of temperature on pH pzc relates to the decomposition of acidic oxygen containing f unctional groups starting at 773K R emov al of oxygen containing functional groups to create a more basic carbon through the use of high temperatures in inert environments is a common practice (Leon y Leon et al., 1992; Menndez et al., 1996; Barton et al. 1997; Considine et al., 2001; Lillo Rodenas et al., 2005) While inert condition s at high temperatures were not deliberately employed for this purpose in the current research, following activation the reactor wa s allowed to cool to room temperature in a flow of nitrogen gas before the carbon sample wa s removed. During the cooling of the sample from as high as 973K to less than 373K the carbon is briefly exposed to high temperatures in an inert environment which may strip the surface of the functional g roups

PAGE 90

90 formed during the activation process A less acidic carbon with increasing activation temperature supports the finding of Figure 7 3, where improved MIB removal was predicted as temperature of activation increases; a less acidic surface would attrac t fewer water molecules leavings the adsorption sites free to remove MIB ( Pendleton et al. 1997 ; Considine et al. 200 1) Though the pH pzc is predicted to be constant regardless of activation time, as activation time was not found to be a significant fac tor in pH pzc there may be changes in the quantity of basic groups with time that are not evident in this study of surface chemistry but manifest in improved removal of MIB with activation time at even the lowest temperature employed (Figure 7 3). The occu rrence of an increase in steam flow relating to an increase in pH pzc (i.e., an increase in basicity) may be explained by an associated increase in the formation of hydrogen as the steam flow is increased. Recall that hydrogen is created as a result of the gas shift reaction. The free hydrogen molecule would then react with active sites to form hydrogen containing functional groups (Walker et al., 1996). These hydrogen containing function al groups would occur on sites that could have contained an oxygen containing functional group, as Bansal et al. (1974) determined that both oxygen and hydrogen adsorb to the same reactive sites, so that the formation of an acidic oxygen containing functio nal group is avoided. Aside from possible direct competition of the hydrogen and oxidant for adsorption sites, the hydrogen also blocks active sites from further reactions with atmospheric oxygen to create functional groups following activation. The pro cess of preventing oxygen chemisorption by preadsorption with hydrogen has been proposed by Bansal et al. (1974) and confirmed by Mendendez et al. (1996) where hydrogen exposure was used to create a stable basic carbon.

PAGE 91

91 As a less acidic surface is favo rable for the removal of MIB, an increase in steam flow, the creation of hydrogen, and the corresponding increase in pH pzc would be expected to correspond to improved MIB removal. In support of the link between steam, hydrogen, and MIB removal, Nowack et al. (2004) observed similar performance for removal of MIB in a steam treated carbon compared to a carbon treated in hydrogen at a similar temperature. These authors took the similar performance of the steam and hydrogen treated carbons as support of the influence of the water hydrogen with the carbon surface on improved removal of MIB. The influence of steam flow on surface chemistry supports the results of Figure 7 4 which shows that a t high DO there is an improved removal of MIB with increasing SC. The expected increase in performance as SC is increased is only significant at high DO, at low DO the performance of the carbons remains the same regardless of steam flow. It is possible t hat at high DO, the competition provided by hydrogen is more significant as there are more oxidants to compete with (i.e., both steam and O 2 ). Without this competition from hydrogen, the higher oxygen in contact with the carbon, as the DO devolves from th e water upon creation of steam, results in the creation of more acidic functional groups and therefore a more acidic carbon. At low DO, hydrogen is competing with a smaller quantity of oxygen and the resulting impact of competition with hydrogen is less s ignificant and produces negligible results on total acidity. However, the DO level was not found to significantly impact the total acidity of the carbon, which could indicate that, while the impact on the total number of oxygen containing functional group s remains the same regardless of DO, the type of individual oxygen containing functional groups changes with the DO level. A change in the dominant type of oxygen containing functional group would not necessarily change the total acidity.

PAGE 92

92 Activation wi th oxygen would produce active sites with different energy levels than those produced by steam reactions and these active sites may react to form different functional groups when exposed to the atmosphere. Puri (1970) notes that the activity of the surfac e may determine the type of group formed, with more active surfaces producing lactones while less active sites produce carbonyls. Otake and Jenkins (1993) and Molina Sabio et al. (1996) also make note of differences in the thermal stability of surface oxi des depending on the method of oxidation (air versus HNO 3 activation and steam versus CO 2 activation, respectively). The findings of these studies, along with the knowledge of the relative stability of functional groups created by oxygen and steam reactio ns, where the exothermic nature of the C O 2 adsorption step makes the resulting functional groups short lived, support a hypothesis that different oxidants can yield different types of functional groups. Therefore, by passifying the surface of the carbon in low DO versus high DO, the formation of different functional groups may be avoided though the total acidity remains unchanged. In other words, if oxygen reacts with the surface in the high DO conditions, different types of individual functional groups may result than if mostly steam reacted with the surface, as in the low DO conditions. A preliminary analysis of the precursor and activated carbons demonstrated that these carbons are highly basic, as indicated in Figure 7 6, which makes their analysis vi a the Boehm titration method, a common wet method of evaluating individual acidic functional groups, difficult. In a study by El Sayed and Bandosz (2003) of individual functional groups using the Boehm titration method, when the total basicity was almost 3 times the measured total acidity the Boehm titration method measured negligible quantities of individual acidic functional groups. Therefore, t o evaluate whether surface chemistry is a factor in the observed changes in Figure 7 4, the same four carbons representing the four corners of Figure 7 4 and described in Table 7 6

PAGE 93

93 were evaluated for their removal of MIB in dichloromethane, a nonaqueous solvent. The goal of this experiment is to determine whether MIB removal changes depending on the activation c has been removed. The MIB removal of these four carbons in the dichloromethane solvent, where the carbon dose was increased to 1100 mg/L, is shown in Table 7 8. The removal of MIB in dichloromethane is low compared to the removal of MIB in water. Considine et al. (2003) also evaluated the removal of MIB from dichloromethane using activated carbon and the decrease in removal of MIB in dichloromethane was ex plained by the difference in the polarity of the solvents and the way the solvent interacts with the carbon surface. Within standard error, the removals of MIB in dichloromethane with the four corner point carbons are similar. The similar removal support s the theory that a difference in surface chemistry is partly responsible for the difference in removal of the SLDH carbon, as when the water molecules and the hypothesized interaction of the water molecules with the acidic functional groups are removed fr om the solution SLDH performs just as well as the other three carbons. Chemical versus Physical Methods of Removing DO Removal of DO can be accomplished either by physical means (e.g., bubbling with nitrogen gas or passing the water through membranes) or c hemical means (e.g., addition of an oxygen scavenger). For use in boiler streams, it has been found that physical methods do not allow for removal of all of the dissolved oxygen and therefore the streams are often polished with an oxygen scavenger (Bizot and Bail e y 1997). Sodium sulfite is a common oxygen scavenger employed in the boiler industry because it is relatively cheap, safe, and effective at removing DO As this study has so far focused on the removal of oxygen through physical means, it was al so of interest to evaluate a chemical means of removing dissolved oxygen to

PAGE 94

94 determine if the results are similar. To this end, the SHDL and SLDL activation scenarios described in Table 7 6 were repeated but with sodium sulfite used to decrease the DO. T he carbons which mimic the DO levels used in the Box Behnken approach (i.e., 3 mg/L DO) are labeled SHDL s and SLDL s where the subscript s is used to denote that these conditions were created as a result of sodium sulfite addition. In addition, two carbon s were created under the same conditions as SHDL and SLDL except that the DO was reduced to 0 mg/L, which is easily done with sodium sulfite compared to physical removal of DO which cannot remove all DO. The 0 mg/L DO carbons are labeled SHDL o and SLDL o The MIB removal and the chemical and physical properties of these four activated carbons are shown in Table 7 9. Both sets of carbon, comparing low and high SC at a given DO concentration, have similar BET surface area and similar trends in pH pzc namely a decrease in pH pzc at low steam flow. The removal of MIB with the SHDL s and SLDL s also showed similar trends to the removal of MIB with SHDL and SLDL (approximately 62% MIB removal) with no significant change in removal with a change in SC. The MIB rem oval with SHDL o and SLDL o however, showed a decrease in removal with SHDL o compared to SLDL o The difference in removal with the 0 mg/L DO activations may relate to the difference in their ideal pore size volume (1.2 10 nm). The PSD plots for these carb ons are shown in Figure 7 7 Of the four carbons created with sodium sulfite as an oxygen scavenger, only SHDL o had a significantly different pore volume in the ideal range of pores than the other three carbons. The decrease in the volume of this ideal p ore range explains the decrease in MIB removal with this carbon It may be that at this lower DO, where more sodium sulfite was added to reach 0 mg/L of DO, and a higher SC, where a higher volume of water vapor containing sodium sulfite

PAGE 95

95 contacted the carb on, more sodium sulfite interacts with the carbon surface, thus impacting adsorption. Phenol Removal The same PAC slurries created from the 29 activated carbons of the Box Behnken where used in a phenol batch adsorption study. The performance of the ca rbons for removal of phenol bonding) versus MIB adsorption. In addition, phenol is a common toxic organic compound in surface waters for which activated c arbon is used. Preliminary experiments demonstrated that a PAC dose of 50 mg/L in contact with 5 mg/L of phenol for 24 hours would yield a range of removal that was easily detectable. The phenol removal data was input into Design Expert software which an alyzed the significance of each factor in the removal of phenol and produced a linear model, indicating a lack of statistically significant interactions between factors, to predict the performance of the carbons over the entire range of levels studied. Fo r a complete list of the phenol removal of each carbon and the statistical results of the Design Expert process, see Appendix C. Pertinent statistical information for the phenol removal response is shown in Table 7 10. Only temperature and time were fou nd to be statistically significant factors influencing phenol removal. Figure 7 8 shows a 3 D model of the predicted removal of phenol versus both time and temperature of activation. As with MIB removal, where both time and temperature were found to be s ignificant factors, it is likely that increases in surface area and/or pore size distribution associated with increased temperature of activation and treatment time simply yielded more adsorption sites for phenol removal. In addition, the decrease in tot al surface acidity found with increasing temperature may also play a role in the increased removal of phenol with temperature as an

PAGE 96

96 increase in surface acidity has also been found to be detrimental to phenol adsorption ( Bansal et al. 2002 ; Ania et al. 20 04 ). Neither steam flow nor DO was found to be a significant factor in phenol removal. It is possible that the influence of surface chemistry and physical changes brought about by DO were overcome by the higher concentrations of phenol employed herein s uch that the concentration gradient drove the phenol onto the carbon surface despite the obstacles Furthermore, as it was suggested that the different functional groups may have been created with different levels of DO and this is what impacts MIB remova l, the particular acidic functional groups formed in the low SC and high DO carbons may not impact phenol adsorption as it impacts MIB adsorption. For example, in a study by Salame and Bandosz (2003) the decreased removal of phenol by wood based activated carbons was shown to have a stronger correlation to an increase in the quantity of carboxylic groups than to total surface acidity. Another possibility is that the wood based activated carbon used herein responds differently than coal based carbons, as t he difference in coal and wood based carbons has been discussed. Specifically, one adsorption mechanism for electrons on the carbon surface interact. However, wood and wood based precursors are seen to have higher H/C elemental ratios, which suggest a lower aromatic structure, than their coal based counterparts (Buczek et al., 1993; Gergove et al., 1993; Pendleton et al., 1997; El Sayed and Bandosz, 2003; Fernades et al., 2003). Therefore, the removal of phenol with the wood based activated carbon used herein may not be impacted by changes in surface chemistry as it may Summary and Conclusions The Box Behnk en approach was employed to evaluate the influence of four factors on the physical and chemical properties and performance of the resulting activated carbons. An

PAGE 97

97 initial evaluation of mass loss following activation using the Box Behnken design confirmed t he validity of this approach as an increase in mass loss with activation time and activation temperatures as well as an interaction between activation temperature and SC are phenomenon supported in the literature. The Box Behnken approach also predicted an increase in MIB and phenol removal with activation time and temperature. An increase in the physical properties of the carbon with activation time and temperature, supported by an observed increase in mass loss under these conditions, as well as a decr ease in surface acidity with an increase in activation temperature were deemed responsible for an increase in removal of these compounds. An interaction between SC and DO was also seen with the removal of MIB where only at high DO and low SC did these fa ctors impact MIB removal in the form of decreased removal. Neither DO nor SC was found to be significant in the removal of phenol. The physical properties of the carbon, namely a decrease in the PSDs of the carbons, explained a decrease in removal at the highest DO and lowest SC level. However, a low SC level was also found to correspond to a more acidic carbon which can impact MIB removal. While the total acidity of the carbons treated at low SC remained unchanged regardless of the DO level, the high D O activated carbon had decreased removal of MIB. It was proposed that different groups are formed at high DO, where the free oxygen devolved from the water upon creation of steam, while the total acidity remains unchanged. A lack of difference in removal found in dichloromethane, a nonpolar solvent, supports that the difference in removal found with low SC and high DO may be a combination of both physical and surface chemistry changes, though this theory would need to be confirmed with new techniques to e valuate surface chemistry on the heterogeneous and basic activated carbons used herein.

PAGE 98

98 An additional study of DO removal with sodium sulfite indicated that high levels of sodium sulfite may be detrimental to MIB removal, though the amount required to de crease DO to 3 mg/L did not impact performance. While the impact of DO on surface chemistry may require further study, it is clear that the DO level in the water used to create steam for activation of carbon does influence the to adsorb certain compounds. Table 7 1. Concentration of dissolved gases in water at 273 K Gas (mg/L) Air (14.7 psi) Air (24.7 psi) Oxygen (24.7 psi) Nitrogen (24.7 psi) Nitrogen 14.8 25.2 0 32.3 Oxygen 9.1 15.5 73.6 0 Oxygen (measured) 9.0 7 8 12 13 2 3 Argon 0.6 0.9 0 0 Carbon Dioxide 0.5 0.9 0 0 Table 7 2. Levels of four factors employed in the Box Behnken design Factor Low Medium High Temperature (K) 673 823 973 Time (min.) 10 30 50 Steam to carbon ratio (SC) (g H 2 O/g c arbon) 0.4 1.4 2.4 DO (mg/L) 3 7.5 12 Table 7 3. Box Behnken statistical results for mass loss Parameter F value Prob > F value Coefficients of Equation Value Lack of fit (F value) 2.90 0.156 DO (mg/L) 0.82 0.37 7 0.91 Temperature (K) 185 .0 < 0.0001 0.14 Time (min. ) 7.54 0.0125 0.31 SC (g H 2 O/ g carbon ) 3.78 0.0662 7.39 Temperature 2 39.6 < 0.0001 0.00013 DO Temperature 3.63 0.0713 0.0018 Temperature Time 9.53 0.0 058 0.00065 Temperature SC 12.1 0.0023 0.015 Intercept 45.79 Standard deviation 1.27 R squared 0.93

PAGE 99

99 Table 7 4. Surface area versus treatment time, treatment temperature, and SC (DO = 7.5 mg/L) Temperature (K) Time (min.) SC (g H 2 O/g carbon) BET Surface Area (m 2 /g) ( 5%) 973 10 1.4 541 973 50 1.4 615 673 10 1.4 470 673 50 1.4 498 673 30 0.4 468 673 30 2.4 496 973 30 0.4 537 973 30 2.4 606 Table 7 5. Box Behnken statistical results for MIB removal Parameter F value Prob > F value Coefficients of Equation Value Lack of fit (F value) 2.44 0.20 09 DO (mg/L) 4.12 0.0559 4.76 Temperature (K) 73.6 < 0.0001 0.038 Time (min. ) 39. 2 < 0.0001 1.12 SC (g H 2 O/ g carbon ) 5.16 0.0344 10.33 DO 2 2.25 0.1491 0.2 Temperature 2 14.5 0.0011 0.00046 DO Time 2.70 0.1158 0.064 DO SC 6.47 0.0193 1.99 Intercept 126.7 Standard deviation 7.0 R squared 0.88 Table 7 6. BET surface area and total pore vol ume of four carbons Activated carbon label SC (g H 2 O/ g carbon) DO (mg/L) BET s urface a rea ( 5 % ) Micro pore v olume (< 2 nm) (cc/g) Meso pore v olume (2 50 nm) (cc/g) Macro pore v olume (>50 nm) (cc/g) Pore volume (1.2 10 nm) (cc/g) SHDL 2.4 3 501 0.232 0.021 0.003 0.0371 SLDL 0.4 3 461 0.217 0.021 0.006 0.0327 SHDH 2.4 12 478 0.221 0.022 0.006 0.0356 SLDH 0.4 12 430 0.202 0.020 0.006 0.0257

PAGE 100

100 Table 7 7. Box Behnken statistical results for pH pzc Parameter F value Prob > F value Coeffic ients of Equation Value Lack of fit (F value) 0.34 0.9544 DO (mg/L) 0.83 0.3699 0.017 Temperature (K) 143.5 < 0.0001 0.0066 Time (min. ) 2.32 0.1409 0.0063 SC (g H 2 O/ g carbon ) 13. 4 0.0013 0.30 Intercept 7.0 St andard deviation 0.28 R squared 0.87 Table 7 8. Activation conditions of four carbons and their removal of MIB in dichloromethane (time = 30 min., temperature = 873K) Activated carbon label MIB r emoval (%) ( 1.6%) SHDL 5.4 SLDL 5.8 SHDH 6.2 SLDH 5.8 Table 7 9. BET surface area and total pore volume carbons treated with sodium sulfite Activated carbon label MIB r emoval (%) ( 2.5%) pH pzc ( 0.1) BET s urface a rea ( m 2 /g ) ( 5 % ) Micro pore v olume (< 2 nm) (cc/g) Meso pore v olume (2 50 nm) ( cc/g) Macro pore v olume (>50 nm) (cc/g) Pore volume (1.2 10 nm) (cc/g) SHDL s 64 11.67 506 0.227 0.025 0.006 0.0381 SLDL s 59 11.31 484 0.230 0.021 0.006 0.0361 SHDL 0 49 11.63 469 0.225 0.019 0.006 0.0305 SLDL 0 57 11.37 494 0.236 0.019 0.006 0.0355 Ta ble 7 10. Box Behnken statistical results for phenol removal Parameter F value Prob > F value Coefficients of Equation Value Lack of fit (F value) 2.64 0.1791 DO (mg/L) 0.21 0.6517 0.10 Temperature (K) 47.8 < 0.0001 0.046 Time (min. ) 4.92 0.0362 0.11 SC (g H 2 O/ g carbon ) 1.47 0.2368 1.22 Intercept 29. 8 Standard deviation 3.47 R squared 0.69

PAGE 101

101 Figure 7 1. Mass loss versus time and temperature (SC = 1.4 g H 2 O/g carbon, DO = 7.5 mg/L) Figure 7 2. Mass los s versus SC and temperature (time = 30 min., DO = 7 mg/L)

PAGE 102

102 Figure 7 3. MIB removal versus time and temperature (SC = 1.4 g H 2 O/g carbon, DO = 7.5 mg/L) Figure 7 4 MIB removal versus SC and DO (time = 30 min., temperature = 873K)

PAGE 103

103 Fig ure 7 5. Pore size distributions of sample carbons Figure 7 6. pH pzc versus SC and temperature (time = 30 min., DO = 7 mg/L)

PAGE 104

104 Figure 7 7. PSDs of carbons treated with sodium sulfite Figure 7 8. Phenol removal versus time and temperature (SC = 1.4 g H 2 O/g carbon, DO = 7 mg/L)

PAGE 105

105 CHAPTER 8 CONTRIBUTIONS TO SCI ENCE First to determine that activated carbon treated with CO 2 yields improved removal of MIB due to a delay in the wetting of the carbon surface; Supported the importance of investigating individual acidic functional groups on the surface of activated carbon as not all of the individual acidic functional groups measured were altered followin g air treatment and yet the water contact pH and the air performance in the removal of MIB and NOM was altered following air treatment; First to demonstrate that variations in the DO content of water used to create steam for activation of a wood based carbon alters the subsequent performance of the activated carbon for removal of MIB through a combination of physical and chemical changes in the carbon surface; Highlighted the necessity to tailor activated carbon for a specific contaminant as the performance of the activated carbons for removal of phenol, unlike MIB, was unaffected by alterations in the DO content of the water used to create steam for activation.

PAGE 106

106 CHAPTER 9 FUTURE WORK Explore the impact of wetting time of activated ca rbon, as it pertains to the slurries by which PAC and GAC are transported, on the resulting adsorption performance of the activated carbon; Investigate the influence of the DO content of water used to create steam for activation of carbon on the subsequen t adsorption performance of the activated carbon for removal of MIB and phenol in natural water (e.g., water containing NOM); Determine a method to either evaluate the quantity and type of individual acidic oxygen containing functional groups on the surf ace of highly basic, heterogeneous carbons or evaluate the influence of the DO content of water used to create steam on the surface chemistry of a more homogenous surface (e.g., silica gel).

PAGE 107

107 APPENDIX A SAMPLE CALCULATIONS FOR DISSOLVED GASES Table A Bunsen coefficient* Air (14.7 psi) Nitrogen 0.01559 7.80E 01 Oxygen 0.03109 2.10E 01 Argon 0.03412 9.00E 03 Carbon Dioxide 0.8981 3.15E 04 er (2006) Sample calculation of dissolved gas concentration in solution Scenario: Air, 24.7 psi = 1277 torr where P O2 = partial pressure B O2 = Bunsen coefficient, and K = molecular weight/molecular volume B O2 = 0.03109 P O2 = (1277torr 17.52 torr) = 264.5 torr O 2 / torr = 15.5 mg/L Sample c alculation of change in pH with change in car bon dioxide in solution Scenario: Atmospheric pressure Bunsen Method: B C O 2 = 0.8981 P O2 = (760torr 17.52 torr) = 0.223 torr O 2 / torr = 0.376 mg/L

PAGE 108

108 Charge Balance: [H + ] = [HCO 3 ] + [CO 3 2 ] + [OH ] assume [CO 3 2 ] and [OH ] are negligible in this scenario Scenario: 1.68 times atmospheric pressure Bunsen Method: B Co2 = 0.8981 mg/L/torr P O2 = (1277torr 17.52 torr) = 0.377 torr O 2 / torr = 0.6381 mg/L Charge Balance: [H + ] = [HCO 3 ] + 2[CO 3 2 ] + [OH ] assume [CO 3 2 ] and [OH ] are negligible in this scenario Scenario: Atmospheric pressure plus calcium chemistry Bunsen Method: B C O 2 = 0.8981

PAGE 109

109 P O2 = (760torr 17.52 torr) = 0.223 torr O 2 / torr = 0.376 mg/L Charge Balance: [H + ] + 2[Ca 2+ ] = [HCO 3 ] + 2[CO 3 2 ] + [OH ] assume [CO 3 2 ] and [OH ] are negligible in this scenario an d the calcium in solution (as measured) is 2.5 mg/L or 8*10 4 M. Moles of dissolved gases versus moles of steam and sample c alculation Table A 2. Moles of dissolved gases and water vapor versus treatment scenario Level of O 2 (mol O 2 ) SC (g H 2 O/ g carbon) mol H 2 O 3 mg/L 7.5 mg/L 12 mg/L 0.4 0.111 1.88*10 7 4.69*10 7 7.50*10 7 1.4 0.389 6.56*10 7 1.64*10 6 2.63*10 6 2.4 0.667 1.13*10 6 2.81*10 6 4.50*10 6 Scenario: 1.4 g H 2 O/g carbon and 3 mg/L oxygen 1.4 g H2O/ g carbon 5 g carbon = 7g H 2 O 7 g H 2 O / (18 g/mol) (molecular weight of H 2 O) = 0.389 moles of H 2 O 7 g H 2 O / (1000 g/L) (density of water) = 0.007 L H 2 O 3 mg/L Oxygen 0.007 L / (32000 mg/mol) = 6.56*10 7 moles of oxygen

PAGE 110

110 APPENDIX B PRELIMINARY STUDIES FOR DETERMINATION OF LEVELS FOR BOX BEHNK EN To begin any experiment, one must first determine the levels (or settings) of the variables to be evaluated. In this study, the variables of activation time, activation temperature, steam to carbon ratio (SC), and the dissolved oxygen (DO) content in the water used to create steam were evaluated to determine the impact of these parameters on the adsorption performance of the resulting carbons for MIB and also the impact of these parameters on both the chemical and physical properties of the resulting carbons. The following section discusses how the levels were determined for the Box Behnken approach. Note that no attempt is made to explain observed phenomenon as that is outside the scope of t his section. For the latter variable, DO content, it was decided to use the same demarcations used in a previous study on the impact of DO content on reactivation of carbon (Chesnutt et al., 2007). In a study by Chestnutt et al. (2007), the DO contents were characterized as being low (less than 4 mg/L), medium (between 7 8 mg/L), and high (greater than 10 mg/L). In order to use the Box Behnken approach, there must be three levels for each variable and these levels must be evenly spaced. Therefore, the DO content levels in this study were set at 3 mg/L, 7.5 mg/L, and 12 mg/L and where achieved by pressuring the water vessel with either nitrogen, air, or oxygen gas. Given that activations are typically carried out at high temperatures (1173 K), prelimina ry experiments were performed at 1023 K and also a lower temperature, as low temperatures are known to manifest differences in surface chemistry, of 823 K. The activation time, 30 min., and SC, 0.68 g H 2 O/g carbon, for these studies were based on steam a ctivations used in a previous study of steam activation by Tennant and Mazyck ( 2003). The results of the first experiment varying both temperature and DO and evaluating MIB removal in batch studies are shown in Figure B 1.

PAGE 111

111 A ctivations at 1023 K did not manifest a differen ce in MIB removal but the activations at 823K demonstrated that higher DO in the water used to create steam for activation does alter the carbons performance for MIB removal It was also noted that in the study of Chesnutt et al. (2007 ) a very low temperature (648 K) was employed which manifested significant difference in MIB removal with altered DO. Therefore, in order to encompass the range of results where DO impacts MIB and conditions where it does not, the temperatures employed fo r the Box Behnken approach were 673 K, 823 K, and 973 K. Next, to shorten activation time and thereby cut experiment time in half, an activation time of 15 minutes was evaluated at the same steam flow as Figure B 1 (0.68 SC) and an activation temperature of 823 K. The performance of these carbons for removal of MIB is shown in Figure B 2. From Figure B 2, it is apparent that there is a change in MIB removal depending on the activation time employed. As 15 and 30 minutes of treatment create different re movals of MIB, and in order to create evenly spaces treatment times that represent extreme conditions, treatment times of 10, 30, and 50 minutes were employed in the Box Behnken study. Finally, an analysis of flow was performed where the steam flow was rai sed to the highest possible in the system (2.84 g H2O/g carbon) and compared to the previously employed flow based on Tennant and Mazyck (2003). The performance of these carbons for removal of MIB is shown in Figure B 3. There is a change in MIB removal depending on the SC employed. As 0.68 g H 2 O/g carbon and 2.84 g H 2 O/g carbon create different removals of MIB, and in order to create evenly spaces treatment times that represent extreme conditions, SC of 0.4 g H 2 O/g carbon, 1.4 g H 2 O/g carbon, and 2.4 g H 2 O/g carbon were employed in the Box Behnken study.

PAGE 112

112 These preliminary studies highlight the importance of studying interactions between factors, as clearly DO is not an influencing factor under all conditions and this underlines why the Box Behnken statis tical design of experiments was employed in this study. Figure B 1. MIB removal versus activation temperature and DO (SC = 0.68 g H 2 O/g carbon, activation time = 30 min.). Figure B 2. MIB Removal versus time and DO (temperature = 823 K, SC = 0.68 g H 2 O/g carbon).

PAGE 113

113 Figure B 3. MIB removal versus steam flow and DO (temperature = 823 K, time = 30 min.).

PAGE 114

114 APPENDIX C BOX BEHNKEN RESULTS Table C 1. Box Behnken carbons activation conditions, MIB removal, mass loss, pH pzc phenol removal Factor A Factor B Factor C Factor D Response 1 Response 2 Response 3 Response 4 Std. Dissolved oxygen (mg/L) Temp erature (K) Time (min.) Steam/ Carbon (g H 2 O/g carbon) MIB r emoval (%) Mass l oss (%) p zc (pH) Phenol r emoval (%) 1 3 673 30 1.4 56.5 8 9.9 56.8 2 12 67 3 30 1.4 42.4 9 10.0 49.6 3 3 973 30 1.4 94.7 21 11.8 65.9 4 12 973 30 1.4 66.9 17 12.1 67.4 5 7.5 823 10 0.4 37.1 11 10.5 52.6 6 7.5 823 50 0.4 65.7 10 11.0 59.4 7 7.5 823 10 2.4 42.4 10 11.0 54.8 8 7.5 823 50 2.4 62.9 12 11.4 62.7 9 3 823 30 0.4 72.6 11 11.0 63.2 10 12 823 30 0.4 49.2 11 10.4 61.6 11 3 823 30 2.4 56.8 11 11.5 62.4 12 12 823 30 2.4 69.3 11 11.7 63.0 13 7.5 673 10 1.4 30.5 10 9.9 45.9 14 7.5 973 10 1.4 75.4 15 12.0 68.1 15 7.5 673 50 1.4 58.7 10 10.1 54. 4 16 7.5 973 50 1.4 95.9 23 11.9 68.8 17 3 823 10 1.4 37.3 10 11.2 53.6 18 12 823 10 1.4 50.5 10 10.9 55.6 19 3 823 50 1.4 76.5 12 11.6 56.4 20 12 823 50 1.4 66.5 11 11.0 55.6 21 7.5 673 30 0.4 38.5 8 9.7 50.7 22 7.5 973 30 0.4 71.1 14 12.0 64.8 23 7.5 673 30 2.4 63.0 8 10.5 59.2 24 7.5 973 30 2.4 95.3 22 12.1 64.8 25 7.5 823 30 1.4 55.3 9 11.2 63.2 26 7.5 823 30 1.4 57.0 11 10.7 59.0 27 7.5 823 30 1.4 56.2 11 11.2 59.0 28 7.5 823 30 1.4 45.2 11 11.8 63.6 29 7.5 823 30 1.4 52.4 11 11.6 62.3

PAGE 115

115 Table C 2. Analysis of variance (ANOVA) results for mass loss (Prob > F of less than 0.05 are significant factors) Source Sum of s quares DF Mean s quare F v alue Prob > F Model 422.83 8 52.85 32.80 < 0.0001 signif icant A 1.32 1 1.32 0.82 0.3769 B 298.75 1 298.75 185.42 < 0.0001 significant C 12.15 1 12.15 7.54 0.0125 significant D 6.08 1 6.08 3.78 0.0662 B 2 63.80 1 63.80 39.60 < 0.0001 significant AB 5.84 1 5.84 3.63 0.0713 BC 15.35 1 15.35 9.53 0.0058 significant BD 19.55 1 19.55 12.13 0.0023 significant Residual 32.22 20 1.61 Lack of f it 29.66 16 1.85 2.90 0.1565 not significant P ure e rror 2.56 4 0.64 Cor t otal 455.05 28 Table C 3. Statistical data for mass loss Parameter Values Std. d ev. 1.26 Mean 12.0 C.V. 10.5 PRESS 99.5 R s quared 0.929 Adj R s quared 0.900 Pred R s quared 0.781 Adeq p recision ( ratio greater than 4 desired) 20.4 Table C 4. Data for coefficient of parameters for mass loss Factor Coefficient e stimate DF Standard e rror 95% CI low 95% CI h igh VIF Actual factor s Intercept 10.82 1 0.31 10.17 11.46 45.79 A DO 0.33 1 0.37 1.10 0.43 1 0.91 B Temperature 4.99 1 0.37 4.23 5.75 1 0.14 C Time 1.01 1 0.37 0.24 1.77 1 0.31 D Steam to c arbon r atio 0.71 1 0.37 0.05 1.48 1 7.39 B 2 3.01 1 0.48 2.01 4.01 1 0.00013 AB 1.21 1 0.63 2.53 0.12 1 0 .0018 BC 1.96 1 0.63 0.63 3.28 1 0.00065 BD 2.21 1 0.63 0.89 3.53 1 0.015

PAGE 116

116 Table C 5. Diagnostic case studies for mass loss Standard o rder Actual v alue Predicted v alue Residual Leverage Student r esidual d istance Outlier t 1 8 7.96 0.17 0.50 0.19 0.00 0.18 2 9 9.72 1.15 0.50 1.28 0.18 1.30 3 21 20.36 0.89 0.50 0.99 0.11 0.99 4 17 17.28 0.09 0.50 0.10 0.00 0.10 5 11 9.10 1.82 0.23 1.63 0.09 1.70 6 10 11.11 0.79 0.23 0.71 0.02 0.70 7 10 10 .52 0.27 0.23 0.25 0.00 0.24 8 12 12.53 0.86 0.23 0.77 0.02 0.76 9 11 10.44 0.56 0.23 0.51 0.01 0.50 10 11 9.77 1.63 0.23 1.46 0.07 1.51 11 11 11.86 0.75 0.23 0.68 0.01 0.67 12 11 11.20 0.60 0.23 0.54 0.01 0.53 13 10 9.79 0.06 0.50 0.06 0.00 0.06 14 15 15.85 0.83 0.50 0.92 0.09 0.92 15 10 7.89 2.35 0.50 2.62 0.76 3.15 16 23 21.78 1.47 0.50 1.63 0.30 1.71 17 10 10.14 0.17 0.23 0.15 0.00 0.15 18 10 9.48 0.52 0.23 0.47 0.01 0.46 19 12 12.15 0.54 0.23 0. 49 0.01 0.48 20 11 11.49 0.16 0.23 0.14 0.00 0.14 21 8 10.34 1.97 0.50 2.20 0.54 2.46 22 14 15.90 2.14 0.50 2.39 0.63 2.75 23 8 7.34 0.89 0.50 0.99 0.11 0.99 24 22 21.74 0.71 0.50 0.79 0.07 0.79 25 9 10.82 1.57 0.06 1.27 0.0 1 1.29 26 11 10.82 0.29 0.06 0.24 0.00 0.23 27 11 10.82 0.23 0.06 0.19 0.00 0.18 28 11 10.82 0.26 0.06 0.21 0.00 0.21 29 11 10.82 0.06 0.06 0.05 0.00 0.05

PAGE 117

117 Table C 6. ANOVA results for MIB removal Source Sum of s quares DF Mean s quare F v alue Prob > F Model 7306 8 913 18.35 < 0.0001 significant A 205 1 205 4.12 0.0559 B 3665 1 3665 73.62 < 0.0001 significant C 1951 1 1951 39.19 < 0.0001 significant D 257 1 257 5.16 0.0344 significant A 2 112 1 112 2.25 0.1491 B 2 724 1 724 14.54 0.0011 significant AC 135 1 135 2.70 0.1158 AD 322 1 322 6.47 0.0193 significant Residual 996 20 50 Lack of f it 903 16 56 2.44 0.2009 not significant Pure e rror 92 4 23 Cor t otal 8302 28 Table C 7. Statistical data for MIB removal Parameter Values Std. d ev. 7.05 Mean 60.0 C.V. 11.7 PRESS 2380.4 R s quared 0.880 Adj R s quared 0.832 Pred R s quared 0.713 Adeq p recision (ratio greater than 4 desired) 15.3 Table C 8. Data for coefficient of parameters for MIB removal Factor Coefficient e stimate DF Standard e rror 95% CI l ow 95% CI h igh VIF Actual f actors Intercept 54.16 1 2.127 49.72 58.59 126.7 A DO 4.13 1 2.037 8.38 0.12 1 4.76 B Temperature 17.48 1 2.037 13.23 21.72 1 0.038 C Time 12.75 1 2.037 8.50 17.00 1 1.12 D Steam to c arbon r atio 4.63 1 2.037 0.38 8.87 1 10.33 A 2 4.03 1 2.685 1.57 9.63 1.019 0.2 B 2 10.24 1 2.685 4.64 15.84 1.019 0.00046 AC 5.80 1 3.528 13.16 1.56 1 0.064 AD 8.98 1 3.528 1.62 16.33 1 1.99

PAGE 118

118 Table C 9. Diagnostic case studies for MIB removal Standard o rder Actual v alue Predicted v alue Residual Leverage Student r esidual d istance Outlier t 1 56.5 55.09 1.41 0.31 0.24 0.00 0.24 2 42.4 46.82 4.42 0.31 0.76 0.03 0.75 3 94.7 90.04 4.66 0.31 0.80 0.03 0.79 4 66.9 81.77 14.87 0.31 2.55 0.33 3.02 5 37.1 36. 78 0.32 0.26 0.05 0.00 0.05 6 65.7 62.28 3.42 0.26 0.56 0.01 0.55 7 42.4 46.03 3.63 0.26 0.60 0.01 0.59 8 62.9 71.53 8.63 0.26 1.42 0.08 1.46 9 72.6 66.67 5.93 0.52 1.21 0.17 1.22 10 49.2 40.45 8.75 0.52 1.78 0.38 1.89 1 1 56.8 57.97 1.17 0.52 0.24 0.01 0.23 12 69.3 67.65 1.65 0.52 0.34 0.01 0.33 13 30.5 34.17 3.67 0.27 0.61 0.01 0.60 14 75.4 69.12 6.28 0.27 1.04 0.04 1.04 15 58.7 59.67 0.97 0.27 0.16 0.00 0.16 16 95.9 94.62 1.28 0.27 0.21 0.00 0.21 17 37.3 43.77 6.47 0.52 1.32 0.21 1.34 18 50.5 47.10 3.40 0.52 0.69 0.06 0.68 19 76.5 80.87 4.37 0.52 0.89 0.09 0.89 20 66.5 61.00 5.50 0.52 1.12 0.15 1.13 21 38.5 42.30 3.80 0.27 0.63 0.02 0.62 22 71.1 77.25 6 .15 0.27 1.02 0.04 1.02 23 63 .0 51.55 11.45 0.27 1.89 0.14 2.04 24 95.3 86.50 8.80 0.27 1.46 0.09 1.50 25 55.3 54.16 1.14 0.09 0.17 0.00 0.17 26 57 .0 54.16 2.84 0.09 0.42 0.00 0.41 27 56.2 54.16 2.04 0.09 0.30 0.00 0.30 28 45 .2 54.16 8.96 0.09 1.33 0.02 1.36 29 52.4 54.16 1.76 0.09 0.26 0.00 0.25 Table C 10. ANOVA results for pH pzc Source Sum of S quares DF Mean s quare F v alue Prob > F Model 12.94 4 3.23 40.01 < 0.0001 significant A 0.07 1 0.07 0 .83 0.3699 B 11.60 1 11.60 143.53 < 0.0001 significant C 0.19 1 0.19 2.32 0.1409 D 1.08 1 1.08 13.36 0.0013 significant Residual 1.94 24 0.08 Lack of f it 1.22 20 0.06 0.34 0.9544 not significant Pure e rror 0.72 4 0.18 Cor t otal 14.88 28

PAGE 119

119 Table C 11. Statistical data for pH pzc Parameter Values Std. d ev. 0.284 Mean 11.0 C.V. 2.56 PRESS 2.57 R s quared 0.869 Adj R s quared 0.847 Pred R s quared 0.826 Adeq p recision (rati o greater than 4 desired) 21.7 Table C 12. Data for coefficient of parameters for pH pzc Factor Coefficient e stimate DF Standard e rror 95% CI l ow 95% CI h igh VIF Actual f actors Intercept 11.09 1 0.053 10.984 11.202 7.01 A DO 0.08 1 0.082 0.24 4 0.094 1 0.017 B Temperature 0.98 1 0.082 0.814 1.153 1 0.0066 C Time 0.13 1 0.082 0.044 0.294 1 0.0063 D Steam to c arbon r atio 0.30 1 0.082 0.131 0.469 1 0.30

PAGE 120

120 Table C 13. Diagnostic case studies fo r pH pzc Standard o rder Actual v alue Predicted v alue Residual Leverage Student ve sidual d istance Outlier t 1 8 7.96 0.17 0.50 0.19 0.00 0.18 2 9 9.72 1.15 0.50 1.28 0.18 1.30 3 21 20.36 0.89 0.50 0.99 0.11 0.99 4 17 17.28 0. 09 0.50 0.10 0.00 0.10 5 11 9.10 1.82 0.23 1.63 0.09 1.70 6 10 11.11 0.79 0.23 0.71 0.02 0.70 7 10 10.52 0.27 0.23 0.25 0.00 0.24 8 12 12.53 0.86 0.23 0.77 0.02 0.76 9 11 10.44 0.56 0.23 0.51 0.01 0.50 10 11 9.77 1.63 0.23 1.46 0.07 1.51 11 11 11.86 0.75 0.23 0.68 0.01 0.67 12 11 11.20 0.60 0.23 0.54 0.01 0.53 13 10 9.79 0.06 0.50 0.06 0.00 0.06 14 15 15.85 0.83 0.50 0.92 0.09 0.92 15 10 7.89 2.35 0.50 2.62 0.76 3.15 16 23 21.78 1.47 0.50 1.63 0.30 1 .71 17 10 10.14 0.17 0.23 0.15 0.00 0.15 18 10 9.48 0.52 0.23 0.47 0.01 0.46 19 12 12.15 0.54 0.23 0.49 0.01 0.48 20 11 11.49 0.16 0.23 0.14 0.00 0.14 21 8 10.34 1.97 0.50 2.20 0.54 2.46 22 14 15.90 2.14 0.50 2.39 0.63 2.75 23 8 7.34 0.89 0.50 0.99 0.11 0.99 24 22 21.74 0.71 0.50 0.79 0.07 0.79 25 9 10.82 1.57 0.06 1.27 0.01 1.29 26 11 10.82 0.29 0.06 0.24 0.00 0.23 27 11 10.82 0.23 0.06 0.19 0.00 0.18 28 11 10.82 0.26 0.06 0.21 0.00 0.21 29 11 10.8 2 0.06 0.06 0.05 0.00 0.05 Table C 14. ANOVA results for phenol removal Source Sum of s quares DF Mean s quare F v alue Prob > F Model 656.55 4 164.14 13.60 < 0.0001 significant A 2.52 1 2.52 0.21 0.6517 B 576.85 1 576.85 47.80 < 0.0001 significant C 59.41 1 59.41 4.92 0.0362 significant D 17.76 1 17.76 1.47 0.2368 Residual 289.60 24 12.07 Lack of f it 269.20 20 13.46 2.64 0.1791 not significant Pure e rror 20.41 4 5.10 Cor t otal 94 6.15 28

PAGE 121

121 Table C 15. Statistical data for phenol removal Parameter Values Std. d ev. 3.47 Mean 59.4 C.V. 5.84 PRESS 434.2 R s quared 0.693 Adj R s quared 0.642 Pred R s quared 0.541 Adeq p recision (ratio greater than 4 desired) 12.6 Table C 16. Data for coefficient of parameters for phenol removal Factor Coefficient e stimate DF Standard e rror 95% CI l ow 95% CI h igh VIF Actual f actors Intercept 59.46 1 0.65 58.13 60.79 29.76 A DO 0.46 1 1.00 2.53 1.61 1 0.10 B Temperature 6.93 1 1.00 4.86 9.00 1 0.046 C Time 2.23 1 1.00 0 .16 4.29 1 0.11 D Steam to c arbon r atio 1.22 1 1.00 0.85 3.29 1 1.22

PAGE 122

122 Table C 17. Diagnostic case studies for phenol removal Standard o rder Actual v alue Predicted v alue Residual Leverage Student r esidual d istance Outlier t 1 56.8 52.99 3.81 0.20 1.23 0.08 1.24 2 49.6 52.07 2.47 0.20 0.80 0.03 0.79 3 65.9 66.85 0.95 0.20 0.31 0.00 0.30 4 67.4 65.94 1.46 0.20 0.47 0.01 0.46 5 52.6 56.02 3.42 0.20 1.10 0.06 1.11 6 59.4 60.47 1.07 0.20 0.34 0.01 0.34 7 54.8 58.45 3.65 0.20 1.18 0.07 1.19 8 62.7 62.90 0.20 0.20 0.07 0.00 0.06 9 63.2 58.70 4.50 0.20 1.45 0.11 1.48 10 61.6 57.79 3.81 0.20 1.23 0. 08 1.24 11 62.4 61.14 1.26 0.20 0.41 0.01 0.40 12 63.0 60.22 2.78 0.20 0.90 0.04 0.89 13 45.9 50.30 4.40 0.20 1.42 0.10 1.45 14 68.1 64.17 3.93 0.20 1.27 0.08 1.28 15 54.4 54.75 0.35 0.20 0.11 0.00 0.11 16 68.8 68.62 0.18 0.20 0.06 0.00 0.06 17 53.6 57.70 4.10 0.20 1.32 0.09 1.34 18 55.6 56.78 1.18 0.20 0.38 0.01 0.37 19 56.4 62.15 5.75 0.20 1.85 0.17 1.96 20 55.6 61.23 5.63 0.20 1.81 0.17 1.91 21 50.7 51.31 0.61 0.20 0.20 0.00 0.19 22 64.8 65.18 0.38 0.20 0.1 2 0.00 0.12 23 59.2 53.75 5.45 0.20 1.76 0.16 1.84 24 64.8 67.61 2.81 0.20 0.91 0.04 0.90 25 63.2 59.46 3.74 0.03 1.10 0.01 1.10 26 59.0 59.46 0.46 0.03 0.14 0.00 0.13 27 59.0 59.46 0.46 0.03 0.14 0.00 0.13 28 63.6 59.46 4.14 0.03 1 .21 0.01 1.22 29 62.3 59.46 2.84 0.03 0.83 0.00 0.83

PAGE 123

123 LIST OF REFERENCES Albers, P.W., Pietsch J., Krauter, J., Parker, S.F ., 2003. Investigations of activated carbon catalyst supports from different natural sources. Phys. Chem. Chem. Phys. 5, 19 41 1949. Alcaniz Monge, J., Cazorla Amoros, D., Linares solano, A., 1994. Effect of the activating gas on tensile strength and pore structure of pitch based carbon fibres. Carbon 32 (7), 1277 1283. Ania, C.O., Parra, J.B., Pis, J.J., 2004. Oxygen indu ced decrease in the equilibrium adsorptive capacities of activated carbons. Adsorption Science and Technology 22 (4), 337 351. Appleton, Wisconsin, 2006. Wastewater treatment facility: Treatment plant tour. Accessed: July 9. . Arriagada, R., Garcia, R., MolinaSabio, M., RodriguezReinoso, F., 1997. Effect of steam activation on the porosity and chemical nature of activated carbons from Eucalyptus globulus and peach stones. Microporous Materials 8 (3 4), 123 130. Bach, M.T., 2004. The role of calcium in and methodologies for overcoming pH excursions for reactivated granular activated carbon. Thesis (M.E.). University of Florida. Benjamin, M.M., 2002. Water C hemistry. McGraw Hill; N ew York. Bansal, R.C., Aggarwall, D., Goyal, M., Kaistha, B.C., 2002. Influence of carbon oxygen surface groups on the adsorption of phenol by activated carbons. Ind. Journal of Chem. Tech. 9, 290. Bansal, R.C., Donnet, J.B., Stoeckli, F., 1988. Active Carbon Marcel Dekker, Inc.; New York. Bansal, R.C., Vastola, F.J., Walker, P.L., 1974. Influence of hydrogen chemisorption on the subsequent chemisorption of oxygen on activated graphon. Carbon 12, 355. Barnett, K.E., 1999. Wastewater treatment method and plant United States Patent Number 5,961,830. Barton, S.S., Evans, M.J.B, Halliop, E., MacDonald, J.A.F., 1997. Acidic and basic sites on the surface of porous carbon. Carbon 35 (9), 1361 1366. Bizot, P.M., Bailey, B.R., 1997. Sodium sulfoxylate f ormaldehyde as a boiler additive for oxygen scavenging. United States Patent Number 5,660,736 Boehm, H., 1966. Functional groups on the surfaces of solids. Angewandte Chemie Inernational Edition 5, 533 522. Box, G.E.P., Behnken, D.W., 1960. Some ne w three level designs for the study of quantitative variables. Technometrics 2 (4), 455 475.

PAGE 124

124 Buczek. B., 1993. Development of properties within particles of active carbons obtained by a steam activation process. Langmuir 9, 2509 2512. Bull, R.J. Robi nson, M., Meier, J.R., Stober, J., 1982. Use of biological assay systems to assess the relative carcinogenic hazards of disinfection by products. Environmental Health Perspectives 46, 215 227. Campbell, A.G. Folk, R.L., Tripepi, R.R., 1997. Wood ash as an amendment in municipal sludge and yard waste composting processes. Compost Science & Utilization 5 (1), 62 73. Cannon, F.S., Snoeyink, V.L., Lee, R.G., Dagois, G., DeWolfe, J.R., 1993. Effect of calcium in field spent GACs on pore development during regeneration. JAWWA 85 (3), 76 89. Cannon, F.S., Snoeyink, V.L., Lee, R.G., Dagois, G., 1994. Reaction mechanism of calcium catalyzed thermal regeneration of spent granular activated carbon. Carbon 32 (7), 1285 1301. Carr, S., Farmer, R.W., 1995. React pH TM improves pH in carbon effluent. The National Environmental Journal 5 (2), 54. Centeno, T.A., Stoeckli, F., 1995. The oxidation of an asturian bituminous coal in air and its influence on subsequent activation by steam. Carbon 33 (5), 581 586. Chen W.F., Cannon, F.S., Rangel Mendez, J.R., 2005a. Ammonia tailoring of GAC to enhance perchlorate removal. I: Characterization of NH 3 thermally tailored GACs. Carbon 43 (3), 573 580. Chen, W.F., Cannon, F.S., Rangel Mendez, J.R., 2005b. Ammonia tailoring of GAC to enhance perchlorate removal. II: Perchlorate adsorption. Carbon 43 (3), 581 590. Chestnutt, T.E., Bach, M.T., Mazyck, D.W., 2007. Improvement of thermal reactivation of activated carbon for the removal of 2 methylisoborneol. Water Res. 41, 7 9 86. Clark, R.M., Lykins, B.W., 1989. Granular Activated: Design, Operation and Cost. Lewis Publishers; Chelsea, MI. Considine, R., Denoyel, R., Pendleton, P., Schumann, R., Wong, S.H., 2001. The influence of surface chemistry on activated carbon adso rption of 2 methylisoborneol from aqueous solution. Colloids and Surfaces A: Physicochemical and Engineering Aspects 179 (2 3), 271 280. Crittenden, J.C. Berrigan, J.K., Hand, D.W., 1986. Design of rapid small scale adsorption tests for a constant diffu sivity. Journal of Water Pollution Control Federation 58 (4), 312 319. Crittenden, J.C. Reddy, P.S., Arora, H., Trynoski, J., Hand, D.W., Perram, D.L., Summers, RS., 1991. Predicting GAC performance with rapid small scale column tests. JAWWA 83 (1), 77 87.

PAGE 125

125 Das, K.C., Melear, N.D., Kastner, J.R., Buquoi, J.Q., 2003. Influence of ash amendment on odor emissions and aerobic biodegradation of biosolids mixes. Transactions of the ASAE 46 (4), 1185 1191. Dussert, B.W., Farmer, R.W., Hayden, R.A., 1995. Oxi dized activated carbon for the control of pH and alkalinity in water treatment applications. United States Patent 5,466,378. El Sayed, Y., Bandosz, T.J., 2003. Effect of increased basicity of activated carbon surface on valeric acid adsorption from aqueo us solution activated carbon. Phys. Chem. Chem. Phys 5 (21), 4892 4898. Farmer, R.W., Dussert, B.W., Kovacic, S.L., 1996. Improved granular activated carbon for the stabilization of wastewater pH. American Chemical Society, Division of Fuel Chemistry 4 1 (1), 456 460. Farmer, R.W., Kovacic, L., Matviya, M., Wadhwa, N.P., 1998. Activated carbon treated by carbon dioxide for the stabilization of treated water pH. United States Patent Number 5,714,433. Fernandes, M.B., Skjemstad, J.O., Johnson, B.B., Wel ls, J.D., Brooks, P., 2003. Characterization of carbonaceous combustion residues. I. Morphological, elemental, and spectroscopic features. Chemosphere 51, 785 795. Frederick, H.T., Cannon, F.S., Dempsey, B.A., 2001a. Calcium and TOC loading: Effect of h ydroxyl and carboxyl substituents. Colloids and Surfaces A: Physicochemical and Engineering Aspects 191 (1 2), 161 177. Frederick, H.T., Cannon, F.S., Dempsey, B.A., 2001b. Calcium loading onto granular activated carbon with salicylate or phthalate. Col loids and Surfaces A: Physicochemical and Engineering Aspects 177 (2 3), 157 168. Frederick, H.T., Cannon, F.S., 2001. Calcium NOM loading onto GAC. JAWWA 93(12), 77 89. Garc a Garc a, A., Greg rio, A., Franco, C., Pinto, F., Boavida, D., Gulyurtlu, I. 2003. Unconverted chars obtained during biomass gasification on a pilot scale gasifier as a source of activated carbon production. Bioresource Technology 88, 27 32. Garten, V.A., Weiss, V.E., 1957. The ion and electron exchange properties of activated carbon in relation to its behavior as a catalyst and adsorbent. Rev Pure Appl Chem 7, 69 122. Gerber, N.N., 1968. A volatile metabolite of actinomycetes, 2 methylisoborneol. Journal of Antibiotics 22 (10), 508. Gergova, K., Eser. S., 1996. Effects o f activation method on the pore structure of activated carbons from apricot stones. Carbon 34 (7), 879 888.

PAGE 126

126 Gergova, K., Petrov, N., Eser, S., 1994. Adsorption properties and microstructure of activated carbons produced from agricultural by products by steam pyrolysis. Carbon 32 (4), 693 702 Gergova, K., Petrov, N., Butuzova, L., Minkova, V., Isaeva, L., 1993. Evolution of the active surface of carbons produced from various raw materials by steam pyrolysis activation. Journal of chemical technology and biotechnology 58 (4), 321 330. Gerth, W.A., 1985. Applicability of Henry's law to hydrogen, helium, and nitrogen solubilities in water and olive oil at 37 C and pressures up to 300 atmospheres. Archives of Biochemistry and Biophysics 241 (1), 187 199. Gillogly, T., Sneoyink, V., Elarde, J., Wilson, C., Royale, E., 1998. 14 C MIB adsorption on PAC in natural water. JAWWA 90(1), 98 108. Gonzalez, M.T., Molina Sabio, M., Rodriguez Reinoso, F., 1994. Steam activation of olive stone chars, development of porosity. Carbon 32 (8), 1407 1413. Hammer, M.J., Hammer, M.J., Jr., 2001. Water and Wastewater Technology Prentice Hall; Upper Saddle River, NJ. Hassler, J.W., 1974. Purification with activated carbon: industrial, commercial, environmental. Chemical Pub. Co.; New York. Herzing, D., Snoeyink, V., Wood, N., 1977. Activated carbon adsorption of the odorous compounds 2 methylisoborneol and geosmin. JAWWA 69, 223. Heschel, W., Klose, E., 1995. On the suitability of agricultural by products for the manufa cture of granular activated carbon. Fuel 74 (12), 1786 1791. Jung, M.W. Ahn, K.H., Lee, Y., Kim, K.P., Rhee, J.S., Park, J.T., Paeng, K.J., 2001. Adsorption characteristics of phenol and chlorophenols on granular activated carbons (GAC). Microchemical Journal 70 (2), 123 131. Karanfil, T., Kitis, M., 1999. Role of granular activated carbon surface chemistry on the adsorption of organic compounds. 2. Natural Organic Matter. Environ. Science and Technology 33 (18), 3225 3233. Kastner, J.R., Das, K.C ., Buquoi, Q., Melear, N,D., 2003. Low temperature catalytic oxidation of hydrogen sulfide and methanethiol using wood and coal fly ash. Environ. Sci. Technol. 37 (11), 2568 2574. Kawamura, S., 2000. Integrated design and operation of water treatment fac ilities John Wiley & Sons; New York.

PAGE 127

127 Khezami. L., Chetouani, A., Taouk, B., Capart, R., 2005. Production and characterisation of activated carbon from wood components in powder: Cellulose, lignin, xylan. Powder Technology 157, 48 56. Kipling, J.J., Gass ser, C.G., 1960. Adsorption from liquid mixtures at solid surfaces. J. Phys.Chem 64 (6), 710. Kirk Othmer, 1977. Encyclopaedia of Chemical Technology, Third Edition, Wiley; New York. Knappe, D.R.U., Snoeyink, V.L., Dagois, G., DeWolfe, J.R., 1992. Effec t of calcium on thermal regeneration of GAC. JAWWA 84 (8), 73 80. Kruyt, H.R., De Kadt, G.S., 1929. Die ladung der kohle. Kolloid Chem. Beihefte 47, 44. Laine, N.R., Vastola, F.J., Walker, P.L., 1963. Importance of active surface area in carbon oxygen reaction Journal of Physical Chemistry 67 (10), 2030. Leon y Leon, C.A., Solar, J.M., Calemma, V., Radovic, L.R., 1992. Evidence for the protonation of basal plane sites on carbon. Carbon 30 (5), 797 811. Lillo Rodenas, M.A., Cazorla Amoros, D., Linar es Solano, A., 2005. Behaviour of activated carbons with different pore size distributions and surface oxygen groups for benzene and toluene adsorption at low concentrations. Carbon 43 (8), 1758 1767. Lussier, M.G., Zhang, Z., Miller, D.J., 1998. Characte rizing rate inhibition in steam/hydrogen gasification via analysis of adsorbed hydrogen. Carbon 36 (9), 1361 1369. MacKenzie, J.A., Tennant, M.F., Mazyck, D.W., 2005. Tailored GAC for the effective control of 2 methylisoborneol. JAWWA 97 (6), 76 87. Mar tin Gullon, I., Asensio, M., Font, R., Marcilla ,A., 1996. Steam activated carbons from a bituminous coal in a continuous multistage fluidized bed pilot plant. Carbon 34 (12), 1515 1520. Mazyck, D.W., Cannon, F.S., 2000. Overcoming calcium catalysis durin g the thermal reactivation of granular activated carbon: Part I. Steam curing plus ramped temperature N 2 treatment. Carbon 38 (13), 1785 1799. Mazyck, D.W., Cannon. F.S., 2002. Overcoming calcium catalysis during the thermal reactivation of granular acti vated carbon. Part II. Variation of steam curing reactivation parameters. Carbon 40, 241 252. Mazyck, D.W., Cannon, F.S., Bach, M.T., Radovic, L.R., 2005. The role of calcium in high pH excursions with reactivated GAC. Carbon 43, 511 518.

PAGE 128

128 Menndez, J. A., Phillips, J., Xia, B., Radovic, L.R., 1996. On the modification and characterization of chemical surface properties of activated carbon: In the search of carbons with stable basic properties. Langmuir 12 (18), 4404 4410. MolinaSabio, M., Gonzalez, M.T ., RodriguezReinoso, F., Sepulveda Escribano, A., 1996. Effect of steam and carbon dioxide activation in the micropore size distribution of AC. Carbon 34 (4), 505 509. Montgomery, D.C., 1991. Design and analysis of experiments. Wiley; New York. Myers, M.H., 1995. Response surface methodology. Wiley; New York. Newcombe, G., Morrison, J., Hepplewhite, C., Knappe, D.R.U., 2002. Simultaneous adsorption of MIB and NOM onto activated carbon. II. Competitive effects. Carbon 40, 2147. Newcombe, G., 1999. Ch arge vs. porosity some influences on the adsorption of natural organic matter (NOM) by activated carbon. Water Science & Technology 40 (9): 191 198. Nowack, K.O., Cannon, F.S., Arora, H., 1999. Ferric chloride plus GAC for removing TOC. J AWWA 91 (2), 65 78. Nowack, K.O., Cannon, F.S., Mazyck, D.W., 2004. Enhancing activated carbon adsorption of 2 methylisoborneol: Methane and steam treatments. Environ. Sci. Technol. 38, 276 284. Nowack, K.O., Cannon, F.S., 1997. Control of calcium buildup in GAC: Effe ct of iron coagulation. Carbon 35 (9), 1223 1237. Otake, Y., Jenkins, R.G., 1993. Characterization of oxygen containing surface complexes created on a microporous carbon by air and nitric acid treatment. Carbon 31 (1), 109 121. Pelekani, C., Snoeyink, V. L., 1999. Competitive adsorption in natural water: Role of activated carbon pore size. Water Research 33 (5), 1209 1219. Pendleton, P., Wong, S.H., Schumann, R., Levay, G., Denoyel, R., Rouquerol, J., 1997. Properties of activated carbon controlling 2 met hylisoborneol adsorption. Carbon 35, 1141. Petrucci, R.H., Harwood, W.S., 1997. General chemistry: principles and modern applications. Prentice Hall; New Jersey. Pereira, M.F.R., Soares, S.F., "rfo, J.J.M., Figueiredo, J.L., 2003. Adsorption of dyes on activated carbons: influence of surface chemical groups. Carbon 41 (4), 811 821. Puri, B.R, 1970. Surface Complexes on Carbon in Chemistry and Physics of Carbon Vol. 6, ed. P.L. Walker, Jr., Marcel Dekker; New York.

PAGE 129

129 Quirilivan, P.A., Li, L., Knappe D.R.U., 2005. Effects of activated carbon characteristics on the simultaneous adsorption of aqueous organic micropollutants and natural organic matter. Water Research 39 (8), 1663 1673. Rangel Mendez, J.R., Cannon, F.S., 2005. Improved activated carbon by thermal treatment in methane and steam: Physicochemical influences on MIB sorption capacity. Carbon 43, 467. Rashash, D.M.C., Dietrich, A.M., Hoehn, R.C., 1997. FPA of selected odorous compounds. JAWWA 89 (4), 131 141. Ravichandran, M., Aiken, G.R., Reddy, M.M., Ryan, J.N., 1998. Enhanced dissolution of cinnabar (mercuric sulfide) by dissolved organic matter isolated from the Florida Everglades. Env. Sci.& Tech. 32, 3305. Rio, S., Faur Brasquet, C., Le Coq, L., Le Cloirec, P., 2005. Production and characterization of adsorbent materials from an industrial waste. Adsorption 11, 793 798. Rodrguez Reinoso, F., Molina Sabio, M., Gonzlez, M.T., 1995. The use of steam and CO 2 as activating agents in the preparation of activated carbons. Carbon 33 (1), 15 23. Rosenfeld, P.E., Henry, C.L., 2000. Wood ash control of odor from biosolids application. Journal of Environmental Quality 29 (5), 1662 1668. Ryu, S.K., Jin, H., Gondy, D., Pusset, N., Ehrburger, P., 1993. Activation of carbon fibers by steam and carbon dioxide. Carbon 31 (5), 841 842. Salame, I.I., Bandosz, T.J., 2001. Surface chemistry of activated carbons: Combining the results of temperature programmed desorption, Boehm, and potentiometric titrations. J. Colloid and Interface Sciences 240, 2 52 258. Salame, I.I., Bandosz, T.J., 2003. Role of surface chemistry in adsorption of phenol on activated carbons. Journal of Colloid and Interface Science 264 (2), 307 312. Sarkar, M., Acharya, P.K., 2006. Use of fly ash for the removal of phenol and it s analogues from contaminated water. Waste Management 26,559 570. Smith, W.R., Polley, M.H., 1956. The oxidation of graphitized carbon black Journal of Physical Chemistry 60 (5), 689 691. Water Quality and Treatment AWWA; Denver, CO. Souza Macedo, J., Costa Jnior, N.B., Almeida, L.E., Silva Vieira, E.F., Cestari, A.R., Ftima Gimenez, I., Carreo, N.L.V., Barreto, L.S., 2006. Kinetic and calorimetric study of the adsorpti on of dyes on mesoporous activated carbon prepared from coconut coir dust. Journal of Colloid and Interface Science 298, 515 522.

PAGE 130

130 Speitel, G.E., Dvorak, B.I,. Morley, M.C., 2001. R apid small scale column test. AEESP Environmental Engineering Processes Lab oratory Manual. Association of Environmental Engineering and Science Professors; Champaign, Illinois. Streat, M., Patrick, J.W., Perez, M.J.C., 1995. Sorption of phenol and para chlorophenol from water using conventional and novel activated carbons. Wat Res. 29 (2), 467 472. Summers, R.S., Haist, B., Koehler, J., Ritz, J., Zimmer, G., Sontheimer, H., 1989. The influence of background organic matter on GAC adsorption. JAWWA 81 (5), 66 74. Suzuki, M., 1990. Adsorption engineering. Elsevier; New York. The effect of the gradual thermal decomposition of surface oxygen species on the chemical and catalytic properties of oxidized activated carbon. Carbon 40 (14), 2627 2639. Taiz, L., Zeiger E., 2006. Plant Physiology, Fourth Edition. Sinauer Associates, Inc.; Sunderland, MA. Tennant, M.F., Mazyck, D.W., 2003. Steam pyrolysis activation of wood char for superior odorant removal. Carbon 41, 2195. Tennant, M.F., 2004. Activation and use of powdered activated carbon for removing 2 methylisoborneol in water utilities. Dissertation (Ph.D.). University of Florida. Tomkw, K., Siemieniewska, T., Czechowski, F., Jankowska, A., 1977. Formation of porous structures in activated brown coal chars u sing O 2 CO 2 and H 2 O as activating agents. Fuel 56 (2), 121 124. Walker, P.L., Rusinko, F., Austin, L.G., 1959. Gas reactions of carbon. Advan. Catal. 11, 133 221. Walker, P.L., 1996. Production of activated carbons: Use of CO 2 versus H 2 O as activating agent. Carbon 34 (10), 1297 1299. Weiss, R.F., 1970. The solubility of nitrogen, oxygen, and argon in water and seawater. Deep Sea Research 17, 721 735. Wiegand, W.B., 1937. pH properties of colloidal carbon. Ind. Eng. Chem. 29, 953. Wigmans, T., 1989 Industrial aspects of production and use of activated carbons. Carbon 27 (1), 13 22. Zhang, Y.Z., Lu, Z., Maroto Valer, M.M., Andresen, J.M., Schobert, H.H., 2003. Comparison of high unburned carbon fly ashes from different combustor types and their st eam activated products. Energy & Fuels 17 (2), 369 377.

PAGE 131

131 Zou, Y., Han, B.X., 2001. High surface area activated carbon from Chinese coal. Energy & Fuels 15, 1383 13 86.

PAGE 132

132 BIOGRAPHICAL SKETCH Morgana Bach was born in St. Petersburg, Florida, where she was rai sed by her mother. She graduated from Gibbs High School in June 1998. From there, she traveled to Gainesville, Florida, to attend the University of Florida where I received she received her bachelor of science degree with highest honors and her Master of Engineering in Environmental Engineering Sciences. She stayed on at the University of Florida to work on her doctor of philosophy degree with Dr. David W. Mazyck in the field of environmental engineering sciences with a focus on water treatment. She look s forward to working towards the betterment of mankind through research into the technologies which provide clean water and air.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E20101114_AAAABP INGEST_TIME 2010-11-14T08:19:42Z PACKAGE UFE0021222_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES
FILE SIZE 26228 DFID F20101114_AAAQNU ORIGIN DEPOSITOR PATH bach_m_Page_128.QC.jpg GLOBAL false PRESERVATION BIT MESSAGE_DIGEST ALGORITHM MD5
77dfbdbc74b40b8baa9acf6227ae7d56
SHA-1
0d21793fb46e730edbe503c77d495d6842886720
53447 F20101114_AAAQOI bach_m_Page_025.pro
7e6950303a89be1caa56a0b25ef2ba2e
06de585fe5dae6b017804b44a65396c5fd755711
16346 F20101114_AAAQNV bach_m_Page_112.QC.jpg
f0c64afcc0ed2457dc5ea47393b1f356
2163fb113922f86f787c0001bb8ce5a2f3b05718
23323 F20101114_AAAQOJ bach_m_Page_018.QC.jpg
8c008de6aa5b825669ff3ecd19033377
acb08e4a0afa762288b079903a8511fc938e103e
67109 F20101114_AAAQNW bach_m_Page_014.jpg
6aa728ad5da322b86761502d31182f82
268bee62502ab4d7e6341661d18edc115c96aca8
10337 F20101114_AAAQOK bach_m_Page_131.jp2
21858626fd637563213ecc46276ede5f
f00fe404317f44b63a865e2221fe16b1d37691d3
1053954 F20101114_AAAQNX bach_m_Page_130.tif
6ee4001284e896f29f8b35a53065b3f5
2fec33083194e27684d3aea3ae7e2e299194c233
24861 F20101114_AAAQPA bach_m_Page_007.jpg
af5b556be08a12b904fcd401530bbab6
ca5ac55c35f8bfb3485f5aababbc40a6a440ee50
1370 F20101114_AAAQOL bach_m_Page_099.txt
f774bcb6e2dfd507b7d0e816786f4e8d
f7d0ab8579eadc129bab51145988e1e3bb0aa513
77355 F20101114_AAAQNY bach_m_Page_093.jpg
437ddb70937bd46cc4d95eeeb97bf23f
9e217a3ef6e70120982c322bea2de5456548b654
64952 F20101114_AAAQPB bach_m_Page_116.jp2
96fe78b35042d4ede7ea85be252c3766
b8981893bcd36a8d653b0d3e9cd8ff9e3e21b8d8
78538 F20101114_AAAQOM bach_m_Page_062.jpg
11e43381f801ea42e09ea45d169549a9
a186892025d961be405d6c6ff01eec306d6d50fb
27929 F20101114_AAAQPC bach_m_Page_004.jpg
18238de1f576e3a2f3cd41bc51356236
f2958511e82d0604d971acedb58db1e02c40242f
35178 F20101114_AAAQON bach_m_Page_004.jp2
31750739aacdb91c892b21d4f644eff6
8af3cf349e959d7eb0e97b1b072724d5661af455
2146 F20101114_AAAQNZ bach_m_Page_080.txt
ec6f3ce12c06d3cccb8aba6109d7daca
a0cda90c22ebf98977a3ecbf243be185d665c969
528727 F20101114_AAAQPD bach_m_Page_007.jp2
57f2efd5ed60a940c6e5f8c10935db32
29c864ecee411d4e1d589a7dff490bc150a8be04
115544 F20101114_AAAQOO bach_m_Page_088.jp2
b1bfa47ab9c601e6e5f7690bae55b40d
7ccbf2059f925b2ea53ab9710c2c24875240315e
428 F20101114_AAAQPE bach_m_Page_103.txt
3782d8d39712ccb9f35f8f0b2741a068
9aea5e71f6e10c402b8335450158119d0161ef28
6154 F20101114_AAAQOP bach_m_Page_042thm.jpg
44f83ef0ade5c4b0cfff7c3ac4db2ef8
13ec56db1c3035fd37ee69802e391a8ca12ea8fd
36642 F20101114_AAAQPF bach_m_Page_117.pro
505c3e6fa21ee7d5b09e6cdd8232e8fc
2c1ae2a9a4dc68eb3cad876c40f2380a778dd24d
F20101114_AAAQOQ bach_m_Page_031.tif
45d863ff75ebe67e741d7f370e933f26
0d579c7941a05449c6d07964cf097f0f2898378a
21752 F20101114_AAAQOR bach_m_Page_073.QC.jpg
4da800e9c05728612ac75aa44d1e31cb
1e261cc8a6f682b605f064ba1c5775995459c9f1
7001 F20101114_AAAQPG bach_m_Page_035thm.jpg
7bef7d450ffd939e7b015aca8a173a18
95f469dff170baca5ed8ba69f279507ed93b9d05
F20101114_AAAQOS bach_m_Page_033.tif
026ede3504ad42b8971deb99452512b8
22f2b997e7539dc95d0b278d4a125842def3e1b8
F20101114_AAAQPH bach_m_Page_020.tif
a29b0160675e0c791c70f995b9076280
90d332c255b1fcdd0a1b459d2169f5647d26962c
75033 F20101114_AAAQOT bach_m_Page_071.jpg
7eb1e0473d67ed1a15bdaced34e517b9
9a778ebca41c602c38742619b45772cb69d6af48
2086 F20101114_AAAQPI bach_m_Page_090.txt
e2228a99f391b2c1f94294498ea45677
285bc5e1904a71559e7664b440d6e59ef39537f9
119798 F20101114_AAAQOU bach_m_Page_023.jp2
62395608a239c8b5f86306b6cc2d56a0
24556b90a4e565d4dc06c5e34586526e5f0cb5d7
74996 F20101114_AAAQPJ bach_m_Page_087.jpg
d51a8f020565150819d35b1db8e70417
3a04a9e665f8f68c32f9a9501b571d00dedbcc13
24973 F20101114_AAAQOV bach_m_Page_027.QC.jpg
5a574aff9f2578a84905139ea142f20e
b5e48b7fd9de8f60d51fc06606d20cae4bbb1f09
2525 F20101114_AAAQPK bach_m_Page_129.txt
689e87a5673a770a35ee6075c9165f23
b75ff2a2a84a1a4423442939e1a3168c51419b04
11044 F20101114_AAAQOW bach_m_Page_011.QC.jpg
641843930d1f22ef943e7dd559e64333
500f7ee6d57e41489490aa77d524892026ec2976
122004 F20101114_AAAQPL bach_m_Page_032.jp2
61638f6856896ef54222df079a5e255a
3bd2ec8b12d5f257361dce6d3aae47a52013ad69
101161 F20101114_AAAQOX bach_m_Page_036.jp2
6b540028dcbb5a9e693be66bf07be59a
1d8d070c980b7c969a10bf28bfa84cbcdea43cf3
53622 F20101114_AAAQQA bach_m_Page_110.pro
c8c35e5b32f53cc45d8332d8c518e983
7d1de8ce2033b2dfc6816915e54d78fe4c17123c
F20101114_AAAQPM bach_m_Page_106.tif
91780edbbe66ce78a1fcb113e95f192c
ad4ffe4058152a3932e1e1d0f66e7ae8fec00553
6654 F20101114_AAAQOY bach_m_Page_071thm.jpg
5baf710c328a4ad643fc9da2598fb07c
2d2c4e1526594c48a571d3eef4657b2422716a16
4194 F20101114_AAAQQB bach_m_Page_131.QC.jpg
6b9604085df750664c3d108af5eede54
88dce67d843be0edce200e561b3096da067d1695
25271604 F20101114_AAAQPN bach_m_Page_112.tif
35a9401e1b7a9dda8eb5a2e7dabca38b
ff6880ffd5de4df3bb97b687750071c129ef811b
51922 F20101114_AAAQOZ bach_m_Page_109.jpg
b54a290f966ef9e57313752aa8c3830d
f754db60cd80b8cdc2deef6e583fd605cd61ec4c
57247 F20101114_AAAQQC bach_m_Page_099.jpg
35a891bc68943d6f9a7a004ca403acfc
96be6432194f9b1d7663411100cbe42ef72e2876
F20101114_AAAQPO bach_m_Page_109.tif
5e2724c9cbddea7ec10150a853dfaded
0adc7039a1cc7964a6ad1d5fc8be8af0d8750d27
24863 F20101114_AAAQQD bach_m_Page_090.QC.jpg
9b44c39b79ef20fd2fa95657ec1026df
e5020afd2727c2dcad73027d90dedf0fa96e1509
F20101114_AAAQPP bach_m_Page_079.tif
9346cd66b463a8f049177d1d4b8614a2
7afc6130377f7c92964d8f4029578fe3159a7291
54157 F20101114_AAAQQE bach_m_Page_089.pro
4225c459796702024ed4444a7051d01b
80aa213857991ae1f97692fe54f42c68f18adf44
76852 F20101114_AAAQPQ bach_m_Page_070.jpg
c18855362c5d85c585f5b858f0dc7bfb
894368f203ff6a375df5173c1fd25f3cd9d95f3a
15803 F20101114_AAAQQF bach_m_Page_119.pro
61914f3749788dfa9ba7118b623292de
6d4b7efdbe8977a09658e656cf3bcfbdc24efb7a
9016 F20101114_AAAQPR bach_m_Page_121.QC.jpg
fa7248a9e2048ccc09e33005a03a9240
5099f4bec5cbc35249e8defa6843f82af250308c
F20101114_AAAQQG bach_m_Page_107.tif
4210d2087ab340a5608cf345501dd099
4142fcd6ae6cc2d63d33c1da42aa1b3157cef719
5455 F20101114_AAAQPS bach_m_Page_112thm.jpg
6e10fe795c0e2d8e22fc6b99c4e0fb50
c1bad5a59c9c0a5163de626ceb177f9482f95af8
2147 F20101114_AAAQPT bach_m_Page_035.txt
372815e4bd8d54ee8aa8f8175ac711e5
c5fd784e8b9ba9df327d7d8e2139d64f71deb19a
7000 F20101114_AAAQQH bach_m_Page_110thm.jpg
899242cc2d599161363ec3f9d8bea9f2
b0f23d71898002bf15abf6c925753dff52384d4c
100351 F20101114_AAAQPU bach_m_Page_014.jp2
3cac1f152445ca4877501d8b21fa73ae
eff165c676b782f9e7d1e40ce3cef056ac4f3d0c
3556 F20101114_AAAQQI bach_m_Page_064thm.jpg
71ada97d256f7c04095ac7fe936384a7
0e3b27a59296a6b948805cb415e199b7f45b3dca
7786 F20101114_AAAQPV bach_m_Page_101.pro
b22249b9337cd991743d94d0a1673d97
9da19dac1c1804b786050706c6724ee518f80303
81589 F20101114_AAAQQJ bach_m_Page_126.jpg
ff1d7078dc1cc92a5a6469ce6d1b2fad
3176e608e89c4dd8b94b5244f7ab24975681bccd
74510 F20101114_AAAQPW bach_m_Page_030.jpg
f768e168a5d0b4da467ff491850bf039
8390e6d783c32715e514c504a98b4fe0c686fdaa
17230 F20101114_AAAQQK bach_m_Page_099.QC.jpg
d7b9fd688f5c17e4015109fd79be2a7d
ab119cf848db2b051b04ccf342874275f0b93960
627 F20101114_AAAQPX bach_m_Page_004.txt
daf481a94b5c20a65a37f8da8c161665
015c53a3a2b3a0f17240400b64bf1d9a0dabae69
9875 F20101114_AAAQRA bach_m_Page_106.QC.jpg
6d28238426481bb611e654c367b2d0c1
ab6ba10e9f67fb56021af5743ea43f7174cb0ea9
3925 F20101114_AAAQQL bach_m_Page_074thm.jpg
0fc7b1af04f975d6f4ed606511a265f2
a353ebd9d4972576dc19aa2b93b81042933b6659
22654 F20101114_AAAQPY bach_m_Page_056.QC.jpg
91b85f1c755d04767a9f2c575e1394a9
1b6bd3a97ab2ba47f78cab91b78f9ee4d786b0e2
54054 F20101114_AAAQRB bach_m_Page_024.pro
6a19c00ac7585b00057a960642853a42
e2ea9153556bb4fb23c0841d8900e19f91941894
60068 F20101114_AAAQQM bach_m_Page_128.pro
af5c7385d02ac9127403d4e4da05c8aa
8160d1294cb04d38e14c0c151cce88e005b46e60
40340 F20101114_AAAQPZ bach_m_Page_064.jpg
d527e1770f077ab28f489154aa664eaa
299c69f5c7e3320394f9c62149732ecfe45836bf
90941 F20101114_AAAQRC bach_m_Page_013.jp2
10d33f3652c1bf3140fe93611d1a45b4
3627f091676dbc4d6ee0177ab74c95a0fbf6a7ee
4858 F20101114_AAAQQN bach_m_Page_109thm.jpg
26e7c63369c228fcbe7148a144377d83
6968f2f88362870a7047fa7419f95e38a90b225b
478 F20101114_AAAQRD bach_m_Page_043.txt
7c27810c3431f6748a51627d6482d10f
013dd774fa6c1f0cf1a5da02ea364fdcf78b9919
F20101114_AAAQQO bach_m_Page_023.tif
cb8e1d42cfa046fbde9abba584565b3d
637af2574fd4b84a3ba15632cb3d91711a354d8c
1871 F20101114_AAAQRE bach_m_Page_046.txt
0b11a6a0ff888c6c1e966ca44722844f
87066191f88599e0b9f18c21b76c3cda05f3fa80
17347 F20101114_AAAQQP bach_m_Page_008.QC.jpg
64c4e92c1858ccc9a33ca16c212fd9f6
6d1e9d09649f820ba85c4b507ec2968a6bade1f1
2222 F20101114_AAAQRF bach_m_Page_092.txt
3f9eab7fd8093e8ceb08d9aed2be48d9
ea17d19d65eb6f8dfecb041ae7bb31c30b574ae3
36630 F20101114_AAAQQQ bach_m_Page_119.jp2
de35940ab2a68facfe46de7f68a29a94
343872c02338f17a71fcce33ed6ce0414f62cf1e
25291 F20101114_AAAQRG bach_m_Page_110.QC.jpg
ba278a884fecfc60846e011d1f842388
6720ab940ed5c00ece6ed4b06578c3a9abdbd132
6953 F20101114_AAAQQR bach_m_Page_058thm.jpg
9dbd4f90131791876fe71f4873fc31ee
79dae70b91e3298586abf43973977c0beb0f7faf
907 F20101114_AAAQRH bach_m_Page_106.txt
fa8d0645d822868f7b17a2d47b763334
d2c7c6aad125b7dbf4a8872e75d20b2aa1f317d5
2148 F20101114_AAAQQS bach_m_Page_070.txt
2040f72d64cb374b10a0784e0a554f4c
19ca52b4f0782deaa563afae078fae843bd7c359
F20101114_AAAQQT bach_m_Page_058.tif
8ceda390eaac0f6cf10b7d322c696aa6
b402c02bd6caf74d5f9c1aad6034c7627944158b
75879 F20101114_AAAQRI bach_m_Page_047.jpg
817f6309e499d85d6a8a29034f448a38
b9ecae75c72ff9c17d990ae0abde28d7d6bbed07
25606 F20101114_AAAQQU bach_m_Page_069.QC.jpg
0c0db60d378896d88326f00b6f5a69ed
90a22900299d7c7717b08664ebe6c7cd7d6d1290
74382 F20101114_AAAQRJ bach_m_Page_025.jpg
a606da3f0e04ffa8b7a6c7e13f1915fc
5ad06e9b27b281f32da7357ba811294667d42aa4
26938 F20101114_AAAQQV bach_m_Page_077.QC.jpg
8787d923c16117597cf5d43a4e89a498
2ab15791ada3b417a7e574b547627009ad854936
1938 F20101114_AAAQRK bach_m_Page_031.txt
f80b39bc53b82eea5555f64155e65657
ed804e5a04ef5365dda5d2ecaea20228bcc3089e
63557 F20101114_AAAQQW bach_m_Page_118.jpg
91c623ddb5c1f28532f667e88c9558f1
9f62c53425ae466f4df85b8ae4f3b1da82951931
F20101114_AAAQSA bach_m_Page_063.tif
a5088a6183518faecf5a153e1b45eea2
669068e01e04f3a1183806cc83590ae47fb3a06f
6796 F20101114_AAAQRL bach_m_Page_080thm.jpg
be9f541189cb7b45b018ae0d4753c7f5
b6bdb89f7a113ec6b00a9278bb03f71694f34993
F20101114_AAAQQX bach_m_Page_065.tif
b6722370734fdcca2e3f46d5fe25c59e
0a7cc9ee349d00bc6f9727e35980a15807da7845
11817 F20101114_AAAQSB bach_m_Page_064.QC.jpg
815b0691b1d059e35ea3aaa9b6c74ca1
f851c7a603540774492faf83c337b1fc87a54bdc
F20101114_AAAQRM bach_m_Page_092.tif
e19107a89c4aff9fb68a62f36b13d070
4ff97244e03ed65db44f368bd55e310523b308f8
74869 F20101114_AAAQQY bach_m_Page_049.jpg
9f51cc66016c3234260c79eac123fe89
13c5fb3737292edf6073961cfc45efdade192cd0
23732 F20101114_AAAQSC bach_m_Page_055.QC.jpg
25e4f190e6c521b45821f025caff122b
55c336b0c916b6b9ec9870b84b8e5802dbec35e2
36850 F20101114_AAAQRN bach_m_Page_116.pro
4a0cd2aabfe48323714ce5412a5da8fa
981c71e72633c4c788e3a02a173e407354c6e7eb
F20101114_AAAQQZ bach_m_Page_099.tif
64017acefc813b1647916358a47a70e7
0523f34eeb039994e8eccb1c3e64c6dfd2598fba
F20101114_AAAQSD bach_m_Page_085.tif
b79fd58499aca3fcacf1511f76e9a11d
a4b0506500e560fa3f87b9f04c7e8c6b86108e1b
24944 F20101114_AAAQRO bach_m_Page_081.QC.jpg
f32ede070e6268ec8bcdbe896b3dceba
a41c2a2253b1dbf0c09a411769795b62661576a0
92160 F20101114_AAAQSE bach_m_Page_118.jp2
7ee3b1e78bc8659c326c9d680a10da80
a23f17b30201fa3825cdeaf95df7eca5f7044947
49139 F20101114_AAAQRP bach_m_Page_065.jpg
ae98c31636c70901637b02a4c1da3bdf
4548d3fcc958224ccb45dab59db1ccb142354b52
80463 F20101114_AAAQSF bach_m_Page_078.jpg
19a91ccdb2a7175572a52a89a8130190
ee5f072efc8bda57704095d36abd936b1bf3a1c7
52429 F20101114_AAAQRQ bach_m_Page_055.pro
e3092ebe331c0a1ee8704919cdac5ca8
9e35e961dcadaf0efd5bc168215df1c4a79c1e57
2160 F20101114_AAAQSG bach_m_Page_069.txt
ad7c546a142f671c0bee5c12ab059bbb
7e991a7bd61e3a559d6e75c778110c4f6af842f4
542017 F20101114_AAAQRR bach_m_Page_065.jp2
2fba1ee9538bd947423005756af12d95
e9fa570fd5bc806ee34196efdc9264535a38164b
199613 F20101114_AAAQSH bach_m_Page_066.jp2
344e2527187824ac5e397a7c8e10acc5
41380ea760d4a651d3a0f595a98802233b334d93
79946 F20101114_AAAQRS bach_m_Page_092.jpg
5d8c49adbd282add557a853e35d43d88
705d7128ba657b86b42897292fbd92e2528cee97
30414 F20101114_AAAQSI bach_m_Page_108.pro
7729d57a468b8c23363bd29b1544b35d
f951ef59938452de318b26f65e72f1fd3c54b787
6769 F20101114_AAAQRT bach_m_Page_085thm.jpg
9b3a1c6b3018531c16bed6603faec193
bec7ae62c61c794e3b88564d01720759fae59fac
2046 F20101114_AAAQRU bach_m_Page_050.txt
2cdf6ba742053dce1031d3113fe4b761
fe5e4681af85a2b58d5da780142812d8a7189140
5362 F20101114_AAAQSJ bach_m_Page_114thm.jpg
e971860ef24ceabbd40d7b83cd7b5c3c
572c851ab757d15ff44cffde5f87a28da3f61611
14267 F20101114_AAAQRV bach_m_Page_102.QC.jpg
b5a433e76659ab3711b5a1ecc4ee6a14
f130632621a76a33126c8deca12a0f2d90b6fa39
85968 F20101114_AAAQSK bach_m_Page_123.jpg
f8057d2a260ebb164d9b11edd2c5f4d2
4286a376526534e113a5a3d19a731283315041ba
F20101114_AAAQRW bach_m_Page_100.tif
4ac47712a967589c7942dc1207dab28d
59e4fdc95d310857dbd731106819a601489f8ce5
1172 F20101114_AAAQSL bach_m_Page_045.txt
82e23ae511dd49816fc8ef5d75b4b662
fa061d7ff4d47624b404ba38debeb972ea9b1205
2074 F20101114_AAAQRX bach_m_Page_095.txt
963fd7ff1d4b7965a6053dbfc047fc41
f75f7f27d170949acd3d5742b8d3c5cf00fd3de5
F20101114_AAAQTA bach_m_Page_016.tif
1b594cb5cb10f9f7f755cb02951e6ee0
0cb713b89882f2bf638d92658ab58e26e00e04e9
F20101114_AAAQSM bach_m_Page_066.tif
277790bd2406268afc344d0a4d5da1be
3646dff67e01a2d29e47bd14930f196cfb215eef
122211 F20101114_AAAQRY bach_m_Page_033.jp2
69c269c4471cff341766e55721bc7546
de63c712b83223c40a1447ae0dec6cf07b5a5cb6
437 F20101114_AAAQTB bach_m_Page_066.txt
d05a6065da8cacabc71a78f0eb731652
fae0a74809881398e23c1042d309e772272f09f3
112121 F20101114_AAAQSN bach_m_Page_016.jp2
7ac13e23776a33f93c7535676d61ee7b
ba7e90aa7f00a25bac93084e5ba38ddcca405b98
2182 F20101114_AAAQRZ bach_m_Page_118.txt
165230f265337ff650aebd8e42741a53
746a6c873bed805187527b7c887ebdbacffd7447
F20101114_AAAQTC bach_m_Page_034.tif
8185cf3555b9a93e7056731888c9d0ca
25a0306771e83723923c9d2504fbdfac0bbaa867
55608 F20101114_AAAQSO bach_m_Page_037.pro
21e60d73512e357de828602cd4535d7c
83dd438338b30be1c4a78bf976ecdf5c98d3d0e9
6341 F20101114_AAAQTD bach_m_Page_010thm.jpg
c87d585567f86c10eab6808d4ddb8ec0
4b34ed32ab55077e704d91d45c3b948c9b955fd9
2052 F20101114_AAAQSP bach_m_Page_067.txt
d7b3e5fd5610d31c9836e556ee58631c
d6446924d1a336aa89c639b32f29d3594a523328
F20101114_AAAQTE bach_m_Page_028.tif
e1d6433ecaf1247d6142f7b2818d3075
a9197b540f2414074e68f09afda13286fbd33ee5
F20101114_AAAQSQ bach_m_Page_008.tif
f49b340999346fa3fbafbd445d7b2e67
89b5cf48b3ec303b681731723847db37bc7eebfe
53869 F20101114_AAAQTF bach_m_Page_115.jpg
847b23a0f21586951b635ad57b8fad02
739d404584216d9579765f8f34b19dd7be7df196
77830 F20101114_AAAQSR bach_m_Page_058.jpg
6590a54158883cd98acba8e2dc7a1919
bcf10024b28f7bec63b6cdb42586f937f43a14e1
16534 F20101114_AAAQTG bach_m_Page_115.QC.jpg
791b60a8e194fadbb4c2a3805c79c0f6
1b6e631e9b0f956c5f4ff5c511343db8cfea10f7
F20101114_AAAQSS bach_m_Page_075.tif
691578d5ff26cae3db14f933b644e639
4681b2199439bd828a7d499b257b5bb822828323
2033 F20101114_AAAQTH bach_m_Page_072.txt
557d4cd14d353d12f3c95a0150c3f8f8
e0532ec4205e2bea247dab73edf53efb39a30ac5
25792 F20101114_AAAQST bach_m_Page_096.QC.jpg
462b59c3209827667dc32c70d5fc77db
864fb72ae865ecf7552b6e67bd5c7503d3e11763
616 F20101114_AAAQTI bach_m_Page_064.txt
2efec194109cb0ac423e2016ada45479
9e61bea823e7d8ed3978f0122a481fd7d0173b8d
5412 F20101114_AAAQSU bach_m_Page_120thm.jpg
1de0dc6ffe5e93f873f0fb52e469c446
358e031b41cc33b7e69bcd9f58b7f495de42a6ca
49816 F20101114_AAAQTJ bach_m_Page_122.jpg
243e682783367094e504700e9e70e4b1
70d302c7628237cb9309a5d0747a43b894e68231
15560 F20101114_AAAQSV bach_m_Page_075.QC.jpg
61bf6663bfb60272df1e5078783ccb5f
4375a73369210db8d0593b3e38bb7f45aecd36e1
F20101114_AAAQTK bach_m_Page_121.tif
c6fb78c0926803a87c1a4b1907b521ec
f1c3e8a022f343fa5ae0c983bef1d7e9eb244c2f
26986 F20101114_AAAQSW bach_m_Page_129.QC.jpg
44163e6e1da08e4aeffcaab0f37adbb8
0015f52e29a77024a34fe387352fa9b81681b03a
8887 F20101114_AAAQUA bach_m_Page_043.QC.jpg
9244f0630392be23af825221292018ba
3ee87032d70e6564907d8ab714fa3b87f6a5098f
F20101114_AAAQTL bach_m_Page_094.tif
c6c60a0e3851cd5b71f6550ee2bcc746
0921065ff46b36e6a605c18805bb58197c4d2896
2747 F20101114_AAAQSX bach_m_Page_076thm.jpg
53c3e4b7bd4f9335677c58c168a28fb7
2e4c7b836749819fc9b312d1cacac8bac2fdece1
36203 F20101114_AAAQUB bach_m_Page_121.jp2
687066ed0a9765d728032b7337183ea7
b26b85bb42295fe7077e52468e38368fc255a61b
2273 F20101114_AAAQTM bach_m_Page_032.txt
2e16c135dbf6de200f0b2c967815142e
5dc7f856fa863ee08f9ee5d55c4295d729403418
75532 F20101114_AAAQSY bach_m_Page_060.jpg
7207e6efcea796fd462071ee27b19901
5fc03819385662b2d04a80b307ebb6f594479622
F20101114_AAAQUC bach_m_Page_014.tif
cca7757b16738882ad55e47e78c9b894
498a938dc3d0bf2bd0ea4177bc96f354461c9af7
4617 F20101114_AAAQTN bach_m_Page_075thm.jpg
7ecb5d3b6106e72172653ee309e35d14
d72dfb2db6eee0834f695856319a0f64d7eb36e4
121332 F20101114_AAAQSZ bach_m_Page_078.jp2
c10533c1b7b56cfc689e3e23a5fbbe30
2ed5a98372f095948ba27edad58ee035415e5260
F20101114_AAAQUD bach_m_Page_053.tif
d4d6964729a35bc67027f875e8206bcf
f2fbd0ed6d20d9a4aaf92f7e48e300461350693a
F20101114_AAAQTO bach_m_Page_037.tif
4b35cd9a20b74f04eb8fcf1c28448492
2c54fbd7070b11b1c5a9df9220834466bb03965f
7092 F20101114_AAAQUE bach_m_Page_125thm.jpg
0f71150f6cb12dbdacde0ed9c7f273f9
cf850f80a9e60a9a7ba9c3e4521b6b54eac77b46
76431 F20101114_AAAQTP bach_m_Page_089.jpg
a1aa07158bae77c487ab2461d8e24287
f157bbba11b1c5fa65e26b27e5a5bccb4bf50add
13512 F20101114_AAAQUF bach_m_Page_074.pro
7072e86c6f9cdcb815c7ff26e38352ee
02ab865d7145ea1ac0bb9138abbda56d547b2ade
129 F20101114_AAAQTQ bach_m_Page_131.txt
530c392bcd2501eae35524ea6284c39a
1c00adf07029ed7e8b2b5de7b971a188cb7cdee7
23896 F20101114_AAAQUG bach_m_Page_072.QC.jpg
b03f0b85fece20bc52457f0590902ef1
bc977b092423fdd391eaabd963dd56fd1c67139c
133128 F20101114_AAAQTR bach_m_Page_125.jp2
36cbf2a3a58c16ad6991350158a9be94
2dce1e9cb7870d634d851d852551b6f051e3ea8d
90530 F20101114_AAARAA bach_m_Page_063.jp2
ecef5d5a9c525eeb946885f309c4f54d
b2ca8fe002472378fa86bb5ee68780e02308aa78
56191 F20101114_AAAQUH bach_m_Page_023.pro
4e06cebb335af9ff47aec0893d4142bc
64486efa71e6ba44f5d4d463c2604bc9f24c7991
46450 F20101114_AAAQTS bach_m_Page_011.jp2
da31fb682a86a88ee20fd1fd76f4b54f
e02479fe1db87d787713a21d260feae332ad1a05
409114 F20101114_AAARAB bach_m_Page_064.jp2
6838003bf015eee3bbda0a4cc5e8c8f2
2b6e8b7ce22f7db0db51160b5f5f45f674871c29
14340 F20101114_AAAQUI bach_m_Page_116.QC.jpg
0448606b62cb56a9af53f2f95ad0f002
e0dab6bd35294203e38bfdee8faabf53e44d006c
2047 F20101114_AAAQTT bach_m_Page_120.txt
5ecea61b433917f55269babd300b7699
197d64df5ddc0f245aac08ad3f0aa9b7e0a6b342
113006 F20101114_AAARAC bach_m_Page_067.jp2
3cb861e44554e9513dca4da787801f38
e56d82e7b6952c96d5c6561d1f5765236d6ace83
F20101114_AAAQUJ bach_m_Page_022.tif
ac2270c224005170fa6b4d11dc6ec3ac
6b889951a70d634f033d7c78189a4c499ef6854d
597 F20101114_AAAQTU bach_m_Page_057.txt
2c55f51262548a532a23fb49b4d0df70
fdb684cd1c78b6a33670acf63dc691cd456e0ccc
118290 F20101114_AAARAD bach_m_Page_068.jp2
a6ccad95b710d1a4f6804341890e12e1
66ebe3f2077f910bc51e27e5c70f02fa0586f333
F20101114_AAAQUK bach_m_Page_091.tif
108f0eb0a0265221eeac8cca80066639
92322e9c67b68fded1051147c11885d6ef0bb608
26229 F20101114_AAAQTV bach_m_Page_062.QC.jpg
9e4f2e6b1f8fc85f091fc5938ed81be9
03c0c6edba7be3e0e7ed7059f1426d56f20d875a
117861 F20101114_AAARAE bach_m_Page_069.jp2
1c09fcf4dd68783d522e5c9f770697ed
61f00869278b573e938e3e9377dff26f46d43811
6773 F20101114_AAAQTW bach_m_Page_090thm.jpg
f827af7307a3a20a79dff652a4a92561
31d957e8b3868263fac0a7e248eb1633879afbe9
110360 F20101114_AAARAF bach_m_Page_071.jp2
e0c6f96e1c7dca934f67f928a9c8c9e6
d567909e4d4e26404cad04165e28298f58d50d9e
F20101114_AAAQUL bach_m_Page_070.tif
874b18431255bf49923c6938d2ff1fad
747d80b6b21dceb72a8aef7ce3de2e6278170638
117945 F20101114_AAAQTX bach_m_Page_091.jp2
07a52389aacec4a31385737167eef212
032d2881cb96d10540e157412fc2a4786a0aff22
433117 F20101114_AAARAG bach_m_Page_074.jp2
6d5a5f56cdc1578aded527624dbaf77f
1cd758c1df17757f281d7f70e57ca4848bb3a4eb
1751 F20101114_AAAQVA bach_m_Page_013.txt
38bac68b86509ded84d0504359b4372e
a5d83bd701bdb663087f534169d95ee207b53957
6659 F20101114_AAAQUM bach_m_Page_018thm.jpg
c04aed60a950b90378c51a4c17326c80
53708cab9926f44bd0098983a7d549840027d6e6
62991 F20101114_AAAQTY bach_m_Page_125.pro
c3221b64759909c8af3f9c596163c8e8
a44a662fb9ec67d3a7a23b8f552f253afba386a2
479444 F20101114_AAARAH bach_m_Page_075.jp2
adf1e3ca52d6bc94e57022dea54d345f
1d50dbbdb6a6daf927dbb5b8e670da6714679491
75085 F20101114_AAAQVB bach_m_Page_051.jpg
ed57f7b66e0dba044868c3490634f4ad
fc83c565ce9cea159a07d3dc7f4c175873998860
6999 F20101114_AAAQUN bach_m_Page_029thm.jpg
d9838cd409cfe9df42ad292f2be3fc17
c27f99877e792f65c52228cd5b15fb38d3962be0
2192 F20101114_AAAQTZ bach_m_Page_053.txt
66c75c1b943b45a2805fc9f5fb3fa673
4a1da3294b36001075deb40de4f6edf57e1992a5
117791 F20101114_AAARAI bach_m_Page_079.jp2
dc03665a399069391acf01dc26c658f7
1da6ab229f72442ccc120e8e7779738479e56bf5
28286 F20101114_AAAQVC bach_m_Page_105.pro
62fd330789ceeba9dab8d8535e00568d
4040542f759d02666452fbd87a9477356a0c9b00
2092 F20101114_AAAQUO bach_m_Page_020.txt
6421792727628f25f766ed53c4ceecad
70580fa4a5eee11bf1c47eb6284acfc428292eae
114206 F20101114_AAARAJ bach_m_Page_082.jp2
e33e16ee4e3424e8731e85410bbc79b3
9ccbd39647b5ceddcf21b988dc3b21cd973cdc57
1336 F20101114_AAAQVD bach_m_Page_002thm.jpg
a0d251938c0cded0338173367aaf8367
e9f3708fa27483524a88f02caf4cd4f62d991a33
F20101114_AAAQUP bach_m_Page_104.tif
a655de68f282bdc63aeef36dd290aedc
ef758ec35167784436491f222aa07703115aa742
115648 F20101114_AAARAK bach_m_Page_084.jp2
a37a8490513b89294616077633c932ec
40d34f18c52cc46497dd1902f6d84aec7c4a4f12
72916 F20101114_AAAQVE bach_m_Page_117.jp2
f76ce29a246c1f1ebfea8aa7b90edb90
58a0e513bc9a9b207c95fe34a37bf2254ac20048
75114 F20101114_AAAQUQ bach_m_Page_016.jpg
ad17c1837778cca9dde0f3f6feb133b6
0d4052b23b1755d25be3a1c7c0b502f8b6d67061
110617 F20101114_AAARAL bach_m_Page_085.jp2
1400f0ef51586d2ba7142ba47a634f32
db4d97ac9ced600daf904cb7c664c6dadadc22b6
58956 F20101114_AAAQVF bach_m_Page_130.pro
69c804d5b33d06c794eeffd0e6b75399
1d200f67d2c7b7eb9419289b973c10a7f1d3553b
F20101114_AAAQUR bach_m_Page_027.tif
ed9a871b1006ec40ed4217207c0a237e
6d3579dace0825ad689f13765fda17162e88035d
64304 F20101114_AAARBA bach_m_Page_108.jp2
662e842241f34128c35f86a72cce73c5
f8f771d261a6d842ad76f9b4e1860ff3c762631c
111567 F20101114_AAARAM bach_m_Page_087.jp2
740c8ebd499fc24686447f994bbec3cb
c3951a7b9eb610d57918150e823fdc914248c602
1462 F20101114_AAAQVG bach_m_Page_108.txt
48248087432b2e9e0c6579db5d6b20cb
e3acb8e38320c2ac0d02c23b10208ae5dc29e0d1
2352 F20101114_AAAQUS bach_m_Page_077.txt
f77d5888718ee88f28894e373942c2c9
2e964a3c3b291ab019261a89e3606b4cac14f07f
116463 F20101114_AAARBB bach_m_Page_110.jp2
bbd3af9e584a0f61361764d634cc928a
a4096110f3422bc2d1ee9c14654b87bad6711408
115711 F20101114_AAARAN bach_m_Page_089.jp2
5e7db3ca9a9730dc9442189733f1a3ad
b9bdbb64c0431416979f372160d49612eff0c05f
110625 F20101114_AAAQVH bach_m_Page_072.jp2
801f036d6ba150b01f24a29bf2c7a950
062e0143b35e0755665e51fc53e22ddae2293770
F20101114_AAAQUT bach_m_Page_021.tif
44cc826f934f6866934e251d1a50c302
f37d7a3c97376f435e8973d0d1a69ddaa3095043
109176 F20101114_AAARBC bach_m_Page_111.jp2
63f5ed8c1408a8c931f67125c0134083
cae668dffe78920090e4a89714296c3396bdc2d5
121852 F20101114_AAARAO bach_m_Page_092.jp2
403d706fac5bfa89fd2002e83742c85d
48a4780fd78897a8e6008443b1088d73df52ad71
27109 F20101114_AAAQVI bach_m_Page_042.pro
15731247cb220377b98bfee679c31fde
7ed0c2a13787e31239e23aa55c79c4c54867541f
24915 F20101114_AAAQUU bach_m_Page_071.QC.jpg
5b58aafb66f11d934fc215e6d021b772
492bf9eaeabdd76eddc092d4525a2852403a2aed
544096 F20101114_AAARBD bach_m_Page_112.jp2
1bf1cb637ce82383ef6dfd93dd4f0910
e509160b0a1cf0b3c9ce34a4de9e863d85001549
117903 F20101114_AAARAP bach_m_Page_093.jp2
626244f5df28b906c9a7bea8be81cec8
7e5e1fe63785a264f5b12808dc2f7a4389877259
23273 F20101114_AAAQVJ bach_m_Page_083.QC.jpg
08c14b669e32e7b4ef922ebc5eb62292
2a34c6edd9a24366630d04bbc0a5e05c56da9b7b
24595 F20101114_AAAQUV bach_m_Page_059.QC.jpg
e8d98dff86793d76576f8941ec08876c
1bd8ad497519dfb793cbf71e405c266c217dd5f9
75759 F20101114_AAARBE bach_m_Page_114.jp2
9880f10f020a68c3deea00eff0c19807
d3ba76504093c4546e79c7f292407b49277f1d6a
110842 F20101114_AAARAQ bach_m_Page_095.jp2
862890c93753afd6c89f694d017dc2cf
e6714abaa4f373145517ea6cdaab99b8fdbe5a0e
308519 F20101114_AAAQVK bach_m_Page_076.jp2
48bf34d0c4648a62ea2021274d271284
01b5af0dcd8b086ff0ae7577a7c9dd6fdfc8ba4b
71544 F20101114_AAAQUW bach_m_Page_109.jp2
c96133b0db7d61d897af252367df26d2
4672765b444b489dd860347663475dab902ae956
76199 F20101114_AAARBF bach_m_Page_115.jp2
e0b7d776793e6d110ff6253812fc9c99
6a552c52c8eeb506fb0c5fdd5e30f8b86eda8d2e
88162 F20101114_AAARAR bach_m_Page_098.jp2
53da29cb7665ae7bb242dd10755e1e61
255e711cea3f301af7b24186e13ccf1419560fe9
2329 F20101114_AAAQVL bach_m_Page_001thm.jpg
232d9e15d9624da48b58d240157baf1d
74d0fa1dc8ca67eb305678c6fc4724e08055b417
89923 F20101114_AAAQUX bach_m_Page_124.jpg
a93857eef707c098348910a13ed1803a
b337237f85d1f49342c5383a57a66fc49f15643c
87455 F20101114_AAARBG bach_m_Page_120.jp2
037902f443b246746e126e59a554e9f5
f17fb52f36a40ea63ed6144f5de54d3bdfc99532
75382 F20101114_AAAQWA bach_m_Page_022.jpg
05695eb89045f4a8ea5011a67df23376
681c181baa185b7887be1cb2e3719bd53bbe9187
87474 F20101114_AAARAS bach_m_Page_100.jp2
b4ca41af76d2937d7f592ff64693ab9a
d1e72a85323d1b09166e2848a582b57d8a35a604
6746 F20101114_AAAQUY bach_m_Page_051thm.jpg
fe6a6ea1af05b6a2f62f842c99b4f608
2c0d293fdf85b970f14fea89989709d0ba6d6cb3
70244 F20101114_AAARBH bach_m_Page_122.jp2
9ee10dc1d3d759c6cc63b1cc1ee4a6ec
2c17393573153d5a38f8ff3bde9f79e176573425
78264 F20101114_AAAQWB bach_m_Page_023.jpg
af2153bcb9ccd76a8d12464195dd8c43
86e9e99c94e9131ef12f69f82ba970bc6d11404f
865592 F20101114_AAARAT bach_m_Page_101.jp2
de7f61e95e85d72822ee610b5715a2ca
b67bb508d81f50b7987922315900251010199adc
1717 F20101114_AAAQVM bach_m_Page_008.txt
6bbe2c0d52b5cbcce30cf741d2b65297
f292b97c5115efad955d8c69f7005d2e2b1b4bb7
83530 F20101114_AAAQUZ bach_m_Page_127.jpg
2f758f6c43711e87c871be54df5c5037
de9d4343faf65f0ce2568c9048472d9c159e6cd9
128850 F20101114_AAARBI bach_m_Page_123.jp2
cd4187e349094c4548b450c810ec55a2
1b9bb51ecf14690c732fa12946be871783163d24
75529 F20101114_AAAQWC bach_m_Page_024.jpg
f2f0ef741835c4309e93643839c21848
d660a25a6c9a6d0bb1e4dfe0769656f718718f8f
748449 F20101114_AAARAU bach_m_Page_102.jp2
9f00e6b91b21019e66ff931049fbd6c9
36d41a5f5387820f360270e5ede50ce733d8df22
51699 F20101114_AAAQVN bach_m_Page_030.pro
e04687f37a13580399d0a821c055cce5
a50b8351fc098de7e323743f9299f047ce27956c
134993 F20101114_AAARBJ bach_m_Page_124.jp2
7c9317948449b3567180b443aef9ed3b
3ab69cc083b5d64ad6cfe56187806b687335e7f7
75617 F20101114_AAAQWD bach_m_Page_027.jpg
89f377a2048b14941ac85906a9a1bbe6
f6833170849f1f224443cac954c76df164a85257
665239 F20101114_AAARAV bach_m_Page_103.jp2
ee521c6e557b151c03b0c30f12a314e2
887cd46ba165b627b70edd37601b27f50f61595e
195894 F20101114_AAAQVO UFE0021222_00001.xml FULL
5879d7ad15b9cb8a794a75ac8e4a5ee4
d759392271c1a979487ffb7826d7fc0bdb1161a7
126956 F20101114_AAARBK bach_m_Page_128.jp2
5b1653272fe139673cbc426f4be875d3
cada5a8a91ff8e4f89d449c58f6af46d05f87f69
74748 F20101114_AAAQWE bach_m_Page_028.jpg
39c926b55e0d5a6bd1af7751b5d82ed2
ed6a350ea1a039e5d950089bec3861d7a4aa19c7
124023 F20101114_AAARBL bach_m_Page_130.jp2
52615d8e8283d31e8314ccf3187cd851
7de6120bca448accf851028f9a568b699e9f6289
71408 F20101114_AAAQWF bach_m_Page_031.jpg
b442bf9ec54c327777824a2fc5c67fad
75b70abab9e82c26f9c7f7cec16b9e59e1032cfe
651937 F20101114_AAARAW bach_m_Page_104.jp2
1becbf75184f96cd78c55a6dae376ea6
16c6b5dcef68d95978be2c3e4dec2d19e27b5b37
F20101114_AAARBM bach_m_Page_003.tif
ab8acbfde357bd2e7eda6f8e697a59c6
4435d06700046223eded222bc0f379654d226f8e
79334 F20101114_AAAQWG bach_m_Page_032.jpg
cf599f536c2b6f0a793266346a09b3aa
a01d7177b69aa0a17789240997b3b96899b10458
62093 F20101114_AAARAX bach_m_Page_105.jp2
a4524e3775e5337715773336bca169f9
f624abe2ea4d29063389dae989205dc7b31bbd27
23874 F20101114_AAAQVR bach_m_Page_001.jpg
445d8e6dcf12c3d47ba6cf6b41fc4ef7
f241ea560b948fc0eaa797c28da38a9eed0227b3
F20101114_AAARCA bach_m_Page_035.tif
ec216abd024e9c31e5ea43046898aec4
ceb12bb62bf045144bc5be5c2ca245ddb063d410
F20101114_AAARBN bach_m_Page_004.tif
f834a99e44473127ced67cf43f1d5fed
9f73ab5f33bb9407cf52f6c7bf494480e780e890
80494 F20101114_AAAQWH bach_m_Page_033.jpg
114343ed162fd6e59e369a92b13fe02e
66e36258a8f15cb88a3b3e4b3cb27a1ed49a33bf
46940 F20101114_AAARAY bach_m_Page_106.jp2
bfb352d15ab90ef969eae02abae06454
a45efab187d9cf4a5d0e770cc778fd89e33a83e6
83757 F20101114_AAAQVS bach_m_Page_009.jpg
df75f7f3aad21c4d362adaaf5440f788
317ce278cb4f25bc54ce3838a627ba8bb1ec9e27
F20101114_AAARCB bach_m_Page_036.tif
d0c68cfbf289fe8d931185a21ccfea52
328aa8084b1c5d39747e850204993bb15619eb75
F20101114_AAARBO bach_m_Page_007.tif
a34ac9e62ad67d6534d9b42fdef81eaa
6e6660f52bb70bc295c4ac1e4a2a5ec2827a358a
78992 F20101114_AAAQWI bach_m_Page_035.jpg
c68bd328436cc827fe3d7ea747adfd4d
1916738b53315ac7d4002836f394c664dfaffbd7
74050 F20101114_AAARAZ bach_m_Page_107.jp2
ef8687d2c6e872c6333a004473583a44
64d089d782a68e94fe2a6239f8ffa8481df5a2a1
69661 F20101114_AAAQVT bach_m_Page_010.jpg
10085d38137d4cd3d58072b2e46076de
f2e95758573877e0d7c28c39df18fdb3f32137b6
F20101114_AAARCC bach_m_Page_038.tif
92c37c2f4d6b86357f733cc2a8aacb35
deba6d031bae879e7b01f30f041f7926d2e07c1d
F20101114_AAARBP bach_m_Page_009.tif
a3177beb0240ca254621ec87fe1fb38f
ada30361ac11bb874fc2de4cd2224ab0137e6968
67065 F20101114_AAAQWJ bach_m_Page_036.jpg
062d0f24dcc4c03223e32f8bafd8e520
bfeb58ed734b509c706a976d1a196e0de7dddc2d
33730 F20101114_AAAQVU bach_m_Page_011.jpg
b1f5355f69a730b443d3354ea93f9b7c
79928de9e9d232848fddb33d5f05d6adf9200854
F20101114_AAARCD bach_m_Page_039.tif
bd39e97a83f8c1dda722f4035312926c
3c6987df4cc7f6d336c9abdfc9c8353043edb1bb
F20101114_AAARBQ bach_m_Page_010.tif
bca4ba160b83d9c6e5c221441d283151
f2a691a48ddddcb01c4ec17e98432bacb4f2c146
79545 F20101114_AAAQWK bach_m_Page_037.jpg
7f0782fc96474d11bea7f79c67e0e773
954c67793b8f2e427fe317c80bf72a645c5721db
63812 F20101114_AAAQVV bach_m_Page_012.jpg
1ec754e333b69457eeb7220186874e44
ab883ed76b894b31b4c8160bee3615be828b0dc6
F20101114_AAARCE bach_m_Page_041.tif
26d5a8f3e51ff1157022941f311ec145
ddb7d91fbb447d0628315b32f99b21658051a73a
F20101114_AAARBR bach_m_Page_011.tif
b6939d9a7f3c7d66a6e4595ec871d290
e7d05b3378944606100ca9d24a1d4dcb48789a9d
76337 F20101114_AAAQWL bach_m_Page_038.jpg
fc533f23f77696d7b089998f743f439b
f0bb2a38a51aff3f7ed2c4d50696e371edbc31c1
76805 F20101114_AAAQVW bach_m_Page_015.jpg
090dc09b78a291cb39b0c722b3f3fc8f
b3f9cbc32d5d7438d353d97069d651f11b68ed96
F20101114_AAARCF bach_m_Page_042.tif
683ec343790a076bcb2050fb6ce52533
5a18640dc4c51e4cd05395d44e760336d64d5f70
72865 F20101114_AAAQXA bach_m_Page_059.jpg
b6907eb5e51b572626f9afac51fd0d27
bb5bd380dc2cb7fac679cdcc7dd7fe7184db4e84
F20101114_AAARBS bach_m_Page_017.tif
6de669e67a504278d6511d2ee464f9a6
6759a142835224c113991a069dc33e89cbc394a5
77887 F20101114_AAAQWM bach_m_Page_039.jpg
ca544a09cc9edb6aa8f63aad8771cfcc
8b2f63351a1769652b00d97d05c2efe795ec2e02
71150 F20101114_AAAQVX bach_m_Page_018.jpg
e0c003af97a7bc3d69e25c9499049f85
ec29b5e6c363716e9db024cdb54b9a33ee10da96
F20101114_AAARCG bach_m_Page_045.tif
139f72f677f308e507ce5ac335fa67cb
b3caa2e6140e47debcf73575565a5cdc16efa682
75680 F20101114_AAAQXB bach_m_Page_061.jpg
9e5e82266b6df974d79c053c7c41f5c8
a1383e5a0877972b6943d735b50dd2ab41d82be6
F20101114_AAARBT bach_m_Page_018.tif
4a5b2aafb38d6a8bf0d5f1615ecb395a
764be981ce79b18e4e98a3b2a2556094ec130813
75493 F20101114_AAAQVY bach_m_Page_020.jpg
79b78b24cf51ea71781fdf540edf1797
360005c7b190b158796c941d3261c3043a6213b7
F20101114_AAARCH bach_m_Page_047.tif
864b1f5ea44b5ece9241db6862f29b57
acd0cbef9e7213a2967a732e5b3b7533663b70e2
61225 F20101114_AAAQXC bach_m_Page_063.jpg
f326f4603c43677671ece68f2e095b40
7cb98ea3c954cadacff245d6f87b820ad3e79665
F20101114_AAARBU bach_m_Page_019.tif
99ba75f43411d193d73c8ac9b677f02d
cee53c598beddfb8b0800b07d70b27c06ff031ab
80061 F20101114_AAAQWN bach_m_Page_040.jpg
606d9decabf5366adf375e4a56649210
e9ee7d9294ca0521fe8ecbb97fbf90c5e6fe24f3
77015 F20101114_AAAQVZ bach_m_Page_021.jpg
2867c7ad015a26c081040a7abe490400
4839a52c7d85dae6689f44f506767e086dd1930f
F20101114_AAARCI bach_m_Page_049.tif
b30366b1c91b1aaa1fd7e3047c71b0c9
7f1b7482c17d5c8e84b4bc02e65940a606c338dd
24525 F20101114_AAAQXD bach_m_Page_066.jpg
6d3ae0e4288423c9e380cc261638795c
8ee527a4995fb710027883c187868f5f74e101a6
F20101114_AAARBV bach_m_Page_024.tif
e7f23400f32def056147e066e677f1bb
7c011a0275c2f1624f2e016351eae9df42a9054e
61372 F20101114_AAAQWO bach_m_Page_042.jpg
29d7082c75771d79776ef34e7535aa72
54b9fc6ffaecf68473965522cdffe7feb8d5133e
F20101114_AAARCJ bach_m_Page_050.tif
4074003bdfaecf8a5eb328ff3dffd054
606f7f8a85e547fe4ab37bd5ceb6ef92595baa65
76650 F20101114_AAAQXE bach_m_Page_067.jpg
20d87bf599420ea525e6d5c70d98994e
1e94b9fd73aaac3b71a9a3b39d8cbbae47b02e94
F20101114_AAARBW bach_m_Page_025.tif
edbd8d8735f176e4da9ae54af7455c08
c76062dd61549d07e776fbc25c58b92f9c48c0c1
26415 F20101114_AAAQWP bach_m_Page_043.jpg
6f3f6e5d39963083f44dd44c91f07e67
c6bf9fbb0525481814a0f193f1fe6831e68a08df
F20101114_AAARCK bach_m_Page_054.tif
428d3965585c67711eea9e8dc47124d2
1461a5ccbdf8ee84b1a591c9f9dbf8051fd4617e
78478 F20101114_AAAQXF bach_m_Page_068.jpg
72052b804fda91b39a8599292721c6d2
de8fd4392b849db72ab9014eeeda7864b6d70f7e
71486 F20101114_AAAQWQ bach_m_Page_044.jpg
b61f1c3fd651dd7b19b109965c54d44d
39042d32476fff17ce37eaed322b423f9c25c1d5
F20101114_AAARCL bach_m_Page_056.tif
d4fea9ce5cbfce5af4b6bf8758dfb0d1
85756f41ca4b0a8cbd4f549887462051d3b37bd3
73191 F20101114_AAAQXG bach_m_Page_072.jpg
b1efe6b04f77aec77e640ccfadfe3491
0ab5cbe9167b4d9e6c7f859c03c792018b76d7e2
F20101114_AAARBX bach_m_Page_026.tif
417ad6b903dfcfd7f920ad8659273297
84af2dac9b72d33c3ab87b3695a4800f2a377006
43419 F20101114_AAAQWR bach_m_Page_045.jpg
e3205b2e660e906090eab30c6f970efd
c9e4f8c10c8c040716e0aff185c1b3bdbd9ef154
F20101114_AAARDA bach_m_Page_089.tif
6d56eebc50ea78a2dc9a59f9eecf179f
4210afe492fd4d5e81f1063d111a4b77c2df0b4c
F20101114_AAARCM bach_m_Page_059.tif
9ea5c586fa0b258d474df2eb3eefdfea
3d1f7fc5eaedee34352a9d9dbf8ed0ab5f8f7add
65163 F20101114_AAAQXH bach_m_Page_073.jpg
4059217ea18238a390976687b4bd8f75
46d11c8476391d994b6285a1cecb5c5a84959053
F20101114_AAARBY bach_m_Page_030.tif
81403252ba7e748ee419b5a5a59891c2
d983de0e97bca2e16293ae2b1e230a588d1b53a5
66636 F20101114_AAAQWS bach_m_Page_046.jpg
d7a00eacaa4eb64240bb0295f3dba54a
725285659e36317750d9c28129260c54ee7d1a45
F20101114_AAARDB bach_m_Page_093.tif
c445ebe1214e8332c55abdecc1b9c720
f9bd3b4ff04b4349b1fb9cbf198af70dda314bde
F20101114_AAARCN bach_m_Page_061.tif
6f7d02a9499e8376f68fdb455db66912
925dfeed16d72a9273152f19c194e47a1734da89
46761 F20101114_AAAQXI bach_m_Page_075.jpg
0352c543e64d855f998d52d6bd5cf5bc
dcca4a5c84800868e3bceb6e44d10338a501ed5f
F20101114_AAARBZ bach_m_Page_032.tif
8fa775383fe7a3e29b7a9233fc9f94c5
366d96a7afcc99e0b3d50f92f19c6472dd70633b
73765 F20101114_AAAQWT bach_m_Page_048.jpg
0ef3a74642a8c88fd541b91d2e6e1234
bf42c27f72404f65b3bb7698905c3765bec1ba5b
F20101114_AAARDC bach_m_Page_095.tif
8515cf55c8d291ded4b3846ba14d5185
b084c59389909ec3071348b860b39205b0ab50ce
F20101114_AAARCO bach_m_Page_062.tif
6eaa5b7ced76359aeb7dba1c2eac6e0b
59745baa2fb557b46cf026285e23e187ae8553dc
23987 F20101114_AAAQXJ bach_m_Page_076.jpg
69ac5a37e55dd429a0a9ed79be1c9b1d
40e818dbfd76902062f199524c571caa2d929491
75020 F20101114_AAAQWU bach_m_Page_050.jpg
4d9646783a82262e09ae79064d1900b3
310df9cecdc5675216863f1e21cc73e7eda1f068
F20101114_AAARDD bach_m_Page_096.tif
682fcb35d2384050a2ea6617848f9603
fd4a6d41ae23d5ac8d7b2b6efe92e0b17a6724fe
F20101114_AAARCP bach_m_Page_067.tif
a7e9f2ff21d1dd7742221ddedcac552c
fed04c4516752283b872f55937cf892053c67f9f
83262 F20101114_AAAQXK bach_m_Page_077.jpg
14e729cc3c94b5d4d2c4e7a2f4011e19
6d67664ea01e807fd9898d2d93f37946a99680de
69937 F20101114_AAAQWV bach_m_Page_052.jpg
be607bc12aab3f7f5f1e94c1bf4db179
1b7084d17482b7c03ea6d24c8dbacb7341ceffa2
F20101114_AAARDE bach_m_Page_102.tif
0ec4ec75a564946491a4535566ccbadd
832c72c0bca1d5991426e20cdb5f3bd81014df52
F20101114_AAARCQ bach_m_Page_069.tif
3da8c78cc9a7d703586a1368485c7213
92a91e94eea943cfcf735e72e7392a92c62dd7dd
77530 F20101114_AAAQXL bach_m_Page_079.jpg
5f18ed24972e46e34d65d9cb6aa8f46d
5e3161be6bcf8ea9e6cf62371a1a931bf2f5e248
78080 F20101114_AAAQWW bach_m_Page_053.jpg
8d74c023117d138e2ef319695bebd407
7c6324b2cd11c33753003ab8db1e48a0a3885087
F20101114_AAARDF bach_m_Page_103.tif
5b27281864d62aec9370ecda92660497
7d22dce314b1f237277d7cd251f51db154d82f4e
F20101114_AAARCR bach_m_Page_071.tif
6ce604612a357c79133481a329338feb
7954d323ceeb100bff3fcdc9a65919fa13d57f7b
75589 F20101114_AAAQXM bach_m_Page_081.jpg
0b2718dcd02b8adf090922f2de573bc4
117e35ff3d52f1b744d4ec485e99f70e959cfb56
75268 F20101114_AAAQWX bach_m_Page_054.jpg
2bea95f5ec7cc028ce3f6ea7f2dab426
98fa7c8e3f17e3c46b9b0717082c2237318dcc70
F20101114_AAARDG bach_m_Page_105.tif
8c0396f56e3a679e751e3e5f4d6947ca
3bd7c7b262cda690be5a4e9247d8de2271fed6a0
45284 F20101114_AAAQYA bach_m_Page_104.jpg
53ff14961af24f6925b64c13cf223c5a
03393c8c9e40cac8f73972675d265069d226651b
F20101114_AAARCS bach_m_Page_072.tif
25283bc0253b497534b30843d0c22322
7fad950c8abfb82c18e8bca5d4d5812a1bfbf32d
75707 F20101114_AAAQXN bach_m_Page_082.jpg
2575397a0a7c1ade129ff7a262a6d403
f060616371434deb926e9d6a80cf48b7ce0fc53f
72503 F20101114_AAAQWY bach_m_Page_055.jpg
7d1891907895448f24714ae4ed7e73aa
67659f46db6b7217521f80d1a9a8653fbbc6f022
F20101114_AAARDH bach_m_Page_108.tif
68d7dfc559004744e321e8fe55e6a0b6
7e4730f7ff9715b4b54f0070bcc9d43b35c90e00
33582 F20101114_AAAQYB bach_m_Page_106.jpg
b526e3566dc87f56958882e310f34ed2
52fecb457237c8ace533ae7ec5eaba95e3261b41
F20101114_AAARCT bach_m_Page_074.tif
e32a2ef0c8621a0117910146a194da2b
fd94c84ae2360d1d03ab607200d264e782717da3
69421 F20101114_AAAQWZ bach_m_Page_056.jpg
e9969ca7298475e1182a0b232de42c52
87c5c3847e51e887523a88ccb6ef2b71abbeb14f
F20101114_AAARDI bach_m_Page_110.tif
f0228759e696fa01b12117f7b831f452
7919e341d67f359d34dbcb040fcafee9f77670ff
51670 F20101114_AAAQYC bach_m_Page_107.jpg
2a859f76a81ea87fc72dbbbd0ca04ef3
c5beb445dc264009e83279fdad44ec167b6a7195
F20101114_AAARCU bach_m_Page_078.tif
75a36fac8fcbe87b61a36d3fc4d014a5
cd10675f671504a7e0e5057384b6b3be339364f5
72183 F20101114_AAAQXO bach_m_Page_083.jpg
16e166e14351c0233848dd768bd0bae0
d6c545fbc228d832fc5211c706847484217e2bce
F20101114_AAARDJ bach_m_Page_113.tif
ab6bcef59a752c1c52c7afc11aae49e9
8cca0de08ebc28cd73d58dc9a018087ad1b57303
46794 F20101114_AAAQYD bach_m_Page_108.jpg
75c481bf980651e6d5c14c3a20d48028
bf919eb1c757d48d324812250d66b114abf7a98a
F20101114_AAARCV bach_m_Page_080.tif
e35fc841abdbf94baac3c58736a93644
4036be6940ec61b3da4b5a4e925d01846f296b31
76657 F20101114_AAAQXP bach_m_Page_084.jpg
14fd4676ac7dfe70d43d562be995344e
47b106419053b4b511f5ada70fa9c3bfa8ae8435
F20101114_AAARDK bach_m_Page_114.tif
13b04ace965409062c28056984221e70
f40964c57a2c4602a21a6b569d07c9bc30e668e8
77994 F20101114_AAAQYE bach_m_Page_110.jpg
e253b277a672241639ee405382efe1ae
17080382ef4fee1391fc33a0e7547569641440ae
F20101114_AAARCW bach_m_Page_082.tif
4940e641fc2024323eb64a73dd238a0e
d47e50b5061e1119246e42d805845b25d402dd9f
75517 F20101114_AAAQXQ bach_m_Page_090.jpg
a62d8ab9c8ceb700a280f80a965a91c5
8cfb1016ea42f2c1fd73d6702bd252af83913167
F20101114_AAARDL bach_m_Page_115.tif
854488e3cbb989f8f9f78080d9d6d610
dc73617c5d25cfc24657c9b67b8cc947cc788beb
48576 F20101114_AAAQYF bach_m_Page_112.jpg
380e38c25ca3fd7f16f02fab2ed24040
aeadaaa562c5a311e072bb48b5eae524f84356da
F20101114_AAARCX bach_m_Page_083.tif
28170300cbe30395aabf8f02db40feaf
82992719bf3ed0f09db7e6ce073e04f52a683b00
77331 F20101114_AAAQXR bach_m_Page_091.jpg
2ed13e72d32e378d481946cf237e8963
deb13ab5e251e1b58c9f44cfa13ce393c10d7bcf
41981 F20101114_AAAREA bach_m_Page_008.pro
4554b3f209cfca6c8c5887e150e09bc0
54d6c433a0d04f60adf5ec584ac2d0c72f8e35de
F20101114_AAARDM bach_m_Page_116.tif
4392ffe4d83d0ef6f75e2b514cec04b9
b73915c0784fff174b8c269e7b1b4cf6156be5f2
53407 F20101114_AAAQYG bach_m_Page_114.jpg
5b349bc59bd826cc0fd722c4dc04bc2d
509e23a98f798e17fe05d62811bc8e7407c1b09b
76060 F20101114_AAAQXS bach_m_Page_094.jpg
7b22337d70c2848487ba4399b0d76c4e
6c7c377595725bfb024afff8f47ec43a22d9b48e
45699 F20101114_AAAREB bach_m_Page_010.pro
fdaa706985848b08a10ecf87a3ef5321
2aa9725efa05c7277dbd86080f3cf7aadf7cf7dd
F20101114_AAARDN bach_m_Page_117.tif
4bf38431b740cc2bb1928790538d0446
99f53d2ae658d13c14e24a8dec330f561720c6b1
46908 F20101114_AAAQYH bach_m_Page_116.jpg
7c65fa8b9d61c6112fd2fd85895ad463
1a37e5c42e3c89c8e851cd7d9584107c74fdebf1
F20101114_AAARCY bach_m_Page_084.tif
c58fa1118ed98f8ad11043b923f99136
d695aab1d0a384dd32c7f5399a387abf52ba7ae1
43709 F20101114_AAAREC bach_m_Page_012.pro
2602d3e1f3af0d4deb9ec3cee09545f7
708363f39c8c640e27b40ce111e1e903653a9a74
F20101114_AAARDO bach_m_Page_119.tif
58e9fb4c4cf808ac3813099ae3a7317f
d0132d84d80dab1c62dfa9c04bef7472426a80f5
61231 F20101114_AAAQYI bach_m_Page_120.jpg
7183d85d5b8dad38205dfd51ec70b97f
74453dd153a3f15c947a488e29976a88d61cbd36
F20101114_AAARCZ bach_m_Page_087.tif
e98d00f84c18c67a80044847a4134f5e
54f57c9de1f1a71eb578f3c876699a9ce7890103
73756 F20101114_AAAQXT bach_m_Page_095.jpg
98eae5ec00ea11aef361402d25ef77a9
311d6eaa783796f6b7daad58ac57b9b27552fb30
40414 F20101114_AAARED bach_m_Page_013.pro
fe75bfcd192a7cc19878bf0acc0e1784
a72706575df28d4c6f5509f50c63d12667b7efbe
F20101114_AAARDP bach_m_Page_120.tif
63911d4b893deac95b51f756eb251390
edb5f09978af8cb2ad5f556596a7311d135b430b
29050 F20101114_AAAQYJ bach_m_Page_121.jpg
5d5ab26f3dba4a2dfc6bd5a4fe08fe6c
30903ec77bf04b1313ca2257cf17aa50f666fbd9
78815 F20101114_AAAQXU bach_m_Page_096.jpg
05b090fcc4edf4b57bd58ce5a715a01b
dbd2f632f1378a7eaad036ae9029d327a222e945
54745 F20101114_AAAREE bach_m_Page_015.pro
b6fe1bc26b483ffed2ee8d62cc3ccdca
25d2b57188d7159bd93d082bbf7f649c19bb1e50
F20101114_AAARDQ bach_m_Page_123.tif
b29e11a74ab0d866e164b3e06a538568
51ea243fb71309b21d47ec576074cc5b5968e892
89058 F20101114_AAAQYK bach_m_Page_125.jpg
4e46174cf4240a9fd1860b18c30ce691
c2417a28585568686660e1c00a8b655fa3f7f613
75304 F20101114_AAAQXV bach_m_Page_097.jpg
cc9d1767f4037f232dab22fd0b8c5a9e
4eb965baf762bb378a884e0af593e088a5eecb48
52321 F20101114_AAAREF bach_m_Page_017.pro
23fafe9e0172d9cac254ca801d8fc05a
0d10aa282c9e9a417b53b0c882a2cd8476813c21
F20101114_AAARDR bach_m_Page_124.tif
494ac03c6360786d2375e01ca1d6a3ab
4b3e73b8f3eb891f6d975a350e9693e37ec1fa61
85254 F20101114_AAAQYL bach_m_Page_128.jpg
fc0d401e2233aa63a137fc9060097071
42ff4b79bf243bfd1d91a65f93575c05167c73fa
65461 F20101114_AAAQXW bach_m_Page_100.jpg
0f2b2bf518d1231dfc6df3b94b6e50b6
7046cbb51b53d267277c9dc5a1f52d6b819ccdc1
50522 F20101114_AAAREG bach_m_Page_018.pro
07994fa42cae367b8007a53c5f24c049
d4030049edfaa3688ea5f792812f5fb98853cc88
120809 F20101114_AAAQZA bach_m_Page_026.jp2
24acd147b76030d461e892f98ad0e48f
3b1d607e696a33428d9d184d58240e758924def0
F20101114_AAARDS bach_m_Page_125.tif
f4a63c51c41cf08bb1cbb7a0d16107a9
16e931af99cdc260dca6ab197ca0461f55e4ffed
81821 F20101114_AAAQYM bach_m_Page_130.jpg
9c29301aefb073fb9e1b5cb5f33765d9
ee28f25aa09e6079c9d89ea35f2c153d85d6a6bf
53025 F20101114_AAAQXX bach_m_Page_101.jpg
b6fa065a620b7e34d04ef7bcc81ca308
af7f9ab632fb97686529f8c856e787b2ae54cfc5
53072 F20101114_AAAREH bach_m_Page_020.pro
0eece67893ad9be894484f3fef6a2693
cbf6fbf7a9bd9fbe3984c9fd761539d1fd9b66f0
114660 F20101114_AAAQZB bach_m_Page_027.jp2
8a0c6a16c774675fdb44cfb3b53536fb
57b047d92ddc5cbf24498691ff4fbd7cb2f9245d
F20101114_AAARDT bach_m_Page_126.tif
5227afa63cbd90fab26fb9bf02272fa1
bb15c339ff1dee67e7089b7b3a44c541680c2ad8
12815 F20101114_AAAQYN bach_m_Page_131.jpg
27425e57a044a0c0d9bd1cf744b91d88
7c0e268896f327cc4154f52678d88c4d11e0c5ca
49536 F20101114_AAAQXY bach_m_Page_102.jpg
1518fd76af81e8ef6d4290ce429dd8db
d632fe89ee5351747bbad5edb1c541bf7e6a9b5a
53849 F20101114_AAAREI bach_m_Page_021.pro
ef0bdb781c804231535458ada7456717
f3dcd20900bc36cd256d5ebb99006e63beaa68cf
114416 F20101114_AAAQZC bach_m_Page_028.jp2
5f110a67dc960cd112048b8a3f44489a
525200be5d3b27fb8ed743c159bca22f87605b06
F20101114_AAARDU bach_m_Page_127.tif
7ab200e55ffd3af322774563edfcdf9a
3afe2dfabde7cd3188c45421e78cd9c4561dcfbf
23702 F20101114_AAAQYO bach_m_Page_001.jp2
536d09a07610ff1e809ed6c91064e419
42aa9a7a1ca6bb87919fa131acd1db9efc8650cb
45127 F20101114_AAAQXZ bach_m_Page_103.jpg
60bb7dce5a0e0cb7ab2d46da0e86a01e
04ada6d7560178c5caee66d8df4e532e693df3ed
52979 F20101114_AAAREJ bach_m_Page_022.pro
5cc2e6f558bec761a4f0809d91f3b4f3
38e6fdc7c277d1939fdcfb85afd2e1c3b1c3034d
122188 F20101114_AAAQZD bach_m_Page_029.jp2
aa736454c9bba058c060a0ff1f727a5b
36ab9d91d589313762e8dd7e9a642f1b3867be71
F20101114_AAARDV bach_m_Page_128.tif
52e4fec89651612fcce020043c616b1e
0f9416debed6fc2f643f3a6b815ba65c36456f3e
56514 F20101114_AAAREK bach_m_Page_026.pro
f9d27e79d928dc9f1472966cf96fe4a9
882da9197d9fbc9a424add4565bc35afa2232683
112853 F20101114_AAAQZE bach_m_Page_030.jp2
b05693d715aa05592884a97cefb5b1cf
4a07adb5e8e319103c1f8bfc974d7022927fa4b0
F20101114_AAARDW bach_m_Page_131.tif
bc9b55445f07bb9d238c8454fc293f2a
11050630f6f6ba5dd9171dcc90ca035ebe448e39
5832 F20101114_AAAQYP bach_m_Page_002.jp2
22a9a70345bdc35edcfda6d5814ef1b2
0e420c4caf88cbbaa6d3fe0496c061f632ee7247
53083 F20101114_AAAREL bach_m_Page_027.pro
059a6d088aa750815bfced51a5a0cb7c
bd1844b2c2e81f710a7fcee645ba5ff48360236a
107432 F20101114_AAAQZF bach_m_Page_031.jp2
f70c0c4f75873727ae9957340778eb7d
cb0fb16ef92276b42eaa436cfeb5473adb03ea7a
F20101114_AAARDX bach_m_Page_132.tif
ba9a4fd6bbb22131bd67287dab135f7f
44138c89f6827129112fe3b9e3c1d1ee5f11ccd3
1051981 F20101114_AAAQYQ bach_m_Page_005.jp2
6db40dacd866fb17f1980c6d2ac33970
d7fd9d7bdf254b5e2bfea0a5234211656ca348ec
52666 F20101114_AAAREM bach_m_Page_028.pro
796e6a84f9c2a1b083503cf54ccf3c9e
12607116042aee3ecec5b45920b29716914bcba4
116905 F20101114_AAAQZG bach_m_Page_034.jp2
cf38b490350cb5da805a2c7cf73d2131
4f534fd6b48cab7656ef8e0e3a8185308f06a7ae
880 F20101114_AAARDY bach_m_Page_003.pro
9a4dcd1e7cb70ebe54ca11fdd3e21218
3b7afda452d5c9459a458acffabe5388ee58c289
1051971 F20101114_AAAQYR bach_m_Page_009.jp2
48196755a85ebe4ea085ee98833663ab
a9720172fee88685b18b832565a909aadb0db29a
52675 F20101114_AAARFA bach_m_Page_051.pro
9caf617725a01cda0632c984ca8b8b54
aab98d8323f0b2a455e4d7023e52408787d83df0
55828 F20101114_AAAREN bach_m_Page_029.pro
722338006698d9f7ab8cbe6041f51aa1
f0d6ba45d4ce2697c0553109fba2a95e9a6cd564
116186 F20101114_AAAQZH bach_m_Page_038.jp2
90883de0b15652ebc702d88cabafffbd
b48118711730f71cf08721f9fc58965de4e780be
95513 F20101114_AAAQYS bach_m_Page_012.jp2
6b81004dd4728f4ec72274a096e99c91
1e5b87af95d825e42758122480e5a97e25c49e90
54276 F20101114_AAARFB bach_m_Page_053.pro
105295ada570ffce3d702e6098514a28
a09240d157222e082c2c083ab7cb5f853ce89547
48912 F20101114_AAAREO bach_m_Page_031.pro
5b8e2a9b8cbbf690a67e77fb10542c50
fdf3c6fa78623e479424d64c3447421d3724f502
122026 F20101114_AAAQZI bach_m_Page_040.jp2
24386dcc29b85345476e649a3c6a47bc
cf7d0e57f9e481ca425e4c7a5989f1b81aa67955
73865 F20101114_AAARDZ bach_m_Page_005.pro
695a2c80e3fc7824a139d649d3059da9
f590d87784ed08d5bf398300675bb605b8b5af4c
117477 F20101114_AAAQYT bach_m_Page_015.jp2
699d28e10a1834df753658a4b40719d9
d862092328e179cf10ca0ca26d97beb002adda08
13398 F20101114_AAARFC bach_m_Page_057.pro
61285a7aa026e442f2c0a158f71be4b5
ab608560300416f94cb7efab3179396883e259b1
56994 F20101114_AAAREP bach_m_Page_033.pro
ca4a2c50bba7a086840709aa0b63d50d
4bdb3cc5a3eb593517f2796377cf8bdef18c2386
114445 F20101114_AAAQZJ bach_m_Page_041.jp2
d848085ce4c9419bb3baf380948b0df0
4f08da2592096706141699700134611dd61327d9
113688 F20101114_AAAQYU bach_m_Page_017.jp2
e9ac4e2e86ac8fe323b6c836a452b524
751b143571d575fcfebedd2ef37c4765b510c5dc
52422 F20101114_AAARFD bach_m_Page_058.pro
12c22120ae1db0646bb013dc3d622abd
05a9c3f6541572832ab7c19c92f9f90b1897b613
54714 F20101114_AAAREQ bach_m_Page_035.pro
81a2169893ddd688aa70643cf0a2277c
749a4b96fd83aa83788a3e156933c9fbc312f5ae
334201 F20101114_AAAQZK bach_m_Page_043.jp2
4f142cedd783ad40333906622a30fa18
e1c3b0ede3ff37b4e9e6a58211583ff6ff902f7e
108994 F20101114_AAAQYV bach_m_Page_018.jp2
dc3724bcf78b453e0669015a150e285f
62c6816d5b9e8e8855a81850ce01327f7dcc9cfa
52629 F20101114_AAARFE bach_m_Page_060.pro
143b4bd8293321925d09b74c5781d914
65c7096d5a5982bbe96efdd5890b0e99ac1dbb12
113295 F20101114_AAAQZL bach_m_Page_044.jp2
7f5ed5d8ca8dd60bc186c2e135559c6f
09018107d4df05f4b0e2d1024254087df2b678ad
119524 F20101114_AAAQYW bach_m_Page_019.jp2
c9cf2cb47ca8b35686623b2bbb9854d7
7ed2b63e4afae9418a8ba15c130e750472ab0be0
52056 F20101114_AAARFF bach_m_Page_061.pro
a3a3df4dcb02d95295789a4604513d57
f23e9b2adf40bc97d0aebc7050384f846c25b029
45619 F20101114_AAARER bach_m_Page_036.pro
1f91a2fb2f930f246c21382ff168b49b
d76f42de463c3d33c0d1f0cf43cde086bcc5158c
62835 F20101114_AAAQZM bach_m_Page_045.jp2
0de10882aaaef95df545b7d858ba282f
2e80b1371d71808de1271106d2e971bacd667bf1
117243 F20101114_AAAQYX bach_m_Page_021.jp2
2d07e6413386ed3647be49266195eb6e
9ee4a520eba4d0e58f8dd41c0b27d0b409c0f1ca
55566 F20101114_AAARFG bach_m_Page_062.pro
2cb6d67b99c01c23de35d18221902a9a
9395a12e09cc63326d4760578a44d6e4b19a1e21
53744 F20101114_AAARES bach_m_Page_038.pro
dc177a079c96523eeee074661043428e
3507ce4dcd642666bfeb75bc9b6b30e5353103e1
99238 F20101114_AAAQZN bach_m_Page_046.jp2
6423cb2f269b6e29b759255d69cbf0e6
61cb71418569f9fc38b28522cded4cd4ac3fde2e
114476 F20101114_AAAQYY bach_m_Page_022.jp2
d1442b04562007136ae9a1fcb457d506
f0a3cc37848eeb70915b1f822d62a991eb0c8f15
40189 F20101114_AAARFH bach_m_Page_063.pro
75328ea809a9492e6f7a7c9754d7e6c7
277e0a1dfac1a04143af99c27839878a6e15f949
54723 F20101114_AAARET bach_m_Page_039.pro
a99d93334bc71d1e755ba0b4ebef421f
4d452ea25d246e5994ade0fde5b0c007b91108e9
112399 F20101114_AAAQZO bach_m_Page_049.jp2
d21ec7c0ff05bdacd9feb9f27dfdf33e
a05f1f5836be2acaf18e839c559b83e1ad830461
114497 F20101114_AAAQYZ bach_m_Page_025.jp2
37840557a9f19212e297c36973591850
e978ce4370b0c444ecbc699b34040db8c64c4143
12358 F20101114_AAARFI bach_m_Page_064.pro
d4edf2070cdbea4a0784fa31e46a09a0
74c27917b6784f9d6eb665a347bc4c3a8c47e348
52986 F20101114_AAAREU bach_m_Page_041.pro
ce5680693a3b202bc9b12500167b75a0
aaaa2b54c2e2ded81bf69b249fece2ae23e7b031
111609 F20101114_AAAQZP bach_m_Page_050.jp2
bd0915df400aca292c67d1987da1e903
4b4de4733ebe90ab1a111a3a60a9ff89f1b45c2f
5333 F20101114_AAARFJ bach_m_Page_066.pro
43358ee0caa1467c742267d8538ed4ad
b1055f67d49c213bd1d7c6149d8d451a4253f927
7020 F20101114_AAAREV bach_m_Page_043.pro
2f3435a521cd5a1a96aac4942bd5e257
bb761db191feb63cc3af19fec4b2e0a61479f6eb
50602 F20101114_AAARFK bach_m_Page_067.pro
05d4c6d6acce1496b476ebeff4845c11
05d162c0815153ea996d69e9e6be83ad57c22d55
53201 F20101114_AAAREW bach_m_Page_044.pro
baf700b87d32f91d19d0b56d63bbf053
1903f51954e812b54565700a5b16dac253e6ae09
114018 F20101114_AAAQZQ bach_m_Page_051.jp2
3aee302a5104c7dd85e7777a7edc4cb5
667608f8f209e6a84efd0a06b6f0adc1763987e2
54938 F20101114_AAARFL bach_m_Page_069.pro
c5e11d495550d1d209de40205e61134b
9b2f4e2a992e97cbf62cf2c3a5b41d1c538cb5f8
28132 F20101114_AAAREX bach_m_Page_045.pro
ad78877f8af1f813ff315d702b72a6e5
c45ef20aa6ebf33fd11ca53dbd703cca1965812f
105640 F20101114_AAAQZR bach_m_Page_052.jp2
f1ce6cfd18e1f04532c8acdf3b41ffbe
e4f14ee29dcc91f167bf9f02d5d0b8442e8305bb
54391 F20101114_AAARGA bach_m_Page_088.pro
c7cd412ea4860b89dc529a4bee949d61
b8c4d29a2e3882023badeb43812e82f758a32ab1
54073 F20101114_AAARFM bach_m_Page_070.pro
ec0b159ca985b52f2e3bb2d72d8046cd
2078223983e9db0746b83e842e8dec90b40af83e
51391 F20101114_AAAREY bach_m_Page_049.pro
94a8a26f26e67a995506f88393d490f5
1a901e42c8e4f6d735b4ddee665e97d07eb9ba55
117215 F20101114_AAAQZS bach_m_Page_053.jp2
932d86ed50882322f2c33a5c785d6009
d6d7a77d29bec507f58dd9744e0e8c1b36ba3139
52981 F20101114_AAARGB bach_m_Page_090.pro
1d410cf47f5cf42cc6f0475f62120937
27236ea6f2cf5d72c23d1caebd4107ec2efab6c0
43119 F20101114_AAARFN bach_m_Page_073.pro
72841998559eb1afaab3751eb3f7880e
753b1b210e4b14b42534457bc5e0a1e23bf8a1d1
51159 F20101114_AAAREZ bach_m_Page_050.pro
a755f8a258c86e4df353966d031bb9cb
fc137023a8d8d07a324cab52645814a0f82b5e3a
112503 F20101114_AAAQZT bach_m_Page_054.jp2
c3b0f49340c36f9f72292443367c2b9d
076fbc3d2534388cfb4d79616e153047dec3ea80
56584 F20101114_AAARGC bach_m_Page_092.pro
abc17ea8c7f82f964b915644e9c946b6
a5f6d7916148861130486e7fea348f1bcd7dbda2
12735 F20101114_AAARFO bach_m_Page_075.pro
76ff724f2bb19a50dfb1aa8fa6d8e00b
b8420181f10865c528a5d8b9e4dcf42798942395
109225 F20101114_AAAQZU bach_m_Page_055.jp2
4bb1664f320ad2dcc02452241ccea10e
689ba8969f3f3bb6109862087c9250818b94c27b
51482 F20101114_AAARGD bach_m_Page_094.pro
5fb204a83967d431ba7de51b6131b112
f8accfbd3d471c922615fe6c63efd5128a4b0dd5
4692 F20101114_AAARFP bach_m_Page_076.pro
b0a814b38a8ba33fa9512f60357edd59
657f8c87a62ce200ac3b839cf3e4f2f6f671aadc
411330 F20101114_AAAQZV bach_m_Page_057.jp2
3db99ea803752954b8508526a848f8c2
fd17b4eb850db686f4a6b7c46abe4fa9a31a61f9
51591 F20101114_AAARGE bach_m_Page_095.pro
69c14e415a0cea05532e8b6b7f340323
fdfe42148b201cd9b6a68f94f7860114db54b9ad
57990 F20101114_AAARFQ bach_m_Page_077.pro
2d04854499bd6e9409f80adbcc872be2
254a648f5be70bedccaf28017743a84bd6332b2a
115720 F20101114_AAAQZW bach_m_Page_058.jp2
359f62428ec07094c2d1815eab9ab869
ed170f78e181ac58541019f42b2e597a064bb23c
52206 F20101114_AAARGF bach_m_Page_097.pro
558530678f6ddb983a14eadf7c116c7e
6d85a9f8c255be8a4af97e5fc82a6ab9d180d031
57009 F20101114_AAARFR bach_m_Page_078.pro
f27bd207a862e7c1974bc62300f125ca
9753917ade4dc2d9239dd5720ee48ad31be8e843
113960 F20101114_AAAQZX bach_m_Page_060.jp2
fa746020f92b05b596ab64b4ba3c400b
18879a11c34ccfb6cc6651d9329af3e1766eb1cc
37960 F20101114_AAARGG bach_m_Page_098.pro
0a16aad2811e87d8d8fcb09bc43c338a
072eff5162a2de06b9033ea3263359aff665a825
55001 F20101114_AAARFS bach_m_Page_079.pro
ecad0e6e9df87abefd5bb1500c7b45b6
e95189c8b03780aa877903f9588ed68da1c62598
113015 F20101114_AAAQZY bach_m_Page_061.jp2
d4b4f4fd27cb68442463d4f75f5b1d6d
9919a0d051cdd5e09ee992d02441a51111d1ded0
31595 F20101114_AAARGH bach_m_Page_099.pro
66bea5b2acee0284622d862f30c4b5d0
7130c550406e612c63c78e70e60b9fd3582a248a
54445 F20101114_AAARFT bach_m_Page_080.pro
adace7da1503da8c99f3f270151976b3
0c847d95bf6daf138820ab560331aeea8b2b984e
120029 F20101114_AAAQZZ bach_m_Page_062.jp2
1a5cd3fb6f94874dc4e1ed3648791a4c
3eefe66b201f31ecf59bfbaa50b035f557d82762
37216 F20101114_AAARGI bach_m_Page_100.pro
c222a392c9aebf8f0fb4b512d79db8c4
8d78c12baddac9241b0c1f3bab9922c42d72f69b
54074 F20101114_AAARFU bach_m_Page_081.pro
9a6c39b66be94f311f82e66904a18f3c
a39941e4c7b735f5210a3ea75cbddc7c212cd76c
9178 F20101114_AAARGJ bach_m_Page_102.pro
dd0f8c22b9ada4a0ae4a854e2c5798ff
cf2fd072c5f927e1b506a75e1936dbed556e0fed
53368 F20101114_AAARFV bach_m_Page_082.pro
b3f20a9a89169d807fc74a1114aac857
db8f687adcdf1c938a5028d70ef75017b67d09a1
5879 F20101114_AAARGK bach_m_Page_103.pro
bc89ef8c6da030b46738eee4d0995fc1
eae8335eff52cc268f84fb7b6f28f477f80ad0a0
51253 F20101114_AAARFW bach_m_Page_083.pro
5dc952208a9a4fcf7491708b49e1d642
7f59d3f9de0ac6e5cb497019b65f874b1be30910
20676 F20101114_AAARGL bach_m_Page_106.pro
ffba6d312645568d1f422b76a09f733f
88e22f710fb312ec8f10eefad175b443759b725d
54330 F20101114_AAARFX bach_m_Page_084.pro
927fc99616a655fb78a82d51cbb2d06a
1c2c58008cf9839dae53e1ef04a99abdf274b6fe
33348 F20101114_AAARGM bach_m_Page_107.pro
7c960a399c170774cfdbf5244fad39d0
bc4730a854574436996804c9db4bbd3f1c39e382
51040 F20101114_AAARFY bach_m_Page_085.pro
fe25530091ae644962a1bc4e338e8520
39b51d9aa4353ffe88020195724b925ed5dd95d3
19242 F20101114_AAARHA bach_m_Page_132.pro
fbe11ea31728c698677ddcc542b9e131
06a554da717378a0896cac9ce041a3fcf1ceb77d
32408 F20101114_AAARGN bach_m_Page_109.pro
458ca70d7fe147a4ea27ff711225ab98
d6e2ec5eb44a920b61c23113d134534be50a7cec
51792 F20101114_AAARFZ bach_m_Page_087.pro
2c5a2e87f8a39531fbcca0f5f9ddcd8e
87fa66f1076a64651403172f5c100362f407a6e5
92 F20101114_AAARHB bach_m_Page_002.txt
e2dbd4249262927d1caa3c548d3aeab8
505d138b67ac543f01c0ee1241082074e2a3d518
49693 F20101114_AAARGO bach_m_Page_111.pro
a29313d659ff56466e542ad9c25f3196
88aaefb4de153ed88d15868a518556bd8c214211
3256 F20101114_AAARHC bach_m_Page_005.txt
f7d6c194a06924379ace4b124873eff2
3329c6df3f9632437cfefc250480a48296d83ef6
17463 F20101114_AAARGP bach_m_Page_112.pro
3da9c79177714940af70927c139cf87e
0d8a1e6dd6f46db814501fe68b9bcfa24153c935
3248 F20101114_AAARHD bach_m_Page_006.txt
b23b6c46d21537ffd25db464ec1cd51d
18f6af4e40dda4a0bcc3625dc55590a95950a902
4981 F20101114_AAARGQ bach_m_Page_113.pro
5c438196aca865fd46875679d9ec2e07
d22cea1ea759d5ccbb67804a8a0bc3c873de31a2
528 F20101114_AAARHE bach_m_Page_007.txt
88e7a713d12de6244d8ab58f6711473c
36ea7b5c003c8f7f559930852eb6f46c0ff14fa7
36612 F20101114_AAARGR bach_m_Page_114.pro
e9c8564051a8a90e4713a2719f5de371
0fc627409a9a8724ee2d7959e65ead75d13d2b2b
2495 F20101114_AAARHF bach_m_Page_009.txt
f8ff832d3d7bc69de297a58e05c7d777
eb1eef1ca7f22d4a24980a9fc1a886fec54560d8
15809 F20101114_AAARGS bach_m_Page_121.pro
b101b85bdbaee73d72ec9bcae51a3ca2
0ac3ebad84ab6e3f91be0df52a26817100242476
2014 F20101114_AAARHG bach_m_Page_010.txt
9fda33e4829a5b5a86a06df3a3acc587
96476794a555320f300cc8845b64aff4d65cdfa0
39083 F20101114_AAARGT bach_m_Page_122.pro
6b56b4c9c2531a0f56ad3b3a751f3f9f
d4eae33b93041a5d580d1e18b9786ca2b7a847e7
801 F20101114_AAARHH bach_m_Page_011.txt
cae40c560843fa5db80d25e40b3157d4
09928235269a07521816b99dfab7d549cf986958
61176 F20101114_AAARGU bach_m_Page_123.pro
cd7796651e4d11860bfc900e4b0297bf
2c91e2389b2bce5b305c8e432e64790f336288fd
1822 F20101114_AAARHI bach_m_Page_012.txt
2ce0e54bacb7df8131c7ac22a15f6fd2
1f5cfb86ac0afbe77c59c4d31a20cf8f9f57aed1
64975 F20101114_AAARGV bach_m_Page_124.pro
898f9a8c391169b9cca9aeaceee1fe1d
0884c1067fa110c1b5560c4a66c025b883f3bc0f
1849 F20101114_AAARHJ bach_m_Page_014.txt
9d179e579b2218c9baba668f44379cbd
4e73db93694016ba3b5c40c0719da9ef7bec11c7
58483 F20101114_AAARGW bach_m_Page_126.pro
25da80ae058929b454c53960f2b687f4
c33c574c6ad01d5f63d3879b59848f2d565bea3f
2030 F20101114_AAARHK bach_m_Page_016.txt
f9902be6ba28dccd7e8fd2801cb80369
549cc6c340dcec61592788771fd99f72ae0ec421
59200 F20101114_AAARGX bach_m_Page_127.pro
ab0e85b8620a28c38b9b27bec176ff33
974bdfe87b8bfa647056093957081a7da555664a
2118 F20101114_AAARHL bach_m_Page_021.txt
499421bf0c7565dc0b56e855c1353110
dc17923feb20678c9b3299fdbbd71342df7f7595
65311 F20101114_AAARGY bach_m_Page_129.pro
07baeec2cb665547d36e5f23ebd6018b
e26891193143ee6b58dd74e2565e1b1bf3c3bb0b
2076 F20101114_AAARIA bach_m_Page_055.txt
5ebaf2aae46ce366a06c00a9701884b5
8644aa1bcccee80a08cbe90c459502b71a26c571
2211 F20101114_AAARHM bach_m_Page_023.txt
43b40fdfedfba704245d53e19d9bed51
4a3e035b3c1a1248dcfbaaf93694a585cfbc4f52
3131 F20101114_AAARGZ bach_m_Page_131.pro
71feacc1a6f3704d3b831caf7d45cbd7
c485d9aaee1fec9d8807185afa504e0dab586d29
2104 F20101114_AAARIB bach_m_Page_060.txt
74c3fd02608f26de9ffdcf1a993092e1
033f227a8a95db55fc1322f909047652efead1d7
2144 F20101114_AAARHN bach_m_Page_025.txt
4b31ddd65419ce96a93bc770886d6526
ef27589abe673fabc4fd007721c72e4039e278bf
F20101114_AAARIC bach_m_Page_062.txt
783c64d9b12061e2563f7f25e74cf89a
694daa770527bc6efbafcdb76bf1b25d2be94703
2213 F20101114_AAARHO bach_m_Page_026.txt
91c02c88781d3736bba4a7d5eca58d41
1a10f2aa107d22817cdbb90b104b482b343896cd
1644 F20101114_AAARID bach_m_Page_063.txt
2b92c1970b5462fd19c26e483fb127ce
8fc3b4e8182811530ac35f18208261f44110acd6
2084 F20101114_AAARHP bach_m_Page_028.txt
634ec47dbbd0536d31ad21d44d3b2463
7776423f5ec09fb0b41d2c73ee74c70f3ad78486
596 F20101114_AAARIE bach_m_Page_065.txt
a284ae93f6f9837e39ed53f843b88f23
c0a88da040fd478f9f415ec3c70b5c4cbb1ba828
2220 F20101114_AAARHQ bach_m_Page_029.txt
eae7a4a7a98c77087afddd118d0c02ac
8f71f5985403b1afe7e6ab3817e235be1a399670
2140 F20101114_AAARIF bach_m_Page_068.txt
33f104e3ba284bca6ac1ade28bd01c04
3c89bb2aacd3945a543a5c72a4099c72741029d8
2072 F20101114_AAARHR bach_m_Page_030.txt
4cdbeeac63b65cec6280df4809684ad6
cb70b8973107c740fb072cff16e96a5937c40b1b
2025 F20101114_AAARIG bach_m_Page_071.txt
e931df19cf93fc085d9a0995023f8c49
ceba49436d280d34e9ca73fd31f71aa9b6f3f0be
2242 F20101114_AAARHS bach_m_Page_033.txt
7fb1102fbc46670dc4b7eec25472d5a3
8d8d502c92e8cb773153c38e3cbd620b1087c0eb
586 F20101114_AAARIH bach_m_Page_074.txt
f49d80a8ecabb21c9d2be448f32abff3
3f6de988765dddc128d749410d2233c3e5369491
F20101114_AAARHT bach_m_Page_034.txt
cca283cfd181d747afadcdd210798e28
a8c2182057b1f2d069c98e00f2bbacc4c5dfe92e
324 F20101114_AAARII bach_m_Page_076.txt
5aa6cdad9bc3326381420836a56133c6
fbc771226f3c55da7e0bb001cad87fc13c493f3b
1832 F20101114_AAARHU bach_m_Page_036.txt
702ea2db01fe7569aeac1132076b8520
68449253ea97339b130647a65129ffd59b8159a2
2268 F20101114_AAARIJ bach_m_Page_078.txt
9f582451e4b18b95aabd0cd63d35fe3e
cb37baa039c755c0d2914bd7e8806b9d1b30e932
F20101114_AAARHV bach_m_Page_047.txt
5231dcaab8982088b5424c646e36ed10
7f1e768927dd047bdb9a26ced1b94f278e7315e2
2155 F20101114_AAARIK bach_m_Page_081.txt
6205b4b2472120293ef9db9e407322ed
e946126ed729971a7461d7a34ac9cb46d42881e7
2097 F20101114_AAARHW bach_m_Page_048.txt
a9bfff1d50c52eecc28ed6a22d8a1095
a2083599f4dc540ff44ec205cf8a8a5d8b91432f
2109 F20101114_AAARIL bach_m_Page_082.txt
5e5bfebc0114d42fe239cd6cadac46a7
63638c19cc9d951791f95a6e408e30ec686c3e26
2110 F20101114_AAARHX bach_m_Page_051.txt
f0c8803a54546a14619b0f19a0eb4950
ac3c66c71a6e4ec7425f1c194af4670881f52087
1979 F20101114_AAARJA bach_m_Page_111.txt
d0b87e122ed7bfb5b017bb07a8742ebb
43a71ba79abd4d952d26f33b6b37477c6211a0fd
2056 F20101114_AAARIM bach_m_Page_087.txt
ace9197aed873b8f730f684a84a6e168
8c3cd3da433d95d4d5f30e9c53a3a0124f57dcc5
1966 F20101114_AAARHY bach_m_Page_052.txt
84f3488cced39463ad65b045c66f92b8
d7b4b3d61e9030dba38ad16e692000eac395a28b
778 F20101114_AAARJB bach_m_Page_112.txt
4f3f317b64569b2bab9cb4ae6616b517
c138b132237fe27a32ce69ed93735ee81beb7492
2149 F20101114_AAARIN bach_m_Page_088.txt
879f4ba019428abea4a8ca63e7f18342
c69b541868c187bbccb3ec05a37c47a8aa52e1f8
2103 F20101114_AAARHZ bach_m_Page_054.txt
8c27f6500812da65b06a813d61b65b25
0ecd38ca065dc2797798e83f339d84e46cb4c924
6956 F20101114_AAAQGA bach_m_Page_039thm.jpg
dc6decf9ffda079e1384da393ee023cb
6cb86bdaadb9f315e8def00fa91279447042c0ae
253 F20101114_AAARJC bach_m_Page_113.txt
1bf79a3bdded4573a9cc731061c2ae4e
109897f54b5e48ee998d51d0b010dd1bf6b4f386
2130 F20101114_AAARIO bach_m_Page_089.txt
5e7ea81015782c7721724ddca2740960
03261a6fc38f39887c75ed53dbc7a3e0f87d0e49
7216 F20101114_AAAQGB bach_m_Page_077thm.jpg
f67de79960b2c5407947ca90247581e4
35a877b654472d5f9b9b3a27bc5c832a1305ec1f
1579 F20101114_AAARJD bach_m_Page_114.txt
75d235d79f1b8d88a1caeff8593f779d
682d23971ab8f225e54faf82dd5ec395baa0bb52
2151 F20101114_AAARIP bach_m_Page_091.txt
a37e51a5e2342f3ce7b5f6cc8c3a8d4b
7595c9304d9aa534d64e4e4684d84b0a2af3406d
184597 F20101114_AAAQGC bach_m_Page_113.jp2
cbdb5b064607b189abbc7d9ef592b7c0
9b9454b0396e8a370ee4569a711b74f99ecf20b1
1605 F20101114_AAARJE bach_m_Page_115.txt
67b0e35f81a4b37dd0edf40ced195d7f
7071c4397b1119c4589ba0604de74a9183c6ffad
2159 F20101114_AAARIQ bach_m_Page_093.txt
a9e407ac0231de653cd511ba1ffc40fd
a22b0c6d6b4ec0adb5c50191a6e42670d70e4752
677 F20101114_AAARJF bach_m_Page_119.txt
63509203275f84df793e816a66df2d7b
e37c5255c6225fc2bcde0f26c1dcd4563aadccda
2032 F20101114_AAARIR bach_m_Page_094.txt
52a39526ffebf93e03ab60ce3354cb27
2e463104610af61e91c6089b6b411295f5ea79d5
54466 F20101114_AAAQGD bach_m_Page_034.pro
34037f4b9d097bc163a4890216e9a40f
bf7a41a51c5013032d55fe352575b2d89f141055
F20101114_AAARJG bach_m_Page_121.txt
f38be67d4bcbaa35e9fd0808b7f9c8be
30015d3ed4673988ec09189a514491233ecdb807
55227 F20101114_AAAQFQ bach_m_Page_096.pro
693a1174dd3c5993dabf79a228a4f94c
54b97f0a88f8af15edf11eae010335df539d1298
2205 F20101114_AAARIS bach_m_Page_096.txt
446dfa5663ab52d4d0f92ea5ae6fb82c
acad2933104e361ac4f77de4ec198bd32daffed3
2279 F20101114_AAAQGE bach_m_Page_126.txt
20e0052c90cfcb8f4e1381c007855452
9356f29f41898657634698c1d03da892f81c1bfe
1631 F20101114_AAARJH bach_m_Page_122.txt
0e6e56f0f678ea112e7ed4fe0e2f6c8f
82713769003f78d2680a5d50eb1aae0aa26d3dfa
52077 F20101114_AAAQFR bach_m_Page_117.jpg
fd09da11f32d970629e92f9934852126
dbef6369b235d5617a687ab169a032e75de02f37
1558 F20101114_AAARIT bach_m_Page_098.txt
fdaad77c154df34c80a74efa11048e57
b7efbcd98ab357a52783d2615dc74402d7a54a28
F20101114_AAAQGF bach_m_Page_083.txt
b184d18995aba4f43d0f785c41d3dd3e
3df5333b9a1df32d3d79f1cbce94f1e281ba250b
2414 F20101114_AAARJI bach_m_Page_123.txt
b6ed32aecfd11352b9b90f997916c28c
aef9cb432f67cad5213b70f6d2fc4b82f9d3db35
2156 F20101114_AAAQFS bach_m_Page_039.txt
33e6c083a45aee9d11d7bf6fd3302292
f2a547068f975c48a29c68ad390ba2feb91de68d
629 F20101114_AAARIU bach_m_Page_101.txt
132828ecce0b9534d42c9d0c46828c42
e89ed094995086458be05d64df3ac119fbcb9008
48299 F20101114_AAAQGG bach_m_Page_052.pro
d0d364335c8f5dcf054f934e3bc76971
effd184fc4938d38947088dc8b6ce861f7c092d4
2517 F20101114_AAARJJ bach_m_Page_124.txt
dc0e62f215940a084609c2ba66becbf5
4c4264631ca7cf207692883d268eb0461195cce1
F20101114_AAAQFT bach_m_Page_088.tif
edc2ac34d6c944622f0c837b0222e507
5e628ab5cc54d54e6f9fbe1679147e6d99486559
712 F20101114_AAARIV bach_m_Page_102.txt
bbb2ddaf5a3ed1cb77f74346d7103cb0
bd1ac6d2b8fb8a9e609fcdd23d4f54924b196a06
78219 F20101114_AAAQGH bach_m_Page_019.jpg
d384f3edc5ccd778b46310cf70d85de9
f8d36dd7dfb5be9ddf311efd87d7e657a2c97624
2448 F20101114_AAARJK bach_m_Page_125.txt
bb5881a67505cb27dbe9f5c1024075cb
bdf87860120159c8cee746fabf743e637137002f
F20101114_AAAQFU bach_m_Page_012.tif
025e94807ce0c9c1d12772501aee09a1
ee37f51949cf7eb0095e206677285e27a1007f60
1207 F20101114_AAARIW bach_m_Page_105.txt
87cc9c534062e88fae595aac22efd001
fc3c91c50aa58500b72c68bd09f5a6b94fa57c92
44230 F20101114_AAAQGI bach_m_Page_046.pro
f8f45090c2abda9f201300a3c52b67c3
e963c44b3b8bff1c291bfac535c771cd4c921e4c
2304 F20101114_AAARJL bach_m_Page_127.txt
4565bd3bd3ac1308a5e873e31df13f6a
90631aafef545f7c9ed11ad14c8568e2788c3abf
46265 F20101114_AAAQFV bach_m_Page_132.jp2
e5571672d8f75b5ac2f53aab061570ac
dbe1620889cb26c949ef535cc4f7b458001fa04d
1535 F20101114_AAARIX bach_m_Page_107.txt
3e61f524ad780c5aa76474ed5a86c4d1
01b54d469ac33af9d9e5b2f2974e5b9fa92d1a7e
F20101114_AAAQGJ bach_m_Page_002.tif
a739c17f171f1e873da5e31ffae01940
c7f7e1f1a0006b4e5fb97689df117ae4acaf6395
6376 F20101114_AAARKA bach_m_Page_009thm.jpg
d8bd100e97cd56f90498047389299ddf
555e12a83a29be0047c1a0ea1203abcb07cc177f
2335 F20101114_AAARJM bach_m_Page_128.txt
43ffcb0b5243974b228117ae5b5c4c48
634cad3235bb9bfafd8f8370e8b2c7dfa06f5d6d
74268 F20101114_AAAQFW bach_m_Page_085.jpg
4665c0157c85dbc3f44bb384dc912b3d
00d1cbb86e3db2a41ed7b958096d127f6dd57ad3
1418 F20101114_AAARIY bach_m_Page_109.txt
d2ec094fabbf57f32fcb145b726abfac
d0a2dba735ddb2997813c4c76c33fc4ddb64a753
115529 F20101114_AAAQGK bach_m_Page_047.jp2
be1421d19a853ed3ca1f160491193c2c
37f2ba6a72e22f00165f3a11f10adceb3dc97f62
21604 F20101114_AAARKB bach_m_Page_010.QC.jpg
36e49353c314090d5eaeecb3a7db9880
df3971564faeec1dd6960334c4c8d2221685cd1b
2287 F20101114_AAARJN bach_m_Page_130.txt
7a9139c1fdb204259ac0c4522823bdcf
2c696db7a8c13cee51ee551e2dfaf43d96035f12
F20101114_AAAQFX bach_m_Page_075.txt
4c6873d83f233b0fb823926208b3d88a
f196ee6d7486b3bd370542195358b56f8a14997c
F20101114_AAARIZ bach_m_Page_110.txt
233d7cb5d84669a79abc2dcceb0568c5
a67926cdb925d3da4ab398551e01757150165e49
90 F20101114_AAAQGL bach_m_Page_003.txt
732fb7c497e2a8d54c65ab2c5e8570a7
df9824b3def4d8f1a9769dcd95ec9863bada8229
3477 F20101114_AAARKC bach_m_Page_011thm.jpg
fa729c1712f3a3614a4290187316d961
0db1ff1b154ee1f722468f69b11e5d50c51f9ce6
16973 F20101114_AAAQFY bach_m_Page_114.QC.jpg
1e20b4bda863a32a4f9e071d08a628fe
afb833fec205cfb1e53a7ce023c20b789df871e5
81326 F20101114_AAAQHA bach_m_Page_006.pro
ed0f75a217248665cc4b4967e44e9f84
fe6a1890034af8485037ee6ef1b6071cd5bfff24
123358 F20101114_AAAQGM bach_m_Page_126.jp2
bc51a98515d402efdc3d43db660d1b55
1afbb4f57b7cbfae3be591a909c893f2e3325702
812 F20101114_AAARJO bach_m_Page_132.txt
b922e2c91ed57beea315c7330a58d328
bc9b1affe7e86b00d7b4af6ae154a1455aee0d60
20590 F20101114_AAARKD bach_m_Page_012.QC.jpg
3e1cda6264e7eb9b9e1dbddc485c945d
fae9c3f9af041651f5614c5622ad47d14e5369f3
112160 F20101114_AAAQFZ bach_m_Page_086.jp2
c6e8fec2c08657c53cdb546df8537459
ff254d6efced89e4f8eee92759b42e5ce578e154
49271 F20101114_AAAQHB bach_m_Page_118.pro
ff1f5c2dea64b6a856bbd2b4a12d4af4
a6c61f8f11d23d6054bff4db02320444b920824e
37783 F20101114_AAAQGN bach_m_Page_115.pro
cd1eddea723664a0b1700e0e579d8769
659861c465a087094bed2343e05f1923a8eb4ded
1038935 F20101114_AAARJP bach_m.pdf
057bd21fa61ca9102ef31fe0bb4fb723
296d41438ae158bdea9707f11d54692ddc606da8
5858 F20101114_AAARKE bach_m_Page_012thm.jpg
63dd9b1f58551eb447c01a25c15e3847
c53b13f1563ae7a19426d6297cfb760a8429a294
2224 F20101114_AAAQHC bach_m_Page_040.txt
2e8101fef8d74a7d8d20e3254f945c42
4d68ce8cef8f59b121de98fdb4c4f6e6d28f20a1
F20101114_AAAQGO bach_m_Page_013.tif
d45dbdaffddbf2d30c40449e26e177c0
05c0b185c59d9633de568ba8d58760fe2ba2ea33
3139 F20101114_AAARJQ bach_m_Page_002.QC.jpg
3d3589e6911c9f278aae566f0985699a
36924428a7925679b11d96667356f76006f9ce8e
5877 F20101114_AAARKF bach_m_Page_013thm.jpg
5a3b16f7ceb0b971d72398526af02ad9
000ab800ab4b81934a265e5754e2991e99cce14d
23370 F20101114_AAAQHD bach_m_Page_052.QC.jpg
259535717cded4a47dd81729bf29e315
69bee0d0f0bacf28407f8439766a2e4aeb8465d6
2981 F20101114_AAAQGP bach_m_Page_043thm.jpg
3338af1b6ba30f446541ba7ef0fcd79b
a5449cc377aed75f4f7172ed214bbc668de38a54
1334 F20101114_AAARJR bach_m_Page_003thm.jpg
e74a3c3973d4cb198a640a06b9b0dae4
26906f5919e042397f38b8fc4ef7ba733c77c04e
21835 F20101114_AAARKG bach_m_Page_014.QC.jpg
51ef198b857c4003a7af2b013ef35482
74d745c45dca214192e85341ba53976afcc9331b
6970 F20101114_AAAQHE bach_m_Page_020thm.jpg
6fab3b7f9672970336e8eeab159149b0
28d0907396cca25ffc12cb17ab5e3a19455cf06a
52415 F20101114_AAAQGQ bach_m_Page_086.pro
9e629342d665a02ad7bab57873ef8fa5
e053d7b2efc89e65d261afdbd28a4eab9bc27a04
9097 F20101114_AAARJS bach_m_Page_004.QC.jpg
460ffe6078b18dca3d1f12ec60d4a373
e44b4b08d7ceb7acb2d7cecf290b077ca5b7eed0
25578 F20101114_AAARKH bach_m_Page_015.QC.jpg
3dfea76e8890d6d36f2f9c2d5faf8e45
c69a3447ae3c763e557429aeeddcc22c5feb699f
24761 F20101114_AAAQHF bach_m_Page_060.QC.jpg
64fd725fa1595b29788defa035cf1fcd
85ad8a16337904a50eef304dd2b0213bdd56dc82
10107 F20101114_AAAQGR bach_m_Page_065.pro
787dc33054e59e53b033a3ebf10097f4
360d6dd61278441a1393d71a11d6e7f5d1aef0c5
2808 F20101114_AAARJT bach_m_Page_004thm.jpg
ad69483e54e41fc440853e3783c5f809
de3e6ac3df67b5c9ee8812a57a7f14148b8e9542
6992 F20101114_AAARKI bach_m_Page_015thm.jpg
3473fc1aa1c6100bc8ef2c162954cf0d
fb46ed5125f99cd8ca09d1b55bc58653a7cca4a0
26054 F20101114_AAAQHG bach_m_Page_125.QC.jpg
6d76f146cd37ed8b12551f30f17ff16a
a9ef39556c5be13c8af95d141bee4d55c1ab37ee
77703 F20101114_AAAQGS bach_m_Page_069.jpg
d4da4c1e66ed0e1c912f1e5347d7be02
7c2fe0085415b1e9f83697831a069e09a89b7276
25596 F20101114_AAARJU bach_m_Page_006.QC.jpg
59bc247f41df3af8267eab684ee0b604
c77c3f8041d350a18bda031a777964530183872e
6883 F20101114_AAARKJ bach_m_Page_016thm.jpg
f8a3e17082f74efec0356c655b87cf42
98d969123376beea66e5a0adabbd24f909a57e78
34189 F20101114_AAAQHH bach_m_Page_132.jpg
ea7d57d36eef5d699d3832d4ad37f09e
7918251b17456bf0731c2bd453a65c49d7734d6b
F20101114_AAAQGT bach_m_Page_101.tif
c92e2fbd83961e93a78d5957bc2334e8
7e033d190089212738fc986ea166efcf8629fa78
6283 F20101114_AAARJV bach_m_Page_006thm.jpg
e88b45e427b3cd88f0f5783b7ab5918b
34778cb2b53c37fb8b760a490231c5f5211eee41
F20101114_AAARKK bach_m_Page_017.QC.jpg
46afcecc4cf24a58fc701353b7a21e44
bef702baa593a73e76c3c189e26039764974bb39
7444 F20101114_AAARJW bach_m_Page_007.QC.jpg
ac488a6549cce0c5215b3f99a07e7884
0466ec5de61c51d82d92f12ddd18f8e3edaec0c0
F20101114_AAAQHI bach_m_Page_051.tif
3fead400e8d28a6fee45be31b713a0cb
edee9eea00e3a26a0850be65230aac9a255b958f
F20101114_AAAQGU bach_m_Page_056.txt
4745f53c4ea31a06c9995e4d18ce3f14
a7e23742427e064fd010b86831b5581fed2bb549
6575 F20101114_AAARLA bach_m_Page_031thm.jpg
c98680e4262a420768bd8bd6e6c0fc7f
83b6a36c2a4cf25d42be5ef2d03eda09bea59e19
6923 F20101114_AAARKL bach_m_Page_017thm.jpg
fdafca79577a55f948ea87d77c1c30a7
d9c013621f35439fcd1bda4b5c5ee509d09084c1
2464 F20101114_AAARJX bach_m_Page_007thm.jpg
c8ad88d1683eb11887a21b5808a42b55
39d886504283ba79c8e3457a1ae1958925bd5520
F20101114_AAAQHJ bach_m_Page_098.tif
a872346678d0a16c4f58355f7a14cbc5
f2e6ee57cc54bb3dc86f729c3644a081892361d9
26314 F20101114_AAAQGV bach_m_Page_092.QC.jpg
c80109e64b94a0cfdcb782ae42886a2a
c54c516569a569402766213bff8a0d29c955fe55
25603 F20101114_AAARKM bach_m_Page_019.QC.jpg
5ff6bec55149257104b5adf08485e05b
ec93e8050b7558bed7a3aaa6246c27bea8a40035
4336 F20101114_AAARJY bach_m_Page_008thm.jpg
edf341c28f7a4dec90f62ec607969b92
6420f9253277719347bf4d1b68dd90217f33fd3a
2091 F20101114_AAAQHK bach_m_Page_041.txt
1bf6c123b1f4c11080ec162e12b3a49f
b0ddca21e364c9b095ed60eb2bb5383280c3ae3a
6623 F20101114_AAAQGW bach_m_Page_086thm.jpg
17f633b26efc8ec01451d19ae9f3f554
96de91a0f74413766d6cf58441e7761591c80eeb
26218 F20101114_AAARLB bach_m_Page_032.QC.jpg
b271b927174fde110cc5b60941ba114e
f78d32b707e6eec32e5e215302a5d082e20e22a3
7046 F20101114_AAARKN bach_m_Page_019thm.jpg
1105c12c7da10746300f6ec36d4b0f89
48ecdb3159cc59b51b29e2d48e940e6fcb7f9c2f
24514 F20101114_AAARJZ bach_m_Page_009.QC.jpg
11c60f51988d304bd0485598cca202c8
f47bae0bac9edc2b0fe7961fe1ecc95cdd46eda2
7094 F20101114_AAAQHL bach_m_Page_096thm.jpg
704de318aa4624e2720e11b601383b5e
3e5777be0c145791e67fca8d2a17a298d7cfedc2
115866 F20101114_AAAQGX bach_m_Page_081.jp2
d480e69002c5edc1774f7aa60fb5ddb8
080c6c3efcf70654026791ba58cd8576f1fb9916
6959 F20101114_AAARLC bach_m_Page_032thm.jpg
63bf321b1e0206f07acf8265a694c53a
8eb73373d04275944a23d76b2a3ce08a627afeb6
25437 F20101114_AAARKO bach_m_Page_021.QC.jpg
1955e4f6576781bc768567f7199532a2
18bd5ab4c5a89035c53c5df2a30578d165c8d9ec
135592 F20101114_AAAQIA bach_m_Page_129.jp2
94f6a46ee1d4e2fd005b115208229b3a
958f9f87eb72cb2cac86e40be923358b67fcf061
115686 F20101114_AAAQHM bach_m_Page_080.jp2
fc953f6dbeaa03b9e5a622ff72f39c3b
94f0ea7d69032dae07465543ff71b0f44350b085
F20101114_AAAQGY bach_m_Page_006.tif
282dd0beb16004a5a3f10d06ce6286bd
5a75819d713a58968629aed8de27bdbe69704d13
7171 F20101114_AAARLD bach_m_Page_033thm.jpg
7620085fcdf0bf6827f2ef661abe55d7
a231d06d872e236013652f84b4d7a6734f2b1062
26300 F20101114_AAARKP bach_m_Page_023.QC.jpg
2852d9a4d6afa136ccfac81baf815c2b
c140a84b16ca124341b0bf379d4557ee0e657bf6
6820 F20101114_AAAQIB bach_m_Page_025thm.jpg
68c44e3b87df0e0c6ff75a3c332c2cde
f7d94cef8caec97e41a8cca05d275568c5a789a3
6805 F20101114_AAAQHN bach_m_Page_041thm.jpg
1f4211c6b1d5be0897d2e15441088c70
5630f7a71d74c78a77d015d1ff6fecf327a542f2
5025 F20101114_AAAQGZ bach_m_Page_108thm.jpg
25e0164b113da68ed732b86d0ecb26c5
6419fd7b2ed0f6370fb3e8481993d808a6affe3f
25699 F20101114_AAARLE bach_m_Page_034.QC.jpg
97451f73f7a80dfce56b6d9027605327
fb0b06ec3322158813b2e42639ce92199c6b1e2b
7214 F20101114_AAARKQ bach_m_Page_023thm.jpg
56a57fab8b03b2f06f9d6252cba29ea9
03d9d2a1548bcaae79a4ba02c5fd9923feafb440
F20101114_AAAQIC bach_m_Page_117.txt
3b558fc4a8821522315a299f09b9e93d
c7a13b2406c7eff71a02b44fa8d37e22830a935e
80197 F20101114_AAAQHO bach_m_Page_029.jpg
965aa967765c41110c9906fefe70608b
a2ef9d7f699211d45c00e24c4c203f3ff725ebf7
25768 F20101114_AAARLF bach_m_Page_035.QC.jpg
21b1ff2fd0cf8122b812d840f165b17e
af494202810d53e9669cc6e6c24919ef93f416a9
25322 F20101114_AAARKR bach_m_Page_024.QC.jpg
fff90847b9f8d9c1fca30ddeb3276275
97c4086d99fab99c22d11339494d0c79112b37b2
F20101114_AAAQID bach_m_Page_048.tif
9ccc17f513d7764f6adec3d3d7ff99b4
656b0175b60d0d9ef533ea07daa36b47c4c3dae8
25011 F20101114_AAAQHP bach_m_Page_038.QC.jpg
2b4db7ca1740d83f68e4561b52f621c7
15795a9458c57ebf582b86c2bd8db1a3addc3849
6108 F20101114_AAARLG bach_m_Page_036thm.jpg
5a19f7d4200778a01d4660feae07ad04
84e2ca6b4ef3205ecd1917b5eb342b41d8cbccd5
F20101114_AAARKS bach_m_Page_024thm.jpg
0a210fe45a73b1370b31e1af51891d53
d5455ed157db49203fcbab2dee8bb33f2188b462
14072 F20101114_AAAQIE bach_m_Page_104.QC.jpg
33246bf29be158e1d050d35a79106bcd
af58e19798ec90b779a2e6a4e5eb296624b628df
7152 F20101114_AAAQHQ bach_m_Page_088thm.jpg
639548adc1bf792772224fe056f2858f
92e38ac2533e99c18cfd40afd7cb8a5731cb1aa0
26358 F20101114_AAARLH bach_m_Page_037.QC.jpg
7fa41bd58b93df6b68e274379f884436
99bbae08dbe6ce3b7a765120280c3bca78dc4491
24568 F20101114_AAARKT bach_m_Page_025.QC.jpg
35ec9abce099c7aa7b27d6fc0a2c92b6
d4ea2cd20b4a5089e1ec66857e19968e30399324
538 F20101114_AAAQIF bach_m_Page_104.txt
a81de6caba537e80c909f4f3d2101ac6
6ea3df6b43a9f6f8759c25c249a4dd57269efeec
18755 F20101114_AAAQHR bach_m_Page_118.QC.jpg
78ba39bcba5dc36522b8771594f92319
a830e498815604e3006d5f5035548acb514f1e10
7033 F20101114_AAARLI bach_m_Page_037thm.jpg
d24faf3e56ffb88e86095e7bd4e34334
686b673296408badc02b62feca7ed0559936512f
7127 F20101114_AAARKU bach_m_Page_026thm.jpg
014896b234ca861024b2bd7bbb769c8a
fb963668fa23cc4618d6c32eec540645516cc0d8
22551 F20101114_AAAQIG bach_m_Page_113.jpg
3a3a5a5ddff20baab622d84caa5e24a1
eb5d94ffa22b2e61e78f2e74322eeaf9c943df4c
5397 F20101114_AAAQHS bach_m_Page_100thm.jpg
2339113a0bd4a9fa1758619b8cbd4a71
b4babb5b589e7bfaea8e87b071fe7fef5d9d9d5e
6665 F20101114_AAARLJ bach_m_Page_038thm.jpg
baacd5c7d6b2ec05bbb32d84bc2eba82
89de21b1e420604120b0704fe1ca5386421b84aa
6921 F20101114_AAARKV bach_m_Page_027thm.jpg
6a7d6fe4e11fff82d0834cc2bb738456
7b36a6d660155393d00ab8e9cd88db0b52402505
13667 F20101114_AAAQIH bach_m_Page_007.pro
b55d75dc7a165a1236821eceb434d3a5
23fa7e5cfa5fa044148b14b080736abeba2fe96f
F20101114_AAAQHT bach_m_Page_077.tif
ad8e75fbba11e39609837fa952f84347
42674aca63dc461dfff687661865b90118135bb8
26448 F20101114_AAARLK bach_m_Page_040.QC.jpg
b4a99729efa9eee6af14ca3792f68c2e
6dc556d9f3b57f1dd09ae70c35405c7e734b08e7
6533 F20101114_AAARKW bach_m_Page_028thm.jpg
4a2e6ad4b8b875fe969fc07e6cd0cc4d
55394a3ab5259c247ccc99752ce0ea80124734ab
2162 F20101114_AAAQII bach_m_Page_079.txt
f6864b4204d36c0c7fc3a2a36e24aec4
2dce8a723556f08fa78a0ab31d478b8292448d5b
24189 F20101114_AAAQHU bach_m_Page_085.QC.jpg
8f8d6fcc3777209cf9fd8fed3e39502b
92e712142f5d74d52c63a163c06235ba99ec0130
6686 F20101114_AAARMA bach_m_Page_050thm.jpg
991d01afb8a2cf44b19b39c268a46820
2ae47e5e4e8af4a4bd004e89ea833dcea2c5b719
7006 F20101114_AAARLL bach_m_Page_040thm.jpg
9789d5641453d92e89999f21c3b8c7e6
14199e2d540211fdc2e8c1ea0b78d869acc6f040
26600 F20101114_AAARKX bach_m_Page_029.QC.jpg
a5fbe04cc1a7ed28398eaf9ac613f797
e2181b54de5d9943e04c68cc76b203a9717b2341
24596 F20101114_AAAQIJ bach_m_Page_087.QC.jpg
8c825b8f60f12299d6aedea6a6ec9664
2d56df317467e1cb112b08fa6da2fcf303c8716d
6755 F20101114_AAAQHV bach_m_Page_022thm.jpg
993f2de58bb7547da2eab13f5bad59cd
62ffa068c5fdf999dbcdbc7334e2c109e475b1ea
24968 F20101114_AAARMB bach_m_Page_051.QC.jpg
8cd07a78e55da41db10d0ef8925800ca
996d95c71c30f9b8f746f52a4c33f14c6bbf08fc
25002 F20101114_AAARLM bach_m_Page_041.QC.jpg
417685193b07a98c1ffbaa4345451864
af9495507a01658d712013260bd987625d2aa7a6
6904 F20101114_AAARKY bach_m_Page_030thm.jpg
67d5bc2aa12e04d96c3f3938aa0be9ed
9d69d3be4d5e1dc62772d603735828f81752a029
25301 F20101114_AAAQIK bach_m_Page_130.QC.jpg
9c697c8d0cdef109c163246bae43abc5
b8fdc65bb5930d7bdf2ecc90d1ac9dc2335a67a4
54678 F20101114_AAAQHW bach_m_Page_091.pro
6c76adbad43ea45e75011dd977d6b05b
464d9dd26b391ac24aeb497a5e854089959096c5
19587 F20101114_AAARLN bach_m_Page_042.QC.jpg
6ca200dc7258ed8c88ff2d67892fe6cb
a69ae1817794e2fd66acdf0bb8e43993ec36fa0c
23455 F20101114_AAARKZ bach_m_Page_031.QC.jpg
34f3f9372e11df14a832bbaed9f53be1
65428a4f3a9d52637ddaad3071f469b02872fc24
F20101114_AAAQIL bach_m_Page_055.tif
2ea36bc2d0108d4a3f7f0546031c1de8
919b432b8c03ae2085bba3ccd203f90d6bd4ae9f
6836 F20101114_AAAQHX bach_m_Page_095thm.jpg
f771e24aac522029ab99398dadbe5295
83224a5c39b854f44776165d106a55b4e736f6a6
25607 F20101114_AAARMC bach_m_Page_053.QC.jpg
1800c2a718bdce1a2e07e6ff7d56a599
4ba6eb555e40545215e327c71b350ddf39390e70
21578 F20101114_AAARLO bach_m_Page_044.QC.jpg
b6a3bc1855521d00b137638caad42f48
4db417e85b2fea5a322cac39e0c07d91acc374ee
7266 F20101114_AAAQIM bach_m_Page_124thm.jpg
228c3d573701706e71762a98cd2120c5
2204222f938b711d9fc1883b4dd22a8c1d243c53
4896 F20101114_AAAQHY bach_m_Page_003.jp2
dc3fe1dcc91bc23c1fda6c13741e7114
c8038984848ef9cb04c4b1517115480433e02128
6621 F20101114_AAAQJA bach_m_Page_059thm.jpg
e20d091e0b57b46abfd03f48a4a78ce0
74fc92a4e073314659652304852cb09c15351eeb
6857 F20101114_AAARMD bach_m_Page_053thm.jpg
5abd6e08534fe1d889adf55fe6d5ea7d
1f7d838fa61c277e55e4a157f70beeda3f82bf7c
6469 F20101114_AAARLP bach_m_Page_044thm.jpg
31610fcb725a4c2a38a3b47ebe9fbea1
c4b83639df9891f2116be599f91f02f40302f324
F20101114_AAAQIN bach_m_Page_086.txt
87191a64e56c29263261b8a0631b2306
dc194201fae2a83216c711fe3267510bac2584f4
F20101114_AAAQHZ bach_m_Page_122.tif
3c7c08fab7708443b73b4c8a5bba97a3
a7eacf27afbae902e44789edec094f5ea7df3098
54209 F20101114_AAAQJB bach_m_Page_068.pro
baa6059ef47015c91e97a9f2923a3e05
822862df6949c5da81d7c170c33d87b580a802f9
24093 F20101114_AAARME bach_m_Page_054.QC.jpg
e497ad34c9142d3204ebaab951e5778d
f9d0dabd929929a0e713a3349d4a9d5e8fbe7ffb
13312 F20101114_AAARLQ bach_m_Page_045.QC.jpg
d7ace7868648b0033cb2b05ae8488d8a
4b5293e2c34e385bd9900b5bdd820bb7c2b7320b
116475 F20101114_AAAQIO bach_m_Page_024.jp2
cf86a15b64d534b5bd6bdbbfdd1417b8
fb272d629464e754020e935cae5e5b9eb8ae1ee9
111467 F20101114_AAAQJC bach_m_Page_094.jp2
40929863fbdd85bdff9c52a02aa21283
d1a710652b30fd7b4782f6bae2be17246dd2ff41
6515 F20101114_AAARMF bach_m_Page_055thm.jpg
5cc06617b106b8fdad1302d848a52374
42c621e891e66ea309254962617952dffc869c71
3963 F20101114_AAARLR bach_m_Page_045thm.jpg
aec840a0329d8d5e7a1221027198da7b
a98026f9579d0df23c8a507f7f5663c16c4b0edd
F20101114_AAAQIP bach_m_Page_111.tif
4984795c87aa403f1ba1faf1350b4e4e
2779aa8508a7d1580f0b760e6cc09981af7f3290
52249 F20101114_AAAQJD bach_m_Page_048.pro
7faa22bc087239a758872d5c944dd5bf
ae8779442758867e06f23193ce95828e54b3eaa6
11387 F20101114_AAARMG bach_m_Page_057.QC.jpg
27fc6bf0657312ab74d638e9161d17b8
598aac6cde41ee6417070aa20253517b4cbc5af6
22111 F20101114_AAARLS bach_m_Page_046.QC.jpg
fa8885544696c537d4f3551cafe60157
2dd8266aba3ffc1b768649b0bfabcb3bd8c15d04
24669 F20101114_AAAQIQ bach_m_Page_022.QC.jpg
3d26fd150324db4a6c39a592f62f74d8
216dec170303f86aa7cb0220144f35b884ede622
100632 F20101114_AAAQJE bach_m_Page_056.jp2
5a983ce9263cadf57aae752173d0ed76
d6faf2b165ba5b67080784bb88a31bb669177c28
3853 F20101114_AAARMH bach_m_Page_057thm.jpg
6eb03e9429da7d2a79888f0f173e4543
004a07432536030de3c4b47be140aeca3a36c7e1
6190 F20101114_AAARLT bach_m_Page_046thm.jpg
14a885405b1c0682dde54dfbfed7dfb7
4ccf4e1cf6342d440a2d2ce3eb9dcbdbe880800d
21834 F20101114_AAAQIR bach_m_Page_005.QC.jpg
c032e3018d9149abc7c874d1e61ed304
2d6402179fef09b7f5d07e6c9f2c76415d4bec2b
F20101114_AAAQJF bach_m_Page_040.tif
029d18cf447018c1edf6cc0becdbb873
c457e6f3357041036a63841b254d34f3be31ebd8
25199 F20101114_AAARMI bach_m_Page_058.QC.jpg
5fb60fa6f6d14a2a02c33a0addd39bac
a7194fbafa110b41454242c282a1353409642d4e
24959 F20101114_AAARLU bach_m_Page_047.QC.jpg
b60112e6656c4bfb2eb55c95232cb74a
128ff2b61e5e8203fce87ffefe346b0d27739686
815980 F20101114_AAAQIS bach_m_Page_042.jp2
73c2021872778584a1e08b4c234c3d12
7e23e7f7d0d8a31568cfb33fa959b4c68a54c769
47823 F20101114_AAAQJG bach_m_Page_120.pro
2a1c97390ad8698119a8003c6e43d479
b59c371f178366cee43fcf5061e64d8d4631bb1c
7002 F20101114_AAARMJ bach_m_Page_060thm.jpg
0eebf45ff2caa1d0af5161074ea8c7ab
6e6078e7062ea6a649764c1527a05ea21b38813d
6957 F20101114_AAARLV bach_m_Page_047thm.jpg
9c56153048a60c4d633092559a23a0a1
943ca6e61f496617d6ba68b059a7db65bbe6cfda
2138 F20101114_AAAQIT bach_m_Page_058.txt
18ce2af544f660b83b902ed6230242a5
4b3195a7c87520b93d5735429395de438c48a784
F20101114_AAAQJH bach_m_Page_061.txt
bbf98f3ed6be3351898e6c96fbb916cd
77b80a094f1c1ac40e528e56eb4e6f97841897b1
24756 F20101114_AAARMK bach_m_Page_061.QC.jpg
c2d3bb45b57d700c8d9f7cc71a0fa650
92eef70983d2e91dc5ea447d2a19a28693ceb60b
24578 F20101114_AAARLW bach_m_Page_048.QC.jpg
d161350b8477fc3c39b72181df8bc0ee
4652860efcb548d85c9d737e43badc30ab3cda61
F20101114_AAAQIU bach_m_Page_052.tif
dad9e27465666331c0851991f54cf4eb
9005df9e96b689e95b3ab21a18fa60579c1ccc53
2057 F20101114_AAAQJI bach_m_Page_097.txt
df3203addd6ca780fbc98b7c3ad06063
3e04da1761585cd494020165d2efda303d2fbd2b
8015 F20101114_AAARNA bach_m_Page_076.QC.jpg
6e54fb4b3b695b9beebdd778f69b20ff
52022f4d9f7a81829b87f786694276a934cfcbb1
6763 F20101114_AAARML bach_m_Page_061thm.jpg
ed3c7026fde24cd361d28784bf76afa3
66a3bc4820ece4a3a60901fc0a341ec8faffb537
F20101114_AAARLX bach_m_Page_048thm.jpg
1acfc5329d54552b2161785c4de927f2
3497c004253e481748b8ca79b68eb2261c0685df
51581 F20101114_AAAQIV bach_m_Page_016.pro
fb93171e32d02123db07589169fe963e
0dce5be8021ce73f1a4b37db8e198b8abc41b2de
14593 F20101114_AAAQJJ bach_m_Page_004.pro
125a48cd360844c1a5560722f1961316
bd6b838e5f2636a6b9d79631a726430c69bab9d7
26975 F20101114_AAARNB bach_m_Page_078.QC.jpg
1ab26283d117fcf63f0db436e599f1b6
97966fe5a623371914cbebbec8a088f72708229f
7110 F20101114_AAARMM bach_m_Page_062thm.jpg
696d1d05f2858add72c8069e0fc8f2cc
9ec934b55d88a9ba279a40cc9455a72e72eeb8ba
24989 F20101114_AAARLY bach_m_Page_049.QC.jpg
25ed4cd83f7ba3cd0f3713f35799e468
b733e73e69ba514c9da8181ca14c2ef7ecc98f28
F20101114_AAAQIW bach_m_Page_118.tif
a32213b187eeb195cbcbd48efab26ad9
7cdd6a870e2e6594bc1fab5741de8f46da301c5c
485 F20101114_AAAQJK bach_m_Page_001.txt
7052c6e70041651059bf7d23e612d2cd
99f5c93aac0190f7908a2d99c9a52cf9a74a2e3c
25755 F20101114_AAARNC bach_m_Page_079.QC.jpg
c91d595cddaf990ad90ca692097560c7
f93563ad3fcf7a02bc4c415a9732d424fbabc514
20212 F20101114_AAARMN bach_m_Page_063.QC.jpg
0cf2d4ea7dc7062997610577e904cc13
af3731f255e113aff6d0aeb970424299a22aa4f6
7037 F20101114_AAARLZ bach_m_Page_049thm.jpg
44e00590bb493e835f74887c8a7df9a9
00f01f11d657e3b1f5b9a2b8be2d9b1d2d054bf3
76386 F20101114_AAAQIX bach_m_Page_041.jpg
6fd64d519e0873b7670802f42d5b383e
0ad2ffe3b3db7739ba9f7773b021dba896aa5af4
F20101114_AAAQKA bach_m_Page_029.tif
0a674d2120434b20031ba7317f6c1c91
7b93e0349699e73236cdadafde169ee153c14000
45767 F20101114_AAAQJL bach_m_Page_014.pro
bd6709cb5d325f0cb2481266bdb17bef
f107622b1a0f3119be2d529bb7fc707ccccca0b4
5670 F20101114_AAARMO bach_m_Page_063thm.jpg
d623766b37d7a3d52c3b04e128a3072c
cd423c977b3340ef4da7d6534df34ae7e60f88e6
19799 F20101114_AAAQIY bach_m_Page_013.QC.jpg
59fd37738acc7b9cf3b074ec79c103d9
0aedb2a9045c073469d6cae61949364b9a0bba2c
F20101114_AAAQJM bach_m_Page_005.tif
375614dbf8866b079cb098208b9781ed
2ff574f421cfb3cf5bdedf3d12f7db763ee59a02
25168 F20101114_AAARND bach_m_Page_080.QC.jpg
0926f75b4ad3ee6709d52b9da71c6f8b
93c5e5dc208724f3b67ec47bc22daedb3b99e82d
14857 F20101114_AAARMP bach_m_Page_065.QC.jpg
62efa7425344c7a092a7b4bc8e23adb7
af6cbec97a25e41a2076a5dfe0c9d497eaac49e2
F20101114_AAAQIZ bach_m_Page_020.QC.jpg
6f89c755869ef05a44e5eb48434311bb
2d8be2106bd8ae99b228e1cd5bf2eb26925b117a
25655 F20101114_AAAQKB bach_m_Page_089.QC.jpg
ee2918c9aca2c7d64d1e4940690140a3
1b2e22d6c6b4273f80f66f433b1f8dcb6131ede4
50014 F20101114_AAAQJN bach_m_Page_071.pro
a66edf2af325886d4f345c7ef87a6b6c
f22ef9366c2b0009767207fd11e0407c96192f0e
6624 F20101114_AAARNE bach_m_Page_081thm.jpg
3b4df05397050e7abbd0e19722440ffa
0604c5a942cdf0135d2ef4dda7010cf7f5edf3c2
4342 F20101114_AAARMQ bach_m_Page_065thm.jpg
830ae0c87c6d91a4f85b73a209366dcf
0b83f099d6301995007ea1fb9f0a4024ff92a10f
6964 F20101114_AAAQKC bach_m_Page_034thm.jpg
d4dce910a6f4c50c06d532d9722a15e7
0eeec5ea3f1082273d9b0b5eaa4894521e0cc7b9
984 F20101114_AAAQJO bach_m_Page_002.pro
d3e0c156d442270a133984f50c99e018
3a13420389520218bc37cebb2921f6786528bcf7
25048 F20101114_AAARNF bach_m_Page_082.QC.jpg
bb57135903b3ed99acf12aa1f039531f
845e6d2073f293ccf6f6e66396a97401aa0290c8
7614 F20101114_AAARMR bach_m_Page_066.QC.jpg
d8b21a5e38a42e8f7dc7ae052c5d6e74
be1a0e54612cb36288c7fb616b89b7e3b1812eff
61126 F20101114_AAAQKD bach_m_Page_009.pro
4ca92eb39dd5f788b64e774c4f2c413a
74db6cbcb30b58ef313640dde14501ee9c91abb2
F20101114_AAAQJP bach_m_Page_044.tif
0e526648880b12cd194dc65398e7add2
64a9479be561a7765c31b2e96987f969e83d5f57
6918 F20101114_AAARNG bach_m_Page_082thm.jpg
4b25a95c3ba509867c926e6e7f3a2dc2
bee33ca1708850747934603fb7958d9adb491a3c
25221 F20101114_AAARMS bach_m_Page_067.QC.jpg
49015a35ea42c7379f2c90fd998db709
9d0d1f6e919c800fa96bf1c9d5b060719c9a7f3c
2150 F20101114_AAAQKE bach_m_Page_015.txt
37099c8d1bf2fd0fdd6b4ae6a775a507
e681a1e3b09816485e8abbddad44ac60a7cc2b0d
56300 F20101114_AAAQJQ bach_m_Page_040.pro
bb17d8bd3f10fcb443e0ffd40e5a9c30
3c0121598c39bdd24f769fc7c983ded8ce02e34f
6653 F20101114_AAARNH bach_m_Page_083thm.jpg
cd657c8c529404c2cea95cbabe5369e6
6f2139da4342fc823cd4e72225fae70d0d65b6d4
25666 F20101114_AAARMT bach_m_Page_068.QC.jpg
daea52bee4a0a5eb5527f1a73eafde6a
1f9c1fb284d00cb203319333787975e1ad9b49b2
2194 F20101114_AAAQKF bach_m_Page_084.txt
3e3f3c2baca1c3efebcd87734e809b4d
7fb8b4d9daef41eb044d306c0c6c04907c033e77
2127 F20101114_AAAQJR bach_m_Page_038.txt
d9ef24d71589d0e98e00b40a94698743
91ffbac286d9073d11a53c4c5f6463422891f266
25404 F20101114_AAARNI bach_m_Page_084.QC.jpg
5130cb2276944463eb718ee16580bc00
710d47a676f63ec339f2fbc7f5f5ccfac9f3f087
6851 F20101114_AAARMU bach_m_Page_068thm.jpg
4a854dabf7ee085ed9b89c0ca8d114a0
e74db6d3c025b9d8dcc52ccc0c134457fe912b6d
74365 F20101114_AAAQKG bach_m_Page_086.jpg
176c0d30e2c56ac7824e230290295725
81f341f8f7b4f60da3b201dbada3785801a48bd5
96509 F20101114_AAAQJS bach_m_Page_073.jp2
c381d0c75cdf321329cc88d62dd914cc
ae699e006e2a5fdd2f34bb45ca40c1002c6a5c72
7209 F20101114_AAARNJ bach_m_Page_084thm.jpg
70c619ca2a58ed04e36809650ef58bdb
842772114ebb5a3c7826204fd0a3d4a10afd5259
6925 F20101114_AAARMV bach_m_Page_069thm.jpg
13961af30c61d75c74adf90843f5e2ef
77785d04b459923946c6156079f9cada0783c59e
54489 F20101114_AAAQKH bach_m_Page_093.pro
56c7691af581d1f648d7631ea4634a31
dce81ed92a3e72e603b740fde9d2d00277637bb4
6513 F20101114_AAAQJT bach_m_Page_052thm.jpg
49a9ebe5c0cb8d70e52a69581769ee13
0bd02dbe276c3544f0e9cb430d4137f1b7075829
6922 F20101114_AAARNK bach_m_Page_087thm.jpg
7bdf318308ca590eb988318b9cfef5bc
2bff23e04acd79bc29b9fe647dbdcf5b4106aaec
25638 F20101114_AAARMW bach_m_Page_070.QC.jpg
5c4b7ab6da971987963826d676a6eb0e
807d5ecae32b33875f782fe2dfab9b83ccc9387a
83719 F20101114_AAAQKI bach_m_Page_005.jpg
1683b8359ab8f7637e14c68a9b89d633
f8f6dbb8fb8cc5601f8625cd4cf584698595855c
F20101114_AAAQJU bach_m_Page_043.tif
8515fa89fe56853c880086168202f42e
fdc4d9b9ce434ea5da93ff2cf8f384099ba45018
15351 F20101114_AAAROA bach_m_Page_101.QC.jpg
70d08afb130bfc7c8fa36f8aa6c099da
8eec05a5f8236fd10d9184a9b6fb02f7ec751317
6758 F20101114_AAARNL bach_m_Page_089thm.jpg
510de4e097723ecda09a29cb65857073
83aeae8e8abf315173856eb94dd22cd6e3404fcb
6975 F20101114_AAARMX bach_m_Page_070thm.jpg
a22f1b5a54c1d84ab145e51a04bbc796
e447950a77ec14f11a921242bcaada3b585741d7
56985 F20101114_AAAQKJ bach_m_Page_032.pro
66a22bfc0fa038e42261177926f911e4
fcc64927cbfa5b235118ec87ee809aaedb9a55e8
26262 F20101114_AAAQJV bach_m_Page_026.QC.jpg
c7950a45bd30a1041240305f55862904
08fbea6264a26d59dc4ba4ae1269dfcb1c02c595
4556 F20101114_AAAROB bach_m_Page_101thm.jpg
42ba78bd4f07afaff9803e335831b634
8693846ebbab56bd0d77df3ded333ffc55ecc7b3
25386 F20101114_AAARNM bach_m_Page_091.QC.jpg
5e9f47ba5ffa17ab615b2fa8a9ceca9a
e112a295ccd3cdfe868fc40cd2c723eb3dd6cdef
6004 F20101114_AAARMY bach_m_Page_073thm.jpg
cb6febff1e66315b0c97b44035329229
26e3b52147df2ae3a0fa9b2849587c0961da65ec
F20101114_AAAQKK bach_m_Page_047.pro
3e2eeb858be2304c405c748da454c1de
7735f91e873563b310064b61e6fa7e7dc833d870
119787 F20101114_AAAQJW bach_m_Page_035.jp2
5fa645d221981721bf65054ead57a46e
5957c3a448294b0f7cedabead5c203ed4f8813a6
4229 F20101114_AAAROC bach_m_Page_102thm.jpg
987ad8449884e96badf5ef257a2c7ea9
c4b9564bfa1f982265bf888144a283af1c4414ff
F20101114_AAARNN bach_m_Page_091thm.jpg
4420a010701d983632f1dd170592c5ef
90ddd7f8e4bcfcbccf860603671da7ec8b1d1a29
12772 F20101114_AAARMZ bach_m_Page_074.QC.jpg
38c2f0e6eb191038d0f0faeb658327ba
0c538cdd033fcdd60cb4814e909b3c3ae14059bb
51520 F20101114_AAAQLA bach_m_Page_072.pro
9ad577df4b137bcb31bb43b0f0a58c4e
26481f3e8ee9d52060b6a4eb9e4203ec1948025d
1555 F20101114_AAAQKL bach_m_Page_131thm.jpg
377b17bf24f058ae4bc5c619fb331e0a
4d47c3eb0cbbe76f480ae33501eeb446dabf1aa1
19977 F20101114_AAAQJX bach_m_Page_011.pro
32f6a3962e03c53510a9a8a05c3d396f
1a5dcd5723ba7eeb0e1d43288a120a23302e5b39
13841 F20101114_AAAROD bach_m_Page_103.QC.jpg
94a006a61c800ac38cae9622efe9c312
d0a380a1ac4120b7935f81f06a7294baabb958c4
7160 F20101114_AAARNO bach_m_Page_092thm.jpg
0810ddf4fbc0b43915116e23103d8ee7
12472904fbb31a351dbbf266b0c04299d623ae8a
F20101114_AAAQKM bach_m_Page_081.tif
532e89c9d3ccb169668fe460690afd23
8431db1aac8063f17303ce0bf0b367399f6db530
64182 F20101114_AAAQJY bach_m_Page_098.jpg
d536ce3ccdcadf72b636c7bfe6873293
edcc9d9735402580029c07aa0772dcb7b891f083
6685 F20101114_AAAQLB bach_m_Page_072thm.jpg
58a1854da6d525ff531c3cc6b39e0565
9e38b99afaeef12f12ce6e53fdd4851887782835
25731 F20101114_AAARNP bach_m_Page_093.QC.jpg
3b5b04e7ada8a8e7896cad6321024dcc
3de2ab1610093b6de7c3dc672fc8202208931b5a
2203 F20101114_AAAQKN bach_m_Page_019.txt
23f50f67fe92a2855069d93795da5eb2
47a6326501e355e1412a5ed29b2b908cb6dda5bf
76896 F20101114_AAAQJZ bach_m_Page_099.jp2
88b9ae9f45686339cb2c554489dda00c
6623cdb5183198a32fa1e5fae5d78e431b72c20f
4144 F20101114_AAAROE bach_m_Page_103thm.jpg
c99484befe7c8e812aad5f075f3c1d48
1ef215c40ce0c5f66a1b3b5e687208a22b6c28dd
7098 F20101114_AAARNQ bach_m_Page_093thm.jpg
e46eb405af5630983f24e5b6d68d506d
415422c4cc3e06ad2035c48a857d384a5de7668c
46840 F20101114_AAAQKO bach_m_Page_056.pro
d3022d0758049379ff5d9756d89a2c15
4a443709cd2d434162ab559e0b1859948247a9b9
110664 F20101114_AAAQLC bach_m_Page_059.jp2
d070bfff474f9cad5ab425d7383ad77f
c8a23eca9e82df0a68162924fff6142479bb0c5a
4162 F20101114_AAAROF bach_m_Page_104thm.jpg
d248b9ac7ac74598008fdd6f548338f0
fb82ec06edd8a56de79d31e4538dbc2dde01e66d
24903 F20101114_AAARNR bach_m_Page_094.QC.jpg
c021ed4d0fcf23ff8b4ac547e0c5568a
c621171f87cc3d04555aabb65d3e90b3a4ed2792
119637 F20101114_AAAQKP bach_m_Page_037.jp2
666eeca8df8f51aa8415f26bb27a300f
499d301d4edc37c776d3627501be90fcd3cdbbc0
15028 F20101114_AAAQLD bach_m_Page_108.QC.jpg
7cbe976584eef798f51b4815ec45e3d9
05d2d3f161f49fb2de95da026b4f9421e4cd748c
12550 F20101114_AAAROG bach_m_Page_105.QC.jpg
ef8990deae4d645c66826e1a6860b1a0
cafe76bb95f4d8a7541ca58fe8d86e667146dd3f
F20101114_AAARNS bach_m_Page_094thm.jpg
1cff968f7fa949f5d09b2b35a518b70d
4cb7b8253072632868b8e7540207a98013d76178
1051947 F20101114_AAAQKQ bach_m_Page_008.jp2
9a53b918c6e24d456d5cad070639c75a
f90de57fb268e9b7e10c0f52402a2adf3721e68e
79231 F20101114_AAAQLE bach_m_Page_026.jpg
66c46c3717cbf3a4df6c643494e5844c
dbeebb42c87f57c8816f8fc327d9a13b10ee783b
3758 F20101114_AAAROH bach_m_Page_105thm.jpg
f69c2d2d92dd4214a7930c7ec30b360f
69d7343eebe88e6f8473f7261a40161d922e6459
24154 F20101114_AAARNT bach_m_Page_095.QC.jpg
795fc0a1cda0ae73c98d15f13cb52c1e
c300e7f4c5d6827c86a205444b8165abd4cd84df
114739 F20101114_AAAQKR bach_m_Page_020.jp2
3fe78e47b63cfdcd47bbaebad631961b
4c4ee071295a989841ec6410c4c2586866350fc2
F20101114_AAAQLF bach_m_Page_076.tif
4c76da3561c696aff5622eec47c798e4
5b8e992a09e3f833586578a80f0cb89479a62f52
3070 F20101114_AAAROI bach_m_Page_106thm.jpg
0da328a7c39fd3d03348eb1f6c612c4f
8bb08eca3c18f33be1acc86afb69f455710cc1e9
25046 F20101114_AAARNU bach_m_Page_097.QC.jpg
b95ba02bccb2dbafbfeb049247206dd3
c30c2b4e0ee74171f74c7724b62333689255aba9
17258 F20101114_AAAQKS bach_m_Page_107.QC.jpg
09104220d1d1789b2e9f3b54e1fd0fde
707048142cdc29a4013730683896bf4f7cb47c55
56081 F20101114_AAAQLG bach_m_Page_019.pro
0f426bbf3017b3ef108b0e4e31c83dd4
0d65b9b7f137b11fcf76439567fe53f03c09efc5
5125 F20101114_AAAROJ bach_m_Page_107thm.jpg
8b5e44059ccc93e6f11fc2092230c04c
58918cd6802d9cc34830e4fbaebbfd4c892781ae
6898 F20101114_AAARNV bach_m_Page_097thm.jpg
e116bbfc22dc73c706d0b2374a785a1a
892cfe09c04018ed352a5b537eac135e014722f7
119410 F20101114_AAAQKT bach_m_Page_096.jp2
ca6826fde80f068ae79fccfc63486cd0
bd0cafbd3ef2104bc60bd79f9a89012cca00fd07
6235 F20101114_AAAQLH bach_m_Page_104.pro
38b123c09126150305226a0bf7f47020
3c34783424b337462fde1cfff90fdce824ef576b
16162 F20101114_AAAROK bach_m_Page_109.QC.jpg
a1465766010c70946325cd9111d5df64
d5eaf8b0f2a2b4272fa6ff7eb30c9dc669437395
20620 F20101114_AAARNW bach_m_Page_098.QC.jpg
d0c985c9cd32a19e6f091f430f0d2b47
e6665690eeed3de185c6c3c84dca0c53ef77de9d
2916 F20101114_AAAQKU bach_m_Page_119thm.jpg
bdcbd1db0a32b7f84e7b28ab0ccfc27c
95ce2ae33320e63eb7aeee497d3f089ec5a83140
F20101114_AAAQLI bach_m_Page_024.txt
76cd494b23ac233c88be540e49bd70b9
4d3d4c29ff104e725a45f79a418d693e294c1c29
6873 F20101114_AAARPA bach_m_Page_126thm.jpg
f4a7ec19c298440b795403ef423d36a3
ba61c28b5f0dbc22e80506c2804b7a5c05ea4dad
F20101114_AAAROL bach_m_Page_111.QC.jpg
60d7896f0229e15e3e8f81984eac744e
f8e5c3ded4069737c8f03ea6b426092aebdab4fd
5699 F20101114_AAARNX bach_m_Page_098thm.jpg
4fed0674fcc45e9498390078ec2badb6
ea0e2ab4a3bb82193832ebb713429eca30548ddb
2083 F20101114_AAAQKV bach_m_Page_022.txt
6e4770e3b1b16f5ded1e6926fd7b75a4
f9267ae0287f371e6f446c99c6177bfacd759bcc
6968 F20101114_AAAQLJ bach_m_Page_079thm.jpg
9dc6285a7b696815e8411297fbe21bc4
af3364d436f3cce15169ce5ddbf3df38c9ba8819
24791 F20101114_AAARPB bach_m_Page_127.QC.jpg
4b39d68460bc7b09a243087a4b11ef69
e33f4bf177909d76d7310f8466571b83853c20b1
6641 F20101114_AAAROM bach_m_Page_111thm.jpg
d5f7e6355a7fd4bfc57e01af7103954b
e4012df78189fb3f96212428da0c2f199df33230
4815 F20101114_AAARNY bach_m_Page_099thm.jpg
e2883994705a026eea77f1bc7f9694e3
48bb233dba14ee758c3a007ec1a23f9c99ab5c37
F20101114_AAAQKW bach_m_Page_057.tif
e9d6c45d0a530b2274183db74d0092df
66cd23a75a191f6d301fd6874a7bb8ef8ede7e19
7682 F20101114_AAAQLK bach_m_Page_113.QC.jpg
6753dc0788821385d663adb7568233ab
e615f1e30fc616523c88c48cdcfc8728b222a419
F20101114_AAARPC bach_m_Page_127thm.jpg
01ebae7ee49e957c53fb2b8727fa5fc6
001dcb2dacdb2c6706c69793e69d568737f125b5
2716 F20101114_AAARON bach_m_Page_113thm.jpg
69c4389e2353f1e6cd1d23312d49179e
fdd66c930a22081a82dde96dedb7999fae693270
20553 F20101114_AAARNZ bach_m_Page_100.QC.jpg
013a8749192fdda09c7570b08943a22e
510487b4b29945c0237d71ad266145a466a9b3db
102213 F20101114_AAAQKX bach_m_Page_010.jp2
796d3b8968ebdadef3e4beb240ceac38
a0592fe5bb06661775651fac02de159a600e0a49
6234 F20101114_AAAQMA bach_m_Page_014thm.jpg
2a4c1d72cd2049d5d4694df07722719f
bab2ed97f85090be6358d093af334a79a3ea69af
2256 F20101114_AAAQLL bach_m_Page_044.txt
2a4e4f0118be6e2c7c818ff367ff09a2
b504a0105935dcb3191f518c1bec212867cafefc
7161 F20101114_AAARPD bach_m_Page_128thm.jpg
191d9ce3b8ee804d6c6ecf8a1b5ba02c
95edbaca8491ac3157e0a7e63ae3c61b104c9ce4
4777 F20101114_AAAROO bach_m_Page_115thm.jpg
ce76d4a0b8d6114ab63896bba52592d2
621697f9bb34eff9bd9ffc82a9dbd1b47439d35c
51118 F20101114_AAAQKY bach_m_Page_059.pro
dfc61a2958cdc214ec148623352faf65
ae15bfa4a7901398e7594ad909222668545d284c
7757 F20101114_AAAQMB bach_m_Page_001.pro
e608d9e0758b29c17a02a7ee7503387e
7ca4c403f27a498939b356c96e98b25564e5f4a7
6743 F20101114_AAAQLM bach_m_Page_021thm.jpg
035940b14155e29272c1598a9881faa7
5052cdbea8f9010ab4115fafc36a9ccdf8f12162
7366 F20101114_AAARPE bach_m_Page_129thm.jpg
1b270e6149bb3e5b00d67a3c11f4d602
0958c7e5a542d7cedb4721489c04feffd016a08a
4184 F20101114_AAAROP bach_m_Page_116thm.jpg
7cde59ff0972b023ccefec761a1585f3
625c0893528e2acd3a80c439fdb9583ea7bfa985
F20101114_AAAQKZ bach_m_Page_060.tif
a554035e4d24abaca63f8365743515c4
48efc717317c23690d6d819d4182212661d184e8
1545 F20101114_AAAQMC bach_m_Page_116.txt
b59067456819c280c31156419d41fddc
fc72f057780d0592806ee8b397473a90f480e260
F20101114_AAAQLN bach_m_Page_015.tif
2a1089fbdec614e38a5c9e0f88cd41a6
57da52bae3d2c161771f6f5cf121d78660ac6d45
4798 F20101114_AAAROQ bach_m_Page_117thm.jpg
54697512448475e0c883a304dfbdc3ce
cf1b11760cd0c9f47415a1cd3c36ec62247a4b42
73667 F20101114_AAAQLO bach_m_Page_017.jpg
11fe7b3d44ee7c952e30e5e4e3b40676
73d562e86d240cda16e52b123d9ff05c8ca59f0a
6900 F20101114_AAARPF bach_m_Page_130thm.jpg
43bc9c8e6651de056ed3fb5406e1f9e1
88f6fac9fcbace95be8b2f2e1ebb22a67ace75c6
9071 F20101114_AAAROR bach_m_Page_119.QC.jpg
d2a03118361eb00a1b05d00b6bbed3d4
56e420ae2f9a7e24e423a0e315381d29a371e50b
F20101114_AAAQMD bach_m_Page_086.tif
92ea72f012914dc78f1a8046ffdd512e
c826cd539c203683f5ce81421d7b0aa50e4def0e
102578 F20101114_AAAQLP bach_m_Page_006.jpg
bb76db83fd827d3f89c73a17c66f4bfc
086a0152ada9d319a1c050d3382217d05871de27
11011 F20101114_AAARPG bach_m_Page_132.QC.jpg
7ced1970ed54286687d8396ecbb228c0
0db73dc408fcac94912d9a474dc3aaf9a1982a20
18660 F20101114_AAAROS bach_m_Page_120.QC.jpg
7a8ff0e5a3accac326cbf15851140e04
19ee191669da20c59bde39aa7685b6b738d99bab
1051977 F20101114_AAAQME bach_m_Page_006.jp2
d1605eafbf129cdddb43c281b6a72631
1db1a87ac2e77fd79c15875e0016b02372be20f0
6753 F20101114_AAAQLQ bach_m_Page_054thm.jpg
6c6cf0aff4c9eba1f7fce56b9d28f364
6cedf44eca7348a012ee815f85c3b59e6b9afa09
3331 F20101114_AAARPH bach_m_Page_132thm.jpg
9ad28b126f6ad3ed15f9d85b21a99006
f4cb9e185efd44d2349bec49b43dfeb7fc59bb21
2920 F20101114_AAAROT bach_m_Page_121thm.jpg
e14d397bc6b53d439cb1f3e2e44e7ea5
3b0e3c0543c6fad99465167a39b52d8589688a3e
F20101114_AAAQMF bach_m_Page_001.tif
efa7679fc4d31dc5bbca1157d3a65dcf
7d0d81c93a3587f9d71d83ac02537e65e606af74
24633 F20101114_AAAQLR bach_m_Page_050.QC.jpg
bc8638ba3fffa08213744d7021b9bb5b
2ff5c22ebca23d68bd4f7d250c48e9ebb8e37421
151426 F20101114_AAARPI UFE0021222_00001.mets
0ca1f29ad39faeeb97cddea24474b1ce
d49a8209a9162281c37198bf2e90dea6401f5a8e
14715 F20101114_AAAROU bach_m_Page_122.QC.jpg
c95d678ccd5ac125e50526d3c80962a3
2c9af1c344a55a346e6d0af67a23a842f3af9df3
3067 F20101114_AAAQMG bach_m_Page_003.QC.jpg
3b743237689ef1dfd0d7a3bd37bd50ab
7244ab5bd7c7ba4ae625a430e5de5bcddae4494a
1758 F20101114_AAAQLS bach_m_Page_073.txt
ea060c0306ac187707927d36af009c70
7338dcca0c410f956b0da877f4686a328b625429
4348 F20101114_AAAROV bach_m_Page_122thm.jpg
1f2635f5ec5e96f16f9183c0b72380f5
9dde32a7700425ce1dbc23c4d018c53d51526c05
16280 F20101114_AAAQMH bach_m_Page_117.QC.jpg
9339d2b65ff3a12274216e9c65a5d5d5
48b40be737b192fbad353cb34b83cc9cc2524f08
10121 F20101114_AAAQLT bach_m_Page_002.jpg
b264a7d3d2bdecbebff10025f9b0ae26
fe5ab2e4cc27a4bdf4b7b3c1edbdb3a531825dec
26376 F20101114_AAAROW bach_m_Page_123.QC.jpg
d553dd50153238c9844d190e85b95682
07bd3c2eb53b8293475572feda3316ec3b73de46
124222 F20101114_AAAQMI bach_m_Page_127.jp2
d586f6550286c4cdf118a5a8fa42ca4c
63e492cb59258f4c919c2f1b9d2268feb894237e
29225 F20101114_AAAQLU bach_m_Page_119.jpg
5de1d0c6b4d57eb18b761272326ff1a9
7d7207bfb39179d51f25d48af8cf0b98125d31c7
7019 F20101114_AAAROX bach_m_Page_123thm.jpg
6ec28161fea0e8fca49f77f1648d2c92
b3fd8bcce733bb05f8802d024eeaf16977db5feb
90434 F20101114_AAAQMJ bach_m_Page_129.jpg
c06ae652c4c60dc6efd33bad3c587ce8
ccb9cd3abd8c6cf25b9348da8b40ad85241a64d2
F20101114_AAAQLV bach_m_Page_068.tif
7f29a0932aeaff933c80b92cb537fdf0
caaaccf2c852f5084fea043a46e98ddcdaee0fca
26953 F20101114_AAAROY bach_m_Page_124.QC.jpg
0b392a90223c36af65df50ec5778740a
265f2022951357bd24ab8b14a9507f05834a4de4
1993 F20101114_AAAQMK bach_m_Page_018.txt
c5adf02ca35eb86cd7e0ff55c6d9743e
932a21647c8bbe1d1d37cd8f6f0f66b0237ef036
9881 F20101114_AAAQLW bach_m_Page_003.jpg
9d055406f26a9d92517e86b5f889f4e1
7e4a3c19ad494036205b96fc93cd1681feb538bf
24319 F20101114_AAAROZ bach_m_Page_126.QC.jpg
0aefa2965d186dc5d51839587b16827a
aaa3b26d1b107d4ffcbad4f3f780b1f53f75f268
24546 F20101114_AAAQNA bach_m_Page_086.QC.jpg
b57207a555e3732fe8f4bb9561258932
305df9bd19ed34fa37a77e62f8242cec2450ad92
118654 F20101114_AAAQML bach_m_Page_039.jp2
c8213d4a2b2b8e9426d2041cbc42106a
12cd693b097fd353ce82c870ca227184314e3950
36753 F20101114_AAAQLX bach_m_Page_057.jpg
3fdfe6df70fc57f9dc84884c7d5db1e8
aa3c8582f1db6ccd4b773c40e3d9b8f961d51c07
53101 F20101114_AAAQNB bach_m_Page_054.pro
bcbe716a2649fda00a8f55d83a86a49e
07549b2294a394ec4fe5aa5832be11b5d0c7981c
2604 F20101114_AAAQMM bach_m_Page_066thm.jpg
ef843a133a2006e4bd1c2ecd668720f9
b42d8223007aed78a9d0b5f93b470def353e7ed1
1239 F20101114_AAAQLY bach_m_Page_042.txt
c167a63e1456a10b0257e28e545d0a3f
b6af8c53efd970cdf91b3655d432581be211ed66
F20101114_AAAQNC bach_m_Page_027.txt
7f5d3a11c7540acf74d17aef557c56ff
fd2439edef15b08b2937f4744128fb56f3ab69a9
7259 F20101114_AAAQMN bach_m_Page_001.QC.jpg
1b3b73fb4aec66a4a574933cdc7dab63
3d56ce88b1aac00d1c30e02259ae34d5a2fd8647
61035 F20101114_AAAQLZ bach_m_Page_013.jpg
6d3b92f24acb3afde9b8d5c87618c595
74fdb19025e5403b7e20ad00949dc17b1b727b0c
F20101114_AAAQND bach_m_Page_064.tif
96377cac2844245095376a6cc724459f
0c76fd90464d3215fc7471fb45fcdd0e7fa19f5c
2077 F20101114_AAAQMO bach_m_Page_017.txt
653b050a111011abba379ffaee2fc12b
f822985ddd7e0d59a9c5175856554a13ecc28018
24290 F20101114_AAAQMP bach_m_Page_016.QC.jpg
be8b37d46c015fe19d45baf59fb65fca
064e4341830a376c9accd3fd099f8fa0ae46188e
5301 F20101114_AAAQNE bach_m_Page_005thm.jpg
c6568674a6b4f426460d14373dae9876
29e0c56a62d158629826e047dfa8dbf08b307367
6752 F20101114_AAAQMQ bach_m_Page_067thm.jpg
33db995882a0faebcd0a6cb8503e7664
9cb7a48759d58321e7e10ee59b16ea97dc2bbbb7
115106 F20101114_AAAQNF bach_m_Page_090.jp2
9c32d5714ff90f1e7654bac7d0e27fa1
0f156970a190e99edc6b8d3292bb631cc63eba30
72043 F20101114_AAAQMR bach_m_Page_111.jpg
5fbacda420995ed7578d69bd565fa631
761bbdf7f6a91922ec308eefcb82bf64a0bcb146
25836 F20101114_AAAQNG bach_m_Page_088.QC.jpg
880e815e9870ca1967d3bbd0a6afe083
3dbd3597b41b04bdf82f7bea4ebd3dc85b39b6f8
5571 F20101114_AAAQMS bach_m_Page_118thm.jpg
dd3429de856ab5d61f7d62724fde1194
3e63584075884214300a30dd2796b5e7001a9ae7
113696 F20101114_AAAQNH bach_m_Page_097.jp2
0587fad197c97a19cb43c17a6555b56d
6f5a9699bedd0a1d986665b444b8ec0dd6a6cf96
22009 F20101114_AAAQMT bach_m_Page_036.QC.jpg
40ea7d4acd382cd254e4bce7f0653750
fb0ad4a85ee6c909b062d458999afc0495324a01
76708 F20101114_AAAQNI bach_m_Page_088.jpg
9ac45de4396c5ee48b2e71626f40eb86
0c51f31f9ea2622cc4ddfee3852f9dc9e370beee
112393 F20101114_AAAQMU bach_m_Page_048.jp2
ea399b39e746d7dc512a47f954d41e16
89411df589b583c2390a3a05636bda00db3822a4
24409 F20101114_AAAQNJ bach_m_Page_030.QC.jpg
ea614edebc6da18bf60d9ef4ca947503
466ef227be35c2f873aca8aee08b5658c8b01ed0
24554 F20101114_AAAQMV bach_m_Page_028.QC.jpg
258ec43051f1415d148a2231799cc72b
6d8ffe60429d5357a14a1b60e401db78ebdeeaf3
77050 F20101114_AAAQNK bach_m_Page_034.jpg
be815586d61abae900a9198306917b43
f57797b2fa1b53ce4392b611c9008f3a49820171
F20101114_AAAQMW bach_m_Page_046.tif
64881f92c8381840a20067ee34403c11
236d899cd8e334c0708f725d35fc991adf2cc8db
124456 F20101114_AAAQNL bach_m_Page_077.jp2
b26c877cc4c1295c11368935f1599f93
f7048e1a10fc57e9988ec5c8b0fbba60c76152a3
2049 F20101114_AAAQMX bach_m_Page_059.txt
70eccea0581be9dc34e0e60a2adab200
dd57394661c7f8a75dc346341081db6a7feee2b7
58925 F20101114_AAAQOA bach_m_Page_008.jpg
289ff104200161efb40dcdf617d564c8
bb3fdb45fb48f4a3e41d1c0fb87908ddf68cd854
7276 F20101114_AAAQNM bach_m_Page_078thm.jpg
694fae875f76638036bec47bfb0b0427
5fddeb90def47fff6b0f8e557606d4840569afb3
75841 F20101114_AAAQMY bach_m_Page_080.jpg
2b4ae985d0bcc38d34d151cbb545bb2f
918bee7f1321eee0e0065aeeda4ca1df80a5a11b
1691 F20101114_AAAQOB bach_m_Page_100.txt
513a880fc6b3e6303444804ebd814d6f
61451ed85a0fc7a60f878187bff6a9e01394becf
6149 F20101114_AAAQNN bach_m_Page_056thm.jpg
139cef41537646bb1c1827dc520bd641
38219333d49d7d9a00423b9b610806dc79a55ab3
F20101114_AAAQMZ bach_m_Page_097.tif
00d0d7da3924d9b42422efe34b6502de
f95b734fb1a85d1940f645c55b4c05c14afe2927
26648 F20101114_AAAQOC bach_m_Page_033.QC.jpg
9cbf23cfbd67eaec24939f237ff11001
7ad5c129b3fa3f917d99c57a6296b673dc40f9d8
44906 F20101114_AAAQNO bach_m_Page_105.jpg
cf406ce7ae3e26f01b62678638b1cd3d
32c0ba2c9082d1c6e82aa77af7c092d93ec9b90f
115947 F20101114_AAAQOD bach_m_Page_070.jp2
323c1285a1f797f5315abad91e9250ba
d5c3c2f9fdaaf139c4b0fd097b2a217376738d04
109096 F20101114_AAAQNP bach_m_Page_083.jp2
2cddc7f9e7b49cb4d9a61a6172a43bb7
7f039b2c29522eaab49ca156277c1f0df4538b1f
2019 F20101114_AAAQOE bach_m_Page_085.txt
642a7a0013250f17bc67a6936cb739cb
e9e8d5155618f6fd1626557bf4a3568bb1a2edd7
2181 F20101114_AAAQNQ bach_m_Page_037.txt
3f2ee4a71c53962ecfd387acdb5ea8c4
ce10222b44aeb09fe782180b21e83e497559cb07
F20101114_AAAQNR bach_m_Page_090.tif
1903726290ea068fb6f4ec2f114b9659
b24bf20499af9ad9ba3ef1d973712502cd399535
25947 F20101114_AAAQOF bach_m_Page_039.QC.jpg
6330f6979f1b7f9b57cff8033566f342
911ba832fd19219f7e888dcf96be44692384bfdf
40417 F20101114_AAAQNS bach_m_Page_074.jpg
75b92e6030c85f71416b9ca9fca0ae5e
2e42ad4280a2ae39ae97730c0c8b35dbef1c7b58
F20101114_AAAQOG bach_m_Page_073.tif
c66e2b08190446fd843af730039de5bd
efb6cd1ef69fc2f972595f70b09ba33c6e338798
2080 F20101114_AAAQNT bach_m_Page_049.txt
8f6837c129ddc56f203dcc460e650eb9
01161bb992bf49772204e0d74acc2ec134bb6006
F20101114_AAAQOH bach_m_Page_129.tif
b5b77186e8fd160a84bfacd86ef193ee
59709f94efea3803bb5c48ed678a07a3f9538890