laboratory and pilot-scale bioremediation of pentaerythritol tetranitrate (petn) contaminated soil

8
Journal of Hazardous Materials 264 (2014) 261–268 Contents lists available at ScienceDirect Journal of Hazardous Materials jou rn al hom epage: www.elsevier.com/locate/jhazmat Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil Li Zhuang a,, Lai Gui b,c , Robert W. Gillham b , Richard C. Landis d a Guangdong Institute of Eco-environmental and Soil Sciences, Guangzhou 510650, China b Department of Earth and Environmental Sciences, University of Waterloo, Waterloo, Ontario N2L 3G1, Canada c Pest Management and Regulatory Agency, Health Canada, Ottawa, Ontario K1A 0K9, Canada d E. I. du Pont de Nemours and Company, Wilmington, DE 19880, USA h i g h l i g h t s Laboratory and pilot-scale bioremediation of PETN-contaminated soil. PETN-contaminated soil treated with granular iron and organic carbon materials. The amendment of carbon source effectively stimulated PETN biodegradation. The low efficiency of iron method was caused by the high concentration of nitrate. a r t i c l e i n f o Article history: Received 4 September 2013 Received in revised form 31 October 2013 Accepted 14 November 2013 Available online 22 November 2013 Keywords: Pentaerythritol tetranitrate Explosive Bioremediation Organic carbon source Granular iron a b s t r a c t PETN (pentaerythritol tetranitrate), a munitions constituent, is commonly encountered in munitions- contaminated soils, and pose a serious threat to aquatic organisms. This study investigated anaerobic remediation of PETN-contaminated soil at a site near Denver Colorado. Both granular iron and organic carbon amendments were used in both laboratory and pilot-scale tests. The laboratory results showed that, with various organic carbon amendments, PETN at initial concentrations of between 4500 and 5000 mg/kg was effectively removed within 84 days. In the field trial, after a test period of 446 days, PETN mass removal of up to 53,071 mg/kg of PETN (80%) was achieved with an organic carbon amendment (DARAMEND) of 4% by weight. In previous laboratory studies, granular iron has shown to be highly effective in degrading PETN. However, for both the laboratory and pilot-scale tests, granular iron was proven to be ineffective. This was a consequence of passivation of the iron surfaces caused by the very high concentrations of nitrate in the contaminated soil. This study indicated that low concentration of organic carbon was a key factor limiting bioremediation of PETN in the contaminated soil. Furthermore, the addition of organic carbon amendments such as the DARAMEND materials or brewers grain, proved to be highly effective in stimulating the biodegradation of PETN and could provide the basis for full-scale remediation of PETN-contaminated sites. © 2013 Elsevier B.V. All rights reserved. 1. Introduction Intensive military activities over the past century have resulted in widespread contamination of soil and water with residues of explosives and related compounds [1]. It is currently estimated that explosives-contaminated sites could occupy millions of hectares of land in the U.S., while the global extent of contamination is difficult to assess [2]. Explosives are anthropogenic nitro-organic Corresponding author at: Guangdong Institute of Eco-Environmental and Soil Sciences, No. 808 Tianyuan Road, Guangzhou 510650, China. Tel.: +86 20 87025872; fax: +86 20 87025872. E-mail address: [email protected] (L. Zhuang). compounds with very few naturally occurring analogs. As a conse- quence, natural microbial populations are generally not acclimated and thus explosives tend to be persistent in the environment. The conventional method for remediation of explosives-contaminated soil is incineration. Because of the high cost and the associated dis- advantages such as production of large volumes of unusable ash, increased attention has turned to alternative remediation methods, such as chemical reduction with metallic iron [3–7] and bioreme- diation including composting, bioslurry and landfarming [8]. Granular iron containing a zero-valent iron core has been shown to be an effective reductant for 2,4,6-trinitroluene (TNT) and hexahydro-1,3,5-trinitro-1,3,5 triazine (RDX) [3–5], both of which are nitrated munitions. In a field trial involving 70 kg of soil from a munitions wastewater disposal site, using iron (5%, w/w), RDX 0304-3894/$ see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.jhazmat.2013.11.035

Upload: richard-c

Post on 21-Dec-2016

216 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

LT

La

b

c

d

h

••••

a

ARRAA

KPEBOG

1

ieeod

Sf

0h

Journal of Hazardous Materials 264 (2014) 261– 268

Contents lists available at ScienceDirect

Journal of Hazardous Materials

jou rn al hom epage: www.elsev ier .com/ locate / jhazmat

aboratory and pilot-scale bioremediation of Pentaerythritoletranitrate (PETN) contaminated soil

i Zhuanga,∗, Lai Guib,c, Robert W. Gillhamb, Richard C. Landisd

Guangdong Institute of Eco-environmental and Soil Sciences, Guangzhou 510650, ChinaDepartment of Earth and Environmental Sciences, University of Waterloo, Waterloo, Ontario N2L 3G1, CanadaPest Management and Regulatory Agency, Health Canada, Ottawa, Ontario K1A 0K9, CanadaE. I. du Pont de Nemours and Company, Wilmington, DE 19880, USA

i g h l i g h t s

Laboratory and pilot-scale bioremediation of PETN-contaminated soil.PETN-contaminated soil treated with granular iron and organic carbon materials.The amendment of carbon source effectively stimulated PETN biodegradation.The low efficiency of iron method was caused by the high concentration of nitrate.

r t i c l e i n f o

rticle history:eceived 4 September 2013eceived in revised form 31 October 2013ccepted 14 November 2013vailable online 22 November 2013

eywords:entaerythritol tetranitratexplosiveioremediation

a b s t r a c t

PETN (pentaerythritol tetranitrate), a munitions constituent, is commonly encountered in munitions-contaminated soils, and pose a serious threat to aquatic organisms. This study investigated anaerobicremediation of PETN-contaminated soil at a site near Denver Colorado. Both granular iron and organiccarbon amendments were used in both laboratory and pilot-scale tests. The laboratory results showedthat, with various organic carbon amendments, PETN at initial concentrations of between 4500 and5000 mg/kg was effectively removed within 84 days. In the field trial, after a test period of 446 days, PETNmass removal of up to 53,071 mg/kg of PETN (80%) was achieved with an organic carbon amendment(DARAMEND) of 4% by weight. In previous laboratory studies, granular iron has shown to be highlyeffective in degrading PETN. However, for both the laboratory and pilot-scale tests, granular iron was

rganic carbon sourceranular iron

proven to be ineffective. This was a consequence of passivation of the iron surfaces caused by the veryhigh concentrations of nitrate in the contaminated soil. This study indicated that low concentration oforganic carbon was a key factor limiting bioremediation of PETN in the contaminated soil. Furthermore,the addition of organic carbon amendments such as the DARAMEND materials or brewers grain, provedto be highly effective in stimulating the biodegradation of PETN and could provide the basis for full-scaleremediation of PETN-contaminated sites.

. Introduction

Intensive military activities over the past century have resultedn widespread contamination of soil and water with residues ofxplosives and related compounds [1]. It is currently estimated that

xplosives-contaminated sites could occupy millions of hectaresf land in the U.S., while the global extent of contamination isifficult to assess [2]. Explosives are anthropogenic nitro-organic

∗ Corresponding author at: Guangdong Institute of Eco-Environmental and Soilciences, No. 808 Tianyuan Road, Guangzhou 510650, China. Tel.: +86 20 87025872;ax: +86 20 87025872.

E-mail address: [email protected] (L. Zhuang).

304-3894/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jhazmat.2013.11.035

© 2013 Elsevier B.V. All rights reserved.

compounds with very few naturally occurring analogs. As a conse-quence, natural microbial populations are generally not acclimatedand thus explosives tend to be persistent in the environment. Theconventional method for remediation of explosives-contaminatedsoil is incineration. Because of the high cost and the associated dis-advantages such as production of large volumes of unusable ash,increased attention has turned to alternative remediation methods,such as chemical reduction with metallic iron [3–7] and bioreme-diation including composting, bioslurry and landfarming [8].

Granular iron containing a zero-valent iron core has been

shown to be an effective reductant for 2,4,6-trinitroluene (TNT) andhexahydro-1,3,5-trinitro-1,3,5 triazine (RDX) [3–5], both of whichare nitrated munitions. In a field trial involving 70 kg of soil froma munitions wastewater disposal site, using iron (5%, w/w), RDX
Page 2: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

2 rdous

d5

tclpnaalA(g

sogBcescm

fRtprlfutP

ohPurcpa

2

2

atio1ic2ua1

trf

62 L. Zhuang et al. / Journal of Haza

ecreased from an average initial concentration of 12,100 mg/kg to40 mg/kg, a 96% removal, within 120 days [6].

Explosive compounds have also been shown to be suscep-ible to biodegradation, with acclimated indigenous microbialommunities which have been exposed to contaminants for aong period of time showing the greatest promise. For exam-le, three species of the family Enterobacteriaceae, isolated fromitramines explosives-contaminated soil, can reduce both RDXnd octahydro-1,3,5,7-tetranitro-1,3,5,7 tetrazocine (HMX) undernaerobic conditions [9]. Similarly, Marshall and White [10] iso-ated four anaerobic bacterial species of Pseudomonas putida,rthrobacter, Klebsiella and Rhodococcus from glycerol trinitrateGTN)-contaminated soil, all of which can use GTN as the sole nitro-en source.

The electron-withdrawing character of nitro groups in explo-ives is responsible for their low susceptibility to typical advancedxidative catabolism [11]; therefore anaerobic bacteria holdreater potential for remediation of explosives-contaminated sites.ecause of the low solubility of explosive compounds, an externalo-substrate is commonly added to stimulate the growth of thexplosive-degrading bacteria. For example, explosive compoundsuch as TNT, HMX and tetryl can all be effectively removed fromontaminated soil by indigenous bacteria with the addition ofolasses as a carbon source for bacterial growth [12–15].To date, most studies on the degradation of explosives have

ocused on nitroaromatics and nitramines, particularly on TNT andDX, but very little work has been reported on pentaerythritoletranitrate (PETN), a nitrate ester compound. From laboratory testserformed with aqueous solutions, we previously reported [16]apid degradation of PETN in the presence of granular iron. Half-ives were on the order of minutes, and the process appeared toollow sequential denitration. In a further laboratory study [17],sing an anaerobic consortium from a contaminated site and withhe addition of an organic carbon amendment, biodegradation ofETN was shown to be an effective process.

Our previous work, using laboratory-synthesized PETN in aque-us solution, showed that granular iron and biodegradation bothad the potential to be effective technologies for remediation ofETN-contaminated sites. The goal of this study was to further eval-ate the potential and possible limitations of both technologies foremediation of PETN present in solid and aqueous forms at heavilyontaminated sites. In particular, both laboratory and small-scaleilot tests were performed using contaminated soil materials from

particular site located near Denver Colorado.

. Materials and methods

.1. PETN-contaminated soil

The contaminated site considered in this study is locatedpproximately 40 km south of Denver, Colorado. The site consists ofwo settling ponds near a currently inactive explosive manufactur-ng facility, which received wastewater from PETN production forver 20 years (between 1967 and 1989). The ponds, each 119 m by19 m, are lined with high-density PVC membrane with approx-

mately 22–36 cm of clayey soil on top of the liner. The PETNoncentration in the soil is highly variable, ranging from 1 to00,000 mg/kg of dry soil. Because nitric acid and sulfuric acid weresed in the process of PETN synthesis, concentrations of nitratend sulfate are also high and generally variable, ranging from 1 to0,000 mg/kg of dry soil.

The soil used in the laboratory experiments was obtained fromhe southwest corner of the south pond. The PETN concentrationanged from 65 to 600 mg/kg and high levels of nitrate and sul-ate (8000–10,000 mg/kg) were also present. For the laboratory

Materials 264 (2014) 261– 268

tests, the PETN concentration was increased to 4500–5000 mg/kgby spiking with pure PETN powder. Prior to PETN addition, a portionof the soil was leached several times with Millipore water to reducethe levels of nitrate and sulfate. On average, the concentrations ofnitrate and sulfate were decreased to 1500 and 2500 mg/kg, respec-tively. The soil was air dried and ground to pass a 2 mm sieve beforeuse. Soil used in the sterile controls was triple-autoclaved (1 h at121 ◦C on three consecutive days). The soil had a total organic car-bon content of 0.41%, including PETN.

2.2. Iron and organic materials

The granular iron was obtained from Connelly-GPM Inc.(Chicago, Illinois) and used without pretreatment. The iron mate-rial was characterized previously as containing 89.8% metalliciron with a surface layer of various forms of iron oxide (dataprovided by Connelly-GPM Inc.). The specific surface area ofthe iron used in this study was 1.02 m2/g, measured by theBET (Brunauer–Emmett–Teller) method. The sources of organiccarbon used in the laboratory tests included two different DARA-MEND materials, D6390Fe20 and ADM-298500. The materials wereprovided by ADVENTUS Remediation Technologies (Mississauga,Ontario, Canada). The DARAMEND products are manufactured fromplant materials rich in carbon and nutrients, though the precisecomposition is proprietary. D6390Fe20 and ADM-298500 are iden-tical except that D6390Fe20 is fortified with 20% (by weight) finegranular iron. Brewers grain, a residual of the brewing process wastested as an alternative carbon source in the field pilot tests. Thismaterial was obtained from Teague Diversified, Inc., Ft. Morgan, CO.The brewers grain had a moisture content of 65% and is fibrous witha high protein content.

2.3. Laboratory batch experiment

A total of 12 treatments were conducted; set-1, set-2 and set-7were control samples with no amendment added; set-3 to set-6contained different percentages of granular iron ranging from 2 to10% (by weight); and set-8 to set-12 tested the two types of organicmaterials (D6390Fe20 and ADM-298500) at 1% and 2% levels. Thecomposition of each treatment is given in Table 1. Each treatmentinvolved the same laboratory conditions with identical set-up andsampling and analysis procedures. Tests were conducted in 40 mLglass vials with screw caps fitted with Teflon-lined septa. Each vialcontained 15 g of PETN-contaminated soil and the desired amountof a particular amendment (iron or DARAMEND materials), and wasfilled to the top with deoxygenated Millipore water before trans-ferred to an anaerobic glovebox. A headspace was created once inthe glovebox (5% H2 + 5% CO2 + 90% N2) by removing 10 mL of water.The vials were vortexed for one minute and then incubated in thedark at room temperature (25 ◦C). Triplicate vials were sacrificedfor analysis. Before analysis, the vials were centrifuged for 15 minat 1500 rpm. The aqueous solution was removed for inorganic anal-ysis including nitrite, nitrate and sulfate. The soil was analyzed forPETN following the acetonitrile-sonication extraction method (USEPA Method 8330).

2.4. Field pilot test

The pilot test consisted of the following 10 treatments (% bydry weight): a control (no amendment), 10% iron, 1%, 2% and 4% ofD6390Fe20, 2% and 4% of ADM-298500, and 1%, 2% and 4% of brew-

ers grain. Because the dry weights of soil were not known precisely,the final percentages of the amendments in the treatments weresomewhat different from the objective values, at 1.19%, 2.63% and4.37% for D6390Fe20, 2.32% and 5.81% for ADM-298500, and 1.33%,
Page 3: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

L. Zhuang et al. / Journal of Hazardous Materials 264 (2014) 261– 268 263

Table 1Composition of the anaerobic microcosms of laboratory batch experiment.

Method Treatment No. Soil/Pretreatment Amendment

Iron tests 1 leached soil None2 Leached soil, autoclaved, None3 Leached autoclaved soil 10% Connelly iron4 Leached soil 10% Connelly iron5 Leached soil 5% Connelly iron6 Leached soil 2% Connelly iron

Biodegradation tests 7 Original soil None8 Autoclaved, leached soil 2% ADM-2985009 Leached soil 1% D6390Fe20

2c

fotec6rcmt

lacwpstM3n

2

uwAwwf

10

11

12

.99% and 6.38% for brewers grain. For convenience, the text willontinue to use the objective concentrations (1%, 2% and 4%).

For each treatment, contaminated sediment was first collectedrom the pond, and then thoroughly mixed with the desired amountf amendment material in a cement mixer. The mixture was thenransferred to a plastic tub (45 cm wide × 90 cm long × 25 cm deep)quipped with a float and valve to control the water level. Each tubontained an approximately 18 cm layer of sediment material with–8 cm of standing water. The purpose of the water layer was toeduce oxygen invasion and thus assist in maintaining anaerobiconditions within the sediment. The tubs were buried in the sedi-ent in the settling pond and the floats were connected by a hose

o a nearby water source (tap water) (Fig. 1).At each sampling time, triplicate samples of sediment were col-

ected at random locations in each tub at depths between 0–10 cmnd 10–18 cm using a soil probe. The samples from each depth wereombined, subsampled in duplicate, and placed in plastic bottlesith screw caps. Samples were stored in a freezer prior to ship-ing to the University of Waterloo. In the laboratory, each sedimentample was first air-dried in a fume hood to a constant weight, andhen PETN was extracted for analysis in accordance with the US EPA

ethod 8330. The remaining soil (air-dried soil passed through a0 mesh sieve) was used for inorganic analyses including nitrite,itrate and sulfate.

.5. Chemical analyses

Concentrations of nitrate, nitrite and sulfate were analyzedsing a Dionex ion chromatograph (Dionex ICS 2000) equippedith a conductivity detector, an Ion-Eluent Generator and a Dionex

S-40 autosampler. A Dionex IonPac AS18 column (4 × 250 mm)as used. The injection volume was 25 �L and the mobile phaseas 30 mM KOH at a flow rate of 1.2 mL/min. The detection limit

or nitrate, nitrite and sulfate was 0.5 mg/L.

Fig. 1. Configuration of in situ pilot test at Colorado,

Leached soil 2% D6390Fe20Original soil 2% D6390Fe20Original soil 2% ADM-298500

Analyses for PETN and intermediate products were performedusing a series 1100 Hewlett-Packard high performance liquid chro-matograph (HPLC) equipped with a UV visible diode array detector.A Zorbax SB-C18 column (3.5 �m particles, 4.6 × 150 mm) and aZorbax guard column (5 �m particles, 4.6 × 12.5 mm) were used.Aqueous samples were centrifuged in 1.5 mL HPLC vials for 5 minat 10,000 rpm prior to being loaded onto the atuosampler. A water-methanol-acetonitrile mixture (40:50:10, v/v/v) was used as themobile phase at a flow rate of 1.0 mL/min. The injection volumewas 100 �L and the absorbance was measured at a wavelength of210 nm. The detection limit for PETN was 0.1 mg/L. A series of PETNexternal standards was prepared using an analytical standard andanalyzed with the samples for PETN quantification. Intermediatedegradation products were identified as described elsewhere [16].Because standards for the intermediates were not commerciallyavailable, they were not quantified.

3. Results and discussion

3.1. Laboratory results: iron treatments

The iron tests were all conducted using the leached soil. Thechanges in PETN concentrations over the 93-day incubation periodare shown in Fig. 2a. Reductions in PETN concentrations of approx-imately 17%, 19% and 26% were observed in the samples containing2%, 5% and 10% iron, respectively; while little or no PETN removalwas observed in the two controls (soil without iron amendmentand autoclaved soil without iron). In the third control (autoclavedsoil amended with 10% iron), the concentration of PETN declinedby 19% during the incubation. These results suggest that granular

iron is capable of degrading PETN in the contaminated sediments,and that the contribution of microbial activity to the decline in PENTconcentration was insignificant. However, while the degree of PETNremoval in the soil was in the order of increasing iron content, it was

U.S.A.: (a) individual set-up and (b) field view.

Page 4: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

264 L. Zhuang et al. / Journal of Hazardous Materials 264 (2014) 261– 268

0.6

0.7

0.7

0.8

0.8

0.9

0.9

1.0

1.0

1.1

0 10 20 30 40 50 60 70 80 90 100

Time (d)

PE

TN

Co

nc

. (C

/C0

)

autoclaved, leached soil leached soil

autoclaved, leached soil, 10% iron leached soil, 2% iron

leached soil, 5% iron leached soil,10% iron

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

0 10 20 30 40 50 60 70 80 90 100

Time (d)

Nit

rate

Co

nc

. (C

/C0

)

a

b

Fo

nrfitp

iitmit

rmcotioct[tnl

smlciw

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 10 20 30 40 50 60 70 80 90 100

Time (d)

Nit

rate

Co

nc. (m

g/k

g)

leached soil

original soil

leached soil, 1% D6390Fe20

leached soil, 2% D6390Fe20

original soil, 2% D6390Fe20

original soil, 2% ADM-298500

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

0 10 20 30 40 50 60 70 80 90 100 110 120

Time (d)

Su

lfate

Co

nc. (m

g/k

g)

a

b

ig. 2. Changes in PETN (a) and nitrate (b) concentration over time in the presencef iron at 0, 2, 5 and 10% level in the laboratory soil microcosm experiments.

ot proportional. Furthermore, the rate of PETN removal declinedapidly over time, with most of the removal occurring within therst 20 days of the experiment. These results are in sharp contrasto the very rapid rates of removal in aqueous solution as reportedreviously [16].

Fig. 2b shows substantial declines in the nitrate concentrationsn all samples with the exception of those that had been autoclaved,ndicating that nitrate reducing bacteria were highly active in allests. In the autoclaved tests, nitrate reduction, though present, was

uch less, on the order of 10–20% reduction. In the presence of 10%ron in the autoclaved test most of the reduction occurred withinhe first 10 days.

Previous research has shown that although granular iron caneduce nitrate to ammonia [18,19], the reaction results in the for-ation of iron oxide films on the iron particle surfaces [20,21]

ausing passivation of the iron material. Thus the declining ratesf PETN removal shown in Fig. 2a and the incomplete removal overhe 93-day incubation period are believed to be a consequence ofron passivation caused by the formation of surface films as a resultf nitrate reduction. As a consequence of the PETN synthesis pro-ess, there are very high concentrations of nitrate and sulfate inhe settling ponds. Although, the initial tests in aqueous solution16] suggested that granular iron could have considerable poten-ial in the remediation of PETN-contaminated sediments, the highitrate in the sediments of the site in Colorado could pose a serious

imitation.Granular iron may however be effective for PETN-contaminated

oil if nitrate or other competing oxidants are not present. Further-ore, Lu [22] showed that the change in the iron surface and the

oss of iron reactivity due to nitrate oxidation is a reversible pro-

ess. Thus, given sufficient time for the nitrate to be depleted, theres reason to expect that more effective degradation of PETN by iron

ould proceed.

Fig. 3. Changes in nitrate (a) and sulfate (b) concentration over time in the labora-tory soil microcosms with carbon amendments (D6390Fe20 and ADM-298500).

3.2. Laboratory results: enhanced biodegradation by carbonamendment

Tests of PETN biodegradation with amendment of organic mate-rials involved 6 treatments (set-7 to set-12), as described in Table 1.The changes in concentrations of nitrate and sulfate in the treat-ments are shown in Fig. 3a and b, respectively. Nitrate removalin the control samples was very slow and incomplete, with 894 of1058 mg/kg removed in the leached soil, and 1341 of 9296 mg/kg inthe unleached soil by the end of the 93-d incubation period. In con-trast, nitrate concentrations in the treatments with the DARAMENDmaterials decreased to below the detection limit (0.5 mg/L) withinthe first 5 days, in both the leached and unleached soil samples. Thiswas consistent with the period of gas production observed in thesample vials. Clearly, in the controls, there was insufficient organiccarbon available to support high rates of anaerobic bioactivity.There was no decrease in the sulfate concentration in the controlswithout carbon amendment. This is not surprising because the car-bon source in the soil was not sufficient to support the energeticallyfavorable reaction of nitrate reduction and thus would not be likelyto support the thermodynamically less favorable reaction of sul-fate reduction. In contrast, following removal of nitrate, appreciabledeclines in sulfate concentration were observed in the treatmentsamended with the DARAMEND materials. As shown in Fig. 3b, anunexpected rise in sulfate concentration occurred over early timefor all treatments. A supplemental experiment confirmed that thiswas a consequence of the dissolution of sulfate-containing mineralspresent in the soil (data not shown).

PETN concentrations in the organic-amended treatments areshown in Fig. 4. Rapid and consistent rates of PETN removal are

shown for all tests receiving DERAMEND materials without auto-claving, while little or no loss was observed in the controls. Theautoclaved control that received 2% ADM-298500 is an exception,
Page 5: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

L. Zhuang et al. / Journal of Hazardous Materials 264 (2014) 261– 268 265

0.0

0.2

0.4

0.6

0.8

1.0

0 10 20 30 40 50 60 70 80 90 100

Time (d)

PE

TN

Co

nc. (C

/C0)

autoclaved, leached soil

leached soil

original soil

autoclaved, leached soil, 2% ADM-298500

leached soil, 1% D6390Fe20

leached soil, 2% D6390Fe20

original soil, 2% D6390Fe20

original soil, 2% ADM-298500

Fig. 4. Changes in PETN concentration over time in the laboratory soil microcosmsw

setpt

trMdccdcsTmkrmsw22rtdw

esrtob[sccrnwIunn

val i

n

the

pil

ot-s

cale

trea

tmen

ts

amen

ded

wit

h

orga

nic

mat

eria

ls

and

gran

ula

r

iron

.

Am

end

men

ts

Init

ial c

once

ntr

atio

n

(mg/

kg)

74-d

incu

bati

on

per

iod

446-

d

incu

bati

on

per

iod

Nit

rate

Sulf

ate

PETN

PETN

Rem

oval

(%)

PETN

Mas

s

Rem

oved

(mg/

kg)

PETN

Rem

oval

(%)

PETN

Mas

s

Rem

oved

(mg/

kg)

Surf

ace

Bot

tom

Ave

r.

Surf

ace

Bot

tom

Ave

r.

Surf

ace

Bot

tom

Ave

r.

Surf

ace

Bot

tom

Ave

r.

Con

trol

(non

e)

223

1926

147,

835

9

32

20

1303

1

4755

2

3029

2

20

23

22

2982

4

3387

4

3184

910

%

Con

nel

ly

iron

172

1019

84,8

20

15

−0.0

2

7

1288

8

−178

9

5550

41

32

36

3455

0

2674

8

3064

91%

D63

90Fe

20

114

898

62,8

37

26

27

26

1611

1

1672

4

1641

7

67

50

58

4181

8

3149

7

3665

82%

D63

90Fe

20

130

1232

80,5

91

28

25

27

2289

8

1987

9

2138

9

62

51

56

5008

4

4085

9

4547

24%

D63

90Fe

20

254

1225

66,4

38

52

31

42

3465

5

2063

4

2764

4

96

64

80

6383

3

4230

9

5307

12%

AD

M29

8500

118

1452

111,

044

15

11

13

1620

9

1249

2

1435

0

53

53

53

5871

4

5894

2

5882

84%

AD

M29

8500

218

1622

121,

808

25

9

17

3083

0

1094

0

2088

5

69

35

52

8465

5

4259

4

6362

41%

brew

ers

grai

n

107

1741

130,

602

11

15

16

1468

7

1908

9

1688

8

43

37

40

5598

8

4769

4

5184

12%

brew

ers

grai

n

89

1890

108,

177

18

3

10

1954

0

2851

1119

6

46

32

39

4985

4

3410

0

4197

74%

brew

ers

grai

n

180

2399

167,

407

26

14

20

4319

8

2359

7

3339

8

68

37

53

1146

39

6212

7

8838

3

ith carbon amendments (D6390Fe20 and ADM-298500).

howing about a 20% loss of PETN at the conclusion of the test. How-ver, in this test no losses were observed during the first 35 days andhus the late-time losses are believed to be a consequence of incom-lete sterilization or contamination of the vials over the course ofhe test.

PETN was almost completely removed in the active treatmentshat received DARAMEND materials (set-9 to set-12). The lack ofesponse in the controls and the similarity in response to the DARA-END materials with and without granular iron indicate that PETN

egradation is dominated by biological processes, with abiotic pro-esses playing an insignificant role. It is also clear that the organicarbon amendment was necessary in order to stimulate the degra-ation processes. In general, the kinetics of biological processesan be represented by the Monod equation, which includes con-ideration of substrate concentration and bacterial growth rates.he kinetics of PETN degradation in the DARAMEND-added treat-ents (Fig. 4) were highly consistent with the pseudo-first-order

inetic model. This suggests that the biological factors remainedelatively constant over the course of the experiment. The esti-ated first-order half-lives were: 17.5 d (R2 = 0.984) for leached

oil with 1% D6390Fe20 (set-9), 8.9 d (R2 = 0.960) for leached soilith 2% D6390Fe20 (set-10), 14.4 d (R2 = 0.986) for original soil with

% D6390Fe20 (set-11) and 15.8 d (R2 = 0.984) for original soil with% ADM-298500 (set-12). The results showed that the degradationate in the leached soil amended with 2% D6390Fe20 (set-10) waswice that of the test with 1% D6390Fe20 (set-9). Though the evi-ence is sparse, it appears that the rate of PETN removal increasesith increasing amounts of carbon amendment.

It is proposed that organic carbon amendments played twossential roles in stimulating PETN degradation. It served as aubstrate for biological growth and respiration and it was alsoesponsible for the development of anaerobic conditions in theest materials. For example, the addition of molasses at a ratiof 1:20 increased the bacterial density in TNT-contaminated soily two orders of magnitude relative to the soil without molasses15]. Roberts et al. [23] examined the ability of glucose, solubletarch, insoluble starch and acetate as external carbon sources toreate anaerobic conditions for TNT removal. The addition of glu-ose induced the fastest decline and lowest redox potential andesulted in rapid and complete removal of TNT in the contami-ated soil. On the other hand, the added organic carbon substrateould serve as the electron donor for reduction of nitro groups.

n previous work [17], we conducted PETN biodegradation tests

sing an enriched anaerobic consortium from the same contami-ated site and demonstrated that denitrifying bacteria possessingitrite reductase were capable of using PETN and its intermediates Ta

ble

2PE

TN

rem

o

Tub

#

1 2 3 4 5 6 7 8 9 10

Page 6: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

266 L. Zhuang et al. / Journal of Hazardous Materials 264 (2014) 261– 268

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

0 50 100 150 200 250 300 350 400 450

Time (d)

Cu

mu

lati

ve

PE

TN

ma

ss

re

mo

va

l (m

g/k

g)

0.1

0.3

0.5

0.7

0.9

1.1

0 50 100 150 200 250 300 350 400 450

Time (d)

PE

TN

Co

nc

. (C

/C0

)

control 10% Iron 1% D6390Fe202% D6390Fe20 4% D6390Fe20 2%ADM2895004% ADM298500 1% Brewers grain 2% Brewers grain4% Brewers grain

a

b

F oval (t

atntepn(oaPrcp

3

dsohcwr

ig. 6. Decline in average PETN concentration (C/C0) (a) and average PETN mass remhe average of removal in the upper and lower sediment samples.

s terminal electron acceptors. The removal of alternative elec-ron acceptors in aqueous batch experiments followed the order ofitrate, PETN, PETN transformation intermediates and sulfate; andhe majority of PETN degradation was not concurrent with othernergy-yielding reactions, which is a requirement for co-metabolicrocesses. A comparison of Figs. 3a, 3b and 4 shows that removal ofitrate, PETN and sulfate occurred predominantly during the early0–5 d), middle (0–36 d) and late (30–105 d) phases, respectively,f the incubation period. Though overlap among the three electron-ccepting processes occurred, a reduction sequence of nitrate overETN followed by sulfate is evident. Though not supported by theesults of the aqueous tests [17], microbial degradation of PETN byo-metabolic pathways cannot be excluded considering the com-lex characteristics of the contaminated soil.

.3. Pilot test results

The laboratory study demonstrated the occurrence of anaerobicegradation of PETN by indigenous bacteria of the contaminatedite, provided the sediments were amended with readily usablerganic matter. However, the laboratory tests did not simulate the

eterogeneous conditions in the field and may not have adequatelyonsidered other environmental conditions. A small-scale pilot testas conducted to evaluate the transferability of the encouraging

esults obtained in the laboratory to field conditions. This was

mg/kg) (b) in the pilot-scale treatments within 446-day incubation. Data represents

viewed as an essential step prior to considering design for full-scaleremediation. While the field tests focused on the DARAMEND mate-rials that had proven to be successful in the laboratory, a second,and less costly organic material (brewers grain) was also tested,and though granular iron proved to be unsuccessful in the labora-tory tests, one treatment in the field test was reserved for the ironmaterial.

The distribution of PETN contamination at the field site was veryheterogeneous, resulting in different initial concentrations for eachtreatment, ranging from 60,000 to 170,000 mg/kg (Table 2). Thelarge variation in initial concentration complicated data interpre-tation, and thus the results are summarized in two ways. Fig. 5ashows the decline in relative PETN concentration, which is themeasured residual PETN concentration in the soil (C) normalizedby its initial concentration (C0). Fig. 5b shows the cumulativePETN mass removed (mg/kg) over the test period. In both fig-ures, the data represent the average of the PETN concentrationsin the upper and lower soil samples, each of which is the aver-age of duplicate samples. Table 2 summarizes results of both theupper and lower samples in the treatments with organic amend-ments, along with the initial concentrations of PETN, nitrate and

sulfate.

As indicated in Fig. 5a and b, there was a great deal of varia-tion over time in the data for the control treatment. Initially thehigh degree of small-scale variation in PETN concentration in the

Page 7: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

L. Zhuang et al. / Journal of Hazardous Materials 264 (2014) 261– 268 267

-18000

-8000

2000

12000

22000

32000

42000

0 10 20 30 40 50 60 70 80

Time (d)

Cu

mu

lati

ve

PE

TN

ma

ss

re

mo

va

l (m

g/k

g)

0.5

0.6

0.7

0.8

0.9

1.0

1.1

0 10 20 30 40 50 60 70 80

Time (d)

PE

TN

Co

nc

. (C

/C0

)

Con trol 10 % I ron 1% D6390 Fe202% D6390Fe20 4% D6390 Fe20 2% ADM2985004% ADM298500 1% Brewers grain 2% Brewers grain4% Brewers grain

a

b

F oval (t

swrit

tdTit

tc1mwtAe(boa(i

ig. 5. Decline in average PETN concentration (C/C0) (a) and average PETN mass remhe average of removal in the upper and lower sediment samples.

ediment materials was not recognized, and the control, whichas prepared first, received minimum mixing. The treatments that

eceived amendments were mixed much more thoroughly, result-ng in a more homogeneous material and thus more consistentrends in concentration over time.

Of the nine treatments, the decline in PETN was the least inhe granular iron test. Though the variability makes interpretationifficult, the trend in the iron results is dissimilar from the control.hus, as suggested by the laboratory results, it appears that theron was rapidly passivated by the high nitrate concentration inhe sediments.

Appreciable declines in PETN concentration were observed inhe treatments amended with organic materials (Table 2). The per-entage removal at the end of the 74-day test period varied from0% to 42%, with the greatest removal in the 4% D6390Fe20 treat-ent. In terms of PETN mass removal, 11,196–33,398 mg/kg PETNere removed within 74 days, with the greatest mass removal in

he 4% brewers grain treatment. In tub-5, 7 and 10, D6390Fe20,DM-298500 and brewers grain were added as carbon sourcesach at 4%. In terms of the decline in relative concentrationFig. 5a), D6390Fe20 was the most favorable (42%), followed byrewers grain (20%) and ADM-298500 (17%). However, in terms

f mass removal, the order changed to brewers grain, D6390Fe20nd ADM-298500 (Fig. 5b). Considering PETN has a low solubility<40 mg/L), the initial concentration of crystallized PETN in the sed-ment is expected to have minimal effect on microbial activity and

mg/kg) (b) in the pilot-scale treatments within 74-day incubation. Data represents

biodegradation rate. While we do not feel that the data allows usto select a preferred treatment, it does appear that all three carbonamendments were effective in stimulating PETN bioremediation,which is consistent with the laboratory results. Clearly the indige-nous bacteria in the sediments have acclimated to PETN and thusthe results suggest that the persistence of PETN in the settling pondis a consequence of an insufficient supply of labile organic car-bon. Similar to the laboratory findings, the trend of greater PETNremoval with higher loading of carbon sources was observed for allthree types of organic materials. For instant, approximately 16,417,21,388 and 27,644 mg/kg of PETN were removed in the respective1%, 2% and 4% treatments with D6390Fe20 (Fig. 5b). In the labora-tory, with initial PETN concentrations of 4500–5000 mg/kg, nearly100% removal occurred in 84 days and the kinetics of removalwere highly consistent with the pseudo-first-order equation. In thefield however, there was generally rapid removal during the first40 days, but by the end of the experiment, the rates of removaldecreased and deviated from the initial pseudo-first-order behav-ior. The field test proceeded from early August to mid-October,during which time the average ambient daytime temperature var-ied from a high of 32 ◦C to a low of 16 ◦C, and decreased to below25 ◦C after 37 days of incubation. This decline in temperature may

have contributed to the decline in the rate of PETN removal in thelatter stages of the test.

Though the experiment was monitored in detail over the ini-tial 74 days of the test, the experiment was left in tact into the

Page 8: Laboratory and pilot-scale bioremediation of Pentaerythritol Tetranitrate (PETN) contaminated soil

2 rdous

fmetamgic

iaoginprstchms

4

raptpsPsoowatcpa

A

Nw

[

[

[

[

[

[

[

[

[

[

[

[

[22] Q. Lu, Effect of Oxidant (Nitrate) on TCE Degradation by Granular Iron, M.Sc.

68 L. Zhuang et al. / Journal of Haza

ollowing year, allowing for evaluation of much longer-term perfor-ance. As shown in Fig. 6 and Table 2, the trend in PETN removal in

ach treatment was maintained. After the 446-d incubation period,he highest decline in relative concentration of PETN (80%) waschieved in the 4% D6390Fe20 treatment, and the greatest PETNass removed (88,383 mg/kg) was observed in the 4% brewers

rain treatment. Indeed, if the observed trends were to persist,n another year several of the treatments could be approachingomplete removal of the PETN.

Within the 74-d incubation period, algal growth was observedn the treatments with organic amendments, with the greatestlgal production in the treatments receiving the largest amountsf amendment (4% D6390Fe20, 4% ADM298500 and 4% brewersrain). There was no algal growth in the control treatment. Thust appears that the algal blooms were caused by the addition ofutrients associated with the carbon amendments. No tests wereerformed to identify the limiting nutrient(s). Interestingly, refer-ing to Table 2, PETN removal was generally greater in the surfaceamples than in the bottom samples and this difference increased ashe amount of a particular amendment increased. Algal growth willontribute organic exudates and dead algal cells, both of which areighly labile carbon sources for bacteria. This may result in greatericrobial activity and growth in the upper soil than in the lower

oil, and thus explain the greater PETN removal in the surface soil.

. Conclusions

This study examined the feasibility of abiotic and biologicalemediation of PETN-contaminated soil using both granular ironnd organic carbon amendments in both laboratory and pilot-scalerocedures. In the laboratory tests, the effectiveness of the ironreatments was demonstrated to be seriously comprised by ironassivation caused by the presence of high levels of nitrate in theediments. In the treatments using organic carbon amendments,ETN at between 4500 and 5000 mg/kg was removed by indigenousoil bacteria within 84 days. In the pilot tests, significant removalf PETN was observed in the treatments in which a source of labilerganic carbon, such as DARAMEND materials or brewers grain,as provided. The results of the laboratory tests and pilot tests

re qualitatively similar, providing evidence and support for fur-her larger-scale field tests and full-scale treatment using organicarbon amendments. Though not suitable for remediation at thisarticular site, granular iron may nevertheless be an effective agentt sites where the nitrate concentration is low.

cknowledgments

Financial support for this research was provided through theSERC/Dupont/EnviroMetal Industrial Research Chair in Ground-ater Remediation held by Dr. Robert W. Gillham.

[

Materials 264 (2014) 261– 268

References

[1] A. Eisentraeger, G. Reifferscheid, F. Dardenne, R. Blust, A. Schofer, Hazardcharacterization and identification of a former ammunition site using microar-rays, bioassays, and chemical analysis, Environ. Toxicol. Chem. 26 (2007)634–646.

[2] E.L. Rylott, A. Lorenz, N.C. Bruce, Biodegradation and biotransformation ofexplosives, Curr. Opin. Biotech. 22 (2011) 434–440.

[3] L.S. Hundal, J. Singh, E.L. Bier, P.J. Shea, S.D. Comfort, W.L. Powers, Removal ofTNT and RDX from water and soil using iron metal, Environ. Pollut. 97 (1997)55–64.

[4] J. Singh, S.D. Comfort, P.J. Shea, Remediating RDX-contaminated water and soilusing zero-valent iron, J. Environ. Qual. 27 (1998) 1240–1245.

[5] J. Singh, S.D. Comfort, P.J. Shea, Iron-mediated remediation of RDX-contaminated water and soil under controlled Eh/pH, Environ. Sci. Technol.33 (1999) 1488–1494.

[6] S.D. Comfort, P.J. Shea, T.A. Machacek, T. Satapanajaru, Pilot-scale treatmentof RDX-contaminated soil with zero-valent iron, J. Environ. Qual. 32 (2003)1717–1725.

[7] G. Naja, A. Halasz, S. Thiboutot, G. Ampleman, J. Hawari, Degradation ofhexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) using zero-valent iron nanopar-ticles, Environ. Sci. Technol. 42 (2008) 4364–4370.

[8] T.A. Lewis, D.A. Newcombe, R.L. Crawford, Bioremediation of soils contami-nated with explosives, J. Environ. Manage. 70 (2004) 291–307.

[9] C.L. Kitts, D.P. Cunningham, P.J. Unkefer, Isolation of three hexahydro-1,3,5-trinitro-1,3,5-triazine-degrading species of the family Enterobacteriaceae fromnitramine explosive-contaminated soil, Appl. Environ. Microbiol. 60 (1994)4608–4711.

10] S.J. Marshall, G.F. White, Complete denitration of nitroglycerin by bacteriaisolated from a washwater soakaway, Appl. Environ. Microbiol. 67 (2001)2622–2626.

11] O. Isayev, B. Rasulev, L. Gorb, J. Leszczynski, Structure-toxicity relationships ofnitroaromatic compounds, Mol. Divers. 10 (2006) 233–245.

12] R. Boopathy, Bioremediation of explosives contaminated soil, Int. Biodeter.Biodegr. 46 (2000) 29–36.

13] R. Boopathy, Bioremediation of HMX-contaminated soil using soil slurry reac-tors, Soil Sediment Contam. 10 (2001) 269–283.

14] R. Boopathy, Bioremediation of tetryl-contaminated soil using sequenc-ing batch soil slurry reactor, Int. Biodeter. Biodegr. 55 (2005)293–297.

15] K.M. Lamichhane, R.W. Babcock Jr., S.J. Turnbull, S. Schenck, Molasses enhancedphyto and bioremediation treatability study of explosives contaminatedHawaiian soils, J. Hazard. Mater. 243 (2012) 334–339.

16] L. Zhuang, L. Gui, R.W. Gillham, Degradation of Pentaerythritol Tetran-itrate (PETN) by granular iron, Environ. Sci. Technol. 42 (2008)4534–4539.

17] L. Zhuang, L. Gui, R.W. Gillham, Biodegradation of Pentaerythritol Tetrani-trate (PETN) by anaerobic consortia from a contaminated site, Chemosphere89 (2012) 810–816.

18] I.F. Cheng, R. Muftikian, Q. Fernando, N. Korte, Reduction of nitrate to ammoniaby zero-valent iron, Chemosphere 35 (1997) 2689–2695.

19] C.P. Huang, H.W. Wang, P.C. Chiu, Nitrate reduction by metallic iron, Water Res.32 (1998) 2257–2264.

20] O. Schlicker, M. Ebert, M. Fruth, M. Weidner, W. Wüst, A. Dahmke, Degradationof TCE with iron: the role of competing chromate and nitrate reduction, GroundWater 38 (2000) 403–409.

21] K. Ritter, M.S. Odziemkowski, R. Simpgraga, R.W. Gillham, D.E. Irish, An in situstudy of the effect of nitrate on the reduction of trichloroethylene by granulariron, J. Contam. Hydrol. 65 (2003) 121–136.

Thesis, University of Waterloo, Ontario, Canada, 2005.23] D.J. Roberts, F. Ahmad, S. Pendharkar, Optimization of an aerobic polishing stage

to complete the anaerobic treatment of munitions-contaminated soils, Environ.Sci. Technol. 30 (1996) 2021–2026.