Chemosphere 111 (2014) 169–176

Contents lists available at ScienceDirect

Chemosphere journal homepage: www.elsevier.com/locate/chemosphere

A XAFS study of plain and composite iron(III) and chromium(III) hydroxides N. Papassiopi a,⇑, F. Pinakidou b, M. Katsikini b, G.S.E. Antipas a, C. Christou a, A. Xenidis a, E.C. Paloura b a b

School of Mining and Metallurgical Engineering, National Technical University of Athens, Zografou Campus, Athens 15780, Greece School of Physics, Department of Solid State Physics, Aristotle University of Thessaloniki, Thessaloniki 54124, Greece

h i g h l i g h t s  Stability of plain and mixed Fe/Cr hydroxides depends on their structure.  XAFS studies were used to examine structure of typical remediation Cr/Fe products.  Plain Cr hydroxide was crystalline in XRD, but local order was undetectable in XAFS.  Mixed Fe/Cr hydrolysis or redox products resembled amorphous 2-line ferrihydrite.  Better arrangement of CrO6 octahedra was observed in the Fe/Cr hydrolysis product.

a r t i c l e

i n f o

Article history: Received 26 August 2013 Received in revised form 27 February 2014 Accepted 12 March 2014

Handling Editor: X. Cao Keywords: XANES EXAFS Cr(VI) remediation Cr(III) hydroxide Fe(III)–Cr(III) mixed hydroxides

a b s t r a c t Reduction of hexavalent Cr(VI) to the trivalent state is the common strategy for remediation of Cr(VI) contaminated waters and soils. In the presence of Fe the resulting compounds are usually mixed Fe(III)–Cr(III) phases, while, under iron-free conditions, reduction leads to formation of plain Cr(III) hydroxides. Environmental stability of these compounds depends on their structure and is important to understand how different precipitation conditions affect the local atomic order of resulting compounds and thus their long term stability. In current study, typical Cr(VI) environmental remediation products, i.e. plain and mixed Fe(III)–Cr(III) hydroxides, were synthesized by hydrolysis and redox reactions and their structure was studied by X ray diffraction and X ray absorption fine structure techniques. Plain Cr(III) hydroxide was found to correspond to the molecular formula Cr(OH)33H2O and was identified as crystalline in XRD. However, the same compound when examined by EXAFS did not exhibit any clear local order in the range of EXAFS detectable distances, i.e. between 0 and 5 Å. Namely, EXAFS spectroscopy detected only contribution from the first nearest neighboring (Cr–O) shell, suggesting that CrO6 octahedra interconnection is loose, in accordance with the suggested anti-bayerite structure of this compound. Mixed Fe(III)–Cr(III) systems resembled 2-line ferrihydrite irrespective of the synthesis route. Analysis of Fe-K-EXAFS and Cr-K-EXAFS spectra indicated that FeO6 octahedra are bonded by sharing both edges and corners, while CrO6 octahedra seem to prefer edge sharing linkage. EXAFS data also suggest that Fe– Cr hydroxide produced by hydrolysis presents a better arrangement of CrO6 octahedra compared to the redox product. Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction Classified as a confirmed human carcinogen, hexavalent chromium, Cr(VI), typically in the form of oxyanion, CrO2 4 , is one of the common pollutants in soils and aquifers. As Cr(VI) is highly soluble in the environment, its reduction to the relatively innocuous Cr(III) cation is an important remediation strategy. Cr(VI) reduction ⇑ Corresponding author. Tel.: +30 210 772 2298. E-mail address: [email protected] (N. Papassiopi). http://dx.doi.org/10.1016/j.chemosphere.2014.03.059 0045-6535/Ó 2014 Elsevier Ltd. All rights reserved.

is normally achieved either by inorganic agents (e.g. Na2S2O4, CaSx, Fe0, FeSO47H2O) (Eary and Rai, 1988; Puls et al., 1999; Dermatas et al., 2006; Kumpiene et al., 2006; Chrysochoou et al., 2010) or by microorganisms possessing a direct or indirect chromate reducing capacity, such as sulfate and iron reducing bacteria which produce strong reductants, like S2 or Fe2+, through their metabolism (Michel et al., 2001; Wielinga et al., 2001; Battaglia-Brunet et al., 2004; Papassiopi et al., 2008, 2009). For iron-free reductants the resulting compound is the Cr(OH)3 hydroxide (Battaglia-Brunet et al., 2004), while iron-based agents reduce Cr(VI) to a mixed

170

N. Papassiopi et al. / Chemosphere 111 (2014) 169–176

Fe(1x)Crx(OH)3 hydroxide (Eary and Rai, 1988; Puls et al., 1999; Wielinga et al., 2001). The environmental stability of the reduction products, taken to be inversely proportional to solubility in water reservoirs, reflects on the success of the remediation process. Studies of the Cr(OH)3 solubility over a wide pH range indicate good stability at neutral and alkaline pH and increased solubility at pH below 5 (Rai et al., 1987, 2002, 2004, 2007; Papassiopi et al., 2012). On the other hand, the solubility of Fe(1x)Crx(OH)3 mixed hydroxides has been established as orders of magnitude lower to that of plain Cr(OH)3, depending on the molar ratio between Fe and Cr (Sass and Rai, 1987; Papassiopi et al., 2012). The structure of plain Cr(III) hydroxide has been an issue of controversy between the researchers. The team of Rai et al. (1987, 2002, 2004, 2007) systematically investigated the thermodynamic properties of plain Cr(III) hydroxide and its solubility in aquatic systems of variable chemical composition. In all their studies the Cr(III) hydroxide was characterized as amorphous in X-ray diffraction (XRD). Many other researchers studying the mixed Fe–Cr hydroxides produced plain Cr(III) hydroxides as reference end product, which were also characterized as amorphous (Sass and Rai, 1987; Wielinga et al., 2001; Tang et al., 2010; Papassiopi et al., 2012). The molecular formula attributed to this amorphous compound coincides with that of anhydrous Cr(III) hydroxide, i.e. Cr(OH)3. Another group of researchers make reference to a crystalline compound, corresponding to the molecular formula of a trihydrate hydroxide, Cr(OH)33H2O (Giovanoli et al., 1973; Giovanoli and Stadelmann, 1973; Von Meyenburg et al., 1973; Spiccia and Marty, 1986; Allen and Poeppelmeier, 1994). According to Giovanoli et al. (1973) and Spiccia and Marty (1986), this compound is stable at room temperature, but upon aging in an aqueous suspension or after heating at relatively mild conditions, turns into an amorphous phase. The local atomic order of the amorphous Cr(III) hydroxide was studied applying XANES and EXAFS techniques by Charlet and Manceau (1992), Fendorf et al. (1994) and Rai et al. (2004, 2007). XANES spectra provide information about the oxidation state of the central atom. Contrary to Cr(III) compounds, the Cr-K-edge XANES of Cr(VI) containing compounds (e.g. K2Cr2O7) exhibit a characteristic pre-edge peak which is assigned to 1s ? 3d electronic transitions (Ankudinov et al., 1998). When Cr(VI) belongs to a non centro symmetric environment, i.e. it is tetrahedrally coordinated, the characteristic preedge peak gains intensity due to p–d hybridization. Based on XANES results, Rai et al. (2004, 2007) confirmed that their products contained only trivalent chromium. They also compared the Cr-Kedge EXAFS spectra with model compounds of Cr(III) oxides and oxyhydroxides, such as the crystalline compounds Cr2O3, a-CrOOH and b-CrOOH and the amorphous c-CrOOH. They concluded that the local structure of amorphous hydroxide Cr(OH)3 is similar to that of c-CrOOH, in accordance with the previous studies by Charlet and Manceau (1992) and Fendorf et al. (1994). The structure appears to consist of chains of two edge-sharing CrO6 polyhedra linked to other dimer chains through an apical hydroxyl group. To our knowledge the local atomic order of the crystalline Cr(OH)33H2O has not been studied by XAFS techniques until now. In most cases, X-ray diffraction of mixed Fe(III)–Cr(III) hydroxides has characterized them as amorphous (Sass and Rai, 1987; Puls et al., 1999; Wielinga et al., 2001; Tang et al., 2010; Papassiopi et al., 2012). From the Fe-K-edge EXAFS studies conducted by Hansel et al. (2003), the authors concluded that the local atomic structure of mixed Fe(III)–Cr(III) oxides resembles that of 2-line ferrihydrite; the latter is characterized by a local atomic order between a-Fe2O3 and a/b-FeOOH. The Cr-K-edge EXAFS spectra is in agreement with local atomic ordering of the form a/b-MOOH (M = Fe/Cr). The environmental stability of plain Cr(III) and mixed Fe(III)-Cr(III) hydroxides depends closely on their structure and it is important to understand how different precipitation conditions

affect the local atomic order of the resulting compounds and thus their long term stability. Here, we use XRD and EXAFS/XANES to examine the structural environment of Cr and Fe in three typical remediation products: (1) a plain Cr(III) hydroxide, (2) a mixed Fe–Cr hydroxide, Fe0.75Cr0.25(OH)3 produced via hydrolysis and (3) a mixed Fe–Cr hydroxide, of the same stoichiometry as (2), produced via redox reaction. A plain Fe(III) hydroxide was also synthesized as reference compound and examined with the same techniques. 2. Materials and methods For the production of Fe(III) hydroxide, 40.4 g of Fe(NO3)39H2O were mixed in 500 mL of deionized water with 330 mL KOH 1 M to achieve a pH value within a range of 7–8. The solution was then filtered and the cake of solid was treated by dialysis in order to remove the residual electrolyte ions. The clean cake of solids was freeze-dried and the resulting powder was sealed in a N2–atmosphere container. Synthesis of the Cr(III) hydroxide involved dissolution of 40 g Cr(NO3)39H2O in 500 mL of deionized water followed by the same procedure used in the synthesis of Fe(III) compound. The mixed Fe0.75Cr0.25(OH)3 hydroxide was synthesized by dissolving 30.3 g of Fe(NO3)39H2O in 500 mL of water followed by the addition of 10 g of Cr(NO3)39H2O and 330 mL of KOH. For the production of Fe0.75Cr0.25(OH)3 via the redox route, 750 mL of FeSO4 and 200 mL of a K2Cr2O7 solution were mixed in a spherical reactor under a N2 flux; NaOH 5 M was added gradually up to a pH of between 7 and 8 and the solution was then stirred for 24 h after which it was subjected to filtration, dialysis for ionic species removal and freeze-drying. All synthesis experiments were carried out at ambient temperature conditions (22 ± 1 °C). For the elemental analysis of the compounds 0.2 g of each solid was dissolved in 20 mL of 6 N HCl acid and iron and chromium content was determined by atomic absorption spectrophotometry (flame emission) using a Perkin Elmer 2100 spectrophotometer. To determine the number of structural water molecules, 1 g of solid was heated at 1000 °C for approximately 30 min under a N2 atmosphere. The weight loss was attributed to chemically bound water. All chemical analyses and physical measurements were conducted in duplicate. The XRD spectra of the solids were obtained using a Bruker D8 diffractometer equipped with software EVA 100. Analysis was carried out with a Cu source (k = 1.5418 Å). Diffraction data were collected between two-theta values of 5–70° with a step size of 0.02° and an average counting time of 1 s per step. The Fe and Cr K-edge EXAFS measurements were conducted at the storage ring BESSY in Berlin (Germany), at the KMC beamline which is equipped with a double SiGe crystal monochromator. The spectra were recorded simultaneously in the transmission and fluorescence yield modes. After correcting the latter for selfabsorption effects, the two modes resulted in equivalent spectra. Prior to analysis, the EXAFS and XANES spectra were normalized to the impinging photon flux using the signal from an ionization chamber positioned in front of the sample. After the subtraction of the atomic absorption, the v(k) EXAFS spectra were fitted with the FEFFIT program with the photoelectron scattering paths constructed using the FEFF8 code (Ankudinov et al., 1998). Finally, the XANES spectra were normalized to the edge jump after subtraction of a linear background. 3. Results and discussion 3.1. Composition The chemical composition of plain and mixed Fe(III) and Cr(III) hydroxides and their approximate molecular formula as estimated

171

N. Papassiopi et al. / Chemosphere 111 (2014) 169–176

from the elemental analysis and water content measurements are presented in Table 1. The plain Fe(III) hydroxide was found to contain 52.2% Fe and 25.7% H2O. These values are close to the theoretical composition of Fe(OH)3 corresponding to 52.3% Fe and 25.2% H2O. The chemical composition of plain Cr(III) hydroxide, i.e. 36.1% Cr and 50.1% H2O, indicates the presence of excess water molecules and corresponds approximately to the formula of trihydrate Cr(III) hydroxide, Cr(OH)33H2O, whose Cr and H2O content are 33.1% and 51.6% respectively. The two mixed hydroxides were synthesized so as to contain iron (III) and chromium (III) in molar ratio Fe(III)/Cr(III) = 3/1. As seen in Table 1, the assay of the two elements is in agreement with the expected molar ratio. The water content of the mixed hydroxides, namely 27.8% and 32.0%, is relatively higher compared to that expected from the theoretical formula Fe0.75Cr0.25(OH)3, i.e. 25.5%. The fact that plain Cr(III) hydroxide has a higher water content compared to all three other hydroxides, i.e. 50% against 26–32%, is also reflected to the lower specific gravity of the compound, i.e. 2.19 g cm3 for Cr(OH)33H2O against 2.70–2.80 g cm3 for the plain Fe(III) and the mixed Fe(III)– Cr(III) hydroxides (see Table 1). 3.2. X-ray diffraction characteristics The XRD spectrum of the pure Fe(III) hydroxide is shown in Fig. 1a. It is typical of amorphous 2-line ferrihydrite, with the two broad bands at approximately d = 2.56 Å and 1.48 Å (2h = 35 and 62.8° with Cu-K radiation). The XRD spectrum of plain Cr(III) hydroxide, shown in Fig. 1b, reveals a high degree of crystallinity. It follows the XRD pattern of Cr(OH)33H2O as reported by Giovanoli et al. (1973). According to Giovanoli et al. the structure of this crystalline compound can be described as ‘‘anti-bayerite’’. Representation of ‘‘anti-bayerite’’ structure in comparison with that of bayerite is given in Fig. 2. The structure of trihydrate chromium hydroxide consists of isolated Cr(OH2)3(OH)3 octahedra that are interlinked by hydrogen bonds. As seen in Fig. 2 the octahedral positions that are filled in the bayerite structure are empty in the structure of Cr(OH)33H2O. Despite the fact that the formation of a crystalline Cr(III) hydroxide under ambient temperature conditions has been reported since 1952 (Laubengayer and McCune, 1952) and this compound has been characterized in details by Giovanoli et al. since 1973, the majority of researchers, particularly in the field of environmental studies, make reference to an amorphous Cr(III) hydroxide (Rai et al., 1987, 2002, 2004, 2007; Wielinga et al., 2001; Tang et al. 2010; Papassiopi et al., 2012). This is obviously due to the low stability of Cr(OH)33H2O structure. The crystalline network of tri-hydrate Cr(III) hydroxide is stable at room temperature, but collapses at relatively mild temperature conditions. When the initial precipitate is heated in air or under water at temperatures greater than 60 °C the crystalline compound is transformed within a relatively short period, e.g. 24–72 h, to an amorphous phase in which no lattice order is observed (Giovanoli et al., 1973; Giovanoli and Stadelmann, 1973; Allen and Poeppelmeier, 1994). According to Giovanelli et al. the hydrogen-bonded isolated octahedral units condense to edge or corner

sharing polynuclear groups, by releasing water molecules. This transformation can also take place under ambient temperature conditions, when the solids are kept in the form of an aqueous suspension (Spiccia and Marty, 1986). The researchers who didn’t identify any crystalline phase during the characterization of Cr(III) hydroxides either had dried the solids at temperatures above 60 °C (Tang et al., 2010) or had left them in suspension for a long period of time (Rai et al., 1987, 2002, 2004, 2007; Wielinga et al., 2001; Hansel et al., 2003; Papassiopi et al., 2012). The XRD spectra of the mixed hydroxides, presented in Fig. 1c and d, appear largely similar, both exhibiting 2-line ferrihydrite amorphicity, with information related to Cr(OH)33H2O altogether absent in the spectra. The intensity of the two broad peaks at 2h = 35° and 63° is lower compared to plain Fe(OH)3 (Fig. 1a), suggesting that the presence of Cr causes some loss of the ferrihydrite structure order. A similar effect was observed by Hansel et al. (2003) and Tang et al. (2010). Obviously co-precipitation of Cr with Fe, either through hydrolysis or redox mechanism, suppresses the formation of crystalline tri-hydrated Cr hydroxide. It is not clear if the product is a physical mixture of amorphous Cr(OH)3 and ferrihydrite or a single phase resulting from the chemical substitution between Cr(III) and Fe(III). Most studies opt for the second alternative, i.e. the formation of a solid solution resembling to ferrihydrite when the mixed product is rich in Fe(III) and to amorphous Cr(OH)3, when it is rich in Cr(III) (Sass and Rai, 1987; Hansel et al., 2003; Tang et al., 2010). 3.3. X-ray absorption near-edge structure The oxidation state and type of polyhedron that is formed around Fe and Cr were assessed using the XANES spectra. The Fe K-edge XANES spectra of the Fe(OH)3 and the mixed Fe/Cr hydroxides are shown in Fig. 3. The low-intensity peak that is detected at approximately 10 eV below the adsorption edge has been attributed to electronic 1s ? 3d transitions which become dipole allowed due to mixing of the 3d with the 4p orbitals. The characteristics of the pre-edge peak are strongly affected by both the oxidation state of Fe as well as coordination symmetry around Fe. In particular, when Fe atoms form centro-symmetric octahedra, the pre-edge peak tends to be wide and rather faint; on the contrary, this feature is narrower and of increased intensity when Fe is tetrahedrally coordinated. Additionally, the pre-edge peak position is indicative of the Fe ions oxidation degree, i.e. it undergoes a violet shift when the oxidation state of Fe increases form (II) to (III) (Westre et al., 1997). The pre-edge peak, in the case of the studied samples, was fitted using three Voigt functions (P1, P2, P3) as shown in the inset of Fig. 3. The peaks P1 and P2, appearing at 7113.7 and 7114.8 eV, respectively, were attributed to the presence of Fe(III) centers in deformed octahedra, much in accord with the structure of Fe(OH)3. The peak P3 at 7117.4 eV is due to Fe-Fe interactions related to the clustering of the Fe octahedra and was not considered in the estimation of the total area of the pre-edge peak. The total area under the pre-edge peak is nearly identical (0.39 ± 0.01 arbitrary units) in all samples, a feature that denotes similarity in the degree of distortion of the FeO6 octahedra.

Table 1 Chemical composition and specific gravity measurements of the compounds (mean value ± relative difference of duplicate measurements).

a b

Sample

Fe, %

Fe(OH)3 Cr(OH)33H2O FeCr hydrolysis FeCr redox

52.2 ± 1.9 36.7 ± 1.6 37.0 ± 0.7

Cr, %

H2O, %

Molecular formulaa

Specific gravityb (g cm3)

36.1 ± 0.3 11.0 ± 0.3 10.5 ± 0.5

25.7 ± 1.4 50.1 ± 1.6 27.8 ± 1.1 32.0 ± 2.0

Fe2O33H2O Cr2O38.5H2O (Fe2O3)0.76(Cr2O3)0.243.4H2O (Fe2O3)0.77(Cr2O3)0.233.7H2O

2.77 ± 0.12 2.19 ± 0.07 2.70 ± 0.07 2.80 ± 0.16

Approximate molecular formula, estimated from elemental analysis and water content. Determined with the use of water displacement pycnometer.

172

N. Papassiopi et al. / Chemosphere 111 (2014) 169–176

Fig. 1. XRD spectra with Cu-K radiation (k = 1.5418 Å). (a) Fe(OH)3, (b) Cr(OH)33H2O, (c) and (d) hydrolysis- and redox-synthesized Fe0.75Cr0.25(OH)3 hydroxide respectively.

The Cr-K-edge XANES spectra of the Cr(OH)33H2O and the mixed Fe/Cr hydroxides are shown in Fig. 4. As in the case of the Fe-K-edge spectra, a pre-edge peak is also present in the Cr-K-edge XANES spectra of all studied Cr-hydroxides which is also attributed to 1s ? 3d electronic transitions. In the case of tetrahedrally coordinated Cr, the intensity of the pre-edge peak is strong and only a single peak component is present. However, two contributions are detected in the pre-edge peak of octahedrally coordinated Cr, and the pre-edge peak is broad and of low intensity (Arcon et al., 1998). Furthermore, the position of the absorption edge (Eabs) can be used for the determination of the Cr oxidation state. More

specifically, a blue shift of approximately 11 eV is observed in the case of Cr(III)-compounds compared to Cr metal. The XANES fitting analysis results reveal that the position of Eabs of plain Cr(III) hydroxide, i.e. Cr(OH)33H2O, and of the mixed Fe/Cr hydroxides is shifted by +10.9 and +12.1 eV, respectively, compared to the Eabs of the Cr foil. Using the calibration curve for the Eabs proposed by Arcon et al. (1998), the oxidation state of Cr is found equal to +3 and +3.6 in Cr(OH)33H2O and Fe/Cr hydroxides, respectively. The oxidation state of +3.6, suggesting the presence of a small amount of Cr(VI), could be justified for the case of mixed hydroxide that was produced via the redox synthesis procedure. However,

173

N. Papassiopi et al. / Chemosphere 111 (2014) 169–176

Fig. 2. (a) Representation of the anti-bayerite type layer of Cr(OH)33H2O. Lines joining the octahedra represent hydrogen bonds. (b) Comparative view of a bayerite, Al(OH)3, layer (adapted from Allen and Poeppelmeier, 1994).

Cr metal Cr metal 2 1.5

FeCr redox redox

FeCr redox redox

1.0

Intensity (arb. units)

Intensity (arb. units)

hydrolysis hydrolysis

Fe(OH) Fe(OH)33

hydrolysis hydrolysis Cr(OH)3

Cr(OH)3.3H2O

1

0.2

0.2

0.5

P1'

P2' P3'

0.1

0.1 P1 P3 P2

0.0 7105

7115

0.0 7140

7160

5988

7120 5960

7120

0.0

0 7110

7180

5980

6000

6020

5992 6040

5996 6060

Energy (eV)

Energy (eV) Fig. 3. Fe-K-edge XANES spectra normalized to the edge jump of the Fe hydroxide [Fe(OH)3] and the hydrolysis- and redox-synthesized Fe0.75Cr0.25(OH)3 samples. The inset shows the fitting of the isolated pre-edge peak using three Voigt functions, P1–P3.

existence of Cr(VI) is less probable in the case of Fe/Cr produced by hydrolysis. As far as the pre-edge peak of the Cr-K-XANES spectra is concerned, the pre-edge peak was fitted using three Lorentzian functions. More specifically, as shown in the inset of Fig. 4, the Lorentzian P10 , P20 and P30 appear at 5990.9, 5992.8 and 5994.2 eV, respectively. P10 and P20 are attributed to distorted octahedral Cr(III) due to ls ? 3d(t2g) and ls ? 3d(eg) electronic transitions, respectively (Eeckhout et al., 2007). Peak P30 , which is assigned to metal-metal transitions (Frommer et al., 2010) is eliminated in the case of pure Cr(OH)33H2O. 3.4. Extended X-ray absorption fine structure The short-range order of the Fe hydroxides can be studied by means of EXAFS spectroscopy. It is expected that Fe will form

Fig. 4. Cr-K-edge XANES spectra normalized to the edge jump of the Cr hydroxide [Cr(OH)33H2O] and the hydrolysis- and redox-synthesized Fe0.75Cr0.25(OH)3 samples. The inset shows the fitting of the isolated pre-edge peak using three Lorentzian functions, P10 –P30 . In the case of the Cr(OH)33H2O sample the contribution of P30 is diminished.

Fe-centered octahedra linked by edge, face or corner and each linkage between the FeO6 octahedra corresponds to different Fe–Fe distances (Manceau and Drits, 1993). The Fourier Transform (FT) of the k3-weighted v(k) Fe-K-EXAFS spectra of all studied solids are shown in Fig. 5a. The fitting was performed in the first four nearest neighboring (nn) shells using the photoelectron scattering paths of the crystalline model of akaganeite (b-FeOOH). More specifically, a Fe–O path at 2.01 Å, two Fe–Fe paths at 2.98 and 3.40 Å that correspond to edge and double corner sharing configuration and a Fe–O path at 3.53 Å were used. It should be noted that splitting of the neighboring shells in Fe and Cr sub shells was not considered during the fitting since EXAFS spectroscopy is not able to discriminate between elements with small difference in the atomic numbers. The fitting results of the Fe-K-edge EXAFS spectra are listed in Table 2a. Fe is found 6-fold coordinated in both the pure

174

N. Papassiopi et al. / Chemosphere 111 (2014) 169–176

0.1

FT amplitude (arb. units)

χ (k)

6

0.0

-0.1 3

4

5

6

7

4

8

9

10 11 12

k (Å-1)

FeCr redox 2

FeCr hydrolysis hydrolysis

Fe(OH) Fe (OH) 3 3 0 0

1

2

3

4

5

6

7

8

R (Å)

(a) Fe-K-edge EXAFS

0.3

χ (k)

15

FT amplitude (arb. units)

0.0

3

4

5

6

7

8

9 10 11 12 13

k (A-1) 10

FeCr redox redox

5

FeCr hydrolysis hydrolysis

Cr(OH)Cr(OH) 3⋅3H 2O 0

0

1

2

3

4

5

6

7

3

8

R (Å)

(b) Cr-K-edge EXAFS Fig. 5. Fourier transform amplitude of the k3 weighted v(k) EXAFS spectra recorded at (a) the Fe-K-edge and (b) the Cr-K-edge. The experimental curve and the fitting are shown in thin and thick lines, respectively. The insets show the corresponding fitting in the k-space.

and mixed Fe/Cr oxides, whereas the Fe–O distance is characteristic of Fe-centered octahedral coordination. The Fe–Fe distances in the second and third nn shells are characteristic of edge- and double corner-sharing FeO6 octahedra. The ratio of the number of edge vs corner sharing octahedra is slightly higher in the Fe(OH)3 hydroxide compared to the mixed Fe/Cr samples. The results are in agreement with Hansel et al. (2003), who established that mixed Fe(III)–Cr(III) hydroxides had a local order similar to ferrihydrite.

The Fourier Transform (FT) of the k3-weighted v(k) of the Cr-Kedge EXAFS spectra for the Cr(OH)33H2O phase and the mixed Fe/ Cr hydroxides are shown in Fig. 5b. The fitting was performed using the structure of b-CrOOH (Pernet et al., 1977) where Cr forms octahedra that share edges and double corners. The structure of akaganeite, where the central absorbing atom of Fe was substituted by Cr, was also tested and yielded similar fitting results. As already mentioned, the separation between Fe and Cr atoms is difficult owing to the small Z difference between the two, which in turn induces very similar backscattering amplitude. The fitting results of the Cr-K-edge EXAFS analysis are listed in Table 2b. The Cr–O distance of 1.98 Å with coordination number N = 6 is characteristic of octahedrally coordinated trivalent Cr and was observed in all three compounds. Cr–Cr(Fe) distances were observed with an appreciable intensity mainly in the 2 mixed Fe(III)–Cr(III) hydroxides in the distance d = 3.06 Å, corresponding to edge sharing octahedra. The contribution due to corner sharing CrO6 or FeO6 octahedra at d = 3.4–3.5 Å is not detected and the edge sharing is the dominant linkage way, irrespective of the synthesis route. The absence of corner sharing octahedra at d = 3.4– 3.5 Å was further verified by back–Fourier transform; the existence of two Cr–Cr(Fe) scattering paths at distances that differ by 0.35 Å is expected to yield a beat-node in the back-Fourier transformed second peak in the FT at approximately 4.5 Å1, which was not observed. The main difference between the products synthesized under hydrolysis and redox conditions is that in the former case the coordination number in the 2nd nn shell is almost twice the corresponding in the latter case. This finding is indicative of the better arrangement of the CrO6 octahedra in the case of the sample synthesized by hydrolysis in accordance with the slightly sharper XRD patterns shown in Fig. 1. The Cr-K-edge EXAFS spectra of the mixed Fe/Cr hydroxides in our study are contrasted with the results of Hansel et al. (2003) and Charlet and Manceau (1992), who studied similar mixed hydroxides. Hansel et al. analyzed a series of Fe/Cr hydroxides produced by hydrolytic precipitation, and two Fe/Cr hydroxides produced via a microbially induced redox process. For all studied solids, the Cr-K-edge EXAFS analysis indicated the existence of a Fe(Cr) shell around the central Cr atom at a distance of d = 3.46– 3.51 Å, which is characteristic of the a-FeOOH type local structure of ferrihydrite. Charlet and Manceau (1992) produced a mixed Fe/ Cr hydroxide by hydrolytic co-precipitation and also detected the characteristic peak of double corner sharing octahedra at d = 3.46 Å. They have thus concluded that Cr substitutes isomorphically Fe in the structure of ferrihydrite. It should be noted that Hansel et al. (2003) and Charlet and Manceau (1992), after the hydrolytic co-precipitation of mixed Fe/Cr hydroxides, left the solids in suspension for 3 and 7 d respectively. This ageing procedure has probably allowed a better incorporation of Cr(III) atoms in the structure of ferrihydrite and can justify the differences observed between the ‘‘fresh’’ hydroxides examined in our study and the ‘‘aged’’ hydroxides examined by the previous researchers. The Cr-K-edge EXAFS spectra of our ‘‘fresh’’ Fe/Cr products have similarities with the spectra observed by Charlet and Manceau (1992) in the case of ferrihydrite samples which had been subjected to Cr(III) adsorption experiments. This finding suggests that probably, even during co-precipitation, Cr are initially retained on the surface of ferrihydrite nanoparticles by adsorption or surface precipitation and are subsequently incorporated in the bulk structure with a more slow mechanism upon ageing. Contrary to what was observed for Fe(III)–Cr(III) mixed hydroxides, in the case of plain Cr(OH)33H2O it was not possible to detect any neighboring Cr–Cr shells. As seen in Table 2b and Fig. 5b, the scattering contribution of edge sharing octahedra at d = 3.0 Å in the Cr-K-edge EXAFS spectrum was very faint, corresponding to a

175

N. Papassiopi et al. / Chemosphere 111 (2014) 169–176

Table 2 Fitting results of the EXAFS analysis of plain and mixed Fe(III) and Cr(III) hydroxides. R is the interatomic distance, N is the coordination number and r2 is the Debye–Waller factor. This notation is followed throughout. Quoted errors correspond to the uncertainty of fitting. Values marked with an asterisk were kept fixed during fitting. Sample

1st nn shell [Fe–O] N

(a) Fe-K-edge Fe(OH)3 FeCr hydrolysis FeCr redox

6* 6* 6*

2nd nn shell [Fe–Fe]

R (Å) ±0.01

r2 (Å2)

1.97 1.97 1.98

3rd nn shell [Fe–Fe]

R (Å) ±0.03

r2 (Å2)

±10%

N ±20%

0.016 0.014 0.013

4.1 3.4 2.9

3.04 3.05 3.03

r2 (Å2)

±30%

R (Å) ±0.03

0.017 0.015 0.014

3.6 4.9 3.8

3.45 3.44 3.41

1st nn shell [Cr–O] N (b) Cr-K-edge Cr(OH)33H2O FeCr hydrolysis FeCr redox

6* 6* 6*

4th nn shell [Fe–O]

N ±20%

R (Å) ±0.05

r2 (Å2)

±30%

N ±50%

0.018 0.016 0.015

5.9 5.7 6.1

3.62 3.63 3.62

0.021 0.020 0.018

±50%

2nd nn shell [Cr–Cr] R (Å) ±0.01

r2 (Å2)

1.98 1.98 1.98

coordination number N = 0.9. Despite the crystallinity of this compound, indicating a long range order (see Fig. 1b), this order is not detectable in the EXAFS spectrum. This finding, though unusual, is in accordance with the ‘‘anti-bayerite’’ structure of crystalline Cr(OH)33H2O. As seen in Fig. 2a, in this structure the CrO6 octahedra are relatively isolated and interconnected only via hydrogen type bonds. According to the study of Rustad and Casey (2006), the Cr–Cr distances in the structure of Cr(OH)33H2O are equal to 5.02 Å and 5.65 Å. These distances are in practice outside of the usual range of EXAFS measurements. As Newville (2004) mentions XAFS is an inherently local probe, not able to see much further than 5 or so Angstroms from the absorbing atom. The loose interconnection of Cr atoms in the Cr(OH)33H2O structure, which is reflected to the fact that it does not possess any detectable short range order in EXAFS spectra, indicates that it is an unstable solid phase. This is in accordance with the finding of many researchers that upon aging and in contact with an aqueous solution this hydroxide becomes amorphous via a dehydration process (Giovanoli et al., 1973; Giovanoli and Stadelmann, 1973; Spiccia and Marty, 1986; Allen and Poeppelmeier, 1994). 4. Conclusions Upon Cr(III) solution exposure to an alkaline environment in the absence of iron, the crystalline Cr(OH)33H2O phase precipitates; the latter exhibits XRD Bragg peaks but is EXAFS inactive. Hence, CrO6 octahedra do not exhibit characteristic face, edge or corner sharing up to 5 Å, due to their loose anti-bayerite structure. The differences in the bonding configuration of Cr in the plain and mixed Fe/Cr hydroxides are also evident in the XANES spectra. The plain Fe(III) hydroxide and the mixed Fe(III)-Cr(III) systems resembled amorphous 2-line ferrihydrite in XRD analysis, irrespective of the synthesis route. Fe-K-edge EXAFS spectra indicated also that the local structure of mixed hydroxides is similar to that of ferrihydrite, with characteristic edge and corner sharing FeO6 octahedra at d = 3.0 Å and d = 3.4 Å respectively. Based on Cr-K-edge EXAFS spectra, the local structure around Cr is slightly different from that of ferrihydrite, consisting primarily of edge sharing CrO6 (or FeO6) octahedra at d = 3.0 Å and without detectable corner sharing octahedra at d = 3.4–3.5 Å. EXAFS data also indicated that the mixed Fe–Cr hydroxide produced by hydrolysis presents a better arrangement of CrO6 octahedra compared to the redox product. Despite the absence of long range order, the Fe–Cr mixed hydroxides present a local order, which justifies their environmental stability. Ageing seems to improve their stability by promoting the isomorphic incorporation of Cr(III) in the ferrihydrite structure.

R (Å) ±0.03

r2 (Å2)

±10%

N ±20%

0.003 0.002 0.002

0.9 6.1 3.3

3.00 3.06 3.06

0.007 0.013 0.012

±30%

On the contrary crystalline Cr(OH)33H2O seems to be an unstable solid phase, without detectable interconnections between the CrO6 octahedra in the EXAFS spectra. This finding is in accordance with the fact that upon aging and in contact with an aqueous solution the anti-bayerite structure collapses, through a dehydration process, and the hydroxide is transformed to an amorphous but more condensed and stable phase. Acknowledgement The measurements at BESSY were funded by the European Community’s Seventh Framework Programme (FP7/2007-2013) under Grant Agreement No. 226716. References Allen, J.L., Poeppelmeier, S., 1994. Synthesis of lithium chromium oxide from a hydroxide precursor. Polyhedron 13, 1301–1310. Ankudinov, A.L., Ravel, B., Rehr, J.J., Conradson, S.D., 1998. Real-space multiplescattering calculation and interpretation of X-ray-absorption near-edge structure. Phys. Rev. B 58, 7565–7576. Arcon, I., Mirtic, B., Kodre, A., 1998. Determination of valence states of chromium in calcium chromates by using X-ray Absorption Near-Edge Structure (XANES) spectroscopy. J. Am. Ceram. Soc. 81, 222–224. Battaglia-Brunet, F., Foucher, S., Morin, D., Ignatiadis, I., 2004. Chromate (CrO2 4 ) reduction in groundwaters by using reductive bacteria in fixed-bed bioreactors. Water Air Soil Poll. Focus 4, 127–135. Charlet, L., Manceau, A.A., 1992. X-ray absorption spectroscopic study of the sorption of Cr(lll) at the oxide–water interface. J. Colloid Interface Sci. 148, 443– 458. Chrysochoou, M., Ferreira, D.R., Johnston, C.P., 2010. Calcium polysulfide treatment of Cr(VI)-contaminated soil. J. Hazard. Mater. 179, 650–657. Dermatas, D., Chrysochoou, M., Moon, D.H., Grubb, D.G., Wazne, M., Christodoulatos, C., 2006. Ettringite-induced heave in chromite ore processing residue (COPR) upon ferrous sulfate treatment. Environ. Sci. Technol. 40, 5786– 5792. Eary, L.E., Rai, D., 1988. Chromate removal from aqueous wastes by reduction with ferrous ion. Environ. Sci. Technol. 22, 972–977. Eeckhout, S.G., Bolfan-Casanova, N., McCammon, C., Klemme, S., Amiguet, E., 2007. XANES study of the oxidation state of Cr in lower mantle phases: periclase and magnesium silicate perovskite. Am. Mineral. 92, 966–972. Fendorf, S.E., Lamble, G.M., Stapleton, M.G., Kelley, M.J., Sparks, D.L., 1994. Mechanisms of chromium(III) sorption on silica. 1. Cr(III) surface structure derived by extended absorption fine structure spectroscopy. Environ. Sci. Technol. 28, 284–289. Frommer, J., Nachtegaal, M., Czekaj, I., Kretzschmar, R., 2010. The Cr X-ray absorption K-edge structure of poorly crystalline Fe(III)–Cr(III)-oxyhydroxides. Am. Mineral. 95, 1202–1213. Giovanoli, R., Stadelmann, W., 1973. Über kristallines chrom(III)hydroxid II. Thermische Zersetzung. Thermochim Acta 7, 41–55. Giovanoli, R., Stadelmann, W., Feitnecht, W., 1973. Über kristallines Chrom(III)hydroxid. I. Helv. Chim. Acta 56, 839–847. Hansel, C.M., Wielinga, B.W., Fendorf, S., 2003. Structural and compositional evolution of Cr/Fe solids after indirect chromate reduction by dissimilatory iron-reducing bacteria. Geochim. Cosmochim. Acta 67, 401–412.

176

N. Papassiopi et al. / Chemosphere 111 (2014) 169–176

Kumpiene, J., Ore, S., Renella, G., Mench, M., Lagerkvist, A., Maurice, C., 2006. Assessment of zerovalent iron for stabilization of chromium, copper, and arsenic in soil. Environ. Pollut. 144, 62–69. Laubengayer, A.W., McCune, H.W., 1952. New crystalline phases in the system chromium (III) oxide-water. J. Am. Chem. Soc. 74, 2362–2364. Manceau, A., Drits, V.A., 1993. Local structure of ferrihydrite and feroxyhite by EXAFS spectroscopy. Clay Miner. 28, 165–184. Michel, C., Brugna, M., Aubert, C., Bernadac, A., Bruschi, M., 2001. Enzymatic reduction of chromate: comparative studies using sulfate-reducing bacteria: key role of polyheme cytochromes c and hydrogenases. Appl. Microbiol. Biot. 55, 95–100. Newville, M., 2004. Fundamentals of XAFS. University of Chicago, Chicago, IL (Revision 1.7). Papassiopi, N., Kontoyianni, A., Vaxevanidou, K., Xenidis, A., 2008. Evaluation of Fe(III) reducing microorganisms for the biostabilisation of chromium in contaminated soils. In: Proceedings of International Conference GeoCongress 2008, New Orleans, Louisiana, USA, pp. 535–542. Papassiopi, N., Kontoyianni, A., Vaxevanidou, K., Xenidis, A., 2009. Assessment of chromium biostabilization in contaminated soils using standard leaching and sequential extraction techniques. Sci. Total Environ. 407, 925–936. Papassiopi, N., Gaitanarou, Z., Xenidis, A., 2012. Stabilization of chromium in the form of mixed Fe(III)–Cr(III) hydroxides. Fresen. Environ. Bull. 21, 2399–2405. Pernet, M., Berthet-Colominas, C., Franco, M.A.A., Christensen, A.N., 1977. Etude par diffraction neutronique de l’oxyhydroxyde de chrome b-CrOOH. Phys. Status Solidi (a) 43, 81–88. Puls, R.W., Paul, C.J., Powell, R.M., 1999. The application of in situ permeable reactive (zero-valent iron) barrier technology for the remediation of chromatecontaminated groundwater: a field test. Appl. Geochem. 14, 989–1000. Rai, D., Sass, B.M., Moore, D.A., 1987. Chromium (III) hydrolysis constants and solubility of chromium(III) hydroxide. Inorg. Chem. 26, 345–349.

Rai, D., Hess, N.J., Rao, L., Zhang, Z., Felmy, A.R., Moore, D.A., Clark, S.B., Lumetta, G.J., 2002. Thermodynamic model for the solubility of Cr(OH)3(am) in concentrated NaOH and NaOH–NaNO3 solutions. J. Solution Chem. 31, 343–367. Rai, D., Moore, D.A., Hess, N.J., Rao, L., Clark, S.B., 2004. Chromium(III) hydroxide solubility in the aqueous Na+OH H2PO4 HPO24 PO34 H2O system: a thermodynamic model. J. Solution Chem. 33, 1213–1242. Rai, D., Moore, D.A., Hess, N.J., Rosso, K.M., Rao, L., Heald, S.M., 2007. Chromium(III) 2 hydroxide solubility in the aqueous K+–H+–OH–CO2–HCO 3 –CO3 –H2O system: a thermodynamic model. J. Solution Chem. 36, 1261–1285. Rustad, J.R., Casey, W.H., 2006. A molecular dynamics investigation of hydrolytic polymerization in a metal–hydroxide gel. J. Phys. Chem. B 110, 7107–7112. Sass, B.M., Rai, D., 1987. Solubility of amorphous chromium(III)–iron(III) hydroxide solid solutions. Inorg. Chem. 26, 2228–2232. Spiccia, L., Marty, W., 1986. The fate of active chromium hydroxide, Cr(OH)33H2O, in aqueous suspension. Study of the chemical changes involved in its aging. Inorg. Chem. 25, 266–271. Tang, Y., Michel, F.M., Zang, L., Harrington, R., Parise, J., Reeder, R., 2010. Structural properties of the Cr(III)–Fe(III) oxyhydroxide compositional series: insights for a nanomaterial ‘‘solid solution’’. Chem. Mater. 22, 3589–3598. Von Meyenburg, U., Široky´, O., Schwarzenbach, G., 1973. Die Deprotonierung von Metallaquoionen II: Aquochrom(III)-ion. Zur Struktur aktiver Chromhydroxide. Helv. Chim. Acta 56, 1099–1114. Westre, T.E., Kennepohl, P., DeWitt, J.G., Hedman, B., Hodgson, K.O., Solomon, E.I., 1997. A multiplet analysis of Fe K-Edge 1s ? 3d pre-edge features of iron complexes. J. Am. Chem. Soc. 119, 6297–6314. Wielinga, B., Mizuba, M.M., Hansel, C.M., Fendorf, S., 2001. Iron promoted reduction of chromate by dissimilatory iron-reducing bacteria. Environ. Sci. Technol. 35, 522–527.