Environ. Sci. Technol.
2001, 35, 4054-4059
Hydrogen Peroxide Effects on
Chromium Oxidation State and
Solubility in Four Diverse,
Chromium-Enriched Soils
MELANIE L. ROCK,†
BRUCE R. JAMES,‡ AND
G E O R G E R . H E L Z * ,†
Department of Chemistry and Biochemistry and Water
Resources Research Center and Soil Chemistry Program,
Department of Natural Resources Science and Landscape
Architecture, University of Maryland,
College Park, Maryland 20742
High concentrations of H2O2 are being tested for in situ
oxidation and remediation of buried organic contaminants
in soils and groundwater. Peroxide is being considered
as a direct chemical oxidant in Fenton-type reactions or as
a source of oxidizing equivalents in bioremediation
schemes. How H2O2 affects the oxidation state and solubility
of Cr(III) and Cr(VI), common co-contaminants with
organic chemicals, is explored here in four chemically
diverse soils containing elevated levels of Cr. Soil
contaminated with soluble Cr(VI) from chromite ore
processing residue and soil containing high levels of
recently reduced Cr (III) from electroplating waste both
released dissolved Cr(VI) after single applications of up to
24 mM H2O2. In no case was there evidence that H2O2
reduced preexisting Cr(VI) to Cr(III), even though this would
be allowed thermodynamically. Chromate in the leachates
exceeded the U.S. EPA drinking water standard for
total dissolved Cr (2 µM) by a factor of 10-1000. Anaerobic
conditions in an organic-rich, tannery waste-contaminated
soil protected Cr(III) from oxidation and mobilization.
Mineral forms of Cr in serpentinitic soil near a former
chromite mine also resisted oxidation on the time scale of
days. Mobilization of Cr(VI) could be a hazardous
consequence of using H2O2 for in situ remediation of
chemically complex wastes, but H2O2 could prove attractive
for ex situ treatment (i.e., soil washing). This paper
demonstrates marked differences among Cr-contaminated
soils in their capacity to release Cr(VI) upon chemical
treatment with H2O2.
Introduction
Hydrogen peroxide is a potentially useful reagent for in situ
waste remediation. It can be used with high Fe(II) concen-
trations to effect Fenton-type oxidation of refractory organic
contaminants in soils and groundwater. In one field test, 436
m3 of 50% H2O2 was injected into an aquifer to oxidize an
* Corresponding author e-mail: gh17@umail.umd.edu; phone:
(301)405-1797; fax: (301)314-9121.
† Department of Chemistry and Biochemistry and Water Resources
Research Center.
‡ Soil Chemistry Program, Department of Natural Resources
Science and Landscape Architecture.
4054 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 20, 2001
estimated 33 000 kg of trichloroethylene (1). In contrast, H2O2
can be used at lower levels (3-300 mM) to deliver oxidizing
equivalents to support in situ bioremediation (2).
Such uses of H2O2 raise concern about possible side
reactions with co-contaminants. A common co-contaminant
in hazardous wastes is Cr, which is reported at 43% of the
hazardous waste sites on the National Priorities List in the
United States (3). The ability of H2O2 to act both as an
oxidizing agent and as a reducing agent is well-known. Under
appropriate conditions H2O2 can oxidize Cr(III) to Cr(VI) or
reduce Cr(VI) to Cr(III) (4, 5). What its effect, if any, might
be on Cr at hazardous waste sites is an open question.
Chromium(VI) is a human carcinogen that predominates in
soils and natural waters as mobile anions, HCrO4- or CrO42-.
In contrast, Cr(III) is an essential trace element for humans
and is relatively immobile in the same environments because
of low solubility and propensity to sorb to natural solids (6).
Cr contamination in soils has resulted from the disposal of
industrial wastes produced by several different processes
(e.g., electroplating, leather tanning, chromite ore mining
and processing). Cr-contaminated sites are therefore chemi-
cally diverse with respect to forms, oxidation state, and
solubility of this element.
We explore in this paper the potential for solubilization
of Cr by aqueous H2O2 in selected soils that incorporate some
of the chemical and pedologic diversity found at contami-
nated sites. Differences include chromium deposition pro-
cesses, total Cr levels, soil pH, redox conditions, chromate/
bichromatespeciation,andorganic/inorganicCr(III)speciation.
Low millimolar H2O2 levels were chosen for the experimental
treatments based on the concentration range being con-
sidered for enhanced bioremediation treatments (2).
Site Descriptions
Sampling. A similar sampling protocol was followed at each
site: an undisturbed area about 1 m2 was cleared of leaf and
plant covering, a pit was dug, soil horizons were marked and
identified, and samples were taken from each horizon.
Samples were kept intact as large blocks (∼2 L volume), sealed
in 0.004 in thick plastic bags, transported to the laboratory
in coolers, and stored in a refrigerator at 4° C in the field-
moist condition (0 to ∼0.1 kPa water potential). Intact blocks
of soil from individual horizons were prepared for the
experiments by passing them through a polyethylene sieve
using gentle hand pressure to obtain a relatively homoge-
neous subsample from each horizon. Soil from the chromite
ore processing site was prepared using a 0.40 cm sieve; a 0.25
cm sieve was used on all other soils.
Electroplating Waste Site. National Chromium, Inc. (a
small electroplating facility near Putnam, CT) discharged
chromic acid-containing wastewater directly into an adjacent
mixed hardwood wetland and flood plain from 1939 to 1975
(7). The peat-like uppermost horizon contained the highest
total Cr levels of any soil in this study (Table 1). Green
chromium(III) hydroxide coatings were evident on fallen
branches and plant debris surrounding the site, and some
soil samples contained as much as 60 g of Cr/kg of soil. The
soil is very poorly drained, and its pH is low: 4-5. Even
though its high levels of organic matter (200 g of C/kg of soil)
and low pH might be expected to reduce Cr(VI) to Cr(III), it
nonetheless contains leachable Cr(VI) (60-90 µM) in the
uppermost horizon. The X-ray diffraction spectrum indicated
the presence of quartz and Cr(III)-rich chlorite. Soluble Cr-
(VI) was not detected in the middle horizon at this site but
did appear in the glacial till below 100 cm. Mattuck (8)
10.1021/es010597y CCC: $20.00  2001 American Chemical Society
Published on Web 08/30/2001
TABLE 1. Soil Properties
horizon soil organic carbon
soil designation (cm) pH (g/kg of soil)
electroplating 0-14 5.4 280d
waste 14-40 5.0 118d
>100 4.9 5.2d
ore processing residue surface 8.8 66
tannery 0-20 6.7 154d
20-40 5.7 202d
chromite mine 53-75 6.5 4.2d
a Atomic absorption analysis, H SO -H O -HF digestate. b DPC colorimetry, heated, strongly alkaline, OH--CO 2- leach. c DPC colorimetry,
2 4 2 2
aqueous leach, 10:1 solution:soil ratio. d Measurements by Typrin (36).
reported Cr(VI) levels of up to 950 µM in groundwater from
a well about 10 m downslope from the sampling locality.
Chromite Ore Processing Site. Soil incorporating chromite
ore processing residue in Kearney, NJ, had a comparatively
high pH (8.8), a relatively large Cr(VI):Cr(III) ratio (0.10), and
high ambient levels of soluble Cr(VI) (940 µM) (9-11). During
ore processing, chromite ore was crushed to less than 100
mesh size, mixed with soda ash, and roasted to produce
soluble Na2CrO4. The Na2CrO4 was leached from the roasted
material several times with water, after which the alkaline
residue was discarded in low-lying areas and wetlands. The
residue, which still releases Cr(VI) decades later (9), retains
residual levels of sparingly soluble salts including CaCrO4,
Ca4Al2CrO10, and FeOHCrO4 as well as insoluble Cr(III)
resistant to the original processing (10). Soil samples were
from a flood plain about 600 m from the Hackensack River
where the depth to groundwater was 2-3 m. Powder X-ray
diffraction patterns revealed only quartz and calcitesno
crystalline Cr minerals (12).
Tannery Site. The Aberjona watershed near Woburn, MA,
was the site of over 100 chrome tanning operations between
1838 and 1988. Inorganic arsenical and lead-based insecti-
cides were manufactured in the same locale from the 1860s
until the 1920s (13, 14). In the tanning process, Cr(VI) was
reduced with organic acids over the surface of animal hides,
forming stable Cr(III) complexes that preserved the leather.
Waste Cr(III,VI) was mainly discharged into lagoons that were
used to dispose of the sludge from leather production (15).
Because of the high levels of animal-derived organic waste
at these sites, natural biodegradation processes produce
highly reducing environments that evolve H2S and CH3SH
(13).
Two horizons of a riverbank soil were sampled at the
Superfund site downstream from the tannery disposal
lagoons. Sulfidic conditions in the shallow groundwater near
the site could be expected to reduce Cr(VI) (16). Although Cr
appears exclusively as Cr(III) in these sites (Table 1), it has
a surprising degree of mobility in the groundwater (13),
probably due to the enhanced solubility of Cr(III) organic
complexes formed with dissolved organic C provided by the
decomposing hides (17).
Mine Site. Unlike the three other sites, the Maryland
serpentine barrens, about 20 km northwest of Baltimore,
contain only naturally elevated levels of Cr. This site lies
within the belt of serpentinites scattered in the Piedmont
Plateau from New Jersey to Alabama (18). The site first became
a source of industrial chromite around 1840. Soils here have
been studied extensively to determine the cause of their low
soil fertility (19). Soil was sampled at Soldier’s Delight Natural
Environmental Area, 100 m from an old mine. Samples used
in this study contained 2.5 g of total Cr/kg of soil (Table 1).
X-ray diffraction revealed only antigorite, a serpentine
mineral. No chromite was identified, even though this mineral
is reportedly the Cr host phase at the site. If all the Cr in our
sample were present in chromite, this phase would comprise
total Cra total Cr(VI)b soluble Cr(VI)c soluble Cr(III)d
(g/kg of soil) (g/kg of soil) (µM) (µM)
61 ( 5 0.071 ( 0.007 95 ( 6 11 ( 1
21 ( 1 0.079 ( 0.005 <0.1 <1.0
0.40 ( 0.03 0.079 ( 0.003 32 ( 1 <1.0
8.6 ( 0.5 0.914 ( 0.022 940 ( 40 <1.0
1.3 ( 0.1 <0.00005 <0.1
4.7 ( 0.2 <0.00005 <0.1 2.0 ( 0.2
2.5 ( 0.2 <0.00005 <0.1 2.0 ( 0.2
3
only 0.5% of the samplestoo little to observe by X-ray
diffraction.
Methods
Procedures. Batch experiments on soils were conducted in
triplicate in 50 mL, polycarbonate centrifuge tubes. Suspen-
sions containing solution:soil ratios of 10:1 by mass were
prepared by placing 3.00 g of field-moist soil in each tube
and equilibrating for 24 ( 1 h at 25 °C on an orbital shaker
with 30.0 mL of 0.01 M NaNO3. After equilibration, H2O2 was
added to the soil suspensions in small volumes to obtain the
desired initial concentration in the soil suspension while not
diluting significantly the original Cr concentrations. To
monitor soluble Cr(VI), samples were extracted, centrifuged
(15000g, 15 min, 25 °C), and analyzed as described below.
Experiments were also conducted on synthetic chromium-
(III) (hydr)oxides. Solutions containing 280 µM Cr(NO3)3 and
0.01 M NaNO3 were prepared and neutralized with 0.005 M
NaOH to give total OH-/Cr ratios of 0, 1.0, 2.0, 3.0, and 4.0.
To minimize high, localized OH- concentrations during
preparation, NaOH was added just below the liquid surface
at a rate of 2 mL min-1 while stirring. Formation of particles
(blue-green flocs) was observed only in the 3:1 and 4:1 OH:
Cr solutions. The solutions were aged with gentle swirling
on an orbital shaker (50 cycles min-1) for at least 1 week,
after which time the pH values remained stable for several
months. To avoid possible effects that have been noted from
the interaction of buffers with Cr/H2O2 intermediate species
(5, 20) and to observe pH changes in the systems we
investigated, solutions were prepared without buffers. Reac-
tions of these Cr(III) solutions were studied with a procedure
like that used with the soil suspensions: peroxide was added
to the OH-/Cr(III) solutions, and the production of Cr(VI)
was monitored over time.
Analysis. The pH was measured with an Orion flat surface
combination pH electrode inserted directly into the super-
natant solution in each centrifuge tube. Soluble Cr(VI) was
measured by the diphenylcarbazide (DPC) colorimetric
method (21) using a 4.5 mL aliquot of the centrifugate.
Diphenylcarbazide reacts with Cr(VI) in acidic solution to
form a Cr(III)-diphenylcarbazone complex that absorbs at
540 nm (detection limit 0.1 µM). At concentrations higher
than 10-5 M, H2O2 causes a negative interference with the
DPC determination of Cr(VI) by competitively reducing Cr-
(VI) under the pH 1.8 conditions of the test. Pettine et al. (22)
found that the negative interference of H2O2 can be avoided
at H2O2 < 10-4 M by increasing reagent concentrations. In
our experiments, however, much higher concentrations of
H2O2 were used (up to 0.1M). Removal of peroxide by adding
catalytic (10-8 M) amounts of the enzyme, catalase, to a
sample and waiting 30 min prior to analysis was found to be
effective. Where catalase was used, Cr(VI) standards were
also prepared with 10-8 M catalase to correct for the slight
absorbance of catalase at 540 nm. Results for standards using
DPC and catalase corresponded well with results obtained
VOL. 35, NO. 20, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4055
FIGURE 1. Dissolved Cr(VI) released from the 0-14 cm (peaty) horizon
of the electroplating waste site upon treatment with various
concentrations of H2O2, which was added at time ) 0. Solution to
soil ratio, 10:1. Peroxide decayed rapidly, as indicated in Table 2.
Analyses were made in triplicate; the scatter at each sampling
point in the figure is contributed both by soil heterogeneity and by
analytical error.
using a direct optical method (23) at 350 nm for HCrO4-.
Total Cr(VI) in soil samples was determined using a heated
carbonate-hydroxide aqueous extraction method with quan-
tification by DPC colorimetry (24). Slight optical interference
due to solubilized humic materials from the soils was
corrected by subtracting the absorbance measured in soil
extracts containing no DPC. Soluble total Cr in the soil was
determined by atomic absorption spectroscopy following a
dilute salt extraction (0.01 M CaCl2, 1:5 soil:solution ratio, 1
h equilibration) (36). Soluble Cr(VI) was determined by DPC
colorimetry, and soluble Cr(III) was estimated by difference
between soluble total Cr and soluble Cr(VI). Total Cr in soil
was determined by atomic absorption following a digestion
procedure using H2SO4-H2O2-HF dissolution of oven-dried
(105 °C), crushed (35-mesh) soil samples (25).
A modification of the 4-aminoantipyrene horseradish
peroxidase method, reviewed in (26), was used to determine
H2O2 concentrations. H2O2 oxidatively couples with 4-ami-
noantipyrene (AAP) and phenol in the presence of horserad-
ish peroxidase to produce a quinoneimine dye with a
maximum absorption at 505 nm. The linear range was 5-300
µM H2O2; for many of our assays, it was necessary to dilute
the sample into this range before analysis. High concentra-
tions of phenol provided stearic hindrance to complexation
of AAP with organic hydroperoxides in soil solution, ensuring
specificity for H2O2 (26). The modified aqueous reagent was
mixed in a 500 mL volumetric flask as follows: 0.0010 g of
horseradish peroxidase (type VI) from Sigma (about 2 × 10-8
M), 0.50 g of AAP, 1.17 g of phenol, 5.0 mL of 0.1 M phosphate
buffer (pH 6.9), and 100 µL of 0.01 M H2O2 (or 2 µM) to give
more stable readings. Reagent (2.0 mL) and aqueous sample
(3.0 mL) were vortexed; absorbance readings at 505 nm
reached a maximum in 30-120 s, after which time they were
no longer stable. Sample color faded at a rate that varied
with H2O2 concentrations, so it was necessary to make
measurements promptly. Further details on methods and
tabular presentations of the data are available in Rock (12).
Results
Soil Experiments. As shown in Figure 1, H2O2 additions to
the surface soil at the electroplating site produced rapid,
dose-dependent increases in soluble Cr(VI) in the uppermost,
peaty horizon. Without added H2O2, little rise in soluble Cr-
4056 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 20, 2001
FIGURE 2. Dissolved Cr(VI) release from the 14-40 cm and >100
cm horizons of the electroplating waste site upon treatment with
H2O2. Analyses made in triplicate.
TABLE 2. Decay H2O2 upon Contact with Several Soilsa
soil initial 1.5 h 24 h 48 h
electroplating waste site, 0-14 cm 24 11.7 <0.005 <0.005
electroplating waste site, 14-40 cm 24 18.5 2.6 <0.005
ore processing residue 24 10.1 <0.005 <0.005
(VI) was observed after the water-soluble fraction of Cr(VI)
had initially dissolved. Increases in Cr(VI) levels peaked 4 h
after H2O2 application. Thereafter, Cr(VI) drifted back toward
pre-peroxide levels over a period of 3 weeks. Within 90 min
after application, ∼50% of the H2O2 in samples spiked with
24 mM had disappeared, and peroxide was not detectable
in any sample after 1 day (Table 2). Thus, the decline in
Cr(VI) after 4 h probably reflects reduction of Cr(VI) by organic
matter in the peaty soil after the H2O2 has disappeared. On
a molar basis, the maximum amount of Cr(VI) released from
the soil was approximately 1% of the H2O2 lost, indicating
that these processes are not linked stoichiometrically.
At the two highest peroxide doses, the maximum Cr(VI)
concentration exceeded all the Cr(VI) in the soil sample, based
on total Cr(VI) (Table 1). As discussed later, this suggests that
H2O2 was oxidizing Cr(III) in the sample. Alternately, it would
be possible that H2O2 is simply a more efficient extractor of
preexisting Cr(VI) in the soil than is the hot alkaline leach
used to quantify total Cr(VI) (Table 1), although this seems
less likely since H2O2 contains no ions capable of desorbing
or dissolving Cr(VI).
As shown in Figure 2, Cr in the 14-40 cm horizon at the
electroplating site was dissolved by a 12 mM H2O2 dose, but
to a very limited extent. This sample initially contained no
soluble Cr(VI), but total Cr(VI) was similar to that in the
surface layer at this site (Table 1). Chromium(VI) levels
reached 5-6 µM. H2O2 took longer to disappear in the 14-40
cm horizon as compared to the 0-14 cm horizon (Table 2).
Application of only 3 mM peroxide to soil from the >100
cm horizon at the electroplating site produced an interesting
effect (Figure 2). Initially, chromate levels fell in this low
organic carbon soil. Chromate extracted from this soil was
determined by direct optical absorption (350 nm) rather than
by DPC. The initial loss of chromate absorption does not
necessarily indicate reduction of Cr(VI) to Cr(III). Conceiv-
ably, the loss of HCrO4- absorbance was due to the formation
of another, nonabsorbing Cr(VI) species, for example, a
peroxo-Cr(VI) complex. We will return to this possibility later.
Peroxide in the >100 cm soil took about 2 d to reach the
detection limit of 5 µM.
FIGURE 3. Dissolved Cr(VI) concentrations released from ore
processing residue (10:1 solution/soil by mass) upon single
applications of H2O2 at time ) 0. Peroxide decayed rapidly and was
undetectable after 1 day. Analyses made in triplicate.
The soil sample enriched with chromite ore processing
residue released 900 µM Cr(VI) when leached for 24 h simply
with 0.01 M NaNO3 containing no H2O2 (Figure 3). As much
as 200 µM additional Cr(VI) was released over 24 h by
leachates containing up to 24 mM H2O2. Peroxide dropped
by more than 50% within 90 min after addition and was
undetectable after 24 h (Table 2). In contrast to the
electroplating site (Figure 1), Cr(VI) in the ore processing
residue did not decline appreciably even over a period of 1
week after H2O2 was undetectable. The slight Cr(VI) decline
observed in the most concentrated sample in Figure 3 was
only 6%, nearly within analytical uncertainty. It seems likely
that the very high soluble Cr(VI) concentrations and the
moderate organic C content of this soil overwhelmed a re-
adsorption mechanism like that observed at the electroplating
site.
Release of Cr(VI) was not observed in the highly reducing
environment of the Massachusetts tannery waste site soil at
peroxide application levels of up to 100 mM. Samples from
the chromite ore site also did not produce detectable soluble
Cr(VI) after addition of up to 100 mM H2O2 (data not shown
for either soil).
Treatment of Synthetic Chromium(III) Hydroxides. To
prove that H2O2 can oxidize Cr(III) to Cr(VI) under conditions
similar to those used in our soil leaching experiments, we
prepared chromium(III) (hydr)oxides by neutralizing 280 µM
Cr(NO3)3 with 1-4 equiv of NaOH. At OH/Cr ratios of 1and
2, the pH was 3-4 and all of the Cr(III) remained dissolved
as cationic hydroxide complexes (e.g., CrOH2+). At OH/Cr
ratios of 3 and 4, blue green flocs were present. The Cr(III)
solubility reached a minimum at OH/Cr ) 3, corresponding
to the stoichiometry of Cr(OH)3(s); at higher OH/Cr ratios,
some Cr(III) redissolved, presumably as anionic hyrdoxo-
chromates (e.g., Cr(OH)4-).
Figure 4 shows Cr(VI) production by oxidation of the Cr-
(III) in these mixtures after 100 µM H2O2 was added.
Production continued slowly over 4 months, although most
of the Cr(VI) appeared within 3 weeks. The greatest extent
of oxidation occurred under mildly alkaline conditions at
OH-/Cr ) 4, where at the end of 115 days, 71% of the
stoichiometric yield of Cr(VI) expected from 100 µM H2O2
was obtained. The smallest yield occurred in the 3:1 mixture
where Cr(III) solubility was minimal. Note that the reversal
of Cr(VI) production seen in Figure 1 does not occur in these
synthetic mixtures, presumably due to the absence of a
reducing agent or sorption surface.
FIGURE 4. Production of dissolved Cr(VI) from synthetic Cr(III)
hydrolysis products aged 1 week before addition of 100 µM H2O2.
Time measured from H2O2 addition. Bold numbers indicate the initial
stoichiometric ratios of NaOH to Cr(NO3)3. Indicated pH measure-
ments made at day 17; pH started higher in each case but stabilized
by day 17. Oxidation of Cr(III) by H2O2 generates acidity.
Discussion
To summarize our observations, responses to H2O2 were
varied. The highly reducing tannery soil and the soil
developed on a natural mineral deposit at the mine site
released no Cr(VI), while soils at the electroplating and ore
processing sites released large amounts of Cr(VI). However
in the acidic, peaty electroplating site surface soil, Cr(VI) was
lost from solution within days after H2O2 had decayed. In the
alkaline soil with moderate organic C at the ore processing
site, most of the Cr(VI) remained dissolved after H2O2 had
decayed. In no case was there evidence that H2O2 was
reducing preexisting Cr(VI) to Cr(III), even though this would
be allowed thermodynamically. In the samples where H2O2
treatments had any effect, that effect was always a higher,
not a lower, Cr(VI) yield as compared to treatment with a
redox-inert electrolyte.
Total Cr and total Cr(VI) proved poor predictors of Cr(VI)
dissolution after H2O2 treatment. For example, in the 14-40
cm horizon at the electroplating site, with 2.1% total Cr, very
little Cr(VI) was released (Figure 2); whereas, the >100 cm
horizon at the same site, with 0.04% total Cr, released more
Cr(VI) despite treatment with 25% as much H2O2. These two
horizons were essentially identical with respect to initial Cr-
(VI) content and soil pH (Table 1).
In our limited group of samples, the best predictor of
Cr(VI) release was the presence of soluble Cr(VI) in the soil
prior to H2O2 treatment. The tannery and mine soils as well
as the 14-40 cm horizon at the electroplating waste site
contained no Cr(VI) that was extractable in the aqueous leach
(Table 1). Each of these soils released little or no Cr(VI) upon
subsequent treatment with H2O2. Each of the three soils that
contained substantial leachable Cr(VI) prior to H2O2 treat-
ment also released much more Cr(VI) after treatment.
Several explanations for observed increases in Cr(VI) after
H2O2 treatment can be proposed: H2O2 may have oxidized
Cr(III) in the soil or it may have simply solubilized existing
Cr(VI). The first can be represented as follows:
3H2O2(aq) + 2Cr(OH)3(s) f 2HCrO4- + 2H+ + 4H2O
(1)
The second might occur through formation of aqueous
chromium(VI) peroxide species (5, 27-30), as represented
by the following reaction in which a solid Cr(VI) source is
represented by chromatite:
VOL. 35, NO. 20, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4057
2H2O2(aq) + CaCrO4(s) + H+ a
- 2+
Cr(O)(O2)2(OH) + Ca + 2H2O (2)
Additionally, Cr(VI) mobilization might be accomplished
simply physically, as oxidation of organic matter promotes
disintegration of soil structure, thereby increasing access of
the solution to Cr(VI) mineral phases.
Although we were unable to observe direct evidence for
formation of chromium(VI) peroxide complexes, several lines
of evidence favor this complexation phenomenon as an
important process at the ore processing site. First, as already
noted, this soil is known to contain sparingly soluble Cr(VI)
salts such as chromatite (CaCrO4). Second, Cr(III) in this soil
has been found previously to be resistant to other oxidants.
James (11) found that neither the Cr(III) in ore residue soils
nor the soluble Cr(III) added to these soils oxidizes when
exposed to manganese(III,IV) (hydr)oxides, despite favorable
conditions (e.g., high pH and low organic matter content).
Third, if sparingly soluble Cr(VI) salts dissolve through
complexation with H2O2, subsequent decomposition of the
complex would release CrO42- into the soil solutions as H2O2
levels dropped. In Figure 3, Cr(VI) concentrations appear to
continue rising beyond the first day, by which time all the
free H2O2 is gone, suggesting slow release of CrO42- as
peroxochromates decompose. (At the electroplating site
(Figure 1) this phenomenon, if it existed, was obscured by
the rapid removal of Cr(VI) by organic matter after H2O2 was
gone.)
On the other hand, complexation of Cr(VI) by peroxide
cannot be the sole process explaining our results. Oxidation
of Cr(III) by H2O2 is thermodynamically feasible and has been
demonstrated in Figure 4 over the pH range relevant to our
soils. A mass balance analysis indicated that some Cr
oxidation must have taken place in the surface horizon at
the electroplating site, at least at the highest H2O2 treatment
levels. Soluble Cr(VI) extracted from this soil surpassed the
initial Cr(VI) concentrations reported in Table 1. Further
support for Cr oxidation at the electroplating site is provided
by the pH, which declined from 5.5 to as low as 4.6 after
applying H2O2, despite the buffering capacity of the soil. As
shown by reaction 1, oxidative dissolution of Cr(III) lowers
pH. However, oxidation of other soil components (e.g.,
organic matter) could also contribute to acid generation.
A key point demonstrated by the experiments reported
in Figure 4 is that the yield of Cr(VI) obtained by H2O2
treatment depends on the history of the Cr(III) starting
material. The aged Cr(III) solutions used in this study likely
contained mixtures of soluble and insoluble oligomers, e.g.,
Crn(OH)3n-22+ and Crn(OH)3n0 (31). At OH-/Cr(III) ) 3:1, flocs
of Crn(OH)3n formed. Figure 4 shows that the 3:1 mixture was
relatively unreactive to H2O2. The soluble Cr(III) products,
stable at both lower and higher OH-/Cr(III) ratios, were more
easily oxidized by H2O2. It is likely that yields of Cr(VI) from
soils are also dependent upon such factors as the pH of the
soil, pH of the Cr-bearing contaminant, redox status, and
aging time.
Soil organic matter clearly plays important roles in Cr
mobility. At the electroplating waste site, a reducing fraction
of the high levels of organic matter in the surface horizon
probably accounted for the decline of soluble Cr(VI) after
H2O2 decay. Wittbrodt and Palmer (32) observed the reduction
of Cr(VI) by soil humic and fulvic acids across a pH range
of 2-7. Nakayasu et al. (33) found that gallic and tannic acids
(polyphenols that originate in decaying leaves, likely precur-
sors to humic substances) reduced Cr(VI) at pH 5 at even
faster rates than humic and fulvic acids. The refractory
character of Cr at the tannery site is also likely due to organic
matter.
An issue not answered by our work is the fate of H2O2. The
limited data in Table 2 indicate that the half-life of H2O2 in
4058 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 35, NO. 20, 2001
contact with our soils was on the order of hours. As noted
above, the quantity of peroxide lost greatly exceeded the
amount of Cr(VI) released in all samples. A recent investiga-
tion of peroxide decomposition in a number of natural soils
(34) has shown that the principal product of the H2O2 reaction
with ordinary soils is O2, possibly indicating pervasive catalase
activity. This implies that most H2O2 introduced into soils
disproportionates without oxidizing soil components. Nev-
ertheless, measurable amounts of hydroxyl radical can be
produced (34). At high concentrations, dissolved Cr may
catalyze H2O2 decomposition (5, 12). In this process, Cr(VI)
oxidizes H2O2 to O2 while lower-valent Cr products reduce
H2O2 to H2O.
It should be kept in mind that our experiments involved
only one application of peroxide. Most remedial peroxide
treatments call for continuous delivery of peroxide over many
days and at much higher concentrations than those used in
these experiments. The extent and persistence of Cr(III)
oxidation under such conditions would be much greater than
we have observed, as would be the possibility of forming
reactive peroxochromium intermediate species.
Our results raise a cautionary flag regarding the use of
H2O2 for in situ remediation of contaminated soils associated
with hazardous waste sites. On the other hand, H2O2 leaching
could be attractive under the conditions of ex situ remediation
(i.e., soil washing; 35) provided that recovery of leachates
could be assured. Our work shows that peroxide mobilizes
more anthropogenic Cr than can be obtained with a neutral
aqueous leach (NaNO3 solution) and that it does not mobilize
tightly bound Cr in natural forms such as chromite (FeCr2O4).
Excess H2O2 decomposes to harmless products, H2O and O2.
Given the great variability we have observed from one site
to another, the suitability of H2O2 in any remediation strategy
would have to be evaluated site specifically.
Acknowledgments
This work was partially supported by the Maryland Water
Resources Research Center with grants from the U.S.
Geological Survey and from the National Science Foundation
(DGE9354935).
Literature Cited
(1) In Situ Remediation Technology: In Situ Chemical Oxidation;
EPA/542/S-98/008; Office of Solid Waste and Emergency
Response, U.S. Environmental Protection Agency: Washington,
DC, 1998.
(2) Pardieck, D. L.; Bouwer, E. J.; Stone, A. J. Contam. Hydrol. 1992,
9, 221-242.
(3) U.S. Environmental Protection Agency. http://www.epa.gov/
superfund/sites/ (Accessed December 2000).
(4) Pettine, M.; Millero, F. J. Limnol. Oceanogr. 1990 35, 730-736.
(5) Perez-Benito, J. F.; Arias, C. J. Phys. Chem. A 1997, 101, 4726-
4733.
(6) James, B. R. Environ. Sci. Technol. 1996, 30, 248A-251A.
(7) Nikolaidis, N. P.; Robbins, G. A.; Scherer, M.; McAninch, B.;
Binkhorst, G.; Asikainen, J.; Suib, S. L. Ground Water Monit.
Remed. 1994, 14, 150-159.
(8) Mattuck, R. Master’s Thesis, University of ConnecticutsStorrs
(Technical Report ERI-94.05), 1994.
(9) Burke, T.; Fagliano, J.; Goldoft, M.; Hazen, R. E.; Iglewicz, R.;
McKee, T. Environ. Health Perspect. 1991, 92, 131-137.
(10) Weng, C. H.; Huang, C. P.; Allen, H. E.; Cheng, A. H. D.; Sanders,
P. F. Sci. Total Environ. 1994, 154 (1), 71-86.
(11) James, B. R. J. Environ. Qual. 1994, 23, 227-233.
(12) Rock, M. L. Ph.D. Dissertation, University of Maryland, College
Park, 1999.
(13) Davis, A.; Kempton, J. H.; Nicholson, A.; Yare, B. Appl. Geochem.
1994, 9, 569-582.
(14) Spliethoff, H. M.; Hemond, H. F. Environ. Sci. Technol. 1996,
30, 121-128.
(15) Interim Report on the Groundwater Quality of East and North
Woburn, Massachusetts; Contract 68-01-6056, TDD F1-8010-
04B, U.S. Environmental Protection Agency: Washington, DC,
1981.
(16) Thornton, E. C.; Amonette, J. E. Environ. Sci. Technol. 1999, 33,
4096-4101.
(17) James, B. R.; Bartlett, R. J. J. Environ. Qual. 1983, 12, 169-172.
(18) Pearre, N. C.; Heyl, A. V. Chromite and Other Mineral Deposits
in Serpentine Rocks of the Piedmont Upland, Maryland,
Pennsylvania and Delaware; Bulleting 1082-K; U.S. Geological
Survey: Reston VA, 1960.
(19) Rabenhorst, M. C.; Foss, J. E.; Fanning, D. S. Soil Sci. Soc. Am.
J. 1982, 46, 607-616.
(20) Beck, M. T.; Nagy, I. P.; Székely, G. Int. J. Chem. Kinet. 1991, 23,
881-896.
(21) Bartlett, R.; James, B. J. Environ. Qual. 1979, 8, 31-35.
(22) Pettine, M.; La Noce, T.; Liberatori, A.; Loreti, L. Anal. Chim.
Acta 1988, 209, 315-319.
(23) Hug, S. J.; James, B. R.; Laubscher, H.-U. Environ. Sci. Technol.
1996, 31, 160-170.
(24) James. B. R.; Petura, J. C.; Vitale, R. J.; Mussoline, G. R. Environ.
Sci. Technol. 1995, 29, 2377-2381.
(25) Bowman, R. A. Soil Sci. Soc. Am. J. 1988, 52, 1301-1304.
(26) Frew, J. E.; Jones, P.; Scholes, G. Anal. Chim. Acta 1983, 155,
139-150.
(27) Funahashi, S.; Uchida, F.; Tanaka, M. Inorg. Chem. 1978, 17,
2784-2789.
(28) Witt, S. N.; Hayes, D. H. Inorg. Chem. 1982, 21, 4014-4016.
(29) Dickman, M. H.; Pope, M. T. Chem. Rev. 1994, 94, 569-584.
(30) Zhang, L. B.; Lay, P. A. Inorg. Chem. 1998, 37, 1729-1733.
(31) Stunzi, H.; Marty, W. Inorg. Chem. 1983, 22, 2145-2150.
(32) Wittbrodt, P. R.; Palmer, C. D. Eur. J. Soil Sci. 1996, 47, 151-162.
(33) Nakayasu, K.; Fukushima, M.; Sasaki, K.; Tanaka, S.; Nakamura
H. Environ. Toxicol. Chem. 1999, 18, 1085-1090.
(34) Petigara B. R. Hydrogen Peroxide Decomposition Mechanisms
in Natural Soil Systems. Ph.D. Dissertation, University of
Maryland, College Park 2000, 182 pp.
(35) Rock, M. L.; Kearney, P. C.; Helz, G. R. In Pesticide Remediation
in Soils and Groundwater; Kearney, P. C., Roberts, T., Eds.; John
Wiley Sons: Chichester, England, 1998.
(36) Typrin, L. R. Master’s Thesis, University of Maryland, College
Park, 1998.
Received for review January 31, 2001. Revised manuscript
received July 16, 2001. Accepted July 16, 2001.
ES010597Y
VOL. 35, NO. 20, 2001 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4059