Water, Air, and Soil Pollution

, Volume 197, Issue 1, pp 49–60

Reduction of Hexavalent Chromium in Soil and Ground Water Using Zero-Valent Iron Under Batch and Semi-Batch Conditions


    • Institute of ChemistryLQA—UNICAMP
  • Leonardo M. Da Silva
    • Department of ChemistryFACESA—UFVJM
  • Wilson F. Jardim
    • Institute of ChemistryLQA—UNICAMP

DOI: 10.1007/s11270-008-9790-0

Cite this article as:
Franco, D.V., Da Silva, L.M. & Jardim, W.F. Water Air Soil Pollut (2009) 197: 49. doi:10.1007/s11270-008-9790-0


Chemical remediation of soil and groundwater containing hexavalent chromium (Cr(VI)) was carried out under batch and semi-batch conditions using different iron species: (Fe(II) (sulphate solution); Fe0G (granulated elemental iron); ZVIne (non-stabilized zerovalent iron) and ZVIcol (colloidal zerovalent iron). ZVIcol was synthesized using different experimental conditions with carboxymethyl cellulose (CMC) and ultra-sound. Chemical analysis revealed that the contaminated soil (frank clay sandy texture) presented an average Cr(VI) concentration of 456 ± 35 mg kg−1. Remediation studies carried out under batch conditions indicated that 1.00 g of ZVIcol leads to a chemical reduction of ∼280 mg of Cr(VI). Considering the fractions of Cr(VI) present in soil (labile, exchangeable and insoluble), it was noted that after treatment with ZVIcol (semi-batch conditions and pH 5) only 2.5% of these species were not reduced. A comparative study using iron species was carried out in order to evaluate the reduction potentialities exhibited by ZVIcol. Results obtained under batch and semi-batch conditions indicate that application of ZVIcol for the “in situ” remediation of soil and groundwater containing Cr(VI) constitutes a promising technology.


Contaminated soilHexavalent chromiumCarboxymethyl celluloseColloidal zerovalent ironChromium immobilization

1 Introduction

Chromium and other heavy metals can contaminate soil and groundwater by means of several anthropogenic processes involving industrial activities, e.g., electroplating process, wood, pulp and tannery processing, corrosion inhibition and pigment production (Tokunaga et al. 2001; Xu and Zhao 2007).

Chromium can be considered a pollutant in its different oxidation states (Cr(0)–Cr(VI)) (Calder 1988; Palmer and Wittbrodt 1991). However, in natural environments, the more stable forms of chromium are available as soluble and insoluble compounds containing the Cr(IV) and Cr(III) species (Calder 1988; Palmer and Wittbrodt 1991). Hexavalent chromium (Cr(VI)) presents the higher standard oxidation potential and exhibits toxic and carcinogenic properties for different biological systems (Calder 1988; Palmer and Wittbrodt 1991; Kimbrough et al. 1999).

According to the literature (Calder 1988; Palmer and Wittbrodt 1991) most compounds containing the Cr(III) species present a very low solubility in water and, therefore, their transport in soil and groundwater is limited when compared with Cr(VI) species. Besides, the toxicity exhibited by compounds containing Cr(III) is considerably lower than that of compounds containing Cr(VI) (Calder 1988; Palmer and Wittbrodt 1991; Kimbrough et al. 1999).

The anthropogenic occurrence of chromium in different polluted environments is regulated and controlled by the different environmental protection agencies around the world. For instance, the U.S. Environmental Protection Agency (EPA) established a maximum concentration value (MCV) for total chromium in potable water of 0.1 mg l−1, while in Brazil the National Sanitary Agency (ANVISA—resolution MS n° 518) established a MCV for total chromium in potable water of 0.05 mg l−1.

Hexavalent chromium, in the anionic form, is soluble in water at different pH values (Calder 1988; Palmer and Wittbrodt 1991); the chromate anion (CrO42−(aq)) is the predominant form of Cr(VI) in natural environments presenting a pH in the range of 6 to 9. On the contrary, the ion Cr(III) remains stable in natural environments as an insoluble hydroxide (Cr(OH)3(S)) (Calder 1988; Palmer and Wittbrodt 1991).

The “in situ” chemical reduction of soil containing Cr(VI) following the direct application of the reductant (solution or suspension) has been reported in the literature using Fe(II) and ZVIcol (Tokunaga et al. 2001; Su and Ludwig 2005; Seaman et al. 1999; Ponder et al. 2000; EPA 2000). When compared to the passive redox remediation process using permeable reactive barriers that uses Fe0, the alternative remediation technology comprising the direct application of the redox solution onto the soil containing Cr(VI) can offer several advantages (He et al. 2007). For instance, the redox agent solution can be applied directly in the aquifer zone without further modifications in the natural flux of the groundwater, thus reducing costs and remediation time (He et al. 2007).

Studies have revealed that Cr(VI) can be reduced “in situ” and precipitated as an insoluble hydroxide compound (Cr(OH)3 and/or FexCr1-x(OH)3) following the direct application in soil of a suspension containing ZVIcol (Xu and Zhao 2007; Ponder et al. 2000). Some reports (He et al. 2007; He and Zhao 2005; Schrick et al. 2004) revealed that the direct application of a suspension containing ZVIcol in soil can be accompanied by an undesirable agglomeration of the ZVI particles. In this case both the redox power and the mobility in soil of the redox agent are considerably reduced or inhibited due to particle agglomeration.

Several studies have been carried out in order to obtain stable suspensions of metallic particles (e.g. Au0, Ag0, Fe0 and iron oxide) (Xu and Zhao 2007; He et al. 2007; Schrick et al. 2004; Kataby et al. 1997, 1999; Sun and Zeng 2002; Kim et al. 2003; Si et al. 2004; Magdassi et al. 2003) and different stabilizing agents have been used to prevent agglomeration of the iron oxide nanoparticles, including thiols, carboxylic acids, surfactants and different polymers (Xu and Zhao 2007; He et al. 2007; He and Zhao 2005; Schrick et al. 2004; Kataby et al. 1997, 1999; Sun and Zeng 2002; Kim et al. 2003). Despite of these efforts, only a few stabilizing agents are indeed available for stabilization of the ZVI particles when this species is devoted to environmental applications, since for in situ application only environmentally friendly species that result in a low cost-effect ratio for the remediation process are viable alternatives (He and Zhao 2005; He et al. 2007).

An ideal stabilizing agent for ZVI nanoparticles must present the following characteristics (He and Zhao 2005; He et al. 2007): (1) be able to interact specifically with the ZVIcol nanoparticles and to suppress the agglomeration process; (2) be environmentally friendly; (3) low cost and (4) provide a high mobility of the ZVIcol nanoparticles in the soil microstructure. Most of these pre-requisites can be indeed attained in practice using stabilizing agents based on natural polymers (e.g. amide, modified amide, alginates, xantanes, etc) (He and Zhao 2005; He et al. 2007). In fact, these species have been used with success during preparation and stabilization of supermagnetic iron oxide nanoparticles (SPIONs), Fe0 and Ag0 (Xu and Zhao 2007; He et al. 2007; He and Zhao 2005; Schrick et al. 2004; Kataby et al. 1997, 1999; Sun and Zeng 2002; Kim et al. 2003; Si et al. 2004; Magdassi et al. 2003).

Concerning the laboratory preparation of stable suspensions containing metallic nanoparticles, the synthesis assisted by ultra-sound has proved to be a powerful tool in order to obtain ZVI nanoparticles presenting a low average size and, in some cases, causing the formation of porous ZVI particles with very high active surface area for the redox process (Suslick et al. 1996).

This paper describes a comparative redox treatment study using different iron species (Fe(II) as well as macro and nanoparticles of Fe0) for the reduction of the hexavalent chromium present in soil and groundwater. The remediation study was carried out under batch (slurry reactor) and semi-batch (packed bed column reactor) conditions.

2 Experimental

2.1 Soil Analysis

Chemical analysis of the contaminated site (an industrial waste landfill located in Brazil) revealed an average total chromium concentration in groundwater and soil (frank clay sandy texture) of 550 ± 12 mg dm−3 and 456 ± 35 mg kg−1, respectively.

Hexavalent chromium extracted from groundwater and soil was analyzed via the colorimetric technique using the diphenylcarbazide complexation procedure (maximum absorption at 540 nm) following the method 7196A of EPA. Determination of total Cr and Fe was carried out using atomic absorption (AAS) according to method 3050B of EPA.

Analytical grade chemicals were used throughout: ferrous sulphate (FeSO4·7H2O—Synth); sodium boronhydride (NaBH4—Acros Organic); sodium carboxymethyl cellulose (CMC—Synth); sec-diphenylcarbazide (Synth); granulated iron (Fe0G—Aldrich) and acetone (Mallinkrodt).

2.2 Synthesis of Zero-valent Iron (ZVI)

Non-stabilized (ZVIne) and stabilized (ZVIcol) ZVI-nanoparticles were synthesized in the aqueous phase via chemical reduction of a Fe(II) solution in the absence and the presence of CMC and ultra-sound. In all cases, NaBH4 was used as the reductant (Xu and Zhao 2007; He et al. 2007; He and Zhao 2005; Schrick et al. 2004; Suslick et al. 1996).

2.2.1 Synthesis of ZVIcol

A Fe(II) solution prepared using FeSO4·7H2O was added to a reaction flask containing the CMC solution in order to obtain a colloidal suspensions of Fe0 in the desired concentration: 1–2 g l−1 Fe0 stabilized with 0.25% CMC (w/w). The flask containing the colloidal suspension was allowed to stand for 20 min. After that, the ZVIcol suspensions were prepared by adding NaBH4 directly to the reaction flask using the [Fe(II)]:[NaBH4] molar ratio of 1:2. This procedure was assisted by ultra-sound for 5 min.

2.2.2 Synthesis of ZVIne

ZVIne suspensions were prepared as described previously (see item 2.2.1), except for the fact that CMC was not used. Figure 1 shows the different macroscopic appearance presented by the ZVIcol and the ZVIne suspensions.
Fig. 1

Macroscopic appearance presented by the ZVIcol and the ZVIne suspensions. a ZVIne (absence of ultrasound). b ZVIne (presence of ultrasound). c ZVIcol (absence of ultrasound). d ZVIcol (presence of ultrasound)

2.3 Chemical Remediation of Groundwater Containing Hexavalent Chromium via Redox Process Using ZVI Under Batch Conditions

Comparative batch experiments were carried out in order to evaluate the redox performance for the remediation process exhibited by ZVIcol in comparison with the other iron species (Fe0G and ZVIne). A study concerning the performance of Fe(II) for the redox remediation process was also evaluated in the presence and in the absence of CMC (see Table 1). In this study a fixed volume of contaminated groundwater was added to the solutions containing the reductant in order to obtain the appropriate molar ratio to result in [Cr(VI)] = 1,000 mg l−1 (see Table 1).
Table 1

Experimental conditions adopted during chemical reduction of hexavalent chromium using different iron species


Reductant mass/g





Fe(II) + CMC






ZVIne—natural pH



ZVIne—pH 5



ZVIcol—natural pH



ZVIcol—pH 5







aStoichiometric ratios for the redox process: Cr(IV):Fe(II) = 1:3 and Cr(VI):Fe0 = 1:1.

Fe(II) solutions and ZVI suspensions were placed in an orbital shaker (120 rpm for 24 h). All suspensions were centrifuged (3,600 rpm for 10 min) and filtered using a cellulose acetate membrane (0.2 μm). The supernatant was characterized based on the parameters: pH, redox potential (EH), [Cr(VI)] and [Cr(total)].

2.4 Chemical Remediation of Soil Containing Hexavalent Chromium via Redox Process Using ZVI Under Semi-Batch Conditions

Semi-batch studies were carried out to investigate the overall removal rate of Cr(VI) under flow conditions using a packed bed column reactor containing the contaminated soil. A scheme representing the experimental setup is presented in Fig. 2.
Fig. 2

Scheme showing the experimental setup used in studies carried out using ZVI under semi-batch conditions

The vertically disposed column reactor (height: 5.5 cm, diameter: 3.55 cm and volume of 52.9 ml) was carefully filled with previously dried and homogenized contaminated soil. The density ρ(packed soil) was ≅1.4 ± 0.2 g ml−1, thus confirming reproducible conditions concerning the packing procedure was rather good.

Before the experiments, the packed soil samples were slowly saturated under down flow conditions using synthetic ground water (SGW) pumped at a volumetric flow rate (G) of 9.00 ± 0.60 ml h−1. SGW was synthesized to obtain the following composition (in mg l−1) (Seaman et al. 1999): 1.00 of Ca2+; 0.37 of Mg2+; 0.21 of K+; 1.40 of Na+ and 0.73 of SO42−.

The volume of SGW filling the permeable porous soil microstructure was defined as the pore volume (PV) of the packed soil (Seaman et al. 1999). In the light of this approach, an average PV of 25.5 ± 1.9 ml was obtained. Application of Darcy’s law using the falling head test (Domenico and Schwartz 1979) revealed a hydraulic conductivity of the contaminated soil of (3.78 ± 0.71) × 10−2 cm h−1.

Soil samples containing Cr(VI) were treated under semi-batch conditions using different suspensions of ZVIcol. The [Cr(VI)]:[ZVIcol] molar ratios of 1:4 and 1:8 were used throughout. These molar ratios were chosen considering a packed soil mass of 70.0 ± 3.2 g. In all cases, due to the instability of the redox suspension in an aerated medium (solution exposed to the atmosphere) the suspension was prepared immediately before its application.

ZVIcol suspensions were pumped at 18.0 ± 0.60 ml h−1. In all cases the suspension was pumped using a model Minipuls 3 Peristaltic Pump from Gilson. Aliquots of the aqueous phase were withdrawn at the outlet of the packed bed column as functions of the treatment time (see Fig. 2). These aliquots were filtered using a cellulose acetate membrane (0.45 μm) and measurements of pH, EH and [Cr(VI)] were promptly carried out.

2.5 Extraction of the Residual Hexavalent Chromium Present in Soil after the Redox Treatment Using ZVIcol in Semi-Batch Conditions

Three extractions for each soil sample were carried out by gradually increasing the strength of the extractor reagent (James et al. 1995). The first extraction was carried out using distilled water (pH 5.7) in order to obtain the soluble and labile chromium present in soil. The second one was carried out using phosphate buffer (pH 7.0) in order to determine the amount of chromium available in the exchangeable form (the more strongly adsorbed fraction of chromium present in soil). The fraction defined as total hexavalent chromium, Cr(VI)total, was obtained from a hot extraction procedure (80–90°C) using a solution containing NaOH and CaCO3 (pH > 12) as the extractor phase. The fraction of the hexavalent chromium defined as non-exchangeable (insoluble), which represents the amount of Cr(VI) present as a precipitate and/or chemically adsorbed in soil, was obtained by subtracting the amount of Cr(VI)labile plus Cr(VI)exchangeable from Cr(VI) total.

3 Results and Discussion

3.1 Chemical Reduction and Immobilization of Chromium Present in Groundwater Under Batch Conditions

A comparative remediation study using different iron species (Fe(II), ZVIne, ZVIcol and Fe0G) was carried out under batch conditions in order to evaluate their redox performance for chemical reduction of Cr(VI).

Analysis of the data presented in Table 2 reveals that lower pH values were obtained using the Fe(II) solution. In the other cases involving the different iron species one can observe that pH changed from 4 to 9. Analysis of the EH-values shows that ZVIcol presents better reducing conditions in the reaction medium.
Table 2

Parameters obtained in the redox treatment process carried out under batch conditions (t = 24 h; 120 rpm and 25°C)


Reacting medium

pH (final)

EH (final)/mV

[Cr(VI)]/mg l−1

[Cr(total)]/mg l−1

[Fe(total)]/mg l−1


Distilled water

6.00 ± 0.01

496 ± 30

932 ± 25

1,160 ± 20


Distilled water + CMC + NaBH4

5.90 ± 0.01

476 ± 30

908 ± 25

1,150 ± 20


Fe0G (30×)

6.10 ± 0.12

470 ± 18

957 ± 25

1,260 ± 23


Fe(II) (1×)

2.35 ± 0.15

636 ± 7


560 ± 10

740 ± 25


Fe(II) + CMC (1×)

2.30 ± 0.10

627 ± 5


640 ± 8

1,170 ± 28


ZVIne—pHnat (4×)

7.30 ± 0.05

390 ± 28

652 ± 15

640 ± 15


ZVIne—pH 5 (4×)

6.30 ± 0.06

471 ± 17

465 ± 13

520 ± 13


ZVIcol—pHnat (4×)

8.60 ± 0.06

299 ± 31

441 ± 10

550 ± 13


ZVIcol—pH 5 (1.5×)

3.90 ± 0.60

700 ± 35

297 ± 15

340 ± 17


ZVIcol—pH 5 (3×)

4.60 ± 0.60

380 ± 16



220 ± 15


ZVIcol—pH 5 (4×)

4.60 ± 0.60

375 ± 8



475 ± 21

It is desirable that the reduction of Cr(VI) present in soil and groundwater is accompanied by immobilization in soil of the corresponding insoluble reduced chromium species. The efficiency presented by ZVI for the remediation of groundwater containing Cr(VI) can be evaluated comparing the different findings shown in Fig. 3. A blank study using distilled water and also the CMC+NaBH4 mixture revealed that chemical reduction of Cr(VI) is negligible in the absence of iron species.
Fig. 3

Experimental findings (chromium fractions) obtained after application of ZVI for the reduction of groundwater containing Cr(VI)

According to Fig. 3, the remediation based on application of Fe0G does not result in reduction of Cr(VI) present in groundwater even after a remediation time of 24 h. Figure 3 also reveals that Fe(II) presented a higher efficiency for reduction for Cr(VI). However, in this case, considerable quantities of Cr(III) remained soluble in the aqueous phase (immobilization ∼50%). Results obtained for ZVIne at different pH (natural and 5), indicate that the reduced chromium was totally converted into insoluble compounds. On the contrary, other chromium fractions not reduced (around 45%) during treatment remained dissolved as Cr(VI).

A comparative study based on results presented in Fig. 3 clearly shows that application of ZVIcol (3 and 4×) comprises the more promising scenario for the remediation process, since in this case all fractions of Cr(VI) were totally converted into insoluble Cr(III) compounds.

Studies have shown that CMC can suppress formation of the insoluble hydroxides containing Cr(III) due to its stabilizing properties (Xu and Zhao 2007). In the present case (pH 5) the formation of a gelatinous precipitate (brown color) during remediation indicates the formation of an insoluble compound containing Cr(III)–Fe(III) and the hydrolyzed CMC. Therefore, one can assume that in the present case the use of CMC does not significantly suppress the formation of insoluble Cr(III) compounds. One can notice that precipitate formation was not observed for the other cases where CMC was used without further pH control.

According to the literature (He et al. 2007), the stabilizing properties presented by CMC ensure the ZVI nanoparticles do not undergo magnetic attraction (an agglomeration process), since CMC acts creating a protecting surface coverage for the ferromagnetic particles.

Despite the advantages discussed above regarding the use of CMC, the formation of a protecting surface coverage can results in an undesirable effect. The CMC coverage reduces the occurrence of the redox process taking place at the reactive Fe0/solution interface, since the surface coverage creates an impedance to the electron transfer from iron (donor) to chromium (acceptor) atoms. The occurrence of such an impedance by ZVIcol particles stabilized by CMC was shown by findings obtained when CMC was used without pH control (see Fig. 3).

An increase in the redox activity of ZVIcol was verified at low pH values. As sketched in Fig. 4, this result indicates that an increase in the H+ activity promotes a partial hydrolysis of CMC, thus leading to an increase in the redox activity of ZVIcol.
Fig. 4

Scheme representing the partial hydrolysis of CMC in acidic medium and the release of ZVI

The presence of an electron donor in the aqueous suspension containing ZVIcol can lead to a reduction in the amount of ZVIcol available for the redox process with Cr (VI). Therefore, a pH adjustment in order to the obtain more acid conditions should be carried out slowly and very carefully, since this procedure can lead to the following undesirable reaction (Xu and Zhao 2007; Ponder et al. 2000):
$${\text{Fe}}^0 _{\left( {{\text{col}}.} \right)} + 2{\text{H}}^{\text{ + }} _{\left( {{\text{aq}}} \right)} \to {\text{Fe}}^{2 + } _{\left( {{\text{aq}}.} \right)} + {\text{H}}_{2\left( {\text{g}} \right)} $$
Another redox process leading to oxidation of ZVIcol concerns the presence of dissolved oxygen in the water (Xu and Zhao 2007; Ponder et al. 2000):
$$2{\text{Fe}}^0 _{\left( {{\text{col}}.} \right)} + {\text{O}}_{2\left( {{\text{aq}}.} \right)} + 2{\text{H}}_2 {\text{O}} \to 2{\text{Fe}}^{2 + } _{\left( {{\text{aq}}.} \right)} + 4{\text{OH}}^ - _{\left( {{\text{aq}}.} \right)} $$

The occurrence of these side reactions to a considerable extent in the present study is supported by considering that the ZVIcol particles were not able to provide the chemical reduction of Cr(VI) according to the expected stoichiometric ratio (see Table 2—exp. 9).

A careful analysis of findings obtained under batch conditions revealed the best results regarding chemical reduction of Cr(VI) using ZVI were obtained using ZVIcol (3×), since in this case a quantitative reduction of Cr(VI) was verified followed by a total immobilization of the reduced chromium species (see Fig. 3). In this case the application of 1000 mg of ZVIcol resulted in the reduction of 284 mg of Cr(VI) present in groundwater.

As expected, the analysis of findings obtained for ZVIcol synthesized using ultra-sound (pH 5) revealed this reductant is more efficient for chemical reduction of Cr(VI) when compared with the ZVIne species also synthesized using ultra-sound.

3.2 Chemical Reduction and Immobilization of Chromium in Soil under Semi-Batch Conditions Using a Packed Bed Column

In order to reproduce as much as possible the field conditions, a dynamic reduction process was carried out in the laboratory using a packed bed column reactor and the reductant ZVIcol. The packed soil containing Cr(VI) was firstly saturated using SGW. After that, the suspension containing ZVIcol was applied at 18.0 ± 0.60 ml h−1.

Figure 5 shows the experimental findings obtained during application of ZVIcol using the [Cr(VI)]:[ZVIcol] molar ratios of 1:4 and 1:8. The dynamic behavior of [Cr (VI)] and pH as functions of PV was investigated.
Fig. 5

Dependence of [Cr(VI)] and pH on PV. Molar ratio: a [Cr(VI)]:[ZVIcol] = 1:4 and b [Cr(VI)]:[ZVIcol] = 1:8. G = 18.0 ± 0.60 ml h−1

Analysis of Fig. 5 reveals that ∼3.0 and 1.5 PV-values are necessary for reducing the initial Cr(VI)-concentration from ∼300 mg l−1 down to 0.05 mg l−1, using the [Cr(VI)]:[ZVIcol] molar ratios of 1:4 and 1:8, respectively.

A test using pure SGW (absence of redox reaction) indicated that more than 35 PV are necessary in order to achieve the MCV via a leaching process (see Fig. 5). Analysis of the dynamic behavior of pH showed that the application of ZVIcol leads to a sudden decrease in pH (4.5 to 2.9) when the condition [Cr(VI)] ≅ MCV is achieved.

The pH changes during redox remediation carried out under semi-batch conditions can be attributed to different processes (Calder 1988; Palmer and Wittbrodt 1991; Kimbrough et al. 1999): (1) chemical reduction of Cr(VI); (2) hydrolysis of Cr(III) and/or Fe(II) species resulting in the formation of insoluble compounds and (3) presence of an excess of the reductant in the reaction medium.

3.2.1 Overall Removal Rate of Hexavalent Chromium Present in Soil via Reductive Process Using ZVIcol: A Kinetic Study

Kinetic studies concerning models for remediation of soil containing Cr(VI), under semi-batch conditions, are rather scarce. As previously proposed, the overall removal rate for Cr(VI) present in soil carried out via redox and/or leaching processes can be described using the pseudo-first order kinetic equation (Franco et al. 2008):
$$\ln \left( {{{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]} \mathord{\left/{\vphantom {{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]} {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]}}} \right.\kern-\nulldelimiterspace} {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]}}_0 } \right) = - k_{{\text{obs}}}^* t$$
where [Cr(VI)] and [Cr(VI)]0 are the instantaneous and the initial Cr(VI) concentrations (mg L-1), respectively; t is the remediation time (h) and kobs* is the apparent overall kinetic rate constant representing the overall removal rate for remediation (h−1).
The kinetic model representing the overall removal rate of Cr(VI) from soil, under semi-batch conditions (ν > 0), can be described by the following rate equation (Franco et al. 2008):
$$\ln \left( {{{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]} \mathord{\left/{\vphantom {{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]} {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]}}} \right.\kern-\nulldelimiterspace} {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]}}_0 } \right) = - z\left( {k_{{\text{VD}}} } \right)^n \alpha \left( \nu \right)^{\text{X}} \gamma \left( {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]_0 \left[ {{\text{ZVIcol}}} \right]_0 } \right) \cdot t$$
where: z and n are constants representing the contribution of the spatial dispersion on remediation process; kVD is the volumetric mass dispersion coefficient for ZVIcol (h−1); ν and [ZVIcol]0 are the linear velocity of the fluid (cm h−1) and the initial concentration of the ZVIcol (mg l−1), respectively. Parameters α and X are empirical constants specific for the particular heterogeneous redox system (Franco et al. 2008). Also γ ≡ kSAas, where kSA is the specific reaction rate constant based on the surface area of the nanoparticles (l h−1 m−2) and as is the specific surface area of the nanoparticles (m2 g−1) (He et al. 2007).

Equation 4 can be simplified in two different cases: (1) in the absence of the redox process (when only the leaching process takes place) and (2) when the concentration of the reductant (ZVIcol) is present in excess at the soil/solution interface ([ZVIcol]surface ≫ [Cr(VI)]surface).

In the limiting case where only the leaching process takes place Cr(VI) is solely removed from soil due to convection imposed by the dynamic conditions of the circulation fluid (ν > 0) inside the soil microstructure. In this case n = 0, z = 1 and γ = 1, and Eq. 5 is obtained (Franco et al. 2008):
$$\ln \left( {{{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]} \mathord{\left/{\vphantom {{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]} {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]_0 }}} \right.\kern-\nulldelimiterspace} {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]_0 }}} \right) = - \alpha \left( \nu \right)^{\text{X}} t$$
where kobs* = α(ν)X.
In the other limiting case, the chemical redox process where mass transport of ZVIcol from the main stream (net flow direction) to the soil (active sites)/solution interface due to spatial dispersion can be considered negligible (n = 0 and z = 1), which is valid when the concentration of ZVIcol is in excess at the soil/solution interface, yields Eq. 6 (Franco et al. 2008):
$$\ln \left( {\frac{{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]}}{{\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]_0 }}} \right) = - \alpha \left( \nu \right)^{\text{X}} \gamma \left( {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)} \right]_0 \left[ {{\text{ZVIcol}}} \right]_0 } \right) \cdot t$$
where \(k_{{\text{obs}}} * = \alpha \left( \nu \right)^{\text{X}} \gamma \left( {\left[ {{\text{Cr}}\left( {{\text{VI}}} \right)_0 } \right.\left[ {{\text{ZVIcol}}} \right]_0 } \right)\).

Analysis of the kinetic data shows that the overall removal rate of Cr(VI) from soil follows a pseudo-first order kinetic model (r > 0.965).

Kinetic data regarding reduction of the Cr(VI) in soil using ZVIcol are presented in Table 3. According to these findings, the apparent overall kinetic rate constant obtained at shorter treatment times was denoted as kobs1* while the one verified at longer treatment times was denoted as kobs2*. A careful analysis of the kinetic data concerning the temporal dependence of kobs* for the different reducing solutions (see Table 3) permits to proposing that this complex kinetic behavior can be attributed to a combination of physical (leaching and mass transfer) and chemical (redox reaction) processes which, depending on the experimental conditions, can take place simultaneously. According to Table 3, best results for soil remediation were achieved applying ZVIcol at pH 5.
Table 3

Influence of the [oxidant]:[reductant] molar ratio and the pH on soil treatment

Reacting medium

Molar ratio


PV ([Cr(VI)] < MCV)

























The interaction between ZVIcol particles and Cr(VI), leading to chromium reduction, constitutes a heterogeneous electron transfer reaction (Zhang et al. 1998). Fundamental processes taking place during reduction of Cr(VI) at the surface of Fe0-particles are described in Fig. 6.
Fig. 6

Scheme representing the redox processes taking place at the ZVI surface

As indicated in Fig. 6, the ZVIcol particle acts as an electron donor similar to the process taking place during iron corrosion. This behavior explain the influence of the [oxidant]:[reductant] molar ratio on the remediation kinetics observed in the present case. Due to the heterogeneous nature of this redox process one can expect that the lower the average size of the ZVI particles for a given ZVI mass, the higher will be the rate of the chemical reaction inside the packed bed column reactor (packed soil).

3.2.2 Hexavalent Chromium Removal from Soil Using ZVIcol: A Reaction Scheme

Fundamental processes implicit in the kinetic model described by Eq. 6 can be alternatively represented using a reaction scheme (Franco et al. 2008). The scheme presented below represents the reduction of Cr(VI) present in soil by ZVIcol nanoparticles.
$$\begin{aligned} & \\ & \begin{array}{*{20}l} {{\begin{array}{*{20}c} {{ \equiv {\text{S}} - {\text{Cr}}{\left( {{\text{VI}}} \right)}_{{{\text{ads}}}} + {\text{H}}_{2} {\text{O}}}} & {{\xrightarrow{{k_{1} }}}} & {{ \equiv {\text{S}} - {\text{H}}_{2} {\text{O}} + {\text{Cr}}{\left( {{\text{VI}}} \right)}_{{{\text{bulk}}}} {\text{ }}{\left( {leaching} \right)}\quad \quad \quad \quad \;\;\;{\text{ }}\;{\text{ }}{\left( {\mathbf{I}} \right)}\quad }} \\ \end{array} } \hfill} \\ {{\begin{array}{*{20}c} {{ \equiv {\text{S}} - {\text{Cr}}{\left( {{\text{VI}}} \right)}_{{{\text{ads}}}} + {\text{ZVIcol}}_{{{\left( {{\text{ads}}} \right)}}} }} & {{\xrightarrow{{k_{2} }}}} & {{ \equiv {\text{S}} - {\text{Cr}}{\left( {{\text{III}}} \right)} - {\text{Fe}}{\left( {{\text{III}}} \right)}{\text{ }}{\left( {redox\,\,reaction} \right)}\quad \quad {\text{ }}{\left( {{\mathbf{II}}} \right)}}} \\ \end{array} } \hfill} \\ {{\begin{array}{*{20}c} {{ \equiv {\text{S}} - {\text{Cr}}{\left( {{\text{III}}} \right)} - {\text{Fe}}{\left( {{\text{III}}} \right)} + w{\text{H}}_{2} O}} & {{\xrightarrow{{k^{\prime }_{3} - {\text{pH}} \geqslant 3}}}} & {{ \equiv {\text{S}} - w{\text{H}}_{2} {\text{O}}}} \\ \end{array} + {\text{Cr}}{\left( {{\text{III}}} \right)} + {\text{Fe}}{\left( {{\text{III}}} \right)}\;{\left( {release} \right)}\quad \;{\text{ }}{\left( {{\mathbf{IIIa}}} \right)}} \hfill} \\ {{\begin{array}{*{20}c} {{ \equiv {\text{S}} - {\text{Cr}}{\left( {{\text{III}}} \right)} - {\text{Fe}}{\left( {{\text{III}}} \right)} + w{\text{H}}_{2} {\text{O}}}} & {{\xrightarrow{{k^{{\prime \prime }}_{3} - {\text{pH}} \leqslant 3}}}} & {{ \equiv {\text{S}} - {\left( {{\text{Cr}}_{x} {\text{Fe}}_{{x - 1}} {\text{OH}}} \right)}_{{y{\left( {\text{s}} \right)}}} + y{\text{H}}^{ + } {\left( {immobilization} \right)}{\text{ }}{\left( {{\mathbf{IIIb}}} \right)}}} \\ \end{array} } \hfill} \\ {{{\mathbf{where }} \equiv {\mathbf{S is an active surface site present in the soil microstructure}}{\mathbf{.}}} \hfill} \\ \end{array} \\ \end{aligned}$$

The reaction scheme points out an ideal remediation process involving the chemical reduction of Cr(VI) and the immobilization of Cr(III) as an insoluble mixed compound (step IIIb) using ZVIcol as the reducing agent must provide an experimental scenario where the redox reaction (step II) is more important than the leaching process (k1 << k2). According to this theoretical analysis, the remediation process also must ensure that only a negligible release of the adsorbed ≡S–Cr(III)–Fe(III) fraction will takes place (k3″ >> k3′). Under these idealized conditions, Cr(VI) can be chemically reduced at the soil microstructure and quantitatively converted into an insoluble mixed hydroxide (chromium immobilization in soil).

The “in situ” application of a suspension containing ZVIcol for soil remediation can present an advantage when compared with the alternative of using Fe(II), since the presence of an excess of the ZVIcol particles in the soil microstructure can provide an extra reducing power (Cao and Zhang 2006). As a consequence, the reduced chromium does not undergo re-oxidation and, therefore, the occurrence of chromium leaching in the aquifer zone can be considerably minimized.

3.2.3 Determination of the Hexavalent Chromium Present in Soil after the Remediation Process Carried Out Using ZVIcol Under Semi-Batch Conditions

Figure 7 shows the different fractions of Cr(VI) determined before and after the reduction process using ZVIcol. It can be seeing that, after soil remediation using ZVIcol, only ∼2.5% of Cr(VI)total remained in the soil as Cr(VI)residual. One can also notice that neither labile nor exchangeable Cr(VI) fractions were detected after the remediation process.
Fig. 7

Different fractions of Cr(VI) determined before and after the redox treatment process using ZVIcol

Analysis of Fig. 7 indicates application of the ZVIcol can be efficient for reduction of the Cr(VI) fractions weakly adsorbed in soil microstructure (labile and exchangeable fractions). It was also found in this study that a considerable amount of the Cr(VI) fraction denoted as non-exchangeable (the fraction more strongly attached in the soil microstructure) was removed from this porous medium. As a result, an efficient immobilization of the reduced Cr(III) species was obtained.

One can argue based on the discussion presented above that application of ZVIcol for soil remediation can constitute a promising alternative technology for the “in situ” treatment of contaminated sites containing Cr(VI), not only due to the immobilization of Cr(III), but also due to the fact that the colloidal solution can be injected in both saturate an non-saturated soils with much higher efficiency than that using granular zerovalent iron.

4 Conclusions

Treatment of soil containing hexavalent chromium was investigated under batch and semi-batch conditions using different zerovalent iron forms: stabilized (ZVIcol) and non-stabilized (ZVIne).

These findings led to the following conclusions:
  1. (1)

    ZVIcol reduces the Cr(VI) present in groundwater and soil. A comparison between the results presented by ZVIcol and ZVIne showed the stabilization of Fe0 nanoparticles using carboxymethyl cellulose (CMC) ensures a high redox performance for ZVIcol particles;

  2. (2)

    Studies carried out using a packed bed column reactor (semi-batch conditions) revealed that application of ZVIcol can constitute a promising alternative technology for “in situ” remediation of contaminated sites containing hexavalent chromium;

  3. (3)

    Application of ZVIcol for reduction of Cr(VI) under optimized experimental conditions for the “in situ” remediation process is achieved using a 1[Cr(VI)]:4[ZVIcol] molar ratio and pH 5. In this case ZVIcol particles present a high mobility in the soil microstructure and show good reactivity with the Cr(VI).



The authors wish to thank the National Scientific Council for Research and Development (CNPq—Brazil) and Dr. Carol Collins for her technical assistance with the English.



Maximum concentration value for Cr(VI)


Synthetic groundwater


Non-stabilized zerovalent iron


Colloidal zerovalent iron stabilized using CMC


Granulated zerovalent iron


Ferrous sulfate solution


Carboxymethyl cellulose


Pore volume (−)


Volumetric flow rate (ml min−1)


Linear velocity of the fluid (cm h−1)


Volumetric mass dispersion coefficient (h−1)


Pseudo-first order kinetic rate constant for the redox reaction (h−1)


Overall pseudo-first order kinetic rate constant for the redox reaction (h−1)

Copyright information

© Springer Science+Business Media B.V. 2008