|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
THE ROLE OF CALCIUM IN AND METHODOLOGIES FOR OVERCOMING pH
EXCURSIONS FOR REACTIVATED GRANULAR ACTIVATED CARBON
MORGANA T. BACH
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
Morgana T. Bach
This document is dedicated to my mother and sister
TABLE OF CONTENTS
LIST OF TABLES ....v......... .................. .....i......... ... vi
LIST OF FIGURES ........................ ................................... vii
ABSTRAC T ............ ................... ............. .... .......... viii
1 INTRODUCTION/LITERATURE REVIEW.............. ............................... 1
1.1 A activated C arbon ........................................ 1
1.2 Adsorption............................ .. ... ....... ........ 1
1.3 Current U ses of A activated Carbon................................................................. 3
1.4 Reactivation of Activated Carbon..................................... .............. 4
2 HYPOTHESIS AND OBJECTIVES..................................................
2.1 Hypothesis.............................. .......... ........ 7
2.2 Objectives ........................................ .......... 7
3 THE ROLE OF CALCIUM IN PH EXCURSIONS FOR REACTIVATED GAC........ 9
3.1 Literature Review ............... .. .. ............... 10
3.1.1 ph Excursions in W ater Treatment ............................... ......... 10
3.1.2 Carbon B asic Properties............................... ............................................... 13
3.1.3 pH Excursion Mechanism for Virgin Carbons .................................... 14
3.2 Experim mental .............. ................... .................. ............. ......... 16
3.2.1 Activated Carbon Samples............. ........ ................... 16
3.2.2 Water Contact pH ............................. .............. 16
3.2.3 Calcium Solution Concentration....................... ................. 17
3.2.4 G A C Sulfate C apacity................................................................... ........ 17
3.3 Results and Discussion ..................................... ............ ......... 18
3.3.1 Calcium Content Versus W ater Contact pH .................................................. 18
3.3.2 Predicted pH Based on Calcium Leached from GAC ..................................... 20
3.3.3 pH of Steam-Cured Plus Ramped-Temperature Reactivated Carbons......... 23
3.3.4 Influence of Acid and Base Washing on Water Contact pH ........................... 24
3.3.5 Comparisons of pH in Milli-Q and Sulfate-Containing Water........................ 25
3.4 Conclusions ................................................ 26
4 CREATING A pH STABLE CARBON.................... .... ................... 28
4.1 Literature Review ..... ................... ...... ....... ............... 29
4.1.1 Methods to Overcome pH Excursions............................. 29
4.1.2 Effect of Oxidation on Activated Carbon ...................................... 33
4.2 Experimental ................ .................. .... ......... 36
4.2.1 Air Treatments and CO2 Treatments............... ................... 36
4.2.2 Boehm Titration ............... ........ ......... .............. 36
4.2.3 Equilibrium Isotherms .................................. ................... 37
4.3 Results and Discussion ....................................... ......................39
4.3.1 Air and CO2 Treatments Versus W ater Contact pH ........................................ 39
4.3.2 Total Surface Acidity of Treated Carbons................................. ................... 40
4.3.3 Comparison of MIB Adsorption Capability ............................................... 41
4 .4 C onclu sions........................................... 46
5 CONTRIBUTIONS TO SCIENCE ...................................... ............. 49
REFERENCES .. ............................ ...... ............. 50
BIOGRAPHICAL SKETCH ................ .................... ........ 54
LIST OF TABLES
1 Metal concentrations leached from the activated carbon samples.......................... 23
2 Freundlich values for MIB removal in Manatee water........................................... 46
LIST OF FIGURES
1 Full-scale water treatment plant effluent pH and calcium concentration data after
reactivated GAC was returned to service............... ................... 12
2 Correlation between log GAC calcium content and water contact pH exhibited by
reactivated carbon .............. ........ .... .... ............................ 18
3 Predicted pH based on the calcium that leached out of the reactivated carbon in an
isolated and an open system versus the water contact pH measured in Milli-Q water
........................................................................... . ..... 2 2
4 Comparison of water contact pH for conventionally reactivated and steam-cured
reactivated carbons to 3 different temperatures................... .............. 24
5 Sulfate uptake for several reactivated carbons versus their water contact pH as
measured in 80 mg/L sulfate water........... .......... .................. 26
6 Proposed mechanism for the formation of COOH .................................. 32
7 Effect of air and CO2 treatments on water contact pH.................................. 40
8 Effect of air and C02 treatments on total surface acidity....................................... 41
9 Equilibrium adsorption of MIB in deionized water versus the dose of the composite
.......................................................................... . ...... 4 2
10 Equilibrium adsorption of MIB in Manatee water versus the dose of the composite
.......................................................................... . ...... 4 4
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
THE ROLE OF CALCIUM IN AND METHODOLOGIES FOR OVERCOMING pH
EXCURSIONS FOR REACTIVATED GRANULAR ACTIVATED CARBON
Morgana T. Bach
Chair: David W. Mazyck
Major Department: Environmental Engineering Sciences
Water contact pH experiments employing several reactivated granular activated
carbons (GAC) showed that reactivated carbons exhibit pH excursions (i.e., pH > 8.5)
when wetted. Moreover, when the water contact pH was predicted based on the log of
the quantity of calcium that leached out of the GAC during the pH experiments, good
agreement was obtained with the measured water contact pH of reactivated GACs that
contained a range of calcium content, typically 0 to 1% wt. Ca. Treatments employing
air exposure at 4000C for 1 hour and carbon dioxide exposure at ambient temperature for
20 minutes were found to create pH stable carbons (i.e., water contact pH < 8.5). An
evaluation of total surface acidity suggested that the pH excursion was balanced by the
acidic oxygen-containing functional groups created on the surface of the air treated
reactivated GAC. Physisorption of carbon dioxide and subsequent desorption forming
carbonic acid upon water immersion was proposed as the means behind balancing pH
excursions with the carbon dioxide treated reactivated GAC.
In a comparison of adsorption performance with 2-methylisoborneol (MIB), a
taste and odor causing compound often found in surface water, in deionized water the
untreated reactivated GAC removed more MIB than the air treated reactivated GAC. It is
proposed that the increase in surface acidity with the air treated reactivated GAC created
preferential adsorption of water onto the GAC surface over MIB adsorption. In
adsorption experiments with MIB in water containing natural organic matter (NOM), a
ubiquitous surface water component shown to compete with MIB for adsorption sites, the
air treated reactivated GAC removed more MIB than did an untreated reactivated GAC.
The proposed explanation for this is that the negatively charged NOM is attracted to the
positively charged surface of the untreated carbon via electrostatic forces, thus out-
competing MIB for adsorption sites. In both deionized water and water containing NOM,
the carbon dioxide treated reactivated GAC removed the most MIB, though the
mechanism behind this is unclear and warrants further research. Regardless of the
mechanism, the results show that treatments to create pH stable carbons will change the
carbon's adsorption capabilities and, thus, use of these treatments should be based on
more than their ability to create pH stable carbons.
1.1 Activated Carbon
Activated carbon is created when a carbonaceous precursor, generally wood, peat,
lignite-coal, bituminous-coal, or other material possessing relatively high carbon content,
undergoes a thermal process involving pyrolysis and oxidation to increase the surface
area and develop a range of pore sizes within the precursor. Pyrolysis is carried out in an
inert environment, or oxygen starved, to ensure that the raw material does not gasify but
rather is transformed into char. The oxidation process, or simply activation, is performed
in the presence of an oxidizing gas, such as air, steam, carbon dioxide (C02), or a
combination of these, at temperatures between 800-900'C in order to oxidize the surface
of the carbon pores and further develop the internal pore structure. In physical activation,
the pyrolysis and oxidation steps are separate, while in chemical activation the two
processes are combined (Snoeyink and Summers 1999). Powdered activated carbon
(PAC) is activated carbon ground in various types of mills so that 65-90% passes through
a number 325 mesh (45[tm) sieve. Granular activated carbon (GAC) consists of any
larger activated carbon particles and is the portion of activated carbon used in the packed
bed contact basins of water treatment plants.
Adsorption of a substance onto activated carbon in water occurs when the
substance is concentrated more on the surface of the activated carbon than it is in the bulk
solution. The substance that accumulates is called the adsorbate and the activated carbon
is termed the adsorbent. Physisorption, or physical adsorption, occurs when the
accumulation results from dispersion forces while chemisorption, or chemical adsorption,
results from an exchange or sharing of electrons between the adsorbate and the surface of
the activated carbon. The properties of activated carbon that make it ideal as an
adsorbent are a significant surface area to which the adsorbate may accumulate and an
extensive internal pore structure in which the adsorbate may become trapped, thus
encouraging its accumulation.
The relative hydrophobicity of the adsorbate is a driving force for its
accumulation on the adsorbent. Thus, the adsorption of a substance generally decreases
with increased contaminant solubility in water (Snoeyink and Summers 1999). The pore
size of the activated carbon also plays a major role in adsorption as, for example, a large
fulvic acid molecule will be unable to fit into small pores, or micropores, which have
diameters less than 2 nm. An activated carbon with a large volume of mesopores, with
diameters between 2 and 50 nm, and macropores, with diameters greater than 50 nm,
would be better suited for fulvic acid removal. Conversely, a larger microporous volume
is preferred when the target compound is a small molecule. The surface chemistry of the
activated carbon also affects adsorption, particularly the presence of surface oxygen
functional groups as will be discussed extensively in Section 4.1.2.
The rate of removal, or adsorption kinetics, is governed by the rates of the four
steps composing physical adsorption. The first step, bulk transport, where the adsorbate
moves from the bulk solution to the boundary layer of the adsorbate, is encouraged by the
turbulent flow created in a packed bed of GAC. Film resistance to transport follows bulk
transport and involves the transport of the adsorbate through the stationary layer of water
surrounding the adsorbent. Film resistance may also be minimized by increased flow rate
past the adsorbent. The third step is internal pore transfer where the adsorbate travels
through the adsorbent's pores via surface or pore diffusion to existing adsorption sites.
The third step is often the rate-limiting step as the final step, physical adsorption, occurs
almost instantaneously, so it has little effect on the rate of the overall reaction (Snoeyink
and Summers 1999). The rate of adsorption, therefore, is influenced by the same
variables affecting diffusion rate, including the concentration gradient and the
temperature of the system (Bansal et al. 1988).
1.3 Current Uses of Activated Carbon
GAC can be used as a substitute for typical granular filter media, in that it
removes suspended matter, though, through adsorption, GAC is also capable of removing
organic compounds such as synthetic organic chemicals, natural organic matter, and taste
and odor causing compounds. Synthetic organic chemicals (SOCs), such as most
pesticides, may be contaminants in both ground and surface water. SOCs at chronic
exposure levels may cause neurological and kidney effects and, for some pesticides,
cancer (Hammer and Hammer 2001). Natural organic matter (NOM), a mixture of fulvic
and humic acids, as well as hydrophilic acids, and carbohydrates, can impart color to
water and may also react with chlorine, added for disinfection purposes, to create
disinfection by-products (DBPs). DBPs, such as chloroform, a trihalomethane, are
considered carcinogenic (Hammer and Hammer 2001). It is, therefore, important to
reduce the NOM concentration in the water before disinfection with chlorine occurs. The
adsorption of NOM onto GAC is enhanced if calcium complexes with the NOM prior to
adsorption by activated carbon (Frederick et al. 2001a). The effects of this enhanced
adsorption relationship will be discussed throughout this paper. Taste and odor causing
compounds, considered the primary reason water treatment utilities use activated carbon,
may be consequences of biological growth or industrial activities. An example of a taste
and odor causing compound resulting from biological growth is 2-methylisoborneol
(MIB), which can be produced by cyano-bacteria, or blue-green algae. MIB, described as
having an earthy-musty odor, enters a water treatment plant when the source water is
surface water experiencing a bloom in the growth of the MIB producing organism.
1.4 Reactivation of Activated Carbon
During normal use, activated carbon will eventually become saturated with
adsorbates so that the treated water exceeds the desired level of adsorbate removal, or
reaches breakthrough. At this point, the spent activated carbon can be reactivated to
restore the adsorption capacity of the GAC so that it can be reused rather than simply
disposed of and replaced with virgin GAC. The time to reactivate a GAC bed can range
from 6 months to 5 years, depending on the type of adsorbate, the influent concentration,
and the desired treatment level.
Steam or chemical regeneration may be employed, though thermal reactivation is
the most commonly employed activation scheme. Thermal reactivation consists of four
steps: drying, desorption, pyrolysis, and gasification. The drying process removes water
and some highly volatile adsorbates. At higher temperatures, thermal desorption occurs
with vaporization of volatile adsorbates and decomposition of unstable adsorbates to
more volatile components. A pyrolytic process follows using high temperatures, between
650-850'C, in an inert environment that converts heavy or non-volatile adsorbates to
char. The final step, gasification at temperatures above 700'C in steam, C02, or a
combination of both, involves desorption of the vapors and gaseous products of char and
their exit from the pores of the reactivated carbon (Clark and Lykins 1989, Cannon et al.
1993, Snoeyink and Summers 1999).
Reactivation has several effects on the activated carbon, aside from the intended
effect of, ideally, restored adsorption capacity. Mass loss during the activation process
may range from 10-15% of the original GAC and is the result of loss through the transfer
of GAC from the treatment site to the regeneration site and/or burning of the GAC during
reactivation (Clark and Lykins 1989). The loss is compensated for by the addition of
virgin GAC before the reactivated GAC is returned to service. The surface chemistry of
the activated carbon is also affected by reactivation as most oxygen-containing functional
groups are stripped during reactivation at temperatures greater than 700'C and in an inert
environment (Menendez et al. 1996, Pereira et al. 2003).
Inorganic molecules, such as calcium, may be adsorbed during normal operation
of a GAC filter and are not removed during the activation or reactivation processes. The
inorganic molecules remaining on the surface of the reactivated carbon prevent these
adsorption sites from being freed for further adsorption and thus compromise the ability
of the reactivation process to restore the adsorption capacity of the carbon (Clark and
Lykins 1989). One well-documented effect of calcium presence in the reactivation
process is calcium catalysis (Knappe et al. 1992, Cannon et al. 1993, Mazyck and Cannon
2000, Mazyck and Cannon 2002). Calcium catalysis increases the number of active sites
on the surface of the carbon rather than, in the traditional sense of a catalyst, reducing the
activation energy. Calcium catalyzes the gasification step of reactivation and is
characterized by increased mass losses during the reaction process and pore enlargement
(Knappe et al. 1992). As discussed in Section 1.2, the pore size of the activated carbon
can affect its ability to adsorb certain compounds, and an increase in pore size may make
it unsuitable for certain applications. There is also cost associated with the addition of
virgin GAC to replace the increased mass lost during the reactivation process.
Another result of reactivation is the incidence of pH excursions, an increase in the
pH of treated water when reactivated carbon is returned to service (Farmer et al. 1996,
Farmer et al. 1998). The duration and extent of the pH excursion will determine its
impact on the system. Water with high pH levels is unfit for distribution to customers
and also exceeds national pollution discharge elimination system (NPDES) discharge
limits due to environmental impact concerns. An anion-exchange mechanism has been
proposed for pH excursions with virgin activated carbons, in which anions present in the
water displace hydroxyl (OH-) ions from the GAC surface (See Section 3.1.3) (Dussert et
al. 1995). Methods to reduce the pH of the treated water on-site can be both time-
consuming and expensive. Oxidation of the reactivated carbon using air has been
proposed for the creation of a pH stable carbon as well as treatment with carbon dioxide
(C02) (Dussert et al. 1995, Farmer et al. 1996, Farmer et al. 1998). Methods to overcome
pH excursions will be discussed in-depth in Section 4.1.1.
HYPOTHESIS AND OBJECTIVES
* Calcium chemistry (CaO + H20 Ca(OH)2; Ca(OH)2 Ca2+ + 20H-),
rather than the chemistry of other inorganic constituents of carbon or
carbon basicity, dominates in pH excursions;
* Air treatments
The addition of carboxylic acid groups to the carbon surface via air
treatment between 350C and 450'C will balance the basicity effect
created by calcium chemistry;
Air treatments add acidic oxygen-containing functional groups to
the carbon surface via chemisorption of oxygen;
Air treatments decrease adsorption capacity for MIB due to the
preference of the acidic oxygen-containing functional groups for
* CO2 treatments
The physisorption of CO2 to activated carbon will combat pH
excursions because the CO2 will desorb from the carbon upon
wetting, forming carbonic acid (H2CO3);
CO2 treatments do not add oxygen-containing functional groups to
the carbon surface, but rather the CO2 is physisorbed to the carbon;
* Determine whether calcium within reactivated GAC contributes to pH
excursions, due to the transformation of calcium oxide to calcium
hydroxide, which can further dissociate to calcium ions and hydroxides;
* Determine if an anion exchange phenomenon that is associated with the
carbon surface (versus with calcium) contributes to pH excursions via
anion adsorption and subsequent release of hydroxide ions for reactivated
* Create pH stable carbons using air treatments and CO2 treatments;
* Determine the total acidity ([teq/g) of the air and CO2 treated carbons and
evaluate the effect of total acidity on the carbon's adsorption of MIB
THE ROLE OF CALCIUM IN PH EXCURSIONS FOR REACTIVATED GAC
Water and wastewater utilities thermally reactivate their spent granular activated
carbon (GAC) because thermal reactivation offers an economical approach for reusing
the exhausted material. These utilities save money by reactivating the spent GAC versus
wasting it and replacing it with virgin GAC. For example, together, American Carbon
Services (a subsidiary of the American Water Works Company), the Cincinnati Water
Works, and NORIT Americas reactivate more than 25 million pounds of spent GAC per
year. Because thermal reactivation is more economical than virgin carbon replacement,
and because it has been shown that reactivated carbon performs as well as or better than
virgin carbon (Goins 2000), thermal reactivation is an increasingly attractive industrial
Over the years, research has been carried out on two recurring issues that pertain
to thermal reactivation. The first is related to calcium catalysis (Knappe 1992, Cannon et
al. 1993, Cannon et al. 1994) and has been addressed through the development of a new
thermal reactivation protocol that overcomes the deleterious effects of calcium catalysis
(Mazyck and Cannon 2000, Mazyck and Cannon 2002). This new protocol not only
reduced the mass, volume, and micropore loss compared to conventional reactivation, but
also has been shown to increase GAC performance for removing the common odorant
MIB compared to conventionally reactivated and virgin GAC (Goins 2000).
The second issue is pH excursion: when reactivated carbon is returned to service,
several water and wastewater utilities have found that the pH of the subsequently treated
water is excessively high. Indeed, some water utilities have reported pH excursions
greater than pH 11 that have lasted for several weeks. Furthermore, many carbonaceous
precursors such as bituminous and sub-bituminous coals, wood, coconut, peat, and acid-
washed carbons have been found to produce pH excursions (Dussert et al. 1995). If left
unmitigated, this deviation in pH could cause some water utilities that discharge
backwash water to rivers to operate out of compliance, and risk being fined because their
NPDES permits often require that water pH be maintained between 6 and 9. Therefore,
some utilities request that their reactivated GAC maintain a stable pH of 6.5 to 8.5 when
it is placed back in service (i.e., with a certain "safety buffer"). Currently, to overcome
pH excursions, water utilities often either a) backwash their GAC until the pH is less than
8.5-9, b) add acid, or c) recirculate their water until the pH excursion is abated. All of
these solutions are time-consuming and expensive.
The objective of this study was to elucidate the role that calcium plays in pH
excursions. Several other pH excursion mechanisms have also been proposed in the
literature (see Section 3.1). The intent of this paper has not been to disprove any of these
other mechanisms. It is deemed plausible that calcium could cause pH excursions, thus,
the role of calcium on pH excursions was the primary focus of this study.
3.1 Literature Review
3.1.1 ph Excursions in Water Treatment
During potable water treatment, water-soluble calcium can adsorb onto GAC as a
complex with natural organic matter (NOM) (Nowak and Cannon 1997, Nowak and
Cannon 1999, Frederick and Cannon 2001, Frederick et al. 2001a, Frederick et al.
2001b). Moreover, calcium is abundant in many coal-based carbons. Thermal
reactivation of spent GAC at temperatures above 850'C decomposes the oxygenated
functionality of the NOM that complexes with Ca. This process results in the formation
of CaO, which is thermodynamically the most stable species at these high temperatures
(Mazyck and Cannon 2000). When the reactivated carbon is returned to service, the
calcined CaO can react with water to form Ca(OH)2 (equation 1). Furthermore, Ca(OH)2
can dissolve and subsequently dissociate as Ca2+ and 20H- (equation 2). The release of
OHf would elevate the pH of the subsequently treated water.
CaO + H20 -> Ca(OH)2 (1)
Ca(OH)2 Ca2+ + 20H- (2)
Calcium chemistry and thermodynamics as related to GAC and thermal
reactivation have been previously discussed at length (Cannon et al. 1994, Mazyck and
Cannon 2000, Mazyck and Cannon 2002).
A water utility in the Virginia-American Water Company, under the umbrella of
the American Water Works Company, measured increases in both pH and Ca2+
concentration in their backwash water when a reactivated GAC was placed back in
service (Figure 1). These data are representative of a full-scale GAC filter bed adsorber.
As shown, the reactivated GAC posed an adverse effect on water quality in that it
elevated the filter bed adsorber's effluent pH to greater than 11. Even after 6 days, the
pH was still above 8.5. In addition to the pH, the utility also measured the concentration
of calcium in the effluent for two days. As shown in Figure 1, as the calcium
concentration decreased, the water contact pH likewise decreased. The influent calcium
concentration in their backwash water was relatively constant over the two days. No
other operational changes could have contributed to the higher Ca2+ and OHf levels; and
these results are fairly typical, as reported by American Water Works Company
personnel. These results led to the hypothesis that the manner by which the reactivated
GAC was contributing both calcium and hydroxyl to the water was via the mechanisms
described in equations 1 and 2 above. Furthermore, it was rationalized that the deviation
from a [OH'] (molar concentration of hydroxyl ions) to [Ca2+] (molar concentration of
calcium ions) ratio of 2, in that the released OHf that originated from Ca(OH)2 would
have also been consumed to some extent by the natural buffering capacity of the water.
Thus, the data from the Virginia-American Water Company could be taken as a semi-
quantitative example of the mechanisms described in equations 1 and 2.
I 11.0 o Calcium (mg/L)
10.0 ra EC E
o O j=
8.5 + 10
0 1 2 3 4 5 6
Figure 1. Full-scale water treatment plant effluent pH and calcium concentration data
after reactivated GAC was returned to service
3.1.2 Carbon Basic Properties
Leon y Leon et al. (1992) investigated the basic (alkaline) nature of carbons and
proposed a mechanism that involved electron donor/acceptor and pyrone-type
interactions. They showed that the oxygen content of the carbon dictated the importance
of either interaction, and that these basic sites were created as a result of high temperature
(> 700C) heat treatments. Indeed, Papirer et al. (1991) have also shown that the number
of basic sites can be maximized by heat treatment between 750'C and 850'C.
Temperatures higher than 700'C are commonly employed when manufacturing and
Leon y Leon et al. (1992) used highly pure (i.e., inorganic-free) carbons, which
were chemically or thermally treated, in order to determine the surface basicity. Through
HCI adsorption, stepped temperature-programmed desorption, electrophoresis and mass
titration, they argued in favor of the following mechanism:
C, + 2H20 <- CH30 + OH- (3)
where Cn is a graphene layer in which the itinerant 7t electrons may become partially
localized as a result of H30+ addition. Leon y Leon et al. (1992) showed that these
oxygen-free Cn sites can adsorb protons from solution, and thus render the surface
positively charged. Therefore, this electron donor/acceptor interaction was proposed to
be responsible for the basicity of the carbon surface. In addition to this mechanism, they
could not discount that pyrone-type surface groups were also responsible for carbon
basicity. The electron donor/acceptor mechanism is predominant in carbons with low
oxygen concentrations, while interactions with pyrone groups could predominate in
carbons with high oxygen content. Barton et al. (1997) also investigated acidic and basic
sites on carbon and confirmed that acid functionalities are associated with chemisorbed
oxygen, whereas basic sites are concentrated at delocalized electrons (Cr) in the basal
plane. However, they have not supported the notion that pyrones can contribute to
carbon basicity. Darmstadt and Christian's (2003) findings support those of Barton et al.
(1997) in that no significant correlation between oxygen groups and basic equivalents on
the surface of carbon blacks was found and that the dominant portions of the basic sites
present on the carbon blacks are associated with the basal planes of the graphine layer.
Menendez et al. (1999) did show, however, that complex tricyclicc) pyrones, which are
located at the edges of graphene layers, could have a pKa around 12.7 and that they can
contribute significantly to carbon basicity. It is evident that there is no general consensus
as to the dominant contributor to carbon basicity or to the basic strength of the
contributor. Montes-Moran et al. (2004) provide a summary of the virtues and
drawbacks of the two proposed contributors to carbon basicity discussed herein, basic-
oxygen-containing groups and delocalized xt-electrons of the basal planes.
3.1.3 pH Excursion Mechanism for Virgin Carbons
A mechanism involving anion exchange with hydroxide ions following carbon
protonation has been proposed by Farmer and coworkers to explain pH excursions (Carr
and Farmer 1995, Dussert et al. 1995, Farmer et al 1996). These authors suggested that
pH excursions are a function of the carbon surface that is altered during high temperature
activation or reactivation. Activated carbons exposed to high temperatures in a reducing
atmosphere during manufacturing and reactivation tended to adsorb strong acids in water.
Garten and Weiss (1957) classified these as H-type carbons. In contrast, L-type carbons
are those produced by surface oxidation; and these are known to adsorb strong bases in
water (Garten and Weiss 1957). Farmer et al. (1996) argued that this acid adsorption
during water treatment might involve the protonation of pyrone-type surface groups or
other similar functionalities on the carbon surface, in agreement with Leon y Leon et al.
(1992). In addition, they suggested that loading the carbon with sulfate, chloride, or other
anions that are present in water could neutralize the positive charge on the carbon
surface. Furthermore, they proposed that this charge neutralization could also occur
through an anion exchange process involving sulfates and hydroxides on the carbon
surface. In other words, they surmised that the carbon could adsorb sulfates or other
anions from water while releasing hydroxyls, and that this ion exchange process could
cause the pH to rise. Farmer et al. (1996) did acknowledge that this anion exchange
mechanism requires additional testing and confirmation. To date, however, there has
been no further evaluation presented in the literature that would either prove or disprove
Dussert et al. (1995) supported this anion exchange mechanism because they
found no significant pH excursion when virgin GAC was immersed in Milli-Q (anion-
free) water, and they concurred that anions were required to trigger a pH effect. In
contrast, as discussed below, the results show that pH excursions occur even in Milli-Q
water when using reactivated carbons. Dussert et al. (1995) came to their conclusions
after studying virgin GACs that removed 2 to 9 mg of sulfate/g GAC. In Milli-Q water
spiked with 80 mg sulfate/L, they measured the water contact pH and sulfate capacity for
several virgin carbons. Not only did they find that pH excursions only occurred when
anions such as sulfate were present, but they also reported that, as the sulfate capacity
increased, pH also increased. In their study, the carbons that did not exhibit pH
excursions also did not significantly remove sulfate ions from solution (< 0.7 mg/g).
They also found that sulfate (SO4 ) removal was greater than nitrate (NOs3) or chloride
(CYl) removal. They did not measure calcium concentrations for their carbons; however,
they did find that the ash content of the carbon could not be correlated with pH
3.2.1 Activated Carbon Samples
Several coal-based reactivated GAC samples were received from the NORIT
Americas, Inc. plant in Pryor, OK. These carbons had served in potable water treatment,
had become spent, and had been subsequently reactivated in Pryor, OK. The calcium
content of these carbons, or the GAC calcium concentration, was determined by ashing
the GAC, dissolving the ash in acid, then analyzing for calcium via inductively coupled
plasma (ICP) emission spectroscopy (Leeman Labs PS3000UV, Framingham, MA)
following the procedure described by Nowack and Cannon (1997), except that emission
spectroscopy was used instead of absorption spectroscopy. (Analyses were conducted at
the Materials Characterization Laboratory, PSU.) The reproducibility of the results was
3.2.2 Water Contact pH
The water contact pH for these carbons was determined by mixing approximately
2 grams of as-received GAC with 80 mL of Milli-Q water for 30 minutes, then measuring
the pH. This ratio of GAC to solution was the same used by Dussert et al. (1995) and
Farmer et al. (1996). In initial exploratory experiments, pH measurements (Standard
Methods for the Examination of Water and Wastewater procedure 4500-H ) were taken
every five minutes, and it was observed that after 20 or 30 minutes the pH reached a
pseudo-equilibrium level (less than 0.1 change in pH over 5 minutes). Dussert et al.
(1995) used a similar technique and concluded that this protocol provides a good
prediction of the pH that would be measured during water treatment service. For the
experiments described herein, each reactivated carbon was sampled twice and the pH for
these two samples was within 0.4 pH units. (Six of the seven samples varied by less
than 0.1 pH units.)
3.2.3 Calcium Solution Concentration
After the water contact pH experiment was completed, the GAC was separated
from the solution by utilizing a vacuum filtration apparatus (0.45 |tm Whatman filter
paper), and the 80 mL of Milli-Q water was analyzed for calcium by inductively coupled
plasma (ICP) emission spectroscopy, as above. Calcium levels leached out of the GAC
during the 30-minute water contact pH experiment were determined. The measured
concentrations of the replicas were within 3.5% of one another. Aluminum, iron, and
magnesium concentrations leached from the samples were also measured.
3.2.4 GAC Sulfate Capacity
In selected experiments, the water contact pH was measured in Milli-Q water that
contained 80 mg SO4 /L. The sulfate concentration of this solution was verified using a
HACH DR/2000 spectrophotometer (Loveland, CO) by utilizing their sulfate method 680
(SulfaVer 4). After each experiment, the carbon was separated from the solution using
vacuum filtration with a 0.45 |tm Whatman filter paper. The filtrate was then analyzed
for sulfate concentration as described above. The difference in sulfate concentration
between the stock solution and the filtrate represented the sulfate adsorbed on the
activated carbon. The reproducibility of sulfate concentrations was within 5%.
3.3 Results and Discussion
3.3.1 Calcium Content Versus Water Contact pH
Water contact pH experiments were conducted for seven reactivated carbons. The
duplicated results are summarized in Figure 2. As shown, all the pH values were above
8.5 and none of the carbons exhibited pH greater than 11.0. It is anticipated from
empirical experience that most activated carbons that had become loaded with calcium-
NOM complexes will exhibit pH excursions with pH values greater than 8.5. These data
confirmed the pH excursions that water utilities face when they use reactivated carbons.
'E R2 = 0.98
y = 0.5x 5.56
(D 0 -0.80
8.5 9.0 9.5 10.0 10.5 11.0 11.5
Water Contact pH
Figure 2. Correlation between log GAC calcium content and water contact pH exhibited
by reactivated carbon
Calcium chemistry is obviously important for pH excursions, since the water
contact pH increased linearly (R2 = 0.98) with the log of the calcium content of the GAC.
Also, the slope of the line, 0.5, is reminiscent of equation 2, which designates a 1:2 ratio
of moles of calcium to moles of hydroxyl ion. The slope, therefore, supports the
relationship between calcium and pH excursions.
It should be noted that not all of the GAC calcium content would be accessible to
water. The portion that would be fully accessible is that which resulted from
complexation with NOM and subsequent adsorption onto the GAC surface where it
would be expected to undergo the reactions shown in equations 1 and 2. In contrast, the
calcium that was a component of the original activated carbon may or may not have been
accessible to the water, since it could have been imbedded within the graphene layers of
the carbon. For the reactivated GACs of somewhat consistent parent source that were
evaluated herein, the inaccessible carbon posed a fairly uniform background calcium
Figure 2 showed that the contact pH increased as calcium accumulated during
normal operation in a water treatment plant. Despite the interesting correlation between
calcium content of the GAC and water contact pH, only calcium leached from the GAC,
as discussed in Section 3.3.2, should be used to predict pH excursions. Whether there
exists a cause-and-effect relationship in the correlation of the calcium content of the GAC
to the water contact pH is the focus of the ensuing discussion. Each one of these
reactivated carbons exhibited a pH excursion in Milli-Q water; and this contrasted with
the behavior of virgin GACs that were studied by Dussert et al. (1995). Those authors
did not measure a significant pH excursion for virgin carbons in Milli-Q water. This
discrepancy highlights the distinction between the accessible calcium that originates from
adsorbed calcium-NOM, versus the partially accessible calcium that occurs in virgin
3.3.2 Predicted pH Based on Calcium Leached from GAC
After each water contact pH experiment, the concentration of calcium that leached
out of the GAC and into solution was measured. This amount of calcium was used to
predict the pH of the water based on equations 1 and 2: for every mole of Ca2+ released
from dissociated Ca(OH)2, two OH- moles would be released. This relation is shown in
2[Ca2+] = [OH-] (4)
To calculate the water contact pH in an open system, with contributions to acidity
from dissolved C02, and the resulting formation of carbonate (C032-) and bicarbonate
(HC03-) ions, a charge balance was used to relate the molar concentrations of the
pertinent ions according to
[H ] + 2[Ca2+] = [OH-] + 2[C032-] + [HCO3-] (5)
The equation was then solved for the concentration of H and the resulting water contact
pH, through trial and error for each sample and its related concentration of calcium in
The results are presented in Figure 3. Duplicate analyses are plotted; and a good
linear relationship is evident. The dotted line in Figure 3 corresponds to a perfect match
where the predicted water contact pH (excluding reactions with carbonate acidity) is
exactly equal to the measured water contact pH. The predicted water contact pH in an
isolated system is seen to be always higher than the measured value by about 0.3 0.1
pH units. The discrepancy reflects a uniform amount of bicarbonate anion and H+ that
was generated by the side reaction CO2 + H20 -> HC03 + H'. This reaction occurred in
the mixing beaker that was open to the atmosphere. An additional experiment was
performed on the sample with the highest GAC calcium content to test this explanation.
The sample was split to perform water contact pH experiments in open air and an
analogous experiment in a nitrogen filled glove box. The open-air experiment yielded a
water contact pH of 10.8 and the glove box experiment yielded a water contact pH of
11.3. The water contact pH predicted by equations 1 and 2, based on leached calcium for
this sample, was found to be 11.3. This experiment demonstrates the role of ambient CO2
in altering the measured pH values compared to the predicted values. As the slope of 1.0
in Figure 3 suggests, if the acidic effects of CO2 were removed, the line would be
expected to lie on the perfect match line represented by the dotted line in Figure 3.
It should be noted that the predicted water contact pH based on an open
atmosphere is far below the water contact pH measured in the open atmosphere, with an
approximate 2.7 unit decrease in pH compared to t he predicted water contact pH in an
isolated system. It is possible that the system is not at equilibrium when the water
contact pH is measured, and therefore not enough time has passed for the water contact
pH to decrease to the level predicted. It would therefore be expected that if the water
contact pH experiments were carried out for longer than 30 minutes, that the water
contact pH would decrease and eventually reach the theoretical water contact pH as
predicted by Equation 4. Contributions to basicity via EDA or pyrone-type groups on the
surface of the carbon may also affect the pH. The carbon's contribution to basicity is not
included in the theoretical calculations of an open atmosphere water contact pH, and
therefore this theoretical calculation would be expected to under predict the final water
11.5 y = 1.0x + 0.20
C R2= 0.93
M 7E 10.5
S =~ 9.5
g O 9.0
8.0 8.5 9.0 9.5 10.0 10.5 11.0 11.5 12.0
Measured Water Contact pH (Milli-Q Water)
Figure 3. Predicted pH based on the calcium that leached out of the reactivated carbon in
an isolated and an open system versus the water contact pH measured in Milli-Q water
To verify that calcium is primarily responsible for the pH excursions, rather than
other inorganics, the leaching of other inorganics was also measured namely aluminum,
iron, and magnesium (Table 1). Aluminum, iron, and magnesium appear ubiquitously in
raw water and they may be added during chemical treatment in the water treatment plant
(Knappe et al. 1992, Cannon et al. 1993). The minimum detection limit (MDL) for iron
on the ICP is 0.02 ppm and for aluminum the MDL is 0.03 ppm. As was proposed, it is
evident that calcium accumulates onto the GAC samples, and therefore leaches, to a
greater extent than aluminum, magnesium, or iron. In fact, calcium concentrations are at
least ten times higher than the concentrations of other metals leached from the samples,
signifying that the chemistry of calcium indeed dominates the pH excursion effects.
Table 1: Metal concentrations leached from the activated carbon samples
Water Aluminum Iron Magnesium Calcium
Contact pH (ppm) (ppm) (ppm) (ppm)
8.9 Below MDL Below MDL 0.03 0.88
9.6 0.11 Below MDL 0.17 4.85
10.1 0.80 Below MDL 0.20 7.00
10.6 1.17 Below MDL 1.01 16.8
10.6 1.82 Below MDL 1.38 29.0
10.7 3.80 Below MDL 1.37 34.2
10.9 3.46 Below MDL 0.76 38.5
3.3.3 pH of Steam-Cured Plus Ramped-Temperature Reactivated Carbons
A spent bituminous coal-based carbon that contained 2.4 wt. % calcium was
thermally reactivated by means of the steam-curing plus ramped-temperature protocol
that was previously developed to overcome calcium catalysis (Mazyck and Cannon 2000,
Mazyck and Cannon 2002). Briefly, this protocol entailed steam-curing at 3750C for 1 h
followed by ramping in N2 to either 650'C, 750'C, or 850'C. The carbon was not held at
these high temperatures for any extended time, and reactivation was terminated as soon
as the final temperature was reached. Immediately after reactivation, the water contact
pH was measured in each case and the results are shown in Figure 4. It is well known
from the literature that as the final ramped-temperature increases above the calcining
temperature of 800C-850C; CaCO3 would have increasingly decomposed to CaO
(Boynton 1980, Babushkin et al. 1985, Cannon et al. 1994, Mazyck and Cannon 2000).
Because of this calcining effect, the spent GAC exposed to 8500C had the highest pH.
The significance of the highest temperature is that CaO is formed and is available to react
via equations 1 and 2. Other calcium species (e.g., CaCO3 or CaSO4) are less likely to
participate in these reactions because they tend to dissolve less readily in water that has a
neutral or basic pH. Samples exposed to 6500C or 750'C had lower pH values because
neither was heated enough to promote complete CaO formation, as this calcining reaction
is diffusion limited (Boynton 1980).
Pyrolyzed Steam-Curing Steam-Curing Steam-Curing
and Oxidized plus Ramped- plus Ramped- plus Ramped-
at 850C Temperature Temperature Temperature
to 850C to 750C to 650C
Figure 4. Comparison of water contact pH for conventionally reactivated and steam-
cured reactivated carbons to 3 different temperatures
The same spent carbon (from above) that was steam-cured was also reactivated
conventionally using pyrolysis and steam-activation at 8500C. Its water contact pH was
11.5 (see Figure 4); this significantly higher pH occurred because this GAC had been
exposed to steam at 8500C for 15 minutes. At temperatures above ca. 527C for any
partial pressure of steam, Ca(OH)2 decomposes to CaO (Mazyck and Cannon 2000).
Total calcium concentration provided a stronger correlation with pH than did CaO
surface area (data not shown). This is because CaO surface area measurements reflect
only the external fraction of the total CaO available for reaction with water.
3.3.4 Influence of Acid and Base Washing on Water Contact pH
It is noted that high-temperature pyrolysis or steam exposure could also gasify
oxygen-containing functionality on the carbon surface, and this could also elevate the pH.
Thus, to further distinguish the role of calcium from the role of diminished surface
acidity, a non-oxidizing acid was employed to remove the calcium from the GAC to
isolate the surface acidity effect. Specifically, a spent carbon was treated with ten bed
volumes of 1 M HC1. This treatment reportedly removes essentially all of the metals,
including calcium, from the GAC (Knappe et al. 1992, Cannon et al. 1993). After acid
washing, this sample was washed with ten bed volumes of 15.8 M NaOH to remove any
residual H+ then rinsed with 30 bed volumes of deionized distilled water to neutralize the
pH and desorb any previously sorbed HF or OHf. Then, the carbon was reactivated. The
pH of this conventionally reactivated carbon, that had been acid-base treated, was
considerably lower (8.5) than the pH of the conventionally reactivated Ca-containing
carbon (11.5) as shown in Figure 4. These results are as would be expected by the
mechanism of equations 1 and 2.
3.3.5 Comparisons of pH in Milli-Q and Sulfate-Containing Water
Dussert et al (1995) hypothesized that an ion exchange mechanism was
responsible for pH excursions on virgin GAC. To discern if this mechanism was also
valid in describing pH excursions on reactivated GAC, this protocol was tested on the
reactivated carbons herein. The concentration of sulfate was measured and plotted versus
the water contact pH as measured in water that contained 80 mg sulfate/L (Figure 5).
Sulfate uptake was very low for the reactivated GACs; and in some cases it was negative,
indicating that the reactivated GAC itself contributed sulfate to the solution. In light of
these results, it was concluded that the uptake of sulfate and the subsequent release of
hydroxyl ions were not responsible for the observed pH excursions on the reactivated
GACs. Commercial GAC contains a considerable level of sulfur in its pyrite content
(Boynton 1980, Cannon et al. 1997). It is proposed that during thermal reactivation,
sulfur dioxide formed from the inorganic pyrite impurities in GAC, and then the residual
SO2 (i.e., that which remained present as the reactivated GAC cooled) adsorbed onto CaO
to form CaSO3 or CaS04 (Munoz-Guillena 1994), and these subsequently leached out of
the GAC and dissolved in the effluent water. Also, this sulfate could have adsorbed, to a
limited extent, out of the municipal water source. Then, at the high water contact pH that
was incurred by the Ca(OH)2 mechanism, the base ion-exchanged with the adsorbed
R2 = 0.91
( L M) 0.0
8.0 8.5 9.0 9.5 10.0 10.5 11.0
Water Contact pH (80 mg/L sulfate water)
Figure 5. Sulfate uptake for several reactivated carbons versus their water contact pH as
measured in 80 mg/L sulfate water
The data shown here support the contention that calcium chemistry (i.e., equations
1 and 2) is a plausible mechanism to account for pH excursions of reactivated carbons
because as the calcium concentration increased, the water contact pH also increased.
Indeed, the correlation coefficient (R2) between the log of the calcium content and water
contact pH exceeded 0.9.
When calcium is removed from reactivated carbons, the magnitude of the pH
excursion is drastically decreased. The acid (HC1) and base (NaOH) washing procedures
removed the calcium, but did not significantly influence the carbon's functional groups
that existed prior to this treatment, since the HCI and NaOH were neither oxidizing nor
reducing agents. Typically, organic free water has a pH of 5-6. When the acid and base
washed reactivated carbon, assumed to be free of accumulated calcium, was tested, it
altered the pH to 8.5. In contrast, when calcium was present on the reactivated carbon,
the pH rose to 11.5. Therefore, one could conclude that for these reactivated carbons,
calcium exerted the greater influence in causing high pH excursions.
Sulfate uptake and hence an anion exchange mechanism did not provide an
explanation for pH excursions for reactivated carbons because as the sulfate uptake
increased, the water contact pH decreased.
CREATING A pH STABLE CARBON
Previous experiments have shown that calcium chemistry is a plausible
mechanism behind the pH excursions experienced in water treatment when reactivated
GAC is returned to service. In addition, the anion exchange mechanism proposed by
other researchers as the mechanism behind pH excursion in virgin GAC was not found to
explain the pH excursions experienced with reactivated GAC. An understanding of the
mechanisms behind pH excursions with reactivated GAC will aid in the formulation of a
method to overcome these pH excursions.
The option of not reactivating the GAC, thus avoiding the reaction process and
the resulting pH excursions, is not a cost-saving solution, as reactivation of GAC is a
more economical approach than simply replacing the spent GAC with virgin GAC.
Furthermore, as demonstrated by Farmer et al. (1996), virgin carbons can cause pH
excursion as well. Currently, water treatment plants treat the pH excursions as they
occur, including adjustment of the pH of the treated water and backwashing the GAC
until the GAC no longer creates a pH excursion. Altering the reactivation process to
avoid the creation of calcium species that will participate in the proposed calcium
chemistry reactions, using a reactivation scheme proposed by Mazyck and Cannon (2000
and 2002) has been investigated in this paper as a successful means of avoiding pH
excursions experienced with reactivated GAC. Patents are held on methods to overcome
pH excursions that propose treatments of reactivated GAC with either air or carbon
dioxide will create pH stable carbons that will not create pH excursions when the
reactivated GAC is returned to service (Dussert et al. 1995, Farmer et al. 1998)
The objective of this study was to investigate the effects of air and carbon dioxide
treatments on the water contact pH of a reactivated GAC. The mechanism responsible
for the creation of a pH stable carbon via these treatments was also investigated, with a
focus on the effect of the treatments on the surface chemistry of the reactivated GAC.
Finally, the performance of the treated GAC was investigated through equilibrium batch
studies using MIB as the adsorbate.
4.1 Literature Review
4.1.1 Methods to Overcome pH Excursions
The three broad categories of methods to overcome pH excursions are on-site
treatments, tailoring of the reactivation process, and post-reactivation treatments
performed before the reactivated GAC is returned to service.
One on-site method that would occur prior to reactivation is the reduction of the
calcium adsorbed onto the activated carbon during water treatment. To achieve this,
either the NOM complexes with a stronger cation, thus out-competing calcium, or the
calcium-NOM complex is removed before it comes into contact with the GAC. The use
of an iron coagulant would be ideal as it could facilitate both of the above goals. Iron
(Fe3+) is a stronger cation than calcium and would displace the calcium from the NOM
though competition with NOM by hydroxyl species may hinder the process near a neutral
pH. Nowack et al. (1999) proposed that, at near neutral pH, soluble iron hydroxide
would precipitate depleting the amount of iron available for further reactions with NOM.
At adequate pH levels, the iron would also coagulate the calcium-NOM complexes
already formed and this floc would be allowed to settle out before reaching the GAC.
Nowack et al. (1999) and Frederick et al. (2001) have shown that the addition of iron
coagulant to a system containing both calcium and NOM yielded a decrease in the
amount of calcium-NOM complexes adsorbed to the carbon as well as a relatively level
iron loading onto the GAC. These results indicate that, not only is iron addition
successful in decreasing calcium loading, but that the primary mechanism responsible for
this is coagulation of the iron-NOM and/or calcium-NOM complexes. The accumulation
of iron is not detrimental to the reactivation process as iron catalysis is suppressed by the
presence of sulfur, which is ubiquitous in coal-based carbons (Nowack et al. 1999). A
near neutral pH was indeed found to hinder iron-NOM formation as the hydroxide
species out-competed the NOM for complexation with the iron ions. Therefore, Nowack
et al. (1999) and Frederick et al. (2001) concluded that a combination of pH adjustment
to approximately 6.0 pH and iron addition, on the order of 5 mg/L, are the most effective
in reducing calcium loading. Water treatment facilities that operate above this range of
pH are expected to accumulate higher levels of calcium onto their GAC.
Other on-site treatments to overcome pH excursions after the reactivated carbon is
returned to service include backwashing, recycling of the high pH water, and pH
adjustments before the treated water is discharged. Both backwashing and recycling
involve running water through the GAC until the pH reaches acceptable levels, which
reduces the yield of treated water through the system (Farmer et al. 1998). In addition,
backwashing may require a holding tank, as the high pH water cannot be discharged,
which will restrict the allowable backwashing time (Carr and Farmer 1995). Acid
treatment of the high pH water prior to discharge to stabilize the pH can be time-
consuming and expensive, due to additional monitoring equipment
Acid-washing a traditionally reactivated carbon prior to use, aside from being
time-consuming and expensive, may not even produce a pH stable carbon (Dussert et al.
1995). In addition, oxygen surface complexes are introduced to the surface of the carbon
if an oxidizing acid is used and this can impact the adsorption capabilities of the
reactivated carbon, to be discussed further in Section 4.1.2. Depending on the acid, the
micropore volume may also decrease as a result of treatment.
Section 3.3.3 described a reactivation protocol that can overcome calcium
catalysis and the resulting pH excursions by reactivating the carbons at temperatures
below that at which CaO is formed, 850'C. The CaO form participates in the calcium
chemistry mechanism discussed in Section 3. A decrease in the amount of CaO available
to participate in the calcium chemistry reactions will result in a decreased pH excursion.
Oxidizing the reactivated carbon, outlined in a patent by Dussert et al. (1995), or
treating the reactivated carbon with carbon dioxide, outlined in a patent by Farmer et al.
(1998), have been offered as means to create a pH stable carbon. Both patents propose
that oxidation or CO2 treatment, depending on the patent, neutralizes the surface of the
reactivated carbon, thus inhibiting any anion-exchange during water treatment. As the
focus of Section 3 was the role of calcium in pH excursions and the result was that the
anion-exchange mechanism does not fully explain the excursions found in reactivated
carbons, there must be an additional explanation for why these treatments are able to
create pH stable carbons.
The optimum temperature for chemisorption of oxygen, using air as the oxidizing
agent, occurs around 400'C (Wiegand 1937, Puri 1970). The maximum capacity for a
carbon to adsorb bases, or its acidic nature, has also been shown to lie close to 4000C
(Puri 1970). The fact that the maximum chemisorption of oxygen and the maximum
capacity to adsorb bases occur at the same temperature has led many researchers to the
conclusion that the acidic nature of carbon is closely related to its oxygen content.
Subsequent research has supported this conclusion (Puri 1970, Otake and Jenkins 1993,
Barton 1997, Ania et al. 2004). The carboxyl group (COOH) has been suggested as the
oxygen-containing functional group responsible for the acidic nature of carbon by several
researchers (Krut and de Kadt 1931, Puri 1970). The temperature at which these
carboxyl groups were developed was 400'C in either oxygen or air. More recently, Otake
and Jenkins (1993) researched air treated carbons and found a positive linear relationship
between the chemisorbed oxygen concentration desorbed as a CO2 complex and total
acidity of the carbon. Otake and Jenkins (1993) proposed that the oxygen functional
groups existing on the air-oxidized char were carboxyl groups and were formed following
the mechanism of Figure 6.
0 0 0 + H20 \ \ OH OH
\ C/ \/
Formed during air treatment
Figure 6. Proposed mechanism for the formation of COOH (Otake and Jenkins, 1993)
Given this information, it can be proposed that air treatment of reactivated
carbons adds acidic oxygen-containing functional groups, in the form of carboxyl groups,
which upon addition to water form carboxylic acid. Thus, air treatment balances the
increased pH caused by either the anion-exchange or the calcium chemistry mechanism
rather than inhibiting these mechanisms.
Treatment with CO2 is advantageous in that it takes place at ambient
temperatures, thereby avoiding problems with elevated temperature oxidation, while
avoiding the lengthy treatment times and high gas volumes required for air oxidation at
ambient temperatures (Farmer et al. 1998). Here, the mechanism that creates a pH stable
carbon is believed to be physisorption of CO2 to the carbon surface. Upon immersion
into water, the CO2 desorbs and reacts with water to form carbonic acid (H2CO3). As
with air treatment, it is proposed that the CO2 treatment balances out the basic effects of
the anion-exchange or calcium chemistry mechanisms.
4.1.2 Effect of Oxidation on Activated Carbon
The addition of oxygen to the surface of an activated carbon has been shown to
alter its wettability, catalytic and electronic properties, as well its adsorption capabilities
(Puri 1970, Bansal et al. 2002, Szymafiski 2002). Acidic oxygen-containing functional
groups are mainly present on the outer surface or edge of the basal plane due to limited
diffusion into the micropores (Puri 1970, Menendez et al. 1996, Ania et al. 2004). As
these outer sites constitute the majority of the adsorption surface, the concentration of
oxygen on the surface has a great impact on the adsorption capabilities of the carbon
The presence of oxygen imparts a polar nature to the activated carbon, which
results in a preference for removal of the more polar component of a solution, e.g., H20
(Kipling and Gasser 1960). El-Sayed et al. (2003) found that the adsorption of valeric
acid onto activated carbon decreased as the density of acid groups on the surface
increased. El-Sayed et al. (2003) suggested that the polar nature of the activated carbon
surface, associated with its high surface oxygen group content, resulted in the preferential
adsorption of water molecules. In addition, the presence of carboxylic groups acted as
physical obstacles to the interaction of the valeric acid and the activated carbon surface.
Bansal et al. (2002) also attributed the decrease in phenol adsorption with increased
oxidation of activated carbon to the increasing hydrophilic character of the oxidized
carbons. Ania et al. (2004) found a decrease in the adsorption of salicylic acid as the
oxygen-containing functional groups increased at low adsorbate concentrations. At high
levels of salicylic acid, however, Ania et al. (2004) found that the adsorption capacity of
the carbon was similar to an activated carbon that had not been oxidized and proposed
that the higher concentration of valeric acid allowed the water molecules adsorbed onto
the hydrophilic carbon sites to be displaced.
Pendleton et al. (1997) researched the effects of surface chemistry on the
adsorption of MIB and linked decreased adsorption with an increase in the hydrophilic
nature of the oxidized carbon. They also found a strong correlation (R2=0.99) between
oxygen content and the number of hydrophilic sites. A subsequent study by the same
research team (Considine et al. 2001) further supported the link between oxygen content
and decreased adsorption of MIB. In this study, the adsorption of MIB in a
dichloromethane solvent, a non-polar solvent, was investigated as well, as opposed to
previous experiments in water. The results showed that adsorption of MIB was
independent of oxygen content in this solvent; further supporting the contention that
adsorption of water as a result of oxidation impedes MIB adsorption.
Karanfil and Kitis (1999) found that surface oxidation, and the resulting density
of strong acid functionalities, also decreased the adsorption of dissolved organic matter
(DOC) (the dissolved portion of NOM). Water clusters forming around the polar
hydrophilicc) functional groups on the surface reduced the ability of the DOC to reach the
smaller pore sizes needed for its adsorption.
The chemisorption of oxygen and the resulting functional groups have been found
to destroy 7t-electrons due to electron localization (Leon y Leon et al. 1992, Ania et al.
2004). The adsorption of phenol has been attributed to the electron-donor acceptor
complexes formed between the basal planes on the surface of the carbon and the aromatic
ring of the phenol (Bansal et al. 2002, Ania et al. 2004). A decrease in phenol adsorption
on oxidized samples has been shown by a number of researchers (Bansal et al. 2002,
Ania et al. 2004). As the bonding sites are destroyed through chemisorption of oxygen,
the adsorption of phenol decreases (Ania et al. 2004). Pereira et al. (2003) performed a
study on anionic dye (acidic), where the presumed adsorption mechanism involved the
delocalized 7t-electrons of the carbon surface and the free electrons of the dye molecule,
and found that the adsorption of the anionic dye was hindered by the presence of acidic
oxygen-containing functional groups on the surface of the activated carbon. It should be
noted that, in addition to both the increased hydrophilicity and the destruction of 7-bonds,
the chemisorption of oxygen may lead to the formation of acidic-oxygen containing
functional groups creating a weakening in the donor-acceptor (acid/base) adsorption
mechanism (Ania et al. 2004).
4.2.1 Air Treatments and CO2 Treatments
Air treatments were conducted in ambient air using a muffle furnace; a front-
loading box-type oven used for high temperature applications up to 12000C. A composite
of the three reactivated NORIT samples containing the highest calcium content was
created and three grams of this composite were placed in crucibles for 15, 30, and 60-
minute increments at the desired temperature inside the muffle furnace. The air treatment
temperatures used were 350'C, 400'C, and 450'C. These temperatures were chosen
based on research showing that the maximum chemisorption of oxygen and maximum
capacity to adsorb bases occurs around 400'C (Puri 1970). After treatment, the samples
were allowed to cool in a dessicator before water contact pH and Boehm titration
experiments were performed.
A clamshell furnace similar to that used by Mazyck and Cannon (2000 and 2002)
held a quartz-fluidized bed wherein CO2 treatments were carried out at ambient
temperatures. A porous plate near the center of the quartz-fluidized bed allowed a flow
of 0.5 L/min of CO2 to contact three grams of sample for 20 minutes for an exposure of
0.15 mole C02/g composite.
4.2.2 Boehm Titration
The concentrations of acidic oxygen complexes created on the surface of the
reactivated composite samples were determined according to the Boehm titration method
(Boehm 1966). An allotment of 0.5 g of composite and 25 mL of 0.1 N NaOH were
shaken in sealed vials for three days. The suspensions were then filtered using a 0.45 [tm
filter and 20 mL of the filtrate was titrated with 0.1 N H2SO4. The amount of base
(NaOH) consumed by each sample was determined by comparing each titration to that of
a blank. Any base consumed by the sample is expected to result from the neutralization
of carboxylic, lactonic, and phenolic groups, thereby quantifying the surface acidity of
the composite samples.
4.2.3 Equilibrium Isotherms
Equilibrium isotherms were used to compare the adsorption capabilities of the
composite samples. The target compound was MIB as the removal of this substance
represents the primary use of GAC; removal of taste and odor causing compounds. First,
a 0.5 L stock solution of 153 ng radiolabeled MIB/L was created by blending 95.7 pL of
an 800,000 ng MIB/L-methanol stock solution with 500 mL of either deionized water or
raw water influent to the Manatee County Water Treatment Plant. From the 153 ng
radiolabeled MIB/L solution, 4-50 mL allotments were added to separate gas tight
syringes followed by addition of enough of a powdered composite sample to create 5, 10,
20, or 40 ppm powdered reactivated carbon solutions. Grinding the treated composites in
a crucible and sieving the resulting particles through a 325 mesh (45 pm) sieve created
the powdered composites. The syringe was then sealed, permitting for a small amount of
headspace and allowed to mix on a rotisserie style mixer. After 24 hours, a 0.45 [im luer-
lock nylon filter, attached to the syringe, separated the powdered composite from 3 mL of
the solution which was then tested according to previous studies using radiolabeled MIB
(Gillogly et al. 1998, Tennant and Mazyck 2003, Nowack et al. 2004). Each sample was
measured twice for MIB and the mean of these two runs was used for the final value.
The measurements of the replicates were within 2% of one another. The entire process,
starting with 4-50 mL samples of either deionized or raw water influent to the Manatee
County Water Treatment Plant were repeated for each composite for which an
equilibrium isotherm was desired.
Adsorption is usually modeled by isotherms, which relate the relative
concentrations of solute adsorbed to the solid to the concentration in solution. The
adsorption isotherm is defined as the constant temperature relationship between the
quantities of adsorbate that is adsorbed per unit of adsorbent combined with the
equilibrium concentration of adsorbate in solution.
When a mass of adsorbent and a waste stream are in contact for a sufficiently long
time, equilibrium between the amount of pollutant adsorbed and the amount remaining in
solution will develop. For any system under equilibrium conditions, the amount of
material adsorbed onto the media can be calculated using the mass balance of
q, =(C,, C,) (6)
where qe is the mass of pollutant per mass of media [expressed as mg target contaminant
/g powdered composite], Cin is the initial target contaminant concentration in solution, Ce
is the concentration of the target contaminant in solution after equilibrium has been
reached, V is the volume of the solution to which the powdered composite mass is
exposed, and M is the mass of the powdered composite.
The most commonly used isotherm to model the adsorption of contaminants onto
activated carbon is the Freundlich isotherm (Snoeyink and Summers 1999). The general
form of the Freundlich model is
q, = KFCelin (7)
where KF and n are calculated from graphing the linear form of the equation shown
ln(q,) =ln(KF)+ lnC, (8)
4.3 Results and Discussion
4.3.1 Air and CO2 Treatments Versus Water Contact pH
A composite of the three reactivated GACs with the highest calcium content was
subjected to several treatments with either air or carbon dioxide in an attempt to create a
pH stable carbon, i.e., a composite with a water contact pH less than 8.5 pH units, as this
pH falls between the NPDES permitted discharge pH of 6 to 9 while still allowing for a
small safety factor. The resulting water contact pH of the treated composites are shown
in Figure 7. A dashed line at a water contact pH of 8.5 highlights those treatments that
created a pH stable carbon, those beneath this dashed line. The first bar shows the water
contact pH of the untreated composite. A treatment time of 15 minutes in the presence of
air did not create a pH stable carbon regardless of the temperature used, though these
water contact pHs are less than the untreated composite. Treatment times of both 30 and
60-minutes in the presence of air are shown to create pH stable carbons regardless of the
treatment temperature. It is expected that the longer treatment times would yield a
decreased water contact pH compared to the 15-minute treatment time, as more oxygen
would be chemisorbed to the surface to create more of the acidic oxygen-containing
functional groups that will overcome the addition of hydroxyl ions due to the proposed
calcium chemistry mechanism. The carbon dioxide treatment scenario, carried out at
ambient temperature with a carbon dioxide flow of 0.5L/min for 20 minutes also created
a pH stable carbon. The results of these treatments support the patents that claim to
overcome pH excursions via air treatments (Dussert et al. 1997) and carbon dioxide
treatments (Farmer et al. 1998).
E 350C, Air
10.0 E E400C, Air
S0. E 450C, Air
. .*-- Carbon Dioxide
0 15 20 30 60
Figure 7. Effect of air and CO2 treatments on water contact pH. Dashed line represents a
pH of 8.5.
4.3.2 Total Surface Acidity of Treated Carbons
The proposed mechanism behind air treatments yielding a pH stable carbon is that
oxygen chemisorbs to the surface of the GAC creating acidic oxygen-containing
functional groups. The total acidity of the treated composites is shown in Figure 8. As
expected, the air treated composites show an increase in total acidity compared to the
untreated composite. Air treatment of increased duration yields an increasing total
acidity, reflecting the finding that increased duration of air treatments yields pH stable
carbons. The highest total acidity for each time-increment is at an air treatment
temperature of 400'C. This result is expected from Puri's (1970) finding that the
maximum chemisorption of oxygen and creation of acidic carboxylic groups on the
surface of the GAC occur around 400C.
The carbon dioxide treatment yielded a total acidity close to the total acidity of
the untreated composite. The result supports the hypothesis that a pH stable carbon is
created with carbon dioxide treatment as a result of physisorbed carbon dioxide which
dissolves upon contact with water, rather than by the addition of acidity, as is the case
with air treatments.
o 400C, Air
E3 450C, Air +
3 Carbon Dioxide +
0 15 20 30 60
Figure 8. Effect of air and C02 treatments on total surface acidity. A pH stable carbon is
indicated by a (+).
4.3.3 Comparison of MIB Adsorption Capability
To determine the effect that the treatments designed to create pH stable carbons
have on adsorption, the air treated composite that yielded the highest total acidity, the
carbon dioxide treated composite, and an untreated composite were evaluated for their
removal of MIB in both deionized water and natural water (containing NOM). The
percent removal of MIB in deionized water versus the dose of composite is shown in
m Carbon Dioxide
20 400C, 1 hr, Air
0 --- i -- i --------- -.
0 1 2 3 4 5 6 7
Carbon Dose (mg/L)
Figure 9. Equilibrium adsorption of MIB in deionized water versus the dose of the
The air treated composite does not remove as much MIB as the untreated
composite because the air treated composite has a higher total surface acidity. An
increase in total surface acidity has been shown to have a negative effect on the
adsorption of MIB due to the preference of the created acidic functional groups for water
adsorption, as opposed to the adsorption of MIB (Pendleton et al. 1997, Considine et al.
2001). Given this explanation, one would expect that the carbon dioxide treated
composite and the untreated composite would have comparable MIB removal, as their
total surface acidity is similar (Figure 8). The results of Figure 9, however, show that the
carbon dioxide treated composite removes significantly more MIB than the untreated
A preliminary explanation for this occurrence was that some of the carbon dioxide
to which the activated carbon was exposed chemisorbed to the active sites on the carbon
surface, rendering them inactive. Active carbon sites would form acidic functional
groups upon water contact and, as has been discussed, the resulting sites will prefer
adsorption of water to MIB adsorption. By inactivating these sites, the carbon dioxide
treated composite would have less water adsorption than an untreated sample. To
determine if this proposed mechanism is plausible, approximately 0.14 g of sample were
suspended in an enclosed bottle containing 150 mL of deionized water. After 36 hrs, the
sample was weighed and the percent of water vapor adsorbed was determined. The
carbon dioxide treated composite was found to adsorb 6% less water vapor than the
untreated composite. However, the 6% difference in water vapor adsorption may not
adequately explain the approximate 15% difference in MIB removal between the carbon
dioxide treated composite and the untreated composite. Also, Figure 8 shows comparable
surface acidity between the carbon dioxide treated and untreated composite and the
carbon dioxide treated composite should have a lower surface acidity than an untreated
composite if the inactivity theory is correct and such is not the case. Though, it may be
that the type of acidic functional group is important, rather than the total acidity. The
mechanism behind the increase in adsorption of MIB with a carbon dioxide treated
composite remains unclear and warrants further research.
In typical water treatment, other constituents in the influent water may compete
with MIB for adsorption onto the carbon surface. NOM is ubiquitous in surface waters
and the concentration of NOM can be 100,000 times that of the trace contaminant MIB.
The presence of NOM has been shown to decrease the removal of MIB via competitive
adsorption (Newcombe et al. 2002, Tennant 2004). To determine how the composites
perform in natural water, i.e., that containing NOM, a solution was made with influent
water from the Manatee County Water Treatment Plant and subsequently dosed with
enough MIB to make 153 ng/L MIB. As with the deionized water experiments, four
separate doses of the composites were added to the Manatee solution and allowed to mix
for 24 hrs. The MIB removal results in Manatee water are found in Figure 10.
20 *400C, 1 hr, Air
0 10 20 30 40 50 60 70
Carbon Dose (mg/L)
Figure 10. Equilibrium adsorption of MIB in Manatee water versus the dose of the
Unlike the deionized water scenario, the air treated composite removed more MIB
in natural water than did the untreated composite. NOM can adsorb through electrostatic
forces, wherein the negative charge of the NOM is attracted to a positive charge on the
carbon surface (Newcombe et al. 1999). Tennant (2004) has shown that an increase in
adsorption of NOM onto carbon is linked to a decrease in the adsorption of MIB. The
point of zero charge (PZC) is the pH at which the carbon surface is neutral. Any water
pH below the PZC will produce a positively charged surface and a water pH above the
PZC will create a negatively charged surface. An acidic PZC is often indicative of a
wealth of surface acidic groups, while a basic PZC has a low number of surface
functional groups. The PZC can be related to the water contact pH in that both
experiments employ conditions where carbon dominates the system, 10% by weight for
the PZC measurements and 2.5% by weight for the water contact pH experiment. A
water pH below the water contact pH will create a positively charged surface that will
attract the negatively charged NOM. The untreated composite, with a water contact pH
of 10.6, in contact with a Manatee water pH of approximately 7, will have a significant
positive surface charge. The untreated composite will therefore adsorb more NOM and,
as there are fewer adsorption sites available after NOM adsorption, less MIB adsorption
would occur. The air treated composite, with a water contact pH of 7.75, in contact with
a Manatee solution pH of approximately 7, will have less of a positive surface charge and
will therefore not attract as much NOM, leaving more adsorption sites available for MIB.
As in deionized water, the carbon dioxide treated composite removes more MIB
than both the air treated composite and the untreated composite in the natural water
system. The mechanism may or may not be the same mechanism that causes carbon
dioxide treated composites to remove more MIB than other composites in deionized
water. Again, the mechanism behind increased adsorption of MIB with a carbon dioxide
treated composite is unclear and warrants further research.
The conversion of the removal data into a Freundlich isotherm supports the
finding that carbon dioxide treatment creates a carbon better suited for the removal of
MIB compared to air treated composites and untreated composites. A linear trendline
was added to each of the composite data points in order to determine the Freundlich
values. The equations of the linear trendlines, as well as the R2 values for the trendlines
and the calculated KF and 1/n values, Equation 8, for each of the composites are shown in
Table 2: Freundlich values for MIB removal in Manatee water
Composite Linear Equation R2 KF 1/n
Carbon Dioxide y = 0.47x + 0.14 0.78 1.56 0.47
Air Treated y = 0.45x + 0.07 0.77 1.08 0.45
Untreated y = 0.64x 0.69 0.95 0.93 0.64
The trend in KF follows the removal trend shown in Figure 10, in that the carbon
dioxide treated composite has a higher adsorption capacity for MIB than either the air
treated or untreated composite and the air treated composite has slightly less of an
adsorption capacity for MIB compared to carbon dioxide treated composites. While the
untreated composite shows a much higher 1/n value than either the air or carbon dioxide
treated, which is expected, the difference in the 1/n value for air and carbon dioxide
treated is small. The similar 1/n values with air and carbon dioxide treated composites
suggest a common adsorption mechanism that yields similar strengths of adsorption.
The results of this research support the patents of Dussert et al. (1995) and Farmer
et al. (1998), which state that air treatments and carbon dioxide treatments can create pH
stable carbons. A comparison of the total surface acidity of the treated carbon showed
that air treatments have a higher surface acidity than both carbon dioxide treated and
untreated carbons. These findings suggest that air treatments add acidic oxygen-
containing functional groups, which balances the basic solution pH resulting from the
calcium chemistry mechanism. Carbon dioxide treatments, on the other hand, were not
found to yield an increase in total surface acidity. Carbon dioxide is thought to physisorb
to the carbon surface and upon immersion in water it desorbs forming carbonic acid in
solution that balances the basic solution pH produced by the calcium chemistry
The performance of the treated composites was evaluated via removal of MIB.
The air treated carbon removed the least amount of MIB in deionized water as the acidic
oxygen-containing functional groups on the surface of the carbon create a preferential
adsorption of water over MIB. MIB in deionized water was adsorbed more by carbon
dioxide treated carbons than either the untreated or air treated carbons. The mechanism
behind the increase in MIB adsorption with carbon dioxide is not clear but it was not
found to result from a decrease in the number of active sites, as this would be linked with
a decrease in total surface acidity for the carbon dioxide treated carbon compared to the
untreated carbon, which was not supported by the data.
Evaluation of MIB removal in a natural water containing NOM, found that, unlike
MIB removal in deionized water, the air treated composite removed more MIB than the
untreated composite. The difference here is that NOM out-competes MIB for adsorption
sites. The surface of the untreated composite has a positive charge that attracts the
negatively charged NOM, thus leaving fewer sites for MIB adsorption. Air treatment
creates a more acidic surface, where the positive charge of the surface is not as strong,
thus not as much NOM is attracted to the surface, so that MIB may adsorb with minimal
competition from NOM. In both deionized water and natural water containing NOM,
carbon dioxide treatment performed better in removal of MIB than the untreated carbon
or the air treated carbon. The reason behind this is unclear and warrants further research.
The results do indicate, however, that treatments to create pH stable carbon affect the
adsorption capacity of the carbon and, as such, a treatment to create pH stable carbons
should be evaluated for its affect on adsorption capabilities before being used to create
pH stable carbons for use in a water treatment plant.
CONTRIBUTIONS TO SCIENCE
* Calcium chemistry (CaO + H20 Ca(OH)2; Ca(OH)2 Ca2+ + 20H-) is a
plausible mechanism behind pH excursion when reactivated carbon is
returned to service. Also, the concentration of other metals in the reactivated
carbon, such as aluminum, iron, and magnesium, were minimal compared to
the concentration of calcium in solution, thus highlighting the role of calcium
chemistry in pH excursion over the chemistry of other metals;
* The anion exchange mechanism was not found to explain pH excursions with
reactivated carbon as pH excursions occurred even when nanopure water was
used in the experiments. Nanopure water is not expected to contain the anions
required to participate in the anion exchange mechanism;
* Air and carbon dioxide treatments created pH stable carbons. Air treatments
at 350-450'C for 0.5-1 hr balanced pH excursions through the creation of
acidic oxygen-containing functional groups on the carbon surface. Carbon
dioxide treatment at ambient temperatures for 20 min. were proposed to allow
carbon dioxide to physisorb to the carbon surface and then desorb upon
contact with water, resulting in the creation of carbonic acid, which balances
the pH excursions;
* Carbon dioxide treated GACs removed more MIB, a trace contaminant, than
either air treated or an untreated composite. The relative removal of MIB for
air treated and untreated composites likely depended on the preference of the
carbon for water adsorption and also surface charge, which will have an effect
on the electrostatic attraction of the carbon for NOM, which competes with
MIB for adsorption onto the carbon surface. The mechanism behind the
increased removal of MIB with carbon dioxide treated composites is unclear
and warrants further research.
Ania CO, Parra JB, Pis JJ. Oxygen-induced decrease in the equilibrium adsorptive
capacities of activated carbons. Adsorption Science and Technology 2004; 22(4): 337-
Babushkin VI, Matveyev GM, Mchedlov-Petrossyan OP. Thermodynamics of Silicates.
New York, NY; Springer-Verlag, 1985: 265-276.
Bansal RC, Aggarwall D, Goyal M, Kaistha BC. Influence of carbon-oxygen surface
groups on the adsorption of phenol by activated carbons. Indian Journal of Chemical
Technology 2002; 9(4): 290-296.
Bansal RC, Donnet JB, Stoeckli F. Active Carbon. New York, NY; Marcel Dekker, Inc.,
Barton SS, Evans MJB, Halliop E, MacDonald JAF. Acidic and basic sites on the
surface of porous carbon. Carbon 1997; 35(9): 1361-1366.
Boehm H. Functional groups on the surfaces of solids. Angewandte Chemie Inernational
Edition 1966; 5(6): 533-622.
Boynton RS. Chemistry and Technology of Lime and Limestone. New York, NY; John
Wiley and Sons, Inc., 1980.
Cannon FS, Snoeyink VL, Lee RG, Dagois G, DeWolfe JR. Effect of calcium in field-
spent GACs on pore development during regeneration. J AWWA 1993; 85(3): 76-89.
Cannon FS, Snoeyink VL, Lee RG, Dagois G. Reaction mechanism of calcium-catalyzed
thermal regeneration of spent granular activated carbon. Carbon 1994; 32(7): 1285-1301.
Cannon FS, Snoeyink VL, Lee RG, Dagois G. The effect of iron and sulfur on thermal
regeneration of GAC. J AWWA 1997; 89(11): 111-122.
Carr S, Farmer RW. React-pHTM improves pH in carbon effluent. The National
Environmental Journal 1995; 5(2): 54.
Clark RM, Lykins BW. Granular Activated: Design, Operation and Cost. Chelsea, MI;
Lewis Publishers, 1989.
Considine R, Denoyel R, Pendleton P, Schumann R, Wong SH. The influence of surface
chemistry on activated carbon adsorption of 2-methylisobomeol from aqueous solution.
Colloids and Surfaces A: Physicochemical and Engineering Aspects 2001; 179(2-3): 271-
Darmstadt H, Christian R. Surface spectroscopic study of basic sites on carbon blacks.
Carbon 2003; 41(13): 2662-2665.
Dussert BW, Farmer RW, Hayden RA. Oxidized activated carbon for the control of pH
and alkalinity in water treatment applications. United States Patent 5,466,378 (Calgon,
El-Sayed, Bandosz TJ. Effect of increased basicity of activated carbon surface on valeric
acid adsorption from aqueous solution activated carbon. Phys. Chem. Chem. Phys 2003;
Farmer RW, Dussert BW, Kovacic SL. Improved granular activated carbon for the
stabilization of wastewater pH. American Chemical Society, Division of Fuel Chemistry
1996; 41(1): 456-460.
Farmer RW, Kovacic L, Matviya M, Wadhwa NP. Activated carbon treated by carbon
dioxide for the stabilization of treated water pH. United States Patent Number 5,714,433
Frederick HT, Cannon FS. Calcium-NOM loading onto GAC. J AWWA 2001; 93(12):
Frederick HT, Cannon FS, Dempsey BA. Calcium loading onto granular activated
carbon with salicylate or phthalate. Colloids and Surfaces A: Physicochemical and
Engineering Aspects 2001a; 177(2-3): 157-168.
Frederick HT, Cannon FS, Dempsey BA. Calcium and TOC loading: Effect of hydroxyl
and carboxyl substituents. Colloids and Surfaces A: Physicochemical and Engineering
Aspects 2001b; 191(1-2): 161-177.
Garten VA, Weiss VE. The ion and electron exchange properties of activated carbon in
relation to its behavior as a catalyst and adsorbent. Rev Pure Appl Chem 1957; 7: 69-
Gillogly T, Sneoyink V, Elarde J, Wilson C, Royale E. 14C-MIB adsorption on PAC in
natural water. JAWWA 1998; 90(1): 98-108.
Goins KM. The removal of tastes and odors in drinking water using granular activated
carbon. Thesis (M.S.). Pennsylvania State University, 2000.
Hammer MJ, Hammer MJ Jr. Water and Wastewater Technology. Upper Saddle River,
NJ; Prentice Hall, 2001.
Karanfil T, Kitis M. Role of granular activated carbon surface chemistry on the
adsorption of organic compounds. 2. Natural Organic Matter. Environ. Science and
Technology 1999; 33(18): 3225-3233.
Kawamura S. Integrated Design and Operation of Water Treatment Facilities. New
York, NY; John Wiley & Sons, 2000.
Kipling JJ and Gassser CG. Adsorption from liquid mixtures at solid surfaces. J.
Phys.Chem 1960; 64(6): 710.
Knappe DRU, Snoeyink VL, Dagois G, DeWolfe JR. Effect of calcium on thermal
regeneration of GAC. J AWWA 1992; 84(8): 73-80.
Kruyt HR and De Kadt GS. Die ladung der kohle. Kolloid Chem. Beihefte 1929; 47: 44.
Leon y Leon CA, Solar JM, Calemma V, Radovic LR. Evidence for the protonation of
basal plane sites on carbon. Carbon 1992; 30(5): 797-811.
Mazyck DW, Cannon FS. Overcoming calcium catalysis during the thermal reactivation
of granular activated carbon. Part I. Steam-curing plus ramped-temperature N2 treatment.
Carbon 2000; 13(38): 1785-1799.
Mazyck DW, Cannon FS. Overcoming calcium catalysis during the thermal reactivation
of granular activated carbon. Part II. Variation of Steam-Curing Reactivation
Parameters. Carbon 2002; 40(3): 241-252.
Menendez JA, Phillips J, Xia B, Radovic LR. On the modification and characterization
of chemical surface properties of activated carbon: In the search of carbons with stable
basic properties. Langmuir 1996; 12(18): 4404-4410.
Menendez JA, Suarez D, Fuente E, Montes-Moran MA. Contribution of pyrone-type
structures to carbon basicity: Theoretical evaluation of the pKa of model compounds.
Carbon 1999; 37(6): 1002-1006.
Montes-Moran MA, Suarez D, Menendez JA, Fuente E. On the nature of basic sites on
carbon surfaces: An overview. Carbon 2004; 42(7): 1219-1225.
Munoz-Guillena MJ, Linares-Solano A, Salinas-Martinez de Lecea C. A study of CaO-
SO2 interaction. Applied Surface Science 1994; 81(4): 409-415.
Newcombe G, Morrison J, Hepplewhite C, Knappe DRU. Simultaneous adsorption of
MIB and NOM onto activated carbon. II. Competitive effects. Carbon 2002; 40(12):
Newcombe G. Charge vs. porosity-some influences on the adsorption of natural organic
matter (NOM) by activated carbon. Water Science & Technology 1999; 40(9): 191-198.
Nowack KO, Cannon FS, Arora H. Ferric chloride plus GAC for removing TOC J
AWWA 1999; 91(2): 65-78.
Nowack KO, Cannon FS. Control of calcium buildup in GAC: Effect of iron
coagulation. Carbon 1997; 35(9): 1223-1237.
Otake Y, Jenkins RG. Characterization of oxygen-containing surface complexes created
on a microporous carbon by air and nitric-acid treatment. Carbon 1993; 31(1): 109-121.
Papirer E, Dentzer J, Li S, Donnet JB. Surface groups on nitric acid oxidized carbon
black samples determined by chemical and thermodesorption analyses. Carbon 1991;
Pendleton P, Wong SH, Schumann R, Levay G, Denoyel R, Rouquerol J. Properties of
activated carbon controlling 2-methylisobomeol adsorption. Carbon 1997; 35(8): 1141-
Pereira MFR, Soares SF, Orflao JJM, Figueiredo JL. Adsorption of dyes on activated
carbons: influence of surface chemical groups. Carbon 2003; 41(4): 811-821.
Puri BR. Surface Complexes on Carbon in Chemistry and Physics of Carbon, Vol. 6, ed.
P.L. Walker, Jr., New York, NY; Marcel Dekker, 1970.
Snoeyink VL, Summers RS. Adsorption of Organic Compounds in Water Quality and
Treatment, 5th Ed., Letterman, R., McGraw-Hill, 1999.
Szymanski GS, Karpifiski Z, Biniak S, Swiqtkowski A. The effect of the gradual thermal
decomposition of surface oxygen species on the chemical and catalytic properties of
oxidized activated carbon. Carbon 2002; 40(14): 2627-2639.
Tennant MF, Mazyck DW. Steam-pyrolysis activation of wood char for superior odorant
removal. Carbon 2003; 41(12): 2195-2202.
Tennant MF. Activation and use of powdered activated carbon for removing 2-
methylisoborneol in water utilities. Dissertation (Ph.D.). University of Florida, 2004.
Wiegand WB. pH properties of colloidal carbon. Ind. Eng. Chem 1937; 29: 953.
I was born on July 25, 1980, in St. Petersburg, Florida, where my mother, Teresa
Bach, also raised me. I graduated from Gibbs High School in June 1998. From there, I
traveled to Gainesville, Florida, to attend the University of Florida where I received my
Bachelor of Science degree in Environmental Engineering Sciences with highest honors.
I stayed on at the University of Florida to work on my Master of Engineering
degree with Dr. David W. Mazyck in the field of environmental engineering sciences.
Subsequent to the completion of my master's work, I will endeavor to attain a Doctor of
Philosophy degree in environmental engineering sciences, also under the guidance of Dr.
David W. Mazyck.