Chemical Speciation of Arsenic-Accumulating Mineral in a

Sep 4, 2008 - Chemical Speciation of Arsenic-Accumulating Mineral in a Sedimentary Iron Deposit by Synchrotron Radiation Multiple X-ray Analytical Tec...
1 downloads 10 Views 5MB Size
Environ. Sci. Technol. 2008, 42, 7152–7158

Chemical Speciation of Arsenic-Accumulating Mineral in a Sedimentary Iron Deposit by Synchrotron Radiation Multiple X-ray Analytical Techniques SATOSHI ENDO,† YASUKO TERADA,‡ Y A S U H I R O K A T O , § A N D I Z U M I N A K A I * ,† Department of Applied Chemistry, Faculty of Science, Tokyo University of Science, Kagurazaka 1-3, Shinjuku-ku, Tokyo 162-8601, Japan, Japan Synchrotron Radiation Research Institute (JASRI), 1-1-1, Kouto, Sayocho, Sayogun, Hyogo 679-5198, Japan, and Department of Geosystem Engineering, University of Tokyo, 7-3-1 Hongo, Bunkyoku, Tokyo 113-8655, Japan

Received March 5, 2008. Revised manuscript received July 17, 2008. Accepted July 28, 2008.

The comprehensive characterization of As(V)-bearing iron minerals from the Gunma iron deposit, which were probably formed by biomineralization, was carried out by utilizing multiple synchrotron radiation (SR)-based analytical techniques at BL37XU at SPring-8. SR microbeam X-ray fluorescence (SR-µ-XRF) imaging showed a high level of arsenic accumulation in the iron ore as dots of ca. 20 µm. Based on SEM observations and SR X-ray powder diffraction (SR-XRD) analysis, it was found that arsenic is selectively accumulated in strengite (FePO4 · 2H2O) with a concentric morphology, which may be produced by a biologically induced process. Furthermore, the X-ray absorption fine structure (XAFS) analysis showed that arsenic in strengite exists in the arsenate (AsO43-) form and is coordinated by four oxygen atoms at 1.68 Å. The results suggest that strengite accumulates arsenic by isomorphous substitution of AsO43- for PO43- to form a partial solidsolution of strengite and scorodite (FeAsO4 · 2H2O). The specific correlation between the distribution of As and biominerals indicates that microorganisms seems to play an important role in the mineralization of strengite in combination with an arsenic-accumulating process.

Introduction Arsenic is a ubiquitously distributed toxic element in the natural environment, soils, rocks, natural waters, and organisms (1). It is easily dissolved into surrounding solutions due to the oxidative dissolution of arsenic-containing minerals by acidic waters such as hot springs, geothermal springs, and mine waste drainages (2). However, Fukushi et al. (3) recently pointed out that arsenic levels in the drainage of an abandoned arsenic mine in Nishinomaki, Japan, have decreased to below the maximum allowable contaminant level (0.01 mg/L for drinking water, Japan) without any * Corresponding author phone: +81-3-3260-3662; fax: +81-3-32352214; e-mail: [email protected]. † Tokyo University of Science. ‡ Japan Synchrotron Radiation Research Institute. § University of Tokyo. 7152

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008

artificial treatment. It is well-known that the co-occurrence of iron and arsenic can lead to the formation of hydrous iron oxides containing significant amounts of arsenic (4). Dissolved arsenic is often adsorbed by iron hydrosulfate, oxyhydroxides, and hydroxides such as jarosite [KFe3(SO4)2(OH)6] (5, 6), schwertmannite [Fe8O8(OH)6SO4] (7), hematite (R-Fe2O3) (8), goethite (R-FeOOH), and lepidocrocite (γ-FeOOH) (9). In addition, Ve¨ronique et al. (10) found that synthesized Fe(III) phosphate, either amorphous or crystalline, can remove arsenic from waters. It is generally believed that arsenic is retained through adsorption onto the surface of these iron minerals and incorporation into the lattices of the mineral crystals (6); as such, the arsenic content is usually quite low in rivers and lakes (2). It is also known that the oxidation of Fe(II) to Fe(III), often catalyzed by the metabolic activity of microorganisms, such as Acidithiobacillus ferrooxidans, leads to the precipitation of iron hydroxides that are capable of absorbing toxic elements on their surface or of incorporating them within their crystal structures (11). Under acidic water conditions, microbial Fe(II) to Fe(III) oxidation and progressive neutralization of acid waters frequently brings about a precipitation of iron sulfates and hydroxides (12, 13). Furthermore, Islam et al. (14) found that vivianite [Fe3(PO4)2 · 8H2O] in biomats in Bangladesh contains a small amount of arsenic by isomorphous substitution for phosphorus and that some microorganisms may play an important role in the mineralization of this iron phosphate. Thus, biological activity is very effective for ion retention and much attention has been focused on the existence of some microorganisms from the perspective of bioremediation. Kato et al. (15) and Akai et al. (16) demonstrated that biomineralization (biologically induced mineralization; BIM) is responsible for the formation of the Gunma sedimentary iron deposit (a target of the present study) in Japan. It is reported that these sediments contain relatively high levels of arsenic (16), but the relationship between the biomineralization and arsenic accumulation has not yet been clarified. The iron ore contains up to 1 wt% As2O3, and the origin of arsenic is most likely derived from acidic source, such as spring waters. Recently, we have found that the sedimentary iron ore contains an extraordinarily high level of As during the examination of the BIM sediments. Thus, we started the present study to reveal the mechanism of the As accumulation in the iron ores. Since these sedimentary iron ores have a complex fine texture, spatially resolved analytical techniques are required. In most previous studies on the arsenic solid phases associated with biomineralization, microanalyses are performed using electron microprobe techniques, such as SEM-EDS, TEM-EDS, and EPMA. Electron microprobe analysis is not suitable for the analysis of arsenic because of the low excitation efficiency of the As-KR line and overlapping of the As-LR line with the Mg-KR line. In addition, a strong electron beam often causes damage to hydrated minerals (17). In the present study, the sedimentary iron ores from the Gunma iron deposit have been analyzed utilizing multiple SR X-ray techniques to characterize the minerals accumulating arsenic and to reveal the formation mechanism of these minerals. The purpose of the present study is to apply the unique combination of XRF imaging, XAFS, and XRD techniques to the same samples at a single beamline of SPring-8 utilizing X-ray microbeam in order to reveal various characteristics, including the two-dimensional distribution of elements, local structures around the arsenic atom, and mineral phases of the iron ores containing arsenic. 10.1021/es8006518 CCC: $40.75

 2008 American Chemical Society

Published on Web 09/04/2008

FIGURE 1. (a) Optical micrograph of the sample (the black frame shows the analyzed area). SR-XRF imaging area of (b) As and (c) Fe. X-ray beam size, 50 × 50 µm2; step size, 50 × 50 µm2; image size, 323 × 63 pixels.

FIGURE 2. (a) Optical micrograph of the sample. SR-µ-XRF imaging of As and Fe is shown in Figure (b) and (c), respectively. The analyzed area is shown with a black frame in Figure (a): X-ray beam size, 1.8 (V) × 2.8 (H) µm2, step size, 2.0 (V) × 3.0 (H) µm, image size, 241 × 334 pixels. High-resolution µ-XRF imaging of As and Fe is shown in Figure (d) and (e), respectively.

Experimental Section Site Description. The Gunma iron deposit of quaternary age is located at the eastern base of the Kusatsu-Shirane Volcano in Gunma Prefecture, central Japan (N36°34′; E138°30′). The Kusastu-Shirane Volcano is an active volcano surrounded by many acid hot springs. It is thought that the Gunma iron deposit was formed from the activity of some of these acid spring waters (15, 16). The spring water forming the iron deposit was ferruginous, cold (less than 30 °C), and acidic (pH 2-4). The ore body extends over 2 km along a valley with a maximum width of 150 m and a thickness of 20 m. The iron deposit precipitated on the Pleistocene andesitic lava flows and pyroclastic rocks. The iron ore was primarily composed of goethite and jarosite with various proportions of silicified volcanic (andesitic) detritus, and occurs mostly as one of two types; bedded, vuggy, goethite-rich ore and massive jarosite-rich ore. Algal fossils are often well-preserved

in the goethite-rich ore. These fossils are likely green algae, such as Zygnematales sp., or cyanobacteria, such as Lyngbya and Oscillatoria, on the basis of their morphological analogies (16). The target sample of the present study is one of these iron ores. In addition to goethite replacing these microorganisms, goethite occurs as a massive precipitate that is considered to have been inorganically formed from the source water. The microorganisms mentioned above are not observed in the jarosite-rich ore. The relatively large sizes and crystalline forms of jarosite as well as the jarosite overgrowths on algal fossils having been replaced by goethite indicate that the jarosite is of abiotic origin and was inorganically precipitated from the spring water during the later stage (15, 16). Therefore, the iron ores are basically a mixture of goethite-replacing microorganisms, goethite directly precipitated from the water, jarosite, and silicified volcanic detritus in various proportions. VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7153

FIGURE 3. (a) SEM image of concentric morphology and EDS spectra measured at (b) P1 and (c) P2. The arrows in Figure (a) show the analyzed points. The EDS spectra were measured for 300 s excited at 20 kV.

FIGURE 4. As K-edge XANES spectra measured at P1 and P2 in Figure 2d of the iron ore sample and that of strengite containing As(V). XANES spectra of KH2AsO4 and KAsO2 are also shown as reference spectra of As(V) and As(III) states, respectively. SR X-ray Analyses and Sample Preparation. Multiple SR analyses were performed at the BL37XU at the SPring-8, thirdgeneration synchrotron facility, Hyogo prefecture, Japan (18). For the SR analysis, the iron ore was sliced into 1 mm thick 7154

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008

sections using a diamond cutter. The sample was then glued onto a pure quartz glass plate with epoxy adhesive and ground into a thin section by an abrasive (silicon carbide) with a thickness of ca. 20 µm. To reveal the distribution of Fe and As in the iron ore, the thin sections were subjected to two-dimensional SR-XRF imaging. The incident beam was monochromatized with a Si(111) double-crystal monochromator to an energy of 12.8 keV. This incident energy was chosen in order to efficiently excite the As KR line and to avoid excitation of the Pb L line. At first, XRF imaging of the sample in a large area was performed with an X-ray beam of 50 × 50 µm2 produced by a simple set of 4-quadrant slits. SR-µ-XRF imaging was then performed using an X-ray beam obtained by utilizing a K-B mirror system (19), which produced a focused beam of 1.8 (V) × 2.8 µm (H) with a rectangular form at the sample position. An ionization chamber filled with air measured the intensity of the incident X-rays before entering the sample. The fluorescence X-rays of the Fe-KR and As-KR lines generated from the sample were measured by a silicon drift detector (SDD, Rontec Co.) positioned perpendicular to the incident X-ray beam. The measurement time was 0.1 s/pixel. Since SR-XRF imaging was measured by scanning the sample stage, the pixel size was adjusted to be equal to the step size of the sample stage. The XRF intensity from the sample was normalized by the incident X-ray intensity to compensate for the variations in the incident X-ray intensity. The microscopic morphology was observed using a highresolution scanning electron microscope (SEM, JEOL JSM7000F) equipped with an energy-dispersive spectrometer (EDS) operated at an accelerating voltage of 20 kV. In order to reveal the microscopic morphology, chemical etching was applied to the samples by treatment with dilute HF under an ultrasonic wave for 10 min. They were then coated with carbon to avoid charging during the measurements. XAFS Analysis and Sample Preparation. The local environment of arsenic in the iron ore was investigated by both X-ray absorption near-edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) analyses. The experimental layout was the same as that of the XRF imaging. The XAFS spectra were measured in the fluorescence mode utilizing a focused beam with the size of 1.8 (V) × 2.8 µm (H) in a rectangular form. The fluorescence X-rays were detected by the single-element SDD. Energy calibration was made using a reference sample of KH2AsO4 with the energy of 11.867 keV at the inflection point of the As K-absorption edge. A careful examination shows that no apparent photooxidation/reduction of arsenic was observed for the solid samples under the present synchrotron experiment conditions. The reference sample of strengite containing As(V) was prepared following the adsorption method of Farqufar et al. (9). A strengite sample was collected from the Shunomata mine (20), Akita prefecture, Japan. The adsorption experiments were carried out in a 50 mL polypropylene tube by placing 0.3 g of strengite in contact with purified water and spiking it with the required volume of an As(V) stock solution (prepared by dissolving KH2AsO4 in purified water). The pH was adjusted to the range of 2.5-3.0, and the final volume of the solution was 30 mL. The solution was kept overnight with continuous agitation and then filtered through a 0.2 µm cellulose membrane. The strengite adsorbed 6 mg/g As, which was determined by inductively coupled plasma-atomic emission spectroscopy. The XANES spectra were directly calculated from the intensity ratio of the fluorescent to incident X-ray intensities without any correction for self-absorption because of the thinness (thichkness:20 µm) of the samples. The normalized XANES spectra were compared to the reference compounds including KH2AsO4 and KAsO2. A pattern fitting of the

FIGURE 5. (a) Normalized k3-weighted EXAFS spectra and Fourier-filtered EXAFS spectra (black line) and least-squares fits(dotted line) for (b) P1, (c) P2, and (d) As in strengite.

FIGURE 6. Fourier transforms of k3(k) spectra of the EXAFS data of P1, P2, and strengite. The data were not corrected for the phase-shift.

TABLE 1. Structural Parameters Obtained from the Curve Fitting Analysis of the As K-edge EXAFS Data sample

atom

Na

r(Å)

σ2(Å)2

R

P1 (iron sediments with high As level) P2 (iron sediments with low As level) As(V) in strengite

O Fe

4.0 4.0

1.68 3.36

0.011 0.017

7.4

O

4.0

1.69

0.012

2.7

O Fe

4.0 4.0

1.68 3.35

0.009 0.025

9.8

O Fe

4 4

1.68 3.36

scorodite(22)

a N: coordination number; r: interatomic distance; σ2: Debye-Waller factor; R: reliability factor for goodness of fit.

experimental XANES spectra to those of the reference samples was carried out to evaluate the As oxidation state in the iron ore. For the EXAFS analysis, the data were backgroundcorrected, and the EXAFS oscillations χ(k) (where k is the photoelectron momentum) were extracted from the measured spectra using a spline-smoothing method performed with the EXAFS analysis program of RIGAKU REX2000 (version 2.5). The χ(k) spectrum was weighted by k3 and

Fourier-transformed to produce a radial structure function (RSF) using a k-range of approximately 3-12.5 Å-1. Distinct coordination shells of the RSF were then backtransformed to isolate the spectral contributions of each atomic shell. FEFF 8.4 was used to calculate the backscattering amplitude and the phase function (21). The interatomic distances (r) were typically determined with an accuracy of (0.02 Å and the coordination numbers (N) ( within 25% for the first shell, but the N for shells with a greater distance was less accurate because the data were obtained by the fluorescence mode utilizing an X-ray microbeam and single element SDD. The R factor, the goodness of fit, was calculated as follows:



R ) ∑{k3χobs(k) - k3χcal(k)}2 ∑{k3χobs(k)}2

(1)

The error of each parameter was estimated from the square root of the diagonal element of the covariance matrix. SR-XRD Analysis. To identify the crystalline phases of the As-accumulating spot on a micrometer scale, the SRXRD measurements were performed. The intense monochromatic X-rays (12.8 keV) with a beam size of 50 × 50 µm2 adjusted by slits were used for the measurement. The diffraction pattern was recorded by an Imaging Plate (IP, Fuji Film Co.) for 20-25 min in the transmission DebyeScherrer geometry without sample oscillation. The focal distance (between the IP and sample) was 229.2 mm, which was calculated by the diffraction pattern of the silicon standard (NIST SRM 640c). The measured IP data were processed with a BAS2500 reader (Fuji Film Co.).

Results and Discussion SR-XRF Imaging. Since the iron ore has a band structure of alternating jarosite-rich yellow layers and goethite-rich brown layers (16), a large X-ray beam was used to reveal the distribution of Fe and As in each layer. The pixel size of the imaging (i.e., step size of the XY stage) was equal to the beam size (50 × 50 µm2). Figure 1 shows the distribution of Fe and As and the optical microscope image of the iron ore sample (thin section), with the portion outlined by the black frame corresponding to the imaging area. It was found that the Fe, which is the major component of the ore, is distributed over entire specimen including both brown and yellow layers. On the other hand, As mainly distribute to the browm layers, which were found to contain many well-preserved algal fossils VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7155

FIGURE 7. Debye-Scherrer SR-XRD pattern measured at (a) P1 (high As level) and (b) P2 (low As level).

TABLE 2. X-ray Powder Diffraction Data of P1, P2, and Reference Minerals P1 d/Å I/I0

P2 d/Å I/I0

a strengite hkl d/Å I/I0

5.93 32 5.75 14 5.49 55

101 5.93 45 003 5.72 25 111 5.509 60

5.10 56 4.95 43 4.37 100 4.00 22

102 5.09 70 020 201 211 121 112

4.954 4.383 3.996 3.959 3.719

30 85 45 13 25

3.63 32 3.27 12 3.12 53 2.99 16 2.95 19 2.56 45

a jarosite hkl d/Å I/I0

3.11 72 3.07 100 2.97 12 2.88 8 2.55 20

110 3.65 40 221 3.281 17 122 3.114 100 311 3.002 45 131 2.949 45 231 2.631 11 132 2.546 50

201 3.11 75 113 3.08 100 202 2.965 15 006 2.861 30 204 2.542 30

a

Data from the International Centre for Diffraction DatasPowder Diffraction File (ICDD-PDF), strengite PDF No. 33-667, jarosite PDF No. 22-827 (25).

and do not significantly accumulated in the yellow layers, which primarily contained jarosite. To reveal the distribution of elements in detail, SR-µ-XRF imaging was performed for the As-accumulating layer by utilizing an X-ray microbeam with the size of 1.8 (V) × 2.8 (H) µm2 (pixel size was 2.0 (V) × 3.0 (H) µm2). These results are shown in Figure 2. High spatial resolution imaging with a fine pixel size (1 × 1 µm2) was used to reveal the detailed distribution of the As-accumulating phase. These results are presented in Figure 2d and e. The As accumulated in the form of a peculiar concentric circle with a typical radius of ca. 20 µm. Because the experiments were carried out in air and the energy of the incident X-rays was high (12.8 keV) in order to efficiently excite the As KR line, the detection efficiency of the light elements, such as K, P, and S, was low. Therefore, analysis of the light elements was performed by SEM-EDS. Figure 3 shows SEM images of the area measured by the SR-µ-XRF imaging (Figure 2d and e). In areas with high levels of As accumulation, a concentric morphology was observed. The concentric morphology was hardly 7156

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008

observed in the yellow layer. The EDS analysis was made as spot analysis with a beam size of ca. 1 µmφ. The analyzed area is shown with the apex of triangle. The EDS spectrum for this portion (P1) showed strong peaks of P and Fe, and a small peak of As (Figure 3b). Highest As portion measured by SEM-EDS was a few wt % level as As2O3. On the other hand, the EDS spectrum of P2 is composed of peaks of K, S, and Fe, but that of As is lacking (Figure 3c). The measured points at P1 and P2 are also shown in Figure. 2d. The SEMEDS analysis and SR-µ-XRF imaging revealed that As was selectively accumulated in the phase showing a concentric morphology. XAFS Analysis. As K-edge XANES spectra of the iron ore at areas with high (P1) and low (P2) levels of As accumulation are shown in Figure 4. The XANES spectra of As(V) in strengite, KH2AsO4, and KAsO2 are also shown as reference spectra of the As(V) and As(III) states. The analyzed points of P1 and P2 are shown in Figures 2d and 3a. The As K-absorption edges for all samples from the iron ore were located at a similar energy to that of the As(V) standards (arsenate). In contrast, arsenite has an absorption edge at approximately 4 eV lower energy than that of the As(V). Consequently, the predominant valence state of As in these solid phases was found to be As(V). Moreover, pattern fitting of the XANES spectra suggested that there was no detectable As(III) species in these phases. The Normalized k3 weighted EXAFS oscillation of the three pahses are shown in Figure 5a. The As radial structure functions (RSFs) obtained from the Fourier transforms of the k3χ(k) EXAFS oscillation are shown in Figure 6. In all cases, the first peak in the Fourier transforms yields information regarding the As-O interaction of the first coordination sphere of the As atom. Fitting of the EXAFS oscillation obtained by inverse Fourier transforms yields the structural parameters of the first coordination shell of As surrounded by the O atoms. The results of the curve-fitting analysis of the EXAFS data are given in Table 1 together with the best least-squares fits shown in Figure 5b-d. The local As structure is typical of the tetrahedral coordination in the arsenate ion (AsO43-) with an As-O distance of approximately 1.68 Å. More distant shells could not be extracted from the Fourier transform of the P2 point because of the low As level. The second peaks of P1 and strengite in Figure 6 are due to the As-Fe interaction. Fitting the inverse Fourier transform of the Fe shell of the P1 data yields a fit with 4 As-Fe bonds at 3.36 Å (Table 1) These structural parameters for P1 and As(V) in strengite are similar to those of As(V) in

scorodite as in Table 1 (22). A similar As-Fe distances are also reported for the jarosite containing significant amount of AsO4 octaherada (23, 24). SR-XRD Analysis. Figure 7 shows the Debye-Scherrer XRD ring patterns measured at P1 (high As level) and P2 (low As level). The measured points were the same as those used for the EXAFS measurements. Table 2 shows the diffraction data for each point and those of reference samples for comparison. As shown in Table 2, the P1 phase is identified as strengite, and the P2 phase is identified as jarosite based on the diffraction data. The Debye rings for the P1 and P2 phases were successfully indexed with strengite and jarosite, respectively (Figure 7). No other clear diffraction lines were observed. The high backgrounds are due to the quartz glass substrate and may be partly due to unknown amorphous or poorly crystalline phase such as ferrihydrite. The results show that Goethite was not observed by this analysis. Based on these results, it is found that As had selectively accumulated in the strengite. The EXAFS analysis of P1 and As(V) in the strengite suggests that As exists as the typical tetrahedral coordination in the form of arsenate ion (AsO43-). Based on the results of the SR-XRD and XAFS analyses, it may be concluded that As exists in strengite by isomorphous substitution of AsO43- for PO43-, and is mineralogically expressed as a solid solution of strengite and scorodite, though they form only a partial solid-solution series. The As-Fe interaction at a distance of 3.36 Å observed for the P1 phase and As(V) in strengite (Table 1) is indicative of the existence of the corner-sharing coordination between the AsO4 tetrahedra and the Fe(III) octahedra. The EXAFS analysis and SR-XRD data of the P2 phase suggest that As adsorbed on jarosite also exists in an arsenate form. This finding is consistent with previous finding by Foster et al. (6), Paktunc and Dutrizac (23),Savage et al. (24) that jarosite contain As by isomorphous substitution of the AsO4 tetrahedron for the SO4 tetrahedron. They studied the substitution by XAFS and diffraction analysis. The As-O distance of 1.69 Å for the P2 phase is in good agreement with the As-O distance in jarosite (1.69 Å (6), 1.70 Å (24)) reported by them, though the As-Fe bond at 3.29 Å (6), 3.27 Å (7) was not detected in our study because of the low level of As. In conclusion, the SR multiple analyses (XRF imaging, XAFS, and XRD) were effectively combined to elucidate the As solid-state speciation in the iron ore. It was found from the SR-XRF imaging and SEM-EDS analysis that As accumulated in the brown layer together with P and Fe, while Fe was distributed to both layers (Figure 1), and many biologically induced minerals with a concentric morphology were about 20 times more frequently seen in the brown layer than in the yellow layer (Figure 3a). Though goethite and jarosite also have an As adsorption ability, our SR-µ-XRF imaging and SR-XRD analysis revealed that As was selectively accumulated in the strengite (see Figures 2d and 7a, Table 2)and exists as AsO4 octahedra being substituting for PO4 octahedra of strengite, which is a solid solution of strengite FePO4 · 2H2O and scorodite FeAsO4 · 2H2O. However, because of the size difference between AsO4 and PO4, no continuous solid solution was observed. Furthermore, it is interesting to note that strengite containing high levels of As has a peculiar concentric morphology (Figure 3a), implying a relationship with biological processes. In fact, concentric structures in two-dimensional micrographs and ball-like spherical grains in three-dimensional micrographs have been reported for algal and moss-aggregate ores from the Gunma iron deposit (16). The biomineralization of hydrous Fe(III) oxides can be accompanied by incorporation of other dissolved constituents, such as PO4 (26), and this incorporation appears to be a common occurrence in natural systems where microbial cells serve as nucleation sites for the oxidation and or precipitation of Fe(III) minerals (27). Strengite may have been

formed by the following process: Fe2+ is oxidized by bacteria (16), and Fe (III) phosphates are formed using PO43- ions derived from the nearby environment or the bacterial cells themselves (28). The AsO43- ions may be simultaneously incorporated into the strengite crystal lattice to form a slid solution during this process or they may substitute for the PO43- ions in strengite during a later alternation process. The specific correlation between the distribution of As and biominerals, and the morphological evidence of strengite suggest that some biological effect seems to play an important role in the As-accumulating process of this area. This study has provided us with a better understanding of the natural production processes of the ores and may help in designing efficient As-removal processes for environmental remediation.

Acknowledgments The synchrotron radiation experiments were performed at the BL37XU at SPring-8 with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (Proposal No. 2006A1684). We are grateful to Dr. A. Hokura, Tokyo University of Science, for her kind support and advice.

Literature Cited (1) Matschullat, J. Arsenic in the geospheresA review. Sci. Total Environ. 2000, 249, 297–312. (2) Shimada, N. Arsenic in groundwater: The chemical behavior, source and relevant problems. Shigen-chishitsu 2002, 53 (2), 161–172. (3) Fukushi, K.; Sasaki, M.; Sato, T.; Yanase, N.; Amano, H.; Ikeda, H. A natural attenuation of arsenic in drainage from an abandoned arsenic mine dump. Appl. Geochem. 2003, 18, 1267– 1278. (4) Langner, H. W.; Jackson, C. R.; McDermott, T. R.; Inskeep, W. P. Rapid oxidation of arsenite in a hot spring ecosystem, Yellowstone National Park. Environ. Sci. Technol. 2001, 35, 3302–3309. (5) Tomioka, Y.; Hiroyoshi, N.; Kubo, Y.; Tsunekawa, M. Effect of jarosite on the removal of arsenic ions in sulfuric acid solution. Shigen-to-Sozai. 2005, 121, 597–602. (6) Foster, A. L.; Brown, G. E., Jr.; Tingle, T. N.; Parks, G. A. Quantitative arsenic speciation in mine tailings using X-ray absorption spectroscopy. Am. Mineral. 1998, 83, 553–568. (7) Carlson, L.; Bigham, J. M.; Schwertmann, U.; Kyek, A.; Wagner, F. Scavenging of As from acid mine drainage by schwertmannite and ferrihydrite: A comparison with synthetic analogues. Environ. Sci. Technol. 2002, 36, 1712–1719. (8) Redman, A. D.; Macalady, D. L.; Ahmann, D. Natural organic matter affects arsenic speciation and adsorption onto hematite. Environ. Sci. Technol. 2002, 36, 2889–2896. (9) Farqufar, M. L.; Charnock, M. J.; Livens, R. F.; Vaughan, J. D. Mechanisms of arsenic uptake from aqueous solution by interaction with goethite, lepidocrocite, mackinawite, and pyrite: An X-ray absorption spectroscopy study. Environ. Sci. Technol. 2002, 36, 1757–1762. (10) Ve¨ronique, L.; Christelle, L.; Ve¨ronique, D.; Bernard, D.; JeanClaude, B. Arsenic removal by adsorption on iron(III) phosphate. J. Hazard. Mater. 2005, 123, 262–268. (11) Brown, G. E. How minerals react with water. Science. 2001, 294, 67–70. (12) Morin, G.; Juillot, F.; Casiot, C.; Bruneel, O.; Personne, J.; ElbazPoulichet, F.; Leblanc, M.; Ildefonse, P.; Calas, G. Bacterial formation of tooeleite and mixed arsenic(III) or arsenic(V)iron(III) gels in the carnoule’s acid mine drainage, France. A XANES, XRD, and SEM study. Environ. Sci. Technol. 2003, 37, 1705–1712. (13) Savage, K. S.; Tingle, T. N.; O’Day, P. A.; Waychunas, G. A.; Bird, D. K. Arsenic speciation in pyrite and secondary weathering phases, Southern Mother Lode Gold District, Tuolumne County, California. Appl. Geochem. 2000, 15, 1219–1244. (14) Islam, R.; Shiraki, K.; Miyata, K.; Koji, N.; Fujisawa, A.; Tazaki, K. Clay minerals and phosphates found in biomats by the side of the well. J. Clay Sci. Soc. Jpn. 2003, 43, 14–22. (15) Kato, Y.; Ikeda, T.; Moriguchi, E. Unique geochemistry of sedimentary iron deposit formed by biologically induced mineralization. Resour. Geol. 2002, 52, 123–134. (16) Akai, J.; Akai, K.; Ito, M.; Nakano, S.; Maki, Y.; Sasagawa, I. Biologically induced iron ore at Gunma iron mine, Japan. Am. Mineral. 1999, 84, 171–182. VOL. 42, NO. 19, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

7157

(17) Pan, Y.; Brown, A.; Brydson, R.; Warley, A.; Li, A.; Powell, J. Electron beam damage studies of synthetic 6-line ferrihydrite and ferritin molecule cores within a human liver biopsy. Micron 2006, 37 (5), 403–411. (18) Terada, Y.; Goto, S.; Takimoto, N.; Takeshita, K.; Yamazaki, H.; Shimizu, Y.; Takahashi, S.; Ohashi, H.; Furukawa, Y.; Matsushita, T.; Ohata, T.; Ishizawa, Y.; Uruga, T.; Kitamura, H.; Ishikawa, T.; Hayakawa, S. Construction and Commissioning of BL37XU at SPring-8. AIP Conference Proceedings, 2004; 705, 376-379. (19) Takeuchi, A.; Suzuki, Y.; Takano, H. X-ray Focusing Mirror Fabricated with Bent-Polishing Method. AIP Conference Proceedings, 2004; 705, 760-763. (20) Suo, T.; Nakata, M. Tinticite from Shunomata mine, Akita prefecture. Chigakukenkyuu. 2004, 53, 67–73. (21) Ankudinov, A. L.; Ravel, B.; Rehr, J. J.; Conradson, S. D. Realspace multiple-scattering calculation and interpretation of x-rayabsorption near-edge structure. Phys. Rev. B. 1998, 58, 7565– 7576. (22) Waychunas, G. A.; Rea, B. A.; Fuller, C. C.; Davis, J. A. Surface chemistry of ferrihydrite: Part 1. EXAFS studies of the geometry of coprecipitated and adsorbed arsenate. Geochim. Cosmochim. Acta 1993, 57, 2251–2269.

7158

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 19, 2008

(23) Paktunc, D.; Dutrizac, J. E. Characterization of arsenate-forsulfate substitution in synthetic jarosite using X-ray diffraction and X-ray absorption spectroscopy. Can. Mineral. 2003, 41, 905– 919. (24) Savage, K. S.; Dennis, K. B.; O’Day, P. A. Arsenic speciation in synthetic jarosite. Chem. Geol. 2005, 215, 473–498. (25) Mineral Powder Diffraction File. The International Centre for Diffraction Data, 1986. (26) Karl, D. M.; McMurtry, G. M.; Malahoff, A.; Garcia, M. O. Loihi Seamount Hawaii: a mid-plate volcano with a distinctive hydrothermal system. Nature. 1988, 335, 532–535. (27) Konhauser, K. O. Diversity of bacterial iron mineralization. EarthSci. Rev. 1998, 43, 91–121. (28) Konhauser, K. O.; Fyfe, W. S.; Shultze-Lam, S.; Ferris, F. G.; Beveridge, T. J. Iron phosphate precipitation by epilithic microbial biofilms in Arctic Canada. Can. J. Earth Sci. 1994, 31, 1320–1324.

ES8006518