Environ. Sci. Technol. 2000, 34, 438-443
Trapping of Cr by Formation of Ferrihydrite during the Reduction of Chromate Ions by Fe(II)-Fe(III) Hydroxysalt Green Rusts S T EÄ P H A N I E L O Y A U X - L A W N I C Z A K , † P H I L I P P E R E F A I T , * ,§,# JEAN-JACQUES EHRHARDT,† PAUL LECOMTE,‡ AND J E A N - M A R I E R . G EÄ N I N § Laboratoire de Chimie Physique pour l’Environnement, UMR 7564 CNRS - Universite´ Henri Poincare´-Nancy 1, Equipe de Chimie des Surfaces, Equipe sur la Re´activite´ des Espe`ces du Fer, and De´partement Mate´riaux et Structures, ESSTIN, Universite´ Henri Poincare´-Nancy 1, 405 rue de Vandoeuvre, F-54600 Villers-Le`s-Nancy, France, and Centre National de Recherche sur les Sites et Sols Pollue´s, 930 bd Lahure, BP 537, F-59505 Douai Cedex, France
Hexavalent chromium, a byproduct of many industrial processes, is toxic and produces mobile aqueous oxyanions, whereas Cr(III) is relatively immobile in the environment and, moreover, essential in human glucidic metabolism. For this reason, Fe(II)-Fe(III) layered double hydroxysalt green rusts, recently identified as a mineral in hydromorphic soils, were evaluated as potential Fe(II)-bearing phases for hexavalent chromium reduction. Both considered synthetic varieties, the hydroxysulfate GR(SO42-) and the hydroxychloride GR(Cl-), proved to be very reactive; their interaction with potassium chromate solutions leads to the rapid and complete reduction of Cr(VI) into Cr(III). The Cr(III)bearing solid phase, studied by X-ray diffraction, Mo¨ ssbauer, X-ray photoelectron, and Raman spectroscopies, was determined to be a poorly ordered Cr(III)-Fe(III) oxyhydroxide, similar to the “2 the line ferrihydrite”. The comparison between the experimental redox potential and pH values for a theoretical equilibrium diagram bearing Cr and Fe phases indicated that the solubility of this solid solution, which may govern the behavior of chromium in the environment, is of the same order as that of pure Cr(OH)3.
Introduction Chromium is involved in many industrial processes, e.g. in metallurgy (alloys, corrosion inhibitors), electroplating, painting pigments and chemical compounds, leather tanning, ...; therefore, it has become a common contaminant of soils and ground and surface waters (1, 2). It is found most of the * Corresponding author phone: +33 (0)5-46-45-82-17; fax: +33 (0)5-46-45-72-72; e-mail:
[email protected]. †Equipe de Chimie des Surfaces, Universite ´ Henri Poincare´-Nancy 1. §Equipe sur la Re ´ activite´ des Espe`ces du Fer and De´partement Mate´riaux et Structures, ESSTIN, Universite´ Henri Poincare´-Nancy 1. ‡Centre National de Recherche sur les Sites et Sols Pollue ´ s. #Present address: Laboratoire d’Etudes des Mate ´ riaux en Milieux Agressifs, Universite´ de La Rochelle, avenue Marillac, F-17042 La Rochelle Cedex 1, France. 438
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 3, 2000
time as Cr(III) or Cr(VI), its most stable oxidation states. Cr(III) is essential in human glucidic metabolism, and, since its oxides and hydroxides are characterized by very low solubility products (3, 4), it is relatively immobile in the environment. In contrast, Cr(VI) compounds are known to be toxic and carcinogenic (2, 5), rather soluble, and produce aqueous Cr(VI) oxyanions which can move easily through the aquifer. Experiments, evaluating the capacity of natural soil systems for reducing Cr(VI) into Cr(III), have suggested the responsibility of Fe(II)-bearing minerals and organic matter (6-8). Several species have already been considered for laboratory experiments, such as Fe(II) sulfide (9), magnetite Fe3O4 (10), and Fe2+aq (11-14). Metallic R-iron, Fe(0), was also investigated in terms of a possible component for reactive walls (15, 16). In any case, the capacity of Fe(0) and Fe(II) to reduce Cr(VI) was confirmed, but the analyses of the resulting Cr(III)-bearing solids were omitted or not achieved. However, the nature of these compounds is of outmost importance since their physicochemical properties will govern the subsequent behavior of Cr in the environment. They were briefly described from X-ray diffraction (XRD) experiments as amorphous (13) and assumed to be Fe(III)Cr(III) hydroxides (11). Even though such solid solutions were already known, they were usually prepared by precipitation from Cr(III) and Fe(III) in alkaline media (e.g. refs 17 and 18), and the compounds resulting from Cr(VI) reduction may be different, since they can be topotactically related to the initial Fe(II)-bearing phases (19). Green rust Fe(II)-Fe(III) compounds, GRs, are double layered hydroxides built upon Fe(OH)2-like sheets containing from 25 to 66% of Fe(III) which alternates with interlayers composed of anions and water molecules. The exact nature of the GR depends on the interlayer anion, and various forms have been synthesized and studied (20-27), e.g. GR(CO32-), GR(Cl-), and GR(SO42-). Whatever their form, they are very reactive, and their ability to reduce pollutants such as NO3or CCl4 has already been investigated (28, 29). Finally, a GR mineral assumed to be GR(OH-) was identified recently in hydromorphic soils (30-32). In this article, which completes a preliminary approach (19), the interactions between Cr(VI) and GRs, namely GR(SO42-) and GR(Cl-), were thoroughly investigated. In particular, the influence of the Fe(II)/Cr(VI) ratio was described, and the nature of the resulting Cr(III)-bearing solid was fully investigated.
Experimental Methods Chemical Processes. Aqueous suspensions of the Fe(II)Fe(III) hydroxysalts green rusts, that is GR(SO42-) and GR(Cl-), were prepared according to the methodology previously developed (23, 24, 26, 27, 32). It involved the oxidation of Fe(II) hydroxide at 25 °C in aerated aqueous solutions containing the appropriate anions. Fe(OH)2 was first precipitated from 0.12 mol L-1 FeSO4‚7H2O or FeCl2‚4H2O and 0.2 mol L-1 NaOH solutions. All chemical products were provided by Prolabo and verified a 98% minimum purity. The suspension, maintained at a constant temperature of 25 ( 0.5 °C, was stirred vigorously to ensure homogeneous oxidation. The process was monitored by recording the pH and the redox potential Eh of a platinum electrode, using the saturated calomel electrode as a reference. The oxidation led first to the complete transformation of Fe(OH)2 to the GR compound, a process that resulted in a sharp increase of Eh and a sharp decrease of pH. At the very time, when such variations of Eh and pH occurred, a 50 mL K2CrO4 solution was added to the GR 10.1021/es9903779 CCC: $19.00
2000 American Chemical Society Published on Web 12/31/1999
suspension. The amount of chromate ions provided to the system was expressed in terms of the initial Fe(II)/Cr(VI) ratio, and the values 3, 6, and 12 were considered. For Fe(II)/Cr(VI) ) 3, the electron balance is met. Consequently, for values larger than 3, Fe(II) is found in excess. Solid-Phase Analyses. Transmission Mo¨ssbauer spectroscopy (TMS) was used to characterize the suspension immediately after chromate reduction. Therefore, the filtration of the solid, the preparation of the samples, and their setting in the Mo¨ssbauer apparatus were performed under inert nitrogen atmosphere to prevent oxidation of the residual Fe(II)-Fe(III) hydroxysalts. Analyses were performed at 15 K with a closed Mo¨ssbauer cryogenic workstation equipped with vibration isolation stand manufactured by Cryo Industries of America, by use of a constant acceleration Mo¨ssbauer spectrometer with a 50 mCi source of 57Co in Rh, calibrated with a 25 µm foil of R-Fe at room temperature. Further analyses of the solid phases were performed on the end products of the overall oxidation process, after the O2 from the air completed the oxidation of the remaining Fe(II) into Fe(III). X-ray diffraction (XRD), micro-Raman, and X-ray photoelectron spectroscopies could then be used. In any case, the products were filtered, air-dried, and ground to powder. X-ray diffractograms were recorded in the 1-50° range of 2θ using Mo KR1 radiation (λ ) 0.070 926 nm). The Raman spectra were obtained using a Jobin-Yvon T64000 triple monochromator and a charge coupled device (CCD) detector. Excitation of the samples was carried out with 632.8 nm radiation from a Spectra Physics 2017 argon ion laser. The Raman spectra were obtained via a confocal microscope (objective ×50; numerical aperture 0.55; spatial resolution of about 3 µm) in a backscattering geometry. The spectral resolution was 2 cm-1 with a precision of about 0.3 cm-1 on the Raman wavenumber. Surface analyses were done by X-ray photoelectron spectroscopy (XPS) using Mg KR radiation (1253.6 eV); charge effects were corrected with C(1s) binding energy of the contamination carbon (284.6 eV) for reference. Chemical analyses of the end products were finally performed by inductively coupled plasma-atomic emission spectroscopy (ICP-AES). The solid phases were dissolved using a triacid (HF + HClO4 + HCl) method. Iron and Cr concentrations were calibrated by means of two series of four standard solutions of FeSO4‚7H2O and K2CrO4 dissolved in 5% HCl solutions.
Experimental Results Eh and pH vs Time Curves. The curves presented in Figure 1 describe the formation of GR(SO42-) from Fe(OH)2. The thermodynamic meaning of such Eh(t) and pH(t) curves has been discussed and established recently (24). Each plateau corresponds to the electrochemical equilibrium between the initial compound and its product of oxidation. Then, the parameters measured on the first plateau verify the equilibrium conditions between Fe(OH)2 and GR(SO42-). The sharp variations of these parameters testify of the disappearance of the initial phase. At time tg, Fe(OH)2 has been totally consumed, and this moment was chosen for adding the Cr(VI) solution. The second part of the curve, after tg, involved the subsequent action of chromate ions, possibly followed by the oxidizing action of atmospheric O2. Whatever the value of ratio Fe(II)/Cr(VI), the interaction of GR(SO42-) with CrO42- resulted in an instantaneous increase of Eh and pH. For Fe(II)/Cr(VI) ) 3, Eh went from about -0.5 to +0.27 V (referred to the standard hydrogen electrode), whereas pH increased from about 8 to 10.1. In this case, both parameters stabilized rapidly, indicating that Fe(II) has been totally consumed and oxidized into Fe(III) during the reaction. For Fe(II)/Cr(VI) larger than 3, the action of Cr(VI) was followed by an oxidation process due to atmospheric O2. This second stage ended at time tf, and Eh and pH stabilized definitively thereafter.
FIGURE 1. Eh and pH vs time curves plotted during the formation of GR(SO42-) and its subsequent oxidation for various values of the initial Fe(II)/Cr(VI) ratio. Eh is given with respect to the standard hydrogen electrode (SHE). The arrows point out the addition of chromate ions, which is performed at time tg where the transformation of Fe(OH)2 into GR is achieved. tf is the overall oxidation time when a subsequent action of O2 takes place after Cr(VI) reduction. Mo1 ssbauer Spectroscopy of the Products Resulting from Adding Chromate Ions. The suspensions were filtered 40 s after adding the CrO4 solutions, then prepared, and immediately analyzed. The resulting Mo¨ssbauer spectra measured at 15 K are presented in Figure 2 and their computer fittings listed in Table 1. They display two types of components: magnetic sextets S1, S2, and S3, observed whatever the value of Fe(II)/Cr(VI), and paramagnetic quadrupole doublets D1 and D2, only observed for Fe(II)/Cr(VI) above 3. The presence of the various sextets must be interpreted as a distribution of hyperfine fields with a maximum absorption near 480 kOe due to a poor crystallinity rather than three distinct sites, since the other hyperfine parameters, isomer shift δ and quadrupole splitting ∆EQ, are identical. Such features indicate that the magnetically ordered compound is ferrihydrite (33), a poorly ordered form of Fe(III) oxyhydroxides. The hyperfine parameters of doublets D1 and D2 are characteristic of a GR compound at 15 K (24, 27, 34) and must be attributed to the remaining GR(SO42-) left unoxidized by the chromate ions. The relative intensity D1/D2, equal to the Fe(II)/Fe(III) ratio is about 2, in agreement with the composition, FeII4FeIII2(OH)12SO4‚nH2O (23, 24). The relative area of the GR components (D1 + D2) is found at 34 and 67% for the Fe(II)/Cr(VI) values of 6 and 12, respectively. As expected, it increases when the chromate amount decreases. VOL. 34, NO. 3, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
439
FIGURE 2. TMS spectra at 15 K of the products obtained after addition of chromate solutions to GR(SO42-) suspensions for Fe(II)/Cr(VI) ) 3, 6, and 12. • •• •• •: experimental curve; s: elementary components; - - - : global computed curve. Doublets D1 and D2 are not observed when Fe(II)/Cr(VI) ) 3; this implies that all the Fe(II) gets oxidized by Cr(VI) in the form of the ferrihydrite-like oxyhydroxide. Identification of the End Products. XRD and Raman spectra of the end products obtained for Fe(II)/Cr(VI) ) 3 and 12 are presented in Figure 3. XRD patterns look different. For Fe(II)/Cr(VI) ) 3, only two broad diffraction lines can be distinguished, corresponding to interplanar distances of about 0.255 and 0.148 nm, typical of the so-called “2-line ferrihydrite” (35). Ferrihydrite is the least crystalline of the hydrous Fe oxides and oxyhydroxides, and even its stoichiometry, often assumed at Fe5HO8‚4H2O (36), is somewhat uncertain. The XRD pattern varies with crystallinity; poorly crystalline samples lead only to two broad diffraction peaks, whereas more crystalline samples can display up to six peaks. In contrast, the pattern obtained for Fe(II)/Cr(VI) ) 12 seems only composed of the diffraction lines of R-FeOOH, goethite (37). This Fe(III) oxyhydroxide is a usual byproduct of the aerial oxidation of GRs in solution (22, 24, 38) and is obtained here from the fraction of GR(SO42-) that was left unoxidized by Cr(VI). Some broad and small peaks of ferrihydrite are more likely overlapping within the angular regions where the (021), (040) and (002), (250), (320) and (061) peaks of R-FeOOH are found. The literature concerning Raman studies of ferrihydrite is scarce, and no reference could be found, but since both XRD and TMS analyses designate unambiguously the product obtained when Fe(II)/Cr(VI) ) 3 as ferrihydrite, the corresponding Raman spectrum must be considered as that of ferrihydrite or, more specifically, “2-line ferrihydrite”. As a 440
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 3, 2000
matter of fact, this spectrum is also composed of only two broad bands, at 415 and 705 cm-1. The Raman spectrum of the end product at Fe(II)/Cr(VI) ) 12 displays the bands characterizing R-FeOOH (39, 40), but that at 705 cm-1 is still visible. This shows that the ferrihydrite which results from Cr(VI) reduction did not evolve during the second oxidation stage due to O2. It is noticeable that the only detected bands are those which characterize Fe(III) compounds. More precisely, neither the band of Cr(OH)3 at 525 cm-1 (41) nor those of Cr2O3 at about 300, 350, 550, and 610 cm-1 (42) are observed, implying that Cr(III) is incorporated in each case in ferrihydrite. To confirm it, micro-Raman analyses were performed on various spots of the samples; they always gave the same spectrum. This homogeneity was also confirmed by the optical observations through the microscope. Analyses of Chromium in the End Products. The mass balance analyses of Fe and Cr performed by ICP-AES on the final products, listed in Table 2, show the presence of Cr inside the solid matrix. This is confirmed by XPS, which is also a convenient tool for discriminating Cr(VI) from Cr(III) at the surfaces. The binding energy of the Cr(2p3/2) line is generally reported at 577-577.5 eV in Cr(III) oxyhydroxides and hydroxides, while it is found at about 580 eV in Cr(VI) salts such as Na2CrO4 or Na2Cr2O7 (43-45). Moreover, the spin-orbit split between 2p1/2 and 2p3/2 peaks is found at 9.7-9.9 eV in Cr(III) compounds and 8.7-9.4 eV in Cr(VI) compounds (44). The Cr(2p) spectra of the end products obtained for Fe(II)/Cr(VI) ) 3 and 12 are presented in Figure 4. In every case, each Cr(2p3/2) or Cr(2p1/2) line could be adjusted with only one component, implying that almost all Cr atoms (>95%) were in the same oxidation state. The corresponding Cr(2p3/2) binding energies are found close to 577.2 eV, and the 2p spin-orbit splits at 9.8 eV; these values are characteristic of Cr(III). The Case of GR(Cl-). Similar investigations were performed using the chloride containing green rust. GR(Cl-) is characterized by an Fe(II)/Fe(III) ratio which varies with the Cl- concentration of the medium (27, 46). In the conditions chosen here, it is close to 2.3 (46), leading to a chemical composition of about FeII2.3FeIII(OH)6.6Cl‚nH2O. Various results are summarized in Figure 5, demonstrating that the mechanisms are unchanged. As illustrated by the Mo¨ssbauer spectrum for Fe(II)/Cr(VI) ) 3, ferrihydrite remained the Fe(III) compound which results from the oxidizing action of chromate ions. XPS analyses for Fe(II)/Cr(VI) ) 6 demonstrated that chromium was completely reduced as Cr(III). Finally, no Cr(III) phase was observed by Raman spectroscopy besides the ferrihydrite-like compound (broad and minute band at 730 cm-1) and the lepidocrocite, formed by the subsequent action of dissolved oxygen. This implies that Cr atoms are present in the ferrihydrite compound along with Fe(III). Finally, the Fe/Cr ratios measured by ICP-AES confirmed the presence of chromium inside the solid matrix (Table 2). In this case, experimental values are found larger than theoretical ones; it may be due to a partial oxidation of the dissolved Fe(II) that increases the Fe(III) content of the solid.
Discussion Whatever the considered GR, GR(SO42-), or GR(Cl-), the reduction of Cr(VI) resulted in the formation of Cr(III)substituted ferrihydrite. The nature of the initial GR proves to have no influence on the process, an outcome mainly due to the structural properties which define these compounds. This behavior could be extended to the GR mineral, assumed to be the hydroxyl form, GR(OH-) (30-32), more closely related to GR(Cl-) than to any other GR compound because of the likeness between Cl- and OH- anions. The crystal structure of GRs consists of positively charged [FeII(1-x)FeIIIx(OH)2]+x hydroxide sheets separated by negatively charged
TABLE 1. Hyperfine Parameters Measured at 15 K by TMS of the Products Obtained from GR(SO42-) Immediately After the Addition of the Chromate Solutionsa Fe(II)/Cr(VI) ) 3 site
∆E Q
δ
D1 D2 S1 S2 S3
0.50(1) 0.49(1) 0.46(1)
0 0 0
Fe(II)/Cr(VI) ) 6
Fe(II)/Cr(VI) ) 12
H
RA
δ
∆ EQ
H
RA
δ
∆ EQ
H
RA
489(1) 456(1) 412(1)
0 0 53(2) 34(2) 13(2)
1.34(1) 0.51(1) 0.48(1) 0.49(1) 0.48(2)
2.87(1) 0.42(1) 0 0 0
0 0 492(1) 457(1) 414(2)
23(1) 11(1) 28(2) 26(2) 12(2)
1.34(1) 0.52(1) 0.47(1) 0.50(1) 0.41(2)
2.85(1) 0.40(1) 0 0 0
0 0 484(1) 444(1) 392(2)
46(1) 21(1) 14(2) 14(2) 5(2)
a δ (mm s-1), isomer shift taking R-iron as a reference. ∆E (mm s-1), quadrupole splitting. H (kOe), hyperfine field. RA (%) relative abundance. Q Errors on the last digit are given in parentheses.
FIGURE 3. X-ray diffraction and Raman spectra of the end products obtained from GR(SO42-) for Fe(II)/Cr(VI) values of 3 and 12.
TABLE 2. Atomic Ratio (Fe/Cr) Measured by ICP-AES in Various End Products and Compared to the Theoretical Valuesa Fe(II)/Cr(VI) initial product 2-)
GR (SO4 GR (Cl-)
3
6
12
Fe/Cr:exp. theor. Fe/Cr:exp. theor. Fe/Cr:exp. 4.1 5.2
4.5 4.3
7.5 10.5
8.25 7.96
14.3 12.4
theor. 15.75 11.61
a Five experiments were made on some samples, revealing that the error could be up to (10%. Computed from eqs 1-3 for GR(SO42-) and on the basis of similar equations for GR(Cl-) with composition of FeII2.3FeIII(OH)6.6Cl‚nH2O.
interlayers which restaure neutrality, [x/nA-n‚m/nH2O]-x. Formation of Cr-substituted ferrihydrite implies the departure of the interlayer anions from the solid, the insertion of Cr atoms, and the subsequent rearrangement of the hydroxide sheets. Such a drastic perturbation could only lead to the most disordered form of Fe(III) oxyhydroxide, i.e., “2-line ferrihydrite”, which is an undefined random stacking of few sheets of Fe(O)6 octahedra where oxygen atoms can be found in O2-, OH-, or even H2O species [cf. ref 47 for detailed explanations]. Influence of the Fe(II)/Cr(VI) Ratio and Chemical Balance. The interaction between GRs and Cr(VI) proves to be independent of the initial Fe(II)/Cr(VI) ratio, and an amount of GR proportional to the amount of Cr(VI) is
FIGURE 4. Cr(2p) XPS spectra of the end products obtained from GR(SO42-) for Fe(II)/Cr(VI) ) 3 and 12. • •• •• •: experimental curve; s: background and global computed curve. transformed while the excess of GR stays unchanged and can be involved in various redox processes afterward. This can be ascertained by the chemical balance, e.g., in the case VOL. 34, NO. 3, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
441
FIGURE 5. Survey of the results obtained with respect to GR(Cl-): TMS analysis at Fe(II)/Cr(VI) ) 3, XPS at Fe(II)/Cr(VI) ) 6 and Raman study at Fe(II)/Cr(VI) ) 12. The bands denoted L are those of lepidocrocite γ-FeOOH (39, 40). of GR(SO42-). First, the transformation of the GR into Crsubstituted ferrihydrite must be considered. By omitting the crystal water molecules of both the GR and ferrihydrite structures, the reaction at stoichiometry, i.e., with Fe(II)/ Cr(VI) ) 3, is as follows:
15FeII4FeIII2(OH)12SO4 + 20CrO42- f 2-
22Fe45/11Cr10/11HO8 + 15SO4
-
+ 10OH + 74H2O (1)
This reaction is accompanied by a production of OH- ions, in agreement with the increase of pH observed experimentally. Moreover, an Fe/Cr ratio of 4.5 is predicted inside the ferrihydrite-like compound, close to the value of 4.1 found by means of ICP-AES (Table 2). When Fe(II)/Cr(VI) is above 3, some GR(SO42-) remains after the Cr(VI) reduction and is oxidized by dissolved oxygen according to the reaction (24)
FeII4FeIII2(OH)12SO4 + 3/4O2 f 5R-FeOOH + Fe2+ + SO42- + 7/2H2O (2) The overall oxidation process involves successively reactions 1 and 2. For Fe(II)/Cr(VI) ) 6, the action of chromate must be written as
30FeII4FeIII2(OH)12SO4 + 20CrO42- f
22Fe45/11Cr10/11HO8 + 15SO42- + 10OH- + 74H2O + 15FeII4FeIII2(OH)12SO4 (3)
442
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 3, 2000
At this time, there would be 90 Fe atoms in the Cr-substituted ferrihydrite, whereas 90 others would remain in the GR. The GR oxidizes thereafter into R-FeOOH according to eq 2. The corresponding 90 Fe atoms are split into 75 atoms inside FeOOH and 15 as Fe2+aq left in solution. The Fe/Cr ratio of the end product would be 165/20, i.e., 8.25, close to the value 7.5 determined by ICP-AES. Similarly, the chemical balance at Fe(II)/Cr(VI) ) 12 predicts an Fe/Cr ratio of 15.75 inside the end products, close to the experimental determination of 14.3. The relative areas of the Mo¨ssbauer spectral components of ferrihydrite and green rust are similar to the theoretical values. For Fe(II)/Cr(VI) ) 6, eq 2 implies that 90 Fe atoms are trapped in the ferrihydrite-like compound, whereas 90 Fe atoms remain in the GR compound. It is found 66/34 instead of the theoretical 50/50 proportion, and, similarly, 33/67 instead of 25/75 for Fe(II)/Cr(VI) ) 12. This discrepancy is due to the relative Mo¨ssbauer factors fA/B, linking the proportions of resonant 57Fe atoms to the areas of the Mo¨ssbauer components, which may differ from unity. If similar compounds A and B were considered, this factor is close to 1, for instance 1.22 ( 0.20 if A is R-FeOOH goethite and B is γ-FeOOH lepidocrocite (48). If the compositions and structures of compounds A and B differ, this factor can be up to 3; this is the case for FeCl2‚4H2O and metallic R-Fe which give fA/B ) 2.96 ( 0.27 (48). The spectra obtained here allowed us to estimate the fFH/GR relative factor of Crsubstituted ferrihydrite and GR(SO42-) at 1.68 ( 0.26. The Cr(III)-Bearing Compound. The Fe/Cr ratio of the Cr-substituted ferrihydrite is constant at 9/2 whatever the initial Fe(II)/Cr(VI) ratio, in view of eqs 1 and 3. However, this value depends on the type of GR due to the Fe(II)/Fe(III) ratio. The Fe(III) content of a GR may actually depend on the interlayer anion. Most of GRs are found with such a ratio set at 2, namely GR(CO32-) (21, 22) and GR(SO42-) (23, 24); however, some contain more Fe(II) with a ratio set at 3 such as the hydroxyoxalate GR(C2O42-) (49). Moreover, GR(Cl-) and the GR mineral fougerite proved to be exceptions with ratios varying in a large domain (27, 32, 46). When considering the GR mineral with a composition often met at FeIIFeIII(OH)5 (30-32), the reduction of chromate would lead to
15FeIIFeIII(OH)5 + 5CrO42- f
7Fe30/7Cr5/7HO8 + 29H2O + 10OH- (4)
The Cr-substituted ferrihydrite would then contain 6 Fe per 1 Cr instead of 4.5 per 1 in the compound issued from GR(SO42-). Thermodynamics. A thermodynamic approach of the behavior of the Cr(III)-substituted ferrihydrite was made by comparing the experimental Eh and pH values with the conditions given by the theoretical potential Eh-pH equilibrium diagram displayed in Figure 6 and drawn from selected thermodynamic constants (4, 32). In this graph, two diagrams were superimposed, one related to Fe and the other one to Cr. The Fe part was drawn considering the solid phases Fe, Fe(OH)2, GR (SO42-), and R-FeOOH, which, except Fe(0), intervene in the processes described here. The extent of the region of stability of GR(SO42-) depends on the activity of sulfate anions which was set, as an example, at 0.1. In the Cr part, only one solid phase appears, namely Cr(OH)3. Since thermodynamic constants concerning the Cr(III)-substituted ferrihydrite are not available, Cr(OH)3 was considered instead. For clarity, several simplifications were made. The lines related to the equilibrium conditions between two dissolved species (1′ to 6′) were not prolongated inside the region of insolubility of the solid phases, defined as the zone where the activity of the dissolved species in equilibrium with the solid is less than 10-6. The solubility of each solid phase was designated by a set of three lines corresponding to dissolved species activities of 10-2, 10-4, and 10-6. The equilibrium
FIGURE 6. Potential-pH equilibrium diagrams at 25 °C for iron in sulfate containing solution (a[SO42-] ) 0.1) and Cr(OH)3 in water. Consistent thermodynamic constants were selected from refs 4 and 32. The ringed crosses indicate the experimental conditions met at the end of the oxidation of GR(SO42-) for various Fe(II)/Cr(VI) values. conditions involving iron compounds were displayed as usual by solid lines, whereas chain-dotted lines were used for those involving Cr(OH)3. Ringed crosses represent experimental conditions at the end of oxidation for various Fe(II)/Cr(VI) values. They lay in two distinct regions, depending on the presence of goethite. For Fe(II)/Cr(VI) ) 6 and 12, when goethite is present, they are found in a zone where the solubility lines of goethite [14] and Cr(OH)3 [19] intersect, at a pH about 4.5. For Fe(II)/Cr(VI) ) 3, when goethite is absent, the ringed cross is located inside the set of lines representing the equilibrium between Cr(OH)3 and CrO42- anions. Since the experimental conditions appear to be similar to those implied by Cr(OH)3, the solubility of the Cr(III)-substituted ferrihydrite can be considered to be similar to that of the Cr(III)-hydroxide.
Acknowledgments We thank Ms. M. Lelaurain of the University Henri Poincare´Nancy 1 for analyzing the samples by XRD and Drs. B. Humbert and M. Abdelmoula for their help in the Raman and Mo¨ssbauer spectroscopy analyses. Rhodia, France, provided financial support. The advice of the reviewers is fully acknowledged.
Literature Cited (1) Richard, F. C.; Bourg, A. C. M. Water Res. 1991, 25, 807. (2) Puls, R. W.; Clark, D. A.; Paul, C. J.; Vardy, J. J. Soil Contam. 1994, 3, 203. (3) Rai, D.; Sass, B. M.; Moore, D. A. Inorg. Chem. 1987, 26, 345. (4) Beverskog, B.; Puigdomenech, I. Corros. Sci. 1997, 39, 43. (5) Nriagu, J. O.; Nieboer, E. Chromium in the natural and human environments; John Wiley & Sons: New York, 1988. (6) Bartlett, R. J.; Kimble, J. M. J. Environ. Qual. 1976, 5, 383. (7) Wittbrodt, P. R.; Palmer, C. D. Environ. Sci. Technol. 1992, 29, 255. (8) Eary, L. E.; Rai, D. Soil Sci. Soc. Am. J. 1991, 55, 676. (9) Patterson, R. R.; Fendorf, S.; Fendorf, M. Environ. Sci. Technol. 1997, 31, 2039.
(10) Peterson, M. L.; Brown, G. E., Jr.; Parks, G. A. Colloids Surf. A 1996, 107, 77. (11) Eary, L. E.; Rai, D. Environ. Sci. Technol. 1988, 22, 972. (12) Fendorf, S. E.; Li, G. Environ. Sci. Technol. 1996, 30, 1614. (13) Buerge, I. J.; Hug, S. J. Environ. Sci. Technol. 1997, 31, 1426. (14) Sedlak, D. L.; Chan, P. G. Geochim. Cosmochim. Acta 1997, 61, 2185. (15) Pratt, A. R.; Blowes, D. W.; Ptacek, C. J. Environ. Sci. Technol. 1997, 31, 2492. (16) Powell, R. M.; Puls, R. W., Hightower, S. K.; Sabatini, D. A. Environ. Sci. Technol. 1995, 29, 1913. (17) Amonette, J. E.; Rai D. Clays Clay Miner. 1990, 38, 129. (18) Schwertmann, U.; Gasser, U.; Sticher, H. Geochim. Cosmochim. Acta 1989, 53, 1293. (19) Loyaux, S.; Refait, Ph.; Lecomte, P.; Ehrhardt, J.-J.; Ge´nin, J.-M. R. Hydrol. Earth Syst. Sci., in press. (20) Taylor, R. M. Clay Miner. 1980, 15, 369. (21) Hansen, H. C. B. Clay Miner. 1989, 24, 663. (22) Drissi, S. H.; Refait, Ph; Abdelmoula, M.; Ge´nin, J.-M. R. Corros. Sci. 1995, 37, 2025. (23) Ge´nin, J.-M. R.; Olowe, A. A.; Refait, Ph.; Simon, L. Corros. Sci. 1996, 38, 1751. (24) Refait, Ph.; Bon, C.; Simon, L.; Bourrie´, G.; Trolard, F.; Bessie`re, J.; Ge´nin J.-M. R. Clay Miner. 1999, 34, 499. (25) Hansen, H. C. B.; Borggaard, O. K.; Sørensen, J. Geochim. Cosmochim. Acta 1994, 58, 2599. (26) Refait, Ph.; Ge´nin, J.-M. R. Corros. Sci. 1993, 34, 797. (27) Refait, Ph.; Abdelmoula, M.; Ge´nin, J.-M. R. Corros. Sci. 1998, 40, 1547. (28) Hansen, H. C. B.; Koch, C. B.; Nancke-Krogh, H.; Borggaard, O. K.; Sørensen, J. Environ. Sci. Technol. 1996, 30, 2053. (29) Erbs, M.; Hansen, H. C. B.; Olsen, C. E. Environ. Sci. Technol. 1999, 33, 307. (30) Trolard, F.; Abdelmoula, M.; Bourrie´, G.; Humbert, B.; Ge´nin, J.-M. R. Compt. Rend. Acad. Sci. IIA 1996, 323, 1015. (31) Trolard, F.; Ge´nin, J.-M. R.; Abdelmoula, M.; Bourrie´, G.; Humbert, B.; Herbillon, A. J. Geochim. Cosmochim. Acta 1997, 61, 1107. (32) Ge´nin, J.-M. R.; Bourrie´, G.; Trolard, F.; Abdelmoula, M.; Jaffrezic, A.; Refait, Ph.; Maıˆtre, V.; Humbert, B.; Herbillon, A. Environ. Sci. Technol. 1998, 32, 1058. (33) Murad, E.; Schwertmann, U. Am. Min. 1980, 65, 1044. (34) Ge´nin, J.-M. R.; Abdelmoula, M.; Refait, Ph.; Simon, L. Hyper Interact. (C) 1998, 3, 313. (35) Carlson, L.; Schwertmann, U. Geochim. Cosmochim. Acta 1981, 45, 421. (36) Towe, K. M.; Bradley, W. F. J. Colloid Interface Sci. 1967, 24, 384. (37) Szytula, A.; Burewicz, A.; Dimitrijevic, Z.; Krasnnicki, S.; Rzany, H.; Todorovic, J.; Wanic, A.; Wolski, W. Phys. Stat. Sol. 1968, 26, 429. (38) Detournay, J.; deMiranda, L.; De´rie, R.; Ghodsi, M. Corros. Sci. 1975, 15, 295. (39) Thibeau, R. J.; Brown, C. W.; Heidersbach, R. H. Appl. Spectrosc. 1978, 32, 532. (40) DeFaria, D. L.; Venaˆncio Silva, S.; deOliveira, M. T. J. Raman Spectrosc. 1997, 28, 873. (41) Melendres, C. A.; Pankuch, M.; Li, Y. S.; Knight, R. L. Electrochim. Acta 1992, 37, 2747. (42) Zuo, J.; Xu, C.; Hou, B.; Wang, C.; Xie, Y.; Qian, Y. J. Raman Spectrosc. 1996, 27, 921. (43) Allen, G. C.; Curtis, M. T.; Hooper, A. J.; Tucker, P. M. J. Chem. Soc., Dalton Trans. 1973, 1675. (44) Ikemoto, I.; Ishii, K.; Kinoshita, S.; Kuroda, H.; Franco, M. A. A.; Thomas, J. M. J. Solid State Chem. 1976, 17, 425. (45) Shuttleworth, D. J. Phys. Chem. 1980, 84, 1629. (46) Refait, Ph.; Ge´nin, J.-M. R. Corros. Sci. 1997, 39, 539. (47) Drits, V. A.; Sakharov, B. A.; Salyn, A. L.; Manceau, A. Clay Miner. 1993, 28, 185. (48) Meisel, W.; Kreysa, G. Z. Anorg. Allg. Chem. 1973, 395, 31. (49) Refait, Ph.; Charton, A.; Ge´nin, J.-M. R. Europ. J. Sol. State Inorg. Chem. 1998, 35, 655.
Received for review April 2, 1999. Revised manuscript received November 9, 1999. Accepted November 15, 1999. ES9903779 VOL. 34, NO. 3, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
443