Research: Persistent Explosives Present a Problem

Introduction
During winter and spring of 2008 I had the opportunity to participate in the Women in Science Project (WISP) internship program for first-year students. As an intern in the Biochemical Sciences branch at the Cold Regions Research and Engineering Laboratory (CRREL), I conducted research in the laboratory of Dave Ringelberg, sponsored by Jay Clausen. In my research I used state-of-the-art equipment to study contaminate remediation and microorganisms. This analysis stems from an ongoing comprehensive study of the soil mobility of nitroglycerin and 2,6-dinitrotoluene being conducted by a team of researchers at CRREL.

Soil samples were collected from Camp Edwards, Massachusetts, where soldiers train by firing live rounds. New small arms ammunition is typically loaded with a double-based propellant consisting of 84% nitrocellulose (NC), 10% nitroglycerin (NG), and 6% filler compounds (1). Since the propellant does not completely combust when the arms are fired, the initial soil samples were analyzed for the presence of various by-products, such as flash suppressor dinitrotoluene (DNT), which are released into the soil when weapons are fired. In these soils, a high level of NG, 42 mg/kg, was measured (1). The literature suggests that NG typically has a half-life of one to two days. Training at Camp Edwards ceased two years ago. There is still a large quantity of NG left in the soil despite ample time for natural degradation. Thus, the objective of this study is to investigate whether or not certain biological and chemical factors are causing NG to persist in this soil. This is an important issue to resolve, since there is potential for NG to reach ground water.

Indigenous Microbial Activity
The first phase of the analysis was to determine whether there was sufficient microbial activity in the soil. An insufficient indigenous bacterial community could limit biodegradation of NG and 2,6-DNT. A microbial profile of the soil samples was first determined using a respirometer that measures CO2 produced by bacterial metabolism. Five sets of samples of 15 g of air-dried NG- and DNT-contaminated soil samples were analyzed in triplicate, yielding 15 total samples. The five different preparations would compare soil moisture at two different levels (natural moist soil and wet saturated soil), a comparison of a soil with bacteria (live) and without bacteria (kill), and the effect of adding a nutrient source (acetate). The five sample sets were wet-live, wet-kill, wet-acetate, moist-live, and moist-kill. For the wet-live preparation, 30 mL rainwater collected at the same site in March 2007 was added. Thirty mL of a biocide solution of 1% glutareldehyde and 90 mM mercuric chloride was added to the wet-kill set. The wet-acetate set received 30 mL of 50 mM acetate. The added acetate provides excess carbon for bacterial respiration to stimulate growth of carbon-limited bacteria. The moist-live set were given 1.5 mL rainwater, while 1.5 mL of a biocide solution of 1% glutareldehyde and 90 mM mercuric chloride was added to the moist-kill set.

An analysis of microbial respiration indicated that bacterial growth was occurring in the soil. The soil displayed typical exponential growth and approached stationary phase after four days (96 hours). The growth of bacteria was suggested by a notable increase in CO2 and a decrease in O2 levels, indicating that cell respiration was occurring. The acetate treatment enhanced bacterial growth, and bacterial communities flourished much more in wet conditions rather than moist conditions. The results of the bacterial profile display that the soil harbors active bacterial communities under natural conditions. This data ruled out the possibility that a deficient bacterial population was the restricting factor for NG and 2,6-DNT biodegradation in Camp Edwards soil.

Exploring the Biodegradation Potential
In the next phase of the study, a set of samples was examined over a period of one week for NG degradation. Thirty soil slurry flasks were set up with 5 g soil and 20 ml solution. Three live and three kill samples were analyzed at five time intervals: days 0, 1, 2, 5, and 7. Three samples from each set (live and kill) were centrifuged at each time interval to separate the solid and aqueous phases. The solid phase was processed on a sonicator to release materials from the soil surface. All samples were extracted with acetonitrile for processing through high-pressure liquid chromatography (HPLC) and then measured for NG and 2,6-DNT.

The second experimental study (Figure 2) yielded interesting results indicating that the presence of bacterial communities in the live samples were not involved in the biodegradation of NG and DNT. There was no significant difference in degradation rates between the kill and live samples even though the microbial profile study suggested the presence of a healthy microbial community in the live soils, and the absence of microbes in the killed soils. Surprisingly, there was still a large quantity of NG present in the soil samples after a week. There was virtually no NG or 2,6-DNT in the aqueous state.

Future Investigation
Future studies of these soil samples will reveal other mechanisms for NG conservation in field soil. Bacteria can only biodegrade NG in aqueous solution, so something may be preventing the NG from dissolving; there may be an issue relating to a lack of bioavailability because NG is not dissolving into the aqueous phase. One possibility is encapsulation by nitrocellulose. If this is the case, grinding or chemically processing the soil may release NG from within the nitrocellulose for accessibility by the bacteria. Moreover, certain microbes, such as the fungi Sclerotium rolfsii and bacterium Fusarium solani, are known to successfully break down nitrocellulose, and thus could be used as a treatment in these soils (2). Introduction of such bacteria could allow for NG biodegradation returning to normal rates and resulting in elimination from the Camp Edwards site.

Acknowledgements
I would like to thank my sponsor Jay Clausen as well as David Ringelberg and Karen Foley for their assistance at CRREL. Thank you to the Women in Science Project for providing this internship.

References
1. D. Margolis, I. Osgerby, J. Macpherson, and J. Clausen, Site Specific Sorption/Desorption Measurements for Nitroglycerin and Dinitrotoluene. 1, 7 (2007).
2. A. Sharma, S.T. Sundaram, Y. Zhang, and B.W. Brodman, Journal of Industrial Microbiology and Biotechnology. 1 (1984).