Full text of DOI:10.1002/etc.5620210708

Environmental Toxicology and Chemistry, Vol. 21, No. 7, pp. 1384–1389, 2002
Printed in the USA
0730-7268/02 $9.00
ϩ .00
EFFECT OF ADSORPTION TO ELEMENTAL IRON ON THE
TRANSFORMATION OF 2,4,6-TRINITROTOLUENE AND
HEXAHYDRO-1,3,5-TRINITRO-1,3,5-TRIAZINE IN SOLUTION
S
EOK-YOUNG OH,† DANIEL K. CHA,† BYUNG J. KIM,‡ and PEI C. CHIU*†
†Department of Civil and Environmental Engineering, University of Delaware, Newark, Delaware 19716, USA
‡U.S. Army Construction Engineering Research Laboratory, P.O. Box 9005, 2902 Newmark Drive, Champaign, Illinois 61826-9005
(
Received 26 September 2001; Accepted 14 January 2002)
Abstract—The effect of adsorption to elemental iron on the reductive transformation of 2,4,6-trinitrotoluene (TNT) and hexahydro-
1,3,5-trinitro-1,3,5-triazine (royal demolition explosive [RDX]) in aqueous solution was studied with scrap iron and high-purity
iron. In batch experiments with the same total iron surface area and a mixing rate of 100 rpm, TNT and RDX were removed from
the solution within 30 min. With high-purity iron, adsorbed TNT was reduced to 2,4,6-triaminotoluene (TAT) rapidly, with little
accumulation of intermediates at the surface. With scrap iron, the extent of adsorption of TNT and its daughter products was more
significant and reduction of these adsorbed molecules to TAT was slower. Distribution of the intermediates indicated that the reaction
with scrap iron occurred primarily through reduction of the ortho nitro group. Kinetic analysis suggests that mass transfer or
adsorption of TNT controlled the overall rate of TNT reduction to TAT with pure iron, whereas with scrap iron, the rate of TAT
formation was probably limited by other processes. Compared to TNT, transformation of adsorbed RDX was more rapid and less
ϩ
affected by iron type. The RDX was reduced to an unidentified, water-soluble intermediate and NH4 , which accounted for ap-
proximately 50% of the RDX nitrogen. No total organic carbon reduction was observed before and after RDX transformation with
scrap iron.
Keywords—Adsorption
Reduction
2,4,6-Trinitrotoluene
Hexahydro-1,3,5-trinitro-1,3,5-triazine
Elementaliron
INTRODUCTION
[21]. Use of elemental iron to pretreat wastewater containing
TNT and RDX may also be possible, provided the reduction
of these compounds with iron is sufficiently rapid. Studies have
been conducted to investigate the reductive transformation of
TNT, RDX, and other nitroaromatics and nitramines with el-
emental iron. In a batch study, Agrawal and Tratnyek [22]
observed rapid removal of several mono-, di-, and tri-nitroar-
omatic compounds from aqueous solution with elemental iron.
Based on the effect of mixing rate and the minimal influence
of pH and ring substitution on the disappearance rate in so-
lution, these authors proposed that reaction of nitroaromatic
compounds was controlled by mass transfer from solution to
iron surface. Devlin et al. [23] also concluded that electron
transfer was not the rate-limiting step for the reduction of TNT
and other nitroaromatic compounds with scrap iron in a re-
circulating batch experiment. These authors reported that only
traces of 2,4,6-triaminotoluene (TAT) were recovered and that
TNT and its daughter products were strongly or irreversibly
adsorbed to iron. Singh et al. [24] studied RDX reduction with
iron in a batch system and found that RDX was reduced to
nitroso intermediates and ammonium. Oh et al. [25] recently
reported that, in soil microcosms containing master builder
iron, anaerobic sludge, or both, RDX was reduced to small
quantities of nitroso intermediates and methylenedinitramine
(MDNA), a water-soluble product. The RDX removal rate and
the extent of mineralization both increased in the presence of
iron.
2,4,6-Trinitrotoluene (TNT) and hexahydro-1,3,5-trinitro-
1,3,5-triazine (royal demolition explosive [RDX]) are the most
widely used explosives in the world [1]. 2,4,6-Trinitrotoluene
is known to be carcinogenic and mutagenic and is acutely toxic
to microbes, algae, fish, and other organisms [2–4]. Royal
demolition explosive is a heterocyclic nitramine that, together
with octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (high-
melting explosive), is a persistent compound that is toxic to
organisms, including humans [5,6]. Soil and groundwater in
the proximity of munitions-manufacturing plants are often con-
taminated with TNT and RDX [7,8]. In addition, explosives-
laden wastewater from munitions-manufacturing facilities is
difficult to treat because TNT and RDX are resistant to aerobic
biodegradation, because of the presence of electron-withdraw-
ing nitro constituents that inhibit electrophilic attack by ox-
ygenase enzymes [9]. Therefore, these energetic compounds
are often removed from wastewater by costly physicochemical
processes, including advanced oxidation, alkaline hydrolysis,
carbon adsorption, and incineration [10–13].
Elemental iron, usually in the form of scrap iron, has been
increasingly used in recent years in permeable reactive barriers
for the remediation of contaminated groundwater [14].
Through reductive transformation or precipitation, iron has
been shown to remove chlorinated solvents, nitrate, heavy met-
als, and radionuclides from water [15–20]. More recently, el-
emental iron was shown to be able to enhance the degradation
of azo- and nitroaromatic pollutants in water by reductively
transforming them to more biologically amenable compounds
These studies on TNT and RDX reduction with elemental
iron were based on analysis of aqueous samples. However,
molecules that disappeared from the aqueous phase may sim-
ply have been adsorbed to the surface without being trans-
* To whom correspondence may be addressed (pei@ce.udel.edu).
1384

Effect of adsorption on TNT and RDX transformation with iron
Environ. Toxicol. Chem. 21, 2002
1385
formed. Transformation of adsorbed TNT and RDX is difficult
to confirm where aqueous concentrations of intermediates and
products are low [23]. Adsorption may play an important role
in the reduction of organic compounds with iron. Burris et al.
[26] and Allen-King et al. [27] showed that chlorinated ethenes
were adsorbed to nonreactive sites of iron filings. These au-
thors proposed that the adsorption of chlorinated ethenes was
due to exposed graphite in scrap iron and that the extent of
adsorption was related to the hydrophobicity of the adsorbate
[28]. Because scrap iron is typically used for site remediation,
adsorption of hydrophobic organic pollutants may be partic-
ularly important because of high carbon contents of scrap iron.
However, to date, the extent of adsorption and its effect on
the transformation rates of TNT and RDX with elemental iron
have not been characterized.
This study was conducted to determine the extent of ad-
sorption to iron during reductive transformation of TNT and
RDX and the effect of adsorption on the transformation rates.
In addition to extraction from aqueous samples, adsorbed mol-
ecules were extracted and analyzed at different times. The
effect of iron type on adsorption and transformation was in-
vestigated by using high-purity iron powder and scrap iron
having the same surface area. Data also were analyzed to ex-
amine whether TNT reduction and TAT formation rates were
controlled by mass transfer.
periments. The solutions were purged in a glove box to com-
pletely remove oxygen. Initial concentrations of TNT and RDX
were 0.206
Ϯ 0.003 mM and 0.202 Ϯ 0.006 mM, respectively.
After iron was added, the vials were shaken in a horizontal
position by an orbital shaker at 100 rpm. At different elapsed
times, one of the vials was sacrificed and 4.5 ml of the su-
pernatant was filtered through a 0.22-␮m mixed cellulose
membrane filter (Millipore, Bedford, MA, USA) for analysis
by high-performance liquid chromatography (HPLC). Aceto-
nitrile extraction of used filters indicated that TNT and RDX
did not adsorb to the filters.
Extraction of adsorbed compounds
After 4.5 ml of supernatant was removed, the iron (and 0.5
ml of solution) remaining in the vial was extracted once for
high-purity iron powder and twice for scrap iron filings, each
with 2 ml of acetonitrile, to recover adsorbed molecules. Ad-
ditional extraction did not yield meaningful increases in re-
covery. For high-purity iron, after 2 ml of acetonitrile was
added, the vial was vigorously shaken for 1.5 min by a vortex
mixer, 2 ml of the supernatant was withdrawn and passed
through a glass fiber filter, and the filtrate was analyzed by
HPLC. Surface concentrations of adsorbates, defined as ad-
sorbed mass divided by initial solution volume (5 ml), were
calculated with Equation 1
CS
S is the surface concentration (mM; the adsorbed mass
divided by the initial solution volume [ T]), e is the concen-
ϭ
(
Ce
·
Ve
Ϫ
Ca
·
V
a)/VT
(1)
MATERIALS AND METHODS
where
C
Chemicals
V
C
2,4,6-Trinitrotoluene (Ͼ99%) and RDX (Ͼ99%) were pro-
tration in the extract (mM), obtained from HPLC analysis, Ca
vided by Holston Army Ammunition Plant (Kingsport, TN,
USA). 2-Amino-4,6-dinitrotoluene (2A46DNT; 99%) and 4-
amino-2,6-dinitrotoluene (4A26DNT; 99.5%) were obtained
from Chem Service (West Chester, PA, USA). 2,4-Diamino-
6-nitrotoluene (24DA6NT; 0.1 mg/ml standard in acetonitrile),
2,6-diamino-4-nitrotoluene (26DA4NT; 0.1 mg/ml standard in
acetonitrile), and 2,4,6-triaminotoluene trihydrochloride
is the aqueous concentration (mM), obtained from HPLC anal-
ysis, Va is the volume of solution extracted by acetonitrile (0.5
ml), Ve is the extract volume (2.5 ml), and VT is the initial
solution volume (5 ml).
For scrap iron, the extraction was repeated with another 2
ml of acetonitrile. Total surface concentration was taken to be
the sum of the two surface concentrations calculated based on
the two extract concentrations, by using Equations 2 through
(
ϳ
100%) were purchased from Accustandard (New Haven,
CT, USA). -(2-Hydroxyethyl)piperazine- -2-(ethanesulfon-
ic acid) (HEPES) was obtained from Sigma (St. Louis, MO,
USA). Acetonitrile ( 99.9%) was purchased from Fisher Sci-
N
NЈ
4. The first and second extraction efficiencies, defined as CS1
CS and CS2 S, were 82.5 0.3% and 17.5 0.3%, respec-
tively, for TNT.
/
/
C
Ϯ
Ϯ
Ն
entific (Pittsburgh, PA, USA). All chemicals were used as
received.
Two types of elemental iron were used for this study. Master
builder iron was chosen as a representative scrap iron because
it has been widely used in groundwater remediation work and
its elemental composition has been determined [20,29,30].
CS1
CS2
CS
ϭ
ϭ
ϭ
(
Ce1
Ce2
·
·
Ve
Ve
Ϫ
Ϫ
Ca
·
V
a)/VT
(2)
(3)
(4)
(
Ce1
·
V
a)/VT
CS1
ϩ
CS2
where CS1 and CS2 are surface concentrations based on first
and second extractions (mM), respectively, and Ce1 and Ce2
are concentrations in the first and second extracts (mM), re-
spectively.
High-purity iron powder (Ͻ10 ␮m, Ͼ99.5%) was purchased
from Alfa Aesar (Ward Hill, MA, USA). These irons were
used as received without pretreatment. Specific surface areas
of master builder iron and the high-purity iron powder were
1.29 and 0.19 m2/g, respectively, as determined by the Bru-
nauer–Emmett–Teller method with N2.
Analytical methods
The TNT, 2A46DNT, 4A26DNT, 24DA6NT, 26DA4NT,
RDX, and a soluble RDX reduction product (see Results and
Discussion section) were analyzed with a Varian high-perfor-
mance liquid chromatograph (Walnut Creek, CA, USA)
Batch reduction experiments
Batch reduction experiments were conducted in an anaer-
obic glove box (95% N2
ϩ
5% H2, Coy, MI, USA) with 8-ml
equipped with a Supelguard guard column (20
Supelco, Bellefonte, PA, USA), a Supelco LC-18 column (250
4.6 mm, 5- m particle size), an ultraviolet (UV) detector
(2510 Varian), and an isocratic pump (2550 Varian). A meth-
anol:water mixture (55:45, v/v) was used as the mobile phase
at a flow rate of 1.0 ml/min. 2,4,6-Triaminotoluene was ana-
ϫ 4.6 mm,
borosilicate vials containing 5 ml of aqueous solution and
either 1 g of scrap iron or 6.8 g of high-purity iron powder.
The different iron masses were used to give the same surface
area of 1.29 m2. Replicate vials were set up for each experi-
ment. The TNT and RDX solutions contained 0.1 M HEPES
buffer to maintain a constant pH of 7.4 throughout the ex-
ϫ
␮
lyzed by HPLC with an Alltima C18 column (250
ϫ 4.6 mm,

1386
Environ. Toxicol. Chem. 21, 2002
S.-Y. Oh et al.
5-
guard column (7.5
buffer (40 mM, pH 3.2, 10:90, v/v) was used as an eluent at
␮
m particle size, Alltech, Deerfield, IL, USA) and an Alltima
ϫ
4.6 mm, Alltech). Acetonitrile:phosphate
1.0 ml/min. The injection volume for all samples was 10
and the wavelength for the UV detector was 254 nm.
␮l
Total organic carbon (TOC) was analyzed with a TOC an-
alyzer (DC-190 Dohrmann, Santa Clara, CA, USA). The
ϩ
NH4 was analyzed by the salicylate method [31] with a UV-
vis spectrophotometer (DR2010, HACH, Loveland, CO,
USA). Nitrate and nitrite were analyzed with an ion chro-
matograph (Dionex, Sunnyvale, CA, USA) equipped with a
Dionex Ionpac AS11 column.
RESULTS AND DISCUSSION
Reduction of TNT with scrap iron and high-purity iron
powder
Figure 1 shows the aqueous, surface, and total concentra-
tions of TNT and the major reduction intermediates
(2A46DNT, 4A26DNT, 24DA6NT, and 26DA4NT) and end
product (TAT) during reduction with scrap iron. Aqueous TNT
concentration rapidly decreased to zero within 30 min, but
only low quantities of the intermediates were detected in so-
lution during this period (Fig. 1a). More than 75% of the initial
TNT was recovered as TAT after 3 h. However, the surface
concentrations of adsorbed compounds presented a very dif-
ferent picture. As shown in Figures 1a and 1b, almost one half
of the TNT removed from water in the first 5 min (ϳ0.13 mM)
was adsorbed to scrap iron, and the amounts of intermediates
adsorbed were greater than that found in solution. In com-
parison, TAT was only weakly adsorbed because of its rela-
tively high water solubility. The dominant intermediate was
2A46DNT, indicating that reduction of the ortho nitro group
was the principal reaction in the TNT reduction pathway with
scrap iron. Reduction of the ortho nitro group was also a major
reaction for 2A46DNT, as shown by the high yields of
26DA4NT. Figure 1c shows the total concentrations of TNT
and its daughter products, which were obtained by adding the
aqueous and surface concentrations. These profiles represent
concentration changes due only to transformation. Complete
reduction of TNT to TAT took more than 3 h, because the
reaction rate decreased with decreasing number of nitro
groups. Mass balance during the experiment ranged from 70.1
to 97.3%. The lower mass balance at early times might be
ascribed to the TNT reduction products that were not measured
(e.g., nitroso compounds).
Figure 2 shows the aqueous, surface, and total concentration
profiles during TNT reduction with high-purity iron powder.
Similar to the data in Figure 1a, TNT was completely removed
from solution within 30 min (Fig. 2a). The intermediates found
in the aqueous phase were 24DA6NT and 4A26DNT, but only
in minute quantities (Fig. 2a, insert). In contrast to Figure 1a,
aqueous TAT appeared very early and its concentration quickly
approached the initial TNT concentration. This suggested that
TNT and intermediates adsorbed on high-purity iron were
transformed much faster than those adsorbed on scrap iron.
Indeed, Figure 2b shows that all adsorbed TNT reacted within
30 min, adsorbed TAT appeared immediately, and no inter-
mediates accumulated at the surface. These observations are
in sharp contrast to those in Figure 1b. All TNT was fully
reduced and recovered as TAT in solution and the mass balance
ranged from 82 to 108% during the experiment (Fig. 2c).
From the results shown in Figures 1 and 2, adsorption of
TNT and its daughter products clearly was more pronounced
Fig. 1. Aqueous (a), surface (b), and total (c) concentrations of 2,4,6-
trinitrotoluene (TNT), 2,4,6-triaminotoluene (TAT), and four inter-
mediates during TNT reduction with scrap iron. The dashed lines in
Figure 1a represent the aqueous TNT and TAT concentration profiles
based on the fitted first-order rate constant
0.190 0.993) for the disappearance of TNT in
0.013/min (r2
solution. 4A26DNT 4-amino-2,6-dinitrotoluene; 2A46DNT 2-
amino-4,6-dinitrotoluene; 26DA4NT 2,6-diamino-4-nitrotoluene;
24DA6NT 2,4-diamino-6-nitrotoluene.
Ϯ standard deviation of
Ϯ
ϭ
ϭ
ϭ
ϭ
ϭ
with scrap iron. This may be attributed to the carbon in scrap
iron. Master builder iron is known to contain 2 to 4% carbon
and other elements [28–30]. Adsorption of chlorinated ethenes
to scrap iron has been proposed to occur at nonreactive sites,
such as graphite inclusions [28]. However, in contrast to the
studies with chlorinated ethenes [26–28], adsorbed TNT on
scrap iron was reduced much more rapidly. If most TNT was

Effect of adsorption on TNT and RDX transformation with iron
Environ. Toxicol. Chem. 21, 2002
1387
which adsorbed TNT is transformed is currently under inves-
tigation.
The results also suggest that, for pure iron powder, the
overall rate of TNT reduction to TAT was controlled by either
mass transfer or adsorption of TNT in solution. Equation 5
shows a simplified reaction sequence of TNT reduction to TAT.
Note that all steps were assumed to be first-order and desorp-
tion of the intermediates was omitted for simplicity. If the
removal of TNT from solution controlled the overall rate of
the reaction, then either the TNT mass transfer rate constant
(
k
MT,TNT) or the adsorption rate constant (
k
ads) would be ap-
overall, Eqn. 6).
proximately equal to the overall rate constant (
k
In that case, the sequence can be modeled as a single-step,
first-order reaction (Eqn. 6) and the aqueous concentrations of
TNT and TAT can be calculated with Equations 7 and 8. The
predicted concentration profiles of aqueous TAT with scrap
iron and pure iron are shown in Figures 1a and 2a (dashed
lines), by using the fitted pseudo–first-order rate constants for
the disappearance of aqueous TNT (fitted curves also shown
in the figures). Note that these empirical rate constants may
correspond to either kMT,TNT or kads, because mass transfer and
adsorption could not be differentiated in our experiments. For
scrap iron, the model prediction differs greatly from the actual
aqueous concentrations of TAT, suggesting that TAT formation
was controlled not by TNT removal from solution but by sub-
sequent steps in the sequence. For high-purity iron, the model
fits the observed TAT data quite well, which suggests that
either mass transfer or adsorption of aqueous TNT was the
rate-limiting step in the sequence (Eqn. 5).
kMT,TNT
kads
kR1
kR2
TNT(aq)
→
→
→
→
kRn
kdes
kMT,TAT
→ →
→
→
→
TAT(aq)
(5)
koverall
TNT(aq)
→
TAT(aq)
(6)
(7)
(8)
Ϫ
koverall
t
[TNT](aq)
[TAT](aq)
ϭ
ϭ
[TNT]o(aq) ·e
Ϫ
koverallt
[TNT]o(aq) ·(1
Ϫ
e
)
where kR1 kR2 and kRn are the rate constants of sequential re-
actions, kdes is the desorption rate constant, and is time.
t
Reduction of RDX with scrap iron and high-purity iron
powder
Figures 3 and 4 show the changes of aqueous, surface, and
total concentrations of RDX during its reduction with scrap
iron and high-purity iron, respectively. The RDX disappeared
from the aqueous phase within 30 min regardless of the iron
type, as was observed for TNT. However, in contrast to the
TNT data with scrap iron (Fig. 1b), adsorbed RDX was trans-
formed within 30 min with both pure iron and scrap iron
(although slightly faster with pure iron). Thus, the RDX ad-
sorbed to scrap iron apparently either desorbed or reacted more
rapidly than adsorbed TNT.
Fig. 2. Aqueous (a), surface (b), and total (c) concentrations of 2,4,6-
trinitrotoluene (TNT), 2,4,6-triaminotoluene (TAT), and four inter-
mediates during TNT reduction with high-purity iron powder. The
dashed lines in Figure 2a represent the aqueous TNT and TAT con-
centration profiles based on the fitted first-order rate constant of 0.096
Ϯ
0.009/min (r2
4A26DNT 4-amino-2,6-dinitrotoluene; 2A46DNT
dinitrotoluene; 26DA4NT 2,6-diamino-4-nitrotoluene; 24DA6NT
2,4-diamino-6-nitrotoluene.
ϭ
0.990) for the disappearance of TNT in solution.
2-amino-4,6-
ϭ
ϭ
ϭ
ϭ
By using the same liquid chromatography (LC) conditions
as for RDX analysis, we detected a large peak with a retention
also adsorbed to nonreactive sites (e.g., graphite), then the
faster removal of TNT may be explained by the high reactivity
of TNT. Rapid consumption of aqueous TNT could have shift-
ed the equilibrium and caused TNT to desorb from the non-
reactive sites. Alternatively, the more pronounced adsorption
of TNT with scrap iron could be due to adsorption to other
types of reactive sites, which were different from and less
reactive than those found in pure iron. The mechanism through
time of 2.2
Ϯ 0.1 min during RDX reduction with both irons.
The area of the peak increased with time initially and then
slowly decreased (Figs. 3a and 4a), suggesting that it was an
RDX reduction intermediate. This compound was water-sol-
uble and could not be solvent-extracted for analysis by gas
chromatography–mass spectrometry (GC-MS). In soil micro-
cosms containing elemental iron, Oh et al. [25] recently iden-

1388
Environ. Toxicol. Chem. 21, 2002
S.-Y. Oh et al.
Fig. 3. Aqueous (a), surface (b), and total (c) concentrations of hex-
ahydro-1,3,5-trinitro-1,3,5-triazine (royal demolition explosive
[RDX]) during its reduction with scrap iron. Figure 3a also shows
the liquid chromatography (LC) peak area of a possible RDX reduction
product.
Fig. 4. Aqueous (a), surface (b), and total (c) concentrations of hex-
ahydro-1,3,5-trinitro-1,3,5-triazine (royal demolition explosive
[RDX]) during its reduction with high-purity iron powder. Figure 4a
also shows the liquid chromatography (LC) peak area of an uniden-
tified reduction product of RDX.
tified MDNA to be a soluble RDX reduction product by using
LC-MS-MS. Whether the LC peak we observed corresponds
to MDNA remains to be determined.
through enzymatic hydrolysis to form MDNA and
bis(hydroxymethyl)nitramine. These compounds were further
degraded abiotically to give formaldehyde, methanol, nitrous
oxide (N2O), and water. Formaldehyde and methanol were
eventually converted to CO2 and CH4.
In an experiment conducted in deionized water (i.e., the
HEPES buffer was omitted), we found that the TOC remained
constant before (12.4 mg/L) and after (12.2 mg/L) complete
RDX transformation with scrap iron. This indicates that the
carbon in the triazine ring was not mineralized. By using liq-
uid–liquid extraction with hexane or ether followed by GC-
MS analysis, we did not detect either formaldehyde or meth-
anol. In fact, no MS peaks were observed, suggesting that the
carbonaceous products were highly polar or charged.
Although the pathway for RDX reduction with elemental
iron is not resolved, a pathway for biological RDX reduction
has been proposed. Under anaerobic conditions, RDX was
shown by McCormick and coworkers [32] to transform in a
stepwise fashion to hexahydro-1-nitroso-3,5-dinitro-1,3,5-tri-
azine, hexahydro-1,3-dinitroso-5-nitro-1,3,5-triazine, and hex-
ahydro-1,3,5-trinitroso-1,3,5-triazine. Further degradation in-
volved cleavage of the triazine ring and formation of hydra-
zine, 1,1-dimethylhydrazine, 1,2-dimethylhydrazine, formal-
dehyde, and methanol. Another possible reduction pathway
was shown in anaerobic sludge by Hawari and coworkers [33],
who proposed that the triazine ring of RDX was cleaved
Attempts were also made to establish nitrogen balance by

Effect of adsorption on TNT and RDX transformation with iron
Environ. Toxicol. Chem. 21, 2002
1389
ϩ
Ϫ
Ϫ
11. Li ZM, Comfort SD, Shea PJ. 1997. Destruction of 2,4,6-trini-
trotoluene by Fenton oxidation. J Environ Qual 26:480–487.
12. Alnaizy R, Akgerman A. 1999. Oxidative treatment of high ex-
plosives contaminated wastewater. Water Res 33:2021–2030.
13. Bier EL, Singh J, Li Z, Comfort SD, Shea PJ. 1999. Remediating
hexahydro-1,3,5-trinitro-1,3,5-triazine–contaminated water and
soil by Fenton oxidation. Environ Toxicol Chem 18:1078–1084.
14. Vidic RD, Pohland FG. 1996. Treatment walls. TE-96-01. Tech-
nology Evaluation Report. Groundwater Remediation Technolo-
gies Analysis Center, Pittsburgh, PA, USA.
15. Orth WS, Gillham RW. 1996. Dechlorination of trichloroethene
in aqueous solution using Fe0. Environ Sci Technol 30:66–71.
16. Roberts AL, Totten LA, Arnold WA, Burris DR, Campbell TJ.
1996. Reductive elimination of chlorinated ethylenes by zero-
valent metals. Environ Sci Technol 30:2654–2659.
17. Nam S, Tratnyek PG. 2000. Reduction of azo dyes with zero-
valent iron. Water Res 34:1837–1845.
18. Huang CP, Wang HW, Chiu PC. 1998. Nitrate reduction by me-
tallic iron. Water Res 32:2257–2264.
19. Blowes DW, Ptacek CJ, Jambor JL. 1997. In-situ remediation of
Cr(VI)-contaminted groundwater using permeable reactive walls:
Laboratory studies. Environ Sci Technol 31:3348–3357.
20. Gu B, Liang L, Dickey MJ, Yin X, Dai S. 1998. Reductive pre-
cipitation of uranium (VI) by zero-valent iron. Environ Sci Tech-
nol 32:3366–3373.
measuring NH4 , NO2 , and NO3 in solution during RDX trans-
ϩ
formation. The NH4 appeared immediately and, within 1 h,
its concentration reached a plateau, which corresponded to
approximately 50% of the total RDX nitrogen (data not
shown). The rapid formation of NH4 indicated that the nitro-
gen–nitrogen bonds in RDX were cleaved early in the reaction.
No NO2 or NO3 was detected during the experiment. These
ions either were not formed during RDX reduction or were
rapidly reduced with iron to NH4 and did not accumulate
ϩ
Ϫ
Ϫ
ϩ
[18,34].
In summary, our results show that TNT molecules removed
from water may persist for hours at the scrap iron surface,
although all adsorbed TNT was eventually reduced to TAT.
The ortho nitro group of TNT seemed to be preferentially
reduced with scrap iron. Mass transfer seems to be the rate-
limiting step for both TNT reduction and TAT formation with
high-purity iron, but not with scrap iron. In contrast to TNT,
reduction of adsorbed RDX was faster and less affected by
iron type. Approximately one half of the nitrogen in RDX was
ϩ
recovered as NH4 . No TOC removal was observed during RDX
reduction with scrap iron and the carbonaceous products re-
main to be identified. The rapid reduction of adsorbed TNT
and RDX suggests that elemental iron may be useful for treat-
ing munitions-manufacturing wastewater if combined with a
subsequent treatment process to achieve complete minerali-
zation of these energetic compounds.
21. Perey JR, Oh SY, Lubenow BL, Cha DK, Huang CP, Chiu PC.
2001. Enhancing biodegradability of refractory aromatics: Pre-
treatment with elemental iron. In Magar VS, von Fahnestock FM,
Leeson A, eds, Ex-Situ Biological Treatment Technologies. Bat-
telle, Columbus, OH, USA, pp 149–155.
22. Agrawal A, Tratnyek PG. 1996. Reduction of nitro aromatic com-
pounds by zero-valent iron metal. Environ Sci Technol 30:153–
160.
23. Devlin JF, Klausen J, Schwarzenbach RP. 1998. Kinetics of ni-
troaromatic reduction on granular iron in recirculating batch ex-
periments. Environ Sci Technol 32:1941–1947.
Acknowledgement—We thank the U.S. Army Construction Engi-
neering Research Laboratory for funding this research.
24. Singh J, Comfort SD, Shea PJ. 1998. Remediating RDX-contam-
inated water and soil using zero-valent iron. J Environ Qual 27:
1240–1245.
REFERENCES
1. Spain JC, Hughes JB, Knackmuss HJ. 2000. Biodegradation of
Nitroaromatic Compounds and Explosives. Lewis, Boca Raton,
FL. USA.
2. Smock LA, Stoneburner DL, Clark JR. 1976. The toxic effects
of trinitrotoluene (TNT) and its primary degradation products on
two species of algae and the fathead minnow. Water Res 10:537–
543.
3. Won WD, DiSalvo LH, Ng J. 1976. Toxicity and mutagenicity
of 2,4,6-trinitrotoluene and its metabolites. Appl Environ Micro-
biol 31:576–580.
4. Kaplan DL, Kaplan AM. 1982. 2,4,6-Trinitrotoluene–surfactant
complexes: Decomposition, mutagenicity, and soil leaching stud-
ies. Environ Sci Technol 16:566–571.
5. Yinon J. 1990. Toxicity and Metabolism of Explosives. CRC,
Boca Raton, FL, USA.
6. Harvey SD, Fellows RJ, Cataldo DA, Bean RM. 1991. Fate of
the explosive hexahydro-1,3,5,-trinitro-1,3,5-triazine (RDX) in
soil and bioaccumulation in bush bean hydroponic plants. Environ
Toxicol Chem 10:845–855.
25. Oh BT, Just CL, Alvarez PJJ. 2001. Hexahydro-1,3,5-trinitro-
1,3,5-triazine mineralization by zero-valent iron and mixed an-
aerobic cultures. Environ Sci Technol 35:4341–4346.
26. Burris DR, Campbell TJ, Manoranjan VS. 1995. Sorption of tri-
chloroethylene and tetrachloroethylene in a batch reactive me-
tallic iron–water system. Environ Sci Technol 29:2850–2855.
27. Allen-King RM, Halket RM, Burris DR. 1997. Reductive trans-
formation and sorption of cis- and trans-1,2-dichloroethene in a
metallic iron–water system. Environ Toxicol Chem 16:424–429.
28. Burris DR, Allen-King RM, Manoranjan VS, Campbell TJ, Lo-
raine GA, Deng B. 1998. Chlorinated ethene reduction by cast
iron: Sorption and mass transfer. J Environ Eng 124:1012–1019.
29. Reardon EJ. 1995. Anaerobic corrosion of granular iron: Mea-
surement and interpretation of hydrogen evolution rates. Environ
Sci Technol 29:2936–2945.
30. Hardy LI, Gillham RW. 1996. Formation of hydrocarbonsfrom
reduction of aqueous CO2 by zero-valent iron. Environ Sci Tech-
nol 30:57–65.
31. Hach. 1998. DR/2010 Spectrophotometer Handbook. Loveland,
CO, USA.
7. Hundal LS, Shea PJ, Comfort SD, Powers WL, Singh J. 1997.
Long-term TNT sorption and bound residue formation in soil.
Environ Qual 26:894–904.
J
32. McCormick NG, Cornell JH, Kaplan AM. 1981. Biodegradation
of hexahydro-1,3,5-trinitro-1,3,5-triazine. Appl Environ Micro-
biol 42:817–823.
33. Hawari J, Halasz A, Sheremata T, Beaudet S, Groom C, Paquet
L, Rhofir C, Ampleman G, Thiboutot S. 2000. Characterization
of metabolites during biodegradation of hexahydro-1,3,5-trinitro-
1,3,5-triazine (RDX) with municipal anaerobic sludge. Appl En-
viron Microbiol 66:2652–2657.
8. Boopathy R, Manning JF. 2000. Laboratory treatability study on
hexahydro-1,3,5-trinitro-1,3,5-triazine–(RDX–)contaminated soil
from the Iowa army ammunition plant, Burlington, Iowa. Water
Environ Res 72:238–242.
9. Bruhn C, Lenke L, Knackmuss HJ. 1987. Nitrosubstituted aro-
matic compounds as nitrogen source for bacteria. Appl Environ
Microbiol 53:208–210.
10. Heilmann HM, Wiesmann U, Stenstrom MK. 1996. Kinetics of
the alkaline hydrolysis of high explosives RDX and HMX in
aqueous solution and adsorbed to activated carbon. Environ Sci
Technol 30:1485–1492.
34. Kielemoes J, De Boever P, Verstraete W. 2000. Influence of de-
nitrification on the corrosion of iron and stainless steel powder.
Environ Sci Technol 34:663–671.