|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 2200385Ka 4500
controlfield tag 001 002068256
007 cr mnu|||uuuuu
008 100412s2009 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0003105
Environmental remediation of TNT using nanoscale zero-valent iron metal
h [electronic resource] /
by Erica Echols.
[Tampa, Fla] :
b University of South Florida,
Title from PDF of title page.
Document formatted into pages; contains 39 pages.
Thesis (M.S.)--University of South Florida, 2009.
Includes bibliographical references.
Text (Electronic thesis) in PDF format.
ABSTRACT: This research focused on the use of nanoscale zero-valent iron (NZVI) to remediate trinitrotoluene (TNT). Zero-valent iron has demonstrated effective degradation of TNT, however, these particles themselves have significant problems in treating sorbed phase TNT in the aerobic environment. This research was comprised of four areas: degradation studies of neat nano-iron with aqueous TNT, degradation studies of nanoiron emulsion with aqueous TNT, characterization of TNT in Vieques, Puerto Rico sediment, and Solid Phase Microextraction (SPME) technique interface with HPLC. Both neat and emulsion NZVI studies showed TNT degradation. More degradation was seen in studies using fresher iron. The results from our characterization study in Vieques, PR showed no presence of TNT within our detection limits of 0.0625ppm. Also, SPME is a new extraction solvent saving technique being explored because of its reproducible extractions in water. This work also gives a brief history of SPME and possible uses with TNT.
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
Advisor: Kathy Carvalho-Knighton, Ph.D.
Solid phase microextraction
x Environmental Science and Policy
t USF Electronic Theses and Dissertations.
Environmental Remediation of TNT using Nanoscale Ze ro-Valent Iron Metal by Erica Echols A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science Department of Environmental Science and Policy College of Arts and Sciences University of South Florida Major Professor: Kathy Carvalho-Knighton, Ph.D. Henry Alegria, Ph.D. Ashanti Pyrtle, Ph.D. Date of Approval: July 15, 2009 Keywords: Environmental Chemistry, Solid Phase Mic roextraction, Vieques, PR, Explosives, Nitroaromatics Copyright 2009, Erica Echols
i Table of Contents List of Tables i List of Figures iii Abstract v Chapter 1: TNT History 1 1.1 Introduction 1 Chapter 2: Zero-Valent Iron/Emulsion Studies 5 2.1 Introduction 5 2.2 Previous Works 8 2.3 Methods 9 2.3.1 Standard Preparation 9 2.3.2 HPLC Calibration 10 2.3.3 Salting Out Method for Emulsion 11 2.3.4 Nanoscale Zero-Valent Iron Studies (NZVI) 1 3 2.3.5 Nanoscale Zero-Valent Iron Emulsion Studies (EZVI) 13 2.3.6 EZVI/NZVI Results Chapter 3: Vieques, PR 18 3.1 Introduction 18 3.2 Previous TNT Characterization/Sampling Intere st 20 3.3 Field Sampling 24 3.4 Sample handling and storage 25 3.5 Sample Processing Methods 26 3.6 Results and Discussion 27 Chapter 3: Solid Phase Microextraction 30 4.1 Introduction 30 4.2 SPME Methods 32 4.3 SPME Extraction Results and Discussion 34 Chapter 5: Conclusions 36 References 37
ii List of Tables Table 1 Environmental Limits of TNT using Cancer Ri sk 2 Evaluation Guides (CREGs) Table 2 Percentage TNT Recovered from EZVI using 1 2 salting out extraction Table 3 Peak areas of byproducts that were formed d uring 15 the treatment of 2,4,6-Trinitrotoluene with nZVI Table 4 Explosive compounds on the Camp Edwards ana lyte 22 List Table 5 Vieques, PR sediment analysis results 28
iii List of Figures Figure 1. Structure of 2,4,6-Trinitrotoluene 1 Figure 2. Pathways of TNT reduction 5 Figure 3. Photograph and schematic of EZVI droplet 7 Figure 4. Calibration results of TNT Standards on t he HPLC 11 Figure 5. 48 hour TNT degradation with NZVI 14 Figure 6. Results from one week study using NZVI 15 Figure 7. NZVI/EZVI 3 week comparison study 16 Figure 8. Percent grain size in Mosquito Bay, VNWR, Puerto Rico 21 Figure 9. Location of the Massachusetts Military Re servation and 21 Camp Edwards Figure 10. Conceptual Model of Contaminant Transpor t for Camp 23 Edwards Figure 11. Map of Vieques with two sampling locatio ns circled in red 24 Figure 12. Topsoil sampling diagram 25 Figure 13. HPLC Chromatogram of 8330 Nitroaromatics 27 Calibration Standard obtained from Restek Corp. Figure 14. HPLC Chromatogram of Kiani Lagoon Core S ample 27 0-5cm Figure 15. SPME Fiber absorption 30 Figure 16. TNT Concentration vs. Peak Area results from SPME 34 extraction in water
iv Figure 17. TNT Concentration vs. Peak Area results from SPME 34 extraction in acetonitrile Figure 18. TNT Concentration vs. Peak Area results from SPME 35 extraction in methanol
v Environmental Remediation of TNT using Nanoscale Ze ro-Valent Iron Metal Erica Echols ABSTRACT This research focused on the use of nanoscale zero -valent iron (NZVI) to remediate trinitrotoluene (TNT). Zero-valent iron has demonstrated effective degradation of TNT, however, these particles themselves have si gnificant problems in treating sorbed phase TNT in the aerobic environment. This researc h was comprised of four areas: degradation studies of neat nano-iron with aqueous TNT, degradation studies of nanoiron emulsion with aqueous TNT, characterization of TNT in Vieques, Puerto Rico sediment, and Solid Phase Microextraction (SPME) te chnique interface with HPLC. Both neat and emulsion NZVI studies showed TNT degr adation. More degradation was seen in studies using fresher iron. The results fr om our characterization study in Vieques, PR showed no presence of TNT within our detection l imits of 0.0625ppm. Also, SPME is a new extraction solvent saving technique being explored because of its reproducible extractions in water. This work also gives a brief history of SPME and possible uses with TNT.
1 Chapter 1: TNT History 1.1 Introduction Trinitrotoluene (Figure 1) is a solid organic compo und used chiefly as an explosive and is prepared by stepwise nitration of toluene. Figure 1. Structure of 2,4,6-Trinitrotoluene The explosive itself and some of its degradation an d transformation products are considered serious environmental contaminants w ith potential harmful and toxic effects on animals, plants and humans. Due to its relatively water soluble nature, it can migrate through the soil to cause unwanted grou ndwater contamination. TNT is toxic to algae and invertebrates, and chronic expos ure to TNT by humans causes harmful health effects, including anemia, abnormal liver function, cataract development, and skin irritation. Human exposure to ambient levels of TNT above 0.1 mg/m3 for an 8-hour time weighted average (TWA) without appropriate
2 respiratory protection is the toxic (U.S. DOD 1984) The lethal dose for humans is 12 grams and the drinking water unit risk is 9.0E-7 ppb (U.S. DOD 1984). Table 1 lists guidelines for limits of TNT as set by the Ag ency for Toxic Substances and Disease Registry (ATSDR). Table 1. Environmental Limits of TNT using Cancer Risk Evaluation Guides (CREGs) Source Limit Soil 20 ppm Surface Water 1 ppb Ground Water 1 ppb Drinking Water 9.0E-7 ppb TNT has been introduced in the environment by many mechanisms. It is an explosive used in military shells, bombs, and grena des, industrial applications, and underwater blasting. Environmental contamination m ay occur while manufacturing the compound, transport and detonation of bombs, an d recycling and storage of the TNT containing explosives. Contaminated sites inc lude, but are not limited to, production sites, warehouse facilities, and waste d estruction sites, as well as explosives testing sites where TNT-containing produ cts are detonated and dispersed. Additionally, wastewater from processing explosives can be contaminated with TNT. For years, this water was discarded outside of the manufacturing facilities, on the
3 ground or in lagoons that leached explosives into s oil and groundwater (Boopathy 1998). There are several possible strategies for the remov al of TNT contaminants in soil and water. Most cleanup methods include exca vation of the polluted sites, followed by incineration or composting. In soil, s ome remediation strategies include biotechnological applications of bacteria and fungi by way of composting and land farming (Jarvis 1998). Microorganisms cultured wit hin the compost piles degraded the TNT by breaking it down into its fundamental pa rts and using them as a nutritive carbon source (Jarvis 1998). Other remediation tech niques bypass the composting phase and instead add isolated degradative enzymes from the microorganisms themselves directly to the excavated soil, allowing for a speedier approach. Bioremediation is one of many reformative pathways that have been used for TNT removal. One of the newest of these methods is phy toremediation. This simply refers to the use of plants instead of chemicals to decontaminate the environment. Phytoremediation has been a successful tool for the remediation of organic pollutants (Susarla et al. 2002; Burken 2003), radi onuclides and heavy metals (Riley et al. 1992; Stern et al. 1996). Numerous studies have indicated the ability of several plant and algal species to take up and metabolize T NT from water and soil (Schnoor et al. 1995; Hughes et al. 1997; Rivera et al. 1998 ; Hannink et al. 2002). Examples of such plants are stonewort (Nitella), hybrid poplars (Populus), and duckweed (Lemna minor). Studies performed by Schnoor in 1995 found that phytodegradation occurs with the use of plant enzymes, such as nitroreducta se, which have the ability to degrade aromatic rings in the absence of microorgan isms (Schnoor et al. 1995).
4 Although phytoremediation is environmentally friend ly, its main drawback is that it is a slow process. Concerns about the environmental fate of TNT residu es have intensified with identification in vegetation of contaminated sites possibly allowing TNT, TNT metabolites, and plant-produced TNT intermediates t o be introduced into the food chain. (Boopathy 1998). In the US, the Army alone has estimated that over 1.2 million tons of soil have been contaminated with ex plosives, and the impact of explosives contamination in other countries is of s imilar magnitude (Lewis 2004). The ultimate aim of this research is to apply the e mulsion technique to remediate TNT contaminated soils and water. Zero-v alent iron, Fe0, has demonstrated effective degradation of trinitrotolue ne (TNT); however, these particles by themselves have significant problems in treating sorbed phase TNT in an aerobic environment. This research focuses on emulsifying Fe0 particles that are capable of promoting rapid and complete degradation of TNT mol ecules. This work also examines the distribution of TNT in samples collect ed from Mosquito Bay and Kiani Lagoon in Vieques, Puerto Rico. By collecting sedi ment and water samples from these sites, we were able to do a comparative study to examine explosive residues are still in these areas, and if so, how much of this c ontaminant is traveling across the island into the residential area. Experimental studies include: 1) Degradation studies of neat nano-iron with aqueo us TNT (Chapter 2) 2) Degradation studies of nano-iron emulsion with a queous TNT (Chapter 2) 3) Characterization of TNT in Vieques, Puerto Ric o Sediment (Chapter 3) 4) Solid Phase Microextraction (SPME) technique w ith HPLC (Chapter 4)
5 Chapter 2: Zero-Valent Iron/Emulsion Studies 2.1 Introduction Iron has demonstrated effective results in the degr adation of TNT (Agrawal 1996). Although the primary focus is on iron, othe r metals such as zinc, tin, and magnesium have potential to make good zero-valent m etal particles. Iron has been most commonly used in the past due to its low cost and non-toxicity. NZVI has a small particle size (e.g. < 100 nm) and therefore a larger specific surface area (SSA) in comparison to microscale or granular irons (Geig er and Knighton 2009). Figure 2. Pathways of TNT reduction
6 This research explores whether emulsification of ir on particles will promote higher degradation compared to neat, non-emulsified iron particles. An emulsion is the mixture of two liquids that would not be combin ed under normal circumstances. Emulsions provide a crossing point between the two immiscible liquids and increase reaction surface area (Bibette 2002). There are tw o distinguishable classes of emulsification techniques: 1. Emulsification using a colloid mill or a high-press ure homogenizer; and 2. Emulsification using laminar flows or gentle stirri ng. This research utilizes the first technique. Emulsi fying is accomplished by slowly adding one ingredient to another while simul taneously mixing rapidly. This disperses and suspends tiny droplets of one liquid through another (Bibette 2002). However, the two liquids would quickly separate aga in if an emulsifier were not added. Emulsifiers such as surfactants are liaisons between the two liquids and serve to stabilize the mixture (Sjoblom 2006). The drop lets of water in oil are being tested as minireactors for the reactions of contaminants w ith nanoparticles of metals (Sjoblom 2006). Our emulsions were developed at UC F and consist of iron, water, oil, Span 85 (Quinn et al 2005). Figure 3 by Quinn et al. 2005 shows a micrograph of an EZVI droplet and a detailed illustration. This illustration shows the iron particles in the aqueous phase surrounded by an oil layer hel d together by the surfactant.
7 Figure 3. Photograph and schematic of EZVI droplet illustrating particles of nZVI in water surrounded by the oil-liquid membrane (Quinn et al 2005). These heterogeneous mixtures of oil, water, iron, a nd surfactant are great solvents for both polar and non-polar compounds (Sj oblom 2006). The emulsion particles contain the nano zero valet iron (NZVI) i n water surrounded by an oil-liquid membrane. NZVI promotes oxidative, abiotic degrada tion. Therefore, when we combine the capabilities NZVI and the emulsion, we now have possibilities for anaerobic degradation which is driven by the donati on of electrons from the surfactant and oil.
8 2.2 Previous Works In 1996, Agrawal and Tratnyek express the urgency t o move forward in nitroaromatics remediation work. The authors sugge st that the persistence of nitroaromatic compounds continues because of their use in munitions, insecticides, herbicides, and pharmaceuticals. Nitroaromatic com pounds can also be produced naturally in the environment from aromatic contamin ants (Agrawal et al., 1996). Although transformation reaction intermediates and products are commonly hydroxlyamines, nitroso compounds, and aromatic ami nes, these products are still of concern as environmental contaminants. Therefore, it is suggested that these compounds require transformation beyond nitro reduc tion (Agrawal et al., 1996). Their research investigates the use of ZVI to prom ote nitro reduction. Experiments examining effects of iron surface, effe ct of pH, and the effect of mixing rate using nitrobenzene were conducted. The iron s urface had a linear relationship with the rate constant for the degradation of nitro benzene. Agrawal et al. (1996) states that greater surface area was generally asso ciated with a higher rate constant, which yields faster degradation of nitrobenzene. In addition, they indicate no clear effects within the natural environmental pH ranges of 6-8 and a slight decrease in degradation rates for pH below 6. Agrawal et al. ( 1996) found that mixing rates do positively impact degradation of nitro groups. Met hods incorporated continuous shaking between time points in the study. Likewise our methods utilized the aid of a shaker table to promote constant interaction amount analyte and iron.
9 Military underwater testing often occurs directly o ffshore and sometimes may occur in saline or freshwaters. Brannon et al. (2 002) compared the fate and transport of explosives in both saline and freshwater environ ments. There was general agreement in their results between freshwater and s altwater. Their research concludes that the existing freshwater data can be applied to marine systems. The NZVI and EZVI studies were performed to chart d egradation of TNT and also to record formation of byproducts formed. TNT standards were received and calibration curves were conducted in order to obtai n peak areas that correspond to ppm concentrations. EPA method 8330 utilizes salti ng out extraction for low concentration samples. Therefore, salting out stud ies were performed to determine the extraction efficiency. When this method of ext raction proved unreliable, sonication was used to aid in the extraction of TNT from emulsions by breaking the emulsion apart to release any TNT or byproducts tha t may be trapped within the aqueous/NZVI layer. 2.3 Methods 2.3.1 Standard Preparation TNT standards were obtained from Restek Corporation These standards are 5000ppm in acetonitrile. NZVI and Span 85 used for the preliminary studies were obtained from the Department of Chemistry at Univer sity of Central Florida in Orlando, FL. The acetonitrile was obtained from Fi sher Scientific. Standards were prepared for use by diluting the Restek Standard in water to achieve various concentrations used for the experiments.
10 2.3.2 HPLC Calibration Methods A Shimadzu High Performance Liquid Chromatography ( HPLC) VP10 equipped with a Premier C18 column was utilized for sample analysis. All samples were analyzed with a Shimadzu HPLC. Calibrations w ere performed using 8330 nitroaromatics calibration standard obtained from R estek, Inc. This calibration standard contained the following: 2-nitrotoluene; 2 ,4,6-trinitrotoluene; HMX; pnitrotoluene; 2,6-dinitrotoluene; 1,3-dinitrobenzen e; 4-amino-2,6-dinitrotoluene; nitrobenzene; 2-amino-4,6-dinitrotoluene; 1,3,5-tri nitrobenzene; tetryl; 2,4dinitrotoluene; 3-nitrotoluene; and RDX. The 8330 Calibration Standard contains each constituent at 1000ppm concentration. Duplic ate dilutions were made of this standard to in concentrations varying from 0.0625pp m to 100ppm and subsequently a calibration sequence was run. A sample calibration is shown in figure 4 which shows concentrations of 6.25, 12.5, and 25 ppm. Prelimin ary samples were run with 50/50 methanol/water solvent. The mobile phase was chang ed to 50/50 acetonitrile/water to maintain peak visibility and reduce noise. Samples were run with a blank of acetonitrile between each sample. The HPLC method and conditions were as follows: Platinum C18 column Temperature-35 C Pump Flow-1mL/min 50/50 Acetonitrile/water Run time: 20min
11 y = 31943x + 52584 R = 0.986 0 100000 200000 300000 400000 500000 600000 700000 800000 900000 051015202530areappmTNT Calibration Figure 4. Calibration results of TNT Standards on the HPLC 2.3.3 Salting Out Method for Emulsion When extracting TNT from solution with emulsion, sa lting out method was explored. One gram of salt was added to liquid sol ution and allowed to stir. After salt is in solution, three mL of acetonitrile was a dded three times in fifteen minute intervals. Top layer was extracted with a filtered syringe and put into an autosampler vial for HPLC. While using the salting out method, an extraction e fficiency trial was needed. The salting out method was performed in two trials using the same emulsion. The results from the two trials are found below. Initi al concentrations of 100ppm had an average of 47% recovered for trial one and 11.46% f or trial two.
12 Table 2. Percentage TNT Recovered from EZVI using salting out extraction nr n nnr Salting out methods of extraction prove unreliable when used to extract TNT from EZVI. For each sample different recovery amounts w ere obtained and there is high variance among the samples. Since average recovery rates are below 50%, salting out is not recommended for use with EZVI. Although the salting method is a suggested method of extraction in EPA Method 8330, an alterna te extraction technique was utilized. Based upon previous EZVI studies, sonicat ion was found to be an effective extraction technique and was therefore utilized in this research (Quinn et al 2005).
13 2.3.4 Nanoscale Zero-Valent Iron Studies (NZVI) Several studies have been conducted testing the deg radative properties of NZVI. NZVI studies were conducted using one gram o f NZVI and 5mL of various concentrations of TNT in water or acetonitrile in a n amber screw top vial. Control vials contained five mL of TNT standard. Samples w ere run in duplicate, placed on a shaker table, and agitated for varying periods of t ime. Extractions were taken at 0hr, 3hrs, 6hrs, 24hrs, 48hrs, and 1 week. Each extract ion was filtered using Milex Syringe Driven Filter Units, placed in two mL autos ampler vials, and analyzed using High Performance Liquid Chromatography (HPLC). 2.3.5 Nanoscale Zero-Valent Iron Emulsion Studies (EZVI) NZVI emulsions were made by combining one hundred m L of water, twenty grams of NZVI, eighty mL of Mazola corn oil, and th ree grams of Span 85 surfactant. Water and NZVI was added to the blender and allowed to mix. While blending, corn oil and surfactant was added slowly. After all oil /Span was added, the mixture was blended for an additional minute until a homogenous mixture was achieved. Emulsion stability was tested by placing a small am ount (approximately 3g) in a clear vial with DI water. The vial was capped an d shaken by hand for approximately 30 seconds. A stable emulsion leaves the water in the vial clear and may contain smaller amounts of emulsion floating th roughout. However, an unstable emulsion will result in cloudy/milky solution in th e vial and should be discarded. In amber screw top vials, two and a half grams of E ZVI or one gram of NZVI and five mL of varied (50, 100, and 120ppm) concent rations TNT in water were
14 combined. Samples were allowed to shake on a shak er table until extraction time points were reached. Extraction times were at vary ing time points ranging from 0hr to 3weeks, depending on the study. Upon extraction samples were sonicated for 15 minutes and were filtered using Milex 0.45micron sy ringe filters. 2.3.6 EZVI/NZVI Results The first study conducted with the NZVI was in Nove mber 2006 over a 48 hour time period. The results from this preliminary study are illustrated in Figure 5. Figure 5. 48 hour TNT degradation with NZVI In this study, NZVI proves effective in degrading TNT. The concentrations of TNT in solution decreased over 50% more when treate d with NZVI than the control group. The Error bars represent +/one standard d eviation. While the concentration of TNT was monitored, byproducts with retention tim es from 2.3-18.2 minutes after 0 20 40 60 80 100 120 0102030405060 Time (h)Concentration (ppm) 2 day NZVI Control NZVI
15 sample injection were observed. Table 3 lists bypr oduct retention times and associated peak areas. Table 3. Peak areas of byproducts that were formed during th e treatment of 2,4,6Trinitrotoluene with NZVI. By-Product 1 By-Product 2 By-Product 3 By-Product 4 By-Product 5 By-Product 6 Retention time (min) 2.3 3.8 9.8 10.2 10.2 18.2 TIME (h) Peak areas 0 660756 755660.5 n/a n/a n/a n/a 2 1265437.5 2045063.5 n/a n/a n/a 8807246 4 1950294.5 2256897 n/a n/a n/a n/a 24 1014709.5 n/a 81193 115931.5 150139.5 100714 48 1297139.5 n/a n/a 78082.5 119858 n/a Concentrations decreased over time, indicating that as the byproducts are being formed, they are also decomposing as well, some to complete degradation. Further analysis with mass spectroscopy would assist in byp roduct identification. In Spring 2007, a longer study was performed using NZVI. In this study the time period was one week and the results are shown in Figure 6. 0 10 20 30 40 50 050100150200Concentration (ppm)Time (hrs)One Week ZVI Study TNT Control TNT with ZVI Figure 6. Results from one week study using NZVI
16 In this study, the concentration of TNT only decrea sed by 15.5% in the first 48hours in contrast to the preliminary study which showed o ver 50% degradation over the same time period. Upon further investigation it wa s discovered that the reactivity of NZVI used was minimal. Therefore, fresh NZVI was o btained from Toda in Japan. The final study utilizing NZVI and EZVI was conduc ted over a three week time span. In this study, both older NZVI (NZVI UC F) and fresh NZVI (NZVI Toda) were used in neat and emulsion forms. The results are shown in Figure 7. 0 20 40 60 80 100 120 140 01234Concentration (ppm)Time (Weeks)3 Week NZVI/EZVI Study EZVI UCF NZVI UCF EZVI Toda NZVI Toda TNT Control Figure 7. NZVI/EZVI 3 week comparison study In this study, NZVI UCF had 51.32% TNT degradation after three weeks in comparison with NZVI Toda which had 100% degradatio n after only two weeks. Likewise, EZVI UCF had 26.69% TNT degradation as co mpared to EZVI Toda with 57.10% TNT degradation over the same three week tim e period. Both the neat NZVI and EZVI experiments using NZVI Toda proved successful; however, emulsion results using EZVI ma de with NZVI Toda yielded
17 higher degradation than current and prior studies u sing EZVI UCF. When examining EZVI results, it is apparent that when emulsions ar e made using newer, more reactive iron, higher degradation is achieved. When considering using NZVI in-situ for remediation oxidation effects to the non-reacted metal should be considered. Agrawal in dicates ZVI can be easily oxidized to ferrous iron by many substances, which can dramatically decrease the surface area available for degradation. In aqueo us environments, metal corrosion can occur to inhibit the reactivity of ZVI to degra de TNT. Because of this, emulsions show promise for reducing oxidative effects in-situ Emulsions have remained stable in a beaker at room temperature for one year. Alth ough neat metal experiments illustrate rapid reduction in laboratory samples, e mulsions show potential for more efficient degradation over longer periods of time w ith decreased loss of non-reacted surface area because of the oil/water barrier that is created at the emulsion surface. Although longer studies are vital in determining th e long term effectiveness of NZVI and EZVI, most activity occurs in the initial 48hours of exposure. The first 48 hours is when the initial degradation occurs and pr imary intermediates are formed. After longer exposure, these intermediates are brok en down and other transformation products appear. In order to look at the full scop e of TNT degradation, both short term and long term experiments have to be considere d.
18 CHAPTER THREE Vieques, PR 3.1 Introduction During the last 100 years the use of conventional w eapons in training and in combat has left behind a polluted environment with noxious compounds and heavy metals (Simini 1995, DOD 1999, Diaz 2003). The thre at of contamination to soil and groundwater exists because of possible physical mig ration or from uptake by organisms (Rittmann 1994, Mueller 1995). The veget ation of the contaminated sites pose a serious and immediate problem, since plant u ptake would provide an entry of munitions materials, transformation products and pl ant produced metabolites into food webs (Mueller 1995). Remediation of contaminat ed sediment and water is necessary to avoid further distribution of TNT thro ughout the environment and to eliminate any potential future hazards. In Vieques, Puerto Rico, The Atlantic Fleet Weapons Training Area consists of areas and nearby waters that have become contami nated primarily by United States Department of Defense (DOD) activities (US EPA, 200 5). These areas include the Eastern Maneuver Area (including Camp Garcia), the former Surface Impact Area, the Live Impact Area, and the Eastern Conservation Zone on the east end of Vieques (US EPA, 2005). Within the Naval Ammunition Support Detachment (NASD) which is located on the western end of Vieques, the facil ity incorporates, but is not necessarily limited to, eight areas for which the N avy considers the ongoing
19 investigations and remediation to be incomplete (US EPA, 2005). The waters encompassing these areas are widely uncharacterized Known areas of concern include waters influenced by target practice on the eastern side Vieques, areas where ships were anchored north of Vieques, and waters cl oser to the western side of Vieques, including Mosquito Pier (US EPA, 2005). T he east side of the Island of Vieques was used for all facets of naval gunfire tr aining and munitions testing (illustrated in Figure 11). The west side of the i sland was used mostly for ammunition loading and storage, and vehicle and fac ility maintenance. The central section of the island is predominately residential. According to an Environmental Protection Agency (EP A) report, the amount of unexploded ordnance and debris from exploded ord nance was widespread in range areas of Vieques and in the surrounding waters. Haz ardous substances associated with ordnance use may include mercury, lead, copper, mag nesium, lithium, perchlorate, TNT, napalm, and depleted uranium among others (US EPA, 2005). The existing hazardous substances may also include a range of ch emicals such as poly-chlorinated bi-phenyls (PCBs), solvents, and pesticides (US EPA 2005). EPA also affirms that Vieques has a population of a bout 9,300. Since Vieques is becoming a point for active tourism, bot h visitors and residents access beaches, fisheries, and recreational waters. If th ese areas are impacted by past military training, this can have adverse effects on the health of residents and tourists alike. Since Naval activities ended in May 2003 lar ge portions of the impacted areas have been set aside as a wildlife refuge that is ho me to at least 25 endangered species (Federal and Puerto Rican), and other sensitive env ironments (US EPA, 2005).
20 Characterization and remediation of TNT in these ar eas is necessary to protect the health and welfare of the present and future Vieque s residents and visitors. 3.2 Previous TNT Characterization/Sampling/Interes t Our research group collaborates with the Dr. Ashant i Pyrtles Radiogeochemistry Lab in the College of Marine Scie nce at the University of South Florida. Environmental monitoring and assessments of Vieques, PR conducted by the Pyrtle Group have been ongoing for the past three y ears. Because of the military activity that occurred due to the US NAVY Atlantic Fleet Weapons Training Facility, several contaminants, including heavy metals and de pleted uranium, have been released (Ithier-Guzmn et al. 2007). Surface and downcore sediment analysis indicate the presence of Cs-137 in the sediments of Vieques with concentrations ranging from below detection limits to 0.01 Bq/g (I thier-Guzmn et al. 2007). As with TNT, Cs-137 retention is highly influenced by particle size and clay mineralogy. All sediment samples from the Mosquito Bay (Figure 11) site at depths deeper than 12cm contained Cs-137 (Mayo et al. 2006). Notable C s-137 activities were detected at deeper depths for cores from Kiani Lagoon (Mayo et al., 2006). Kiani Lagoon and Mosquito Bay displayed clay size p articles which averaged 4.53 % and 2.12%, respectively (Mayo et al. 2006). Mosquito Bay exhibits differences down core with slight variations in the upper 10 cm and deeper (24-30 cm) sediments while the middle section (11-24 cm) r emain fairly constant, as indicated in Figure 8 (Mayo et al., 2006). Because of common environmental hazard concerns, this site was chosen as a characterizatio n site for TNT and nitroaromatics.
21 Figure 8. Percent grain size in Mosquito Bay, VNWR Puerto Rico (Mayo et al. 2006) Environmental risk assessment of nitroaromatics has been previously conducted at Camp Edwards and the Massachusetts Mil itary Reservation shown in Figure 9 (Clausen 2004). Figure 9. Location of the Massachusetts Military R eservation and Camp Edwards This study consisted of 7,833 shallow surface soil samples representing 1,989 individual locations from 182 areas of investigatio n. Samples consisted of 533 soil
22 boring profile samples came from 146 soil borings, 69 sediment samples from 19 water bodies, and 64 surface water samples from 19 bodies of water. Five storm water samples from the perimeter of the Impact Area were taken and 3,959 groundwater profiling samples from 256 borings. La stly, 1,467 groundwater samples were obtained from 651 monitoring wells at 256 loca tions. The explosive residues that were initially on the C amp Edwards list prior to this study are shown in Table 4. However, over 80 additional residues were actually found on the premises. Table 4. Explosive compounds on the Camp Edwards a nalyte list 1,3-dinitrobenzene 2-nitrotoluene Dinitroso-hexahydro-1,3,5triazine (DNX) 1,3,5-trinitrobenzene (TNB) 3-nitrotoluene Nitroso-dinitro-hexahydro1,3,5-triazine (MNX) 2-amino-4,6-dinitrotoluene (2A-DNT) 4-nitrotoluene Tri-nitroso-hexahydro-1,3,5triazine (TNX) 4-amino-2,6-dinitrotoluene (4A-DNT) Tetryl Octahydro-1,3,5,6-tetranitro1,3,5,7-tetrazocine (HMX) 2,4-diamino-6-nitrotoluene (2,4-DANT) 2,4,6-trinitrotoluene PETN 2,6-diamino-4-nitrotoluene (2,6-DANT) Picric Acid Hexahydro-1,3,5-trinitro-1,3,5triazine (RDX) 2,4-dinitrotoluene (2,4-DNT) Nitrobenzene Nitroglycerine Contaminants included nitroaromatics as well as man y organic pollutants were detected. At Camp Edwards, although primary focus of the researchers was on explosive residues, their research discovered ma ny other contaminant issues. Their findings report that contaminants were found in surface soils in areas where artillery and mortar firing occurred (Clausen et al 2004). Additional
23 contamination was found in deeper soils and groundw ater where impact blasting and detonations took place. Our sites were chosen because of prior research conducted by the Pyrtle group which revealed areas of radionuclide contamination. Figure 10 from Clausen et al. (2004 ) is a schematic of how contaminants can enter and be transported throughou t an environmental system. Figure 10. Conceptual Model of Contaminant Transpo rt for Camp Edwards proposed by J. Clausen et al. 2004 In addition to the topsoil contamination, water sol uble contaminants can percolate through the deeper soil depths and enter groundwater flow. Once the contaminants are fully mobile in the groundwater, t ransport is facilitated by the natural direction groundwater flow.
24 3.3 Field sampling Samples were collected from two locations in Puert o Rico (Figure 11). US Navy Public information site Figure 11. Map of Vieques with two sampling locatio ns circled in red First, in Mosquito Bay, which is located on the sou thern border of the residential section and the naval section of Vieques. Second, in the northern section of the island at Kiani Lagoon sediment samples were collected. C ores were taken using polycarbonate core liners and the top 10cm was coll ected for analysis. Water samples were collected from each site in 1L amber screw cap bottles. Additionally, Lamotte sampling tubes were used to collect topsoil/sedimen t samples A-F around each core (Figure 12). Kiani Lagoon Mosquito Bay
25 Figure 12. Topsoil sampling diagram 3.4 Sample handling and storage After sample collection, the first 5 cm of each co re was put into an amber jar and capped. The rest of the core each core (5-10cm ) was then added to another amber glass jar. All samples were stored in the am ber glass jars and kept cold at approximately 4C and transported back to the labor atory for analysis. Extraction and HPLC analysis was performed within 14 days of sampl e acquisition. In a study performed by Grant (1993), the two nitr amines HMX and RDX were stable for all soils under all storage tempera tures. For the three nitroaromatics (TNB, TNT and 2,4-DNT) the results were very differ ent, in that all three analytes rapidly degraded in spiked soils at room temperatur e, more slowly degraded under refrigerator temperature, and remained quite stable when frozen (Grant, 1993). Of the three, TNB degraded most rapidly, followed by TNT a nd 2,4-DNT. The degradation Core B C A D F E
26 at room temperature and in the refrigerator was muc h faster for one soil than for the others (Grant 1993). For the field-contaminated soi l, the nitroaromatics were much more stable, even at room temperature, although som e degradation occurred (Grant 1993). When frozen, a maximum holding time of eight weeks is recommended (Grant 1993). Although this study deals particularly with soils, we achieved full analysis of the samples within one month. 3.5 Sample Processing Methods Pore water from core samples was extracted using ce ntrifugation at 3,000 rpm for thirty minutes at 4C per sample, following the Unsaturated Flow Apparatus Method for extracting Pore Fluids from relatively i mpermeable and highly unsaturated porous media (UFA 1997). The extracted pore water was then filtered through a 0.45m PVDF Durapore Millex filters using a 5 ml SGE syr inge. Highlevel direct injection and low-level salting out ex tractions were preformed on all surface and pore water samples using EPA method 833 0 coupled with reversed-phase HPLC (RP-HPLC) analysis. After centrifugation, the soil samples were immedi ately dried in a muffle oven at 80C for 24 hours. Samples were ground and homog enized in an acetonitrile-rinsed mortar and passed through 63m mesh sieve. Subsamples were collected and stored in 20mL IChem vials. For analysis, each soil sampl e was extracted with acetonitrile in an ultrasonic bath for 18hours, combined with ca lcium chloride solution and filtered through 0.45m PVDF filters following Method 8330 and stored for later HPLC analysis.
27 3.6 Results and Discussion 8330 Nitroaromatics calibration standards were obta ined from Restek Corporation. Figure 13 below illustrates a chromato gram obtained using a standard. 0 2 4 6 8 10 12 14 mAu 0 50 100 150 mAu 0 50 100 150 Spectrum Max Plot052207 Cal 8330 Retention Time Figure 5. HPLC Chromatogram of 8330 Nitroaromatics Calibration Standard obtained 0 2 4 6 8 10 12 14 mAu 0 50 100 150 mAu 0 50 100 150 Spectrum Max Plot052207 Cal 8330 Retention Time Figure 5. HPLC Chromatogram of 8330 Nitroaromatics Calibration Standard obtained Figure 13. HPLC Chromatogram of 8330 Nitroaromatic s Calibration Standard obtained from Restek Corp. A sample chromatogram from one of the samples from one of the samples taken from Kiani Lagoon is shown in Figure 14. Minutes 0 2 4 6 8 10 12 14 16 18 20 mAu 0 5 10 mAu 0 5 10 Spectrum Max PlotAug 0307 KL Top LT 63 Figure 4. HPLC Chromatogram of Kiani Lagoon Core Sample 0 5cm Minutes 0 2 4 6 8 10 12 14 16 18 20 mAu 0 5 10 mAu 0 5 10 Spectrum Max PlotAug 0307 KL Top LT 63 Figure 4. HPLC Chromatogram of Kiani Lagoon Core Sample 0 5cm Figure 14. HPLC Chromatogram of Kiani Lagoon Core Sample 0-5cm This sample was from the top 5cm of the core. Al l chromatograms contained the two peaks just before two minutes. These peaks occurred because of the salinity of the seawater and were not apparent in blank acet onitrile samples. Table 5 lists the results from sediment analysis. HPLC detection lim its on the HPLC were 0.0625ppm and TNT levels were undetectable in all samples wit hin this limit.
28 Table 5. Vieques, PR sediment analysis results Location Grain Size (63m) Grain Size (63m) Kiani Lagoon TNT Levels TNT Levels Top (0-5cm) Undetectable Undetectable Bottom (5-10cm) Undetectable Undetectable Surface Sample A Undetectable Undetectable Surface Sample B Undetectable Undetectable Surface Sample C Undetectable Undetectable Surface Sample D Undetectable Undetectable Surface Sample E Undetectable Undetectable Surface Sample F Undetectable Undetectable Mosquito Bay Top (0-5cm) Undetectable Undetectable Bottom (5-10cm) Undetectable Undetectable Surface Sample A Undetectable Undetectable Surface Sample B Undetectable Undetectable Surface Sample C Undetectable Undetectable Surface Sample D Undetectable Undetectable Surface Sample E Undetectable Undetectable Surface Sample F Undetectable Undetectable Samples were collected from areas where prior work from our lab has been done. If there were a return sampling trip to Vieq ues, a variation to the sampling area could be implemented by sampling different regions that may be closer to where
29 bombing actually took place, including areas where prior research has previously found TNT. If explosives had been found, we could execute our current remediation techniques with the Vieques sediment. This would m ake it possible for us to assess the effects of using zero-valent iron metal and iro n metal emulsions on contaminated Vieques sediment.
30 Chapter 4: Solid Phase Microextraction 4.1 Introduction Although methods such as solvent extraction and son ication (such methods are employed in EPA Method 8330) have proved succes sful, newer technologies such as Solid Phase Microextraction (SPME) are becoming abundant. SPME, coupled with HPLC, has the potential to facilitate the extr action process illustrated in Figure 15. Figure 15. SPME Fiber absorption (www.labhut.com) SPME extraction is a two part process. Part I cons ists of separation of the analyte between the fiber coating and the sample (P ena-Luengas 2007). Part II consists of desorption of the analyte from the fibe r and analysis on HPLC (PenaLuengas 2007). When using SPME fibers, one can ext ract directly by submersion in a liquid sample, or by sampling the headspace in a soil/soil slurry sample.
31 Gaurav 2007 explores the effect of salinity on the extraction of nitroaromatic compounds using SPME fibers. Their study looked at solutions of HMX, RDX, TNB, DNB, Tetryl, 3,4-DNT, TNT, and 4-ADNT. Each e nergetic compound was examined in solutions of 3, 10, 20, and 30% NaCl (w /v) in water. The results from their study shows that better SPME fiber sorption o ccurred in more saline water, with 30% having the highest extraction amount for every compound. This makes SPME use in ocean/saline samples ideal versus processing the salt water through the HPLC. This research examined the limits of SPME fibers an d utilizes these fibers as a secondary extraction technique. SPME extraction is considered complete when equilib rium is achieved between between the sample and the SPME fiber. Paw liszyn introduces the following equation to describe equilibrium conditio ns: n= K fs V f V s C 0 KfsVf+Vs where n is the extracted amount by the fiber, Kfs is the fiber coating/sample matrix distribution constant, Vf is the volume of the fiber coating, Vs is the volume of sample, and C0 is the initial concentration the given analyte in the sample (Pawliszyn, 1997). This equation also reflects that the relat ionship between the sample concentration and the amount of analyte to be extra cted is directly proportional (Pawliszyn, 1997). In order to ensure rapid and complete extraction, agitation is suggested by Pawliszyn. In aqueous solutions, more effective ag itation techniques such as sonication, stirring, vial movement (facilitated by a shaker table), or faster solvent
32 flow are proposed (Pawliszyn, 1997). During extra ction, the liquid immediately encompassing the fiber becomes depleted of target a nalytes creating a depletion zone. Agitation is required to minimize the effec t caused by the depletion zone surrounding the fiber during extraction (Pawliszyn 1997). SPME shows much promise as an innovative extraction technology. Due to its sampling procedure, it can greatly reduce the a mount of solvent needed to extract the target analyte, as opposed to other methods suc h as EPA Method 8330. Using SPME and polished techniques, extraction may be ach ieved without the addition of any addition solvent to the sample. In contrast, E PA Method 8330 requires the addition of 10-20mL of solvent per sample. In larg e batch studies, this can create much waste for disposal contributing to further env ironmental contamination. 4.2 SPME Methods When implementing SPME techniques, Pawliszyn sugge sts the four stages to be included in method development: extraction stra tegy, hardware selection, initial optimization, and calibration and validation. Extr action strategy incorporates selecting a proper fiber coating for the target ana lyte and selection of extraction mode (direct or headspace). Hardware selection includes selection of separation or detection technique and optimizing the desorption c onditions. Initial optimization of the method begins by determining the optimal sample volume, extraction times, and extraction conditions such as pH, salt, and tempera ture. The instrument calibration and validation is the fourth step in the method dev elopment process. Here, method
33 precision and verification of equilibrium time, sen sitivity, and linear dynamic range can be established. For this research, the fiber coating was selected f rom a prearranged list provided by Supelco. The 65 m polydimethylsiloxane/divinylbenzene fiber was chosen for its selectivity for nitroaromatics. Rea gents used include acetonitrile, methanol, and water and the extraction mode used wa s direct sampling. This study examined extraction efficiency of TNT among the thr ee solvents. 4.3 SPME Extraction Results and Discussion SPME extractions were performed 12, 50, and 100ppm solutions of TNT in water, acetonitrile, and methanol. For each sample the fi ber was directly in solution for 20minutes. Afterwards, SPME fiber was transferred to the HPLC interface and injected. Volume in the desorption chamber was 200 mL. Figures 16, 17, and 18 illustrate SPME extraction of TNT in water, acetoni trile, and methanol respectively. y = 38062x + 1E+06 R = 0.999 0 1000000 2000000 3000000 4000000 5000000 6000000 020406080100120Peak AreaTNT Concentration (ppm)SPME Extraction in Water Figure 16. TNT Concentration vs. Peak Area result s from SPME extraction in water
34 y = -235.9x + 66250 R = 0.697 0 10000 20000 30000 40000 50000 60000 70000 020406080100120Peak AreaTNT Concentration (ppm)SPME Extraction in Acetonitrile Figure 17. TNT Concentration vs. Peak Area results from SPME extraction in acetonitrile y = 623.2x -15860 R = 0.817 -20000 -10000 0 10000 20000 30000 40000 50000 60000 020406080100120Peak AreaTNT Concentration (ppm)SPME Extraction in Methanol Figure 18. TNT Concentration vs. Peak Area results from SPME extraction in methanol Water extraction showed a linear regression with a 99% goodness of fit. Both methanol and acetonitrile extractions lacked linear ity and had relatively low and
35 inconsistent peak areas compared to water extractio ns. Because of the fibers affinity towards water, SPME use proves ideal for extraction in aqueous samples.
36 Chapter 5: Conclusions Environmental contamination due to energetic compo unds such as trinitrotoluene is a growing environmental concern. Explosive residues are not only toxic to humans, but also to plants and animals. This research involves remediation techniques, characterization, and employing efficie nt analytical techniques to reduce solvent waste. Remediation systems include the use of zero-valent iron and zerovalent iron emulsions as a degradative vehicle. St udies also included characterization of trinitrotoluene in Vieques, Puerto Rico and the use of solid phase microextraction as an alternative to EPA Method 8330 soil and water extraction. Results from NZVI studies showed it has capabiliti es to degrade TNT successfully. More degradation was seen in studies using iron that was newer because of oxidation on the surface of the NZVI par ticles. EZVI studies were successful as well, and have the possibilities for degradation over longer periods of time. When used in situ, this can be an advantage for long term remediation of sediments as opposed to neat NZVI which has time li mitations. Although our results from Vieques, PR showed undet ectable amounts of TNT, had there been detectable TNT concentrations in the sediments, both NZVI/EZVI studies would have been performed as well as SPME e xtraction techniques.
37 References Agrawal, A., Tratnyek, P. 1996. Reduction of Nit ro Aromatic Compounds by ZeroValent Iron Metal. Environmental Science Technolo gy. 30, pp. 153-160 Bibette, J., Leal-Calderon, F., Schmitt, V., and Po ulin, P. 2002. Emulsion Science: Basic Principles, An Overview Verlag: Springer. pp.80-93 Boopathy, R. et al. 1998. Biotransformation of e xplosives by anaerobic consortia in liquid culture and in soil slurry. International Biodeterioration & Biodegradation 41 pp. 67-74 Brannon, J.M., and J.C. Pennington. 2002. Environme ntal fate and transport process descriptors for explosives. Technical report TR-0210. U.S. Army Corps of Engineers, Engineer Research and Development Center Vicksburg, MS. Available at http://libweb.wes.army.mil/uhtbin/hyperion/EL-TR-02 -10.pdf Burken JG. 2003. Uptake and metabolism of organic c ompounds: green-liver model. Phytoremediation: Transformation and Control of Con taminants ed. SC McCutcheon, JL Schnoor, pp. 5984. New York:Wiley Clausen, J. et al. 2004. Contaminants on military r anges: A case study of Camp Edwards, Massachusetts, USA. Environmental Pollution 129:13-21. Department of Defense. 1999. BRAC environmental fac t sheet: Unexploded Ordinance (UXO). Office of the Deputy un der Secretary of Defense(Environmental Security). Diaz, E. and Massol-Deya, A. 2003. Trace element co mposition in forage samples from a military target range, three agricul tural areas, and one natural area in Puerto Rico. Caribbean Journal of Science 3 9(2): 215-220. Geiger, C. L. ; Carvalho-Knighton, K. M. 2009. Environmental Applications of Reactive Nanoscale and Microscale Reactive Particle s ;., Eds. ACS Symposium Series American Chemical Society: Washing ton, D C,. In press. Gaurav et al. 2007. SPME-HPLC: A new approach to t he analysis of explosives Journal of Hazardous Materials 147. 691-697.
38 Hannink, N. K.; Rosser, S. J.; Bruce, N. C. 2002. Phytoremediation of Explosives. Critical Reviews in Plant Sciences 21(5):511-538. Hughes JB, Shanks J, Vanderford M, Lauritzen J, Bha dra R. 1997. Transformation of TNT by aquatic plants and plant tissue cultures. Environ. Sci. Technol. 31: 26671. Ithier-Guzmn, W., Mayo, M., Pyrtle, A. 2007. Sed iment Monitoring at Vieques National Wildlife Refuge in Puerto Rico: Assessing the Current Conditions. AGU Ocean Sciences 2007 Joint Assembly Jarvis, A. Susan et al. 1998. Assessment of the E ffectiveness of Composting for the Reduction of Toxicity and Mutagenicity of Explosive -Contaminated Soil. Ecotoxicology and Environmental Safety 39, 131-135 Lewis, T. A. et al. 2004. Bioremediation of soils contaminated with explosives. Journal of Environmental Management 70. 291-307 Mayo, M., Pyrtle, A., Ithier-Guzman, W. 2006. Se diment Quality Baseline Data for Coastal Resource Management in Two Protected Areas in Puerto Rico EOS, Transactions, American Geophysical Union. Mueller, W.F., G.W. Bedell, S. Shojaee, P.J. Jackso n. 1995. Bioremediation of TNT Wastes by Higher Plants. Pr oceedings of the 10th Annual Conference of Hazardous Waste Research. Grea t Plains/Rocky Mountain Hazardous Substance Research Center (pub). Pawliszyn, J. 1997. Solid Phase Microextraction Theory and Practice. Wiley-VCH New York. Pena-Luengas, S. et al. 2007. Development of SPMEHPLC methodology for detection of nitroexplosives Proc. SPIE 6553, 6553 1W, DOI:10.1117/12.720362 Quinn, J., Geiger, C., Clausen, C., et al. 2005. Field Demonstration of DNAPL Dehalogenation Using Emulsified Zero-Valent Iron. Environmental Science & Technology 39 (5), 1309-1318 Riley R.G.; Zachara J.M.; Wobber F.J. 1992.Chemical Contaminants on Doe Lands and Selection of Contaminant Mixtures for Subsurfac e Science Research Washington, DC: US Dep. Energy Off. Energy Res., Su bsur. Sci. Prog. Rittmann, B. E., et. Al. 1994. In Situ Bioremediat ion, second edition. Noyes Publications, Park Ridge, New Jerse y. Pg. 61-63, 205, 219220.
39 Rivera, R.; Medina, V.F.; Larson, S.L.; McCutcheon, S.C. 1998. Phytotreatment of TNT-Contaminated Groundwater. Journal of Soil Contamination 7(4):511529. Schnoor JL, Licht LA, McCutcheon SC, Wolfe NL, Carr eira LH. 1995. Phytoremediation of organic and nutrient contaminan ts. Environ. Sci. Technol. 29:318A-23A. Simini, M., Wentsel, R.S., Checkai, R.T., Phillips, C.T., Chester, N.A., Major, M.A., and Amos, J.A. 1995. Evaluation of soil toxicity at Joliet Army Ammunition Plant. Environmental Toxicology and Chemistry 14( 4): 623-630. Sjoblom, Johan. 2006. Emulsions and Emulsion Stability Second Edition CRC Press. Boca Raton, FL. pp. 263-281. Stern RJ, Moon J,Key T, Amidon T, Sickels F, et al. 1996. A pathway analysis approach for determining generic cleanup standards for radioactive materials. Draft Rep. Comment. Trenton, NJ: NJ Dep. Environ. P rotect., Bur. Environ. Radiat. Susarla, S., Bacchus, S.T., Medina, V.F., McCutcheo n, S.C., 2002. Phytoremediation: an ecological solution to organic chemical contamin ation. Ecological Engineering 18(5), 647658. UFA 1997. The unsaturated flow apparatus (UFA) meth od for extracting pore fluids from relatively impermeable and highly unsaturated porous media, Unsaturated Flow Apparatus (UFA) Ventures Inc., Ric hland, Washington (1997) pp. 17. U.S. DOD (U.S. Department of Defense). 1984. AD-A16 8637. Available from Defense Technical Center. Write to Documents, Came ron Station, Alexanderia, VA 22314, or call (703) 274-7633. US Environmental Protection Agency. 2005. Atlantic Fleet Weapons Training Area-Vieques, Puerto Rico. [online]: http://epa.gov/Region2/superfund/npl/0204694c.pdf http://www.epa.gov/superfund/sites/npl/nar1719.htm