CICEET Progress Report for the period 3/16/06 Through 9/15/06

Project Title: Predicting and Validating the Field Performance of Novel Sorbent-amended Sediment Caps
Principal Investigator(s): Dr. Gregory V. Lowry
Additional Investigator(s): Dr. David A. Dzombak, Dr. Jeanne VanBriesen
Project Start Date: September 1, 2005

Figures

Figure 1

Figure 2

Tables

Table 1

Table 2

Table 3

Tasks to meet objectives
Sorption isotherms to obtain Freundlich isotherm parameters (K_f and n) for:

• System 1: contains PCBs, polyoxymethylene (POM), and de-ionized (DI) water (provides K_POM, the sorption coefficient on POM by extracting water and POM)
• System 2: contains PCBs, POM, DI water, and virgin activated carbon (knowing K_POM provides K_{f_AC}, the sorption coefficient for each PCB on virgin activated carbon by extracting the POM only)
• System 3: contains PCBs, POM, DI water, and biofilm-covered activated carbon (knowing K_POM provides K_{f_Bio}, the sorption coefficient for each PCB on biofilm-covered activated carbon by extracting POM only). The K_f values obtained are compared to those for fresh AC and published literature values (Jonker and Koelmans, 2001).
• System 4: contains PCBs, POM, DI water, and DOM-loaded activated carbon (knowing K_POM provides K_{F_DOM}, the sorption coefficient for each PCB on DOM-covered activated carbon by extracting POM only)

• Nitrogen gas adsorption isotherms at 77 K to obtain total and microporous surface area and total and microporous pore volume for fresh AC and DOM loaded AC.

Progress on Tasks
A biofilm was generated from Staphlococcus epidermidis pure culture and grown on activated carbon for use in the isotherm studies (System 3).

Dissolved organic carbon (DOC) from sediment pore water was generated using ASTM Method D3987-85. AC was mixed with this pore water for 3 days prior to preload with DOC for use in isotherm studies (System 4). The DOM loading, as measured by difference in aqueous concentration, was ~1.2 wt% (dry weight) of AC.

Sorption isotherms Freundlich coefficients (K_F and n) were obtained for POM, virgin AC, biofilm-covered AC, and DOM-loaded AC (these data are included in preliminary data). 11 PCB congeners where chosen for the sorption isotherms ranging from di- to penta-chlorinated congeners, and a method was developed to separate and analyze these congeners by GC-_ECD. These congeners were chosen to cover a wide range of octanol-water partition coefficients, planarity, toxicity, and availability of existing partitioning data to compare our results with results from other studies, and on their occurrence in field samples. The selected PCB congeners are:
2,2’-dichlorobiphenyl (IUPAC #4)
3,4-dichlorobiphenyl (IUPAC #12)
2,2,5’-trichlorobiphenyl (IUPAC #18)
3,3’,4-trichlorobiphenyl (IUPAC #35)
2,2’6,6’-tetrachlorobiphenyl (IUPAC #54)
2,2,5,6’-tetrachlorobiphenyl (IUPAC #53)
2,2,5,5’-tetrachlorobiphenyl (IUPAC #52)
2,3,5,5’-tetrachlorobiphenyl (IUPAC #72)
3,3’,4,4’-tetrachlorobiphenyl (IUPAC #77)
2,2’,3,4,4’-pentachlorobiphenyl (IUPAC #85)
3,3’,4,4’,5-pentachlorobiphenyl (IUPAC #126)

Physical characterizations of virgin AC, biofilm-covered AC, and DOM-loaded AC.
BET surface area, microporous surface area, total pore volume and micropore volume were determined by analyses of nitrogen gas adsorption isotherms at 77 K (these data are included in preliminary data).

Difficulties
Sorption isotherms to obtain Freundlich isotherm parameters (K_f and n).
Only data for which the aqueous concentration was more one order of magnitude less than the water solubility of each PCB congener was used to determine K_f and n ­ all other data was excluded ­ using this rationale, a minimum of 3 sets of triplicates were used to fit the data to regress K_f and n.

Physical characterizations of virgin AC, biofilm-covered AC, and DOM-loaded AC.
The surface area and pore volume of the ACs were determined from analyses of nitrogen gas adsorption isotherms at 77 K performed with a NOVA 2200e Surface Area and Pore Size Analyzer (Quantachrome Instruments, Boynton Beach, FL). In this analysis procedure, AC samples have to be degassed which removes entrapped water from the pore structure. Given the potential to alter the biofilm (which is mostly water) during degassing, no attempt was made to characterize the surface area and pore volume of the biofilm-covered AC. Additionally, since nitrogen gas was used to determine the physical characteristics of the ACs, the surface area and pore volume determined for the DOM-loaded AC may be an overestimate of that available to the much larger PCB molecules. Work is currently underway to determine an appropriate probe molecule to determine surface area accessible to PCBs in DOM-loaded ACs.

Project Objectives for Next Reporting Period

Objectives
Evaluate the effect of groundwater seepage on the assumption of equilibrium sorption in an activated carbon cap

• Design and begin a set of fixed-bed reactor (FBR) experiments to evaluate the relative importance of reaction kinetics and equilibrium sorption in AC caps.
• Given the strong sorption of PCBs to AC, breakthrough is not likely to be reached in practical time scales in the lab using AC particle sizes and flow conditions expected in reactive-core mats. Such problems in drinking water applications have been addressed through use of rapid small-scale column tests (RSSCT). By using smaller AC particles sizes than used in the field, RSSCT is a technique that can be used to design a small-scale physical model that gives identical breakthrough curves to full-scale operation when effluent profiles are plotted as bed volumes fed. This is accomplished by scaling the lab-scale experiments such that the resistances to mass transfer are the same as those in the field.
• Using RSSCT, it will be possible to simulate several years of cap performance in the field (under steady flow conditions) in several weeks or months in the lab.

Work plan to Meet Objectives
Evaluate the effect of groundwater seepage on the assumption of equilibrium sorption in an activated carbon cap

• Design RSSCT columns using previously developed and experimentally-verified scaling equations (Crittenden et al., 1986; Crittenden et al., 1987). This approach is based on equating chemical mass transfer in the field and the lab. The time to breakthrough of the lab-scale columns decreases with decreasing activated carbon particle size ­ approach velocities are adjusted to maintain identical mass transfer conditions as would be in the field with larger activated carbon sizes and realistic approach velocities.
• Use of these scaling equations indicate there are challenges associated with this scaling procedure to run lab-scale studies that simulate capping situations, specifically:
  • Required column length may be impractically small (~1cm), or
  • Unacceptably large water volumes (400L) and operating times (>1yr) may be required
• As such, we are evaluating conditions that could be relaxed within the scaling procedure that would produce a practical RSSCT that is still representative of likely field conditions.
• Setup and run RSSCT ­ collect breakthrough curves using (i) virgin AC and PCB-spiked sediment pore water. Using sediment porewater simulates realistic treatment conditions and allows us to examine the effect of colloidal transport of PCBs.

Anticipated Success in Meeting Project Objectives
We have met the following proposed project objective:
Obtained PCB sorption isotherms (Freundlich K_F and n) for 11 PCB congeners on virgin AC, DOM-loaded AC, and biofilm-covered AC. The sorption strength on virgin AC, quantified by the Freundlich parameter log K_f, ranged from 7.27 to 9.29 (ng/kg)(L/ng)ⁿ. The presence of DOM and biofilm on the AC decreased log K_F by 0.5 to 1.6 log units, depending on the congener. Despite this decrease in log K_F, AC sorbs PCBs so strongly that it is still a viable remediation strategy for PCB-contaminated sediment.

At this time, we anticipate that we will be able to meet the remaining proposed project objective, which is to:
Determine the effect of groundwater seepage on the assumption of equilibrium sorption in the cap through use of RSSCT.

Overall Project Timeline Update
The overall project timeline has not changed from the original proposal (Table 1) for this two year project. Note that we will not run columns with DI water alone given the substantial effort required to perform these column tests correctly. We will run columns with sediment porewater to simulate filed conditions as closely as possible in a lab setting. By measuring particulate bound PCBs and dissolved PCBs we can determine the effect of colloidal transport of PCBs in these studies.

Preliminary Data
Evaluating the effect of DOM and biofilm growth on sorption of PCBs to Activated Carbon
PCB Sorption Isotherms
Jonker and Koelmans (2001) determined log K_{F_POM} values for IUPAC # 18, 52, 77, and 126 to be 3.90, 4.44, 5.01, and 5.20 (ng/kg)(L/ng)ⁿ, respectively. These values fall within the 95% CI of the log K_{F_POM} values obtained in this study (Table 2) and indicate that our experimental techniques and analysis were robust. In addition, four point isotherms for sorption of the 11 congeners to each of three AC treatments (virgin, DOM-loaded, and biofilm covered) were obtained reported in Table 2. To our knowledge this is the first comprehensive data set collected for PCB adsorption to activated carbon that uses a full range of C_w^eq. It is also the first known data set for PCB sorption onto DOM-loaded and biofilm fouled activated carbon. Figure 1 shows a sample of the isotherm data used to determine the coefficients listed in Table 2 for IUPAC #18, 52, 77, and 126. For virgin AC, log K_{F_AC} values for the 11 congeners ranged from 7.27 to 9.29 (ng/kg)(L/ng)ⁿ, indicating that AC sorbs PCBs strongly. As expected, the presence of DOM and biofilm on the AC decreased sorption of PCBs to AC. Depending on the congener, log K_F decreased 0.3 to 1.6 log units. The similarity of log K_{F_DOM} and log K_{F_Bio} was somewhat surprising since no attempt was made to load the AC with similar amounts of DOM and biofilm.

Physical characterizations of the ACs were performed to gain insight as to the cause of the decrease in log K_F upon loading the AC with DOM or biofilm. The surface area and pore volume of the virgin and DOM-loaded AC is shown in Table 3. Given the potential to alter the biofilm during degassing, coupled with the similarity of log K_{F_DOM} and log K_{F_Bio}, no attempt was made to characterize the biofilm-covered AC.

As expected, the surface area and pore volume decreased as a result of DOM-loading. The BET surface area, a measure of the total surface area, decreased approximately 100 m²/g ­ this drop was primarily due to a decrease in micropore area (pores less than 20 Å in width) from 798 to 708 m²/g. Similarly, there was a concomitant decrease in pore volume ­ the total pore decreased from 0.523 to 0.493 cm³/g, primarily due to a decrease in micropore volume from 0.327 and 0.290 cm³/g. As such, despite the presence of a substantial fraction of larger meso- and macro-pores (~37% of a pore volume basis), the DOM-loading primarily occurred in, or blocked access to micropore regions of the AC.

To determine whether the drop in log K_F was due solely to the decrease in site accessibility, the Freundlich parameter, log K_F, was normalized on the basis of BET surface area (SA) for the selected congeners with virgin and DOM-loaded AC (Figure 2).

The reduction in BET surface area upon loading with DOM could not fully account for the difference in K_f for a given congener, i.e. K_f decreased by an order of magnitude while the available surface area decreased by 11.2%. However, this result does not necessarily eliminate the possibility that a decrease in available sorption sites caused by DOM was a primary mechanism for AC fouling with respect to PCB adsorption. TOG is a microporous AC, with ~80% of its surface area contained in micropores (pores <20 Å in width). While only 11.2% of the microporous surface area was inaccessible to nitrogen gas at 77 K following DOM loading (Table 3), a much larger fraction could be inaccessible to the selected PCB congeners since they are larger molecules than nitrogen. The use of probe molecules of similar size as the adsorbate (i.e., the PCBs) would be more appropriate than nitrogen in this instance. Work is currently underway to determine an appropriate probe molecule to determine surface area accessible to PCBs in DOM-loaded ACs. Another possibility is that the PCBs preferentially sorbed to certain parts of the AC surface. It has recently been demonstrated that PCBs adsorb preferentially to specific nm-scale reaction sites on AC (Yoon et al., 2006). As such, it is possible that the DOM is blocking a fraction of these active sites and therefore decreasing sorption of each of the PCB congeners. More experimental work is needed to obtain a mechanistic understanding of the relative importance of DOM pore blockage and DOM competition for specific reaction sites on AC.

In a previous study it was found that coplanar PCBs exhibited stronger sorption to soot and soot-like materials than nonplanar PCBs (Jonker and Koelmans 2002). In this study, we evaluated the planarity effect by comparing log _{KF_AC} values for the tetra-chlorinated biphenyls with varying degrees of ortho-substititution ­ from 4 ortho-chlorines to no ortho-chlorines (Table 2). These compounds have similar KOW’s and the same number of chlorines on the biphenyl ring, so only the placement of the chlorine molecules on the biphenyl ring differs. In order of increasing planarity of the 5 tetra-chlorinated congeners studied, the log K_{F_AC} values were 8.06 (IUPAC # 54-4 ortho chlorines), 7.27 (IUPAC # 53-3 orthochlorines), 7.71 (IUPAC # 52-2 orthochlorines), 8.97 (IUPAC # 72-1 orthochlorine), and 9.07 (IUPAC # 77-0 orthochlorine) (ng/kg)(L/ng)ⁿ. The results indicate that the effect on sorption due to planarity does follow the number of ortho-chlorines on the biphenyl ring with the exception of IUPAC # 54, which has 4 ortho-chlorines.

Previous work in by our research group analyzed the length of time a 1.25 cm thick cap of AC, coke, soil, sand, or sediment could sequester 2,4,5-trichlorobiphenyl using an advective-dispersive transport model (Murphy et al., 2006). A log K_{F_AC} of 6.2 L/kg was used for 2,4,5-trichlorobiphenyl based on published data for Calgon F-100 AC (Kleineidam et al., 2002). Modeling results indicated that even for aquatic systems with high groundwater seepage rates (10 cm/day), AC amended caps could contain 2,4,5-trichlorobiphenyl for 200 years. The current research found the log K_{F_AC} for TOG AC loading with DOM or a biofilm ranged from 6.6 and 8.5 L/kg (Table 2). Based on this modeling effort, these results indicate that TOG AC amended caps could still provide isolation times in excess of 200 years even with biofilm or DOM limiting the AC capacity. However, the RSSCT studies are needed to determine if reaction kinetics are important in capping situations.

Werner et al. (2006) recently modeled PCB mass transfer after amendment of the contaminated sediment with activated carbon. The model was fit to lab data and found that log K_{f_AC} values from Jonker and Koelmans (2001) work needed to be 16 times (1.2 log units) lower in order for the model to fit the experimental data. The log K_{f_AC} values in Jonker and Koelmans work did not take into account the effect of DOM on PCB sorption to activated carbon. The results of this work indicate that the effect of DOM loading on PCB sorption can range from 0.6 to 1.6 log units. Thus, the decrease in sorptive capacity of the activated carbon mixed into sediment in the Werner et al’s (2006) work, is likely due to sediment organic matter sorbing onto the activated carbon and hindering PCB sorption.

Dissemination
Dr. Jeanne VanBriesen, Dr. David Dzombak, Jim Quadrini (QEA), Jim Olsta (CETCO), Dr. John Smith (Alcoa), Ed Neuhauser (Niagara Mohawk), Dr. Tommy Myers (Army Corps of Engineers), Dan Miller (HRNERR), Chuck Nieder (HRNERR), Dennis Timberlake (USEPA).

Publications:
Murphy, P., Marquette, A., Reible, D., Lowry, G. (2006). “Predicting the Performance of Activated Carbon-, Coke-, and Soil-Amended Thin Layer Sediment Caps”, Journal of Environmental Engineering, 132(7), 787-794.

McDonough, K.M., Murphy, P.M., Olsta, J., Zhu, Y., Reible, D.D., and Lowry, G.V. “Development and Placement of a Sorbent-amended Thin Layer Sediment Cap in the Anacostia River”, Journal of Soil and Sediment Contamination (in press).

McDonough, K.M., Fairey, J.L. and Lowry, G.V. “Polychlorinated biphenyl sorption to activated carbon and the impact of fouling by dissolved organic matter and biofilm growth”, in preparation for submission to Environmental Science and Technology, September 2006.

Conferences:
McDonough, K.M., Fairey, J.L., and Lowry, G.V. “Effect of DOC Sorption and Biofilm Growth on the Performance of Activated Carbon-amended Sediment Caps for PCB-Containment”, The Association of Environmental Health and Sciences (AEHS) 17th Annual AEHS Meeting and West Coast Conference on Soils, Sediments, and Water, San Diego, California, March 19-22, 2007.

Lowry, G. V., Murphy, P., Marquette, A., Reible, D. Sorbent-Amended “Active” Caps for PCB-contaminated Sediments: Placement and Performance. The Fifth International Conference on Remediation of Chlorinated and Recalcitrant Compounds, Monterey, CA. May 22-25, 2006.

McDonough, K.M., Lowry, G.V., and Reible, D.D. “Active Capping of Contaminated Sediments,” presented at the American Institute of Chemical Engineers Conference, Cincinnati, OH, November, 2005.

Manuals, Protocols:
None

Outreach Activities:
Conferencing/advising Denise Karabowicz’s science and technology club (group of 6-10th grade kids) to prepare a report on in situ sediment capping for PCB-contaminated sediments. (dkarabowicz@legodogs.com). Their research project on PCBs was chosen to be the one presentation given during the closing ceremonies at the State (Illinois) competition held January 14th, 2006.

Contact with End Users:
I have recently spoken with Jim Olsta (CETCO) and Dave Nakles (Retec) about the project and the potential for an in situ field demonstration using an AC-amended sediment cap. A copy of this report has been submitted to Chuck Nieder (HRNERR), Jim Quadrini (QEA), Jim Olsta (CETCO), Dave Nakles (Retec), Tommy Meyers (USACE), John Smith (Alcoa Inc.), and Dennis Timberlake (US EPA). I am awaiting feedback from end users. Feedback from end users will be included in an updated report to follow shortly.

Patent, Copyright, Invention Disclosure Activity:
None

Expenditures
Expenditures are in line with the proposed budget.

References
Crittenden JC, Berrigan JK, and Hand DW. Design of rapid small-scale adsorption tests for a constant diffusivity. Journal WPCF 1986, 58, 312-319.

Crittenden JC, Berrigan JK, Hand DW, and Lykins B. Design of rapid fixed-bed adsorption tests for nonconstant diffusivities. Journal of Environmental Engineering 1987, 113, 243-259.

Jonker, M.T.O.; Koelmans, A.A. Polyoxymethylene Solid Phase Extraction as a Partitioning Method for Hydrophobic Organic Chemicals in Sediment and Soot. Environ. Sci. Technol. 2001, 35, 3742-3748.

Jonker, M.T.O.; Koelmans, A.A. Sorption of Polycyclic Aromatic Hydrocarbons and Polychlorinated Biphenyls to Soot and Soot-like Materials in the Aqueous Environment: Mechanistic Considerations. Environ. Sci. Technol. 2002, 36, 3725-3734.

Kleineidam, S.; Schuth, C.; Grathwohl, P. Solubility-normalized combined adsorption-partitioning sorption isotherms for organic pollutants. Environ. Sci. Technol. 2002, 36, 4689-4697.

Murphy, P.; Marquette, A.; Reible, D.D.; Lowry, G.V. Predicting the Performance of Activated Carbon-, Coke-, and Soil-Amended Thin Layer Sediment Caps. J. Environ. Eng. 2006, 132, 787-794.

Werner, D.; Ghosh, U.; Luthy, R.G. Modeling Polychlorinated Biphenyl Mass Transfer after Amendment of Contaminated Sediment with Activated Carbon. Environ. Sci. Technol. 2006, 40, 4211-4218.

Yoon, T.H.; Benzerara, K.; Ahn, S.; Luthy, R.G.; Tyliszczak, T.; Brown, G.E. Nanometer-Scale Chemical Heterogeneities of Black Carbon Materials and Their Impacts on PCB Sorption Properties: Soft X-ray Spectromicroscopy Study. Environ. Sci. Technol. (ASAP) 2006.

PI Response to End User Advisor Feedback