![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081107214028im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
2002 Progress Report: Tailoring Activated Carbon Surfaces for Water, Wastewater and Hazardous Waste Treatment Operations
EPA Grant Number: R828157Title: Tailoring Activated Carbon Surfaces for Water, Wastewater and Hazardous Waste Treatment Operations
Investigators: Karanfil, Tanju , Kilduff, James E.
Institution: Clemson University
Current Institution: Clemson University , Rensselaer Polytechnic Institute
EPA Project Officer: Krishnan, Bala S.
Project Period: June 1, 2000 through May 31, 2002 (Extended to July 31, 2004)
Project Period Covered by this Report: June 1, 2001 through May 31, 2002
Project Amount: $223,978
RFA: Exploratory Research - Environmental Engineering (1999)
Research Category: Engineering and Environmental Chemistry
Description:
Objective:The overall objective of this research project is to develop a fundamental understanding of how activated carbon surface chemistry influences adsorption of priority pollutants from complex solution matrices. Granular activated carbon (GAC) treatment is an excellent option for removal of a broad range of synthetic organic compounds (SOCs) from drinking water sources and industrial wastewaters. Specifically, our objectives are to: (1) elucidate how carbon surface chemistry influences adsorption mechanisms (e.g., dispersive, electrostatic, hydrophobic interactions) of SOCs and organic macromolecules; (2) investigate how sorbate molecular properties impacts adsorption by activated carbons with different surface chemistry; (3) examine the role of carbon surface chemistry on competitive (i.e., simultaneous adsorption and preloading) adsorption interactions between target pollutants and organic macromolecules in the background water matrices; and (4) provide a rational basis for selecting or preparing activated carbons for removal (and in some cases regeneration) of SOCs and organic macromolecules from water and wastewaters.
Our hypothesis is that carbon surface chemistry plays a key role in the adsorption of target compounds, especially for small molecular weight organic priority pollutants, and in some cases, may overwhelm the effects of pore structure. Preliminary results suggest that a good understanding of the interactions between carbon surfaces and priority pollutants is a viable route to producing novel sorbents for effective and economic application of adsorption processes in meeting our increasingly stringent water quality standards.
Progress Summary:Effects of Surface Chemistry on Competitive Interactions Between SOCs and Natural Organic Matter (NOM). In our previous progress report, we emphasized the efforts taken to prepare modified activated carbons and carbon fibers outlined in Phase I of the research plan. That work focused on heat treatment in an inert (N2) atmosphere to remove oxygen functionality from the carbon surface (as CO and CO2) and liquid phase oxidation (using HNO3). We investigated the effects of reaction time (2 to 16 hours) and temperature (400 to 1000°C) on the surface properties of an acid-activated, wood-based carbon using a new reactor system that is part of an Autosorb 1 Chemisorption Analyzer (Quantachrome Corp.). The uptake of trichloroethylene (TCE) by the surface treated carbons was a strong function of treatment temperature, increasing with increasing treatment temperature. However, no significant dependence on time, over the range investigated, was observed. We have extended that work to examine the role of surface chemistry on the effects of preloading by NOM. Preloading occurs when the mass transfer zone of NOM moves through a fixed bed at a faster rate than the target compound, and loads the carbon ahead of it. Such preloading by NOM can result in direct competition for adsorption sites, pore blockage, or both. This results in a reduction in equilibrium uptake. The specific objectives of this phase of the work were to: (1) investigate carbon surface chemistry modification as a means to preferentially reduce the uptake of NOM foulants and thus increase the uptake of TCE by GAC preloaded with these foulants; (2) investigate the effects of surface chemistry and pore structure using a series of carbons prepared from different starting materials with a wide range of surface acidity; and (3) evaluate the effects of NOM foulant physicochemical properties (e.g., molecular weight, acidity, and hydrophobicity) by using several different natural and model humic substances (humic and fulvic acids) and natural waters.
The ability to control NOM adsorption through modification of surface chemistry was demonstrated previously; the reactivity of carbon surfaces to NOM uptake depended on the raw material type, activation conditions, and surface treatment. Work described in our previous progress report has shown that surface acidity also plays an important role in the uptake of TCE. Because TCE is adsorbed by a hydrophobic adsorption mechanism, and because increasing surface acidity reduces the hydrophobicity of the surface, it was found that, in general, increasing surface acidity reduced uptake. For the F400 carbon, with initial low oxygen content, uptake decreased significantly upon exposure to nitric acid, which increased acidity. For the WVB carbon, having high acidity initially, solute TCE uptake increased significantly after heat treatment to reduce surface acidity and polarity. Heat treatment temperature had a significant effect on uptake; however, the effects of heat treatment were the same for reaction times ranging from 2 to 16 hours.
It was our goal to optimize the effects of surface treatment; to decrease the uptake of organic matter, thereby minimizing the effects of preloading, but without making the surface so hydrophilic as to reduce the uptake of TCE to an unacceptable degree. In assessing the success of various surface treatments, two comparisons are of interest. For a given surface treatment, a comparison of preloaded and non-preloaded carbons gives the effect of preloading. The effect of preloading can then be assessed for different surface treatments, and the role of surface treatment in mitigating the impact of preloading can be determined. However, only those treatments for which uptake by preloaded carbon is higher than preloaded as-received carbon, are of economic interest. Therefore, a comparison of preloaded as-received carbon with preloaded surface-treated carbon gives the overall effectiveness and economic viability of surface treatment for preloading control.
The impacts of surface acidity on the effects of preloading were found to depend on the surface treatment, carbon type (wood or coal-based), and the organic matter source (i.e., aquatic, soil, or synthetic). In several cases, particularly for the F400 carbon, the effects of preloading are mitigated with increasing surface acidity, a trend consistent with the effects of surface acidity on NOM loading. However, not all experiments showed such trends. Some data suggests that specific interactions may occur between preloaded humic materials and surface functional groups, especially those remaining after heat treatment of oxidized surfaces at 650°C. When specific interactions occur, the high affinity between the carbon surface and organic matter can cause large reductions in uptake resulting from preloading. In general, the effects of preloading were smallest for the most hydrophilic carbons, i.e., the as-received WVB carbon and the most oxidized F400 carbon. However, even though the effects of preloading were low, the uptake also was low. Therefore, even though the effects of preloading were mitigated, the use of highly oxidized carbons to control preloading effects is not economically attractive.
The wood-based carbons exhibited different trends than the coal-based carbons in one important way. A comparison of preloaded and non-preloaded carbons suggests that the impact of preloading on TCE uptake by heat treated carbon is less than the impact on uptake by as-received carbon (i.e., the percentage reduction in single solute uptake by heat treated carbon as a result of preloading is smaller than the reduction observed for as-received carbon). More importantly, a comparison of preloaded, as-received carbon with preloaded, surface-treated carbon reveals that uptake by preloaded, heat treated carbon is significantly higher than both that of single solute uptake by as-received carbon and preloaded, as-received carbon. The observed effects cannot be explained by a reduction in the uptake of NOM. Consistent with previous work, heat treatment of wood based carbons does not significantly affect the uptake of NOM (or other organic macromolecules). New data confirm the efficacy of heat-treated WVB carbon for mitigating preloading effects of both NOM from a surface water, and several different model organic macromolecules. Therefore, the results are valid for organic macromolecules having a wide range of physicochemical properties, including molecular weight, acidity, and polarity.
A comparison of the data reveals that the performance of the heat-treated wood-based carbon, even under some preloading conditions, is similar to single solute TCE uptake by coal-based activated carbons in the absence of preloading. The observed effect may result from some combination of optimum surface acidity, optimal type of surface functional group, and/or pore structure effects. The WVB carbon has a mesoporous pore structure, which has been observed to minimize the impacts of preloading in preliminary comparative experiments designed to isolate this effect. Future work will employ carbon surface characterization techniques that will allow identification of functional groups and more accurate correlation with surface reactivity.
Modeling Competitive Interactions Between SOCs and NOM. In our previous work, we reported on experiments designed to identify the reactive components in solutions containing NOM isolated from surface water sources. This was done to better understand the interactions between these compounds and carbon surfaces, and the competitive interactions between natural (macromolecular) organic matter and low molecular weight SOCs, as outlined in Phases III and IV of the research plan. The Ideal Adsorbed Solution Theory (IAST) was used to predict the uptake of mixtures of NOM fractions. It was found that predictions made using experimentally determined parameters for physically isolated fractions were as good or better than those using fictive components identified in a statistical search. Based on our previous findings, the competitive effects of NOM were modeled by the measured properties of a physically isolated low molecular weight reactive component. The average molecular weight of the proposed reactive fraction operationally defined as those components passing a low molecular weight cut-off ultrafiltration membrane.
Preloading mechanisms include direct site competition (in pores large enough
to admit NOM and the target organic) and pore blockage, which dominates competition
when only the lower molecular weight target compound can access pores. We have
developed a novel competitive adsorption predictive modeling approach, based
on modifications to the IAST, that incorporates these mechanisms. It was hypothesized
that the effects of preloading may be more accurately modeled by a modified
IAST procedure, in which it was assumed that TCE could access sites that the
humic molecules cannot. This hypothesis is represented mathematically by the
parameter, 
,
the fraction of the surface area on which there are no competitive effects.
The total solid phase TCE uptake, qe, and capacity,
Qo, are the sum of uptake and capacity
by competitive and non-competitive surfaces:
Defining a fraction of the surface area on which TCE adsorbs without competition
is equivalent to defining a fraction of the capacity, Qo,
that can be occupied with no competition. Therefore,
Using the 3-parameter Langmuir Freundlich equation to describe TCE sorption, the uptake of TCE by surfaces without competition, q2,non-comp is:
Only the surface area available to both TCE and humic substances is included in the IAST calculations, therefore, the TCE single solute isotherm used in these calculations is:
Similarly, the mass of TCE adsorbed by surfaces without competition is not included in the initial concentration of TCE used in the IAST calculation:
where D is the adsorbent dose, and other terms have been defined previously. All model calculations were performed using an optimization scheme based on a generalized reduced gradient (GRG) non-linear programming algorithm, implemented in a commercial software package (Microsoft Excel).
The fraction of sites on which TCE adsorbs without competition, ,
was calibrated using isotherm data obtained from TCE adsorption on preloaded
carbon. Several representative points spanning a wide range of TCE concentrations,
were used. For each point, the model-calculated TCE qtotal
was compared to the observed isotherm at the same value of Ce.
Optimization was achieved by varying
to minimize the sum of the squared residuals (SSR) for the chosen points. Accounting
for non-competitive surface area increases the accuracy of the data description
significantly. While this approach introduces only one fitting parameter, it
was found that calibrated values of
decreased with increased loading over the range of loading studied. Under the
hypothesis that
represents the fraction of the total surface area that TCE can access but humic
molecules cannot, a decrease in
with humic loading may imply that humic molecules access smaller pores as their
loading increases. This could be explained by a shift in the adsorbed molecular
size distribution of NOM to smaller sizes as loading increases, as observed
in previous work. An alternative explanation is that with increased loading,
surface and pore concentrations become high enough to cause adsorbed humic molecules
to become more compact and better able to penetrate more deeply into the carbon
pore structure.
Using single solute isotherm parameters and -values
calibrated as a function of the loading of the low MW fraction, predictions
for TCE adsorption by carbon preloaded with whole humic acid solutions, were
calculated. To calculate a value for the non-competitive fraction of the surface,
the correlation between
and the low MW fraction loading, was used. TCE isotherms on carbon preloaded
with whole NOM under 10 different loading conditions, were predicted. The model
predictions were reasonably accurate, verifying the modeling approach, and supporting
the non-competitive surface area hypothesis. However, better accuracy was sought
at high TCE concentrations, where the predicted isotherm becomes asymptotic
to the single-solute isotherm and does not reproduce the competitive effects
of preloading. In contrast, measured TCE isotherms are displaced downward, relative
to the qe axis, in a parallel fashion (on log-log
coordinates). This suggests a reduction in TCE capacity, and a corresponding
reduction in the Langmuir-Freundlich value of Qo.
This type of behavior is consistent with a reduction in the total number of
available adsorption sites. These sites may be lost as a result of direct competition,
or as a result of a pore blockage mechanism. As originally formulated, the IAST
model incorporates the effects of direct competition for adsorption sites, but
does not incorporate the effects of pore blockage. Therefore, the IAST model
was modified to account for pore blockage effects.
It was hypothesized that pore blockage can reduce the effective surface area available to TCE, and therefore the single-solute capacity parameter, Qo. To estimate the magnitude of the pore blockage effect, biological, economical, and technical (BET) surface area was measured before and after preloading. Significant reductions in surface area were found, consistent with other studies reported in the literature. The measured surface area reduction caused by NOM preloading was correlated as a function of the whole humic loading. It was assumed that the reduction in the effective surface area available to TCE was in proportion to the reduction in surface area as measured by BET analysis. The reduction in surface area was incorporated into the IAST model by constraining the TCE capacity to reflect the reduced surface area available, which is equivalent to reducing the TCE Langmuir-Freundlich capacity parameter, Qo, in proportion to the measured surface area reduction. Incorporating pore blockage effects in this way does not introduce an additional fitting parameter into the model, retaining its predictive capabilities. Model predictions provided excellent predictions over the entire TCE concentration range studied, and was verified for several different NOM loadings. Calculations were done to show that at most partitioning of TCE to preloaded humic material represented less than 0.02 percent of the total uptake for the largest high humic loadings and high TCE concentrations. Verification of, and possible modification to, this model for other natural water source NOM, and investigation of adsorption phenomena correlated to NOM physicochemical characteristics, constitute further aspects of this research.
Future Activities:Future activities will involve further modification and characterization of activated carbons and carbon fibers. These adsorbents will be used to examine the role of carbon surface chemistry on the adsorption mechanisms of aromatic and aliphatic compounds, as outlined in Phase II of the research plan. Furthermore, the effects of surface chemistry on adsorption and recovery of phenol, will be probed. Based on the data supporting the reactivity of graphite surfaces reported in the first year progress report, further experiments are planned to probe the effects of carbon pore structure on the adsorption and recovery of phenolic compounds. The uptake of whole NOM and NOM fractions by modified carbons, and the subsequent effects on target compound adsorption, will be further investigated as outlined in Phases II and III of the research plan. Finally, the IAST modeling framework for describing/predicting competitive adsorption between NOM and SOC molecules, will be further developed.
Journal Articles on this Report: 2 Displayed | Download in RIS Format
| Other project views: | All 25 publications | 8 publications in selected types | All 8 journal articles |
| Type | Citation | ||
|---|---|---|---|
|
|
Kilduff JE, Karanfil T. Trichloroethylene adsorption by activated carbon preloaded with humic substances: effects of solution chemistry. Water Research, Volume 36, Issue 7, April 2002, Pages 1685-1698. |
R828157 (2001) R828157 (2002) R828157 (Final) |
not available |
|
|
Kilduff JE, Srivastava R, Karanfil T. Preloading of GAC by natural organic matter: effect of surface chemistry on TCE uptake. Characterization of Phosphorus Solids VI Studies in Surface Science and Catalysis 2002;144:553-560. |
R828157 (2002) R828157 (Final) |
not available |
environmental chemistry, analytical measurements, adsorption, toxics, solvents. , Ecosystem Protection/Environmental Exposure & Risk, Water, Scientific Discipline, Waste, RFA, Drinking Water, Chemistry, Hazardous Waste, Fate & Transport, Environmental Chemistry, Hazardous, Engineering, industrial waste, treatment, water quality, wastewater treatment, other - risk management, fate and transport, monitoring, granular activated carbon
Progress and Final Reports:
2001 Progress Report
Original Abstract
Final Report