Chemosphere 63 (2006) 1054–1059

www.elsevier.com/locate/chemosphere

Technical Note

Phytotreatment of propellant contamination

R. Guy Riefler a,*, Victor F. Medina b,1

a Department of Civil Engineering, Ohio University, Athens, 141 Stocker Center, OH 45701, USAb US Army Corps of Engineers, Engineer Research and Development Center, Vicksburg, MS 39180, USA

Received 23 February 2005; received in revised form 15 August 2005; accepted 17 August 2005Available online 9 November 2005

Abstract

Nitroglycerine (NG) and 2,4-dinitrotoluene (2,4-DNT) are propellants often found in soil and groundwater at mil-itary firing ranges. Because of the need for training with live ammunition, control or cleanup of these contaminants maybe necessary for the continued use of these firing ranges. One inexpensive approach for managing sites exposed to thesecontaminants is the use phytoremedation, particularly using common or native grasses. In this study, the uptake of NGand 2,4-DNT from water by three common grasses, yellow nutsedge (Cyperus escalantus), yellow foxtail (Setaria gla-

uca), and common rush (Juncus effusus), was investigated using hydroponic reactors. Rapid removal from solution byall grasses was observed, with yellow nutsedge removal rates being the highest. NG or 2,4-DNT accumulated in thetissues in all of the plants, except yellow foxtail did not accumulate NG. Higher concentrations were observed in killedroots, demonstrating the presence of plant-based enzymes actively transforming the contaminants. Yellow nutsedge wasalso grown in 2,4-DNT spiked soil. Significant uptake into the plants roots and leaves was observed and concentrationsin the soil decreased rapidly, although 2,4-DNT concentration also decreased in the unplanted controls. In summary,the three grasses tested appear to be good candidates for phytoremediation of propellant contamination.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Nitroglycerine; 2,4-DNT; 2,4-Dinitrotoluene; Phytoremediation; Grass

1. Introduction

The US Army maintains numerous ranges through-out the country, which allow soldiers to train with livemunitions. This training greatly improves the perfor-mance of soldiers in actual warfare, and the US Armywill continue to use live fire training in the foreseeablefuture. However, because of incomplete combustion

0045-6535/$ - see front matter � 2005 Elsevier Ltd. All rights reservdoi:10.1016/j.chemosphere.2005.08.044

* Corresponding author. Tel.: +1 740 593 1471; fax: +1 740593 0625.

E-mail addresses: riefler@ohio.edu (R.G. Riefler), victor.f.medina@us.army.mil (V.F. Medina).

1 Tel.: +1 601 634 4283; fax: +1 601 634 3518.

and detonation significant contamination of propellantsand explosives has resulted at some ranges. One area ofconcern is firing points, where artillery, rockets, andguns are fired.

Nitroglycerine (NG) and 2,4-dinitrotoluene (2,4-DNT) are two common propellants used to fire projec-tiles at these locations. These compounds have beenrelatively ignored in the past compared to other firingrange contaminants. However, recent studies have indi-cated that, at least in some cases, NG and 2,4-DNT canaccumulate in the soil at levels in the mg kg�1 range(Clausen et al., 2004). For example, field sampling atCanadian Forces Base (CFB)-Valcartier Arnhem anti-tank rocket range (firing the shoulder light antitankweapon rocket) found soil surface nitroglycerine

ed.

R.G. Riefler, V.F. Medina / Chemosphere 63 (2006) 1054–1059 1055

concentrations of nearly 2000 mg kg�1 at 5 m behind therocket firing line and over 100 mg kg�1 at 25 m behindthe firing line (Jenkins et al., 2004). A rocket firing rangeat CFB Gagetown was described having NG concentra-tions over 1% near the firing location (Jenkins et al.,2004). Similarly, studies at Yakima Training Centerfound propellant residuals at firing ranges for variousartillery weapons (Pennington et al., 2002). 2,4-DNTwas found at concentrations of 1–5 lg kg�1 at 10 mfrom a 155 mm howitzer muzzle location. Much higherconcentrations were found 10 m from tank firing points(120 mm gun), with 2,4-DNT concentrations on theorder of 50 mg kg�1 and NG concentrations about5 mg kg�1.

Both 2,4-DNT and NG have demonstrated toxiceffects to humans and the environment (Trainor andJones, 1966; Smith and Wilcox, 1994; Gong et al.,2003; Medina et al., 2003). 2,4-DNT has been listed bythe US Environmental Protection Agency as a hazard-ous substance and a probably human carcinogen. Bothtend to partition into soil (logKOW for NG = 1.62 andfor 2,4-DNT = 1.98) but are soluble at high concentra-tions (solubility for NG = 1800 mg l�1 and for 2,4-DNT = 270 mg l�1) (Brannon and Pennington, 2002).Therefore, propellant contamination at firing pointareas and the possibility of off-site migration are ofconcern.

Phytoremediation has been studied for the treatmentof high explosives, and it has been shown that plants cantake up and transform TNT (2,4,6-trinitrotol-uene), RDX (hexahydro-1,3,5-trinitro-1,3,5-triazine),and other explosives (Larson et al., 1999; Bhadraet al., 2001; Price et al., 2002; McCutcheon et al.,2003). Although the exact enzymatic mechanisms havenot been fully delineated, it appears that plants are effec-tive at removing and transforming nitro-group contain-ing organic compounds. These processes may includeboth plant degradation and the stimulation of microbialbiodegradation of the contaminants.

Studies have indicated that 2,4-DNT is biodegrad-able at rates comparable to that of TNT (Brannon andPennington, 2002). 2,4-DNT degradation may result inthe formation of amino transformation products,particularly 2-amino-4-nitrotoluene and 4-amino-2-nitrotoluene (Pennington et al., 2003). Bacteria havebeen identified that readily mineralize 2,4-DNT (Smetset al., 1999). Nitroglycerine appears to be easily biode-graded by bacteria through sequential denitration steps(Christodoulatos et al., 1997; Accashian et al., 1998).

Phytodegradation data on 2,4-DNT and NG are lim-ited, but the few studies available are promising. Todd(1996) demonstrated plant degradation of 2,4-DNTand 2,6-dinitrotoluene. McCutcheon et al. (2003) calcu-lated an average pseudo-first-order rate constant of0.5 d�1 from Todd�s data and additional unpublishedresearch. French et al. (1999) produced transgenic

tobacco plants that expressed a bacterial gene for theproduction of pentaerytritol tetranitrate (PETN) reduc-tase, and determined transgenic plants were able to ger-minate effectively in NG spiked media that was toxic tocontrol plants. Goel et al. (1997) demonstrated degrada-tion of NG by non-modified plant cells. Vanek et al.(2000) reacted NG and PETN with a tissue culture ofChinese rhubarb (Rheum palmatum). Within 10 d, theinitial concentration of 50 mg l�1 NG decreased by39%, while the same concentration of PETN was com-pletely removed.

These studies suggest that two of the most commonpropellants found at firing points, NG and 2,4-DNT,may be amenable to phytoremediation. If effective, aphytomanagement strategy for firing points wouldbe inexpensive to implement and would not hinder train-ing activities, two important considerations for activeranges. The use of common or native grasses for phyto-remediation would be desirable and enhance ecosystempreservation at the training ranges (Palazzo et al., 2003).

The objective of this project was to assess the feasibil-ity of using grasses to remove NG and 2,4-DNT fromwater and soil. Yellow nutsedge (Cyperus escalantus),yellow foxtail (Setaria glauca) and common rush (Juncus

effusus), common grasses throughout the US, were usedin this study. Initially hydroponic studies were con-ducted to determine if the plants were capable of remov-ing the contaminants from solutions and to provide datafor kinetic rate calculations. The hydroponic studieswere followed by soil degradation studies using yellownutsedge and 2,4-DNT.

2. Materials and methods

2.1. Explosives and analysis

All samples were analyzed by high-performanceliquid chromatography (HPLC) using an isocratic 50%methanol/DI eluent. NG analysis was performed on aWaters 486 HPLC (Milford, MA) with a Supelco C-l8column (St. Louis, MO) with UV detection at 215 nm.2,4-DNT analysis was performed on a Dionex SummitHPLC (Sunnyvale, CA) with a Supelco C-l8 columnand UV detection at 254 nm. Compounds were identi-fied by comparison with standard elution times. Totalorganic carbon was determined using a ShimadzuTOC-V analyzer (Columbia, MD).

2.2. Grasses

Three commonly occurring grasses were selectedfrom a range of habitats. Yellow nutsedge is found ina variety of habitats including both wetlands anduplands, yellow foxtail is an upland grass, and commonrush is a wetland grass. They also represent the three

0

2

4

6

8

10

12

14

0

2

4

6

8

10

12

14

0 20 40 60 80 100 120Time (h)

Co

nce

ntr

atio

n (

mg

l-1)

Co

nce

ntr

atio

n (

mg

l-1)

(a)

(b)

Fig. 1. Concentrations of NG (a) and 2,4-DNT (b) in hydro-ponic solution with yellow nutsedge (m), yellow foxtail (d),common rush (j), and killed roots ( ).

1056 R.G. Riefler, V.F. Medina / Chemosphere 63 (2006) 1054–1059

families typically referred to as grasses: sedges (yellownutsedge), true grasses (yellow foxtail), and rushes (com-mon rush). Yellow nutsedge plants were grown from rhi-zomes in a greenhouse for eight weeks prior to their usein the experiments. The yellow foxtail and common rushplants were collected from the grounds of the US ArmyCorps of Engineers Research and Development Center(Waterways Experiment Station) in Vicksburg, MS.All grasses were dug from large sections of soil and care-fully washed to remove soil while minimizing rootdisturbance.

2.3. Hydroponic experiments

The experimental units contained 5–15 g of the tar-geted grass at a ratio of 150 g grass l�1 of either 2,4-DNT or NG solution (10 mg l�1) in clean beakers. Tominimize evaporation, the beakers were covered withparafilm, slotted to wrap around the blades of thegrasses. The beakers were placed under grow lights usinga 12:12 light:dark photoperiod. Samples were collectedfrom the hydroponic solutions periodically for up to5 d, diluted 50% with acetonitrile, and analyzed todetermine the propellant concentrations via HPLC.Control beakers contained yellow nutsedge roots thathad been frozen overnight in a solution with sodiumazide (1 g l�1) added to inhibit bacterial activity. Alltreatments were prepared in triplicate.

2.4. Plant analyses

Leaf and root tissues were analyzed by HPLC todetermine the concentrations of NG or 2,4-DNT.Leaves and roots were separated then finely chopped.One to two grams of leaf or root tissue were added to10 ml of acetonitrile in 20 ml vials, sonicated for 8 h at15 �C, filtered through a 0.45 lm syringe filter, thenthe filtrate was analyzed via HPLC. Sub-samples of leafor root tissues were oven dried at 104 �C for 8 h, andtheir moisture contents were calculated.

2.5. Soil experiments

Soil experiments were performed using sieved local(Vicksburg, MS) silt with low organic content, 1.3% byoven analysis and 0.44% by total organic carbon ana-lyzer. Six kilograms of soil was spiked with 2,4-DNTdissolved in acetone to reach a concentration of100 mg kg�1. The soil was thoroughly mixed and keptunder a fume hood for two weeks. Yellow nutsedge(approximately 5 g plant per pot) were planted in 12small pots each containing approximately 120 g ofamended soil. Two types of controls were used in tripli-cate: (1) grass planted in clean soil and (2) pots contain-ing amended soil but no grass. The plants were placedunder grow lights, and a 12:12 light:dark photoperiod

was used. Pots were watered every other day with tapwater to field capacity. Three pots of grasses in 2,4-DNT amended soil were harvested every 5 d for analy-sis. Propellant concentrations were determined viaHPLC on soil from the root zone, leaf tissues, and roottissues. The roots were gently hand washed with tapwater to remove soil particles, then leaves and rootswere separated and finely chopped before analyses. Pro-pellants were extracted and moisture content conductedas previously described.

3. Results

NG was readily taken up by yellow nutsedge, yellowfoxtail, and common rush in hydroponic experiments(Fig. 1a). Within 3 d there was 70% removal of NG inbeakers containing the yellow nutsedge; 83% and 95%of the NG were removed within 5 d by the yellow foxtailand common rush, respectively. NG loss was fit to bothzero order and first order decay models (Table 1). Asshown by the coefficients of determination, R2, bothkinetic models provided adequate fits, with the zeroorder model being slightly better for the yellow nutsedgeand slightly worse for the common rush. There were nostatistically significant differences (a = 0.05) between therates of NG removal by grasses using either set of rate

Table 1Kinetic rate constants for NG and 2,4-DNT removal in hydroponic study

Contaminant Grass Zero order rate (l mg�1 h�1) First order rate (h�1)

Average SD R2 Average SD R2

NG Nutsedge 0.100 0.0236 0.99 0.0165 0.0071 0.94Foxtail 0.071 0.0065 0.79 0.0155 0.0032 0.80Rush 0.071 0.0086 0.71 0.0217 0.0069 0.81

2,4-DNT Nutsedge 0.194 0.0093 0.89 0.0404 0.0118 0.83Foxtail 0.051 0.0137 0.96 0.0086 0.0036 0.95Rush 0.068 0.0029 0.91 0.0159 0.0011 0.97

R.G. Riefler, V.F. Medina / Chemosphere 63 (2006) 1054–1059 1057

constants. Some NG accumulated in the leaves androots of the yellow nutsedge and common rush, withslightly higher concentration in the leaves (Fig. 2). NGdid not accumulate in yellow foxtail tissues. Total accu-mulated NG in leaves and roots only accounted for12.2 ± 1.9% of the initial NG added for the yellow nut-sedge and 4.8 ± 3.7% of the initial NG added for thecommon rush. The presence of NG in the leaves wasclear evidence that NG was taken up and translocatedto the leaves by the grasses. Substantial transformationof the NG occurred either by bacteria in solution or inthe plant tissue itself. In beakers containing yellow fox-tail plants, there was significant (a = 0.05) loss of NGfrom the solution, yet no accumulation in the tissueswas observed. Assuming the microbial cultures in theseflasks were not appreciably different than those in theother flasks of grass, it is likely that enzymes withinthe yellow foxtail tissues were especially effective attransforming NG. Trials with killed grass roots showedhigh concentrations of NG sorbing to the dead roots(Fig. 2). Based on the reported logKOW for NG of1.62, this was not surprising. NG concentrations in thedead roots were significantly higher than NG concentra-tions in the live roots, further indication that enzymeswithin all of the live grasses were actively transformingthe NG. Because of the low volatility of NG (vaporpressure = 3.4 · 10�4 N m�2 at 20 �C compared with

0

50

100

150

200

250

300

350

400

Nutsedge Foxtail Rush Killed Nutsedge Foxtail Rush Killed

Co

nce

ntr

atio

n (

mg

kg

-1)

Fig. 2. Concentrations of NG and 2,4-DNT extracted fromleaves (shaded) and roots (white) at the completion of hydro-ponic uptake experiment.

1600 N m�2 for water), evapotranspiration was consid-ered insignificant.

2,4-DNT was rapidly taken up by the yellow nut-sedge being completely removed from solution within2 d. It was taken up more slowly by the yellow foxtailand common rush with 64% and 86% removed, respec-tively in 4.6 d (Fig. 1b). 2,4-DNT loss was fit to bothzero order and first order decay models (Table 1). Asshown by the coefficients of determination, R2, bothkinetic models provided adequate fits, with the zeroorder model being slightly better for the yellow nutsedgeand slightly worse for the common rush. The rate of2,4-DNT removal by yellow nutsedge was signifi-cantly higher than the rate of 2,4-DNT removal withyellow foxtail and common rush using both sets ofkinetic parameters (a = 0.05). 2-Amino-4-nitrotolueneand 4-amino-2-nitrotoluene, commonly formed reduc-tion products of 2,4-DNT (Pennington et al., 2003),were detected at progressively increasing concentrationsin the solution, although they could not be separatedchromatographically to determine their individual con-centrations (data not shown). 2,4-DNT accumulated inthe leaf and root tissue of the yellow nutsedge, yellowfoxtail, and common rush representing 11.0 ± 0.4%,4.5 ± 1.7%, and 5.8 ± 0.8% of the initial mass added,respectively (Fig. 2). The presence of 2,4-DNT in theleaves is clear evidence that 2,4-DNT is actively takenup and translocated by the grasses. 2,4-DNT was alsodetected in roots of the killed control at a high concen-trations. This demonstrates significant abiotic adsorp-tion of 2,4-DNT to the roots, as would be expectedfrom the reported log KOW of 1.98. The rapid disappear-ance rate in the growth trials compared to the killed con-trols, the lower concentration of 2,4-DNT in live roottissues, and the presence of 2,4-DNT in leaf tissue, alldemonstrated that the grasses transported 2,4-DNT intothe leaves, and plant enzymes degraded it. Because ofthe low volatility of 2,4-DNT (vapor pressure =0.030 N m�2 at 20 �C compared with 1600 N m�2 forwater), evapotranspiration was considered insignificant.

The soil concentrations of 2,4-DNT were low-ered 84% within 5 d in pots planted with yellow nut-sedge, however similar removals were observed in the

0

20

40

60

80

100

120

140

160

180

200

0 5 10 15 20 25

Time (d)

Co

nce

ntr

atio

n (

mg

kg

-1)

Fig. 3. Concentrations of 2,4-DNT extracted from soil in potsgrown with yellow nutsedge (m), from yellow nutsedge roots(h), from yellow nutsedge leaves (s), and from soil in pots withno plant ( ).

1058 R.G. Riefler, V.F. Medina / Chemosphere 63 (2006) 1054–1059

unplanted controls (Fig. 3). Subsequently, there was lit-tle change in the following 15 d. Loss of 2,4-DNT in theunplanted controls was surprising, considering that the2,4-DNT spiked soil was tested several times over twoweeks following spiking with no significant change inconcentration from 92 mg kg�1. Apparently, wateringthe unplanted controls during the experiment stimulatedbacterial growth resulting in a decrease in 2,4-DNT con-centrations. As a result, there was no significant differ-ence (a = 0.05) between planted treatments and theunplanted controls similar to other studies using soilspiked with TNT (Sung et al., 2003; King, 2004). Sub-stantial amounts of 2,4-DNT accumulated in yellownutsedge tissue, exceeding the soil concentration by92% in the roots. The presence of 2,4-DNT in the leavesdemonstrated active uptake and translocation of thecontaminant by the grass. Interestingly, 2,4-DNT pref-erentially bioaccumulated in the roots over the leaves(7:1) during the soil experiments. The opposite trendwas observed in the hydroponic experiments (1:1.5).The 2,4-DNT concentrations in grass tissue decreasedafter the initial uptake, and plant biomass remained rel-atively constant over the three-week duration of theexperiment, further indication that plant enzymesactively transformed the contaminant.

In summary, the three common grasses were effectiveat removing NG and 2,4-DNT from hydroponic solu-tions. The uptake of propellants by grasses may be use-ful for remediating contaminated soil and reducingleachate migration. Although there was some variationamong the grass species, all three species tested wereeffective at removing NG and 2,4-DNT from solution.2,4-DNT was taken up from the soil by yellow nutsedge,although the overall benefits could not be experimentallydetermined because of the difficulty in using spiked soil.Plants show potential as a cost effective means to controlthe migration of propellants from firing points at USArmy training ranges.

Acknowledgements

This research was carried out at the US Army Engi-neer Research and Development Center (WaterwaysExperiment Station) in Vicksburg, MS with the help ofnumerous staff members. The authors particularlywould like to thank Christopher Griggs of Applied Re-search Associates, Inc. and Julie McClure of AnalyticalServices, Inc. who conducted all of the analytical work.This research was funded by US Army EnvironmentalQuality and Technology Program, Distributed SourcesWork Unit.

References

Accashian, J.V., Vinopal, R.T., Kim, B.J., Smets, B.F., 1998.Aerobic growth on nitroglycerin as the sole energy sourceby a mixed bacterial culture. Appl. Environ. Microbiol. 64,3300–3304.

Bhadra, R., Wayment, D.G., Williams, R.K., Barman, S.N.,Stone, M.B., Hughes, J.B., Shanks, J.V., 2001. Studies onplant-mediated fate of the explosives RDX and HMX.Chemosphere 44, 1259–1264.

Brannon, J.M., Pennington, J.C., 2002. Environmental fate andtransport process descriptors for explosives. EngineerResearch and Development Center (ERDC) Publications,Environmental Laboratory (EL). Technical Report ERDC/EL TR-02- 10. United States Army Corps of Engineers.Available from: <http://libweb.wes.army.mil/uhtbin/hype-rion/EL-TR-02-10.pdf>.

Christodoulatos, C., Bhaumik, S., Brodman, B.W., 1997.Anaerobic biodegradation of nitroglycerine. Water Res.31, 1462–1470.

Clausen, J., Robb, J., Curry, D., Korte, N., 2004. A case studyof contaminants on military ranges: Camp Edwards, MA,USA. Environ. Pollut. 129, 13–21.

French, C.E., Rosser, S.J., Davies, G.J., Nicklin, S., Bruce,N.C., 1999. Biodegradation of explosives by transgenicplants expressing pentaerythritol tetranitrate reductase.Nat. Biotechnol. 17, 491–494.

Goel, A., Kumar, G., Payne, G.F., Dube, S.K., 1997. Plant cellbiodegradation of a xenobiotic nitrate ester, nitroglycerine.Nat. Biotechnol. 15, 174–177.

Gong, P., Kuperman, R.G., Sunahara, G.I., 2003. Genotoxicityof 2,4- and 2,6-dinitrotoluene as measured by the Trades-cantia micronucleus (Trad-MCN) bioassay. Mutat. Res.Gen. Toxicol. Environ. 538, 13–18.

Jenkins, T.F., Ranney, T.A., Hewitt, A.D., Walsh, M.E., Bjelia,K.L., 2004. Representative sampling for energetic com-pounds at an antitank firing range. Engineer Research andDevelopment Center (ERDC) Publications, Cold RegionsResearch and Engineering Laboratory (CRREL). Tech-nical Report ERDC/CRREL TR-04-7. United StatesArmy Corps of Engineers. Available from: <http://lib-web.wes.army.mil/uhtbin/hyperion/CRREL-TR-04-7.pdf>.

King, S., 2004. Capillary electrophoresis single-strand confor-mation polymorphism analysis for monitoring bacteriaduring the remediation of TNT-contaminated soil. Ph.D.Dissertation, Ohio University.

R.G. Riefler, V.F. Medina / Chemosphere 63 (2006) 1054–1059 1059

Larson, S.L., Jones, R., Escalon, L., Parker, D., 1999.Classification of explosives transformation products inplant tissue. Environ. Toxicol. Chem. 18, 1270–1276.

McCutcheon, S.C., Medina, V.F., Larson, S.L., 2003. Proof ofphytoremediation for explosives in water and soil. In:McCutcheon, S.C., Schnoor, J.L. (Eds.), Phytoremediation:Transformation and Control of Contaminants. WileyInterscience, Hoboken, NJ, pp. 429–480.

Medina, V.F., Maestri, E., Marmiroli, M., Dietz, A.C., 2003.Plant tolerances to contaminants. In: McCutcheon, S.C.,Schnoor, J.L. (Eds.), Phytoremediation: Transformationand Control of Contaminants. Wiley Interscience, Hobo-ken, NJ, pp. 189–233.

Palazzo, A.J., Hardy, S.E., Jensen, K.B., 2003. Improved nativegrasses and establishment methods for use on militarytraining lands. Engineer Research and Development Center(ERDC) Publications, Cold Regions Research and Engi-neering Laboratory (CRREL). Technical Report ERDC/CRREL TR-03 -20. United States Army Corps of Engi-neers. Available from: <http://libweb.wes.army.mil/uhtbin/hyperion/CRREL-TR-03-20.pdf>.

Pennington, J.C., Jenkins, T.F., Ampleman, G., Thiboutot, S.,Brannon, J.M., Lynch, J., Ranney, T.A., Stark, J.A., Walsh,M.E., Lewis, J., Hayes, C.A., Mirecki, J.E., Hewitt, A.D.,Perron, N., Lambert, D., Clausen, J., Delfino, J.J., 2002.Distribution and fate of energetics on DOD test and trainingranges: Interim Report 2. Engineer Research and Develop-ment Center (ERDC) Publications. Technical ReportERDC TR-02-8. United States Army Corps of Engineers.Available from: <http://libweb.wes.army.mil/uhtbin/cgis-irsi/X8JkpbRbX5/ERDC_VBG/246280018/526/+6886>.

Pennington, J.C., Thorn, K.A., Hayes, C.A., Porter, B.E.,Kennedy, K.R., 2003. Immobilization of 2,4- and 2,6-dinitrotoluenes in soils and compost. Engineer Researchand Development Center (ERDC) Publications, Environ-mental Laboratory (EL). Technical Report ERDC/EL TR-03-2. United States Army Corps of Engineers. Availablefrom: <http://libweb.wes.army.mil/uhtbin/hyperion/EL-TR-03-2.pdf>.

Price, R.A., Pennington, J.C., Larson, S.L., Neumann, D.,Hayes, C.A., 2002. Uptake of RDX and TNT by agronomicplants. Soil Sediment Contam. 11, 307–326.

Smets, B.F., Riefler, R.G., Lendenmann, U., Spain, J.C., 1999.Kinetic analysis of dinitrotoluene biodegradation in anaerobic fluidized bed biofilm reactor. Biotechnol. Bioeng.63, 642–653.

Smith, R.P., Wilcox, D.E., 1994. Toxicology of selected nitricoxide-donating xenobiotics, with particular reference toazide. Crit. Rev. Toxicol. 24, 355–377.

Sung, K.S., Munster, C.L., Rhykerd, R., Drew, M.C., Corap-cioglu, M.Y., 2003. The use of vegetation to remediate soilfreshly contaminated by recalcitrant contaminants. WaterRes. 37, 2408–2418.

Todd, S., 1996. Phytoremediation of dinitrotoluene contami-nated soils using parrotfeather (Myriophyllum aquaticum).M.S. Thesis. Civil Engineering, Auburn University,Alabama.

Trainor, D.C., Jones, R.C., 1966. Headaches in explosivemagazine workers. Arch. Environ. Health. 12, 231–234.

Vanek, T., Nepovim, A., Zeman, S., 2000. Phytoremediation ofselected explosives in a model system of plant tissuecultures. Korean J. Plant Tiss. Cult. 27, 379–389.