Project Title: In-situ Treatment of PCBs in Marine and Freshwater Sediments using Colloidal Zero-Valent Iron
Principal Investigator(s): Kevin H.Gardner

Accomplishments
Scheduled Tasks:
There is an urgent national need for sediment treatment technologies that can be applied in situ, particularly for widespread, recalcitrant organic compounds such as polychlorinated biphenyls (PCBs). Chemical transformation of halogenated organic compounds (HOCs) by nanoscale, colloidal zero-valent iron (ZVI) is one of the latest innovative technologies in environmental remediation. Currently, remediation of PCBs is being performed primarily by dredging, which includes offsite storage and treatment. This research project is developing and evaluating a treatment technology that will build on recent work demonstrating the reductive dechlorination of PCBs to biphenyl in the presence of elemental iron or catalyst-impregnated ZVI. The objective is to develop a robust technology to remediate PCBs in marine and fresh water sediments under ambient conditions in a cost-effective and efficient manner. There is an urgent national need for in-situ sediment remediation methods to be developed that can cost-effectively treat large volumes of contaminated sediment in a timely manner.

Specific objectives include:

1. organizing a set of batch experiments to determine the minimum mass of ZVI required for treatment, the reaction kinetics of different types of iron, and the possibility of adding more than one application of iron;
2. explicate a mass balance by studying iron complexation products or by monitoring the increase of biphenyl and lesser chlorinated congeners in a dechlorinated system;
3. measuring the properties of three different types of ZVI such as BET surface area analysis of the iron particles, and the redox potential, pH, and possible anion interferences of the solution;
4. developing a long term incubation study to investigate the feasibility of this remediation method under conditions that are closer to those in-situ; and
5. establishing the conditions necessary to implement this technology and evaluating an assessment of the economics of the in-situ remediation method.

The tasks set by the research team for the second review period were as follows:

1. elucidate the validity of analytical methods such as quenching, determine rates of reaction, and evaluate different types of zero valent iron through batch experiments;
2. program a macro file to evaluate concentrations of PCBs from the Chem Station chromatograph printout;
3. complete a scientific literature review pertaining to PCBs, degradation, zero-valent iron, and analytical procedures;
4. present up-to-date research at the Batelle Conference in Monterey, California on the Remediation of Recalcitrant Chlorinated Compounds.

Progress on Tasks
The experimental design was delineated in the project proposal presented by Dr. Gardner in March, 2001. The graduate student working on this project, Deana Aulisio, has completed a course in environmental sampling and analysis, which includes quality control methods for laboratory procedures, and was currently refreshed in the 40-hour Occupational Health and Safety (OSHA) training. Biweekly laboratory meetings for the Environmental Research Group are also attended. To ensure the validity of the experimentation, a contract has been established with Northeast Analytical Laboratory in Schenectady, NY to test duplicate samples using state of the art technologies for the measurement of PCBs in sediments. Biphenyl analysis will be performed by Resource Laboratories Inc. in Portsmouth, NH in September, 2002.

A four credit independent study was completed during the Spring 2002 semester with Dr. Gardner and Deana Aulisio to gather all literature pertaining to the physical and chemical properties of polychlorinated biphenyls, the fate and transport in the environment, analytical techniques, and degradation methods. A written summary for each topic is organized in a set of four binders encompassing all journal articles, conference proceedings, and book references relevant to the project.

The sediments were obtained from the New Bedford Harbor Superfund site and the Housatonic River in Pittsfield, MA on November 30, 2001. They have been stored in the dark at 4° C in glass containers with Teflon-lined caps over the past year. Extensive laboratory batch studies are underway, in which varying conditions of iron type, amount, number of applications, and reaction rate are being investigated. PCBs are extracted with ultrasonication and are measured with a GC/ECD. Extractions are performed using the EPA's SW-846 Method 3550B Ultrasonic Extraction. Samples are concentrated to 1 mL in hexane using a Turbovap evaporator with nitrogen gas. A surrogate of congener 207 is added before the extraction to quantify percent recoveries. A sulfuric acid wash is used to remove hydrocarbons and other organic compounds according to SW-846 Method 3665A. GC analysis is performed on an HP 5890 Series II with an electron capture detector and autosampler. The method follows information reported by the SW-846 Method 8082 and the North East Analytical Standard Operating Procedures for the Gas Chromatographic Analysis of Hydrophobic Organic Contaminant Extracts from Great Lakes Water Samples, USEPA, Great Lakes National Program Office, GLNPO Organics SOP-10.

Data accumulation and data analysis are performed using the GC software Chem Station. The program produces a chromatograph with area counts for each peak. Each peak corresponds to one congener at a specific retention time. Using Aroclor standards, a response factor for each congener is determined. These response factors have been entered into an automated macro file for excel using Visual Basic for Applications, developed by research scientist Jeannie Spear. This has decreased the time required for data analysis and ensures a more uniform treatment of data. Area counts for each retention time are inserted into a separate file for each sample and the program calculates the concentrations of each congener. The file has also been programmed to determine the percent recovery of 207, the total concentration, and the concentrations of specific homologues and isomers groups.

Concerns from last time have been mostly resolved. The concern that more dechlorination was occurring after sampling the batch reactors was managed by quenching samples immediately after sampling. An experiment was performed, in which duplicate samples were taken at different times of dechlorination. One sample was dried with 1:1 MgSO4 and Na2SO4 and frozen for one week. The other sample was extracted immediately. The duplicate samples had the same concentration implying that quenching with this procedure is valid.

During the first reporting period, ZVI had been dewatered in a nitrogen freeze dryer and added to the batch experiments dry. This proved to be a hazard, however, because the ZVI would oxidize so quickly in air that while measuring it on the scale, it would ignite. To solve this problem, ZVI is now added wet, and its water content is determined independently.

Difficulties Encountered
The greatest difficulty for the research team at this time is the high standard deviations reported for duplicate samples. It is difficult to determine a mass balance because the samples are heterogeneous and the decrease of specific congeners is not always apparent. Also, as chlorines are removed from the PCB molecule, new congeners are hypothetically formed, therefore, creating additional difficulty in closing the mass balance. Monitoring biphenyl concentrations should help to resolve this problem.

Although proper clean-up steps are being performed on all extracted samples, there is one interference in the chromatograms that cannot be resolved. At 58.0 minutes, a large peak elutes. Consulting with Northeast Analytical Laboratories, it was determined that this peak is possibly the result of using plastics in the analysis such as gloves or pipet tips or an interference in the sediment that cannot be cleaned with sulfuric acid. It has been determined that the peak does not swamp out the peaks of any congener and thus has been ignored at this time.

It was not possible to monitor the corrosion products during the dechlorination process with XRD because of the large mass ratio of sediment to iron. It was attempted to monitor increase in iron species over time to calculate corrosion rates. However, the large ratio of sediment to iron makes it impossible for XRD to measure iron products formed only from the oxidation of ZVI.

Another setback that occurred during this reporting period is the malfunctioning of several instruments. The Labline Orbital Shaker Table purchased for the project had to be sent back to Barnstead International for maintainance, which resulted in two experiments being performed under static conditions, thus resulting in lesser PCB removal. The autosampler also malfunctioned during one batch experiment, which caused a slower turn around time for samples.

Anticipated Success in Meeting Project Objectives in Scheduled Project Period
It is anticipated that all objectives will be met in the scheduled two-year project period.

Preliminary Results
Standard operating procedures for complete analysis of contaminated sediment from the sampling of the batch experiments to the quantification of the gas chromatographic data have been developed and continue to be verified.

Determining Rate Constants
Two experiments were run taking samples daily and hourly. Results show that PCBs can be reduced by 63% in a fine marine sediment and by 95% in a sandy river sediment in approximately one day with one application of 3% Fe to sediment. Rate constants were determined for the daily sampling to be 0.1422 d^-1 and 3.141 d^-1, respectively (Figure 2-1 and Figure 2-2).

Degradation of a Single Congener
The dechlorination mechanism and the breakdown pathway are important in forming a mass balance. PCBs will break down to lesser-chlorinated congeners before reducing to biphenyl. Fortunately, unlike TCE, the breakdown product, biphenyl, is less toxic than the parent compounds. The breakdown of congener 207 was measured in clean sand with ZVI dechlorination. The results were difficult to interpret and a mass balance could not be defined; however, Figure 2-3 shows a chromatogram of a sample taken on day 7. This figure demonstrates that there are peaks for lower congeners, implying dechlorination of 207 to lesser-chlorinated congeners.

Various Iron Types
The objective of this experiment was to use three different types of zero valent iron to determine what one is most effective in dechlorinating PCBs. It is important to determine the significance of the size, purity, and surface area of the iron we will use in a full scale application. A micron scale (~50um) iron powder, 99.9%, supplied by CERAC in Milwaukee, WI would be the least expensive ($4.30/lb) and labor intensive for mass production. However, the colloidal nano-scale iron (1-100 nm, 33.5 m²/g) made at UNH in the lab and a nano-scale iron (~30 nm, 23.6 m²/g) provided by Toda Kogyo Corporation in Japan ($12/lb) produced by a different method than that manufactured at UNH (Wang and Zhang, 1996) were also tested.

The results of this experiment show that all of the ZVI used are equally efficient (Figure 2-4). The percent removals, all less than 50%, were not as high as expected with 5% ZVI. This deficiency could be due to the fact that the flasks were not shaken during the experiment (shaker table out for repair). More tests will be run with the different types of iron.

Tasks and activities for next reporting period

Tasks for the next reporting period

1. Close a mass balance by measuring reduction in total PCB concentration and increase in biphenyl concentration over time;
2. Measure redox potential and pH over time to understand the rate of corrosion of iron in a dechlorinating system;
3. Use a BET Surface Area Analyzer to measure accurate surface area of the three types of zero valent iron used in the last batch experiment;
4. Perform a batch experiment to evaluate the possibility of applying iron at three different steps, at 4 hours and 24 hours, to increase removal of PCBs;
5. Start a static laboratory experiment in a larger vessel, which will mimic in situ conditions.
6. Present a paper at the Second International Conference on the Oxidation and Reduction Technologies for In-Situ Treatment of Soil and Groundwater in Toronto, Canada.

Work plan to accomplish tasks
The work plan is predicted to proceed according to the project proposal. We will continue to develop and refine our laboratory procedures in order to optimize sample analysis and data accumulation. It is our hope to use outside sources to help expand the research project; for example, by purchasing commercially-made zero-valent iron, by establishing a contract with a remediation engineering company to trade ideas on how to implement this technique in the field, by sending samples to contracted laboratories, and by publishing a scientific journal article with new and innovative results. Our most imperative goal is to close a mass balance on the dechlorination of PCBs and we will work diligently to accomplish this task.

Concerns or difficulties
At this time, no concerns or difficulties are expected to obstruct further analyses in this project.

Expenditures
Expenditures were in the range anticipated for the work accomplished to date.