Citation
Photocatalytic Oxidation of Selected Organic Contaminants and Inactivation of Microorganisms in a Continuous Flow Reactor Packed with Titania-Doped Silica

Material Information

Title:
Photocatalytic Oxidation of Selected Organic Contaminants and Inactivation of Microorganisms in a Continuous Flow Reactor Packed with Titania-Doped Silica
Copyright Date:
2008

Subjects

Subjects / Keywords:
Adsorption ( jstor )
Bacteria ( jstor )
Bacteriophages ( jstor )
Bicarbonates ( jstor )
Lamps ( jstor )
Methacrylates ( jstor )
Oxidation ( jstor )
Oxygen ( jstor )
pH ( jstor )
Wastewater ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
7/30/2007

Downloads

This item has the following downloads:


Full Text












PHOTOCATALYTIC OXIDATION OF SELECTED ORGANIC CONTAMINANTS
AND INACTIVATION OF MICROORGANISMS IN A CONTINUOUS FLOW
REACTOR PACKED WITH TITANIA-DOPED SILICA












By

MARY JOANNE GARTON


A THESIS PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
MASTER OF ENGINEERING




UNIVERSITY OF FLORIDA


2005





























Copyright 2005
by

Mary Joanne Garton
















ACKNOWLEDGEMENTS

I would like to thank my advisor, Dr. Paul Chadik of the Department of

Environmental Engineering Sciences at the University of Florida, for giving me the

opportunity to work for him and for his guidance and assistance throughout my

undergraduate and graduate education. I would also like to thank the other members of

my committee, Dr. David Mazyck and Dr. Angela Lindner, for their assistance and

suggestions throughout this research. I would like to acknowledge Dr. Samuel Farrah of

the Department of Microbiology and Cell Sciences, who provided much guidance with

the microbial aspect of this research and allowed me to use his laboratory facilities.

I would like to acknowledge Dr. Matthew Booth of the Department of

Environmental Engineering Sciences, who assisted in the gas chromatography. I also

would like to thank Mickal Witwer, an environmental engineering doctoral candidate,

and Shannon McQuaig, a microbiology graduate student at the University of Florida, who

provided limitless assistance in the experimentation and analyses conducted during the

course of this research.

Additionally, I would like to thank my family for all their love and support.

Special thanks go to my fiance, Dave Friedman, for his constant encouragement and

support throughout my education.
















TABLE OF CONTENTS
Page

A C K N O W L E D G E M E N T S ............................................................................................... iii

LIST OF TABLES ..................................................... vi

LIST OF FIGURES ............................. .. .......... .................................... vii

A B S T R A C T ...................................................................................................... ............ ix

CHAPTER

1 INTRODUCTION ......................... .. .......... ............................ 1

2 LITER A TU RE REV IEW .. .................................................................. .............. 3

P h o to cataly sis ....................................................................................................... 3
T itan iu m D iox id e ................................................................................................. 6
C analyst Supports .............................................................. ....................... .. .... 7
Silica Titania C om posites ................................................................. .............. 8
K in e tic s ..................................................................................................... ............. 1 1
M icro o rg an ism s ........................................................................................................ 12
U V Effect on M icroorganism s..................................... ...................... .............. 16

3 MATERIALS AND METHODS........................................................................ 22

Silica-T itania C om posites ......................................... ......................... .............. 22
R actor D design ......................... .................. .. ........................................ 24
Reactor for Oxidation of Organic Compounds........................................... 24
Reactor for Inactivation of Bacteria and Viruses........................................ 29
Analytical Methods ......................................................... .. ................ 31
A analysis of O rganic Com pounds.................................................. .............. 31
A analysis of B bacteria and V iruses.................................................. .............. 33
Experimental Procedures ................... .............. 34
Oxidation of Organic Compounds.............................................................. 34
Inactivation of Bacteria and Viruses........................................................... 41
S ta tistic s ................................................................................................................. .. 4 6

4 R E S U L T S ................................................................................................................ 4 8

Oxidation of O rganic Com pounds ........................................................ .............. 48









A d so rp tio n ........................................................................................................ 4 8
O x id a tio n .......................................................................................................... 4 9
Inactivation of B bacteria and V iruses..................................................... .............. 58
R ecirculation E xperim ents............................................................ .............. 58
Batch Experiments .................................................... .............. 60

5 SUMMARY AND CONCLUSION ................................................................ 68

S u m m a ry ................................................................................................................... 6 8
C conclusion ....................................................................................................... 69

APPENDIX

A DATA FOR ADSORPTION OF ORGANIC COMPOUNDS............................. 71

B DATA FOR OXIDATION OF ORGANIC COMPOUNDS................ 75

C ANALYSIS OF VARIANCE TABLES FOR ORGANIC COMPOUNDS ............ 79

D TEST FOR EQUAL VARIANCE FOR ORGANIC COMPOUNDS..................... 85

E DATA FOR OXIDATION OF TOC IN SIMULATED NASA WASTEWATER. 88

F DATA FOR INACTIVATION OF MICROORGANISMS..................................... 90

G FIGURES OF DATA FOR INACTIVATION OF MICROORGANISMS .......... 108

LIST O F R EFEREN CE S .... .................................................................. .............. 115

BIOGRAPH ICAL SKETCH ............................................................... .............. 120
















LIST OF TABLES
Table page

1 Comparison of photocatalytic inactivation of microorganisms............................. 20

2 Properties of D egussa P25 TiO 2 ........................................................... .............. 23

3 K ey statistics from the tracer analysis .................................................. .............. 28

4 A average R SD s of SO C analysis............................................................ .............. 33

5 pH values for experiment with 80 mg/L HCO3 solution .................................... 37

6 pH values for experiment with 200 mg/L HCO3 solution .................................. 37

7 pH values for experiment with 3965 mg/L HCO3 solution ................................ 38

8 Sim ulated w astew ater com position....................................................... .............. 40

9 Effluent pH values for solutions containing sodium bicarbonate along with spiked
b acterio p h ag es........................................................................................................... 4 6

10 Effluent pH values for solutions containing sodium bicarbonate along with spiked
b bacteria .................................................................................................... ........... 4 6

11 Removal of toluene and chlorobenzene in attaining adsorption equilibrium in the
re a cto r ...................................................................................................... ........... 4 9

12 pH values for spiked organic contaminant solutions containing differing amounts of
sodium bicarbonate. .. .................... ........ ...... .............. 51

13 Percent rem oval of total organic carbon............................................... .............. 56

14 Effluent pH values for solutions containing sodium bicarbonate along with spiked
b acterio p h ag es........................................................................................................... 6 4

15 Log [No/N] inactivation of bacteriophage under UV radiation ............................. 65

16 Effluent pH values for solutions containing sodium bicarbonate along with spiked
b bacteria .................................................................................................... ........... 6 7

17 Log [No/N] inactivation of bacteria under UV radiation .................................... 67
















LIST OF FIGURES
Figure page

1 Generation of primary radicals at the surface of irradiated TiO2 particles in water... 3

2 L inkages of SiO 2 tetrahedras ..................................... ........................ .............. 10

3 Silica-titania composite suspension being transferred to the 96-well assay plates... 23

4 Organic reactor filled with silica-titania composites ............................ .............. 25

5 Diagram of system setup used in recirculating and single-pass conditions ........... 26

6 E-curve generated from the tracer analysis performed on the reactor................... 29

7 Vertical quartz reactor packed with silica-titania composites ............................. 30

8 Treatment system consisting of reactor packed with SiO2-TiO2 composites, UV
lamp, peristaltic pump, and reservoir of stock solution........................ .............. 31

9 Percent removal of spiked SOCs in a single pass through the reactor with UV
irradiation and without addition of alkalinity ....................................... .............. 50

10 Percent removal of six target analytes in solution containing differing amounts of
bicarbonate operating in single-pass mode after five hours of adsorption in the
recirculation m ode ............. .. ................................ ....... ...... ............ .. 52

11 Percent removal of six target analytes operating in single-pass mode after five hours
of adsorption in the recirculation m ode ................................................ .............. 55

12 Polynomial trend line of desorbing TOC concentration of simulated wastewater ... 57

13 Exponential trend line of desorbing TOC concentration of simulated wastewater .. 57

14 Bacteriophage adsorption and inactivation experiment in a recirculation mode with a
potassium phosphate buffer ....................................... ........................ .............. 59

15 Duplicate sample analysis of adsorption/attachment removal of PRD- 1 for the batch
bacteriophage experim ent with no UV light......................................... .............. 61

16 Duplicate sample analysis of adsorption/attachment removal of MS-2 for the batch
bacteriophage experim ent with no UV light......................................... .............. 62









17 Duplicate sample analysis of inactivation of PRD-1 for the batch bacteriophage
experim ent w ith the 254 nm U V lam p.................................................. .............. 64

18 Percent removal of organic compounds during recirculation in the reactor without
U V lig h t ............................................................................................................... ... 7 4

19 Percent removal of organic compounds during single pass through the reactor with
U V irrad iatio n ........................................................................................................... 7 8
















Abstract of Thesis Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Master of Engineering

PHOTOCATALYTIC OXIDATION OF SELECTED ORGANIC CONTAMINANTS
AND INACTIVATION OF MICROORGANISMS IN A CONTINUOUS FLOW
REACTOR PACKED WITH TITANIA-DOPED SILICA

By

Mary Joanne Garton

August 2005

Chair: Paul Chadik
Major Department: Environmental Engineering Sciences

The NASA Advanced Life Support System (ALS) will provide systems for

recycling water in manned flight space missions. A treatment process train is required in

the ALS to convert wastewater into potable water on these missions. The reactor system

described was designed to fill an ALS need for a final water treatment polishing process

that will oxidize synthetic organic chemicals (SOC) and inactivate pathogenic

microorganisms that are likely to contaminate wastewater systems on these missions. The

reactor does not use expendable oxidation chemicals and uses a catalyst in a fixed bed,

eliminating the need to separate the catalyst from the water after treatment. Two annular

continuous flow reactors with nominal volumes of 440 mL and 150 mL were packed with

silica gel pellets that were doped with titania (TiO2) (12 wt%). The reactors were

configured with ultraviolet (UV) lamps in the center of the reactors. The flow

characteristics of the reactors were determined with tracer analyses. The SOC oxidation









experiments were performed in a single-pass mode with bicarbonate ions present and in a

low dissolved oxygen environment.

Microbial experiments were performed for the inactivation of selected viruses and

bacteria. Cultures of the target microorganisms were prepared in a stock solution, which

was pumped through the reactor for a period of time to allow for microbial attachment to

the reactor and catalyst pellets. The UV lamps were then activated and microbial

concentrations were monitored. The log [No/N] values resulting from two hours of 254-

nm UV irradiation for the bacteriophages OX-174, PRD-1, and MS-2 were 1.67, 1.43,

and 1.65, respectively. The log [No/N] values resulting from two hours of 254-nm UV

irradiation for the bacteria Escherichia coli, Staphylococcus aureus, and Pseudomonas

aeruginosa were 1.70, 1.40, and 1.62, respectively.

The photocatalytic reactor system effectively oxidized the six target organic

contaminants: acetone, carbon disulfide, chlorobenzene, ethyl acetate, methyl

methacrylate, and toluene. The percent removal of each contaminant was largely

unaffected by the addition of bicarbonate ions in the range of 50 120 mg/L as CaCO3

and in a low dissolved oxygen environment between 1 3 mg/L.
















CHAPTER 1
INTRODUCTION

Water recovery is an essential process needed for manned flight space missions.

The ability to regenerate potable water reduces waste storage, promotes the health and

safety of crew members, and minimizes mass of required water resources. A reduction in

costs would also be seen from reduced amounts of mass, volume, and energy required for

water transportation.

The goal of this research is to further develop a finishing process for the treatment

of wastewater for NASA's Advanced Life Support System (ALS). The components of

the wastewater will include shower water, wash water, urine, humidity condensate from

machinery, and wastewater from the solid waste processor. A treatment process train is

required in the ALS to convert wastewater into potable water on the space missions. The

post-processor would provide a final water treatment polishing process that will oxidize

synthetic organic chemicals (SOCs) and inactivate pathogenic microorganisms that are

likely to contaminate wastewater systems on these missions and that may not have been

completely removed in previous treatment processes. These SOCs, which will be

generated on space missions and which will accumulate in recycled water if not removed,

have adverse health consequences if ingested and must be controlled to part per billion

levels as set by NASA's Requirements: Definition and design consideration (Lange and

Lin, 1998). The goal of the reactor system is to inactivate the microorganisms and

completely destroy the SOCs rather than merely change the phase of these contaminants.









The reactor system will not require expendable oxidation chemicals, and will have a low

energy and power demand.

Photocatalysis or photocatalytic oxidation was the mechanism investigated for the

finishing process. Research has shown that photocatalysis is an effective process to

destroy organic compounds, inorganic compounds, and microorganisms. The research

focused on six selected organic compounds as well as selected bacteria and viruses. The

organic compounds were:

acetone
carbon disulfide
chlorobenzene
ethyl acetate
methyl methacrylate
toluene

The target microorganisms were the bacteria Escherichia coli, Staphylococcus aureus,

and Pseudomonas aeruginosa, and the bacteriophages (DX-174, PRD-1, and MS-2.

The research was conducted using two previously designed flow-through annular

reactors (Holmes, 2003, and Ludwig, 2004) and silica-titania composite pellets

(Londeree, 2002) as the photocatalyst in the reactor. By the exposure to ultraviolet (UV)

light, the photocatalyst pellets catalyze the oxidation of the organic compounds and

inactivate the bacteria and bacteriophages. The objectives for the current research were to

investigate the effects of bicarbonate alkalinity on the oxidation of organic compounds,

inactivate selected microorganisms, and investigate the effects of a low dissolved oxygen

concentration environment on the oxidation of organic compounds.

















CHAPTER 2
LITERATURE REVIEW

Photocatalysis

Photocatalysis is an oxidative process in which organic contaminants are

degraded using ultraviolet (UV) light energy. Heterogeneous photocatalysis involves the

use of a solid photocatalyst in contact with either a gas or liquid, while homogeneous

photocatalysis uses a catalyst of the same phase as the contaminated media. In this work,

the media involved is a solid phase photocatalyst, titanium dioxide (TiO2) and an aqueous

solution containing the organic and microbiological contaminants. While UV light or an

oxidant used alone may produce partial degradation of a compound, the combined use of

UV light with an oxidant (03, H202, or 02 with a photocatalyst) has been shown to yield

complete mineralization of organic carbon contaminants to carbon dioxide (Ollis et al.,

1991). The removal also occurs at a faster rate than with the UV or oxidant alone. Figure

1 illustrates the radicals formed on the photocatalyst surface.

HO2, H2O + OH.
TiO2 particle o, + H* o 02 +
,CB \

11%O, hv

H'O

OH. + H+
Figure 1. Generation of primary radicals at the surface of irradiated TiO2 particles in
water (Blake et al., 1991)









Radicals are formed when the photocatalyst absorbs enough light of the

appropriate wavelength for an electron (e-) to jump from the valence band (VB) to the

conduction band (CB) forming a positive electron hole (h ). For titanium dioxide (TiO2),

UV light of less than 388 nm is required. The following equations describe the excitation

of an electron using TiO2 and the subsequent reactions:

TiO2 + hv e- + h+ (1)

e- + 02 02- (2)

h+ + M Mox (3)

h+ + OH -- OH* (4)

where 02- is the superoxide anion, OH* is the hydroxyl radical, and M and Mox are the

organic compound and its oxidized product, respectively (Tanaka et al., 1991).

Both the electron and electron hole are highly energetic and therefore highly

reactive. Most organic photodegradation reactions utilize the oxidizing power of the

electron holes; however, to prevent a buildup of charge, a reducible species must be

present to react with the electrons (Hoffmann et al., 1995). Since oxygen (02) is usually

available in aqueous solutions, it has been shown to be a very important electron acceptor

in the process. Oxygen keeps the electron and electron hole from recombining and thus

leaving open the reaction sites for photocatalysis to take place. Gerischer and Heller

(1991) have demonstrated that with an increase in oxygen concentration, there is a

subsequent increase in the reaction rate of contaminant degradation.

Water and hydroxide ions (OH-) react with the electron holes to form hydroxyl

radicals (OH ). The photocatalytic mineralization of an organic compound that is

adsorbed to the catalyst surface begins with either reaction, with hydroxyl radicals or by









direct oxidation with the holes on the surface of the catalyst. As illustrated in equations 5

- 6, the hydroxyl radical begins to degrade the adsorbed organic compound by either

removing an available hydrogen atom to form water (equation 5) or adding itself to any

unsaturated carbon bonds (equation 6) (Grabner et al., 1991; Hoffman et al., 1995; Mao

et al., 1991, 1992).

OH* + R H20 + R(-H)* (5)

OH* + R ROH* (6)

where R is the organic compound, OH* is the hydroxyl radical, R(-H)* is a radical with a

hydrogen removed, and ROH* is a radical combined with the hydroxyl radical (Mao et

al., 1991).

Turchi and Ollis (1990) found that since the intermediates detected during the

photocatalytic degradation of halogenated aromatic compounds are typically

hydroxylated structures, the primary oxidant was assumed to be the hydroxyl radical. The

oxidative power of the hydroxyl radical is more than twice the power of chlorine

(Goswami and Blake, 1996). After these primary reactions with the hydroxyl radicals,

many intermediate organic radicals and products are formed. From these radical-to-

radical interactions, eventually complete mineralization occurs resulting in CO2 and H20.

Mao et al. (1991 and 1992) have shown that the conditions of a system (pH, oxygen

concentration, competition with other compounds) can affect the pathway a single

compound can take towards complete mineralization. Those system parameters were

investigated in this research. Mao et al. (1991 and 1992) also concluded that the

photocatalytic degradation of halocarbons with TiO2 was initiated via an oxidative

process involving presumably surface-adsorbed, hydroxyl radicals.









Adsorption of compounds to the surface of the catalyst is important in an efficient

photocatalytic system. Competition with other compounds can be detrimental to the

adsorption of specific contaminants on the catalyst surface. Chen et al. (1997) showed the

effect of competitive adsorption on the catalyst surface and its effects on the degradation

rates. The competition with dichloroethane (DCE) for adsorption and the affinity of ions

for TiO2 were in order: chloride < nitrate < (bi)carbonate < sulfate < phosphate. A similar

order was found for the inhibition of photodegradation: nitrate < chloride < (bi)carbonate

< sulfate < phosphate. Carbonate ions can react with the hydroxyl radicals, producing the

less reactive ion radical (CO3*). The researchers postulated that the effect may have

occurred but appears to be of minor importance in comparison to competitive adsorption.

The effect of bicarbonate ions in solution was investigated in this research.

Titanium Dioxide

TiO2 is the photocatalyst most often used because of its low cost, non-toxic

characteristic, insolubility, heat resistance, and photostability compared to other metal

oxides such as zinc oxide (ZnO). ZnO can undergo photocorrosion, decreasing the

effective lifetime of the catalyst (Okamoto et al., 1985). Titanium dioxide has also been

shown to be the most active semiconductor for photocatalysis (Goswami and Blake,

1996).

TiO2 can be synthesized into two different crystalline forms, anatase and rutile.

The anatase crystal form has been found to be the more photocatalytically active of the

two (Ohtani and Nishimoto, 1993). The difference in reactivity can be attributed to the

more positive conduction band of the rutile phase hindering molecular oxygen reduction

(Tanaka et al., 1991). The degradation rate of water-soluble compounds is independent of









the crystal form, while sparingly water-soluble compounds are dependent on the anatase

content of the TiO2 (Tanaka et al., 1991). Degussa P-25 is a commercially available TiO2

with a structure that is 70% anatase and 30% rutile. This particular TiO2 is commonly

used as it has repeatedly demonstrated successful photocatalytic degradation of organic.

Catalyst Supports

TiO2 has been an effective photocatalyst used in slurry form to maximize

exposure to the solution and UV light, but it can be difficult to recover the titania in the

post processing of the solution. Therefore, the TiO2 has been chosen to be incorporated

with a catalyst support to immobilize the titania and ensure removal and recovery.

Several supports have been investigated, including glass beads, sand, clay, activated

carbon, and silica gel (Al-Ekabi and Serpone, 1988; Yoneyama and Torimoto, 2000).

The rate of destruction of organic contaminants is generally controlled by the

concentration of the contaminant; therefore, the rate of mineralization will decrease as the

contaminant concentration decreases. Using a catalyst support that is capable of

adsorbing organic contaminants as well as immobilizing the photocatalyst will enhance

the photodegradation of the contaminants because of the higher concentration of

contaminant present around the loaded TiO2 (Yoneyama and Torimoto, 2000). When the

organic substances are oxidized on the photocatalyst surface, the resulting intermediates

can also be adsorbed to the support and further oxidized. The possibility of creating toxic

intermediates during photocatalysis is a concern, but, if the intermediates are held near

the catalyst, they are more likely to be completely destroyed.

Silica gel (Si02) was chosen for the catalyst support for this study based on the

adsorption capabilities as well as the transparent nature of the material. The silica gel









allows the infiltration of photons to the titania surface. Silica also has high mechanical

strength, thermal stability, and can be synthetically formed into any shape, such as

cylindrical pellets (Londeree, 2002).

Silica Titania Composites

TiO2 and SiO2 can be combined, enabling the formation of highly efficient

photocatalysts. These TiO2 SiO2 photocatalysts allow the placement of the catalyst on

both external surfaces and internal surfaces within the porous silica matrix where

pollutants are adsorbed.

SiO2-TiO2 composites have been successful in the photocatalytic degradation of a

variety of organic compounds (Matthews, 1988; Anderson and Bard, 1995; Anderson and

Bard, 1997; Xu et al., 1999; Vohra and Tanaka, 2003). Xu et al. (1999) synthesized

particles from a titania sol and silica powder and tested for photocatalytic destruction of

acetophenone. Compared to the bare TiO2 prepared in parallel, all the supported TiO2

showed a higher photoactivity. The researcher inferred that this enhanced photoactivity

was related to the increased adsorption of organic substrates, the increased surface area of

the SiO2-TiO2 composite compared to bare TiO2, and perhaps to the absence of rutile

phase in the supported samples. The latter two are closely related to the fact that the TiO2

was highly dispersed over the porous silica. Anderson and Bard (1995) made a similar

comparison of bare TiO2 to a TiO2/SiO2 composite material fabricated through sol-gel

processing by testing for the removal of rhodamine-6G. Again, SiO2-TiO2 was found to

have a faster degradation rate, which was related to the increased adsorption of the

contaminant. These studies demonstrate the idea that increasing adsorption will increase









the concentration of the contaminant around the TiO2 and lead to faster destruction of the

contaminant.

When photo-decomposing phenol (a compound with low adsorption capacity to

silica), Anderson and Bard (1997) found the TiO2 alone performed better, which was to

be expected since the silica did not help with adsorption capabilities and only hindered

the effect of the photocatalyst. However, when the rates were normalized to the TiO2

content of the material, the TiO2/SiO2 material demonstrated a more efficient use of the

TiO2 than the slurry alone. The compounds being researched, such as toluene and

chlorobenzene, have higher adsorption capacities than phenol. In an adsorption batch

experiment performed by Hashizume et al. (2004), the amount of adsorption to a

synthesized mesoporous silicate material was in the order of toluene = benzene > benzoic

acid > phenol. Demeestere (2002) compared the adsorption on TiO2 of chlorobenzene

and toluene in the gaseous phase. Chlorobenzene had a greater adsorption capacity than

toluene.

Silica-titania composites can be efficiently utilized in heterogeneous

photocatalysis systems for the adsorption and subsequent destruction of contaminants in

an aqueous solution. In addition, a dosage of 30 wt% (equal to 12% on a weight per

volume of silica precursor) was also found to be an optimum loading for Londeree

(2002), Holmes (2003), and Ludwig (2004) for the silica-titanium composites created

using a sol-gel doping method.

Sol gel method. There has been much research on the different preparation

methods for creating a solid SiO2-TiO2 material with photocatalytic ability. The

characteristics (e.g., pore size, surface charge, mechanical strength, and adsorption sites)









of the final product are dependent on the synthesis conditions and the type of interaction

between TiO2 and SiO2.

Londeree's (2002) method uses sol-gel hydrolysis to create the silica matrix, then,

during gelation, the solution is doped with the commercially available, highly efficient

Degussa P25 TiO2. The silica network may either form around the titania particle and/or

form a bond with the titania to secure its position in the matrix. The final product

resulting from the sol-gel method steps hydrolysiss, condensation, gelation, aging, drying,

curing) is a gel containing a relatively monodisperse pore size and displaying specific

characteristics associated with the conditions it experienced during processing. The

details of the method are discussed in Chapter 3, Materials and Methods.

Silica gel properties. The silica network consists of four oxygen atoms bonded to

each silicon atom, forming a tetrahedron, and each oxygen atom is shared by two silicon

atoms (Figure 2).


HO OH
HO-Si-OH HO-Si-OH
HO 0 OH
HO-Si-O-Si---0-- Si-O-Si-OH


HO-4i-OH HO-i-OH



Figure 2. Linkages of SiO2 tetrahedras (Hench and West, 1990)

These Si-O-Si, or siloxane, bonds make up the bulk silica structure. The surface

of the silica is hydroxylated in the presence of water, forming Si-OH, or silanol, groups.

It is these silanol groups that make the silica surface hydrophilic and determine the

reactivity of the silica (Nawrocki, 1997)









The hydroxyl concentrations, silanol groups, on the Si02-TiO2 surface contribute

to the ability of the catalyst to adsorb pollutants. The silanol groups can be defined as

single, geminal, or vicinal. Single silanol groups are isolated and formed with a silicon

atom that is bonded to three other oxygen atoms in siloxane bonds. Vicinal silanols have

two hydroxyl groups joined to the gel at two different silicon sites, but a hydrogen group

from one is also bonded to the oxygen group from the other. Geminal silanols exist when

two hydroxyl groups are linked to the same silicon atom. The various silanols can have

different adsorption activity, and Nawrocki (1997) indicated that the isolated silanols are

the more reactive species over the hydrogen-bonded silanols since the hydrogen of the

isolated silanol is not bonded to another atom and is free to react with other species.

Kinetics

Langmuir-Hinshelwood (LH) kinetics has been shown to successfully describe

the photocatalytic degradation of organic contaminants (Al-Ekabi and Serpone, 1988;

Turchi and Ollis, 1989). Since LH kinetics is known to be a good model for solid-gas

reactions, some modifications are necessary to model solid-liquid reactions. The LH

model assumes the reactions take place at the surface of the catalyst, and the reaction rate

is proportional to the fraction of the surface covered by the reactant.

r = -dC/dt = krO = krKC/(1+KC) (7)

where r is the rate of the reaction, C is the concentration of the contaminant in solution, t

is time, kr is the reaction rate constant, 0 is the fraction of the surface covered by the

reactant, and K is the adsorption coefficient of the reactant (Al-Ekabi and Serpone, 1988).









Equation 7 can be simplified when the concentration is either very high (KC >>

1) or very low (KC << 1). Under conditions of low concentration, KC becomes negligible

compared to 1 and equation 7 reduces to a first-order reaction:

-dC/dt = krKC (8)

Since the concentrations used in this research were approximately 200-300 ug/L, the

first-order reaction would likely be applicable; however, this also relies on low K values.

Similarly, under conditions of high concentration, 1 becomes negligible compared

to KC. Equation 7 reduces to a zero-order reaction, where the reaction rate is equal to the

rate constant, kr. This simplification is illustrated in the photocatalytic degradation of 4-

chlorophenol, where no additional increase in rate was observed with an increase in

initial concentration above 0.2 millimoles (Al-Ekabi and Serpone, 1988). It was reasoned

that at a high enough contaminant concentration, the surface sites of the catalyst became

completely saturated, so further increase in concentration cannot further increase the rate

of reaction, according to LH kinetics. Conversely, below the specified concentration, the

catalyst surface sites are not completely saturated; therefore, the fraction of the surface

covered by the contaminant will vary with concentration, thereby varying the reaction

rate. It is important to note that the use of the LH equation to describe photocatalytic

kinetics makes an assumption that adsorption of the contaminant to the TiO2 surface is

the rate-limiting step.

Microorganisms

The goal of NASA's ALS is to remove or inactivate the pathogenic

microorganisms that could be present in the recirculatory water system. The

specifications of potable water include a standard of less than 100 colony forming units









(CFU) per 100 milliliters (mL) for total bacteria. The coliform group and virus

specification is non-detectable or below detectable limit (Lange and Lin, 1998).

Concentrations of pathogens are generally low in water and wastewater, and no

single technique is available to isolate and identify all pathogens. Instead of direct

isolation and enumeration of pathogens, indicator organisms are used to expedite the

analytic procedure. The coliform group of bacteria has been used to indicate the presence

of pathogens in the water. Their presence in water indicates recent fecal contamination,

which is the major source of many enteropathogenic diseases transmitted through water.

Coliforms and pathogens both originate from the same source: fecal matter from humans

and animals.

Bacteria. For the purpose of this research, three different bacteria, Escherichia

coli, Staphylococcus aureus, and Pseudomonas aeruginosa were chosen as indicator

microorganisms. Escherichia coli is one of the enteric bacteria, or Enterobacteriaceae,

which are facultative anaerobic gram-negative rods that live in the intestinal tracts of

animals in both health and disease. The enterics ferment glucose, producing acid and gas,

are typically oxidase-negative, and, when motile, have flagella uniformly distributed over

the body. A number of genera within the Enterobacteriaceae family are major

waterborne pathogens, such as Salmonella, .\/nge//l, and Yersinia. Several others are

normal colonists of the human gastrointestinal tract (e.g., Escherichia, Enterobacter,

Klebsiella), but these bacteria also may occasionally be associated with diseases of

humans (Murray et al., 1994).

Physiologically, E. coli is versatile and well-adapted to its characteristic habitats.

It can grow in media with glucose as the sole organic constituent. The bacterium can









grow in the presence or absence of 02. Under anaerobic conditions it will grow by means

of fermentation, producing acids and gases as end products. However, it can also grow by

means of anaerobic respiration, since it is able to utilize NO3-, NO2-or fumarate as final

electron acceptors for respiratory electron transport processes. In part, this adapts E. coli

to its intestinal (anaerobic) and its extraintestinal (aerobic or anaerobic) habitats. E. coli

can respond to environmental signals, such as chemicals, pH, temperature, osmolarity,

etc. For example, it can sense the presence or absence of chemicals and gases in its

environment and move towards or away from them. In response to change in temperature

and osmolarity, it can vary the pore diameter of its outer membrane to accommodate

larger molecules (nutrients) or to exclude inhibitory substances (Todar, 2005).

Staphylococcus aureus is a gram-positive spherical bacteria, about 1 micrometer

in diameter that occurs in microscopic clusters resembling grapes. S. aureus are

nonmotile, non-spore-forming facultative anaerobes that grow by aerobic respiration or

by fermentation that yields mainly lactic acid. The bacteria are catalase-positive and

oxidase-negative. S. aureus can grow at a temperature range of 15 to 45 degrees and in

aqueous NaCl concentrations as high as 15 percent. S. aureus ferments mannitol, which

distinguishes it from S. epidermidis (Todar, 2005).

Pseudomonas aeruginosa is a gram-negative rod measuring 0.5 to 0.8 [m by 1.5

to 3.0 rm. Almost all strains are motile by means of a single polar flagellum. Its

metabolism is respiratory and never fermentative, but it will grow in the absence of 02 if

NO3- is available as a respiratory electron acceptor. P. aeruginosa has very simple

nutritional requirements. In the laboratory, the simplest medium for growth of P.

aeruginosa consists of acetate for carbon and ammonium sulfate for nitrogen. It is









tolerant to a wide variety of physical conditions, including temperature and high salt

concentrations. Its optimum temperature for growth is 37 degrees, and it is able to grow

at temperatures as high as 42 degrees (Todar, 2005).

Bacteriophages. Bacteriophages, or phages, are parasitic viruses that infect

bacteria instead of plants or animals. Phages are often used in research studies as models

of human enteric viruses instead of plant or animal viruses because one can easily prepare

large populations (up to 1010 PFU/mL) of genetically homogeneous, susceptible host

cells, under well-defined nutritional conditions, as compared to plant and animal hosts

(Matthews, 1971). Many features of phage infection are common to the infective process

of plant and animal viruses.

The three phages chosen for this research were OX-174, PRD-1, and MS-2. These

three organisms include both RNA and DNA phages and have different sizes. The

bacteriophage OX-174 is the principal representative of a group of phages that are simple

icosahedrons (20 faces) with 5-nm spikes extending from all 12 vertices. It is a 27-nm

single-stranded DNA bacteriophage with a circular topology. The bacteriophage OX-174

belongs to the family Microviridae in the genera of Microvirus. Related phages include

S13, OR, and G4 (Birge, 2000). The phage MS-2 is a single-stranded RNA bacteriophage

with icosahedral morphology and a diameter of 26.0 to 26.6 nm (Sjogren and Sierka,

1994). MS-2 belongs to the Leviviridae family in the Levivirus genera. MS-2 has a linear

topology and a unique sequence. Related phages include fr, f2, R17, M12, and QP (Birge,

2000). Both OX-174 and MS-2 are coliphages, or bacteriophages that are infective to

coliform bacteria. The usual hosts for the phages are E. coli strain C for OX-174 and E.

coli F+ for MS-2. The bacteriophage PRD-1 belongs to the Tectiviridae family in the









Tectivirus genera. PRD-1 is a double-stranded DNA phage with icosahedral morphology

and linear topology (Birge, 2000).

UV Effect on Microorganisms

The UV light effective in inactivating bacteria and virus resides in the UV-B and

UV-C ranges of the spectrum (200 to 310 nm) with a maximum effectiveness for most

bacteria and virus species occurring around 265 nm (Malley, 2002). Inactivation of

microorganisms by UV irradiation occurs through the formation of lesions in DNA,

which prevent normal DNA replication, leading to inactivation (Harm, 1980). Thymine

bases on the nucleic acids are particularly reactive to UV light and form dimers

(thymine=thymine double bonds). These dimers then chemically inhibit transcription and

replication of nucleic acids thus rendering the organism sterile (Malley, 2002).

Photoreactivation is a phenomenon in which UV-inactivated microorganisms

recover activity through the repair of lesions in the DNA by enzymes under near-UV

light. In dark repair, UV-inactivated microorganisms repair the damaged DNA in the

absence of light. (Morita et al., 2002). Many factors (UV dose, water quality, length of

exposure to photoreactivating light, species of organism) can affect photoreactivation.

Shaban et al. (1997) found that the enzymatic process of dark repair did not affect the

recovery of irradiated organisms (E. coli, S. aureus, and coliphage). A 2-log

photoreactivation of UV-inactivated E. coli did occur with subsequent exposure (four to

six hours) to light in the visible spectrum. The strategy in UV disinfection has been to

provide a high enough dosage that enough nucleic acid damage occurs to prevent

effective repair by photoreactivation or dark repair.









Photocatalytic inactivation. The hydroxyl radical has been theorized to act as a

potent biocide because of its high oxidation potential and nonselective reactivity. Several

researchers have experimented with the effectiveness of the hydroxyl radical as a

potential biocide to bacteria, viruses, and protozoan. Ireland et al. (1993) used a 300- 400

nm wavelength lamp covered with a photocatalytic sleeve formed of fiberglass mesh,

which was coated with a firmly bonded layer of TiO2. The researchers dechlorinated the

tap water used in the experiment before the introduction of Escherichia coli into the

system. Using a 10-fold stoichiometric excess of sodium thiosulfate, the inactivation of

E. coli was a 0.4 log reduction in concentration in the 6-minute exposure time (flow rate

of 2 L/min). However, after dechlorinating the water using only UV light exposure, the

observed reduction in the concentration of E. coli was 7 orders of magnitude after 6

minutes of cumulative exposure. In the 9-minute cumulative exposure sample, the E. coli

counts were below detection limit, indicative of a total reduction of 9 10 logio units.

Ireland et al. (1993) concluded that TiO2 photocatalysis may be a viable process

for disinfection of bacteria in water treatment systems; though, inorganic-radical

scavengers can have a major negative impact on the efficacy of the process. The presence

of organic matter also degrades the inactivation kinetics, ostensibly by competing with

bacteria for the hydroxyl-radical oxidant.

Sjogren and Sierka (1994) achieved a one-log reduction of the phage MS-2 after

10 minutes of irradiation in a continuously stirred batch reactor containing a TiO2 slurry.

The 365-nm UV lamp emitted an irradiance 2 mW/cm2. The solution was buffered with

22 mM of phosphate (pH = 7.2) and also contained 5 mg/L (0.14 mM) of chloride from a

saline buffer used in the inoculation of MS-2. A 3-log reduction was achieved with the









TiO2 slurry by adding 0.1 mg/L of FeSO4 under the same conditions. This increased level

of inactivation was attributed to supplemental hydroxyl radical oxidations that were

enabled by Fenton reactions, such as:

Fe (II) + H202 "- IC "- Fe (III) + OH- + OH* (9)

where IC is an intermediate complex and Fe (III) and Fe (II) are Fe3+ and Fe2+ and

complexes. The H202 was thought to have been produced by the e- reduction of oxygen.

Cho et al. (2005) found that irradiation for 120 minutes was required for the 0.95-

and 2.25 log inactivation of MS-2 phage and Escherichia coli, respectively. The batch

study included 1 g/L of TiO2, a 20 mM phosphate buffer, and 18W black-light blue lamps

(300 420 nm) as the source of irradiation. The light intensity was measured at 7.9*10-6

einsteins/liter/s (approximately 2.25 3.15 mW/cm2 in the UV wavelength range of 300

- 420 nm). By performing studies in the presence of scavengers, tert-butanol and

methanol, Cho et al. (2005) concluded that the MS-2 phage is inactivated mainly by the

free hydroxyl radical in the solution bulk but Escherichia coli is inactivated by both the

free and the surface-bound hydroxyl radicals. Escherichia coli might also be inactivated

by other ROS (reactive oxygen species), such as 02- and H202.

Wei et al. (1994) found the bacterial inactivation of E. coli (106 cells/mL) adhered

to first-order kinetics. The rate constant (5.5* 10-2 min-1) was proportional to the incident

light intensity in the range of 180-1660 atE/s/m2 and to the TiO2 dose. The researcher

achieved a 6-log reduction in E. coli. by irradiating a 1 g/L TiO2 slurry with 52.3

mW/cm2 of UV intensity for 30 minutes using a 380-nm lamp.

Bekbolet (1997) also found that the rate constants decreased with increasing E.

coli concentration. After one hour of illumination with a 320-420 nm UV lamp, the









inactivation rate calculated was 1.78* 102 min-1. The light intensity of the UV lamp was

67.9 tE/s/m2 (approximately 1.94 2.54 mW/cm2 in the UV wavelength range of 320 -

420 nm), and the TiO2 aqueous suspension dose was 1 mg/mL.

Lee et al. (1997) achieved a 2.2-log inactivation of the phage QP using

immobilized TiO2 and 1-hour irradiation with near UV black light (300-400 nm) at an

intensity of 3.6 mW/cm2. A 0.5-log QP inactivation was observed with the black light

alone. A control experiment was performed using a 0.1 g/L slurry of TiO2. Under the

same irradiation conditions, a 2.4-log inactivation was achieved in only 20 minutes as

opposed to 60 minutes with the immobilized TiO2. No noticeable difference in

inactivation was observed between a 254-nm germicidal lamp irradiation (light intensity

of 0.6 mW/cm2) with and without immobilized TiO2. The rate constant of inactivation

was found to be proportional to the light intensity in the range of 3 8 mW/cm2. At 5

mW/cm2, the rate constant was 0.14 min-' for a 5000X dilution of the QP phage in broth

with autoclaved pure water and 0.02 min-' for dilution of 100X. The lower rate constant

for the 100X broth dilution was attributed to radical scavenging and adsorption on

reaction sites of the broth components.

Shown in Table 1 are the inactivation results for each of these researchers along

with the given parameters for their experiments including UV intensity and dose (when

available). Most of the experiments were performed in a batch reaction with a TiO2

slurry.















Table 1. Comparison of photocatalytic inactivation of microorganisms for various researchers


Loglo TiO2 TiO2 Background
Researcher Organism inactivation UV intensity Time Wavelength UV dose onfiation dose species Flow rates

mW/cm2 min Nm J/cm2 g/L L/min
reland et al. E. coli 7 Unavailable 6 300 400 Unavailable immobilized N/A None 2
(1993)
E. coli 0.4 Unavailable 6 300 400 Unavailable immobilized N/A Sodium thiosulfate 2
Wei et al.
Wei et al. E. coli 6 52.3 30 380 94.14 slurry 1 None batch
(1994)
Bekbolet E. coli 3 1.94 2.54 60 320 -420 6.98 -9.14 slurry 1 None batch
(1997)
Cho etal. E. coli 2.25 2.25- 3.15 120 300- 420 16.20- 22.68 slurry 1 20mM HP042- batch
(2005)
MS-2 0.95 2.25- 3.15 120 300 -420 16.20 22.68 slurry 1 20mM HPO42- batch
Sjogren and 22 mM HPO42-
Sk (9 MS-2 1 2 10 365 1.20 slurry 1 22 mM P04 batch
Sierka (1994) 0.14 mM Cl
Lee et al
(1997) Qp phage 2.2 3.6 60 300- 400 12.96 immobilized N/A broth batch
Qp phage 2.4 3.6 20 300 400 4.32 slurry 0.1 broth batch









For the E. coli studies, the largest reduction (7 orders of magnitude) was seen in

the flow-through reactor using immobilized TiO2. The UV intensity was not reported by

Ireland et al. (1993), so comparisons with other systems are difficult. The major

difference between experiments in Table 1 is the UV dose applied to the batch reactor.

The second greatest reduction, by Wei et al. (1994), of 6 logs can be attributed to the high

UV intensity and therefore UV dose, which is over an order of magnitude above

Bekbolet's UV dose.

For the phage studies, including MS-2 and QP, there was variability in the results.

Sjogren and Sierka (1994) achieved a one-log reduction in MS-2 with a much lower UV

dose than Cho et al. (2005). Both researchers had the same TiO2 dose and approximately

the same background phosphate concentration. In the slurry experiment performed by

Lee et al. (1997), the inactivation was twice as much as the other two researchers, but the

phage was a different type, the TiO2 dose was less, and the UV dose was in between that

of Sjorgren and Sierka (1994) and Cho et al. (2005). Also, the background contents in the

solution were not specified by Lee et al. (1997).

Mechanism for inactivation. Maness et al. (1999) found the lipid peroxidation

reaction was the underlying mechanism of death for E. coli K-12 cells that were

irradiated in the presence of the TiO2 photocatalyst. The occurrence of lipid peroxidation

and the simultaneous losses of both membrane-dependent respiratory activity and cell

viability depended strictly on the presence of both light and TiO2.















CHAPTER 3
MATERIALS AND METHODS

Silica-Titania Composites

The silica-titania composites were created using a sol-gel method (Powers, 1998).

The silica precursor was tetra-ethyl-ortho-silicate (TEOS) (Fisher Scientific, reagent

grade). It was mixed with nanopure water using a water-to-TEOS mole ratio of 16:1.

Ethanol (Aaper Alcohol, 200 proof) was used as the solvent to facilitate the miscibility

between the TEOS and water. Two acid catalysts were used: a 1 M nitric acid solution,

made from 15.8 M nitric acid (Fisher Scientific, certified ACS.) and nanopure water, and

a 3% solution of hydrofluoric acid, formulated from 49% hydrofluoric acid (Fisher

Scientific, reagent ACS.) and nanopure water. The basic formula used to create 10 g of

pellets with a pore size of roughly 140A is as follows: 25 mL nanopure water, 50 mL

ethanol, 35 mL TEOS, 4 mL nitric acid (IM), and 4 mL HF (3%). By changing the

amount of HF added, the pore size and structure within the gel can be altered.

The chemicals were added to a polystyrene container. A magnetic stir plate

provided sufficient mixing while a known mass of Degussa P25 TiO2 was added to the

batch. The TiO2 loadings are reported on a basis of TiO2 weight per TEOS volume. For

example, the 12% TiO2 loading is determined using 4.2 g of TiO2 per 35 mL of TEOS as

a percentage. The properties of Degussa P25 as provided by Degussa are listed in Table 2

(Ludwig, 2004).









Table 2. Properties of Degussa P25 TiO2
Specific surface area (BET) 50 +/- 15 m2/g
Average primary particle size 30 nm
Tapped density 130 g/L
pH value in 4% dispersion 3.5 -4.5


The solution was allowed to mix for 30 minutes and then was transferred by a

pipette into Fisherbrand (Fisher Scientific), polystyrene 96-well assay plates before

gelation to form the titania-doped pellets (Figure 3). The assay plates contained 0.45 mL

in each well. Each batch of chemicals created approximately 4 assay plates of pellets.

















Figure 3. Silica-titania composite suspension being transferred to the 96-well assay plates
used for creating the pellet shape

After 2 hours of gelation, the plates were covered with lids and wrapped in

aluminum foil to limit volatilization during the aging process. The gels were aged at

room temperature for two days, then at 650 C in an Oakton Stable Temperature oven for

two days. After aging, the pellets were removed from the assay plates and placed into

Teflon containers for the next series of heat treatments. A small hole was placed in the lid

of the containers to allow the liquid from the gel's pores to escape as a vapor, thus

providing uniform drying of the gel. Using a Yamato DVS 400 Drying Oven, the









temperature was ramped from room temperature to 1030 C (20/min) and kept constant for

18 hours, resulting in the vaporization of liquid solution within the silica network. Next,

the temperature was ramped to 1800 C (20/min) to remove any physically adsorbed water.

It was kept constant for 6 hours and then was slowly decreased back to room temperature

over a 90-minute period. The resultant size of an individual cylindrical pellet after drying

was approximately 5 mm in length with a diameter of 3 mm. The BET (Brunauer,

Emmett, and Teller equation) surface areas and pore volumes of the gels were analyzed

using a Quantachrome NOVA 1200 Gas Sorption Analyzer.

Reactor Design

Reactor for Oxidation of Organic Compounds

A reactor was designed by Holmes (2003) and constructed by Analytic Research

Systems1 in order to contain the silica-titania composite pellets and provide effective

oxidation of organic compounds and inactivation of pathogenic microorganisms. Holmes'

designed reactor was used in the subsequent experiments involving the oxidation of

selected organic compounds.

The reactor was constructed of quartz for its ability to transmit UV radiation

energy. The reactor's cylindrical shape (Figure 4) was chosen to optimize the exposure of

the pellets to the light. The inside diameter of the reactor was 8.5 cm, and the outside

diameter was 10.5 cm. These dimensions were chosen so that all pellets would be able to

react with the UV radiation. The radiation incident on the pellets would be reduced by a

thicker reactor, thereby not effectively using all of the catalyst. The reactor is 14 cm long,

not including the influent and effluent ends. Each of the influent and effluent cylindrical

ends has a diameter of 1 cm.

1 Analytical Research Systems, Inc. PO Box 140218 Gainesville, FL 32614-0218























Figure 4. Organic reactor filled with silica-titania composites on its support stand and
connected to the system used for the testing

The entrance and exit of the reactor were each threaded to allow for connections

with the PTFE polytetrafluoroethylenee) tubing used in the system. The entire reactor was

made of quartz with the exception of one end where a glass frit was placed to keep the

pellets from flowing out of the reactor with the solution. The reactor had a volume of 436

mL with the pellets taking up 109 mL and interparticle space accounting for the

remaining 327 mL. Holmes (2003) estimated the interparticle bed porosity within the

reactor (75%) by filling a graduated cylinder with known volumes of pellets and

nanopure water.

Reactor system. A system was designed for testing the reactor's capabilities for

degrading the target analytes. The system included the reactor and reactor support, two

sampling ports, a source tank, a pump, a flow dampener, stirrer, a waste beaker, and a

sparge tank for the low dissolved oxygen experiments (Figure 5).





































Figure 5. Diagram of system setup used in recirculating and single-pass conditions

The reactor was placed on a wooden support in the horizontal position to limit the

influence of gravity on the flow. Since NASA specifications require the reactor to work

in a micro-gravity situation, it was necessary to prevent gravity from enhancing the

results of the experiments. Connections were available for four UV lamps to be used in

the center of the reactor. The lamps were 12-inch, 8-watt lamps that each provided

approximately 4.44 W of available UV energy (365 nm wavelength) to the inner surface

of the reactor. The support also included a cover that could be placed over the reactor and

UV lamps to prevent exposure. All of the tubing used was PTFE tubing to prevent

adsorption or desorption of organic compounds during experimentation. All connections









were sealed with PTFE tape, since several of the target analytes are volatile organic

compounds (VOCs).

Since each sample taken for organic analysis from the system was 40 mL, a

source tank was necessary to prevent pockets of air from forming in the tubing or in the

reactor. A 4- L Erlenmeyer flask was used. Glass rods placed through a rubber stopper on

top of the source tank allowed solution to be carried in and out of the flask. The rubber

was covered in Teflon tape to preclude reactions between the stopper and the test

solution. One glass tube through the stopper allowed a small air leak to prevent a vacuum

situation within the system when samples were taken. This did result in a necessary 14.5

mL of headspace at the top of the flask to prevent the loss of solution out of the vent tube.

The total volume of the system was just over 4.5 L with 4.3 L from the source tank and

the reactor. The pump used was a L/S PTFE-Tubing Pump Head powered by a L/S

Variable-Speed Modular Drive. A polyethylene pulse dampener was used to ensure a

steady flow rate in the system.

Reactor hydrodynamics. The hydrodynamic behavior of reactors operates in a

range between two ideal reactor models. These models, the continuously stirred tank

reactor (CSTR) and the plug flow reactor (PFR), represent two extreme cases of fluid

flow behavior. A CSTR consists of a continuous flow of fluid particles, which are

assumed to be completely mixed, and the concentration in all locations within the reactor

is the same as the effluent concentration. A PFR represents a fluid flow with no mixing in

the direction of the flow and infinite mixing perpendicular to the flow direction, and,

therefore, a concentration gradient exists from the influent to the effluent of the reactor.

The PFR is representative of an infinite number of CSTRs in series.









In order to determine the hydrodynamic behavior of the reactor, Holmes (2003)

performed a tracer analysis. Sodium chloride (NaC1) was used as the tracer to prevent

reactions with the composite pellets or the inside of the reactor. The chosen flow-rate for

the tracer analysis was 10 mL/min since this was the design flow for the reactor. With a

reactor volume of 436 mL and 109 mL of that composed of pellets, the mean residence

time of the reactor was predicted to be 32.7 min.

The conductivity of the effluent was measured using a Fisher Scientific

conductivity probe. A linear correlation was found between the conductivity reading on

the probe and the concentration of NaCl in solution by measuring known concentrations

NaCl and comparing them with the conductivity reading.

The mean residence time, variance, and number of tanks in series are shown in

Table 3. The residence time represents the average amount of time the NaCl remained in

the reactor. The variance describes the variability in times over which the contaminant

exited the reactor. The number of CSTRs in series was calculated to model the residence

time distribution of the reactor. An infinite number of CSTRs in series represents a plug

flow reactor.

Table 3. Key statistics from the tracer analysis (Holmes, 2003)
Mean Residence Time 42.1 min
Variance 459
Number of CSTRs in series 3.9


A residence time distribution (E-curve) was generated from the tracer analysis

data and is shown in Figure 6. This distribution was then compared to the tanks-in-series

(TIS) model of CSTRs, which was created using the following equation:









-nt
S \n-l
ne bar nt

t(n -1)! t (10)


Where n is the number of CSTRs in series, t is the time in the reactor, and tbaris the mean

residence time. The tanks-in-series (TIS) model was used because of its ability to

accurately model and simulate the flow in the reactor.


E Curve


25.0

20.0


I.LU

10.0

5.0

n n


- Data
--- Model


0.00 20.00 40.00 60.00 80.00 100.00 120.00
t* (min)

Figure 6. E-curve generated from the tracer analysis performed on the reactor. "Data"
represents the E-curve from the tracer analysis of the reactor, and "Model" is the E-curve
generated from the TIS model (Holmes, 2003)

A comparison with the TIS model revealed that the reactor behaves as

approximately four CSTRs in series.

Reactor for Inactivation of Bacteria and Viruses

The vertical cylindrical reactor, shown in Figure 7, was used for the

microbiological experiments with bacteriophages and bacteria. The reactor was designed









with a hollow center and thin annulus to allow the UV lamp to be placed in the center,

providing maximum UV light exposure to the pellets.













Figure 7. Vertical quartz reactor packed with silica-titania composites.

The inner wall of the annulus was a quartz tube that could be completely

removed, making it simple to remove the pellets after testing. Either a 365 nm or 254 nm

wavelength lamp was placed in the center of the quartz tube. As measured at the center of

the 365nm lamp, the intensity was 7.4 mW/cm2 at the inner diameter of the annulus and

decreased to 4 mW/cm2 at the outer diameter with no pellets in the reactor. For the 254

nm lamp, the intensity was 12 mW/cm2 at the inner diameter and 8 mW/cm2 at the outer

diameter without pellets. With pellets in the reactor, the UV intensity at the outer

diameter was near zero for both the 254-nm and 365-nm lamps. The reactor was 19 cm

long with an inner diameter of 2.5 cm and an outer diameter of 4.2 cm. The empty bed

volume of pellets was 138.6 mL. The reactor was enclosed in a box to provide control

over its exposure to ambient light.

Reactor system. The system, shown in Figure 8, consisted of the reactor, 6 mm

PTFE tubing, a Master flex L/S Digital Standard Drive peristaltic pump, and a 500mL

source tank of stock spiked solution. The system was operated in a closed loop system in

which the solution was recirculated through the reactor. A cover (not shown in Figure 8)









was placed over the front of the reactor to completely enclose it during operation of the

system.

















Figure 8. Treatment system consisting of reactor packed with SiO2-TiO2 composites, UV
lamp, peristaltic pump, and reservoir of stock solution

Reactor hydrodynamics. Ludwig (2004) conducted a tracer analysis to

determine the behavior of the reactor. The tracer analysis method was described

previously in the discussion of the organic contaminant reactor.

The data collected from the tracer analysis was used to calculate a mean residence

time of 12.5 minutes and to model the reactor behavior. A fractional age distribution (E-

curve) of the sodium chloride in the reactor was created and compared to the TIS model.

The comparison revealed that the reactor behaved as five CSTRs in series.

Analytical Methods

Analysis of Organic Compounds

Samples were collected from one of the two three-way luer-lock sampling ports

provided in the system using a gas tight, luer-lock syringe. The samples were then

transferred to 40-mL volatile organic analysis (VOA) vials. The vials were stored at 4C









until analysis. Each sample was viable for up to two weeks at that temperature. A GCQ

gas chromatograph/ mass spectrometer with a Tekmar 3100 purge and trap extraction

system was used to analyze the samples for all six of the constituents. Analyses were

performed in accordance with the USEPA method 524.2, Methods for the Determination

of Organic Compounds in Drinking Water: Supplement 2 (USEPA, 1992). Samples were

purged at room temperature for 11 minutes at 35 mL/min with helium. They were then

dry purged for 2 minutes to remove water. A Supelco k-trap was then back flushed and

heated to 2500C for desorption to the column. The column used was a DB-VRX column

from J&W Scientific. It was a 75-meter long column with a 0.45 mm inner diameter, and

a 2.55-jam-film thickness. Desorption to the column lasted for 6 minutes at 2500C, and

the trap was then baked at 2700C for 10 minutes. The column was taken from a 25C

start temperature and ramped to 2200C at a rate of 60C/min, after an initial hold at 35C

for 6 minutes. The mass spectrometer used is an ion trap that scans from 34 amu and 280

amu with 0.6 seconds/scan.

The reliability of the GC/MS analysis was verified using the percent relative

standard deviation (RSD). The RSD is based on the response factor (RF) of a chemical.

The RF and RSD are calculated using the following equations:

RF = (Aa* Cis)/(Ais* Ca) (11)

Where Aa= GC peak area of the analyte, Ai = GC peak area of internal standard, Ca=

concentration of the analyte, and Ci = concentration of the internal standard.

RSD = 100 (SRF/MRF) (12)

Where SRF = standard deviation of response factors and MRF = average response factor.

The USEPA method requires a RSD of less than 30%, but strongly recommends an RSD









of less than 20%. Table 4 gives the average RSDs for the analyses performed in this

research.

Table 4. Average RSDs of SOC analysis
Compound Average RSD (%)
Acetone 32.59
Carbon Disulfide 13.71
Chlorobenzene 14.91
Ethyl Acetate 18.31
Methyl Methacrylate 20.14
Toluene 12.79

Analysis of Bacteria and Viruses

Bacteriophage. Assays were performed to determine the concentration of

biologically active agents in a sample. A double overlay phage assay was performed for

the three active bacteriophages, OX-174, PRD-1, and MS-2. The bacterial hosts for the

phages were Escherichia coli B (OX-174), Salmonella typimurium (PRD-1), and

Escherichia coli C-3000 (MS-2). The phage assay, as reported by Matthews (1971),

depends upon the ability of an active phage particle to clear a small portion of a "lawn"

of infected bacteria, which has been seeded on a nutrient agar plate. In a tube of melted

nutrient agar, a dilution of phage and sufficient host bacteria to cover the surface of the

plate were combined. The mixture was poured evenly onto the nutrient agar plate. After

the agar has hardened, the plate was incubated at 37C for 8 12 hours in a GCA

Precision Scientific Model 6 Incubator. The phage infective cycle normally terminates

with lysis, or rupture, of the infected cell, with release of the newly formed virus

progeny. Since diffusion of the newly formed virus is prevented in the solid agar

medium, the progeny infect any uninfected cells in their immediate vicinity. Elsewhere

the bacterial lawn grows undisturbed, giving an opaque appearance to most of the plate.









Each plaque is a clear area, resulting from lysis of all of the cells in the immediate

vicinity of an infectious particle. A countable plate contains between 30 and 300 plaques.

Bacteria. A bacteriological assay was performed for the three active bacteria,

Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa, using a

heterotrophic plate count to obtain a viable count of colony forming units (CFUs) of

living bacteria in the samples. As with the phage procedure, a countable plate contains

between 30 and 300 CFUs. A dilution of bacteria was spread evenly using a sterilized

glass stirring rod on a nutrient agar plate. Bacteria-specific media were used to

distinguish between the three bacteria in the samples, MacConkey agar for Escherichia

coli C-3000, Mannitol Salt agar for Staphylococcus aureus, and Pseudomonas isolation

agar for Pseudomonas aeruginosa. The plates were incubated at 37C for 24 48 hours.

Experimental Procedures

Oxidation of Organic Compounds

In order to test the efficiency of the reactor system to degrade the organic

compounds, a spiked solution of the components of interest was made. A 50-mL

volumetric flask was filled with nanopure water. Using a 10-tL syringe, specified

amounts of each compound (acetone, carbon disulfide, chlorobenzene, ethyl acetate,

methyl methacrylate, toluene) were added to the solution. The amounts usually ranged

between 0.70 3.0 pL. Between each addition, the solution was capped to minimize

volatilization, and the syringe was rinsed using methanol. After all the compounds were

added to the solution, the flask was inverted several times to ensure thorough mixing. The

contents of the flask were then transferred to a 50-mL air-lock syringe. A ratio of 1:100

was achieved for the initial solution. A 4-L Erlenmeyer flask (source tank) was filled









with 3960 mL of nanopure water. Using the syringe, 40 mL of the solution was

transferred to the 4-L source tank. The remaining 10 mL of solution was added to a 1-L

volumetric flask containing 990 mL of nanopure water. The source tank contained a

stirring rod and was placed onto a stirrer to ensure proper mixing of the initial solution.

The silica-titania composites packed in the reactor were the same for experiments

performed by Holmes (2003) and remained in the reactor throughout the organic

experiments.

Replication. For the initial degradation experiment, a representative procedure

followed by Holmes (2003) was adopted. Holmes performed experiments testing the

effects of UV radiation intensity, contact time, and adsorption. The system was filled

with the spiked solution at a rate of 10 mL/min. Once the reactor and tubing system were

full, additional spiked solution was added to the source tank to reduce headspace and

volatilization effects. At this time, samples were taken from Port 1 to record the initial

concentrations of the compounds. Holmes (2003) performed exhaustive adsorption

experiments with the reactor system and found that after five hours of recirculation the

adsorptive effect of the photocatalysts is minimal, so the solution was recirculated in the

reactor system for 5 hours in order to measure the adsorption of the compounds onto the

silica-titania pellets. After 4.5 hours of recirculation, two 12-inch, 8-Watt, 365-nm UV

lamps were placed in the center of the reactor and turned on to warm up. The lamps were

shielded with aluminum foil to inhibit early interactions with the photocatalysts. After 5

hours, the shields were taken off. A cover was placed over the reactor to protect

laboratory personnel from the UV light. The reactor system was then switched to the

single-pass system as shown in Figure 2. One liter of solution passed through the system









and into the waste tank. It was assumed that by this point the system had reached a stable

condition. Replicate samples were taken both at Ports 1 & 2 (before and after the reactor).

Alkalinity. Hydroxyl radicals created in the photocatalysis process can be

scavenged by carbonate and bicarbonate ions (Bekbolet and Baleioglu, 1996; Chen et al.,

1997; Munter, 2001). The effect of bicarbonate ions present in solution was investigated.

From bench top batch studies with the silica-titania composites, a buffering concentration

of 80 mg/L of NaHCO3 was found to be needed to hold the pH of the solution at around

6.3. An experiment was performed replicating the conditions of the two previous reactor

experiments. The buffer solution was made in one-liter increments. One liter of nanopure

water was measured using a volumetric flask and 80 mg of NaHCO3 was weighed and

deposited in the water. The flask was inverted for mixing to ensure that all the NaHCO3

was dissolved. HCI was added to the solutions via a Pasteur pipette until the pH of the

solution reached 6.3. The initial pH of the solution was about 8.3. Four liters of solution

were transferred to the 4L-erlenmeyer flask used as the source tank in the reactor

experiment. An additional liter was kept in a 1L-erlenmeyer flask to use to fill up the

system as needed. The flasks were covered with parafilm. The spiked solution was

created as previously discussed using the buffered solution instead of nanopure water.

The reactor experiment proceeded as discussed earlier with the only difference being that

pH measurements were also taken at several points in the experiment along with the

samples for analyzing the organic constituents. When the system was filled initially, the

pH of the solution before entering the reactor (Port 1) was 6.61. Immediately after exiting

the reactor (Port 2), the pH was 4.25. Table 5 shows the corresponding pH values for the

next two sample points in the experiment.
















Increased alkalinity. The reactor experiment was repeated with an increased

buffer concentration of 200 mg/L of NaHCO3. The procedure remained the same as in the

previous experiment. The initial pH of the buffer solution was 6.3, after the addition of

HC1. After the organic contaminants were spiked into the solution, the reactor system was

filled at a rate of 10 mL/min. The initial measured pH in the system was 6.49. The

solution recirculated in the system for five hours. Then, one liter of solution was passed

through the system in a single-pass mode before the samples were taken before and after

the reactor. Table 6 shows the corresponding pH values for the sample points:

Table 6. pH values for experiment with 200 mg/L HCO3 solution
pH t = 0 t = 4.5 hr t = 6.5 hr
Port 1 6.49 6.18 6.16
Port 2 6.48 5.66 5.65

High alkalinity. Using the alkalinity value reported for the simulated NASA

wastewater (Lange and Lin, 1998), 2360 mg/L as CaCO3, the buffer solution was

prepared to this specification. 3965 mg/L of NaHCO3 was used as the concentration of

the buffer solution. The initial pH value for the solution was 8.2. The organic

contaminants were spiked into the solution and the reactor experiment was run as

previously discussed with one exception. After one liter of solution was expended in the

single-pass mode and samples were taken before and after the reactor, the system was

switched back to recirculation for 3.5 hours. A sample was then taken at the effluent of

the reactor. Table 7 shows the corresponding pH values for the sample points in the

experiment.









solution


Low dissolved oxygen. Oxygen is the electron acceptor in the photocatalytic

reaction, preventing electron and electron hole from recombining and thus leaving open

the reaction sites for photocatalysis to take place. Accordingly, an experiment to simulate

a low dissolved oxygen (DO) environment and test the effectiveness of the oxidation of

the organic contaminants was performed.

The procedure used to achieve a low DO environment was to fill the reactor

system with nitrogen-sparged nanopure water, while continually adding more sparged

solution until the reactor system contains only sparged, low DO solution. The 4-L source

tank was filled with nanopure water and sparged with nitrogen to a level of 0.05 mg/L.

The sparged water was pumped into the reactor system and wasted. The initial DO level

at Port 1 was 3.35 mg/L and 4.4 mg/L at Port 2. The level in the source tank remained at

0.05 mg/L. The source tank was continually being sparged with nitrogen gas.

Each sample was taken from the system by extracting the water using the air-tight

syringe, transferring the solution to a 20-mL beaker, and bathing the sample with

nitrogen gas to minimize the transfer of oxygen from the atmosphere into the sample. A

YSI 52 Dissolved Oxygen Meter, previously calibrated to 100% saturation and standard

pressure, was used to obtain the measurements. The DO calibration was achieved by

subjecting the probe to a high relative humidity environment, i.e., placing the probe

inside a plastic cap with a sponge saturated with nanopure water.

The system was being circulated with low DO water from the source tank, and the

DO level was being monitored at Port 1 and 2. After 16 hours of wasting sparged water,









the Port 1 DO level was 0.99 mg/L and Port 2 was 1.34 mg/L. Two liters of solution had

been wasted to reach this point. At this time, the flow of solution was stopped, 200 mg/L

of NaHCO3 was added to the source tank and sparged with nitrogen gas to a level of 0.32

mg/L. The organic contaminants were injected into the source tank via a Tygon plastic

tube through the vent port. After the contaminants were spiked, the continual sparging of

the source tank was stopped because of the volatility of the contaminants. The system

was switched to a recirculating system. The DO was monitored before the five hours of

adsorption as 1.51 mg/L at Port 1 and 1.56 mg/L at Port 2. After adsorption, the system

was switched to single-pass and the UV lamps were turned on to sample for destruction.

After one liter of solution was wasted, samples were taken at Ports 1 and 2 for the

contaminants as well as for DO. The levels were 2.67 mg/L DO at Port 1 and 1.65 mg/L

DO at Port 2.

According to Standard Methods (1998), one method for calibrating a dissolved

oxygen probe for null dissolved oxygen is to supersaturate a water with sodium sulfite in

the presence of cobalt chloride, which acts as a catalyst in the reaction of stripping the

water of oxygen. Na2SO3 was added to nanopure water in excess by double the

stoichiometric amounts (2 moles of Na2SO3 for every mole of 02).

The next experiment with the low dissolved oxygen system involved using the

chemical addition of Na2SO3 and CoCl2 to reduce the DO level just before the single-pass

destruction step of the experiment procedure. The experiment began as the previous runs.

200 mg/L of NaHCO3 was added to nanopure water and the organic contaminants were

spiked into the solution. The solution was recirculated throughout the reactor system for

five hours without UV light for the adsorption step. The DO level after this step was









measured to be 6.88 mg/L and the pH was 8.05. When the reactor system was switched to

single-pass and the UV lights were turned on, an excess mixture of Na2SO3 and CoCl2

was injected into the source tank to lower the DO level. The source tank was also bathed

in nitrogen gas to prevent oxygen diffusing into the solution from the atmosphere. After

the one liter of solution was wasted, samples were taken for the organic and the DO

level monitored at Port 1 was 2.71 mg/L and the pH was 7.91. At Port 2, the DO level

was 4.06 mg/L and the pH was 7.99.

TOC destruction in synthetic wastewater. The next experiment performed

involved testing simulated NASA wastewater for total organic carbon (TOC). The

wastewater feed was created using a formula for typical wastewater provided by Johnson

Space Center in Houston, TX. The chemical composition includes the shower waste,

hand wash waste, oral hygiene waste, urine, and urine flush waste that would be expected

from a crew of four people. Listed in Table 8 are the amounts of each constituent present

in the simulated wastewater.

Table 8. Simulated wastewater composition (Ludwig, 2004)
Wastewater Component Quantity
Pert Plus for Kids 1.2 g
Deionized water 999.4 ml
Ammonium bicarbonate NH4HCO3 2726 mg
Sodium chloride NaCl 850 mg
Potassium bicarbonate KHCO3 378 mg
Creatinine C4H7N30 248 mg
Hippuric acid C9H9NO3 174 mg
Potassium dihydrogen phosphate KH2PO4 173 mg
Potassium bisulfate KHSO4 111 mg
Citric acid monohydrate C6H8O7-H20 92 mg
Tyrosine C9H11NO3 66 mg
Glucuronic acid C6HioO7 60 mg
1.48N Ammonium hydroxide NH4OH 10 ml









The formula represents the raw concentrated wastewater that would be the

influent for the beginning of the water treatment system and would be treated by several

processes before entering the post-processor. To account for pretreatment, the simulated

wastewater was diluted using 9 mL of the concentrated wastewater and 991 mL of

nanopure water for every liter of solution needed. This created a solution of

approximately 3 ppm of TOC, which is within the range of the expected TOC

concentration of the post processor influent (Cambel et al., 2003).

The parameters for the experiment remained the same as in previous runs of the

reactor. The initial pH of the solution was 8.37. After the five hours of recirculation and

single-pass through the reactor, the pH before the reactor was 7, and after the reactor, the

pH was 5.4. The samples taken were analyzed for TOC using a Tekmar Dohrmann

Apollo 9000 TOC Analyzer.

After the experiment with the NASA-simulated wastewater, nanopure water was

sent through the reactor to test for the desorption of TOC from the pellets after contact

with the wastewater. After the reactor was filled in 28 minutes, samples were taken at the

effluent of the reactor at 31, 51, and 71 minutes. The samples were analyzed for TOC.

Inactivation of Bacteria and Viruses

Preparation. Stock cultures were prepared from existing cultures of the

bacteriophages OX-174, PRD-1, MS-2. Confluent plates of each phage were prepared

using the double-overlay phage assay procedure described in the analytical methods

section. The medium for the plates was a DifcoTM Plate Count Agar. A confluent plate

contains around 103 plaques per plate. A 3% solution of BactoTM Tryptic Soy Broth

(TSB) was then poured onto the plates, and the top agar was scraped off using a flame-









sterilized bent glass rod. Using a Beckman J2-HS Centrifuge, the liquid solution was then

centrifuged for 10 minutes at 5,000 rpm. The supernatant was poured off into a second

sterile tube and centrifuged for 10 minutes at 10,000 rpm. The supernatant was then

filtered through a 0.45 [tm plastic filter using a sterile plastic syringe. The resultant stock

cultures were kept refrigerated at 4C. For each of the experiments the stock culture was

diluted from a starting titer of between 108 1010 PFU/mL down to the initial titer of the

experiment of 105 107 PFU/mL, depending on which bacteriophage. The stock cultures

of bacterial hosts used for the phage assay were maintained on plate count agar plates at

room temperature.

The nutrient agar used for the top layer of the double-overlay assay procedure was

made by mixing together BactoTM Tryptic Soy Broth, Fisher Scientific Purified Grade

Agar (CAS 9002-18-0) and deionized water. The mixture was autoclaved for 15 minutes

in an Amsco Eagle Series 2011 Gravity Autoclave to melt the agar, cooled in a National

Appliance Company Model 320 water bath for about an hour, and dispensed into 13mm

glass tubes (5 mL per tube). The tubes were capped, autoclaved for 15 minutes for

sterilization and refrigerated for storage.

To make serial dilutions of a sample, dilution tubes were prepared by mixing

BactoTm Tryptic Soy Broth with deionized water in a 1% concentration. Three mL of

solution was dispensed per 13 mm glass tube. The tubes were capped, autoclaved for 15

minutes for sterilization and refrigerated for storage. After autoclaving, the volume in

each tube is approximately 2.7 mL, so the addition of 0.3 mL of sample will produce a

1/10 dilution of the original sample.









Several agar plates were used in the microbial experiments performed. For the

bacteriophage assay, DifcoTM Plate Count Agar was used as the culture media. All three

of the bacteria were combined in one solution for each experiment. Each of the three

bacteria grows on the plate count agar, so isolating culture media were used to

differentiate between the three bacteria. BBLTM Mannitol Salt Agar was used for the

isolation of Staphylococcus aureus. BactoTM Pseudomonas Isolation Agar combined with

Fisher Scientific Glycerin (CAS 56-81-5) was used as the culture media for Pseudomonas

aeruginosa. DifcoTM MacConkey Agar Base combined with DifcoTM Lactose was used

for the differentiation of Escherichia coli based on fermentation reaction. The procedure

for preparing the agar plates is the same regardless of the media used. The specified agar

was mixed with deionized water and autoclaved for 15 minutes to melt the agar. The

container was cooled in a water bath of 50C for about an hour. The agar is poured onto

the plates, about 20 mL per plate, to cover the entire bottom of the plate. The plates

remain at room temperature overnight to harden and are stored at 4C in the refrigerator.

Each of the three bacteria cultures, Escherichia coli, Staphylococcus aureus, and

Pseudomonas aeruginosa, were grown in 3% TSB for 12 to 18 hours at 37C in an

incubator. The cultures were concentrated and washed twice by centrifugation (10

minutes at 10,000 rpm) with autoclaved deionized water. After each centrifugation, the

supernatant was poured off and autoclaved deionized water was added to the culture. The

solution was mixed together thoroughly using a Fisher Scientific Genie 2 Vortex mixer.

The titer of the bacteria was about 108 CFU/mL. After diluting with the autoclaved

nanopure water for the experiment solution, the titer was diluted to between 106-107

CFU/mL for each of the bacteria.









Recirculation experiments. The internal porosity of the silica-titania composite

could provide capacity for adsorption of the microorganisms. In order to determine this

capacity, the test solutions containing the three bacteriophages, OX-174, PRD-1, MS-2,

were recirculated throughout the system for four hours. Samples were taken initially and

after the four hours without UV light. Samples were taken at the influent to the reactor

(Port 1) and, twelve minutes later, at the effluent of the reactor (Port 2). All of the

samples (10 mL) were injected with 0.5 mL of a 3% TSB solution for stabilization and

kept refrigerated until analysis.

The silica-titania composites have an acidic characteristic in the unused state.

Since the pH of the solution can influence the survival and adsorption of the phages, the

solution was buffered to keep it at a nominal pH. The first bacteriophage run was

buffered using a 1 mM solution of potassium phosphate, and the pH of the solution was

monitored throughout the experiment. The initial stock solution started with a pH of 7.0

and the final sample taken at the effluent had a pH of 6.2.

After the four-hour adsorption step, the UV lamp was turned on and samples were

taken after 1.5 hours and 2.5 hours of recirculation with UV irradiation. The silica-titania

composites were also tested for desorption at the conclusion of the reactor experiment.

Three pellets were taken from the reactor and submersed in 5 mL of 3% TSB solution

until analysis.

The amount of time the solution spent in the reactor was very small compared to

the overall time of the experiment. Based on the flow rate and the volume of the reactor

system, the solution spent 33.6 minutes in the reactor during the 4 hrs of recirculation.

During the 1.5 hours of recirculation with UV light, the solution had spent 50.4 minutes









in the reactor with 12.6 minutes exposed to UV light. During the 2.5 hours of

recirculation with UV light, the solution spent a total of 58.8 minutes in the reactor with

21 minutes of UV irradiation.

Batch experiments. In order to examine the effects of longer UV contact time,

batch experiments were performed. The buffering solution was changed from potassium

phosphate to sodium bicarbonate (100 mg/L as CaCO3); since it is more likely that

bicarbonate will be present in the NASA wastewater system.

Adsorption. As in the previous experiments, the bacteriophages were recirculated

throughout the system for an initial adsorption step. After four hours of recirculation

without UV light, the effluent was sampled. After six hours of recirculation in the dark, a

sample was taken before the reactor (Port 1), the flow of the system was stopped, and the

solution remained in batch conditions for one hour. The flow was started and a sample

was taken at the effluent (Port 2) after five minutes to ensure the sample was from the

batch conditions.

Inactivation. Batch conditions were repeated in the same manner as in the

adsorption procedure for the inactivation steps of the experiment. Samples were taken

before the reactor, the flow of the system was stopped and the UV lamp was turned on

inside the reactor. After 1 hour, the flow was restarted and a sample was taken at the

effluent. The solution was then recirculated through the system for one hour without UV

light to bring the system to a steady state. The batch conditions were repeated for an

additional 2 hours. Three experiments were performed with the bacteriophages using the

batch procedure, including one with a 365-nm wavelength UV lamp, one with a 254-nm

UV lamp, and one without the silica-titania composites inside the reactor (using the 254-









nm lamp). Table 9 presents the effluent pH values for each of the three batch

bacteriophage experiments. The solution was held at a nominal pH by the 100 mg/L as

CaCO3 added.

Table 9. Effluent pH values for solutions containing sodium bicarbonate (100 mg/L as
CaCO3) along with spiked bacteriophages
Experiment ID Effluent pH (t = 0) Effluent pH (t = 9)
365nm lamp 6.9 6.65
254nm lamp 6.94 6.67
254nm lamp (no pellets) 8.18 7.50

Three experiments were also performed with bacteria using the batch procedure,

including two experiments with the 254-nm wavelength UV lamp and one without the

silica-titania composites (using the 254-nm lamp). Table 10 presents the effluent pH

values for each of the batch bacteria experiments.

Table 10. Effluent pH values for solutions containing sodium bicarbonate (100 mg/L as
CaCO3) along with spiked bacteria
Experiment ID Effluent pH (t = 0) Effluent pH (t = 9)
254nm lamp 7.07 6.48
254nm lamp 7.09 6.46
254nm lamp (no pellets) 8.11 7.84

As with the bacteriophages, the solution was held at a nominal pH by the addition

of 100 mg/L as CaCO3.

Statistics

A Poisson distribution event occurs at random in continuous space or time. Events

occur independently, uniformly, and singly. The number of colonies growing on agar

plates in a dilution plating assay is an event that should be randomly distributed in space.

The unit is the volume of sample, which contains organisms held in suspension, on the

plate. If the sample is well-stirred, the particular organisms of interest are distributed









through the sample of liquid at random and independently. Since the total number of

organisms per unit volume in the sample is constant, there is a constant probability of

counting a particular type of organism in plates of a given volume (Clarke and Cooke,

1992). Provided that the cells are randomly distributed and that the count is reasonably

high (greater than 30), then the count conforms to a Poisson distribution. This applies to

all the counts, so no degrees of freedom need to be calculated, the Student's t-value used

for the statistical tests is for an infinite degree of freedom (Deacon, 2005).

Since the count can be treated as part of a Poisson distribution, the microbial

concentration of colonies or plaques per milliliter of sample (CFU/mL or PFU/mL) has a

value of the number of colonies counted multiplied by the dilution of the plate.

A comparison of two Poisson counts can be made by using a hypothesis test for

significant difference. The t-value for the two counts is calculated by Equation 14 and

compared to the critical t-value (the Student's t-value for an infinite degree of freedom).

If the calculated t-value is higher than the critical t-value, then the two Poisson counts are

significantly different. The critical t-value for a probability of 0.01 is 2.58. The t-value is

calculated using the following formula (Deacon, 2005):


X, -(X +X,)* v -0.5
t = (14)





where Xi = first plate count, X2 = second plate count, Vi = volume of sample on first

plate, and V2 = volume of sample on second plate.
















CHAPTER 4
RESULTS

Oxidation of Organic Compounds

Adsorption

The internal porosity and high surface area of the silica-titania composite provide

significant adsorption capacity for some of the contaminants. This adsorption, likely

physisorption resulting from van der Waals forces for organic compounds, is considered

to be advantageous to the oxidation process in collecting the target contaminants at the

surface of the titanium dioxide where the reactive hydroxyl radicals are produced. The

majority of the oxidation is likely to occur near the surface of the silica-titania composite,

therefore, the adsorbed compounds will transport from the inner pores of the silica

structure to the external portion for oxidation.

In order to determine the extent of photocatalytic destruction of the contaminants,

test solutions were recycled through the reactor at 10 mL/min for five hours in the dark,

without UV radiation. Holmes (2003) found the reactor influent and effluent

concentrations were nearly equal after five hours of recycling the test solution through

the reactor system. The representative results for the adsorption of chlorobenzene and

toluene are shown in Table 11. Chlorobenzene and toluene are the two compounds with

the highest octanol-water partition coefficients. All of the adsorption data for each

experiment can be found in Appendix A.









Table 11. Removal of toluene and chlorobenzene in attaining adsorption equilibrium in
the reactor. Solution recycled through reactor with no UV radiation for five hours
(Holmes, 2003)
Contaminant Initial Concentration % Removed by Adsorption
Toluene 76 [tg/L 38
Chlorobenzene 121 tg/L 45

Oxidation

All photocatalytic destruction experiments were preceded by recirculating the test

solution for five hours to allow adsorption effects to stabilize as indicated above. The

flow rate through the reactor during the series of experimental runs remained at an

optimized value of 10 mL/min, with an EBCT of 43.6 minutes (Holmes, 2003). Holmes

also found that for this designed reactor the photocatalytic destruction of the

contaminants was not limited by the number of lamps used beyond the use of two lamps;

therefore, following the preliminary adsorption step, two UV lamps were turned on and

allowed to warm up while they were shielded from the reactor. One hour was allowed for

steady-state operation to be attained before samples were collected at the inlet and outlet

of the reactor.

An increase in the number of UV lamps used could cause the temperature of the

solution in the reactor to rise. This temperature change could have an effect on the UV

output. There is an optimal temperature to maximize the UV intensity of the lamps. If

the temperature changes either above or below this optimal temperature, then the UV

intensity is less than the maximum possible for a given number of lamps.

Replication. Two initial oxidation experiments were conducted by spiking the

selected SOCs into nanopure water. Preceded by the adsorption step, the solution was

irradiated by a 365-nm UV lamp and run through the system in a single-pass mode. The

percent removal of each of the SOCs for the two duplicate experiments and a similar







50


experiment performed by Holmes (2003) are shown in Figure 9. The silica-titania

composites were not removed from the reactor and were the same pellets used for

experiments performed by Holmes. The pellets remained in the reactor throughout the

organic experiments performed in this research.


Oxidation of SOCs
Single Pass at EBCT = 43.6 min
Test 1 Test 2 0 Holmes Experiment
100
90
80
70
> 60
E 50
40
30
20
10
0
Toluene Chlorobenzene Acetone Carbon Ethyl Acetate Methyl
Dsulfide Methacrylate

Figure 9. Percent removal of spiked SOCs in a single pass through the reactor with UV
irradiation and without addition of alkalinity

The percent removals represented in Figure 9 are for the single-pass oxidation

step and do not include the removal by adsorption. The removal achieved by the first two

experiment runs were either greater than or equal to the percent removal seen by Holmes

in a similar experiment set-up with the organic reactor. These results indicate that the

silica-titania pellets in the reactor were still capable of oxidizing the organic contaminants

after multiple experiments (15 performed by Holmes).

The methyl methacrylate data for the first two runs were not reported because the

solution spikes for these experiments were not successful in producing influent

concentrations above 15 tg/L. Therefore, the removal was not statistically acceptable due

to the low range of values.









Alkalinity. The initial oxidation experiments were followed by experiments with

elevated levels of bicarbonate ion concentration. Photocatalytic oxidation may be limited

by bicarbonate ions, which can scavenge hydroxyl radicals resulting in unreactive

bicarbonate radicals. The effect of bicarbonate ions present in solution was investigated

by adding differing amounts of sodium bicarbonate to increase the alkalinity of the

solution and buffer the pH. In initial experiments, without bicarbonate buffering, the

typical pH exiting the reactor system was in the range of 3.8 4, because of the acidic

nature of the pellets.

Experiments with solutions containing a buffering concentration of 80 mg/L of

NaHCO3 (48 mg/L as CaCO3), 200 mg/L of NaHCO3 (120 mg/L as CaCO3), and 3965

mg/L of NaHCO3 (2360 mg/L as CaCO3) were performed. The 2360 mg/L as CaCO3

alkalinity value was chosen because this was the expected alkalinity of the typical NASA

wastewater (formula provided by Johnson Space Center is in Chapter 3: Materials and

Methods). The wastewater feed was created using a formula for typical wastewater. The

pH values for each of the experiments containing bicarbonate ions are presented in Table

12 as well as the pH values for no bicarbonate present.

Table 12. pH values for spiked organic contaminant solutions containing differing
amounts of sodium bicarbonate. Effluent pH value is after five hours of recirculation
through the packed reactor.
Experiment Solution Influent pH Effluent pH
0 mg/L NaHCO3 7.0 3.8
80 mg/L NaHCO3 6.6 4.5
200 mg/L NaHCO3 6.5 5.7
3965 mg/L NaHCO3 8.2 7.5

As expected, as the concentration of sodium bicarbonate increased, the ability of

the solution to buffer the pH increased. The sodium bicarbonate addition provided a










solution that was more realistic and closer in composition to the expected components of

the NASA wastewater. The amount of alkalinity expected to enter the post processor,

according to a revised wastewater formulation, has a nominal value of 58 mg/L as CaCO3

and worst-case value of 75 mg/L as CaCO3 (Verostko et al., 2004). Both of these values

lie between the 80 mg/L as NaHCO3 (48 mg/L as CaCO3) and 200 mg/L as NaHCO3

(120 mg/L as CaCO3) solutions tested. The percent removal of each of the SOCs for the

experiments including bicarbonate alkalinity is shown in Figure 10.




100-

80

0o 60

40-

20 -

0
Toluene Chlorobenzene Acetone Carbon Ethyl Acetate Methyl
Disulfide Methacrylate

0 0 mg/L NaHCO3 E 80 mg/L NaHCO3 0 200 mg/L NaHCO3 0 3965 mg/L NaHCO3


Figure 10. Percent removal of six target analytes in solution containing differing amounts
of bicarbonate operating in single-pass mode after five hours of adsorption in the
recirculation mode (error bars indicate one standard deviation on either side of the mean)

The experiment runs without bicarbonate addition are represented in Figure 10

("0 mg/L NaHCO3") by the average of the two runs. In order to test the organic

contaminant experiments for significant difference, each chemical was considered

independently. A one-way analysis of variance (ANOVA) test was used to compare the

experiments. In order to use the ANOVA test, the data needed to first be tested for equal

variances. The percent removals for each of the experiments were tested using Bartlett's









and Levene's tests for equal variances. The p-value calculated for toluene was 0.152.

Since this value is greater than 0.05, equal variances can be assumed and an analysis of

variance (ANOVA) can be performed. Each of the six analytes had a p-value above 0.05

(Appendix D). The ANOVA p-value calculated for toluene was 0.001, less than 0.05,

indicating that at least one of the experiments was significantly different than the others.

The high bicarbonate concentration experiment (3965 mg/L) was significantly different

than all of the other experiments (0 mg/L NaHCO3, 80 mg/L NaHCO3, 200 mg/L

NaHCO3) for toluene and chlorobenzene. The high bicarbonate experiment was

significantly different than no bicarbonate for ethyl acetate and both no bicarbonate and

200 mg/L of NaHCO3 for carbon disulfide. For acetone and methyl methacrylate, all of

the bicarbonate experiments were not significantly different. All of the oxidation

experiment data are in Appendix B, and the ANOVA tables can be found in Appendix C.

The addition of bicarbonate ion did not have an appreciable effect on the removal

efficiency at the two lower levels of 80 and 200 mg/L of NaHCO3. There was no

significant statistical difference in the percent removed between the experiments with 0

mg/L, 80 mg/L and 200 mg/L of NaHCO3. A decrease in the percent of the contaminants

removed can be seen in the high alkalinity experiment of 3965 mg/L NaHCO3. The

decreased removal can be attributed to the scavenging of hydroxyl radicals by the

abundance of bicarbonate ions in the solution. A decreased removal for the high

bicarbonate concentration is not seen with ethyl acetate or methyl methacrylate.

The results shown in Figure 10 indicate a substantial destruction of most of the

target analytes. Removal of acetone was less effective, although analytical precision was

outside of the acceptable QA/QC range for acetone (relative standard deviation of









replicate injections <20%). Some methyl methacrylate data were not reported because of

solution spikes for these experiments not being successful in producing influent

concentrations above about 15 ug/L.

Low dissolved oxygen. An experiment simulating a low dissolved oxygen (DO)

environment was performed in order to test the effectiveness of the oxidation of the

organic contaminants. The low DO environment was achieved by filling the reactor

system with nitrogen-sparged nanopure water. The influent DO level in the reactor

system was 0.99 mg/L, and the effluent DO was 1.34 mg/L. After one liter of solution

was wasted to obtain a steady state condition, the DO at the inlet of the reactor was

measured to be 2.67 mg/L and 1.65 mg/L DO at the effluent of the reactor. The elevated

dissolved oxygen levels prior to oxidation were attributed to oxygen dissolution from

trapped air in the reactor and pellets. The loss of dissolved oxygen during oxidation was

attributed to use in the photocatalytic reaction. The earlier experiments were presumably

at DO saturation of 9 mg/L; dissolved oxygen measurements were not taken for the

earlier experiments.

It was expected that with a lowered dissolved oxygen concentration, that less

hydroxyl radicals would be available for photocatalysis. The solution also contained 200

mg/L of NaHCO3. The results, shown in Figure 11, indicate that the percent removed still

remained above 80% for toluene, chlorobenzene, carbon disulfide, and methyl

methacrylate. The removal of acetone remained lower than the other contaminants.












100 -

80 -

0 60-

40

20


Toluene Chlorobenzene Acetone Carbon Ethyl Acetate Methyl
Disulfide Methacrylate
So 0 mg/L NaHCO3 D 80 mg/L NaHCO3 U 200 mg/L NaHCO3
3965 mg/L NaHCO3 0 Low DO (2.6 mg/L)

Figure 11. Percent removal of six target analytes operating in single-pass mode after five
hours of adsorption in the recirculation mode (error bars indicate one standard deviation
on either side of the mean)

The low DO experiment was compared to the previous experiments using a one-

way ANOVA statistical test. For chlorobenzene, acetone, carbon disulfide, and ethyl

acetate the low DO run percent removals fall within the standard deviation of the earlier

results. There is no significant statistical difference between the low DO experiment and

all of the previous bicarbonate experiments for these compounds.

For toluene, the low DO experiment was statistically different from the "0 mg/L

NaHCO3" run and the "200 mg/L NaHCO3" run but was not significantly different from

the "80 mg/L NaHCO3" experiment. For methyl methacrylate, the low DO experiment

was statistically different from the "80 mg/L NaHCO3" run and the "3965 mg/L

NaHCO3". All of the ANOVA tables are in Appendix C, and the experimental data for

the oxidation of the organic compounds can be found in Appendix B.

One possible explanation for the low DO environment not having an impact on

the degradation of the organic contaminants stems from the role of acetone. Acetone and









other alcohols are electron hole scavengers. By acetone directly reacting with the electron

holes, the role of the electron acceptor (oxygen) is diminished. Oxygen is important in the

photocatalytic oxidation process for combining with electrons to prevent electron and

electron hole recombination, but if the electron hole is being occupied by acetone (or

another substance), the role of oxygen is less important.

TOC destruction in synthetic wastewater. The next experiment performed

involved testing the simulated NASA wastewater for total organic carbon (TOC). The

simulated NASA wastewater was passed through the system as in the previous

experiments with five hours of adsorption before the single-pass run through the reactor.

The initial TOC concentration of the solution was 2.73 mg/L. After the five hours of

adsorption, the TOC concentration before the reactor was 2.36 mg/L. After the single-

pass through the reactor, the TOC concentration after the reactor was 1.44 mg/L. The

percent removals of TOC during the two stages of the experiment are shown in Table 13.

Table 13. Percent removal of total organic carbon operating in the single-pass mode after
five hours of adsorption in the recirculation mode
Mode of Removal % TOC Removed
Adsorption 13.5
Oxidation 38.8
Total 47.0

The amount of TOC adsorbed to the silica-titania composites was calculated by

multiplying the volume of water passed through the system (3 L) by the difference in

concentration during the adsorption recirculation step (2.73 2.36 mg/L). The TOC

adsorbed by the pellets was 1.01 mg. After the experiment, nanopure water was passed

through the reactor to test for the desorption of TOC from the pellets after contact with

the wastewater. The desorption concentrations were plotted against the amount of








57



nanopure water passed through the system. In Figure 12, the data was fitted by a


polynomial trend line. The data appears to match a polynomial trend for the initial


desorption between 0.03 L and 0.23 L.


Desorption Polynomial Trend


0 0.1 0.2 0.3 0.4 0.5
Volume passed (L)


Figure 12. Polynomial trend line of desorbing TOC concentration of simulated
wastewater versus the amount of nanopure water passed through the system


The TOC desorbed was calculated as the area under the curve between 0 and 0.23


to be 0.23 mg. In Figure 13, the data was fitted by a exponential trend line. The data


appears to match the exponential trend for the desorption between 0.23 L and 0.43 L.


Desorption Exponential Trend


1.6 -
- 1.4-
S1.2 -
1.0-
0.8 -
-
0.6-
-
o 0.4-
0.2 -
0.0


0 0.1 0.2 0.3 0.4 0.5
Volume passed (L)


Figure 13. Exponential trend line of desorbing TOC concentration of simulated
wastewater versus the amount of nanopure water passed through the system


y = 1B,. -' -9 0 .2-., + 1 8-421
R-= 1


y = C0 Oe '
R- =1


I 0









The TOC desorbed was calculated as the area under the curve between 0.23 and

infinity to be 0.06 mg. The total calculated TOC desorbed from the pellets in the

nanopure water rinse was 0.29 mg as compared to the 1.1 mg adsorbed to the pellets. The

remaining TOC could have remained adsorbed to the pellets or have been oxidized along

with the non-adsorbed TOC. The mass of TOC removed in the single pass with UV

irradiation was 2.75 mg. All of the TOC data is contained in Appendix E.

Inactivation of Bacteria and Viruses

Recirculation Experiments

In order to determine the capacity of the silica-titania composites for attachment,

the test solutions containing the three bacteriophages, MS-2, PRD-1, OX-174, were

recirculated throughout the system for four hours. Samples were taken initially and after

the four hours without UV light. Samples were taken at the influent to the reactor (Port 1)

and after twelve minutes, or the approximate mean hydraulic residence time, at the

effluent of the reactor (Port 2).

After the four-hour preliminary step without UV radiation, the UV lamp was

turned on and samples were taken after 1.5 hours and 2.5 hours of recirculation (t = 6 and

t = 7, respectively). The logo concentrations of phage in the samples are shown in Figure

9. The units of concentration are in plaque-forming units (PFU) per milliliter (mL). The

silica-titania composites were also tested for desorption at the conclusion of the reactor

experiment. Three pellets were taken from the reactor and submersed in 5 mL of 3% TSB

solution until analysis. The "pellets" column in Figure 14 represents this desorption or

detachment data.










7.00
UV turned on at t = 4.5 Pr, ..
6.00 PPo I

E 5.00
U-
S4.00
U
0 3.00
S2.00
.J
1.00

0.00
Initial Initial t=4 t=4 t=6 t=6 t=7 t=7 Pellets
Port 1 Port2 Port 1 Port2 Port 1 Port2 Port 1 Port2


Figure 14. Bacteriophage adsorption and inactivation experiment in a recirculation mode
with a potassium phosphate buffer

N represents the concentration of bacteriophage in the sample. No represents the

initial phage concentration. The log removal of the phage in the sample can be calculated

by subtracting the log of the concentration from the log of the initial concentration (log

No- log N). Another way of expressing this removal is by applying a log transformation

to (log No log N) to achieve log[No/N]. For Figure 14, the total log [No/N] values were

calculated by subtracting the log of the "t=7 Port 2" value from the log of the "Initial Port

1" value. The total log [No/N] values were 1.4, 0.0, and 0.82 for OX-174, PRD-1, and

MS-2, respectively. The log [No/N] values resulting from UV irradiation were 1.16, 0.0,

and 0.67, respectively.

The amount of time the solution spent in the reactor was very small compared to

the overall time of the experiment. Based on the flow rate and the volume of the reactor

system, the solution spent 33.6 minutes in the reactor during the 4 hrs of recirculation.

During the 1.5 hours of recirculation with UV light, the solution had spent 50.4 minutes

in the reactor with 12.6 minutes exposed to UV light. During the 2.5 hours of









recirculation with UV light, the solution spent a total of 58.8 minutes in the reactor with

21 minutes of UV irradiation.

Batch Experiments

In order to investigate the effects of longer UV contact time, batch experiments

were performed. Assuming the UV intensity on the reactor is 8 mW/cm2, as measured at

the outer diameter distance of the reactor without pellets or the reactor, the UV dose in

the batch system was 28.8 J/cm2 and 57.6 J/cm2 for the 60 minute and 120 minute

irradiation times, respectively.

The pH buffering solution was changed from potassium phosphate to sodium

bicarbonate (100 mg/L as CaCO3); since it is more likely that bicarbonate will be present

at these concentrations in the NASA wastewater system as compared to phosphate.

Adsorption/Attachment. As in the previous experiments, the bacteriophages

were recirculated throughout the system for preliminary adsorption/attachment step. After

four hours of recirculation without UV light, the effluent was sampled. After six hours of

recirculation in the dark, a sample was taken before the reactor (Port 1), the flow of the

system was stopped, and the solution remained in batch conditions for one hour. The flow

was started and a sample was taken at the effluent (Port 2) after five minutes to ensure the

sample was representative of the batch condition. A representative result from the

adsorptive step of the batch experiment is shown in Figure 15.











Bacteriophage PRD-1 U First Assay
Second Assay
2 8.00
E
M 7.00-
Ll
S6.00
5 5.00
4.00
8 3.00
2 2.00
0 1.00-

Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2

Figure 15. Duplicate sample analysis of adsorption/attachment removal of PRD-1 for the
batch bacteriophage experiment with no UV light

Each of the samples taken from the system was plated in replicate assays on two

separate occasions, represented by the two bars in Figure 15. The second assay was

performed following the analysis of the first assay, 12 24 hours later, to ensure the

proper dilutions were being plated in the analysis. In order to determine if the replicates

were significantly different, a hypothesis t-test for Poisson counts was performed (as

explained in the statistics section of Chapter 3). The critical t-value for significance was

2.58 at a probability level of p = 0.01. The calculated t-values for the replicates shown in

Figure 15 were 2.02, 0.30, 0.25, 0.97, and 1.63 for each of the respective samples. Since

all of these values fell below the critical t-value, they were assumed to be not

significantly different. These results suggest that over the 24-hour period the PRD-1

concentration in the sample did not change. It also supports the QA/QC of the analytical

method by the replication of data.

The results for the bacteriophage MS-2, shown in Figure 16, are from the

adsorption/attachment period of the same batch experiment as the data in Figure 15.











Bacteriophage MS-2 FirstAssay
Second Assay
:2 7.00
E
m 6.00
U-
5.00
0 4.00
3.00
C 2.00
S1.00
0 0.00
-J
Initial PFrt 1 Initial Port 2 t=4 Fbrt 2 t=6 Port 1 t=6 Port 2



Figure 16. Duplicate sample analysis of adsorption/attachment removal of MS-2 for the
batch bacteriophage experiment with no UV light

As with the PRD-1, each of the samples taken from the system was plated in

replicate assays on two separate occasions, 12 24 hours following the analysis of the

first assay, and are represented by the two bars in Figure 16. The calculated t-values for

the replicates shown in Figure 16 were 1.66, 4.83, 3.99, 6.26, and 4.74 for each of the

respective samples. Most of these values fell above the critical t-value (2.58), the

replicates were significantly different for all of the MS-2 samples in Figure 16 except for

"Initial Port 1."


Since there was a difference in replicates for the MS-2 phage, the first and second

platings of the MS-2 samples were compared using a two-sample paired t-test to

determine whether there was a significant difference between the two occasions of

plating. The null hypothesis for the test was that the mean differences of the two platings

were the same. The p-value calculated for this test was 0.005, therefore the null

hypothesis was rejected. These results suggest that over the 24-hour period the MS-2

concentration in the sample decreased, but this difference remained consistent throughout









all of the samples. The results also support the QA/QC of the analytical method. Since

the mean difference between the two platings was not the same, the results should be

compared per plating occasion, i.e., the "Initial Port 1 First Assay" should be compared

with the "Initial Port 2 First Assay" and not the "Initial Port 2 Second Assay."

In order to determine whether there was a significant amount of

adsorption/attachment or other loss of phage during the preliminary recirculation without

UV radiation, the MS-2 plate counts for "Initial Port 1" and "t=6 Port 2" were compared

using a t-test. The t-value calculated for the "First Assay" in Figure 16 was 0.46, which is

less than the critical t-value of 2.58 (p = 0.01), indicating that the two counts are not

significantly different. Therefore, there is little adsorption or attachment of the phages in

the six hours of recirculation and in the one hour of batch conditions at t = 6.

Similar results were obtained for the adsorption of Escherichia coli,

Staphylococcus aureus, and Pseudomonas aeruginosa in the batch experiments

performed with the three bacteria. The adsorptive data and statistical values for all of the

bacteriophage and bacterial experiments can be found in Appendix F.

Inactivation. The batch conditions were repeated in the same manner as in the

adsorption/attachment procedure. The solution flow was stopped and UV irradiation

applied for a one-hour (t = 7) inactivation experiment. After sampling, the solution was

recirculated throughout the system for one hour. The solution flow was then stopped and

an additional two hours of UV irradiation was applied for the two-hour (t = 9)

inactivation experiment. When calculating the log [No/N] inactivation values, No was the

concentration at Port 1 (before the reactor) and N was the concentration at Port 2 (after










the reactor). For the additional two-hour inactivation experiment, No was the Port 1

concentration taken just before the two-hours of continuous UV.

Three experiments were performed with the bacteriophages using the batch

procedure, including one with a 365-nm wavelength UV lamp, one with a 254-nm UV

lamp, and one without the silica-titania composites inside the reactor (using the 254-nm

lamp). The effluent pH values for each of the batch bacteriophage experiments are shown

in Table 14.

Table 14. Effluent pH values for solutions containing sodium bicarbonate (100 mg/L as
CaCO3) along with spiked bacteriophages
Experiment ID Effluent pH (t = 0) Effluent pH (t = 9)
365nm lamp 6.9 6.65
254nm lamp 6.94 6.67
254nm lamp (no pellets) 8.18 7.50

The solution was held at a near neutral pH by an addition of sodium bicarbonate

(100 mg/L as CaCO3) added to the solution. A representative plot of the inactivation

results for the bacteriophages is shown in Figure 17.


U First Assay
Bacteriophage PRD-1 second Assay
7 Second Assay
2 7.00
6.00
Ll
5.00
0
S4.00
3.00
o 2.00-
o 1.00-
o0
0
-J 0.00
t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2


Figure 17. Duplicate sample analysis of inactivation of PRD-1 for the batch
bacteriophage experiment with the 254 nm UV lamp









The samples labeled "t = 7" were the batch conditions of one hour of continuous

UV irradiation. The samples labeled "t = 9" were the batch conditions for two hours of

continuous UV irradiation. As with the adsorption/attachment data, the sample replicates

for the PRD-1 phage are not significantly different. Each of the replicates in Figure 17

was compared using a hypothesis t-test for Poisson counts. The t-values calculated for

each of the pairs of replicates were 0.15, 2.36, 2.20, and 1.40, which are less than the

critical t-value of 2.58 (p = 0.01), indicating that the replicate counts are not significantly

different. The log [No/N] values for PRD-1 resulting from UV irradiation for one and two

hours were 0.78 and 1.43, respectively. The log [No/N] values for the three bacteriophage

experiments are shown in Table 15. The bacteriophage concentration data and statistical

values for each of the three experiments can be found in Appendix F.

Table 15. Log [No/N] inactivation of bacteriophage under UV radiation for three different
experimental conditions
OX-174 PRD-1 MS-2
Irradiation Time 1 hour 2 hours 1 hour 2 hours 1 hour 2 hours
365nm lamp 0.72 0.70 0.10 0.28 0.16 0.95
254nm lamp 0.85 1.67 0.78 1.43 0.85 1.65
254nm lamp (no pellets) 0.65 2.58 0.31 1.98 0.57 1.92

The 254-nm lamp was superior in inactivation of the phages as compared to the

365-nm lamp for both the one- and two-hour batch conditions. The phenomenon was

expected since the peak wavelength that microorganisms absorb UV light is about 265

nm.

In the 254-nm experiment without pellets, by increasing the time the phages were

exposed to UV radiation, the inactivation increased. This could be attributed to the

increased UV dose or the "concentration" (i.e., UV intensity) multiplied by the time. This









is analogous to the CT measured in chlorine disinfection. The UV intensity remained the

same for the two experiments, but the time was increased.

In the 254-nm experiment with pellets, by increasing the time the phages were

exposed to UV radiation, the inactivation increased. This could also be attributed to the

increased UV dose and inactivation by direct exposure to UV light.

In comparing the one-hour inactivation between 254-nm experiments with and

without pellets, the inactivation increased when the pellets were present for all three

phages. The hydroxyl radicals formed aided in the inactivation but only for the shorter

time period. Since the hydroxyl radical reactions for inactivation occur almost

instantaneously, the increased time (to two hours) did not have an impact on inactivation

by hydroxyl radicals. So, for the two-hour experiment with pellets present, the

inactivation was increased from the one-hour experiment with pellets because of the

direct exposure to UV light and not the hydroxyl radical interactions.

In comparing the two-hour inactivation between 254-nm experiments with and

without pellets, the inactivation decreased when the pellets were present for all three

phages. The pellets would aid in the inactivation of the phages to a point, but they may

also block some of the direct UV light, which would decrease the UV intensity seen by

the phages. Ultimately, the results seem to show that the blockage of UV light is out

competing the aiding of inactivation by the hydroxyl radicals.

Three experiments were also performed with selected bacteria using the batch

procedure, including two experiments with the 254-nm wavelength UV lamp and one

without the silica-titania composites (using the 254-nm lamp). Table 16 presents the

effluent pH values for each of the batch bacteria experiments.









Table 16. Effluent pH values for solutions containing sodium bicarbonate (100 mg/L as
CaCO3) along with spiked bacteria
Experiment ID Effluent pH (t = 0) Effluent pH (t = 9)
254nm lamp 7.07 6.48
254nm lamp 7.09 6.46
254nm lamp (no pellets) 8.11 7.84

As with the bacteriophages, the solution was held at a nominal pH by the addition

of 100 mg/L as CaCO3. The log [No/N] values for the three bacteria experiments are

shown in Table 17. The bacteria concentration data for each of the three experiments can

be found in Appendix F.

Table 17. Log [No/N] inactivation of bacteria under UV radiation for different
experimental conditions
E. coli S. aureus P. aeruginosa
Irradiation Time 1 hour 2 hours 1 hour 2 hours 1 hour 2 hours
254nm lamp 0.89 1.00 1.12 1.40 1.04 1.44
254nm lamp 1.41 1.7 1.11 1.62
254nm lamp (no pellets) 1.97 2.53 2.04 1.7 1.24

For the second experiment with the 254-nm lamp, the Staphylococcus aureus was

not present in the initial stock solution. All of the bacterial results, except for the

Pseudomonas aeruginosa 2-hour batch condition, had an increased inactivation when the

silica-titania composites were absent. This result can be attributed to the silica pellets

decreasing the UV irradiation of the microorganisms by blocking the UV radiation. This

blockage is counteracting the inactivation of the bacteria by the hydroxyl radicals.















CHAPTER 5
SUMMARY AND CONCLUSION

Summary

An annular reactor containing silica-titania composites arranged in a packed bed

was capable of degrading all six target analytes, including acetone, carbon disulfide,

chlorobenzene, ethyl acetate, methyl methacrylate, and toluene, in the presence of

bicarbonate ions and in a low (2.6 mg/L) dissolved oxygen environment. Similar

degradation results were obtained for alkalinity values of 80 mg/L NaHCO3 (48 mg/L as

CaCO3) and 200 mg/L NaHCO3 (120 mg/L as CaCO3) and solutions without alkalinity

present as well as solutions in a low dissolved oxygen environment of less than 3 mg/L.

At higher alkalinity (2360 mg/L as CaCO3) the degradation of contaminants decreased by

between 15-20% for toluene, chlorobenzene, and carbon disulfide.

A vertical annular reactor also containing the silica-titania composites arranged in

a packed bed was capable of achieving log inactivation [No/N] values of 1.67, 1.43, and

1.65 for bacteriophages OX-174, PRD-1, and MS-2, respectively, after two hours of

irradiation with a 254-nm UV lamp. Log inactivation [No/N] values for the bacteria

Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa of 1.70, 1.40, and

1.62, respectively, were achieved under similar conditions.









Conclusion

The amount of alkalinity expected to enter the finishing process in the ALS

treatment train, according to a revised NASA wastewater formulation, has a nominal

value of 58 mg/L as CaCO3 and worst-case value of 75 mg/L as CaCO3 (Verostko et al.,

2004). At these lower levels, the addition of bicarbonate ions does not have an

appreciable effect on the photocatalytic oxidation of acetone, carbon disulfide,

chlorobenzene, ethyl acetate, methyl methacrylate, and toluene in the reactor using

titania-doped porous silica pellets. With higher alkalinities, representative of the raw

NASA wastewater of around 2360 mg/L as CaCO3, a decrease in the percent of the

contaminants removed can be expected. The decreased removal can be attributed to the

scavenging of hydroxyl radicals by the abundance of bicarbonate ions in the solution.

A dissolved oxygen concentration in the range of 1 3 mg/L does not decrease

the potential for photocatalytic oxidation for most of the researched organic contaminants

as compared to dissolved oxygen levels near saturation (about 8 mg/L). Toluene and

methyl methacrylate showed a decrease in removal for the lower dissolved oxygen

concentration.

The levels of virus and bacteria inactivation achieved in the reactor are

insufficient for current drinking water standards, which recommend a minimum of 4-log

and 3-log inactivation levels for viruses and bacteria, respectively. In order to increase

the inactivation of the microbial species investigated, higher UV doses and therefore,

intensities would be required.

More microbial inactivation tests need to be performed to fully understand the

relationship between UV intensities, time, and the role of the photocatalyst in the






70


inactivation of both bacteria and bacteriophages. Other continued research could be

performed by combining the organic and microbial contaminants along with the

simulated wastewater to evaluate the performance of the reactors with competition

between the contaminants.















APPENDIX A
DATA FOR ADSORPTION OF ORGANIC COMPOUNDS














TEST 1

0 mg/L NaHCO3


Concentration %
([tg/L) C/Co Removal
t = 0 T = 400 (%) (%)


Toluene 41.0 25.5 62.1 37.9
Chlorobenzene 47.4 23.3 49.1 50.9
Acetone 324.4 196.9 60.7 39.3
Carbon Disulfide 46.8 17.2 36.7 63.3
Ethyl Acetate 45.2 31.1 68.9 31.1
Methyl Methacrylate 5.9 1.8 31.2 68.8


TEST 2
0 mg/L NaHCO3


Concentration (ptg/L) C/Co %
t = 0 ___t = 400 (Port 1) Removal


Vial 1 Vial 2 Avg


Std Dev


Vial 1 Vial 2 Avg


Std
Dev


(%)


Toluene 403.9 331.8 367.9 51.0 251.3 219.9 235.6 22.2 64.0 36.0
Chlorobenzene 626.5 505.6 566.0 85.5 302.7 320.5 311.6 12.6 55.0 45.0
Acetone 300.1 248.6 274.3 36.4 306.6 220.1 263.4 61.1 96.0 4.0
1014.
1713.0 1606.6 1659.8 75.3 1272.6 757.1 9 364.5 61.1 38.9
Ethyl Acetate 149.1 135.3 142.2 9.7 118.6 92.1 105.4 18.7 74.1 25.9
Methyl Methacrylate 13.2 15.1 14.1 1.3 5.9 4.8 5.4 0.8 37.9 62.1

TEST 3 Concentration (Lg/L) C/Co %
80 mg/L NaHCO3 t = 0 t = 400 (Port 1) Removal
Std
Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2 Avg Dev (%) (%)
Toluene 66.9 61.8 64.4 3.6 30.0 26.6 28.3 2.4 44.0 56.0
Chlorobenzene 71.3 72.5 71.9 0.8 24.4 21.5 23.0 2.1 31.9 68.1
Acetone 117.9 111.1 114.5 4.8 109.9 102.6 106.3 5.2 92.8 7.2
Carbon Disulfide 220.9 212.4 216.7 6.0 84.1 73.4 78.8 7.6 36.3 63.7
Ethyl Acetate 129.9 138.3 134.1 5.9 52.4 45.9 49.2 4.6 36.7 63.3
Methyl Methacrylate 123.5 112.8 118.2 7.6 62.6 57.8 60.2 3.4 51.0 49.0














TEST 4
200 mg/L NaHCO3


Concentration (tg/L) C/Co
t = 0 t = 400 (Port 1) Removal


Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2 Avg


Std
Dev


(%)


Toluene 132.6 98.4 115.5 24.2 70.8 70.8 61.3 38.7
Chlorobenzene 173.9 150.1 162.0 16.8 73.9 73.9 45.6 54.4
Acetone 211.6 345.2 278.4 94.5 222.7 222.7 80.0 20.0
Carbon Disulfide 721.3 511.7 616.5 148.2 294.7 294.7 47.8 52.2
Ethyl Acetate 67.3 94.6 81.0 19.3 49.6 49.6 61.3 38.7
Methyl Methacrylate 16.1 62.7 39.4 33.0 0.4 0.4 1.0 99.0


TEST 5
3965 mg/L
NaHCOs


Concentration (gg/L) %
C/Co Removal
t = 0 t = 400 (Port 1)
Vial 1 1 Vial 2 Avg Std Dev Vial 1 | Vial 2 Avg Std Dev (%) (%)


Toluene 114.6 112.8 113.7 1.3 103.1 90.7 96.9 8.8 85.2 14.8
Chlorobenzene 140.8 139.0 139.9 1.3 125.0 112.9 118.9 8.6 85.0 15.0
Acetone 397.8 379.9 388.9 12.7 366.4 376.9 371.6 7.4 95.6 4.4
Carbon Disulfide 462.2 445.2 453.7 12.0 390.8 349.9 370.3 28.9 81.6 18.4
Ethyl Acetate 300.3 251.9 276.1 34.2 107.8 53.6 80.7 38.3 29.2 70.8
Methyl Methacrylate 2040.3 1987.9 2014.1 37.0 1352.7 925.8 1139.2 301.8 56.6 43.4

TEST 6 Concentration (gg/L) C/Co %
Low DO Run t = 0 t = 400 (Port 1) Removal
Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2 Avg Std Dev (%) (%)
Toluene 188.5 220.4 204.5 22.6 171.5 183.0 177.2 8.2 86.7 13.3
Chlorobenzene 125.7 145.2 135.4 13.8 100.4 101.8 101.1 0.9 74.7 25.3
Acetone 184.1 222.6 203.3 27.2 162.6 167.1 164.8 3.2 81.1 18.9
Carbon Disulfide 232.0 260.1 246.1 19.9 155.5 162.3 158.9 4.8 64.6 35.4
EthylAcetate 176.4 221.6 199.0 32.0 137.4 144.6 141.0 5.1 70.9 29.1
Methyl Methacrylate 627.5 708.5 668.0 57.3 580.3 616.3 598.3 25.5 89.6 10.4
















Adsorption / Volatilization 5 hrs in Circulation Mode


* Test 1- No NaHCO3
E 200 mg/L NaHCO3

100.0
90.0
80.0
70.0
60.0
50.0 -
40.0 -
30.0 -
20.0 -
10.0 -
0.0


* Test 2- No NaHCO3
* 3965 mg/L NaHCO3


E 80 mg/L NaHCO3
* Low DO (1.5 mg/L)


Toluene Chlorobenzene Acetone Carbon Ethyl Acetate Methyl
Disulfide Methacrylate


Figure 18. Percent removal of organic compounds during recirculation in the reactor without UV light
















APPENDIX B
DATA FOR OXIDATION OF ORGANIC COMPOUNDS














TEST 1
0 mg/L NaHCO3


Concentration (tg/L) C/Co R
Removed


Port 1


Port 2


Toluene 25.5 1.5 5.9 94.1
Chlorobenzene 105.9 23.3 22.0 78.0
Acetone 196.9 115.7 58.8 41.2
Carbon Disulfide 98.5 17.2 17.5 82.5
Ethyl Acetate 51.1 31.1 60.9 39.1
Methyl Methacrylate 7.5 1.8 24.0


TEST 2


0 mg/L NaHCO3


Concentration (|tg/L) C/Co
Removal


Port 1
V1


Port 1 -
V2


AVG


STD DEV


Port 2 -
V1


Port 2 -


AVG


STD
DEV


Toluene 251.3 219.9 235.6 22.2 2.9 2.3 2.6 0.4 1.1 98.9
Chlorobenzene 302.7 320.5 311.6 12.6 5.3 3.8 4.6 1.1 1.5 98.5
Acetone 306.6 220.1 263.4 61.2 193.5 96.2 144.9 68.8 55.0 45.0
Carbon Disulfide 1272.6 757.1 1014.9 364.5 27.9 20.5 24.2 5.2 2.4 97.6
Ethyl Acetate 118.6 92.1 105.4 18.7 29.5 21.3 25.4 5.8 24.1 75.9
Methyl Methacrylate 5.9 4.8 5.4 0.8 7.0 4.1 5.6 2.1 103.7



TEST 3 Concentration (tg/L) C/Co
Removal
Port 1 Port 1 Port 2 Port 2 STD AV %
80 mg/L NaHCO3 V1 V2 AVG STDDEV V1 V2 AVG DEV
Toluene 30.0 26.6 28.3 2.4 1.9 1.7 1.8 0.1 6.4 93.6
Chlorobenzene 24.4 21.5 23.0 2.1 3.0 3.0 3.0 0.0 13.1 86.9
Acetone 109.9 102.6 106.3 5.2 53.7 75.6 64.7 15.5 60.8 39.2
Carbon Disulfide 84.1 73.4 78.8 7.6 10.5 11.7 11.1 0.8 14.1 85.9
Ethyl Acetate 52.4 45.9 49.2 4.6 10.0 6.2 8.1 2.7 16.5 83.5
Methyl Methacrylate 62.6 57.8 60.2 3.4 2.1 2.7 2.4 0.4 4.0 96.0














TEST 4


200 ma/L NaHCO3


Concentration (|tg/L) C/Co Removal
SRemoval


Port 1 V1


Port 1 V2


AVG


STD
DEV


Port 2 V1


Port 2 V2


AVG


STD
DEV


Toluene 70.8 70.8 2.0 4.6 3.3 1.8 4.7 95.3
Chlorobenzene 73.9 73.9 2.7 4.8 3.8 1.5 5.1 94.9
Acetone 222.7 222.7 125.1 222.4 173.8 68.8 78.0 22.0
Carbon Disulfide 294.7 294.7 17.7 23.7 20.7 4.2 7.0 93.0
Ethyl Acetate 49.6 49.6 10.2 20.3 15.3 7.1 30.7 69.3
Methyl Methacrylate 0.4 0.4 17.0 32.2 24.6 10.7 6150.0 -


TEST 5 Concentration (tg/L) C/Co
Removal
Port 1 Port 1 Port 2 Port 2- STD N N
3965 mg/L NaHCO3 V1 V2 AVG STDDEV V1 V2 AVG DEV
Toluene 103.1 90.7 96.9 8.8 23.7 28.7 26.2 3.5 27.0 73.0
Chlorobenzene 125.0 112.9 118.9 8.6 38.7 41.5 40.1 2.0 33.7 66.3
Acetone 366.4 376.9 371.6 7.4 372.6 91.3 231.9 198.9 62.4 37.6
Carbon Disulfide 390.8 349.9 370.3 28.9 81.8 98.0 89.9 11.5 24.3 75.7
Ethyl Acetate 107.8 53.6 80.7 38.3 0.3 0.0 0.1 0.2 0.2 99.8
Methyl Methacrylate 1352.7 925.8 1139.2 301.8 0.3 25.0 12.7 17.5 1.1 98.9


TEST 6 Concentration (tg/L) C/Co
Removal
Port 1 Port 1 Port 2 Port 2 STD N N
Low DO Run V1 V2 AVG STDDEV V1 V2 AVG DEV
Toluene 171.5 183.0 177.2 8.2 23.4 24.6 24.0 0.9 13.5 86.5
Chlorobenzene 100.4 101.8 101.1 0.9 1.5 15.6 8.6 10.0 8.5 91.5
Acetone 162.6 167.1 164.8 3.2 133.5 138.0 135.7 3.2 82.3 17.7
Carbon Disulfide 155.5 162.3 158.9 4.8 29.4 29.6 29.5 0.1 18.6 81.4
Ethyl Acetate 137.4 144.6 141.0 5.1 33.8 38.9 36.3 3.6 25.8 74.2
Methyl Methacrylate 580.3 616.3 598.3 25.5 91.8 97.1 94.5 3.7 15.8 84.2














HOLMES (2003) DATA


Test 1
0 mq/L NaHCO3


%
Concentration (pg/L) C/Co Removal


Port 1


Port 2


Toluene 53.4 12.1 22.6 77.4
Chlorobenzene 54.6 11.1 20.3 79.7
Acetone -
Carbon Disulfide 31.9 2.4 7.7 92.3
Ethyl Acetate 33.2 20.6 62.0 38.0
Methyl Methacrylate 103.3 42.9 41.6 58.4


Oxidation (Single Pass at EBCT = 43.6 nin)


* Test 1- No NaHCO3
* 200 mg/L NaHCO3


* Test 2- No NaHCO3
* 3965 mg/L NaHCO3


Toluene


Carbon
Disulfide


0 80 mg/L NaHCO3
* Low DO (2.6 mg/L)
---------- 00


Methyl
Methacrylate


Figure 19. Percent removal of organic compounds during single pass through the reactor with UV irradiation


100.0
90.0
80.0
70.0
60.0
50.0
40.0
30.0
20.0
10.0
0.0






















APPENDIX C
ANALYSIS OF VARIANCE TABLES FOR ORGANIC COMPOUNDS



TOLUENE

One-way ANOVA: Percent Removal versus Experiment


Analysis of Variance for Percent
Source DF SS M:


Experiment
Error
Total 1(


Level I
200 mg/L
3965 mg/L
80 mg/L
Low DO
None

Pooled StDev


867.38
59.26
926.65


Mean
95.34
72.69
93.64
86.46
97.31


216.85
9.88



StDev
2.60
6.10
0.04
0.13
2.76


F
21.95


P
0.001


Individual 95% CIs For Mean
Based on Pooled StDev
---+---------+---------+---


---+
3.14 70


Fisher's pairwise comparisons

Family error rate = 0.222
Individual error rate = 0.0500

Critical value = 2.448

Intervals for (column level mean)


(row level mean)


200 mg/L 3965 mg/L 80 mg/L

3965 mg/L 14.951
30.339


80 mg/L


Low DO


None


-5.994
9.394

1.186
16.574

-8.990
5.057


-28.639
-13.251

-21.459
-6.071

-31.635
-17.588


-0.514
14.874

-10.690
3.357


Low DO


-17.870
-3.823


*Intervals indicate significant difference when zero is not inside the interval
Example: The interval for 3965 mg/L and 200 mg/L is 14.951 30.339. This does not
contain zero, so the there is a significant difference for toluene between the 3965 mg/L
experiment and the 200 mg/L experiment.


------+-
-*----)


(--*---)








80



CHLOROBENZENE


One-way ANOVA: Percent Removal versus Experiment


Analysis of Variance for
Source DF SS
Experiment 4 1101.5
Error 6 400.0
Total 10 1501.4


Percent
MS
275.4
66.7


4.13 0.061


Individual 95% CIs For Mean
Based on Pooled StDev


Level
200 mg/L
3965 mg/L
80 mg/L
Low DO
None


Pooled StDev


Mean
94.93
66.16
86.88
91.59
91.69

8.16


StDev
2.02
4.12
1.17
9.82
11.86


Fisher's pairwise comparisons

Family error rate = 0.222
Individual error rate = 0.0500

Critical value = 2.448

Intervals for (column level mean)


3965 mg/L


80 mg/L


Low DO


None


(row level mean)


200 mg/L 3965 mg/L 80 mg/L

8.78
48.75


11.94
28.04

16.65
23.33

15.01
21.48


40.70
-0.73

45.41
-5.44

43.77
-7.28


-24.70
15.28

-23.06
13.43


Low DO


-18.35
18.14








81



ACETONE


One-way ANOVA: Percent Removal versus Experiment


Analysis of Variance for


Source D]
Experiment
Error
Total 1(


Level I
200 mg/L
3965 mg/L
80 mg/L
Low DO
None

Pooled StDev


SS
1229
4471
5700


Mean
21.98
37.03
38.73
17.68


Percent
MS
307
745


F
0.41


P
0. 795


Individual 95% CIs For Mean
Based on Pooled StDev


StDev
30.90
54.78
17.55
0.35


3 44.81 10.18

27.30


Fisher's pairwise comparisons

Family error rate = 0.222
Individual error rate = 0.0500


Critical value = 2.448


Intervals for (column level mean)


3965 mg/L


80 mg/L


Low DO


None


(row level mean)


200 mg/L 3965 mg/L 80 mg/L

-81.9
51.8


-83.6
50.1

-62.5
71.1

-83.8
38.2


-68.5
65.1

-47.5
86.2

-68.8
53.2


-45.8
87.9

-67.1
54.9


Low DO


88.1
33.9








82



CARBON DISULFIDE


One-way ANOVA: Percent Removal versus Experiment


Analysis of Variance for


Source D]
Experiment
Error
Total 1(


Level I
200 mg/L
3965 mg/L
80 mg/L
Low DO
None

Pooled StDev


SS
486.2
183.7
670.0


Mean
92.98
75.53
85.79
81.44
92.55

5.53


Percent
MS
121.6
30.6


3.97 0.066


Individual 95% CIs For Mean
Based on Pooled StDev


StDev
1.44
5.01
2.44
0.50
8.67


*


90 100


Fisher's pairwise comparisons

Family error rate = 0.222
Individual error rate = 0.0500

Critical value = 2.448

Intervals for (column level mean)


(row level mean)


200 mg/L 3965 mg/L 80 mg/L

3.903
30.997


-6.357
20.737

-2.007
25.087

11.938
12.795


23.807
3.287

19.457
7.637


-9.197
17.897


29.388 -19.128
-4.655 5.605


Low DO


3965 mg/L


80 mg/L


Low DO


None


-23.478
1.255








83



ETHYL ACETATE


One-way ANOVA: Percent Removal versus Experiment


Analysis of Variance for


Source D]
Experiment
Error
Total 1(


Level I
200 mg/L
3965 mg/L
80 mg/L
Low DO
None

Pooled StDev


SS
1800
1133
2933


Mean
69.26
99.87
83.71
74.26
63.71

13.74


Percent
MS
450
189


F
2.38


P
0.164


Individual 95% CIs For Mean
Based on Pooled StDev


StDev
14.40
0.19
3.94
1.63
21.30


-*----)


75 100


Fisher's pairwise comparisons


Family error rate =
Individual error rate =

Critical value = 2.448


Intervals for (column level mean)


3965 mg/L


80 mg/L


Low DO


None


(row level mean)


200 mg/L 3965 mg/L 80 mg/L

-64.25
3.03


-48.09
19.19

-38.64
28.63

-25.17
36.25


-17.48
49.80

-8.03
59.24

5.44
66.86


-24.19
43.08

-10.72
50.70


0.222
0.0500


Low DO


-20.16
41.25








84



METHYL METHACRYLATE


One-way ANOVA: Percent Removal

Analysis of Variance for Percent
Source DF SS MS
Experiment 2 236.01 118.01
Error 3 4.46 1.49
Total 5 240.47


Level N Mean StI
3965 mg/L 2 98.64 1.
80 mg/L 2 95.99 0,
Low DO 2 84.21 0.

Pooled StDev = 1.22

Fisher's pairwise comparisons

Family error rate = 0.0983
Individual error rate = 0.0500

Critical value = 3.182

Intervals for (column level me;

3965 mg/L 80 mg/

80 mg/L -1.232
6.532


Dev
.90
.93
.04


an)

/L


versus Experiment


F
79.30


P
0.003


Individual 95% CIs For Mean
Based on Pooled StDev
-----+---------+---------+-----
(___*
(---*--
(---*--4 0)
-----------------------84.0 90.0 96.0-----
84.0 90.0 96.0


(row level mean)


10.548
18.312


102.0
102.0


Low DO


7.898
15.662





















APPENDIX D
TEST FOR EQUAL VARIANCE FOR ORGANIC COMPOUNDS


TOLUENE
Test for Equal Variances


Response
Factors
ConfLvl


Percent Removal
Group
95.0000


Bonferroni confidence intervals for standard deviations

Lower Sigma Upper N Factor Levels


0.92701
2.17394
0.01511
0.04534
1.19914


2.60215
6.10233
0.04243
0.12728
2.76019


415.241
973.785
6.770
20.311
38.986


200 mg/L
3965 mg/L
80 mg/L
Low DO
None


Bartlett's Test (normal distribution)
Test Statistic: 10.721
P-Value : 0.030


Levene's Test (any continuous
Test Statistic: 2.500
P-Value : 0.152



Test for Equal Variances


Response
Factors
ConfLvl


distribution)


CHLOROBENZENE:


Percent Removal
Experiment
95.0000


Bonferroni confidence intervals for standard deviations

Lower Sigma Upper N Factor Levels


0.71793
1.46609
0.41564
3.49897
5.15087


2.0153
4.1154
1.1667
9.8217
11.8563


321.59
656.71
186.18
1567.31
167.46


200 mg/L
3965 mg/L
80 mg/L
Low DO
None


Bartlett's Test (normal distribution)
Test Statistic: 4.553
P-Value : 0.336

Levene's Test (any continuous distribution)
Test Statistic: 0.445
P-Value : 0.774








86



ACETONE


Test for Equal Variances


Response
Factors
ConfLvl


Percent Removal
Experiment
95.0000


Bonferroni confidence intervals for standard deviations

Lower Sigma Upper N Factor Levels


11.0083
19.5151
6.2523
0.1234
4.4226


30.9006
54.7796
17.5504
0.3465
10.1799


4930.98
8741.50
2800.62
55.29
143.79


200 mg/L
3965 mg/L
80 mg/L
Low DO
None


Bartlett's Test (normal distribution)
Test Statistic: 8.759
P-Value : 0.067


Levene's Test (any continuous distribution)
Test Statistic: 23.603
P-Value : 0.001



CARBON DISULFIDE


Test for Equal Variances


Response
Factors
ConfLvl


Percent Removal
Experiment
95.0000


Bonferroni confidence intervals for standard deviations

Lower Sigma Upper N Factor Levels


0.51137
1.78601
0.86907
0.17885
3.76657


1.43543
5.01339
2.43952
0.50205
8.66993


229.059
800.016
389.288
80.114
122.458


200 mg/L
3965 mg/L
80 mg/L
Low DO
None


Bartlett's Test (normal distribution)
Test Statistic: 5.538
P-Value : 0.236

Levene's Test (any continuous distribution)
Test Statistic: 0.396
P-Value : 0.805








87



ETHYL ACETATE


Test for Equal Variances


Response
Factors
ConfLvl


Percent Removal
Experiment
95.0000


Bonferroni confidence intervals for standard deviations

Lower Sigma Upper N Factor Levels


5.13131
0.06801
1.40311
0.57938
9.25312


14.4038
0.1909
3.9386
1.6263
21.2989


2298.49
30.47
628.50
259.53
300.83


200 mg/L
3965 mg/L
80 mg/L
Low DO
None


Bartlett's Test
Test Statistic:
P-Value


(normal distribution)
9.898
0.042


Levene's Test (any continuous distribution)
Test Statistic: 0.549
P-Value : 0.708



METHYL METHACRYLATE


Test for Equal Variances


Response
Factors
ConfLvl


Percent Removal
Experiment
95.0000


Bonferroni confidence intervals for standard deviations

Lower Sigma Upper N Factor Levels


0.718295
0.353787
0.016081


1.89505
0.93338
0.04243


181.440
89.366
4.062


3965 mg/L
80 mg/L
Low DO


Bartlett's Test (normal distribution)
Test Statistic: 4.411
P-Value : 0.110


















APPENDIX E
DATA FOR OXIDATION OF TOC IN SIMULATED NASA WASTEWATER

Single pass adsorption and destruction experiment
EBCT = 43 min Port 1 pH (6.67 hr) = 7.00
pH initial = 8.37 Port 2 pH (6.67 hr) = 5.4


Average St Dev RSD
Rep # (mg/L) (mg/L) (mg/L) (%)


Average
(mg/L)


Port 2 1 1.5687
Vial 1 2 1.5387 1.5340 0.0373 2.4297857
3 1.4946 1.4452
Port 2 1 1.3875
Vial 2 2 1.354 1.3565 0.0299 2.2025213
3 1.3279
Port 1 1 2.3999
Vial 1 2 2.3287 2.3735 0.0390 1.6442665
3 2.392 2.3613
Port 1 1 2.3463
Vial 2 2 2.3422 2.3491 0.0087 0.3704798
3 2.3589
Initial 1 2.7407
Vial 1 2 2.67 2.7089 0.0359 1.3242042
3 2.7159 2.7283
Initial 1 2.7635
Vial 2 2 2.7212 2.7478 0.0232 0.842891
3 2.7587


5ppm
Standard


1
2
3


4.4555
4.3687
4.4253


4.4165


0.0441


0.9977141


Mass of TOC Adsorbed:
M = 3L (2.7283 mg/L 2.3613 mg/L) = 1.10 mg

Mass of TOC Removed in Oxidative Step:
M = 3L (2.3613 mg/L 1.4452 mg/L) = 2.75 mg


C/Co


% Removed


(%) (%)
Adsorption 86.55 13.45
Destruction 61.20 38.80
Total 52.97 47.03







89





Desorption of wastewater-loaded pellets using nanopure water
Single pass at 10 mL/min (EBCT = 43 min)
Average St Dev RSD
Rep # (mg/L) (mg/L) (mg/L) (%)

t = 71 1 0.044
S= 1 1 0.0 0.1470 0.0024 1.63
430mL passed 2 0.1494
3 0.1446
t = 51 1 0.3139
230mL passed 2 0.2946 0.3053 0.0098 3.22
3 0.3074
t = 31 1 1.6251
30mL passed 2 1.5534 1.5595 0.0628 4.03
3 1.4999
5ppm 1 4.6763
Standard 2 4.502 4.5762 0.0900 1.97
3 4.5502

Desorption Polynomial Trend


Volume
passed Conc.
(L) (mg/L)
0.43 0.1470
0.23 0.3053
0.03 1.5595
* omits Rep #1 oft = 71


R- =I





0 01 02 03 04
Volume passed (L)


Using polynomial trendline:
Between 0.03 and 0.43 Mass desorbed = 0 .193 mg
Between 0 and 0.43 Mass desorbed = 0.2438 mg
Between 0 and 0.23 Mass desorbed = 0.2316 mg

Desorption- Exponential Trend Y = 0.7075e-3 6543x
R2 1
R^-
0 1.50-

S ? 1.00
a Eo
o 0.50

0.00
0 0.1 0.2 0.3 0.4 0.5
Volume passed (L)

Using exponential trendline:
Between 0.23 and 0.43 Mass desorbed = 0.03865 mg
Between 0.23 and infinity Mass desorbed = 0.06125 mg
Total Desorbed: 0.2316 mg + 0.06125 mg = 0.29285 mg
















APPENDIX F
DATA FOR INACTIVATION OF MICROORGANISMS




Full Text

PAGE 1

PHOTOCATALYTIC OXIDATION OF SE LECTED ORGANIC CONTAMINANTS AND INACTIVATION OF MICROORGANISMS IN A CONTINUOUS FLOW REACTOR PACKED WITH TITANIA-DOPED SILICA By MARY JOANNE GARTON A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2005

PAGE 2

Copyright 2005 by Mary Joanne Garton

PAGE 3

iii ACKNOWLEDGEMENTS I would like to thank my advisor, Dr Paul Chadik of the Department of Environmental Engineering Scie nces at the University of Florida, for giving me the opportunity to work for him and for hi s guidance and assistance throughout my undergraduate and graduate education. I would also like to thank th e other members of my committee, Dr. David Mazyck and Dr. A ngela Lindner, for their assistance and suggestions throughout this research. I would like to acknowledge Dr. Samuel Farrah of the Department of Microbiology and Cell Sc iences, who provided much guidance with the microbial aspect of this research and allowed me to use his laboratory facilities. I would like to acknowledge Dr. Matt hew Booth of the Department of Environmental Engineering Sc iences, who assisted in th e gas chromatography. I also would like to thank Mickal Witwer, an e nvironmental engineering doctoral candidate, and Shannon McQuaig, a microbiolo gy graduate student at the University of Florida, who provided limitless assistance in the experime ntation and analyses conducted during the course of this research. Additionally, I would like to thank my fa mily for all their love and support. Special thanks go to my fianc, Dave Frie dman, for his constant encouragement and support throughout my education.

PAGE 4

iv TABLE OF CONTENTS Page ACKNOWLEDGEMENTS...............................................................................................iii LIST OF TABLES.............................................................................................................vi LIST OF FIGURES..........................................................................................................vii ABSTRACT....................................................................................................................... ix CHAPTER 1 INTRODUCTION......................................................................................................1 2 LITERATURE REVIEW...........................................................................................3 Photocatalysis.............................................................................................................3 Titanium Dioxide........................................................................................................6 Catalyst Supports........................................................................................................7 Silica – Titania Composites........................................................................................8 Kinetics.....................................................................................................................11 Microorganisms........................................................................................................12 UV Effect on Microorganisms..................................................................................16 3 MATERIALS AND METHODS..............................................................................22 Silica-Titania Composites.........................................................................................22 Reactor Design..........................................................................................................24 Reactor for Oxidation of Organic Compounds.................................................24 Reactor for Inactivation of Bacteria and Viruses..............................................29 Analytical Methods...................................................................................................31 Analysis of Organic Compounds......................................................................31 Analysis of Bacteria and Viruses......................................................................33 Experimental Procedures..........................................................................................34 Oxidation of Organic Compounds....................................................................34 Inactivation of Bacteria and Viruses.................................................................41 Statistics....................................................................................................................4 6 4 RESULTS................................................................................................................. 48 Oxidation of Organic Compounds............................................................................48

PAGE 5

v Adsorption........................................................................................................48 Oxidation..........................................................................................................49 Inactivation of Bacteria and Viruses.........................................................................58 Recirculation Experiments................................................................................58 Batch Experiments............................................................................................60 5 SUMMARY AND CONCLUSION.........................................................................68 Summary...................................................................................................................68 Conclusion................................................................................................................69 APPENDIX A DATA FOR ADSORPTION OF ORGANIC COMPOUNDS................................71 B DATA FOR OXIDATI ON OF ORGANIC COMPOUNDS...................................75 C ANALYSIS OF VARIANCE TABLES FOR ORGANIC COMPOUNDS............79 D TEST FOR EQUAL VAR IANCE FOR ORGANIC COMPOUNDS.....................85 E DATA FOR OXIDATION OF TOC IN SIMULATED NASA WASTEWATER.88 F DATA FOR INACTIVAT ION OF MICROORGANISMS.....................................90 G FIGURES OF DATA FOR INACTIVATION OF MICROORGANISMS..........108 LIST OF REFERENCES................................................................................................115 BIOGRAPHICAL SKETCH..........................................................................................120

PAGE 6

vi LIST OF TABLES Table page 1 Comparison of photocatalytic inactivation of microorganisms................................20 2 Properties of Degussa P25 TiO2...............................................................................23 3 Key statistics from the tracer analysis......................................................................28 4 Average RSDs of SOC analysis................................................................................33 5 pH values for experiment with 80 mg/L HCO3 solution..........................................37 6 pH values for experiment with 200 mg/L HCO3 solution........................................37 7 pH values for experiment with 3965 mg/L HCO3 solution......................................38 8 Simulated wastewater composition...........................................................................40 9 Effluent pH values for solutions containing sodium bicarbonate along with spiked bacteriophages...........................................................................................................46 10 Effluent pH values for solutions c ontaining sodium bicarbonate along with spiked bacteria......................................................................................................................4 6 11 Removal of toluene and chlorobenzene in attaining adsorption equilibrium in the reactor.......................................................................................................................4 9 12 pH values for spiked organic contam inant solutions containing differing amounts of sodium bicarbonate...................................................................................................51 13 Percent removal of total organic carbon...................................................................56 14 Effluent pH values for solutions c ontaining sodium bicarbonate along with spiked bacteriophages...........................................................................................................64 15 Log [N0/N] inactivation of bacteriophage under UV radiation................................65 16 Effluent pH values for solutions c ontaining sodium bicarbonate along with spiked bacteria......................................................................................................................6 7 17 Log [N0/N] inactivation of bacteria under UV radiation..........................................67

PAGE 7

vii LIST OF FIGURES Figure page 1 Generation of primary radica ls at the surface of irradiated TiO2 particles in water...3 2 Linkages of SiO2 tetrahedras....................................................................................10 3 Silica-titania composite suspension be ing transferred to the 96-well assay plates...23 4 Organic reactor filled with silica-titani a composites................................................25 5 Diagram of system setup used in r ecirculating and singl e-pass conditions..............26 6 E-curve generated from the tracer analysis performed on the reactor......................29 7 Vertical quartz reacto r packed with silicatitania composites..................................30 8 Treatment system consisting of reactor packed with SiO2-TiO2 composites, UV lamp, peristaltic pump, and rese rvoir of stock solution............................................31 9 Percent removal of spiked SOCs in a single pass through the reactor with UV irradiation and without addition of alkalinity...........................................................50 10 Percent removal of six target analytes in solution containing differing amounts of bicarbonate operating in singl e-pass mode after five hours of adsorption in the recirculation mode....................................................................................................52 11 Percent removal of six target analytes ope rating in single-pass mode after five hours of adsorption in the recirculation mode....................................................................55 12 Polynomial trend line of desorbing TOC c oncentration of simulated wastewater...57 13 Exponential trend line of desorbing TOC c oncentration of simulated wastewater..57 14 Bacteriophage adsorption and inactivation ex periment in a recirculation mode with a potassium phosphate buffer......................................................................................59 15 Duplicate sample analysis of adsorption/ attachment removal of PRD-1 for the batch bacteriophage experiment with no UV light.............................................................61 16 Duplicate sample analysis of adsorption/ attachment removal of MS-2 for the batch bacteriophage experiment with no UV light.............................................................62

PAGE 8

viii 17 Duplicate sample analysis of inactivation of PRD-1 for the batch bacteriophage experiment with the 254 nm UV lamp......................................................................64 18 Percent removal of organic compounds dur ing recirculation in the reactor without UV light....................................................................................................................74 19 Percent removal of organic compounds during single pass through the reactor with UV irradiation...........................................................................................................78

PAGE 9

ix Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering PHOTOCATALYTIC OXIDATION OF SE LECTED ORGANIC CONTAMINANTS AND INACTIVATION OF MICROORGANISMS IN A CONTINUOUS FLOW REACTOR PACKED WITH TITANIA-DOPED SILICA By Mary Joanne Garton August 2005 Chair: Paul Chadik Major Department: Environmental Engineering Sciences The NASA Advanced Life Support Syst em (ALS) will provide systems for recycling water in manned flight space missions A treatment process train is required in the ALS to convert wastewater into potabl e water on these missions. The reactor system described was designed to fill an ALS need for a final water treatment polishing process that will oxidize synthetic organic chemi cals (SOC) and inactivate pathogenic microorganisms that are likely to contaminate wastewater systems on these missions. The reactor does not use expendable oxidation chem icals and uses a catalyst in a fixed bed, eliminating the need to separate the catalyst from the water after treatment. Two annular continuous flow reactors with nominal volum es of 440 mL and 150 mL were packed with silica gel pellets that we re doped with titania (TiO2) (12 wt%). The reactors were configured with ultraviolet (UV) lamps in the center of the reactors. The flow characteristics of the reactors were determ ined with tracer analyses. The SOC oxidation

PAGE 10

x experiments were performed in a single-pass mode with bicarbonate ions present and in a low dissolved oxygen environment. Microbial experiments were performed for the inactivati on of selected viruses and bacteria. Cultures of the target microorganis ms were prepared in a stock solution, which was pumped through the reactor for a period of time to allow for microbial attachment to the reactor and catalyst pellets. The UV la mps were then activated and microbial concentrations were monitored. The log [N0/N] values resulting from two hours of 254nm UV irradiation for the bacteriophages X-174, PRD-1, and MS-2 were 1.67, 1.43, and 1.65, respectively. The log [N0/N] values resulting from two hours of 254-nm UV irradiation for the bacteria Escherichia coli, Staphylo coccus aureus, and Pseudomonas aeruginosa were 1.70, 1.40, and 1.62, respectively. The photocatalytic reactor system effec tively oxidized the six target organic contaminants: acetone, carbon disulfide, chlorobenzene, ethyl acetate, methyl methacrylate, and toluene. The percent removal of each contaminant was largely unaffected by the addition of bicarbonate ions in the range of 50 – 120 mg/L as CaCO3 and in a low dissolved oxygen environment between 1 – 3 mg/L.

PAGE 11

1 CHAPTER 1 INTRODUCTION Water recovery is an essential proces s needed for manned flight space missions. The ability to regenerate potable water redu ces waste storage, promotes the health and safety of crew members, and minimizes ma ss of required water resources. A reduction in costs would also be seen from reduced am ounts of mass, volume, and energy required for water transportation. The goal of this research is to further de velop a finishing process for the treatment of wastewater for NASAÂ’s Advanced Life Support System (ALS). The components of the wastewater will include shower water, wash water, urine, humidity condensate from machinery, and wastewater from the solid wast e processor. A treatme nt process train is required in the ALS to convert wastewater into potable water on the space missions. The post-processor would provide a final water tr eatment polishing process that will oxidize synthetic organic chemicals (SOCs) and inac tivate pathogenic microorganisms that are likely to contaminate wastewater systems on these missions and that may not have been completely removed in previous treatment processes. These SOCs, which will be generated on space missions and which will accumulate in recycled water if not removed, have adverse health consequences if ingested and must be controlled to part per billion levels as set by NASAÂ’s Requirements: Definition and design consideration (Lange and Lin, 1998). The goal of the reactor system is to inactivate the microorganisms and completely destroy the SOCs rather than mere ly change the phase of these contaminants.

PAGE 12

2 The reactor system will not require expendabl e oxidation chemicals, and will have a low energy and power demand. Photocatalysis or photocatal ytic oxidation was the mechanism investigated for the finishing process. Research has shown that photocatalysis is an effective process to destroy organic compounds, inorganic com pounds, and microorganisms. The research focused on six selected organic compounds as we ll as selected bacteria and viruses. The organic compounds were: acetone carbon disulfide chlorobenzene ethyl acetate methyl methacrylate toluene The target microorganisms were the bacteria Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa and the bacteriophages X-174, PRD-1, and MS-2. The research was conducted using two prev iously designed fl ow-through annular reactors (Holmes, 2003, and Ludwig, 2004) and silica-titania composite pellets (Londeree, 2002) as the photocatal yst in the reactor. By the exposure to ultraviolet (UV) light, the photocatalyst pelle ts catalyze the ox idation of the organic compounds and inactivate the bacteria and bacteriophages. The objectives for the current research were to investigate the effects of bicarbonate alkali nity on the oxidation of organic compounds, inactivate selected microorganisms, and invest igate the effects of a low dissolved oxygen concentration environment on the oxidation of organic compounds.

PAGE 13

3 CHAPTER 2 LITERATURE REVIEW Photocatalysis Photocatalysis is an oxidative proces s in which organic contaminants are degraded using ultraviolet (UV) light ener gy. Heterogeneous photoc atalysis involves the use of a solid photocatalyst in contact with either a gas or liquid, while homogeneous photocatalysis uses a catalyst of the same phase as the contaminated media. In this work, the media involved is a solid phase photocatalyst, tit anium dioxide (TiO2) and an aqueous solution containing the organi c and microbiological contamin ants. While UV light or an oxidant used alone may produce partial degr adation of a compound, the combined use of UV light with an oxidant (O3, H2O2, or O2 with a photocatalyst) ha s been shown to yield complete mineralization of organic carbon c ontaminants to carbon di oxide (Ollis et al., 1991). The removal also occurs at a faster rate than with the UV or oxidant alone. Figure 1 illustrates the radicals formed on the photocatalyst surface. Figure 1. Generation of primary radicals at the surface of irradiated TiO2 particles in water (Blake et al., 1991) TiO2 particle

PAGE 14

4 Radicals are formed when the photoc atalyst absorbs enough light of the appropriate wavelength for an electron (e-) to jump from the valence band (VB) to the conduction band (CB) forming a positive electron hole (h+). For titanium dioxide (TiO2), UV light of less than 388 nm is required. Th e following equations describe the excitation of an electron using TiO2 and the subsequent reactions: TiO2 + hv e+ h+ (1) e+ O2 O2 (2) h+ + M Mox (3) h+ + OHOH* (4) where O2 is the superoxide anion, OH* is the hydroxyl radical, and M and Mox are the organic compound and its oxidized product, respectively (Tanaka et al., 1991). Both the electron and electron hole are highly energetic and therefore highly reactive. Most organic photodegradation r eactions utilize the oxidizing power of the electron holes; however, to prevent a buildup of charge, a reducible species must be present to react with the electrons (Hoffmann et al., 1995). Since oxygen (O2) is usually available in aqueous solutions, it has been s hown to be a very important electron acceptor in the process. Oxygen keeps the electron a nd electron hole from recombining and thus leaving open the reaction sites for photocatal ysis to take place. Gerischer and Heller (1991) have demonstrated that with an in crease in oxygen concentration, there is a subsequent increase in the reaction rate of contaminant degradation. Water and hydroxide ions (OH-) react with the electr on holes to form hydroxyl radicals (OH*). The photocatalytic mineralizati on of an organic compound that is adsorbed to the catalyst surf ace begins with either reaction, with hydroxyl radicals or by

PAGE 15

5 direct oxidation with the holes on the surface of the catalyst. As illustrated in equations 5 – 6, the hydroxyl radical begins to degrade the adsorbed organic compound by either removing an available hydrogen atom to form water (equation 5) or adding itself to any unsaturated carbon bonds (equati on 6) (Grabner et al., 1991; Hoffman et al., 1995; Mao et al., 1991, 1992). OH* + R H2O + R(-H)* (5) OH* + R ROH* (6) where R is the organic compound, OH* is the hydr oxyl radical, R(-H)* is a radical with a hydrogen removed, and ROH* is a radical comb ined with the hydroxyl radical (Mao et al., 1991). Turchi and Ollis (1990) found that since the intermediates detected during the photocatalytic degradation of halogena ted aromatic compounds are typically hydroxylated structures, the prim ary oxidant was assumed to be the hydroxyl radical. The oxidative power of the hydroxyl radical is more than twice the power of chlorine (Goswami and Blake, 1996). After these prim ary reactions with th e hydroxyl radicals, many intermediate organic radicals and pr oducts are formed. From these radical-toradical interactions, eventually complete mineralization occurs resulting in CO2 and H2O. Mao et al. (1991 and 1992) have shown that the conditions of a system (pH, oxygen concentration, competition with other com pounds) can affect the pathway a single compound can take towards complete mineralization. Those system parameters were investigated in this research. Mao et al. (1991 and 1992) also concluded that the photocatalytic degradation of halocarbons with TiO2 was initiated via an oxidative process involving presumably su rface-adsorbed, hydroxyl radicals.

PAGE 16

6 Adsorption of compounds to the surface of th e catalyst is important in an efficient photocatalytic system. Competition with ot her compounds can be detrimental to the adsorption of specific contaminants on the catalyst surface. Chen et al. (1997) showed the effect of competitive adsorption on the catalys t surface and its effects on the degradation rates. The competition with dichloroethane (D CE) for adsorption and the affinity of ions for TiO2 were in order: chloride < nitrate < (b i)carbonate < sulfate < phosphate. A similar order was found for the inhibition of photodegrad ation: nitrate < chloride < (bi)carbonate < sulfate < phosphate. Carbonate ions can re act with the hydroxyl radicals, producing the less reactive ion radical (CO3*). The researchers postulated that the effect may have occurred but appears to be of minor impor tance in comparison to competitive adsorption. The effect of bicarbonate ions in soluti on was investigated in this research. Titanium Dioxide TiO2 is the photocatalyst most often us ed because of its low cost, non-toxic characteristic, insolubility, heat resistance, and photostability compared to other metal oxides such as zinc oxide (ZnO). Zn O can undergo photocorrosion, decreasing the effective lifetime of the catal yst (Okamoto et al., 1985). Tita nium dioxide has also been shown to be the most active semiconductor for photocatalysis (Goswami and Blake, 1996). TiO2 can be synthesized into two different crystalline forms, anatase and rutile. The anatase crystal form has been found to be the more phot ocatalytically active of the two (Ohtani and Nishimoto, 1993). The difference in reactivity can be attributed to the more positive conduction band of the rutile phase hindering molecular oxygen reduction (Tanaka et al., 1991). The degradation rate of water-soluble compounds is independent of

PAGE 17

7 the crystal form, while sparingly water-sol uble compounds are dependent on the anatase content of the TiO2 (Tanaka et al., 1991). Degussa P-25 is a commercially available TiO2 with a structure that is 70% anatase and 30% rutile. This particular TiO2 is commonly used as it has repeatedly demonstrated succe ssful photocatalytic degr adation of organics. Catalyst Supports TiO2 has been an effective photocatalyst used in slurry form to maximize exposure to the solution and UV li ght, but it can be difficult to recover the titania in the post processing of the solution. Therefore, the TiO2 has been chosen to be incorporated with a catalyst support to immobilize the titania and en sure removal and recovery. Several supports have been investigated, including glass beads, sand, clay, activated carbon, and silica gel (Al-Ekabi and Serpone 1988; Yoneyama and Torimoto, 2000). The rate of destruction of organic contaminants is generally controlled by the concentration of the contaminant; therefore, th e rate of mineralization will decrease as the contaminant concentration decreases. Usi ng a catalyst support that is capable of adsorbing organic contaminants as well as immobilizing the photo catalyst will enhance the photodegradation of the contaminants be cause of the higher concentration of contaminant present around the loaded TiO2 (Yoneyama and Torimoto, 2000). When the organic substances are oxidized on the photoc atalyst surface, the resulting intermediates can also be adsorbed to the support and furt her oxidized. The possibili ty of creating toxic intermediates during photocatalys is is a concern, but, if the intermediates are held near the catalyst, they are more likely to be completely destroyed. Silica gel (SiO2) was chosen for the catalyst su pport for this study based on the adsorption capabilities as well as the transp arent nature of the material. The silica gel

PAGE 18

8 allows the infiltration of photons to the titania surface. Sili ca also has high mechanical strength, thermal stability, and can be synthe tically formed into any shape, such as cylindrical pellets (Londeree, 2002). Silica – Titania Composites TiO2 and SiO2 can be combined, enabling the formation of highly efficient photocatalysts. These TiO2 SiO2 photocatalysts allow the placement of the catalyst on both external surfaces and in ternal surfaces within the porous silica matrix where pollutants are adsorbed. SiO2-TiO2 composites have been successful in the photocatalytic degradation of a variety of organic compounds (Matthews, 1988; Anderson and Bard, 1995; Anderson and Bard, 1997; Xu et al., 1999; Vohra and Ta naka, 2003). Xu et al. (1999) synthesized particles from a titania sol a nd silica powder and tested for photocatalytic destruction of acetophenone. Compared to the bare TiO2 prepared in parallel, all the supported TiO2 showed a higher photoactivity. The researcher inferred that this enhanced photoactivity was related to the increased adsorption of or ganic substrates, the in creased surface area of the SiO2-TiO2 composite compared to bare TiO2, and perhaps to the absence of rutile phase in the supported samples. The latter two ar e closely related to the fact that the TiO2 was highly dispersed over the porous silica. Anderson and Bard (1995) made a similar comparison of bare TiO2 to a TiO2/SiO2 composite material fa bricated through sol-gel processing by testing for the removal of rhodamine-6G. Again, SiO2-TiO2 was found to have a faster degradation rate, which was re lated to the increased adsorption of the contaminant. These studies demonstrate the idea that increasing adsorption will increase

PAGE 19

9 the concentration of the contaminant around the TiO2 and lead to faster destruction of the contaminant. When photo-decomposing phenol (a com pound with low adsorption capacity to silica), Anderson and Bard (1997) found the TiO2 alone performed better, which was to be expected since the silica did not help with adsorption capabili ties and only hindered the effect of the photocatalyst. However, when the rates were normalized to the TiO2 content of the material, the TiO2/SiO2 material demonstrated a more efficient use of the TiO2 than the slurry alone. The compounds being researched, such as toluene and chlorobenzene, have higher adsorption capaci ties than phenol. In an adsorption batch experiment performed by Hashizume et al. (2004), the amount of adsorption to a synthesized mesoporous silicate materi al was in the order of toluene benzene > benzoic acid > phenol. Demeestere (2002) compared the adsorption on TiO2 of chlorobenzene and toluene in the gaseous phase. Chlorobenz ene had a greater adsorption capacity than toluene. Silica-titania composites can be efficiently utilized in heterogeneous photocatalysis systems for the adsorption and su bsequent destruction of contaminants in an aqueous solution. In addition, a dosage of 30 wt% (equal to 12% on a weight per volume of silica precursor) wa s also found to be an op timum loading for Londeree (2002), Holmes (2003), and Ludwig (2004) fo r the silica-titaniu m composites created using a sol-gel doping method. Sol gel method. There has been much research on the different preparation methods for creating a solid SiO2-TiO2 material with photocat alytic ability. The characteristics (e.g., pore size, surface charge mechanical strength, and adsorption sites)

PAGE 20

10 of the final product are dependent on the synt hesis conditions and the type of interaction between TiO2 and SiO2. LondereeÂ’s (2002) method uses sol-gel hydrolys is to create the silica matrix, then, during gelation, the solution is doped with th e commercially available, highly efficient Degussa P25 TiO2. The silica network may either form around the titania particle and/or form a bond with the titania to secure its position in the matrix. The final product resulting from the sol-gel method steps (hydr olysis, condensation, gelation, aging, drying, curing) is a gel containing a relatively monodi sperse pore size and displaying specific characteristics associated with the cond itions it experienced during processing. The details of the method are discussed in Chapter 3, Materials and Methods. Silica gel properties. The silica network consists of four oxygen atoms bonded to each silicon atom, forming a tetrahedron, and each oxygen atom is shared by two silicon atoms (Figure 2). Figure 2. Linkages of SiO2 tetrahedras (Hench and West, 1990) These Si-O-Si, or siloxane, bonds make up the bulk silica structure. The surface of the silica is hydroxylated in the presence of water, forming Si-OH, or silanol, groups. It is these silanol groups that make th e silica surface hydrophilic and determine the reactivity of the silic a (Nawrocki, 1997)

PAGE 21

11 The hydroxyl concentrations, silanol groups, on the SiO2-TiO2 surface contribute to the ability of the catalyst to adsorb pollutants. The silanol groups can be defined as single, geminal, or vicinal. Single silanol gr oups are isolated and formed with a silicon atom that is bonded to three other oxygen atom s in siloxane bonds. Vicinal silanols have two hydroxyl groups joined to the gel at two di fferent silicon sites, but a hydrogen group from one is also bonded to the oxygen group from the other. Geminal silanols exist when two hydroxyl groups are linked to the same silic on atom. The various silanols can have different adsorption activity, and Nawrocki (1997) indicated th at the isolated silanols are the more reactive species over the hydrogenbonded silanols since the hydrogen of the isolated silanol is not bonded to another atom and is free to react with other species. Kinetics Langmuir-Hinshelwood (LH) kinetics has been shown to successfully describe the photocatalytic degradation of organic contaminants (Al-Ekabi and Serpone, 1988; Turchi and Ollis, 1989). Since LH kinetics is known to be a good model for solid-gas reactions, some modifications are necessary to model solid-liquid reactions. The LH model assumes the reactions take place at the su rface of the catalyst, and the reaction rate is proportional to the fraction of the surface covered by the reactant. r = -dC/dt = kr = krKC/(1+KC) (7) where r is the rate of the reaction, C is the concentration of the contaminant in solution, t is time, kr is the reaction rate constant, is the fraction of the surface covered by the reactant, and K is the adsorption coefficient of the reactant (Al-Ekabi and Serpone, 1988).

PAGE 22

12 Equation 7 can be simplified when the c oncentration is either very high (KC >> 1) or very low (KC << 1). Under conditions of low concentration, KC becomes negligible compared to 1 and equation 7 reduces to a first-order reaction: -dC/dt = krKC (8) Since the concentrations used in this rese arch were approximately 200-300 g/L, the first-order reaction would likely be applicable; however, this also relies on low K values. Similarly, under conditions of high concen tration, 1 becomes negligible compared to KC. Equation 7 reduces to a zero-order reac tion, where the reaction rate is equal to the rate constant, kr. This simplification is illustrated in the photocatalytic degradation of 4chlorophenol, where no additiona l increase in rate was obser ved with an increase in initial concentration above 0.2 millimoles (A l-Ekabi and Serpone, 1988). It was reasoned that at a high enough contaminant concentratio n, the surface sites of the catalyst became completely saturated, so further increase in concentration cannot further increase the rate of reaction, according to LH kinetics. Convers ely, below the specified concentration, the catalyst surface sites are not completely satura ted; therefore, the fraction of the surface covered by the contaminant will vary with concentration, thereby varying the reaction rate. It is important to note that the use of the LH equation to describe photocatalytic kinetics makes an assumption that adso rption of the contaminant to the TiO2 surface is the rate-limiting step. Microorganisms The goal of NASAÂ’s ALS is to remove or inactivate the pathogenic microorganisms that could be present in the recirculatory water system. The specifications of potable wate r include a standard of less than 100 colony forming units

PAGE 23

13 (CFU) per 100 milliliters (mL) for total bacteria. The coliform group and virus specification is non-detectable or below detectable limit (Lange and Lin, 1998). Concentrations of pathogens are genera lly low in water a nd wastewater, and no single technique is available to isolate and identify all pathogens. Instead of direct isolation and enumeration of pathogens, indi cator organisms are used to expedite the analytic procedure. The colifor m group of bacteria has been used to indicate the presence of pathogens in the water. Their presence in water indicates recen t fecal contamination, which is the major source of many enteropat hogenic diseases transmitted through water. Coliforms and pathogens both originate from the same source: fecal matter from humans and animals. Bacteria. For the purpose of this resear ch, three different bacteria, Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa were chosen as indicator microorganisms. Escherichia coli is one of the enteric bacteria, or Enterobacteriaceae which are facultative anaerobic gram-negative rods that live in the intestinal tracts of animals in both health and disease. The ente rics ferment glucose, producing acid and gas, are typically oxidase-negative, and, when motile have flagella uniformly distributed over the body. A number of genera within the Enterobacteriaceae family are major waterborne pathogens, such as Salmonella, Shigella and Yersinia Several others are normal colonists of the human gastrointestinal tract (e.g., Escherichia Enterobacter Klebsiella ), but these bacteria also may occasiona lly be associated with diseases of humans (Murray et al., 1994). Physiologically, E. coli is versatile and well-adapted to its characteristic habitats. It can grow in media with glucose as the sole organic constituent. The bacterium can

PAGE 24

14 grow in the presence or absence of O2. Under anaerobic conditions it will grow by means of fermentation, producing acids and gases as end products. However, it can also grow by means of anaerobic respiration, si nce it is able to utilize NO3 -, NO2 -or fumarate as final electron acceptors for respirat ory electron transport processe s. In part, this adapts E. coli to its intestinal (anaerobic) and its extraintestinal (aerobic or anaerobic) habitats. E. coli can respond to environmental si gnals, such as chemicals, pH, temperature, osmolarity, etc. For example, it can sense the presence or absence of chemicals and gases in its environment and move towards or away from them. In response to change in temperature and osmolarity, it can vary the pore diamet er of its outer membrane to accommodate larger molecules (nutrients) or to excl ude inhibitory substances (Todar, 2005). Staphylococcus aureus is a gram-positive spherical bacteria, about 1 micrometer in diameter that occurs in microscopic clusters resembling grapes. S. aureus are nonmotile, non-spore-forming facultative anaerobes that grow by aerobic respiration or by fermentation that yields mainly lactic acid. The bacteria are catalase-positive and oxidase-negative. S. aureus can grow at a temperature ra nge of 15 to 45 degrees and in aqueous NaCl concentrations as high as 15 percent. S. aureus ferments mannitol, which distinguishes it from S. epidermidis (Todar, 2005). Pseudomonas aeruginosa is a gram-negative rod measuring 0.5 to 0.8 m by 1.5 to 3.0 m. Almost all strains are motile by means of a single polar flagellum. Its metabolism is respiratory and never fermenta tive, but it will grow in the absence of O2 if NO3 is available as a respiratory electron acceptor. P. aeruginosa has very simple nutritional requirements. In the laborator y, the simplest medium for growth of P. aeruginosa consists of acetate for carbon and amm onium sulfate for nitrogen. It is

PAGE 25

15 tolerant to a wide variety of physical conditions, includi ng temperature and high salt concentrations. Its optimum temperature for grow th is 37 degrees, and it is able to grow at temperatures as high as 42 degrees (Todar, 2005). Bacteriophages. Bacteriophages, or phages, are parasitic viruses that infect bacteria instead of plants or animals. Phages are often used in research studies as models of human enteric viruses instea d of plant or animal viruses because one can easily prepare large populations (up to 1010 PFU/mL) of genetically homogeneous, susceptible host cells, under well-defined nutritional conditions, as compared to plant and animal hosts (Matthews, 1971). Many features of phage in fection are common to the infective process of plant and animal viruses. The three phages chosen for this research were X-174, PRD-1, and MS-2. These three organisms include both RNA and DNA phages and have different sizes. The bacteriophage X-174 is the principal representative of a group of phages that are simple icosahedrons (20 faces) with 5-nm spikes ex tending from all 12 vertices. It is a 27-nm single-stranded DNA bacteriophage with a circular topology. The bacteriophage X-174 belongs to the family Microviridae in the genera of Microvirus Related phages include S13, R, and G4 (Birge, 2000). The phage MS-2 is a single-stranded RNA bacteriophage with icosahedral morphology and a diameter of 26.0 to 26.6 nm (Sjogren and Sierka, 1994). MS-2 belongs to the Leviviridae family in the Levivirus genera. MS-2 has a linear topology and a unique sequence. Related ph ages include fr, f2, R17, M12, and Q (Birge, 2000). Both X-174 and MS-2 are coliphages, or bact eriophages that are infective to coliform bacteria. The usual hosts for the phages are E. coli strain C for X-174 and E. coli F+ for MS-2. The bacteriophage PRD-1 belongs to the Tectiviridae family in the

PAGE 26

16 Tectivirus genera. PRD-1 is a double-stranded DNA phage with icosahedral morphology and linear topology (Birge, 2000). UV Effect on Microorganisms The UV light effective in inactivating bacteria and virus resides in the UV-B and UV-C ranges of the spectrum (200 to 310 nm) with a maximum effectiveness for most bacteria and virus species occurring ar ound 265 nm (Malley, 2002). Inactivation of microorganisms by UV irradiation occurs th rough the formation of lesions in DNA, which prevent normal DNA replication, leadi ng to inactivation (Harm, 1980). Thymine bases on the nucleic acids are particularly reactive to UV light and form dimers (thymine=thymine double bonds). These dimers th en chemically inhibit transcription and replication of nucleic acids thus render ing the organism sterile (Malley, 2002). Photoreactivation is a phenomenon in which UV-inactivated microorganisms recover activity through the repair of le sions in the DNA by enzymes under near-UV light. In dark repair, UV-inactivated micr oorganisms repair the damaged DNA in the absence of light. (Morita et al., 2002). Many factors (UV dose, water quality, length of exposure to photoreactivating light, species of organism) can aff ect photoreactivation. Shaban et al. (1997) found that the enzymatic process of dark repair did not affect the recovery of irradiated organisms ( E. coli, S. aureus and coliphage). A 2-log photoreactivation of UV-inactivated E. coli did occur with subse quent exposure (four to six hours) to light in the visi ble spectrum. The strategy in UV disinfection has been to provide a high enough dosage that enough nucle ic acid damage occurs to prevent effective repair by photoreac tivation or dark repair.

PAGE 27

17 Photocatalytic inactivation. The hydroxyl radical has been theorized to act as a potent biocide because of its high oxidation potential and nonselectiv e reactivity. Several researchers have experimented with the effectiveness of the hydroxyl radical as a potential biocide to bacteria viruses, and protozoan. Irela nd et al. (1993) used a 300400 nm wavelength lamp covered with a photocat alytic sleeve formed of fiberglass mesh, which was coated with a firmly bonded layer of TiO2. The researchers dechlorinated the tap water used in the experime nt before the introduction of Escherichia coli into the system. Using a 10-fold stoichiometric excess of sodium thiosulfate, the inactivation of E. coli was a 0.4 log reduction in concentration in the 6-minute exposure time (flow rate of 2 L/min). However, after dechlorinating the water using only UV light exposure, the observed reduction in the concentration of E. coli was 7 orders of magnitude after 6 minutes of cumulative exposure. In the 9-minute cumulative exposure sample, the E. coli counts were below detection limit, indicat ive of a total reduction of 9 10 log10 units. Ireland et al. (1993) concluded that TiO2 photocatalysis may be a viable process for disinfection of bacteria in water tr eatment systems; though, inorganic-radical scavengers can have a major negative impact on the efficacy of the process. The presence of organic matter also degrad es the inactivation kinetics, ostensibly by competing with bacteria for the hydroxyl-radical oxidant. Sjogren and Sierka (1994) achieved a one -log reduction of the phage MS-2 after 10 minutes of irradiation in a continuously stirred batch reactor containing a TiO2 slurry. The 365-nm UV lamp emitted an irradiance 2 mW/cm2. The solution was buffered with 22 mM of phosphate (pH = 7.2) and also cont ained 5 mg/L (0.14 mM) of chloride from a saline buffer used in the inoculation of MS2. A 3-log reduction was achieved with the

PAGE 28

18 TiO2 slurry by adding 0.1 mg/L of FeSO4 under the same conditions. This increased level of inactivation was attributed to suppleme ntal hydroxyl radical oxidations that were enabled by Fenton reactions, such as: Fe (II) + H2O2 IC Fe (III) + OH+ OH* (9) where IC is an intermediate comp lex and Fe (III) and Fe (II) are Fe3+ and Fe2+ and complexes. The H2O2 was thought to have been produced by the ereduction of oxygen. Cho et al. (2005) found that irradiation for 120 minutes was required for the 0.95and 2.25 log inactivation of MS-2 phage and Escherichia coli respectively. The batch study included 1 g/L of TiO2, a 20 mM phosphate buffer, and 18W black-light blue lamps (300 420 nm) as the source of irradiation. The light intensity wa s measured at 7.9*10-6 einsteins/liter/s (approximately 2.25 – 3.15 mW/cm2 in the UV wavelength range of 300 – 420 nm). By performing studies in the presence of scavengers, tert -butanol and methanol, Cho et al. (2005) concluded that the MS-2 phage is inactivated mainly by the free hydroxyl radical in the solution bulk but Escherichia coli is inactivated by both the free and the surface-bound hydroxyl radicals. Escherichia coli might also be inactivated by other ROS (reactive oxygen species), such as O2 and H2O2. Wei et al. (1994) found the bacterial inactivation of E. coli (106 cells/mL) adhered to first-order kinetics. The rate constant (5.5*10-2 min-1) was proportional to the incident light intensity in th e range of 180-1660 E/s/m2 and to the TiO2 dose. The researcher achieved a 6-log reduction in E. coli by irradiating a 1 g/L TiO2 slurry with 52.3 mW/cm2 of UV intensity for 30 minut es using a 380-nm lamp. Bekbolet (1997) also found that the rate constants decreased with increasing E. coli concentration. After one hour of illu mination with a 320-420 nm UV lamp, the

PAGE 29

19 inactivation rate calculated was 1.78*10-2 min-1. The light intensity of the UV lamp was 67.9 E/s/m2 (approximately 1.94 – 2.54 mW/cm2 in the UV wavelength range of 320 – 420 nm), and the TiO2 aqueous suspension dose was 1 mg/mL. Lee et al. (1997) achieved a 2.2-l og inactivation of the phage Q using immobilized TiO2 and 1-hour irradiation with near UV black light (300-400 nm) at an intensity of 3.6 mW/cm2. A 0.5-log Q inactivation was observe d with the black light alone. A control experiment was perf ormed using a 0.1 g/L slurry of TiO2. Under the same irradiation conditions, a 2.4-log inactiv ation was achieved in only 20 minutes as opposed to 60 minutes w ith the immobilized TiO2. No noticeable difference in inactivation was observed between a 254-nm germicidal lamp irradiation (light intensity of 0.6 mW/cm2) with and without immobilized TiO2. The rate constant of inactivation was found to be proportional to the light intensity in the ra nge of 3 – 8 mW/cm2. At 5 mW/cm2, the rate constant was 0.14 min-1 for a 5000X dilution of the Q phage in broth with autoclaved pure water and 0.02 min-1 for dilution of 100X. The lower rate constant for the 100X broth dilution was attributed to radical scavenging and adsorption on reaction sites of th e broth components. Shown in Table 1 are the inactivation resu lts for each of these researchers along with the given parameters for their experiments including UV intensity and dose (when available). Most of the expe riments were performed in a batch reaction with a TiO2 slurry.

PAGE 30

20 Table 1. Comparison of photocatalytic inactivati on of microorganisms fo r various researchers Researcher Organism Log10 inactivation UV intensityTimeWavelengthUV dose TiO2 Configuration TiO2dose Background species Flow rates mW/cm2 min Nm J/cm2 g/L L/min Ireland et al. (1993) E. coli 7 Unavailable6 300 400 Unavailable immobilized N/ANone 2 E. coli 0.4 Unavailable6 300 400 Unavailable immobilized N/ASodium thiosulfate2 Wei et al. (1994) E. coli 6 52.3 30 380 94.14 slurry 1 None batch Bekbolet (1997) E. coli 3 1.94 2.54 60 320 – 420 6.98 – 9.14 slurry 1 None batch Cho et al. (2005) E. coli 2.25 2.25 3.15 120 300 – 420 16.20 – 22.68 slurry 1 20mM HPO4 2batch MS-2 0.95 2.25 3.15 120 300 – 420 16.20 – 22.68 slurry 1 20mM HPO4 2batch Sjogren and Sierka (1994) MS-2 1 2 10 365 1.20 slurry 1 22 mM HPO4 20.14 mM Clbatch Lee et al. (1997) Q phage 2.2 3.6 60 300 – 400 12.96 immobilized N/Abroth batch Q phage 2.4 3.6 20 300 – 400 4.32 sl urry 0.1 broth batch

PAGE 31

21 For the E. coli studies, the largest reduction (7 orders of magnitude) was seen in the flow-through reactor using immobilized TiO2. The UV intensity was not reported by Ireland et al. (1993), so comp arisons with other systems are difficult. The major difference between experiments in Table 1 is the UV dose applied to the batch reactor. The second greatest reduction, by We i et al. (1994), of 6 logs can be attributed to the high UV intensity and therefore UV dose, which is over an order of magnitude above BekboletÂ’s UV dose. For the phage studies, including MS-2 and Q there was variability in the results. Sjogren and Sierka (1994) achieved a one-l og reduction in MS-2 with a much lower UV dose than Cho et al. (2005). Both researchers had the same TiO2 dose and approximately the same background phosphate concentration. In the slurry experiment performed by Lee et al. (1997), the inactivati on was twice as much as the ot her two researchers, but the phage was a different type, the TiO2 dose was less, and the UV dose was in between that of Sjorgren and Sierka (1994) and Cho et al. (2005). Also, th e background contents in the solution were not specified by Lee et al. (1997). Mechanism for inactivation. Maness et al. (1999) found the lipid peroxidation reaction was the underlying mechanism of death for E. coli K-12 cells that were irradiated in the presence of the TiO2 photocatalyst. The occurr ence of lipid peroxidation and the simultaneous losses of both membrane -dependent respiratory activity and cell viability depended strictly on the presence of both light and TiO2.

PAGE 32

22 CHAPTER 3 MATERIALS AND METHODS Silica-Titania Composites The silica-titania composites were create d using a sol-gel method (Powers, 1998). The silica precursor was tetraethyl-ortho-silicate (TEOS) (Fisher Scientific, reagent grade). It was mixed with nanopure water us ing a water-to-TEOS mole ratio of 16:1. Ethanol (Aaper Alcohol, 200 proof) was used as the solvent to facilitate the miscibility between the TEOS and water. Two acid cataly sts were used: a 1 M nitric acid solution, made from 15.8 M nitric acid (Fisher Scientif ic, certified ACS.) a nd nanopure water, and a 3% solution of hydrofluoric acid, formul ated from 49% hydrofluoric acid (Fisher Scientific, reagent ACS.) and nanopure water. The basic formula used to create 10 g of pellets with a pore size of roughly 140 is as follows: 25 mL nanopure water, 50 mL ethanol, 35 mL TEOS, 4 mL nitric acid (1 M), and 4 mL HF (3%). By changing the amount of HF added, the pore size and st ructure within the gel can be altered. The chemicals were added to a polystyre ne container. A ma gnetic stir plate provided sufficient mixing while a known mass of Degussa P25 TiO2 was added to the batch. The TiO2 loadings are reported on a basis of TiO2 weight per TEOS volume. For example, the 12% TiO2 loading is determined using 4.2 g of TiO2 per 35 mL of TEOS as a percentage. The properties of Degussa P25 as provided by Degussa are listed in Table 2 (Ludwig, 2004).

PAGE 33

23 Table 2. Properties of Degussa P25 TiO2 Specific surface area (BET) 50 +/15 m2/g Average primary particle size 30 nm Tapped density 130 g/L pH value in 4% dispersion 3.5 4.5 The solution was allowed to mix for 30 minutes and then was transferred by a pipette into Fisherbrand (Fisher Scientific), polystyrene 96-well assay plates before gelation to form the titania-doped pellets (Fi gure 3). The assay plates contained 0.45 mL in each well. Each batch of chemicals create d approximately 4 assay plates of pellets. Figure 3. Silica-titania composite suspension bei ng transferred to the 96-well assay plates used for creating the pellet shape After 2 hours of gelation, the plates we re covered with lids and wrapped in aluminum foil to limit volatili zation during the aging process. The gels were aged at room temperature for two days, then at 65o C in an Oakton Stable Temperature oven for two days. After aging, the pellets were remove d from the assay plates and placed into Teflon containers for the next se ries of heat treatments. A small hole was placed in the lid of the containers to allow the liquid from the gelÂ’s pores to escape as a vapor, thus providing uniform drying of the gel. Us ing a Yamato DVS 400 Drying Oven, the

PAGE 34

24 temperature was ramped from room temperature to 103o C (2o/min) and kept constant for 18 hours, resulting in the vaporization of liqui d solution within the silica network. Next, the temperature was ramped to 180o C (2o/min) to remove any physically adsorbed water. It was kept constant for 6 hours and then was slowly decreased back to room temperature over a 90-minute period. The resultant size of an individual cylindric al pellet after drying was approximately 5 mm in length with a diameter of 3 mm. The BET (Brunauer, Emmett, and Teller equation) su rface areas and pore volumes of the gels were analyzed using a Quantachrome NOVA 1200 Gas Sorption Analyzer. Reactor Design Reactor for Oxidation of Organic Compounds A reactor was designed by Holmes (2003) a nd constructed by Analytic Research Systems1 in order to contain the silica-titania composite pellets and provide effective oxidation of organic compounds and inactivati on of pathogenic microorganisms. HolmesÂ’ designed reactor was used in the subseque nt experiments involv ing the oxidation of selected organic compounds. The reactor was constructed of quartz fo r its ability to transmit UV radiation energy. The reactorÂ’s cylindrical shape (Figure 4) was chosen to optimize the exposure of the pellets to the light. The inside diameter of the reactor was 8.5 cm, and the outside diameter was 10.5 cm. These dimensions were chos en so that all pellets would be able to react with the UV radiation. Th e radiation incident on the pe llets would be reduced by a thicker reactor, thereby not eff ectively using all of the catalyst. The reactor is 14 cm long, not including the influent and effluent ends. E ach of the influent a nd effluent cylindrical ends has a diameter of 1 cm. 1 Analytical Research Systems, Inc. PO Box 140218 Gainesville, FL 32614-0218

PAGE 35

25 Figure 4. Organic reactor fille d with silica-titania compos ites on its support stand and connected to the system used for the testing The entrance and exit of the reactor were each threaded to allow for connections with the PTFE (polytetrafluoroethylene) tubing used in the system. The entire reactor was made of quartz with the exception of one e nd where a glass frit was placed to keep the pellets from flowing out of th e reactor with the solution. Th e reactor had a volume of 436 mL with the pellets taking up 109 mL a nd interparticle space accounting for the remaining 327 mL. Holmes (2003) estimated th e interparticle bed porosity within the reactor (75%) by filling a graduated cyli nder with known volumes of pellets and nanopure water. Reactor system. A system was designed for testi ng the reactorÂ’s capabilities for degrading the target analytes The system included the react or and reactor support, two sampling ports, a source tank, a pump, a flow da mpener, stirrer, a waste beaker, and a sparge tank for the low dissolved oxygen experiments (Figure 5).

PAGE 36

26 Figure 5. Diagram of system setup used in recirculating and single-pass conditions The reactor was placed on a wooden support in the horizontal position to limit the influence of gravity on the flow. Since NASA sp ecifications require the reactor to work in a micro-gravity situation, it was necessary to prevent gravity from enhancing the results of the experiments. Connections were available for four UV lamps to be used in the center of the reactor. The lamps were 12-inch, 8-watt lamp s that each provided approximately 4.44 W of available UV energy (365 nm wavelength) to the inner surface of the reactor. The support also included a c over that could be placed over the reactor and UV lamps to prevent exposure. All of the tubing used was PTFE tubing to prevent adsorption or desorption of organic compounds during experi mentation. All connections

PAGE 37

27 were sealed with PTFE tape, since several of the target analytes are volatile organic compounds (VOCs). Since each sample taken for organic an alysis from the system was 40 mL, a source tank was necessary to prevent pockets of air from forming in the tubing or in the reactor. A 4L Erlenmeyer flask was used. Glass rods placed through a rubber stopper on top of the source tank allowed solution to be carried in and out of the flask. The rubber was covered in Teflon tape to preclude reactions between the stopper and the test solution. One glass tube through the stopper allowed a small air leak to prevent a vacuum situation within the system when samples were taken. This did result in a necessary 14.5 mL of headspace at the top of the flask to prev ent the loss of solution out of the vent tube. The total volume of the system was just over 4.5 L with 4.3 L from the source tank and the reactor. The pump used was a L/S PT FE-Tubing Pump Head powered by a L/S Variable-Speed Modular Drive. A polyethylen e pulse dampener was used to ensure a steady flow rate in the system. Reactor hydrodynamics. The hydrodynamic behavior of reactors operates in a range between two ideal reactor models. Th ese models, the continuously stirred tank reactor (CSTR) and the plug flow reactor ( PFR), represent two extreme cases of fluid flow behavior. A CSTR consists of a con tinuous flow of fluid particles, which are assumed to be completely mixed, and the concen tration in all locations within the reactor is the same as the effluent concentration. A PFR represents a fluid flow with no mixing in the direction of the flow and infinite mixi ng perpendicular to the flow direction, and, therefore, a concentration gradient exists from the influent to the effluent of the reactor. The PFR is representative of an infinite number of CSTRs in series.

PAGE 38

28 In order to determine the hydrodynamic behavior of the reactor, Holmes (2003) performed a tracer analysis. Sodium chloride (NaCl) was used as the tracer to prevent reactions with the composite pellets or the insi de of the reactor. Th e chosen flow-rate for the tracer analysis was 10 mL/min since this was the design flow for the reactor. With a reactor volume of 436 mL and 109 mL of that composed of pellets, the mean residence time of the reactor was predicted to be 32.7 min. The conductivity of the effluent was measured using a Fisher Scientific conductivity probe. A lin ear correlation was found between the conductivity reading on the probe and the concentrati on of NaCl in solution by m easuring known concentrations NaCl and comparing them w ith the conductivity reading. The mean residence time, variance, and nu mber of tanks in series are shown in Table 3. The residence time represents the av erage amount of time the NaCl remained in the reactor. The variance describes the vari ability in times over which the contaminant exited the reactor. The number of CSTRs in se ries was calculated to model the residence time distribution of the reactor. An infinite number of CSTRs in series represents a plug flow reactor. Table 3. Key statistics from the tracer analysis (Holmes, 2003) Mean Residence Time 42.1 min Variance 459 Number of CSTRs in series 3.9 A residence time distribution (E-curve) wa s generated from the tracer analysis data and is shown in Figure 6. This distributi on was then compared to the tanks-in-series (TIS) model of CSTRs, which was creat ed using the following equation:

PAGE 39

29 1)! 1 ( n bar t ntt nt n t ne Ebar (10) Where n is the number of CSTRs in series, t is the time in the reactor, and tbar is the mean residence time. The tanks-in-series (TIS) m odel was used because of its ability to accurately model and simulate the flow in the reactor. E Curve 0.0 5.0 10.0 15.0 20.0 25.0 0.0020.0040.0060.0080.00100.00120.00 t* (min)E* (min-1)*103 Data Model Figure 6. E-curve generated from the tracer analysis performed on the reactor. “Data” represents the E-curve from the tracer analysis of the reactor, and “Model” is the E-curve generated from the TIS model (Holmes, 2003) A comparison with the TIS model rev ealed that the reactor behaves as approximately four CSTRs in series. Reactor for Inactivation of Bacteria and Viruses The vertical cylindrical reactor, s hown in Figure 7, was used for the microbiological experiments w ith bacteriophages and bacteria. The reactor was designed

PAGE 40

30 with a hollow center and thin annulus to allo w the UV lamp to be placed in the center, providing maximum UV light exposure to the pellets. Figure 7. Vertical quartz reactor packed with si lica-titania composites. The inner wall of the annulus was a qua rtz tube that could be completely removed, making it simple to remove the pelle ts after testing. Eith er a 365 nm or 254 nm wavelength lamp was placed in the center of th e quartz tube. As measured at the center of the 365nm lamp, the intensity was 7.4 mW/cm2 at the inner diameter of the annulus and decreased to 4 mW/cm2 at the outer diameter with no pe llets in the reactor. For the 254 nm lamp, the intensity was 12 mW/cm2 at the inner diameter and 8 mW/cm2 at the outer diameter without pellets. W ith pellets in the reactor, the UV intensity at the outer diameter was near zero for both the 254-nm and 365-nm lamps. The reactor was 19 cm long with an inner diameter of 2.5 cm and an outer diameter of 4.2 cm. The empty bed volume of pellets was 138.6 mL The reactor was enclosed in a box to provide control over its exposure to ambient light. Reactor system. The system, shown in Figure 8, consisted of the reactor, 6 mm PTFE tubing, a Master flex L/S Digital St andard Drive peristaltic pump, and a 500mL source tank of stock spiked solution. The syst em was operated in a closed loop system in which the solution was recirculated through th e reactor. A cover (not shown in Figure 8)

PAGE 41

31 was placed over the front of the reactor to co mpletely enclose it during operation of the system. Figure 8. Treatment system consis ting of reactor packed with SiO2-TiO2 composites, UV lamp, peristaltic pump, and reservoir of stock solution Reactor hydrodynamics. Ludwig (2004) conducted a tracer analysis to determine the behavior of the reactor. The tracer analysis method was described previously in the discussion of the organic contaminant reactor. The data collected from the tracer analysis was used to calculate a mean residence time of 12.5 minutes and to model the reacto r behavior. A fractiona l age distribution (Ecurve) of the sodium chloride in the reacto r was created and compared to the TIS model. The comparison revealed that the reacto r behaved as five CSTRs in series. Analytical Methods Analysis of Organic Compounds Samples were collected from one of th e two three-way luer-lock sampling ports provided in the system using a gas tight, luer-lock syringe. The samples were then transferred to 40-mL volatile organic analysis (VOA) vials. The vials were stored at 4oC

PAGE 42

32 until analysis. Each sample was viable for up to two weeks at that temperature. A GCQ gas chromatograph/ mass spectrometer with a Tekmar 3100 purge and trap extraction system was used to analyze the samples for all six of the constituents. Analyses were performed in accordance with the USEPA method 524.2, Methods for the Determination of Organic Compounds in Drin king Water: Supplement 2 (USEPA, 1992). Samples were purged at room temperature for 11 minutes at 35 mL/min with helium. They were then dry purged for 2 minutes to remove water. A Supelco k-trap was then back flushed and heated to 250C for desorption to the colu mn. The column used was a DB-VRX column from J&W Scientific. It was a 75-meter long column with a 0.45 mm inner diameter, and a 2.55-m-film thickness. Desorption to the co lumn lasted for 6 minutes at 250C, and the trap was then baked at 270C for 10 minutes. The column was taken from a 25C start temperature and ramped to 220C at a rate of 6C/min, after an initial hold at 35C for 6 minutes. The mass spectrometer used is an ion trap that scans from 34 amu and 280 amu with 0.6 seconds/scan. The reliability of the GC/MS analysis wa s verified using the percent relative standard deviation (RSD). The RSD is based on the response factor (RF) of a chemical. The RF and RSD are calculated using the following equations: RF = (Aa* Cis)/(Ais* Ca) (11) Where Aa = GC peak area of the analyte, Ai = GC peak area of internal standard, Ca = concentration of the analyte, and Ci = concentration of the internal standard. RSD = 100 (SRF / MRF) (12) Where SRF = standard deviation of response factors and MRF = average response factor. The USEPA method requires a RSD of less than 30%, but strongly recommends an RSD

PAGE 43

33 of less than 20%. Table 4 gives the average RSDs for the analyses performed in this research. Table 4. Average RSDs of SOC analysis Compound Average RSD (%) Acetone 32.59 Carbon Disulfide 13.71 Chlorobenzene 14.91 Ethyl Acetate 18.31 Methyl Methacrylate 20.14 Toluene 12.79 Analysis of Bacteria and Viruses Bacteriophage. Assays were performed to de termine the concentration of biologically active agents in a sample. A double overlay phage assay was performed for the three active bacteriophages, X-174, PRD-1, and MS-2. The bacterial hosts for the phages were Escherichia coli B ( X-174), Salmonella typimurium (PRD-1), and Escherichia coli C-3000 (MS-2). The phage assay, as reported by Matthews (1971), depends upon the ability of an active phage pa rticle to clear a small portion of a “lawn” of infected bacteria, which has been seeded on a nutrient agar plate. In a tube of melted nutrient agar, a dilution of phage and sufficien t host bacteria to cover the surface of the plate were combined. The mixture was poured evenly onto the nutrien t agar plate. After the agar has hardened, the plate was in cubated at 37C for 8 – 12 hours in a GCA Precision Scientific Model 6 Incubator. The phage infective cycle normally terminates with lysis, or rupture, of the infected cell, wi th release of the newly formed virus progeny. Since diffusion of the newly formed virus is prevented in the solid agar medium, the progeny infect any uninfected cel ls in their immediate vicinity. Elsewhere the bacterial lawn grows undisturbed, giving an opaque appearance to most of the plate.

PAGE 44

34 Each plaque is a clear area, resulting from lysis of all of the cells in the immediate vicinity of an infectious pa rticle. A countable plate cont ains between 30 and 300 plaques. Bacteria. A bacteriological assay was performed for the three active bacteria, Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa, using a heterotrophic plate count to obtain a viable count of col ony forming units (CFUs) of living bacteria in the samples. As with the phage procedure, a count able plate contains between 30 and 300 CFUs. A dilution of bacteria was spread evenly using a sterilized glass stirring rod on a nutrie nt agar plate. Bacteria-spe cific media were used to distinguish between the three bacteria in the samples, MacConkey agar for Escherichia coli C-3000, Mannitol Salt agar for Staphylococcus aureus, and Pseudomonas isolation agar for Pseudomonas aeruginosa. The plates were incubated at 37C for 24 – 48 hours. Experimental Procedures Oxidation of Organic Compounds In order to test the efficiency of th e reactor system to degrade the organic compounds, a spiked solution of the com ponents of interest was made. A 50-mL volumetric flask was filled with nanopure water. Using a 10-L syringe, specified amounts of each compound (acetone, carbon disu lfide, chlorobenzene, ethyl acetate, methyl methacrylate, toluene) were added to the solution. The amounts usually ranged between 0.70 – 3.0 L. Between each addi tion, the solution was capped to minimize volatilization, and the syringe was rinsed using methanol. After all the compounds were added to the solution, the flask was inverted several times to ensure thorough mixing. The contents of the flask were then transferred to a 50-mL air-lock syringe. A ratio of 1:100 was achieved for the initial solution. A 4L Erlenmeyer flask (source tank) was filled

PAGE 45

35 with 3960 mL of nanopure water. Using th e syringe, 40 mL of the solution was transferred to the 4-L source tank. The rema ining 10 mL of solution was added to a 1-L volumetric flask containing 990 mL of nanopure water. Th e source tank contained a stirring rod and was placed onto a stirrer to ensure proper mi xing of the initial solution. The silica-titania composites packed in the reactor were the same for experiments performed by Holmes (2003) and remained in the reactor throughout the organic experiments. Replication. For the initial degradation experi ment, a representative procedure followed by Holmes (2003) was adopted. Ho lmes performed experiments testing the effects of UV radiation intensity, contact time, and adsorption. The system was filled with the spiked solution at a rate of 10 mL /min. Once the reactor and tubing system were full, additional spiked solution was adde d to the source tank to reduce headspace and volatilization effects. At this time, samples we re taken from Port 1 to record the initial concentrations of the compounds. Holmes (2003) performed exhaustive adsorption experiments with the reactor system and found that after five hours of recirculation the adsorptive effect of the photocatalysts is mini mal, so the solution wa s recirculated in the reactor system for 5 hours in order to meas ure the adsorption of the compounds onto the silica-titania pellets. After 4. 5 hours of recirculation, tw o 12-inch, 8-Watt, 365-nm UV lamps were placed in the center of the reacto r and turned on to warm up. The lamps were shielded with aluminum foil to inhibit early interactions wi th the photocatalysts. After 5 hours, the shields were taken off. A cove r was placed over the reactor to protect laboratory personnel from the UV light. The re actor system was then switched to the single-pass system as shown in Figure 2. One liter of solution passed through the system

PAGE 46

36 and into the waste tank. It was assumed that by this point the system had reached a stable condition. Replicate samples were taken both at Ports 1 & 2 (bef ore and after the reactor). Alkalinity. Hydroxyl radicals created in the photocatalysis process can be scavenged by carbonate and bicarbonate ions (B ekbolet and Baleioglu, 1996; Chen et al., 1997; Munter, 2001). The effect of bicarbonate i ons present in soluti on was investigated. From bench top batch studies with the silicatitania composites, a buffering concentration of 80 mg/L of NaHCO3 was found to be needed to hold the pH of the solution at around 6.3. An experiment was performed replicati ng the conditions of the two previous reactor experiments. The buffer solution was made in one-liter increments. One liter of nanopure water was measured using a volum etric flask and 80 mg of NaHCO3 was weighed and deposited in the water. The flask was inverted for mixing to ensure that all the NaHCO3 was dissolved. HCl was added to the solutions via a Pasteur pipette until the pH of the solution reached 6.3. The initial pH of the so lution was about 8.3. Four liters of solution were transferred to the 4L-erlenmeyer flas k used as the source tank in the reactor experiment. An additional liter was kept in a 1L-erlenmeyer flask to use to fill up the system as needed. The flasks were covered with parafilm. The spiked solution was created as previously discussed using the buffered solution instead of nanopure water. The reactor experiment proceeded as discusse d earlier with the only difference being that pH measurements were also taken at severa l points in the experiment along with the samples for analyzing the organic constituents When the system was filled initially, the pH of the solution before entering the reactor (Port 1) was 6.61. Immediately after exiting the reactor (Port 2), the pH was 4.25. Tabl e 5 shows the corresponding pH values for the next two sample points in the experiment.

PAGE 47

37 Table 5. pH values for experiment with 80 mg/L HCO3 solution pH t = 0 t = 4.5 hr t = 6.5 hr Port 1 6.61 6.04 5.99 Port 2 4.25 4.59 4.52 Increased alkalinity The reactor experiment was repeated with an increased buffer concentration of 200 mg/L of NaHCO3. The procedure remained the same as in the previous experiment. The initia l pH of the buffer solution wa s 6.3, after the addition of HCl. After the organic contaminants were spiked into the solution, the reactor system was filled at a rate of 10 mL/min. The initial measured pH in the system was 6.49. The solution recirculated in the sy stem for five hours. Then, one liter of solution was passed through the system in a single-pass mode before the samples were taken before and after the reactor. Table 6 shows the correspondi ng pH values for the sample points: Table 6. pH values for experiment with 200 mg/L HCO3 solution pH t = 0 t = 4.5 hr t = 6.5 hr Port 1 6.49 6.18 6.16 Port 2 6.48 5.66 5.65 High alkalinity Using the alkalinity value reported for the simulated NASA wastewater (Lange and Li n, 1998), 2360 mg/L as CaCO3, the buffer solution was prepared to this specific ation. 3965 mg/L of NaHCO3 was used as the concentration of the buffer solution. The initial pH valu e for the solution was 8.2. The organic contaminants were spiked into the solution and the reactor experiment was run as previously discussed with one exception. Afte r one liter of soluti on was expended in the single-pass mode and samples we re taken before and after the reactor, the system was switched back to recirculation for 3.5 hours. A sample was then taken at the effluent of the reactor. Table 7 shows the corresponding pH values for the sample points in the experiment.

PAGE 48

38 Table 7. pH values for experiment with 3965 mg/L HCO3 solution pH t = 0 t = 6.5 t = 10.25 Port 1 8.17 7.66 Port 2 6.58 7.49 7.64 Low dissolved oxygen. Oxygen is the electron accep tor in the photocatalytic reaction, preventing electron and electron hole from recombining and thus leaving open the reaction sites for photocatalysis to take pl ace. Accordingly, an experiment to simulate a low dissolved oxygen (DO) environment and te st the effectiveness of the oxidation of the organic contaminants was performed. The procedure used to achieve a low DO environment was to fill the reactor system with nitrogen-sparged nanopure water, while continually adding more sparged solution until the reactor system contains only sparged, low DO solution. The 4-L source tank was filled with nanopure water and sparged with nitroge n to a level of 0.05 mg/L. The sparged water was pumped into the reactor system and wasted. The initial DO level at Port 1 was 3.35 mg/L and 4.4 mg/L at Port 2. The level in the source tank remained at 0.05 mg/L. The source tank was continually being sparged with nitrogen gas. Each sample was taken from the system by extracting the water using the air-tight syringe, transferring the solu tion to a 20-mL beaker, and bathing the sample with nitrogen gas to minimize the transfer of oxyge n from the atmosphere into the sample. A YSI 52 Dissolved Oxygen Meter, previously ca librated to 100% saturation and standard pressure, was used to obtain the measurements. The DO calibration was achieved by subjecting the probe to a high relative humid ity environment, i.e., placing the probe inside a plastic cap with a sponge saturated with nanopure water. The system was being circulated with lo w DO water from the source tank, and the DO level was being monitored at Port 1 and 2. After 16 hours of wasting sparged water,

PAGE 49

39 the Port 1 DO level was 0.99 mg/L and Port 2 was 1.34 mg/L. Two liters of solution had been wasted to reach this point. At this tim e, the flow of solution was stopped, 200 mg/L of NaHCO3 was added to the source tank and sparge d with nitrogen gas to a level of 0.32 mg/L. The organic contaminants were inject ed into the source tank via a Tygon plastic tube through the vent port. Af ter the contaminants were spik ed, the continual sparging of the source tank was stopped because of the volatility of the contaminants. The system was switched to a recirculati ng system. The DO was monitore d before the five hours of adsorption as 1.51 mg/L at Port 1 and 1.56 mg/L at Port 2. After adsorption, the system was switched to single-pass and the UV lamps were turned on to sample for destruction. After one liter of solution was wasted, sa mples were taken at Ports 1 and 2 for the contaminants as well as for DO. The levels were 2.67 mg/L DO at Port 1 and 1.65 mg/L DO at Port 2. According to Standard Methods (1998), one method for calibrating a dissolved oxygen probe for null dissolved oxygen is to supe rsaturate a water with sodium sulfite in the presence of cobalt chloride, which acts as a catalyst in the reac tion of stripping the water of oxygen. Na2SO3 was added to nanopure water in excess by double the stoichiometric amounts (2 moles of Na2SO3 for every mole of O2). The next experiment with the low disso lved oxygen system involved using the chemical addition of Na2SO3 and CoCl2 to reduce the DO level just before the single-pass destruction step of the experi ment procedure. The experiment began as the previous runs. 200 mg/L of NaHCO3 was added to nanopure water and the organic contaminants were spiked into the solution. The solution was r ecirculated throughout th e reactor system for five hours without UV light for the adsorpti on step. The DO level af ter this step was

PAGE 50

40 measured to be 6.88 mg/L and the pH was 8. 05. When the reactor system was switched to single-pass and the UV lights were tu rned on, an excess mixture of Na2SO3 and CoCl2 was injected into the source tank to lower the DO level. The source tank was also bathed in nitrogen gas to prevent oxygen diffusing in to the solution from the atmosphere. After the one liter of solution was wasted, sample s were taken for the organics and the DO level monitored at Port 1 was 2.71 mg/L and the pH was 7.91. At Port 2, the DO level was 4.06 mg/L and the pH was 7.99. TOC destruction in synthetic wastewater. The next experiment performed involved testing simulated NASA wastewater for total organic carbon (TOC). The wastewater feed was created using a formul a for typical wastewater provided by Johnson Space Center in Houston, TX. The chemical composition includes the shower waste, hand wash waste, oral hygiene waste, urine, and urine flush waste th at would be expected from a crew of four people. Listed in Ta ble 8 are the amounts of each constituent present in the simulated wastewater. Table 8. Simulated wastewat er composition (Ludwig, 2004) Wastewater Component Quantity Pert Plus for Kids 1.2 g Deionized water 999.4 ml Ammonium bicarbonate NH4HCO3 2726 mg Sodium chloride NaCl 850 mg Potassium bicarbonate KHCO3 378 mg Creatinine C4H7N3O 248 mg Hippuric acid C9H9NO3 174 mg Potassium dihydrogen phosphate KH2PO4 173 mg Potassium bisulfate KHSO4 111 mg Citric acid monohydrate C6H8O7H2O 92 mg Tyrosine C9H11NO3 66 mg Glucuronic acid C6H10O7 60 mg 1.48N Ammonium hydroxide NH4OH 10 ml

PAGE 51

41 The formula represents the raw concentr ated wastewater that would be the influent for the beginning of the water treat ment system and would be treated by several processes before entering the post-processor. To account for pretreatment, the simulated wastewater was diluted using 9 mL of the concentrated wastewater and 991 mL of nanopure water for every liter of soluti on needed. This created a solution of approximately 3 ppm of TOC, which is within the range of the expected TOC concentration of the post processor influent (Cambel et al., 2003). The parameters for the experiment remained the same as in previous runs of the reactor. The initial pH of th e solution was 8.37. After the fi ve hours of recirculation and single-pass through the reactor, the pH before th e reactor was 7, and af ter the reactor, the pH was 5.4. The samples taken were an alyzed for TOC using a Tekmar Dohrmann Apollo 9000 TOC Analyzer. After the experiment with the NASA-si mulated wastewater, nanopure water was sent through the reactor to test for the deso rption of TOC from the pellets after contact with the wastewater. After the reactor was f illed in 28 minutes, samples were taken at the effluent of the reactor at 31, 51, and 71 mi nutes. The samples were analyzed for TOC. Inactivation of Bacteria and Viruses Preparation. Stock cultures were prepared from existing cultures of the bacteriophages X-174, PRD-1, MS-2. Confluent plat es of each phage were prepared using the double-overlay phage assay proce dure described in the analytical methods section. The medium for the plates was a DifcoTM Plate Count Agar. A confluent plate contains around 103 plaques per plate. A 3% solution of BactoTM Tryptic Soy Broth (TSB) was then poured onto the plates, and th e top agar was scraped off using a flame-

PAGE 52

42 sterilized bent glass rod. Us ing a Beckman J2-HS Centrifuge the liquid solution was then centrifuged for 10 minutes at 5,000 rpm. The supernatant was poured off into a second sterile tube and centrifuged for 10 minutes at 10,000 rpm. The supernatant was then filtered through a 0.45 m plastic filter using a sterile plastic syringe The resultant stock cultures were kept refrigerated at 4C. For each of the experiments the stock culture was diluted from a starti ng titer of between 108 – 1010 PFU/mL down to the initial titer of the experiment of 105 – 107 PFU/mL, depending on which bact eriophage. The stock cultures of bacterial hosts used for the phage assay we re maintained on plate count agar plates at room temperature. The nutrient agar used for the top layer of the double-overlay assay procedure was made by mixing together BactoTM Tryptic Soy Broth, Fisher Scientific Purified Grade Agar (CAS 9002-18-0) and deionized water. The mixture was autoclaved for 15 minutes in an Amsco Eagle Series 2011 Gravity Autoclave to melt the agar, cooled in a National Appliance Company Model 320 water bath fo r about an hour, and dispensed into 13mm glass tubes (5 mL per tube). The tubes were capped, autoclaved for 15 minutes for sterilization and refrigerated for storage. To make serial dilutions of a sample dilution tubes were prepared by mixing BactoTM Tryptic Soy Broth with deionized water in a 1% concentration. Three mL of solution was dispensed per 13 mm glass tube. The tubes were capped, autoclaved for 15 minutes for sterilization and refrigerated fo r storage. After autoclaving, the volume in each tube is approximately 2.7 mL, so the addition of 0.3 mL of sample will produce a 1/10 dilution of the original sample.

PAGE 53

43 Several agar plates were used in the microbial experiments performed. For the bacteriophage assay, DifcoTM Plate Count Agar was used as the culture media. All three of the bacteria were combined in one solu tion for each experiment. Each of the three bacteria grows on the plate count agar, so isolating culture media were used to differentiate between the three bacteria. BBLTM Mannitol Salt Agar was used for the isolation of Staphylococcus aureus. BactoTM Pseudomonas Isolation Agar combined with Fisher Scientific Glycerin (CAS 56-815) was used as the culture media for Pseudomonas aeruginosa. DifcoTM MacConkey Agar Base combined with DifcoTM Lactose was used for the differentiation of Escherichia coli based on fermentation reaction. The procedure for preparing the agar plates is the same regardless of the media used. The specified agar was mixed with deionized water and autocl aved for 15 minutes to melt the agar. The container was cooled in a water bath of 50C for about an hour. The agar is poured onto the plates, about 20 mL per plate, to cover the entire bottom of th e plate. The plates remain at room temperature overnight to harden and are stored at 4C in the refrigerator. Each of the three bacteria cultures, Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa, were grown in 3% TSB for 12 to 18 hours at 37C in an incubator. The cultures were concentrat ed and washed twic e by centrifugation (10 minutes at 10,000 rpm) with autoclaved dei onized water. After each centrifugation, the supernatant was poured off and autoclaved de ionized water was added to the culture. The solution was mixed together t horoughly using a Fisher Scien tific Genie 2 Vortex mixer. The titer of the bacteria was about 108 CFU/mL. After diluting with the autoclaved nanopure water for the experiment solution, the titer was dilu ted to between 106-107 CFU/mL for each of the bacteria.

PAGE 54

44 Recirculation experiments. The internal porosity of the silica-titania composite could provide capacity for adso rption of the microorganisms. In order to determine this capacity, the test solutions cont aining the three bacteriophages, X-174, PRD-1, MS-2, were recirculated throughout the system for four hours. Samples were taken initially and after the four hours without UV light. Samples we re taken at the influent to the reactor (Port 1) and, twelve minutes later, at the effluent of the reactor (Port 2). All of the samples (10 mL) were injected with 0.5 mL of a 3% TSB soluti on for stabilization and kept refrigerated until analysis. The silica-titania composites have an acidi c characteristic in the unused state. Since the pH of the solution can influence th e survival and adsorption of the phages, the solution was buffered to keep it at a nomi nal pH. The first bacteriophage run was buffered using a 1 mM solution of potassium phosphate, and the pH of the solution was monitored throughout the experiment. The initia l stock solution starte d with a pH of 7.0 and the final sample taken at the effluent had a pH of 6.2. After the four-hour adsorpti on step, the UV lamp was turned on and samples were taken after 1.5 hours and 2.5 hours of recircula tion with UV irradiation. The silica-titania composites were also tested for desorption at the conclusion of th e reactor experiment. Three pellets were taken from the reactor and submersed in 5 mL of 3% TSB solution until analysis. The amount of time the soluti on spent in the reactor was very small compared to the overall time of the experiment. Based on th e flow rate and the volume of the reactor system, the solution spent 33.6 minutes in the reactor during the 4 hrs of recirculation. During the 1.5 hours of recirculation with UV light, the solution had spent 50.4 minutes

PAGE 55

45 in the reactor with 12.6 minutes exposed to UV light. During the 2.5 hours of recirculation with UV light, the solution spent a total of 58.8 minutes in the reactor with 21 minutes of UV irradiation. Batch experiments. In order to examine the eff ects of longer UV contact time, batch experiments were performed. The bu ffering solution was changed from potassium phosphate to sodium bicarbona te (100 mg/L as CaCO3); since it is more likely that bicarbonate will be present in the NASA wastewater system. Adsorption. As in the previous experiments, th e bacteriophages were recirculated throughout the system for an initial adsorpti on step. After four hour s of recirculation without UV light, the effluent wa s sampled. After six hours of recirculation in the dark, a sample was taken before the reactor (Port 1), the flow of the system was stopped, and the solution remained in batch conditions for one hour. The flow was started and a sample was taken at the effluent (Por t 2) after five minutes to ensure the sample was from the batch conditions. Inactivation. Batch conditions were repeated in the same manner as in the adsorption procedure for the inactivation step s of the experiment. Samples were taken before the reactor, the flow of the system was stopped and the UV lamp was turned on inside the reactor. After 1 hour, the flow wa s restarted and a sample was taken at the effluent. The solution was then recirculated through the system for one hour without UV light to bring the system to a steady state. The batch condi tions were repeated for an additional 2 hours. Three experiments were pe rformed with the bacteriophages using the batch procedure, including one with a 365-nm wavelength UV lamp, one with a 254-nm UV lamp, and one without the silica-titania composites inside the reactor (using the 254-

PAGE 56

46 nm lamp). Table 9 presents the effluent pH values for each of the three batch bacteriophage experiments. The solution was held at a nominal pH by the 100 mg/L as CaCO3 added. Table 9. Effluent pH values for solutions containing sodium bicarbonate (100 mg/L as CaCO3) along with spiked bacteriophages Experiment ID Effluent pH (t = 0) Effluent pH (t = 9) 365nm lamp 6.9 6.65 254nm lamp 6.94 6.67 254nm lamp (no pellets) 8.18 7.50 Three experiments were also performed w ith bacteria using the batch procedure, including two experiments w ith the 254-nm wavelength UV lamp and one without the silica-titania composites (using the 254-nm la mp). Table 10 presents the effluent pH values for each of the batch bacteria experiments. Table 10. Effluent pH values for solutions containing sodium bi carbonate (100 mg/L as CaCO3) along with spiked bacteria Experiment ID Effluent pH (t = 0) Effluent pH (t = 9) 254nm lamp 7.07 6.48 254nm lamp 7.09 6.46 254nm lamp (no pellets) 8.11 7.84 As with the bacteriophages, the solution was held at a nominal pH by the addition of 100 mg/L as CaCO3. Statistics A Poisson distribution event occurs at ra ndom in continuous space or time. Events occur independently, uniformly, and singly. The number of coloni es growing on agar plates in a dilution plating assa y is an event that should be ra ndomly distributed in space. The unit is the volume of sample, which contains organisms held in suspension, on the plate. If the sample is well-stirred, the part icular organisms of interest are distributed

PAGE 57

47 through the sample of liquid at random and independently Since the total number of organisms per unit volume in the sample is constant, there is a constant probability of counting a particular type of organism in plates of a given volume (Clarke and Cooke, 1992). Provided that the cells are randomly di stributed and that the count is reasonably high (greater than 30), then the count conforms to a Poisson distribut ion. This applies to all the counts, so no degrees of freedom need to be calculated, the StudentÂ’s t-value used for the statistical tests is for an infinite degree of freedom (Deacon, 2005). Since the count can be treated as part of a Poisson distribution, the microbial concentration of colonies or plaques per m illiliter of sample (CFU /mL or PFU/mL) has a value of the number of colonies counted multiplied by the dilution of the plate. A comparison of two Poisson counts can be made by using a hypothesis test for significant difference. The t-value for the two counts is calculated by Equation 14 and compared to the critical t-value (the StudentÂ’s t-value for an infinite degree of freedom). If the calculated t-value is highe r than the critical t-value, then the two Poisson counts are significantly different. The critical t-value for a probability of 0.01 is 2.58. The t-value is calculated using the following formula (Deacon, 2005): ) ( ) ( ) ( 5 0 ) ( ) (2 1 2 2 1 1 2 1 2 1 1 2 1 1V V V V V V X X V V V X X X t (14) where X1 = first plate count, X2 = second plate count, V1 = volume of sample on first plate, and V2 = volume of sample on second plate.

PAGE 58

48 CHAPTER 4 RESULTS Oxidation of Organic Compounds Adsorption The internal porosity and hi gh surface area of the silicatitania composite provide significant adsorption capacity for some of the contaminants. This adsorption, likely physisorption resulting from van der Waals fo rces for organic compounds, is considered to be advantageous to the oxi dation process in collecting the target contaminants at the surface of the titanium dioxide where th e reactive hydroxyl radicals are produced. The majority of the oxidation is likely to occur near the surface of the silica-titania composite, therefore, the adsorbed compounds will tran sport from the inner pores of the silica structure to the external portion for oxidation. In order to determine the extent of photo catalytic destruction of the contaminants, test solutions were recycled through the reactor at 10 mL/min for five hours in the dark, without UV radiation. Holmes (2003) f ound the reactor influent and effluent concentrations were nearly equal after five hours of recy cling the test solution through the reactor system. The representative resu lts for the adsorption of chlorobenzene and toluene are shown in Table 11. Chloroben zene and toluene are th e two compounds with the highest octanol-water pa rtition coefficients. All of the adsorption data for each experiment can be found in Appendix A.

PAGE 59

49 Table 11. Removal of toluene and chlorobenzene in attaining adsorption equilibrium in the reactor. Solution recycled through r eactor with no UV radiation for five hours (Holmes, 2003) Contaminant Initial Concentration % Removed by Adsorption Toluene 76 g/L 38 Chlorobenzene 121 g/L 45 Oxidation All photocatalytic destruction experiments were preceded by recirculating the test solution for five hours to allow adsorption e ffects to stabilize as indicated above. The flow rate through the reactor during the series of experimental runs remained at an optimized value of 10 mL/min, with an EBCT of 43.6 minutes (Holmes, 2003). Holmes also found that for this designed reacto r the photocatalytic destruction of the contaminants was not limited by the number of lamps used beyond the use of two lamps; therefore, following the preliminary adsorption step, two UV lamps were turned on and allowed to warm up while they were shielded from the reactor. One hour was allowed for steady-state operation to be atta ined before samples were collected at the inlet and outlet of the reactor. An increase in the number of UV lamps us ed could cause the temperature of the solution in the reactor to rise. This temperat ure change could have an effect on the UV output. There is an optimal temperature to ma ximize the UV intensity of the lamps. If the temperature changes either above or be low this optimal temperature, then the UV intensity is less than the maximum po ssible for a given number of lamps. Replication. Two initial oxidation experiments were conducted by spiking the selected SOCs into nanopure water. Preceded by the adsorption step, the solution was irradiated by a 365-nm UV lamp and run th rough the system in a single-pass mode. The percent removal of each of the SOCs for the two duplicate experiments and a similar

PAGE 60

50 experiment performed by Holmes (2003) ar e shown in Figure 9. The silica-titania composites were not removed from the react or and were the same pellets used for experiments performed by Holmes. The pellets remained in the reactor throughout the organic experiments performed in this research. Oxidation of SOCs Single Pass at EBCT = 43.6 min0 10 20 30 40 50 60 70 80 90 100TolueneChlorobenzeneAcetoneCarbon Disulfide Ethyl AcetateMethyl Methacrylate% Removed Test 1 Test 2 Holmes Experiment Figure 9. Percent removal of spiked SOCs in a single pass through the reactor with UV irradiation and without addition of alkalinity The percent removals represented in Fi gure 9 are for the si ngle-pass oxidation step and do not include the removal by adsorp tion. The removal achieved by the first two experiment runs were either greater than or equal to the percent removal seen by Holmes in a similar experiment set-up with the orga nic reactor. These resu lts indicate that the silica-titania pellets in the r eactor were still capable of oxi dizing the organic contaminants after multiple experiments (15 performed by Holmes). The methyl methacrylate data for the firs t two runs were not reported because the solution spikes for these experiments we re not successful in producing influent concentrations above 15 g/L. Therefore, th e removal was not stat istically acceptable due to the low range of values.

PAGE 61

51 Alkalinity. The initial oxidation experiments were followed by experiments with elevated levels of bicarbonate ion concentra tion. Photocatalytic oxidation may be limited by bicarbonate ions, which can scavenge hydroxyl radicals resulting in unreactive bicarbonate radicals. The effect of bicarbonate ions present in solution was investigated by adding differing amounts of sodium bicarbon ate to increase the alkalinity of the solution and buffer the pH. In initial experi ments, without bicarbonate buffering, the typical pH exiting the reactor system was in the range of 3.8 – 4, because of the acidic nature of the pellets. Experiments with solutions containing a buffering concentration of 80 mg/L of NaHCO3 (48 mg/L as CaCO3), 200 mg/L of NaHCO3 (120 mg/L as CaCO3), and 3965 mg/L of NaHCO3 (2360 mg/L as CaCO3) were performed. The 2360 mg/L as CaCO3 alkalinity value was chosen because this wa s the expected alkalinity of the typical NASA wastewater (formula provided by Johnson Space Center is in Chapter 3: Materials and Methods). The wastewater feed was created using a formula for typical wastewater. The pH values for each of the experiments contai ning bicarbonate ions are presented in Table 12 as well as the pH values for no bicarbonate present. Table 12. pH values for spiked organic contaminant solutions containing differing amounts of sodium bicarbonate. Effluent pH va lue is after five hours of recirculation through the packed reactor. Experiment Solution Influent pH Effluent pH 0 mg/L NaHCO3 7.0 3.8 80 mg/L NaHCO3 6.6 4.5 200 mg/L NaHCO3 6.5 5.7 3965 mg/L NaHCO3 8.2 7.5 As expected, as the concentration of s odium bicarbonate increased, the ability of the solution to buffer the pH increased. The sodium bicarbonate addition provided a

PAGE 62

52 solution that was more realisti c and closer in composition to the expected components of the NASA wastewater. The amount of alkalinity expected to enter the post processor, according to a revised wastewater formulation, has a nominal value of 58 mg/L as CaCO3 and worst-case value of 75 mg/L as CaCO3 (Verostko et al., 2004). Both of these values lie between the 80 mg/L as NaHCO3 (48 mg/L as CaCO3) and 200 mg/L as NaHCO3 (120 mg/L as CaCO3) solutions tested. The percent removal of each of the SOCs for the experiments including bicarbonate al kalinity is shown in Figure 10. 0 20 40 60 80 100 TolueneChlorobenzeneAcetoneCarbon Disulfide Ethyl AcetateMethyl Methacrylate% Removal 0 mg/L NaHCO3 80 mg/L NaHCO3 200 mg/L NaHCO3 3965 mg/L NaHCO3 Figure 10. Percent removal of six target anal ytes in solution cont aining differing amounts of bicarbonate operating in single-pass mode after five hours of adsorption in the recirculation mode (error bars indicate one st andard deviation on either side of the mean) The experiment runs without bicarbonate addition are represented in Figure 10 (“0 mg/L NaHCO3”) by the average of the two runs In order to test the organic contaminant experiments for significant di fference, each chemical was considered independently. A one-way analysis of varian ce (ANOVA) test was used to compare the experiments. In order to use the ANOVA test, the data needed to first be tested for equal variances. The percent removals for each of th e experiments were tested using Bartlett’s

PAGE 63

53 and LeveneÂ’s tests for equal variances. Th e p-value calculated for toluene was 0.152. Since this value is greater than 0.05, equal variances can be assume d and an analysis of variance (ANOVA) can be performed. Each of the six analytes had a p-value above 0.05 (Appendix D). The ANOVA p-value calculated for toluene was 0.001, less than 0.05, indicating that at least one of the experiment s was significantly different than the others. The high bicarbonate concentration experime nt (3965 mg/L) was signi ficantly different than all of the other e xperiments (0 mg/L NaHCO3, 80 mg/L NaHCO3, 200 mg/L NaHCO3) for toluene and chlorobenzene. The high bicarbonate experiment was significantly different than no bicarbonate for ethyl aceta te and both no bicarbonate and 200 mg/L of NaHCO3 for carbon disulfide. For acetone and methyl methacrylate, all of the bicarbonate experiments were not signi ficantly different. A ll of the oxidation experiment data are in Appendix B, and the ANOVA tables can be found in Appendix C. The addition of bicarbonate ion did not ha ve an appreciable effect on the removal efficiency at the two lower levels of 80 and 200 mg/L of NaHCO3. There was no significant statistical difference in the per cent removed between th e experiments with 0 mg/L 80 mg/L and 200 mg/L of NaHCO3. A decrease in the percent of the contaminants removed can be seen in the high alka linity experiment of 3965 mg/L NaHCO3. The decreased removal can be attributed to the scavenging of hydroxyl radicals by the abundance of bicarbonate ions in the solu tion. A decreased removal for the high bicarbonate concentration is not seen with ethyl acetate or methyl methacrylate. The results shown in Figure 10 indicate a substantial destructi on of most of the target analytes. Removal of acetone was less e ffective, although analytical precision was outside of the acceptable QA/QC range for acetone (relative standard deviation of

PAGE 64

54 replicate injections <20%). Some methyl meth acrylate data were not reported because of solution spikes for these experiments not being successful in producing influent concentrations above about 15 g/L. Low dissolved oxygen. An experiment simulating a low dissolved oxygen (DO) environment was performed in order to test the effectiveness of the oxidation of the organic contaminants. The low DO envir onment was achieved by filling the reactor system with nitrogen-sparged nanopure water. The influent DO level in the reactor system was 0.99 mg/L, and the effluent DO was 1.34 mg/L. After one liter of solution was wasted to obtain a steady state condition, the DO at the inlet of the reactor was measured to be 2.67 mg/L and 1.65 mg/L DO at the effluent of the reactor. The elevated dissolved oxygen levels prior to oxidation we re attributed to oxyge n dissolution from trapped air in the reactor and pellets. Th e loss of dissolved oxygen during oxidation was attributed to use in the phot ocatalytic reaction. The earlier experiments were presumably at DO saturation of 9 mg/L; dissolved oxyge n measurements were not taken for the earlier experiments. It was expected that with a lowered di ssolved oxygen concentration, that less hydroxyl radicals would be available for photo catalysis. The solution also contained 200 mg/L of NaHCO3. The results, shown in Figure 11, indi cate that the percent removed still remained above 80% for toluene, chlor obenzene, carbon disulfide, and methyl methacrylate. The removal of acetone remain ed lower than the other contaminants.

PAGE 65

55 0 20 40 60 80 100 TolueneChlorobenzeneAcetoneCarbon Disulfide Ethyl AcetateMethyl Methacrylate% Removal 0 mg/L NaHCO3 80 mg/L NaHCO3 200 mg/L NaHCO3 3965 mg/L NaHCO3 Low DO (2.6 mg/L) Figure 11. Percent removal of six target anal ytes operating in singlepass mode after five hours of adsorption in the recirculation mode (error bars indicate one standard deviation on either side of the mean) The low DO experiment was compared to the previous experiments using a oneway ANOVA statistical test. For chlorobenzene, acetone, carbon disulfide, and ethyl acetate the low DO run percent removals fall wi thin the standard de viation of the earlier results. There is no significant statistical difference between the low DO experiment and all of the previous bicarbonate experiments for these compounds. For toluene, the low DO experiment was statistically different from the “0 mg/L NaHCO3” run and the “200 mg/L NaHCO3” run but was not signifi cantly different from the “80 mg/L NaHCO3” experiment. For methyl methacrylate, the low DO experiment was statistically different from the “80 mg/L NaHCO3” run and the “3965 mg/L NaHCO3”. All of the ANOVA tables are in Appe ndix C, and the experimental data for the oxidation of the organic compou nds can be found in Appendix B. One possible explanation for the low DO environment not having an impact on the degradation of the organic contaminants stems from the role of acetone. Acetone and

PAGE 66

56 other alcohols are electron hol e scavengers. By acetone directly reacting with the electron holes, the role of the electron acceptor (oxygen) is diminished. Oxygen is important in the photocatalytic oxidation process for combini ng with electrons to prevent electron and electron hole recombination, but if the elec tron hole is being occupied by acetone (or another substance), the role of oxygen is less important. TOC destruction in synthetic wastewater. The next experiment performed involved testing the simulated NASA wastewat er for total organic carbon (TOC). The simulated NASA wastewater was passed thr ough the system as in the previous experiments with five hours of adsorption be fore the single-pass r un through the reactor. The initial TOC concentration of the solu tion was 2.73 mg/L. After the five hours of adsorption, the TOC concentration before th e reactor was 2.36 mg/L. After the singlepass through the reactor, the TOC concentration after the reactor was 1.44 mg/L. The percent removals of TOC during the two stages of the experiment are shown in Table 13. Table 13. Percent removal of to tal organic carbon operating in the single-pass mode after five hours of adsorption in the recirculation mode Mode of Removal % TOC Removed Adsorption 13.5 Oxidation 38.8 Total 47.0 The amount of TOC adsorbed to the s ilica-titania composites was calculated by multiplying the volume of water passed thr ough the system (3 L) by the difference in concentration during the adso rption recirculation step (2 .73 – 2.36 mg/L). The TOC adsorbed by the pellets was 1.01 mg. Afte r the experiment, nanopure water was passed through the reactor to te st for the desorption of TOC from the pellets after contact with the wastewater. The desorpti on concentrations were plo tted against the amount of

PAGE 67

57 nanopure water passed through the system. In Figure 12, the data was fitted by a polynomial trend line. The data appears to match a polynomial trend for the initial desorption between 0.03 L and 0.23 L. Desorption Polynomial Trendy = 13.698x2 9.8324x + 1.8421 R2 = 10.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 00.10.20.30.40.5 Volume passed (L)TOC Concentration (mg/L) Figure 12. Polynomial trend line of desorbing TOC concentration of simulated wastewater versus the amount of na nopure water passed through the system The TOC desorbed was calculated as the area under the curve between 0 and 0.23 to be 0.23 mg In Figure 13, the data was fitted by a exponential trend line. The data appears to match the exponential trend fo r the desorption between 0.23 L and 0.43 L. Desorption Exponential Trendy = 0.7075e-3.6543xR2 = 10.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 00.10.20.30.40.5 Volume passed (L)Concentration (mg/L) Figure 13. Exponential trend line of desorbing TOC concentration of simulated wastewater versus the amount of na nopure water passed through the system

PAGE 68

58 The TOC desorbed was calculated as th e area under the curve between 0.23 and infinity to be 0.06 mg. The total calculat ed TOC desorbed from the pellets in the nanopure water rinse was 0.29 mg as compared to the 1.1 mg adsorbed to the pellets. The remaining TOC could have remained adsorbed to the pellets or have been oxidized along with the non-adsorbed TOC. The mass of TOC removed in the single pass with UV irradiation was 2.75 mg. All of the TO C data is contained in Appendix E. Inactivation of Bacteria and Viruses Recirculation Experiments In order to determine the capacity of the silica-titania composites for attachment, the test solutions containing the th ree bacteriophages, MS-2, PRD-1, X-174, were recirculated throughout the syst em for four hours. Samples we re taken initially and after the four hours without UV light. Samples were ta ken at the influent to the reactor (Port 1) and after twelve minutes, or the approxima te mean hydraulic residence time, at the effluent of the reactor (Port 2). After the four-hour preliminary step without UV radiation, the UV lamp was turned on and samples were taken after 1.5 hour s and 2.5 hours of reci rculation (t = 6 and t = 7, respectively). The log10 concentrations of phage in the samples are shown in Figure 9. The units of concentration are in plaque-forming units (PFU) per milliliter (mL). The silica-titania composites were also tested fo r desorption at the conc lusion of the reactor experiment. Three pellets were taken from th e reactor and submersed in 5 mL of 3% TSB solution until analysis. The “pellets” column in Figure 14 represents this desorption or detachment data.

PAGE 69

59 UV turned on at t = 4.50.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Initial Port 1 Initial Port 2 t=4 Port 1 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 PelletsLog Conc (PFU/mL) Phi X PRD1 MS2 Figure 14. Bacteriophage adsorption and inactiv ation experiment in a recirculation mode with a potassium phosphate buffer N represents the concentration of bacteriophage in the sample. N0 represents the initial phage concentration. The log removal of the phage in the sample can be calculated by subtracting the log of the concentration from the log of the initial concentration (log N0 – log N). Another way of expressing this removal is by applying a log transformation to (log N0 – log N) to achieve log[N0/N]. For Figure 14, the total log [N0/N] values were calculated by subtracting the log of the “t=7 Port 2” value from the log of the “Initial Port 1” value. The total log [N0/N] values were 1.4, 0.0, and 0.82 for X-174, PRD-1, and MS-2, respectively. The log [N0/N] values resulting from UV irradiation were 1.16, 0.0, and 0.67, respectively. The amount of time the soluti on spent in the reactor was very small compared to the overall time of the experiment. Based on th e flow rate and the volume of the reactor system, the solution spent 33.6 minutes in the reactor during the 4 hrs of recirculation. During the 1.5 hours of recirculation with UV light, the solution had spent 50.4 minutes in the reactor with 12.6 minutes exposed to UV light. During the 2.5 hours of

PAGE 70

60 recirculation with UV light, the solution spent a total of 58.8 minutes in the reactor with 21 minutes of UV irradiation. Batch Experiments In order to investigate th e effects of longer UV contact time, batch experiments were performed. Assuming the UV in tensity on the reactor is 8 mW/cm2, as measured at the outer diameter distance of the reactor wi thout pellets or the r eactor, the UV dose in the batch system was 28.8 J/cm2 and 57.6 J/cm2 for the 60 minute and 120 minute irradiation times, respectively. The pH buffering solution was changed from potassium phosphate to sodium bicarbonate (100 mg/L as CaCO3); since it is more likely th at bicarbonate will be present at these concentrations in the NASA wastewater system as compared to phosphate. Adsorption/Attachment. As in the previous experiments, the bacteriophages were recirculated throughout the system for preliminary adsorption/attachment step. After four hours of recirculation wit hout UV light, the effluent was sampled. After six hours of recirculation in the dark, a sample was taken before the reactor (Port 1), the flow of the system was stopped, and the solution remained in batch conditions for one hour. The flow was started and a sample was take n at the effluent (Port 2) after five minutes to ensure the sample was representative of the batch c ondition. A representati ve result from the adsorptive step of the batch experiment is shown in Figure 15.

PAGE 71

61 Bacteriophage PRD-10.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Initial Port 1Initial Port 2t=4 Port 2t=6 Port 1t=6 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Figure 15. Duplicate sample analysis of adso rption/attachment removal of PRD-1 for the batch bacteriophage experiment with no UV light Each of the samples taken from the system was plated in replicate assays on two separate occasions, represented by the two bars in Figure 15. The second assay was performed following the analysis of the firs t assay, 12 – 24 hours la ter, to ensure the proper dilutions were being plated in the analysis. In order to determine if the replicates were significantly different, a hypothesis t-te st for Poisson counts was performed (as explained in the statistics section of Chapte r 3). The critical t-value for significance was 2.58 at a probability level of p = 0.01. The cal culated t-values for the replicates shown in Figure 15 were 2.02, 0.30, 0.25, 0.97, and 1.63 for each of the respective samples. Since all of these values fell below the critical t-value, they were assumed to be not significantly different. These results suggest that over the 24hour period the PRD-1 concentration in the sample did not change. It also supports the QA/QC of the analytical method by the replication of data. The results for the bacteriophage MS-2, shown in Figure 16, are from the adsorption/attachment period of the same ba tch experiment as the data in Figure 15.

PAGE 72

62 Bacteriophage MS-20.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Initial Port 1Initial Port 2t=4 Port 2t=6 Port 1t=6 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Figure 16. Duplicate sample analysis of adso rption/attachment rem oval of MS-2 for the batch bacteriophage experiment with no UV light As with the PRD-1, each of the samples taken from the system was plated in replicate assays on two separate occasions, 12 – 24 hours following the analysis of the first assay, and are represented by the two bars in Figure 16. The calculated t-values for the replicates shown in Figure 16 were 1.66, 4.83, 3.99, 6.26, and 4.74 for each of the respective samples. Most of these values fell above the critical t-value (2.58), the replicates were significantly different for a ll of the MS-2 samples in Figure 16 except for “Initial Port 1.” Since there was a difference in replicates for the MS-2 phage, the first and second platings of the MS-2 samples were compar ed using a two-sample paired t-test to determine whether there was a significant difference between th e two occasions of plating. The null hypothesis for th e test was that the mean diffe rences of the two platings were the same. The p-value calculated fo r this test was 0.005, therefore the null hypothesis was rejected. These results suggest that over the 24-hour period the MS-2 concentration in the sample decreased, but th is difference remained consistent throughout

PAGE 73

63 all of the samples. The results also support the QA/QC of the anal ytical method. Since the mean difference between the two platings was not the same, the results should be compared per plating occasion, i.e., the “Ini tial Port 1 First Assay” should be compared with the “Initial Port 2 Fi rst Assay” and not the “In itial Port 2 Second Assay.” In order to determine whether there was a significant amount of adsorption/attachment or other loss of phage during the preliminary recirculation without UV radiation, the MS-2 plate c ounts for “Initial Port 1” and “t=6 Port 2” were compared using a t-test. The t-value calculated for the “F irst Assay” in Figure 16 was 0.46, which is less than the critical t-value of 2.58 (p = 0.01), indicating that the two counts are not significantly different. Therefore, there is littl e adsorption or attachment of the phages in the six hours of recirculation and in th e one hour of batch conditions at t = 6. Similar results were obtained for the adsorption of Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa in the batch experiments performed with the three bacteria. The adsorptive data and statistical values for all of the bacteriophage and bacterial experi ments can be found in Appendix F. Inactivation. The batch conditions were repeated in the same manner as in the adsorption/attachment procedure. The so lution flow was stopped and UV irradiation applied for a one-hour (t = 7) inactivation experiment. After sampling, the solution was recirculated throughout the syst em for one hour. The solution flow was then stopped and an additional two hours of UV irradiation was applied for the two-hour (t = 9) inactivation experiment. When calculating the log [N0/N] inactivation values, N0 was the concentration at Port 1 (before the reactor) a nd N was the concentration at Port 2 (after

PAGE 74

64 the reactor). For the additional tw o-hour inactivation experiment, N0 was the Port 1 concentration taken just before the two-hours of continuous UV. Three experiments were performed with the bacteriophages using the batch procedure, including one with a 365-nm wavelength UV lamp, one with a 254-nm UV lamp, and one without the silica-titania co mposites inside the r eactor (using the 254-nm lamp). The effluent pH values for each of the batch bacteriophage experiments are shown in Table 14. Table 14. Effluent pH values for solutions containing sodium bi carbonate (100 mg/L as CaCO3) along with spiked bacteriophages The solution was held at a near neutra l pH by an addition of sodium bicarbonate (100 mg/L as CaCO3) added to the solution. A representative plot of the inactivation results for the bacteriophages is shown in Figure 17. Bacteriophage PRD-1 0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 t=7 Port 1t=7 Port 2t=9 Port 1t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Figure 17. Duplicate sample analysis of inactivation of PRD-1 for the batch bacteriophage experiment with the 254 nm UV lamp Experiment ID Effluent pH (t = 0) Effluent pH (t = 9) 365nm lamp 6.9 6.65 254nm lamp 6.94 6.67 254nm lamp (no pellets) 8.18 7.50

PAGE 75

65 The samples labeled “t = 7” were the ba tch conditions of one hour of continuous UV irradiation. The samples labeled “t = 9” were the batch conditions for two hours of continuous UV irradiation. As with the adsorp tion/attachment data, the sample replicates for the PRD-1 phage are not significantly diffe rent. Each of the replicates in Figure 17 was compared using a hypothesis t-test for Po isson counts. The t-values calculated for each of the pairs of replicates were 0.15, 2.36, 2.20, and 1.40, which are less than the critical t-value of 2.58 (p = 0.01), indicating that the replicate counts are not significantly different. The log [N0/N] values for PRD-1 resulting from UV irradiation for one and two hours were 0.78 and 1.43, respectively. The log [N0/N] values for the three bacteriophage experiments are shown in Table 15. The bacter iophage concentration data and statistical values for each of the three expe riments can be found in Appendix F. Table 15. Log [N0/N] inactivation of bacteriophage unde r UV radiation for three different experimental conditions X-174 PRD-1 MS-2 Irradiation Time 1 hour 2 hours 1 hour 2 hours 1 hour 2 hours 365nm lamp 0.72 0.70 0.10 0.28 0.16 0.95 254nm lamp 0.85 1.67 0.78 1.43 0.85 1.65 254nm lamp (no pellets) 0.65 2.58 0.31 1.98 0.57 1.92 The 254-nm lamp was superior in inactiva tion of the phages as compared to the 365-nm lamp for both the oneand two-hour batch conditions. The phenomenon was expected since the peak wavelength that mi croorganisms absorb UV light is about 265 nm. In the 254-nm experiment without pellets by increasing the time the phages were exposed to UV radiation, the inactivation incr eased. This could be attributed to the increased UV dose or the “concentration” (i.e., UV intensity) multiplied by the time. This

PAGE 76

66 is analogous to the CT measured in chlorine disinfection. The UV intensity remained the same for the two experiments, but the time was increased. In the 254-nm experiment with pellets by increasing the time the phages were exposed to UV radiation, the inactivation increa sed. This could also be attributed to the increased UV dose and inactivation by direct exposure to UV light. In comparing the one-hour inactivation between 254-nm experiments with and without pellets, the inactivation increased wh en the pellets were present for all three phages. The hydroxyl radicals formed aided in the inactiva tion but only for the shorter time period. Since the hydroxyl radical re actions for inactivation occur almost instantaneously, the increased time (to two hours) did not have an impact on inactivation by hydroxyl radicals. So, for the two-hour e xperiment with pellets present, the inactivation was increased from the one-hour experiment with pellets because of the direct exposure to UV light and not the hydroxyl radical interactions. In comparing the two-hour inactivati on between 254-nm experiments with and without pellets, the inactiva tion decreased when the pellet s were present for all three phages. The pellets would aid in the inactiv ation of the phages to a point, but they may also block some of the direct UV light, wh ich would decrease the UV intensity seen by the phages. Ultimately, the results seem to show that the blockage of UV light is out competing the aiding of inactiv ation by the hydroxyl radicals. Three experiments were also performed w ith selected bacteria using the batch procedure, including two experiments with the 254-nm wavelength UV lamp and one without the silicatitania composites (using the 254-nm lamp). Table 16 presents the effluent pH values for each of the batch bacteria experiments.

PAGE 77

67 Table 16. Effluent pH values for solutions containing sodium bi carbonate (100 mg/L as CaCO3) along with spiked bacteria Experiment ID Effluent pH (t = 0) Effluent pH (t = 9) 254nm lamp 7.07 6.48 254nm lamp 7.09 6.46 254nm lamp (no pellets) 8.11 7.84 As with the bacteriophages, the solution was held at a nominal pH by the addition of 100 mg/L as CaCO3. The log [N0/N] values for the three bacteria experiments are shown in Table 17. The bacteria concentrati on data for each of the three experiments can be found in Appendix F. Table 17. Log [N0/N] inactivation of bacteria un der UV radiation for different experimental conditions E. coli S. aureus P. aeruginosa Irradiation Time 1 hour 2 hours 1 hour 2 hours 1 hour 2 hours 254nm lamp 0.89 1.00 1.12 1.40 1.04 1.44 254nm lamp 1.41 1.7 1.11 1.62 254nm lamp (no pellets) 1.97 2.53 2.04 1.7 1.24 For the second experiment w ith the 254-nm lamp, the Staphylococcus aureus was not present in the initial stock solution. All of the bacterial results, except for the Pseudomonas aeruginosa 2-hour batch condition, had an increased inactivation when the silica-titania composites were absent. This re sult can be attributed to the silica pellets decreasing the UV irradiation of the microor ganisms by blocking the UV radiation. This blockage is counteracting th e inactivation of the bacter ia by the hydroxyl radicals.

PAGE 78

68 CHAPTER 5 SUMMARY AND CONCLUSION Summary An annular reactor containing silica-titani a composites arranged in a packed bed was capable of degrading all six target analytes, including acet one, carbon disulfide, chlorobenzene, ethyl acetate, methyl methacr ylate, and toluene, in the presence of bicarbonate ions and in a low (2.6 mg/L ) dissolved oxygen environment. Similar degradation results were obtained for alkalinity values of 80 mg/L NaHCO3 (48 mg/L as CaCO3) and 200 mg/L NaHCO3 (120 mg/L as CaCO3) and solutions wi thout alkalinity present as well as solutions in a low dissolved oxygen environment of less than 3 mg/L. At higher alkalinity (2360 mg/L as CaCO3) the degradation of contaminants decreased by between 15-20% for toluene, chlo robenzene, and carbon disulfide. A vertical annular reactor also containi ng the silica-titania co mposites arranged in a packed bed was capable of achieving log inactivation [N0/N] values of 1.67, 1.43, and 1.65 for bacteriophages X-174, PRD-1, and MS-2, resp ectively, after two hours of irradiation with a 254-nm UV lamp. Log inactivation [N0/N] values for the bacteria Escherichia coli, Staphylococcus aureus, and Pseudomonas aeruginosa of 1.70, 1.40, and 1.62, respectively, were achieved under similar conditions.

PAGE 79

69 Conclusion The amount of alkalinity expected to en ter the finishing process in the ALS treatment train, according to a revised NAS A wastewater formulation, has a nominal value of 58 mg/L as CaCO3 and worst-case value of 75 mg/L as CaCO3 (Verostko et al., 2004). At these lower levels, the addition of bicarbonate ions does not have an appreciable effect on the photocatalytic oxidation of acetone, carbon disulfide, chlorobenzene, ethyl acetate, methyl methacr ylate, and toluene in the reactor using titania-doped porous silica pellet s. With higher alkalinities, representative of the raw NASA wastewater of around 2360 mg/L as CaCO3, a decrease in the percent of the contaminants removed can be expected. The d ecreased removal can be attributed to the scavenging of hydroxyl radicals by the abundanc e of bicarbonate ions in the solution. A dissolved oxygen concentration in the range of 1 – 3 mg/L does not decrease the potential for photocatalytic oxidation for most of the researched organic contaminants as compared to dissolved oxygen levels n ear saturation (about 8 mg/L). Toluene and methyl methacrylate showed a decrease in removal for the lower dissolved oxygen concentration. The levels of virus and bacteria in activation achieved in the reactor are insufficient for current drinking water sta ndards, which recommend a minimum of 4-log and 3-log inactivation levels fo r viruses and bacteria, respecti vely. In order to increase the inactivation of the microbial species investigated, higher UV doses and therefore, intensities would be required. More microbial inactivation tests need to be performed to fully understand the relationship between UV intensities, time, a nd the role of the photocatalyst in the

PAGE 80

70 inactivation of both bacteria and bacteriophages. Other c ontinued research could be performed by combining the organic and microbial contaminants along with the simulated wastewater to evaluate the perf ormance of the reactors with competition between the contaminants.

PAGE 81

APPENDIX A DATA FOR ADSORPTION OF ORGANIC COMPOUNDS

PAGE 82

72TEST 1 Concentration ( g/L) C/Co % Removal 0 mg/L NaHCO3 t = 0 T = 400(%) (%) Toluene 41.0 25.5 62.1 37.9 Chlorobenzene 47.4 23.3 49.1 50.9 Acetone 324.4 196.9 60.7 39.3 Carbon Disulfide 46.8 17.2 36.7 63.3 Ethyl Acetate 45.2 31.1 68.9 31.1 Methyl Methacrylate 5.9 1.8 31.2 68.8 TEST 2 Concentration ( g/L) 0 mg/L NaHCO3 t = 0 t = 400 (Port 1) C/Co % Removal Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2Avg Std Dev (%) (%) Toluene 403.9 331.8 367.9 51.0 251.3 219.9235.6 22.2 64.0 36.0 Chlorobenzene 626.5 505.6 566.0 85.5 302.7 320.5311.6 12.6 55.0 45.0 Acetone 300.1 248.6 274.3 36.4 306.6 220.1263.4 61.1 96.0 4.0 Carbon Disulfide 1713.0 1606.6 1659.8 75.3 1272.6 757.1 1014. 9 364.5 61.1 38.9 Ethyl Acetate 149.1 135.3 142.2 9.7 118.6 92.1 105.4 18.7 74.1 25.9 Methyl Methacrylate 13.2 15.1 14.1 1.3 5.9 4.8 5.4 0.8 37.9 62.1 TEST 3 Concentration ( g/L) 80 mg/L NaHCO3 t = 0 t = 400 (Port 1) C/Co % Removal Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2Avg Std Dev (%) (%) Toluene 66.9 61.8 64.4 3.6 30.0 26.6 28.3 2.4 44.0 56.0 Chlorobenzene 71.3 72.5 71.9 0.8 24.4 21.5 23.0 2.1 31.9 68.1 Acetone 117.9 111.1 114.5 4.8 109.9 102.6106.3 5.2 92.8 7.2 Carbon Disulfide 220.9 212.4 216.7 6.0 84.1 73.4 78.8 7.6 36.3 63.7 Ethyl Acetate 129.9 138.3 134.1 5.9 52.4 45.9 49.2 4.6 36.7 63.3 Methyl Methacrylate 123.5 112.8 118.2 7.6 62.6 57.8 60.2 3.4 51.0 49.0

PAGE 83

73 TEST 4 Concentration ( g/L) 200 mg/L NaHCO3 t = 0 t = 400 (Port 1) C/Co % Removal Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2Avg Std Dev (%) (%) Toluene 132.6 98.4 115.5 24.2 70.8 70.8 61.3 38.7 Chlorobenzene 173.9 150.1 162.0 16.8 73.9 73.9 45.6 54.4 Acetone 211.6 345.2 278.4 94.5 222.7222.7 80.0 20.0 Carbon Disulfide 721.3 511.7 616.5 148.2 294.7294.7 47.8 52.2 Ethyl Acetate 67.3 94.6 81.0 19.3 49.6 49.6 61.3 38.7 Methyl Methacrylate 16.1 62.7 39.4 33.0 0.4 0.4 1.0 99.0 TEST 5 Concentration ( g/L) 3965 mg/L NaHCO3 t = 0 t = 400 (Port 1) C/Co % Removal Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2 Avg Std Dev (%) (%) Toluene 114.6 112.8 113.7 1.3 103.1 90.7 96.9 8.8 85.2 14.8 Chlorobenzene 140.8 139.0 139.9 1.3 125.0 112.9 118.9 8.6 85.0 15.0 Acetone 397.8 379.9 388.9 12.7 366.4 376.9 371.6 7.4 95.6 4.4 Carbon Disulfide 462.2 445.2 453.7 12.0 390.8 349.9 370.3 28.9 81.6 18.4 Ethyl Acetate 300.3 251.9 276.1 34.2 107.8 53.6 80.7 38.3 29.2 70.8 Methyl Methacrylate 2040.31987.92014.137.0 1352.7 925.8 1139.2301.8 56.6 43.4 TEST 6 Concentration ( g/L) Low DO Run t = 0 t = 400 (Port 1) C/Co % Removal Vial 1 Vial 2 Avg Std Dev Vial 1 Vial 2 Avg Std Dev (%) (%) Toluene 188.5 220.4 204.5 22.6 171.5 183.0 177.2 8.2 86.7 13.3 Chlorobenzene 125.7 145.2 135.4 13.8 100.4 101.8 101.1 0.9 74.7 25.3 Acetone 184.1 222.6 203.3 27.2 162.6 167.1 164.8 3.2 81.1 18.9 Carbon Disulfide 232.0 260.1 246.1 19.9 155.5 162.3 158.9 4.8 64.6 35.4 Ethyl Acetate 176.4 221.6 199.0 32.0 137.4 144.6 141.0 5.1 70.9 29.1 Methyl Methacrylate 627.5 708.5 668.0 57.3 580.3 616.3 598.3 25.5 89.6 10.4

PAGE 84

74 Adsorption / Volatilization 5 hrs in Circulation Mode0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0TolueneChlorobenzeneAcetoneCarbon Disulfide Ethyl AcetateMethyl Methacrylate% Removal Test 1No NaHCO3 Test 2No NaHCO3 80 mg/L NaHCO3 200 mg/L NaHCO3 3965 mg/L NaHCO3 Low DO (1.5 mg/L) Figure 18. Percent removal of organic compounds dur ing recirculation in th e reactor without UV light

PAGE 85

APPENDIX B DATA FOR OXIDATION OF ORGANIC COMPOUNDS

PAGE 86

76TEST 1 Concentration ( g/L) C/Co % Removed 0 mg/L NaHCO3 Port 1 Port 2 (%) (%) Toluene 25.5 1.5 5.9 94.1 Chlorobenzene 105.9 23.3 22.0 78.0 Acetone 196.9 115.7 58.8 41.2 Carbon Disulfide 98.5 17.2 17.5 82.5 Ethyl Acetate 51.1 31.1 60.9 39.1 Methyl Methacrylate 7.5 1.8 24.0 TEST 2 Concentration ( g/L) C/Co % Removal 0 mg/L NaHCO3 Port 1 V1 Port 1 V2 AVG STD DEV Port 2 V1 Port 2 V2 AVG STD DEV (%) (%) Toluene 251.3 219.9 235.6 22.2 2.9 2.3 2.6 0.4 1.1 98.9 Chlorobenzene 302.7 320.5 311.6 12.6 5.3 3.8 4.6 1.1 1.5 98.5 Acetone 306.6 220.1 263.4 61.2 193.5 96.2 144.9 68.8 55.0 45.0 Carbon Disulfide 1272.6 757.1 1014.9 364.5 27.9 20.5 24.2 5.2 2.4 97.6 Ethyl Acetate 118.6 92.1 105.4 18.7 29.5 21.3 25.4 5.8 24.1 75.9 Methyl Methacrylate 5.9 4.8 5.4 0.8 7.0 4.1 5.6 2.1 103.7 TEST 3 Concentration ( g/L) C/Co % Removal 80 mg/L NaHCO3 Port 1 V1 Port 1 V2 AVG STD DEV Port 2 V1 Port 2 V2 AVG STD DEV (%) (%) Toluene 30.0 26.6 28.3 2.4 1.9 1.7 1.8 0.1 6.4 93.6 Chlorobenzene 24.4 21.5 23.0 2.1 3.0 3.0 3.0 0.0 13.1 86.9 Acetone 109.9 102.6 106.3 5.2 53.7 75.6 64.7 15.5 60.8 39.2 Carbon Disulfide 84.1 73.4 78.8 7.6 10.5 11.7 11.1 0.8 14.1 85.9 Ethyl Acetate 52.4 45.9 49.2 4.6 10.0 6.2 8.1 2.7 16.5 83.5 Methyl Methacrylate 62.6 57.8 60.2 3.4 2.1 2.7 2.4 0.4 4.0 96.0

PAGE 87

77TEST 4 Concentration ( g/L) C/Co % Removal 200 mg/L NaHCO3 Port 1 V1 Port 1 V2 AVG STD DEV Port 2 V1 Port 2 V2 AVG STD DEV (%) (%) Toluene 70.8 70.8 2.0 4. 6 3.3 1.8 4.7 95.3 Chlorobenzene 73.9 73.9 2.7 4. 8 3.8 1.5 5.1 94.9 Acetone 222.7 222.7125.1 222. 4 173.868.8 78.0 22.0 Carbon Disulfide 294.7 294.717.7 23. 7 20.7 4.2 7.0 93.0 Ethyl Acetate 49.6 49.6 10.2 20. 3 15.3 7.1 30.7 69.3 Methyl Methacrylate 0.4 0.4 17.0 32.2 24.6 10.7 6150.0TEST 5 Concentration ( g/L) C/Co % Removal 3965 mg/L NaHCO3 Port 1 V1 Port 1 V2 AVG STD DEV Port 2 V1 Port 2 V2 AVG STD DEV (%) (%) Toluene 103.1 90.7 96.9 8.8 23.7 28.7 26.2 3.5 27.0 73.0 Chlorobenzene 125.0 112.9 118.9 8.6 38.7 41.5 40.1 2.0 33.7 66.3 Acetone 366.4 376.9 371.6 7.4 372.6 91.3 231.9198.9 62.4 37.6 Carbon Disulfide 390.8 349.9 370.3 28.9 81.8 98.0 89.9 11.5 24.3 75.7 Ethyl Acetate 107.8 53.6 80.7 38.3 0.3 0.0 0.1 0.2 0.2 99.8 Methyl Methacrylate 1352.7 925.8 1139.2301.8 0.3 25.0 12.7 17.5 1.1 98.9 TEST 6 Concentration ( g/L) C/Co % Removal Low DO Run Port 1 V1 Port 1 V2 AVG STD DEV Port 2 V1 Port 2 V2 AVG STD DEV (%) (%) Toluene 171.5 183.0 177.2 8.2 23.4 24.6 24.0 0.9 13.5 86.5 Chlorobenzene 100.4 101.8 101.1 0.9 1.5 15.6 8.6 10.0 8.5 91.5 Acetone 162.6 167.1 164.8 3.2 133.5 138.0 135.73.2 82.3 17.7 Carbon Disulfide 155.5 162.3 158.9 4.8 29.4 29.6 29.5 0.1 18.6 81.4 Ethyl Acetate 137.4 144.6 141.0 5.1 33.8 38.9 36.3 3.6 25.8 74.2 Methyl Methacrylate 580.3 616.3 598.3 25.5 91.8 97.1 94.5 3.7 15.8 84.2

PAGE 88

78HOLMES (2003) DATA Test 1 Concentration ( g/L) C/Co % Removal 0 mg/L NaHCO3 Port 1 Port 2 (%) (%) Toluene 53.4 12.1 22.6 77.4 Chlorobenzene 54.6 11.1 20.3 79.7 Acetone Carbon Disulfide 31.9 2.4 7.7 92.3 Ethyl Acetate 33.2 20.6 62.0 38.0 Methyl Methacrylate 103.3 42.9 41.6 58.4 Oxidation (Single Pass at EBCT = 43.6 min)0.0 10.0 20.0 30.0 40.0 50.0 60.0 70.0 80.0 90.0 100.0TolueneChlorobenzeneAcetoneCarbon Disulfide Ethyl AcetateMethyl Methacrylate% Removed Test 1No NaHCO3 Test 2No NaHCO3 80 mg/L NaHCO3 200 mg/L NaHCO3 3965 mg/L NaHCO3 Low DO (2.6 mg/L) Figure 19. Percent removal of organic compounds during single pass through the reac tor with UV irradiation

PAGE 89

79 APPENDIX C ANALYSIS OF VARIANCE TABL ES FOR ORGANIC COMPOUNDS TOLUENE One-way ANOVA: Percent Removal versus Experiment Analysis of Variance for Percent Source DF SS MS F P Experiment 4 867.38 216.85 21.95 0.001 Error 6 59.26 9.88 Total 10 926.65 Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev ---+---------+---------+---------+--200 mg/L 2 95.34 2.60 (----*-----) 3965 mg/L 2 72.69 6.10 (-----*----) 80 mg/L 2 93.64 0.04 (-----*----) Low DO 2 86.46 0.13 (----*-----) None 3 97.31 2.76 (---*----) ---+---------+---------+---------+--Pooled StDev = 3.14 70 80 90 100 Fisher's pairwise comparisons Family error rate = 0.222 Individual error rate = 0.0500 Critical value = 2.448 Intervals for (column level mean) (row level mean) 200 mg/L 3965 mg/L 80 mg/L Low DO 3965 mg/L 14.951 30.339 80 mg/L -5.994 -28.639 9.394 -13.251 Low DO 1.186 -21.459 -0.514 16.574 -6.071 14.874 None -8.990 -31.635 -10.690 -17.870 5.057 -17.588 3.357 -3.823 Intervals indicate significant difference when zero is not inside the interval Example: The interval for 3965 mg/L and 200 mg/L is 14.951 – 30.339. This does not contain zero, so the there is a significant difference for toluene between the 3965 mg/L experiment and the 200 mg/L experiment.

PAGE 90

80 CHLOROBENZENE One-way ANOVA: Percent Removal versus Experiment Analysis of Variance for Percent Source DF SS MS F P Experiment 4 1101.5 275.4 4.13 0.061 Error 6 400.0 66.7 Total 10 1501.4 Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev --------+---------+---------+-------200 mg/L 2 94.93 2.02 (--------*--------) 3965 mg/L 2 66.16 4.12 (-------*--------) 80 mg/L 2 86.88 1.17 (--------*--------) Low DO 2 91.59 9.82 (--------*--------) None 3 91.69 11.86 (------*-------) --------+---------+---------+-------Pooled StDev = 8.16 64 80 96 Fisher's pairwise comparisons Family error rate = 0.222 Individual error rate = 0.0500 Critical value = 2.448 Intervals for (column level mean) (row level mean) 200 mg/L 3965 mg/L 80 mg/L Low DO 3965 mg/L 8.78 48.75 80 mg/L -11.94 -40.70 28.04 -0.73 Low DO -16.65 -45.41 -24.70 23.33 -5.44 15.28 None -15.01 -43.77 -23.06 -18.35 21.48 -7.28 13.43 18.14

PAGE 91

81 ACETONE One-way ANOVA: Percent Removal versus Experiment Analysis of Variance for Percent Source DF SS MS F P Experiment 4 1229 307 0.41 0.795 Error 6 4471 745 Total 10 5700 Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev ---------+---------+---------+------200 mg/L 2 21.98 30.90 (------------*-------------) 3965 mg/L 2 37.03 54.78 (-------------*------------) 80 mg/L 2 38.73 17.55 (------------*-------------) Low DO 2 17.68 0.35 (------------*-------------) None 3 44.81 10.18 (----------*----------) ---------+---------+---------+------Pooled StDev = 27.30 0 35 70 Fisher's pairwise comparisons Family error rate = 0.222 Individual error rate = 0.0500 Critical value = 2.448 Intervals for (column level mean) (row level mean) 200 mg/L 3965 mg/L 80 mg/L Low DO 3965 mg/L -81.9 51.8 80 mg/L -83.6 -68.5 50.1 65.1 Low DO -62.5 -47.5 -45.8 71.1 86.2 87.9 None -83.8 -68.8 -67.1 -88.1 38.2 53.2 54.9 33.9

PAGE 92

82 CARBON DISULFIDE One-way ANOVA: Percent Removal versus Experiment Analysis of Variance for Percent Source DF SS MS F P Experiment 4 486.2 121.6 3.97 0.066 Error 6 183.7 30.6 Total 10 670.0 Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev -----+---------+---------+---------+200 mg/L 2 92.98 1.44 (---------*---------) 3965 mg/L 2 75.53 5.01 (---------*--------) 80 mg/L 2 85.79 2.44 (---------*--------) Low DO 2 81.44 0.50 (--------*---------) None 3 92.55 8.67 (-------*------) -----+---------+---------+---------+Pooled StDev = 5.53 70 80 90 100 Fisher's pairwise comparisons Family error rate = 0.222 Individual error rate = 0.0500 Critical value = 2.448 Intervals for (column level mean) (row level mean) 200 mg/L 3965 mg/L 80 mg/L Low DO 3965 mg/L 3.903 30.997 80 mg/L -6.357 -23.807 20.737 3.287 Low DO -2.007 -19.457 -9.197 25.087 7.637 17.897 None -11.938 -29.388 -19.128 -23.478 12.795 -4.655 5.605 1.255

PAGE 93

83 ETHYL ACETATE One-way ANOVA: Percent Removal versus Experiment Analysis of Variance for Percent Source DF SS MS F P Experiment 4 1800 450 2.38 0.164 Error 6 1133 189 Total 10 2933 Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev ---+---------+---------+---------+--200 mg/L 2 69.26 14.40 (---------*--------) 3965 mg/L 2 99.87 0.19 (---------*--------) 80 mg/L 2 83.71 3.94 (--------*---------) Low DO 2 74.26 1.63 (---------*--------) None 3 63.71 21.30 (------*-------) ---+---------+---------+---------+--Pooled StDev = 13.74 50 75 100 125 Fisher's pairwise comparisons Family error rate = 0.222 Individual error rate = 0.0500 Critical value = 2.448 Intervals for (column level mean) (row level mean) 200 mg/L 3965 mg/L 80 mg/L Low DO 3965 mg/L -64.25 3.03 80 mg/L -48.09 -17.48 19.19 49.80 Low DO -38.64 -8.03 -24.19 28.63 59.24 43.08 None -25.17 5.44 -10.72 -20.16 36.25 66.86 50.70 41.25

PAGE 94

84 METHYL METHACRYLATE One-way ANOVA: Percent Removal versus Experiment Analysis of Variance for Percent Source DF SS MS F P Experiment 2 236.01 118.01 79.30 0.003 Error 3 4.46 1.49 Total 5 240.47 Individual 95% CIs For Mean Based on Pooled StDev Level N Mean StDev -----+---------+---------+---------+3965 mg/L 2 98.64 1.90 (---*----) 80 mg/L 2 95.99 0.93 (----*----) Low DO 2 84.21 0.04 (---*----) -----+---------+---------+---------+Pooled StDev = 1.22 84.0 90.0 96.0 102.0 Fisher's pairwise comparisons Family error rate = 0.0983 Individual error rate = 0.0500 Critical value = 3.182 Intervals for (column level mean) (row level mean) 3965 mg/L 80 mg/L 80 mg/L -1.232 6.532 Low DO 10.548 7.898 18.312 15.662

PAGE 95

85 APPENDIX D TEST FOR EQUAL VARIANCE FOR ORGANIC COMPOUNDS TOLUENE Test for Equal Variances Response Percent Removal Factors Group ConfLvl 95.0000 Bonferroni confidence intervals for standard deviations Lower Sigma Upper N Factor Levels 0.92701 2.60215 415.241 2 200 mg/L 2.17394 6.10233 973.785 2 3965 mg/L 0.01511 0.04243 6.770 2 80 mg/L 0.04534 0.12728 20.311 2 Low DO 1.19914 2.76019 38.986 3 None Bartlett's Test (normal distribution) Test Statistic: 10.721 P-Value : 0.030 Levene's Test (any continuous distribution) Test Statistic: 2.500 P-Value : 0.152 CHLOROBENZENE: Test for Equal Variances Response Percent Removal Factors Experiment ConfLvl 95.0000 Bonferroni confidence intervals for standard deviations Lower Sigma Upper N Factor Levels 0.71793 2.0153 321.59 2 200 mg/L 1.46609 4.1154 656.71 2 3965 mg/L 0.41564 1.1667 186.18 2 80 mg/L 3.49897 9.8217 1567.31 2 Low DO 5.15087 11.8563 167.46 3 None Bartlett's Test (normal distribution) Test Statistic: 4.553 P-Value : 0.336 Levene's Test (any continuous distribution) Test Statistic: 0.445 P-Value : 0.774

PAGE 96

86 ACETONE Test for Equal Variances Response Percent Removal Factors Experiment ConfLvl 95.0000 Bonferroni confidence intervals for standard deviations Lower Sigma Upper N Factor Levels 11.0083 30.9006 4930.98 2 200 mg/L 19.5151 54.7796 8741.50 2 3965 mg/L 6.2523 17.5504 2800.62 2 80 mg/L 0.1234 0.3465 55.29 2 Low DO 4.4226 10.1799 143.79 3 None Bartlett's Test (normal distribution) Test Statistic: 8.759 P-Value : 0.067 Levene's Test (any continuous distribution) Test Statistic: 23.603 P-Value : 0.001 CARBON DISULFIDE Test for Equal Variances Response Percent Removal Factors Experiment ConfLvl 95.0000 Bonferroni confidence intervals for standard deviations Lower Sigma Upper N Factor Levels 0.51137 1.43543 229.059 2 200 mg/L 1.78601 5.01339 800.016 2 3965 mg/L 0.86907 2.43952 389.288 2 80 mg/L 0.17885 0.50205 80.114 2 Low DO 3.76657 8.66993 122.458 3 None Bartlett's Test (normal distribution) Test Statistic: 5.538 P-Value : 0.236 Levene's Test (any continuous distribution) Test Statistic: 0.396 P-Value : 0.805

PAGE 97

87 ETHYL ACETATE Test for Equal Variances Response Percent Removal Factors Experiment ConfLvl 95.0000 Bonferroni confidence intervals for standard deviations Lower Sigma Upper N Factor Levels 5.13131 14.4038 2298.49 2 200 mg/L 0.06801 0.1909 30.47 2 3965 mg/L 1.40311 3.9386 628.50 2 80 mg/L 0.57938 1.6263 259.53 2 Low DO 9.25312 21.2989 300.83 3 None Bartlett's Test (normal distribution) Test Statistic: 9.898 P-Value : 0.042 Levene's Test (any continuous distribution) Test Statistic: 0.549 P-Value : 0.708 METHYL METHACRYLATE Test for Equal Variances Response Percent Removal Factors Experiment ConfLvl 95.0000 Bonferroni confidence intervals for standard deviations Lower Sigma Upper N Factor Levels 0.718295 1.89505 181.440 2 3965 mg/L 0.353787 0.93338 89.366 2 80 mg/L 0.016081 0.04243 4.062 2 Low DO Bartlett's Test (normal distribution) Test Statistic: 4.411 P-Value : 0.110

PAGE 98

88 APPENDIX E DATA FOR OXIDATION OF TOC IN SIMULATED NASA WASTEWATER Single pass adsorption a nd destruction experiment EBCT = 43 min Port 1 pH (6.67 hr) = 7.00 pH initial = 8.37 Port 2 pH (6.67 hr) = 5.4 Average St Dev RSD Average Rep # (mg/L) (mg/L) (mg/L) (%) (mg/L) Port 2 1 1.5687 Vial 1 2 1.5387 3 1.4946 1.5340 0.0373 2.4297857 Port 2 1 1.3875 Vial 2 2 1.354 3 1.3279 1.3565 0.0299 2.2025213 1.4452 Port 1 1 2.3999 Vial 1 2 2.3287 3 2.392 2.3735 0.0390 1.6442665 Port 1 1 2.3463 Vial 2 2 2.3422 3 2.3589 2.3491 0.0087 0.3704798 2.3613 Initial 1 2.7407 Vial 1 2 2.67 3 2.7159 2.7089 0.0359 1.3242042 Initial 1 2.7635 Vial 2 2 2.7212 3 2.7587 2.7478 0.0232 0.842891 2.7283 5ppm 1 4.4555 Standard 2 4.3687 3 4.4253 4.4165 0.0441 0.9977141 Mass of TOC Adsorbed: M = 3L (2.7283 mg/L 2.3613 mg/L) =1.10 mg Mass of TOC Removed in Oxidative Step: M = 3L (2.3613 mg/L 1.4452 mg/L) =2.75 mg C/Co % Removed (%) (%) Adsorption 86.55 13.45 Destruction 61.20 38.80 Total 52.97 47.03

PAGE 99

89 Desorption of wastewater-loade d pellets using nanopure water Single pass at 10 mL/min (EBCT = 43 min) Average St Dev RSD Rep # (mg/L) (mg/L) (mg/L) (%) t = 71 1 0.044 Volume passed Conc. 430mL passed 2 0.1494 (L) (mg/L) 3 0.1446 0.1470 0.00241.63 0.43 0.1470 t = 51 1 0.3139 0.23 0.3053 230mL passed 2 0.2946 0.03 1.5595 3 0.3074 0.3053 0.00983.22 omits Rep # 1 of t = 71 t = 31 1 1.6251 30mL passed 2 1.5534 3 1.4999 1.5595 0.06284.03 5ppm 1 4.6763 Standard 2 4.502 3 4.5502 4.5762 0.09001.97 Desorption Polynomial Trend y = 13.698x2 9.8324x + 1.8421 R2 = 10.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 00.10.20.30.40.5 Volume passed (L)TOC Concentration (mg/L) Using polynomial trendline: Between 0.03 and 0.43 Mass desorbed = 0 .193 mg Between 0 and 0.43 Mass desorbed = 0.2438 mg Between 0 and 0.23 Mass desorbed = 0.2316 mg DesorptionExponential Trend y = 0.7075e-3.6543xR2 = 1 0.00 0.50 1.00 1.50 00.10.20.30.40.5 Volume passed (L)TOC Concentration (mg/L) Using exponential trendline: Between 0.23 and 0.43 Mass desorbed = 0.03865 mg Between 0.23 and infinity Mass desorbed = 0.06125 mg Total Desorbed: 0.2316 mg + 0.06125 mg = 0.29285 mg

PAGE 100

APPENDIX F DATA FOR INACTIVATION OF MICROORGANISMS

PAGE 101

91BACTERIOPHAGE EXPERIMENT BAT CH; 365-NM LAMP (12/04/04) X # of Plaques Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) 132 3 0.1 1.0E+041.32E+06 6.12 Adsorption/Attachment Comparison Initial Port 1 49 3 0.1 1.0E+044.90E+05 5.69 6.10 (between Initial Port 1 and t=6 Port 2) 83 3 0.1 1.0E+04 8.30E+05 5.92 Initial Port 2 83 3 0.2 5.0E+03 4.15E+05 5.62 4.47 t-value 73 3 0.1 1.0E+047.30E+05 5.86 First Plating -5.21 Different t=4 Port 2 28 3 0.2 5.0E+031.40E+05 5.15 8.20 Second Plating -3.57 Different 54 3 0.1 1.0E+04 5.40E+05 5.73 56 3 0.1 1.0E+04 5.60E+05 5.75 t=6 Port 1 49 3 0.2 5.0E+03 2.45E+05 5.39 0.10 59 3 0.1 1.0E+045.90E+05 5.77 t=6 Port 2 47 3 0.2 5.0E+032.35E+05 5.37 4.77 87 3 0.1 1.0E+04 8.70E+05 5.94 INACTIVATION: 147 3 0.5 2.0E+03 2.94E+05 5.47 t=7 Port 1 74 3 0.3 3.3E+03 2.47E+05 5.39 8.33 Label time (hrs) LOG [N 0 /N] 159 2 0.1 1.0E+031.59E+0 5 5.20 t = 7 1.00 0.72 54 2 0.1 1.0E+035.40E+0 4 4.73 t = 9 2.00 0.70 t=7 Port 2 50 2 0.1 1.0E+035.00E+04 4.70 7.13 182 2 0.1 1.0E+03 1.82E+05 5.26 61 3 0.5 2.0E+03 1.22E+05 5.09 t=9 Port 1 50 3 0.5 2.0E+03 1.00E+05 5.00 0.95 48 2 0.1 1.0E+034.80E+04 4.68 126 2 0.5 2.0E+022.52E+04 4.40 t=9 Port 2 44 2 0.3 3.3E+021.47E+04 4.17 3.76 -If t-value is less than 2.58, then no significant difference

PAGE 102

92 BACTERIOPHAGE EXPERIMENT BAT CH; 365-NM LAMP (12/04/04) PRD # of Plaques Plate Dilution Volume Total Dilution Mean Value LOG Mean t-value (10^X) (mL) 40 4 0.1 1.0E+05 4.00E+06 6.60 Adsorption/Attachment Comparison Initial Port 1 110 4 0.2 5.0E+04 5.50E+06 6.74 1.65 (between Initial Port 1 and t=6 Port 2) 69 4 0.1 1.0E+05 6.90E+06 6.84 Initial Port 2 137 4 0.2 5.0E+04 6.85E+06 6.84 -0.02 t-value 42 4 0.1 1.0E+05 4.20E+06 6.62 First Plating -0.47 NOT DIFFERENT t=4 Port 2 75 4 0.2 5.0E+04 3.75E+06 6.57 0.49 Second Plating -0.76 NOT DIFFERENT 45 4 0.1 1.0E+05 4.50E+06 6.65 t=6 Port 1 85 4 0.2 5.0E+04 4.25E+06 6.63 0.22 89 4 0.2 5.0E+04 4.45E+06 6.65 t=6 Port 2 98 4 0.2 5.0E+04 4.90E+06 6.69 0.59 98 4 0.2 5.0E+04 4.90E+06 6.69 t=7 Port 1 93 4 0.2 5.0E+04 4.65E+06 6.67 0.29 INACTIVATION: 47 4 0.1 1.0E+05 4.70E+06 6.67 t=7 Port 2 62 4 0.2 5.0E+04 3.10E+06 6.49 2.07 Label time (hrs) LOG [N 0 /N] 97 4 0.2 5.0E+04 4.85E+06 6.69 t = 7 1.00 0.10 t=9 Port 1 43 4 0.2 5.0E+04 2.15E+06 6.33 4.48 t = 9 2.00 0.28 169 3 0.1 1.0E+04 1.69E+06 6.23 164 3 0.1 1.0E+04 1.64E+06 6.21 t=9 Port 2 178 3 0.1 1.0E+04 1.78E+06 6.25 0.22 -If t-value is less than 2.58, then no significant difference

PAGE 103

93 BACTERIOPHAGE EXPERIMENT BAT CH; 365-NM LAMP (12/04/04) MS-2 # of Plaques Plate Dilution Volume Total Dilution Mean Value LOG Mean t-value (10^X) (mL) 32 4 0.2 5.0E+04 1.60E+06 6.20 Adsorption/Attachment Comparison 118 4 0.5 2.0E+04 2.36E+06 6.37 (betwe en Initial Port 1 and t=6 Port 2) Initial Port 1 53 4 0.5 2.0E+04 1.06E+06 6.03 1.87 189 3 0.1 1.0E+04 1.89E+06 6.28 t-value Initial Port 2 102 3 0.1 1.0E+04 1.02E+06 6.01 5.04 First Plating -8.11 Different 101 3 0.1 1.0E+04 1.01E+06 6.00 Second Plating -3.05 Different t=4 Port 2 102 3 0.1 1.0E+04 1.02E+06 6.01 0.00 73 3 0.1 1.0E+04 7.30E+05 5.86 t=6 Port 1 82 3 0.1 1.0E+04 8.20E+05 5.91 0.64 60 3 0.1 1.0E+04 6.00E+05 5.78 t=6 Port 2 62 3 0.1 1.0E+04 6.20E+05 5.79 0.09 102 3 0.1 1.0E+04 1.02E+06 6.01 INACTIVATION: t=7 Port 1 76 3 0.1 1.0E+04 7.60E+05 5.88 1.87 64 3 0.1 1.0E+04 6.40E+05 5.81 Label time (hrs) LOG [N 0 /N] t=7 Port 2 57 3 0.1 1.0E+04 5.70E+05 5.76 0.55 t = 7 1.00 0.16 114 3 0.1 1.0E+04 1.14E+06 6.06 t = 9 2.00 0.95 t=9 Port 1 43 3 0.1 1.0E+04 4.30E+05 5.63 5.59 109 2 0.1 1.0E+03 1.09E+05 5.04 t=9 Port 2 57 2 0.1 1.0E+03 5.70E+04 4.76 3.96 -If t-value is less than 2.58, then no significant difference

PAGE 104

94 BACTERIOPHAGE EXPERIMENT BAT CH; 254-NM LAMP (01/12/05) X # of Plaques Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (PFU/mL) 53 3 0.1 1.0E+045.30E+05 5.72 Adsorption/Attachment Comparison Initial Port 1 45 3 0.1 1.0E+044.50E+05 5.65 0.71 (between Initial Port 1 and t=6 Port 2) 48 3 0.1 1.0E+04 4.80E+05 5.68 Initial Port 2 38 3 0.1 1.0E+04 3.80E+05 5.58 0.97 t-value 41 3 0.2 5.0E+032.05E+05 5.31 First Plating -4.34 Different t=4 Port 2 52 3 0.2 5.0E+032.60E+05 5.41 1.04 Second Plating -3.24 Different 94 3 0.5 2.0E+03 1.88E+05 5.27 t=6 Port 1 71 3 0.5 2.0E+03 1.42E+05 5.15 1.71 44 3 0.2 5.0E+032.20E+05 5.34 t=6 Port 2 45 3 0.2 5.0E+032.25E+05 5.35 0.00 83 3 0.5 2.0E+03 1.66E+05 5.22 t=7 Port 1 94 3 0.5 2.0E+03 1.88E+05 5.27 0.75 INACTIVATION: 72 2 0.3 3.3E+022.40E+04 4.38 t=7 Port 2 77 2 0.3 3.3E+022.57E+04 4.41 0.33 Label time (hrs) LOG [N 0 /N] 42 2 0.1 1.0E+03 4.20E+04 4.62 t = 7 1.00 0.85 t=9 Port 1 41 2 0.1 1.0E+03 4.10E+04 4.61 0.00 t = 9 2.00 1.67 42 1 0.5 2.0E+018.40E+02 2.92 t=9 Port 2 47 1 0.5 2.0E+019.40E+02 2.97 0.42 -If t-value is less than 2.58, then no significant difference

PAGE 105

95 BACTERIOPHAGE EXPERIMENT BAT CH; 254-NM LAMP (01/12/05) PRD # of Plaques Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (PFU/mL) 71 4 0.1 1.0E+057.10E+06 6.85 Adsorption/Attachment Comparison Initial Port 1 48 4 0.1 1.0E+054.80E+06 6.68 2.02 (between Initial Port 1 and t=6 Port 2) 48 4 0.1 1.0E+05 4.80E+06 6.68 Initial Port 2 52 4 0.1 1.0E+05 5.20E+06 6.72 0.30 t-value 73 4 0.2 5.0E+043.65E+06 6.56 First Plating -3.48 Different t=4 Port 2 69 4 0.2 5.0E+043.45E+06 6.54 0.25 Second Plating -0.32 NOT DIFFERENT 59 4 0.2 5.0E+04 2.95E+06 6.47 t=6 Port 1 48 4 0.2 5.0E+04 2.40E+06 6.38 0.97 80 4 0.2 5.0E+044.00E+06 6.60 t=6 Port 2 103 4 0.2 5.0E+045.15E+06 6.71 1.63 82 4 0.2 5.0E+04 4.10E+06 6.61 t=7 Port 1 85 4 0.2 5.0E+04 4.25E+06 6.63 0.15 INACTIVATION: 85 3 0.1 1.0E+048.50E+05 5.93 t=7 Port 2 56 3 0.1 1.0E+045.60E+05 5.75 2.36 Label time (hrs) LOG [N 0 /N] 339 3 0.3 3.3E+03 1.13E+06 6.05 t = 7 1.00 0.78 t=9 Port 1 184 3 0.2 5.0E+03 9.20E+05 5.96 2.20 t = 9 2.00 1.43 102 2 0.3 3.3E+023.40E+04 4.53 t=9 Port 2 124 2 0.3 3.3E+024.13E+04 4.62 1.40 -If t-value is less than 2.58, then no significant difference

PAGE 106

96 BACTERIOPHAGE EXPERIMENT BAT CH; 254-NM LAMP (01/12/05) MS-2 # of Plaques Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (PFU/mL) 157 3 0.1 1.0E+041.57E+06 6.20 Adsorption/Attachment Comparison Initial Port 1 128 3 0.1 1.0E+041.28E+06 6.11 1.66 (between Initial Port 1 and t=6 Port 2) 219 3 0.1 1.0E+04 2.19E+06 6.34 Initial Port 2 128 3 0.1 1.0E+04 1.28E+06 6.11 4.83 t-value 170 3 0.1 1.0E+041.70E+06 6.23 First Plating -0.46 NOT DIFFERENT t=4 Port 2 103 3 0.1 1.0E+041.03E+06 6.01 3.99 Second Plating -3.57 Different 172 3 0.1 1.0E+04 1.72E+06 6.24 t=6 Port 1 73 3 0.1 1.0E+04 7.30E+05 5.86 6.26 148 3 0.1 1.0E+041.48E+06 6.17 t=6 Port 2 76 3 0.1 1.0E+047.60E+05 5.88 4.74 167 3 0.1 1.0E+04 1.67E+06 6.22 t=7 Port 1 94 3 0.1 1.0E+04 9.40E+05 5.97 4.46 INACTIVATION: 241 2 0.1 1.0E+032.41E+05 5.38 t=7 Port 2 128 2 0.1 1.0E+031.28E+05 5.11 5.83 Label time (hrs) LOG [N 0 /N] 160 3 0.5 2.0E+03 3.20E+05 5.51 t = 7 1.00 0.85 t=9 Port 1 102 3 0.5 2.0E+03 2.04E+05 5.31 3.52 t = 9 2.00 1.65 150 1 0.3 3.3E+015.00E+03 3.70 t=9 Port 2 199 1 0.3 3.3E+016.63E+03 3.82 2.57 -If t-value is less than 2.58, then no significant difference

PAGE 107

97 BACTERIOPHAGE EXPERIMENT BATCH; 254-NM LAMP; NO PELLETS (01/22/05) X # of Plaques Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (PFU/mL) 44 3 0.1 1.00E+04 4.40E+05 5.64 Adsorption/Attachment Comparison 85 3 0.3 3.33E+03 2.83E+05 5.45 (betwe en Initial Port 1 and t=6 Port 2) Initial Port 1 65 3 0.3 3.33E+03 2.17E+05 5.34 2.29 58 3 0.3 3.33E+03 1.93E+05 5.29 t-value Initial Port 2 35 3 0.3 3.33E+03 1.17E+05 5.07 2.28 First Plating -5.50 Different 38 3 0.3 3.33E+03 1.27E+05 5.10 Second Plating -3.17 Different t=4 Port 2 30 3 0.3 3.33E+03 1.00E+05 5.00 0.85 103 2 0.1 1.00E+03 1.03E+05 5.01 53 3 0.5 2.00E+03 1.06E+05 5.03 t=6 Port 1 77 3 0.5 2.00E+03 1.54E+05 5.19 2.02 133 2 0.1 1.00E+03 1.33E+05 5.12 56 3 0.5 2.00E+03 1.12E+05 5.05 INACTIVATION: t=6 Port 2 61 3 0.5 2.00E+03 1.22E+05 5.09 0.37 85 2 0.1 1.00E+03 8.50E+04 4.93 Label time (hrs) LOG [N 0 /N] 49 3 0.5 2.00E+03 9.80E+04 4.99 t = 7 1.00 0.65 t=7 Port 1 43 3 0.5 2.00E+03 8.60E+04 4.93 0.52 t = 9 2.00 2.58 202 1 0.1 1.00E+02 2.02E+04 4.31 59 2 0.3 3.33E+02 1.97E+04 4.29 t=7 Port 2 60 2 0.3 3.33E+02 2.00E+04 4.30 0.00 56 2 0.3 3.33E+02 1.87E+04 4.27 t=9 Port 1 50 2 0.3 3.33E+02 1.67E+04 4.22 0.49 15 0 0.5 2.00E+00 3.00E+01 1.48 t=9 Port 2 72 0 1 1.00E+00 7.20E+01 1.86 3.07 -If t-value is less than 2.58, then no significant difference

PAGE 108

98 BACTERIOPHAGE EXPERIMENT BATCH; 254-NM LAMP; NO PELLETS (01/22/05) PRD # of Plaques Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (PFU/mL) 64 4 0.1 1.00E+05 6.40E+06 6.81 Adsorption/Attachment Comparison 108 4 0.3 3.33E+04 3.60E+06 6.56 (betwe en Initial Port 1 and t=6 Port 2) Initial Port 1 60 4 0.3 3.33E+04 2.00E+06 6.30 3.61 40 4 0.1 1.00E+05 4.00E+06 6.60 t-value Initial Port 2 65 4 0.3 3.33E+04 2.17E+06 6.34 2.99 First Plating -6.82 Different 67 4 0.2 5.00E+04 3.35E+06 6.53 Second Plating -1.89 NOT DIFFERENT t=4 Port 2 63 4 0.3 3.33E+04 2.10E+06 6.32 2.60 53 4 0.2 5.00E+04 2.65E+06 6.42 t=6 Port 1 105 4 0.3 3.33E+04 3.50E+06 6.54 1.58 59 4 0.3 3.33E+04 1.97E+06 6.29 t=6 Port 2 81 4 0.3 3.33E+04 2.70E+06 6.43 1.77 43 4 0.3 3.33E+04 1.43E+06 6.16 INACTIVATION: t=7 Port 1 40 4 0.3 3.33E+04 1.33E+06 6.12 0.22 88 3 0.1 1.00E+04 8.80E+05 5.94 Label time (hrs) LOG [N 0 /N] t=7 Port 2 53 3 0.1 1.00E+04 5.30E+05 5.72 2.86 t = 7 1.00 0.31 54 3 0.1 1.00E+04 5.40E+05 5.73 t = 9 2.00 1.98 t=9 Port 1 41 3 0.1 1.00E+04 4.10E+05 5.61 1.23 66 1 0.1 1.00E+02 6.60E+03 3.82 t=9 Port 2 37 1 0.1 1.00E+02 3.70E+03 3.57 2.76 -If t-value is less than 2.58, then no significant difference

PAGE 109

99 BACTERIOPHAGE EXPERIMENT BATCH; 254-NM LAMP; NO PELLETS (01/22/05) MS-2 # of Plaques Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (PFU/mL) 114 3 0.1 1.00E+04 1.14E+06 6.06 Adsorption/Attachment Comparison 70 3 0.1 1.00E+04 7.00E+05 5.85 (betwe en Initial Port 1 and t=6 Port 2) Initial Port 1 111 3 0.1 1.00E+04 1.11E+06 6.05 3.17 61 3 0.1 1.00E+04 6.10E+05 5.79 t-value Initial Port 2 50 3 0.1 1.00E+04 5.00E+05 5.70 0.95 First Plating -1.73 NOT DIFFERENT 86 3 0.1 1.00E+04 8.60E+05 5.93 Second Plating -3.65 Different t=4 Port 2 53 3 0.1 1.00E+04 5.30E+05 5.72 2.71 72 3 0.1 1.00E+04 7.20E+05 5.86 t=6 Port 1 50 3 0.1 1.00E+04 5.00E+05 5.70 1.90 50 3 0.1 1.00E+04 5.00E+05 5.70 t=6 Port 2 62 3 0.1 1.00E+04 6.20E+05 5.79 1.04 34 3 0.1 1.00E+04 3.40E+05 5.53 INACTIVATION: t=7 Port 1 47 3 0.1 1.00E+04 4.70E+05 5.67 1.33 103 2 0.1 1.00E+03 1.03E+05 5.01 Label time (hrs) LOG [N 0 /N] 100 2 0.1 1.00E+03 1.00E+0 5 5.00 t = 7 1.00 0.57 t=7 Port 2 120 2 0.1 1.00E+03 1.20E+05 5.08 0.14 t = 9 2.00 1.92 68 2 0.1 1.00E+03 6.80E+04 4.83 t=9 Port 1 139 2 0.1 1.00E+03 1.39E+05 5.14 4.87 56 1 0.5 2.00E+01 1.12E+03 3.05 t=9 Port 2 60 1 0.5 2.00E+01 1.20E+03 3.08 0.28 -If t-value is less than 2.58, then no significant difference

PAGE 110

100 BACTERIA EXPERIMENT BATCH; 254-NM LAMP (02/09/05) C3 # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) 65 4 0.1 1.0E+056.50E+06 6.81 Adsorption/Attachment Comparison Initial Port 1 39 4 0.1 1.0E+053.90E+06 6.59 2.45 (between Initial Port 1 and t=6 Port 2) 120 4 0.1 1.0E+05 1.20E+07 7.08 Initial Port 2 67 4 0.1 1.0E+05 6.70E+06 6.83 3.80 t-value 289 4 0.3 3.3E+049.63E+06 6.98 First Plating -4.66 Different 67 4 0.1 1.0E+056.70E+06 6.83 Second Plating -1.74 NOT DIFFERENT t=5 Port 2 47 4 0.1 1.0E+054.70E+06 6.67 2.63 324 4 0.3 3.3E+04 1.08E+07 7.03 47 4 0.1 1.0E+05 4.70E+06 6.67 t=6 Port 1 42 4 0.1 1.0E+05 4.20E+06 6.62 5.43 364 4 0.3 3.3E+041.21E+07 7.08 57 4 0.1 1.0E+055.70E+06 6.76 INACTIVATION: t=6 Port 2 50 4 0.1 1.0E+055.00E+06 6.70 5.37 305 4 0.3 3.3E+04 1.02E+07 7.01 Label time (hrs) LOG [N 0 /N] 46 4 0.1 1.0E+05 4.60E+06 6.66 t = 7 1.00 0.89 t=7 Port 1 41 4 0.1 1.0E+05 4.10E+06 6.61 5.08 t = 9 2.00 1.00 87 3 0.1 1.0E+048.70E+05 5.94 t=7 Port 2 62 3 0.1 1.0E+046.20E+05 5.79 1.97 38 4 0.3 3.3E+04 1.27E+06 6.10 t=9 Port 1 83 4 0.3 3.3E+04 2.77E+06 6.44 4.00 306 2 0.1 1.0E+033.06E+05 5.49 47 3 0.3 3.3E+031.57E+05 5.19 t=9 Port 2 41 3 0.3 3.3E+031.37E+05 5.14 0.53 -If t-value is less than 2.58, then no significant difference

PAGE 111

101 BACTERIA EXPERIMENT BATCH; 254-NM LAMP (02/09/05) PA # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) 91 3 0.1 1.0E+049.10E+05 5.96 Adsorption/Attachment Comparison Initial Port 1 123 3 0.1 1.0E+041.23E+06 6.09 2.12 (between Initial Port 1 and t=6 Port 2) 68 3 0.1 1.0E+04 6.80E+05 5.83 Initial Port 2 93 3 0.1 1.0E+04 9.30E+05 5.97 1.89 t-value 26 4 0.3 3.3E+048.67E+05 5.94 First Plating -0.36 NOT DIFFERENT 65 3 0.1 1.0E+046.50E+05 5.81 Second Plating -0.45 NOT DIFFERENT t=5 Port 2 104 3 0.1 1.0E+041.04E+06 6.02 2.92 124 3 0.1 1.0E+04 1.24E+06 6.09 t=6 Port 1 97 3 0.1 1.0E+04 9.70E+05 5.99 1.75 97 3 0.1 1.0E+049.70E+05 5.99 t=6 Port 2 115 3 0.1 1.0E+041.15E+06 6.06 1.17 43 4 0.3 3.3E+04 1.43E+06 6.16 INACTIVATION: 70 3 0.1 1.0E+04 7.00E+05 5.85 t=7 Port 1 92 3 0.1 1.0E+04 9.20E+05 5.96 1.65 Label time (hrs) LOG [N 0 /N] 124 2 0.1 1.0E+031.24E+0 5 5.09 t = 7 1.00 1.03 t=7 Port 2 66 2 0.1 1.0E+036.60E+04 4.82 4.14 t = 9 2.00 1.44 81 3 0.1 1.0E+04 8.10E+05 5.91 t=9 Port 1 51 3 0.1 1.0E+04 5.10E+05 5.71 2.52 240 1 0.1 1.0E+022.40E+04 4.38 23 2 0.1 1.0E+032.30E+04 4.36 t=9 Port 2 134 2 0.6 1.7E+022.23E+04 4.35 0.02 -If t-value is less than 2.58, then no significant difference

PAGE 112

102 BACTERIA EXPERIMENT BATCH; 254-NM LAMP (02/09/05) SA # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) 54 4 0.3 3.3E+041.80E+06 6.26 Adsorption/Attachment Comparison Initial Port 1 33 4 0.3 3.3E+041.10E+06 6.04 2.14 (between Initial Port 1 and t=6 Port 2) 125 4 0.3 3.3E+04 4.17E+06 6.62 Initial Port 2 56 4 0.3 3.3E+04 1.87E+06 6.27 5.05 t-value 119 4 0.3 3.3E+043.97E+06 6.60 First Plating -6.60 Different t=5 Port 2 61 4 0.3 3.3E+042.03E+06 6.31 4.25 Second Plating -3.30 Different 102 4 0.3 3.3E+04 3.40E+06 6.53 t=6 Port 1 65 4 0.3 3.3E+04 2.17E+06 6.34 2.79 149 4 0.3 3.3E+044.97E+06 6.70 t=6 Port 2 67 4 0.3 3.3E+042.23E+06 6.35 5.51 140 4 0.3 3.3E+04 4.67E+06 6.67 t=7 Port 1 60 4 0.3 3.3E+04 2.00E+06 6.30 5.59 INACTIVATION: 34 3 0.1 1.0E+043.40E+05 5.53 67 3 0.3 3.3E+032.23E+05 5.35 Label time (hrs) LOG [N 0 /N] t=7 Port 2 49 3 0.3 3.3E+031.63E+05 5.21 1.90 t = 7 1.00 1.12 190 3 0.1 1.0E+04 1.90E+06 6.28 t = 9 2.00 1.40 t=9 Port 1 143 3 0.1 1.0E+04 1.43E+06 6.16 2.52 74 2 0.1 1.0E+037.40E+04 4.87 t=9 Port 2 57 2 0.1 1.0E+035.70E+04 4.76 1.40 -If t-value is less than 2.58, then no significant difference

PAGE 113

103 BACTERIA EXPERIMENT BATCH; 254-NM LAMP; REPLICATE (02/13/05) C3 # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) 216 4 0.1 1.00E+05 2.16E+07 7.33 Adsorption/Attachment Comparison Initial Port 1 171 4 0.1 1.00E+05 1.71E+07 7.23 2.24 (between Initial Port 1 and t=6 Port 2) 83 4 0.1 1.00E+05 8.30E+06 6.92 Initial Port 2 65 4 0.1 1.00E+05 6.50E+06 6.81 1.40 t-value 304 4 0.1 1.00E+05 3.04E+07 7.48 First Plating -0.44 NOT DIFFERENT t=4 Port 2 267 4 0.1 1.00E+05 2.67E+07 7.43 1.51 Second Plating -1.06 NOT DIFFERENT 290 4 0.1 1.00E+05 2.90E+07 7.46 t=6 Port 1 246 4 0.1 1.00E+05 2.46E+07 7.39 1.86 206 4 0.1 1.00E+05 2.06E+07 7.31 t=6 Port 2 151 4 0.1 1.00E+05 1.51E+07 7.18 2.86 309 4 0.1 1.00E+05 3.09E+07 7.49 t=7 Port 1 212 4 0.1 1.00E+05 2.12E+07 7.33 4.21 INACTIVATION: 99 3 0.1 1.00E+04 9.90E+05 6.00 t=7 Port 2 100 3 0.1 1.00E+04 1.00E+06 6.00 0.00 Label time (hrs) LOG [N 0 /N] 108 4 0.1 1.00E+05 1.08E+07 7.03 t = 7 1.00 1.41 t=9 Port 1 98 4 0.1 1.00E+05 9.80E+06 6.99 0.63 t = 9 2.00 1.70 45 3 0.3 3.33E+03 1.50E+05 5.18 t=9 Port 2 86 3 0.3 3.33E+03 2.87E+05 5.46 3.49 -If t-value is less than 2.58, then no significant difference

PAGE 114

104 BACTERIA EXPERIMENT BATCH; 254-NM LAMP; REPLICATE (02/13/05) PA # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) 461 3 0.1 1.00E+04 4.61E+06 6.66 Adsorption/Attachment Comparison Initial Port 1 406 3 0.1 1.00E+04 4.06E+06 6.61 1.83 (between Initial Port 1 and t=6 Port 2) 148 3 0.1 1.00E+04 1.48E+06 6.17 Initial Port 2 146 3 0.1 1.00E+04 1.46E+06 6.16 0.06 t-value 190 3 0.1 1.00E+04 1.90E+06 6.28 First Plating -15.03 Different t=4 Port 2 138 3 0.1 1.00E+04 1.38E+06 6.14 2.82 Second Plating -12.65 Different 175 3 0.1 1.00E+04 1.75E+06 6.24 t=6 Port 1 185 3 0.1 1.00E+04 1.85E+06 6.27 0.47 103 3 0.1 1.00E+04 1.03E+06 6.01 t=6 Port 2 116 3 0.1 1.00E+04 1.16E+06 6.06 0.81 127 3 0.1 1.00E+04 1.27E+06 6.10 t=7 Port 1 132 3 0.1 1.00E+04 1.32E+06 6.12 0.25 INACTIVATION: 93 2 0.1 1.00E+03 9.30E+04 4.97 t=7 Port 2 110 2 0.1 1.00E+03 1.10E+05 5.04 1.12 Label time (hrs) LOG [N 0 /N] 108 3 0.1 1.00E+04 1.08E+06 6.03 t = 7 1.00 1.11 t=9 Port 1 110 3 0.1 1.00E+04 1.10E+06 6.04 0.07 t = 9 2.00 1.61 63 2 0.3 3.33E+02 2.10E+04 4.32 t=9 Port 2 101 2 0.3 3.33E+02 3.37E+04 4.53 2.89 -If t-value is less than 2.58, then no significant difference

PAGE 115

105 BACTERIA EXPERIMENT BATCH; 254-NM LAMP; NO PELLETS (02/16/05) C3 # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) 173 4 0.1 1.00E+05 1.73E+07 7.24 Adsorption/Attachment Comparison Initial Port 1 264 4 0.1 1.00E+05 2.64E+07 7.42 4.31 (between Initial Port 1 and t=6 Port 2) 205 4 0.1 1.00E+05 2.05E+07 7.31 Initial Port 2 299 4 0.1 1.00E+05 2.99E+07 7.48 4.14 t-value 60 5 0.1 1.00E+06 6.00E+07 7.78 First Plating -0.18 NOT DIFFERENT t=4 Port 2 191 5 0.3 3.33E+05 6.37E+07 7.80 0.33 Second Plating -2.69 Different t=6 Port 1 48 5 0.1 1.00E+06 4.80E+07 7.68 63 5 0.1 1.00E+06 6.30E+07 7.80 t=6 Port 2 141 5 0.3 3.33E+05 4.70E+07 7.67 1.86 68 5 0.1 1.00E+06 6.80E+07 7.83 t=7 Port 1 132 5 0.3 3.33E+05 4.40E+07 7.64 2.86 54 3 0.1 1.00E+04 5.40E+05 5.73 INACTIVATION: t=7 Port 2 64 3 0.1 1.00E+04 6.40E+05 5.81 0.83 241 4 0.1 1.00E+05 2.41E+07 7.38 Label time (hrs) LOG [N 0 /N] t=9 Port 1 288 4 0.1 1.00E+05 2.88E+07 7.46 2.00 t = 7 1.00 1.97 29 3 0.3 3.33E+03 9.67E+0 4 4.99 t = 9 2.00 2.53 t=9 Port 2 62 2 0.1 1.00E+03 6.20E+04 4.79 N/A -If t-value is less than 2.58, then no significant difference N/A = counts plated at different dilution factors

PAGE 116

106 BACTERIA EXPERIMENT BATCH; 254-NM LAMP; NO PELLETS (02/16/05) PA # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) Initial Port 1 247 3 0.1 1.00E+04 2.47E+06 6.39 Adsorption/Attachment Comparison 892 3 0.1 1.00E+04 8.92E+06 6.95 (between Initial Port 1 and t=6 Port 2) Initial Port 2 253 3 0.1 1.00E+04 2.53E+06 6.40 18.85 134 3 0.1 1.00E+04 1.34E+06 6.13 t-value t=4 Port 2 135 3 0.1 1.00E+04 1.35E+06 6.13 0.00 First Plating -1.98 NOT DIFFERENT 195 3 0.1 1.00E+04 1.95E+06 6.29 Second Plating -1.35 NOT DIFFERENT t=6 Port 1 225 3 0.1 1.00E+04 2.25E+06 6.35 1.42 153 3 0.1 1.00E+04 1.53E+06 6.18 204 3 0.1 1.00E+04 2.04E+06 6.31 t=6 Port 2 217 3 0.1 1.00E+04 2.17E+06 6.34 0.58 197 3 0.1 1.00E+04 1.97E+06 6.29 t=7 Port 1 267 3 0.1 1.00E+04 2.67E+06 6.43 3.20 INACTIVATION: 44 2 0.1 1.00E+03 4.40E+04 4.64 t=7 Port 2 47 2 0.1 1.00E+03 4.70E+04 4.67 0.21 Label time (hrs) LOG [N 0 /N] 101 3 0.1 1.00E+04 1.01E+06 6.00 t = 7 1.00 1.70 t=9 Port 1 86 3 0.1 1.00E+04 8.60E+05 5.93 1.02 t = 9 2.00 1.24 180 2 0.3 3.33E+02 6.00E+04 4.78 t=9 Port 2 145 2 0.3 3.33E+02 4.83E+04 4.68 1.89 -If t-value is less than 2.58, then no significant difference

PAGE 117

107 BACTERIA EXPERIMENT BATCH; 254-NM LAMP; NO PELLETS (02/16/05) SA # of Colonies Plate Dilution Volume Total Dilution Concentration LOG Concentration t-value (10^X) (mL) (CFU/mL) 1172 3 0.1 1.00E+04 1.17E+07 7.07 Adsorption/Attachment Comparison Initial Port 1 141 4 0.1 1.00E+05 1.41E+07 7.15 N/A (between Initial Port 1 and t=6 Port 2) 892 3 0.1 1.00E+04 8.92E+06 6.95 Initial Port 2 128 4 0.1 1.00E+05 1.28E+07 7.11 N/A t-value t=4 Port 2 1068 3 0.1 1.00E+04 1.07E+07 7.03 First Plating -2.44 NOT DIFFERENT t=6 Port 1 1348 3 0.1 1.00E+04 1.35E+07 7.13 t=6 Port 2 1056 3 0.1 1.00E+04 1.06E+07 7.02 t=7 Port 1 1700 3 0.1 1.00E+04 1.70E+07 7.23 INACTIVATION: t=7 Port 2 154 2 0.1 1.00E+03 1.54E+05 5.19 t=9 Port 1 Label time (hrs) LOG [N 0 /N] t=9 Port 2 27 2 0.1 1.00E+03 2.70E+0 4 4.43 t = 7 1.00 2.04 -If t-value is less than 2.58, then no significant difference

PAGE 118

APPENDIX G FIGURES OF DATA FOR INAC TIVATION OF MICROORGANISMS

PAGE 119

109 BACTERIOPHAGE EXPERIMENT – BAT CH; 365-NM LAMP (12/04/04) X-174 Batch Study (365nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Third Assay PRD-1 Batch Study (365nm lamp)5.80 6.00 6.20 6.40 6.60 6.80 7.00Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL First Assay Second Assay Third Assay MS-2 Batch Study (365nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Third Assay

PAGE 120

110 BACTERIOPHAGE EXPERIMENT BAT CH; 254-NM LAMP (01/12/05) X-174 Batch Study (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay PRD-1 Batch Study (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL First Assay Second Assay MS-2 Batch Study (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay

PAGE 121

111 BACTERIOPHAGE EXP. BATCH; 254-NM LAMP; NO PELLETS (01/22/05) X-174 Batch Study without pellets (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Third Assay PRD-1 Batch Study without pellets (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL First Assay Second Assay Third Assay MS-2 Batch Study without pellets (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Third Assay

PAGE 122

112 BACTERIA EXPERIMENT BATCH; 254-NM LAMP (02/09/05) C3 Batch Study (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Third Assay PA Batch Study (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL First Assay Second Assay Third Assay SA Batch Study (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay Third Assay

PAGE 123

113 BACTERIA EXPERIMENT BATCH; 254-NM LAMP; REPLICATE (02/13/05) C3 Batch Study (254nm lamp) Replicate0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay PA Batch Study (254nm lamp) Replicate0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL First Assay Second Assay

PAGE 124

114 BACTERIA EXPERIMENT BATCH; 254-NM LAMP; NO PELLETS (02/16/05) C3 Batch Study without pellets (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 9.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay PA Batch Study without pellets (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL First Assay Second Assay Third Assay SA Batch Study without pellets (254nm lamp)0.00 1.00 2.00 3.00 4.00 5.00 6.00 7.00 8.00 Initial Port 1 Initial Port 2 t=4 Port 2 t=6 Port 1 t=6 Port 2 t=7 Port 1 t=7 Port 2 t=9 Port 1 t=9 Port 2Log of Concentration (PFU/mL) First Assay Second Assay

PAGE 125

115 LIST OF REFERENCES Al-Ekabi, H. and N. Serpone. “Kinetic st udies in heterogeneou s photocatalysis. 1. Photocatalytic degradation of chlorinated phe nols in aerated aque ous solutions over TiO2 supported on a glass matrix.” Journal of Physical Chemistry. 92 (1998): 5726-5731. Anderson, C. and A. Bard. “An improved photocatalyst of TiO2/SiO2 prepared by a solgel synthesis.” Journal of Physical Chemistry. 99 (1995): 9882-9885. Anderson, C. and A. Bard. “Improved photo catalytic activity and characterization of mixed TiO2/SiO2 and TiO2/Al2O3 materials.” Journal of Physical Chemistry B. 101 (1997): 2611-2616. Bekbolet, M. and I. Baleioglu. “Photocatalyt ic degradation kinetics of humic acid in aqueous TiO2 dispersions: The influence of hydr ogen peroxide and bicarbonate ion.” Water Science and Technology. 34:9 (1996): 73-80. Bekbolet, M. “Photocatalytic b actericidal activity of TiO2 in aqueous suspensions of E. coli.” Water Science and Technology. 35:11-12 (1997): 95-100. Birge, E. A. Bacterial and Bacteriophage Genetics. Fourth Edition. New York: Springer, 2000. Blake, D. M., Webb, J., Turchi, C., and K. Magrini. “Kinetic and mechanistic overview of TiO2-photocatalyzed oxidation reac tions in aqueous solution.” Solar Energy Materials. 24 (1991): 584 – 593. Chen, H., Zahraa, O., and M. Bouchy. “Inhi bition of the adsorption and photocatalytic degradation of an organic contaminan t in an aqueous suspension of TiO2 by inorganic ions.” Journal of Photochemistry and Photobiology A: Chemistry. 108 (1997): 37-44. Cho, M., Chung, H., Choi, W., and J. Yoon. “Different inactivation behaviors of MS-2 phage and Escherichia coli in TiO2 photocatalytic disinfection.” Applied and Environmental Microbiology. 71 (2005): 270-275. Clarke, G.M. and D. Cooke. A Basic Course in Statistics. Third edition. London: Edward Arnold, 1992.

PAGE 126

116 Deacon, J. The really easy statistics site. Biology Teaching Organization. University of Edinburgh, United Kingdom. March 2005. . Demeestere, K., Dewulf, J., Van Langenhove, H., and B. Sercu. “Adsorption as initial step in the photocatalytic de gradation of gaseous VOC by TiO2/UV.” Environmental Science and Pollution Research. 8th FECS Conference. Special Issue 2002. Gerischer, H. and A. Heller. “The role of oxygen in photooxidation of organic molecules on semiconductor particles.” Journal of Physical Chemistry. 95 (1991): 5261-5267. Goswami, Y. and D. Blake. “Cleaning up with sunshine.” Mechanical Engineering. August (1996): 56 59. Grabner, G., Li, G., Quint, R., Quint, R., and N. Getoff. “Pulsed laser-induced oxidation of phenol in acid aqueous TiO2 sols.” Journal of the Chemical Society Faraday Transactions. 87 (1991): 1097-1101. Harm, W. Biological Effects of Ultraviolet Radiation. New York: Cambridge University Press, 1980. Hashizume, H. “Adsorption of some ar omatic compounds by a synthetic mesoporous silicate.” Journal of Environmental Science and Health, Part A-Toxic/Hazardous Substances & Environmental Engineering. 39:10 (2004): 2615-2625. Hench, L. and J. West. “The sol-gel process.” Chemical Review. 90 (1990): 33-72. Hoffmann, M., Martin, S., Choi, W., and D. Bahnemann. “Environmental applications of semiconductor photocatalysis.” Chemical Reviews. 95 (1995): 69-96. Holmes, F.R. “The performance of a reactor using photocatalysis to degrade a mixture of organic contaminants in aqueous solution.” Ma ster’s thesis presented to the graduate school of the University of Florida, 2003. Ireland, J.C., Klostermann, P., Rice, E. W., and R.M. Clark. “Inactivation of Escherichia coli by titanium dioxide phot ocatalytic oxidation.” Applied and Environmental Microbiology. 59 (1993): 1668-1670. Lange, K. E. and C.H. Lin. “Advanced lif e support program: Requirements definition and design considerations.” Docu ment # CTSD-ADV-245 (REV A) presented for the Crew and Thermal Systems Division in Houston, TX, 1998. Lee, S., Nishida, K., Otaki, M., and S. O hgaki. “Photocatalytic inactivation of phage Q by immobilized titanium dioxi de mediated photocatalyst.” Water Science and Technology. 35:11-12 (1997): 101-106.

PAGE 127

117 Londeree, D. J. “Silica-titania composites for water treatment.” Master’s thesis presented to the graduate school of the University of Florida, 2002. Ludwig, C.Y. “The performan ce of silica-titania composites in a packed-bed reactor for photocatalytic degradation of gray water.” Ma ster’s thesis presented to the graduate school of the Universi ty of Florida, 2004. Malley, J. P., Jr. “Ult raviolet disinfection.” Control of Microorganisms in Drinking Water. ASCE, 2002. Maness, P., Smolinski, S., Blake, D.M., Huang, Z., Wolfrum, E.J., and W.A. Jacoby. “Bactericidal activity of photocatalytic TiO2 reaction: Toward an understanding of its killing mechanism.” Applied and Environmental Microbiology. 65:9 (1999): 4094-4098. Mao, Y., Schoneich, C., and K. Asmus. “I dentification of organic acids and other intermediates in oxidative degrada tion of chlorinated ethanes on TiO2 surfaces en route to mineralization: A combined photocatal ytic and radiation chemical study.” Journal of Physical Chemistry. 95 (1991): 10080 – 10089. Mao, Y., Schoneich, C., and K. Asmus. “Influence of TiO2 surface on 1,2-chlorine shift in -chlorine substituted radicals as studied by radiation chemistry and photocatalysis.” Journal of Physical Chemistry. 96 (1992): 8522-8529. Matthews, C. K. Bacteriophage Biochemistry. New York: Van Nostrand Reinhold Company, 1971. Matthews, R. “An adsorption water purifier with in situ photocatalytic regeneration.” Journal of Catalysis. 113 (1988): 549-555. Morita, S., Namikoshi, A., Hirata, T., Oguma, K., Katayama, H., Ohgaki, S., Motoyama, N., and M. Fujiwara. “Efficacy of UV irradiation in inactivating Cryptosporidium parvum oocysts.” Applied and Environm ental Microbiology. 68 (2002): 5387-5393. Munter, R. “Advanced oxidation processe s – current status and prospects.” Proceedings of the Estonian Academy of Sciences. Chemistry. 50:2 (2001): 59-80. Murray, P. R., Kobayashi, G. S., Pf aller, M. A., and K.S. Rosenthal. Medical Microbiology. Second Edition. St. Louis: Mosby-Year Book, 1994. Nawrocki, J. “The silanol group and its role in liquid chromatography.” Journal of Chromatography A. 779 (1997): 29-71. Ohtani, B. and S. Nishimoto. “Effect of su rface adsorptions of aliphatic alcohols and silver ion on the photocatalytic activity of TiO2 suspended in aqueous solutions.” Journal of Physical Chemistry. 97 (1993): 920-926.

PAGE 128

118 Okamoto, K., Yamamoto, Y., Tanaka, H., Ta naka, M., and A. Itaya. “Heterogeneous photocatalytic decomposition of phenol over TiO2 powder.” Bulletin of the Chemical Society of Japan. 58 (1985): 2015-2022. Ollis, D., Pelizzetti, E., and N. Serpone “Destruction of water contaminants.” Environmental Science and Technology. 25:9 (1991): 1523 1529. Powers, K. “The development and characterization of sol gel substrates for chemical and optical applications.” Ph.D. dissertation presented to the graduate school of the University of Florida, 1998. Shaban, A.M., El-Taweel, G.E., and G.H. A li. “UV ability to inactivate microorganisms combined with factors affecting radiation.” Water Science and Technology. 35:11-12 (1997): 107-112. Sjogren, J.C. and R.A. Sierka. “Inactivat ion of phage MS-2 by iron-aided titanium dioxide photocatalysis.” Applied and Environmental Microbiology. 60 (1994): 344-347. Standard Methods for the Exami nation of Water and Wastewater. Twentieth Edition. Washington: APHA, AWWA, and WEF, 1998. Tanaka, K., Capule, M., and T. Hisana ga. “Effect of crystallinity of TiO2 on its photocatalytic action.” Chemical Physics Letters. 187 (1991): 73-76. Todar, K. Todar’s Online Textbook of Bacteriology. University of Wisconsin. Department of Bacteriology. Marc h 2005 . Turchi, C. S. and D.F. Ollis. “Mixed reacta nt photocatalysis: Intermediates and mutual rate inhibition.” Journal of Photocatalysis. 119 (1989): 483. Turchi, C. S. and D.F. Ollis. “Photocatalytic degradation of organic-water contaminants: Mechanisms involving hydroxyl radical attack.” Journal of Catalysis. 122 (1990): 178192. United States Environmental Protection Agency (USEPA). “Methods for the determination of organic compounds in drin king water.” EPA/600/R-92/129, Springfield, 1992. Verostko, C.E., Carrier, C., and B.W. Finger. “Ersatz wastewat er formulations for testing water recovery systems.” Document # 2004-01-2448. SAE International. 2004. Vohra, M. and K. Tanaka. “Photocatalytic de gradation of aqueous pollutants using silicamodified TiO2.” Water Research. 37 (2003): 3992-3996.

PAGE 129

119 Wei, C., Lin, W., Zainal, Z., Williams, N.E., Zhu, K., Kruzic, A.P., Smith, R.L., and K. Rajeshwar. “Bactericidal activity of TiO2 photocatalyst in aqueous media: Toward a solar-assisted water disinfection system.” Environmental Science and Technology. 28 (1994): 934-938. Xu, Y., Zheng, W., and W. Liu. “Enhanced photocatalytic activity of supported TiO2 dispersing effect of SiO2.” Journal of Photochemistry and Photobiology A: Chemistry. 122 (1999): 57-60. Yoneyama, H. and T. Torimoto. “Titanium dioxide/adsorbent hybrid photocatalysts for photodestruction of organic substanc es of dilute concentrations.” Catalysis Today. 58 (2000): 133-140.

PAGE 130

120 BIOGRAPHICAL SKETCH Mary Joanne Garton was born in Milton, Florida, on July 22, 1981. She is the daughter of John and Jane Garton and has an older brother, Jack Garton. Her family relocated to Naval bases in Argentia, Newfoundland in Canada and New Orleans, Louisiana, before returning to Milton in 1990. Joanne graduated from Milton High School in 1999. Upon arriving at the University of Flor ida in the Fall of 1999, Joanne was an undecided engineering major. After attending the Introduc tion to Engineering class presentation for the Environmental Engin eering Department, pres ented by Dr. Paul Chadik, Joanne realized her interest in the environment and specifica lly in the water and wastewater field. She pursued a Bachelor of Science degree in Environmental Engineering. Joanne entered graduate school in the Fall of 2003 during her last year of undergraduate study in the combined BS/ME degree program under th e direction of her advising professor, Dr. Paul Chadik. Joanne met her fianc David Friedman dur ing her undergraduate career at U.F. Their mutual love of Gator football, movies and trivia games brought them together, and Joanne and Dave will be married in May 2005.