Microscale Imaging and Identification of Fe Speciation and Distribution

Apr 25, 2011 - Joshua J. LeMonte , Jason W. Stuckey , Joshua Z. Sanchez , Ryan ... Carolyn I. Pearce , Michael J. Wilkins , Changyong Zhang , Steve M...
0 downloads 0 Views 3MB Size
ARTICLE pubs.acs.org/est

Microscale Imaging and Identification of Fe Speciation and Distribution during Fluid Mineral Reactions under Highly Reducing Conditions L. E. Mayhew,*,† S. M. Webb,‡ and A. S. Templeton† † ‡

Department of Geological Sciences, UCB 399, University of Colorado, Boulder, Colorado 80309, United States Stanford Synchrotron Radiation Lightsource, 2525 Sand Hill Road, Menlo Park, California 94025, United States

bS Supporting Information ABSTRACT: The oxidation state, speciation, and distribution of Fe are critical determinants of Fe reactivity in natural and engineered environments. However, it is challenging to follow dynamic changes in Fe speciation in environmental systems during progressive fluid mineral interactions. Two common geological and aquifer materials—basalt and Fe(III) oxides—were incubated with saline fluids at 55 °C under highly reducing conditions maintained by the presence of Fe0. We tracked changes in Fe speciation after 48 h (incipient water rock reaction) and 10 months (extensive water rock interaction) using synchrotronradiation μXRF maps collected at multiple energies (ME) within the Fe K-edge. Immediate PCA analysis of the ME maps was used to optimize μXANES analyses; in turn, refitting the ME maps with end-member XANES spectra enabled us to detect and spatially resolve the entire variety of Fe-phases present in the system. After 48 h, we successfully identified and mapped the major Febearing components of our samples (Fe(III) oxides, basalt, and rare olivine), as well as small quantities of incipient brucite associated with olivine. After 10 months, the Fe(III)-oxides remained stable in the presence of Fe0, whereas significant alteration of basalt to minnesotaite and chlinochlore had occurred, providing new insights into heterogeneous Fe speciation in complex geological media under highly reducing conditions.

’ INTRODUCTION Iron is the most abundant transition metal in Earth surface environments and is dynamically cycled between oxidized and reduced forms in numerous biogeochemical processes.1 Fe minerals often dominate the reactivity of soils, sediments, rock matrices, and engineered environments such as permeable reactive barriers (PRB). Any environmental or geological sample will likely contain iron in one or more of its three valence states (Fe0, Fe(II), and Fe(III)) and in numerous chemical forms. One of the greatest challenges is to identify the most reactive iron phases and secondary precipitates that may be present, even in low abundance. It is also difficult to measure dynamic changes in Fe speciation in environmental systems during progressive fluid mineral interactions. Thus, there is a need to be able to rapidly profile not only the bulk average oxidation state and speciation of Fe, but to identify the distribution and abundance of iron phases at the microscale as a function of time. One of our major goals is to determine the abiotic mineralogical transformations of common Fe-bearing geological substrates, basalt and Fe(III)-oxides, under highly reducing conditions. Moderate temperature (24 h. All experimental apparatus and liquid media were autoclaved and kept sterile prior to assembly. Basalt powder (1.25 g), 0.1 g of Fe0 chips (1 2 mm), 0.25 g of basalt chips (2 5 mm), 35 mL of anaerobic artificial seawater media,16 0.1 mL of 1 M Fe(III) oxides, and 150 mM sulfate were combined in a serum vial with an 80% N2, 20% CO2 headspace. The vial was capped with an airtight rubber stopper and incubated at 55 °C. After 48 h and 10 months, 1 mL of mineral bearing media was removed using a 23 g needle, resulting in the preferential selection of particles 100 °C) of banded iron formations.32 The detection of minnesotaite at 55 °C may be the lowest temperature occurrence that has been documented. Low-Fe clinochlore is often associated with hydrothermal environments and serpentinization processes;33,34 however, low Fe-clinochlore is stable across a wide range of [Si(aq)] at both 25 and 75 °C.35 The detection of minnesotaite and clinochlore in the 10-month sample reflect significant changes in both aqueous and solid phase chemistry as the water rock reaction progressed. Fit of ME Maps with End-Member XANES. Identification of end-member XANES and subsequent least-squares fitting with Fe model compounds enables an understanding of the variety of Fe species present at 48 h and 10 months of reaction progress. Using EM spectra to fit non-EM spectra produces fits with excellent R2 values (e.g., Figure S6), supporting our choice of EM XANES and the validity of characterizing the entire sample with the smaller subset of XANES data. However, fitting the XANES

ARTICLE

spectra only informs us of the Fe-speciation at the points where the XANES spectra were collected. Further data processing is required to understand the spatial distribution of the Fe phases represented by the EM spectra. Ultimately, our goal was to determine the specific distribution and abundance of the Fe phases present in our samples. To accomplish this, we conducted LSF of the multiple energy μXRF maps collected from each sample. With hard X-ray data, the most common procedure is to fit ME maps to represent the distribution of discrete valence states (13 and references therein) (e.g., Figure 1c). In this work, we collected five μXRF maps across the Fe K-edge, and because we identified only four EM XANES it is possible to fit the μXRF maps with the EM XANES. The images generated represent the specific distribution of the four EM XANES spectra in each sample (Figure 3; Tables S2 and S3) and therefore correspond to the distribution of unique Fe-bearing phases. In the 48-h sample, EM XANES maps make it clear that most of the particles possess some amount of basalt glass (48/7 map, Figure 3a), the exception being the Fe(III)-bearing particles (48/9 map, Figure 3b). Compared to basalt, the other Fe phases (48/9, 48/2, and 48/3) are much more limited in occurrence (Figure 3a d). In this relatively simple experimental system it is possible to observe straightforward correspondences between PCA maps and maps of EM XANES. When compared to the component maps (Figure 2), there is a strong correspondence between map component 1 which approximates total Fe distribution (Figure 2a) and basalt (48/7, Figure 3a); this is not surprising because basalt is the primary material reacted with artificial seawater in this experiment. We also observe a strong similarity in the distribution of map component 2 (Figure 2b), which distinguishes phases on the basis of oxidation state, and the Fe(III)-bearing phases (48/9, Figure 3b). To fully describe the speciation of component 3 from PCA of the ME maps, both 48/2 and 48/3 are necessary, suggesting that map component 3 represents a more complex assemblage of Fe-bearing phases that are distinct from the basalt and Fe(III)phases. As discussed above, both 48/2 and 48/3 represent a minor component of the experimental system, olivine, ( brucite, an incipient alteration product. Thus, map component 3 appears to be particularly sensitive to localization of rarer phases and secondary products compared to the dominant basalt and Fe(III)-oxides represented by map components 1 and 2 (Figure 2). However, in order to visualize the distributions of the different Fe-bearing mineral assemblages within map component 3 it was necessary to fit the μXRF maps with the EM XANES (Figure 3c,d). EM XANES maps of the 10-month sample indiate that there is significantly less basalt glass (10/14 map, Figure 3f) relative to the 48-h sample (see Figures 1a and S4f for further evidence). In contrast, the 10/5 map reveals that the secondary Fe-silicates, minnesotaite and clinochlore, are pervasive and widely distributed across the mapped area (Figure 3i). However, it is important to note that the EM maps are not scaled to directly represent mineral abundance; for example, although the minnesotaite/ clinochlore areas in the 10/5 map appear to have significant intensity (Figure 3i), upon examination of the total Fe map (Figure S4f) it is possible to see that these are areas of relatively low Fe abundance. Altogether, our interpretation is that the dissolution of basalt is directly coupled with the precipitation of minnesotaite and chlinochlore as the reaction progressed. Surprisingly, the Fe(III)-bearing phases (10/7 map, Figure 3g) have not significantly changed by 10 months of reaction. Although the particles are smaller and more widely distributed than those in the 48-h sample (Figure 3b), the XANES spectra for the 4472

dx.doi.org/10.1021/es104292n |Environ. Sci. Technol. 2011, 45, 4468–4474

Environmental Science & Technology Fe(III) particles are nearly identical over time, resulting in similar fits for 48/9 and 10/7 (Table S4). Although the Fe-mineralogy associated with ZVI-PRB has been shown to be highly heterogeneous over small spatial scales,9 and ferrihydrite is a common reaction product that can even coat Fe0 surfaces,8 typically oxic fluids must be reacting with the barrier to stabilize Fe(III)-oxides. Instead, we expected to observe reductive dissolution and conversion of the Fe(III)-oxides to magnetite due to large fluxes of Fe(II) produced during anoxic corrosion of the Fe0. However, ferrihydrite will be stabilized in the presence of dissolved ligands such as the high concentrations of calcium, sulfate, and dissolved silica in our experiments (e.g., 10, 36) (Table S4). Moreover, the conversion of ferrihydrite to magnetite only occurs above a threshold of ∼300 μM Fe(II),10,29 whereas the Fe(II) concentration remained at a steady-state value of ∼100 μM throughout these experiments (Table S5). Therefore we have determined that the persistent precipitation of minnesotaite and chlinochlore buffered [Fe(II)(aq)] to low values, resulting in a high Si/Fe ratio (>4) and long-term passivation of the Fe(III)-oxides. Assessment of Integrated Method and Broad Applicability. In this work we have clearly defined an integrated approach to μXRF and μXANES data collection and processing that optimizes for identification of numerous Fe-bearing species in complex samples. PCA of the ME maps identified unique Fe-bearing components and this information guided our μXANES spectra collection and thereby enabled analysis of the variety of phases represented by the components. Using PCA of XANES spectra and LSF of end-member XANES we were then able to specifically identify the diverse Fe-bearing species within each component (to the mineral group level). Using only spectral analysis, we independently corroborated the presence of the initial bulk reaction materials. By conducting PCA of the ME maps prior to μXANES spectroscopy, incipient reaction products such as brucite can be detected in the 48-h sample, even though they represent a very minor proportion of the total Fe-speciation within the sample. The detection of new secondary minerals, minnesotaite and chlinochlore, in the 10month sample illustrates that our integrated approach to data collection and processing enables us to track changes in microscale Fe-speciation during anaerobic water basalt reactions over time. Notably, a number of the particles chosen for μXANES analyses were relatively small and nondescript in terms of Fe fluorescence intensity; for example, the spots containing minnesotaite/clinochlore or olivine/brucite do not necessarily correspond to either the most intense areas of Fe fluorescence (Figure S4e) or welldefined particles (Figure S4f). Thus without conducting PCA of the ME maps we may not have chosen them for μXANES and would have missed sampling unique Fe species within the sample area. Subsequent XANES fitting of the ME maps allows the ME maps to be quantitatively reassessed to reveal the spatial distribution and abundance of each EM XANES spectrum. It was even possible to visualize highly localized incipient reaction products (brucite) that comprise only a small fraction of the total Fe in the sample (but must occur in proportions of g10% in a 2.5  2.5 μm pixel). For the 10-month sample, the loss of basalt and diffuse growth of minnesotaite/chlinochlore regions was clearly visualized (Figure 3). Accurate detection and quantification of fluid mineral reaction products is crucial to understanding changes in Fe speciation and reactivity during secondary mineralization processes. This SRbased integrated approach to data collection and processing can be applied to complex geologic samples (e.g., soils, sediments, rock matrices, and mineral alteration rims) to detect micrometer-scale

ARTICLE

geochemical transformations of a target element. This technique is not limited to Fe-bearing species but rather can be used to detect and interrogate the distribution and speciation of any redox active element (e.g., Mn, U, Cr, As, S) where specific chemical forms can be resolved using XANES analyses, giving rise to species-dependent spatial variability in multiple-energy maps. In addition, since PCA can rapidly be conducted on the ME maps and μXANES data immediately after collection, additional information can be gleaned by targeted μXRD or μEXAFS analyses to better constrain the mineralogy within spots of interest. Ultimately, increasing availability of brighter synchrotron radiation sources and development of new detector technologies (e.g.,37) will enable acquisition of hard X-ray fluorescence data at more energies more quickly, producing image stacks similar to those that are commonly generated in transmission mode for low-Z elements,18 and thereby enabling full-spectral analysis of element speciation in every pixel. In the meantime, the method employed here can be easily and effectively applied to environmental samples to be analyzed at almost any existing microspectroscopy beamline.

’ ASSOCIATED CONTENT

bS

Supporting Information. Further explanation of SRbased data collection and processing, and additional figures and tables that enhance these descriptions. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*Tel: (303) 492 9415; fax (303) 492 2606; e-mail: mayhewl@ colorado.edu.

’ ACKNOWLEDGMENT We thank C. Hansel and P. O’Day for providing reference Fe XANES spectra, E. Swanner for assistance collecting and collating reference Fe XANES spectra, and G. Lau for help with synchrotron μXRF analyses. We are grateful to the Smithsonian National Museum of Natural History for providing minerals for use as Fe model compounds. This work was directly supported by the David and Lucille Packard Foundation (A.S.T.) and the NASA Astrobiology Institute through a cooperative agreement with the University of Colorado. The synchrotron work was conducted on Beamlines 2-3, 11-2, and 4-1 at the Stanford Synchrotron Radiation Lightsource (SSRL), a national user facility operated by Stanford University on behalf of the Department of Energy, Office of Basic Energy Sciences, through the Structural Molecular Biology Program, supported by DOE Office of Biological and Environmental Research and the National Institutes of Health. We also appreciate the constructive comments of three anonymous reviewers. ’ REFERENCES (1) O’Day, P.; Rivera, N.; Root, R.; Carroll, S. X-ray absorption spectroscopic study of Fe reference compounds for the analysis of natural sediments. Am. Mineral. 2004, 89, 572–585. (2) Moody, J. B. Serpentinization: a review. Lithos 1976, 9, 125–138. (3) Fruh-Green, G.; Connolly, J.; Plas, A. In The Subseafloor Biosphere at Mid-Ocean Ridges; Wilcock, W., DeLong, E., Kelley, D., Baross, J., Cary, S., Eds.; American Geophysical Union: Washington, DC, 2004; pp 119 136. 4473

dx.doi.org/10.1021/es104292n |Environ. Sci. Technol. 2011, 45, 4468–4474

Environmental Science & Technology (4) Stevens, T. O.; McKinley, J. P. Abiotic controls on H-2 production from basalt-water reactions and implications for aquifer biogeochemistry. Environ. Sci. Technol. 2000, 34, 826–831. (5) de Combarieu, G.; Schlegel, M. L.; Neff, D.; Foy, E.; Vantelon, D.; Barboux, P.; Gin, S. Glass-iron-clay interactions in a radioactive waste geological disposal: An integrated laboratory-scale experiment. Appl. Geochem. 2010, 26, 65–79. (6) Goldberg, D. S.; Takahashi, T.; Slagle, A. L. Carbon dioxide sequestration in deep-sea basalt. Proc. Natl. Acad. Sci. U. S. A. 2008, 105, 9920. (7) Gu, B.; Phelps, T. J.; Liang, L.; Dickey, M. J.; Roh, Y.; Kinsall, B. L.; Palumbo, A. V.; Jacobs, G. K. Biogeochemical dynamics in zerovalent iron columns: Implications for permeable reactive barriers. Environ. Sci. Technol. 1999, 33, 2170–2177. (8) Furukawa, Y.; Kim, J.; Watkins, J.; Wilkin, R. T. Formation of Ferrihydrite and Associated Iron Corrosion Products in Permeable Reactive Barriers of Zero-Valent Iron. Environ. Sci. Technol. 2002, 36, 5469–5475. (9) Phillips, D. H.; Gu, B.; Watson, D. B.; Roh, Y.; Liang, L.; Lee, S. Y. Performance Evaluation of a Zerovalent Iron Reactive Barrier: Mineralogical Characteristics. Environ. Sci. Technol. 2000, 34, 4169–4176. (10) Hansel, C. M.; Benner, S. G.; Fendorf, S. Competing Fe(II)Induced Mineralization Pathways of Ferrihydrite. Environ. Sci. Technol. 2005, 39, 7147–7153. (11) Dixit, S.; Hering, J. G. Comparison of arsenic (V) and arsenic (III) sorption onto iron oxide minerals: Implications for arsenic mobility. Environ. Sci. Technol. 2003, 37, 4182–4189. (12) Tufano, K. J.; Fendorf, S. Confounding Impacts of Iron Reduction on Arsenic Retention. Environ. Sci. Technol. 2008, 42, 4777–4783. (13) Marcus, M. A. X-ray photon-in/photon-out methods for chemical imaging. TrAC, Trends Anal. Chem. 2010, 29, 508–517. (14) Lombi, E.; Susini, J. Synchrotron-based techniques for plant and soil science: Opportunities, challenges and future perspectives. Plant Soil 2009, 320, 1–35. (15) Ryden, J.; Mclaughlin, J.; Syers, J. Mechanisms Of Phosphate Sorption By Soils And Hydrous Ferric-Oxide Gel. J. Soil Sci. 1977, 28, 72–92. (16) Templeton, A. S.; Staudigel, H.; Tebo, B. M. Diverse Mn(II)oxidizing bacteria isolated from submarine basalts at Loihi Seamount. Geomicrobiol. J. 2005, 22, 127–139. (17) Webb, S. M. SIXPack: a graphical user interface for XAS analysis using IFEFFIT. Phys. Scr. 2005, 2005, 1011. (18) Lerotic, M.; Jacobsen, C.; Schafer, T.; Vogt, S. Cluster analysis of soft X-ray spectromicroscopy data. Ultramicroscopy 2004, 100, 35–57. (19) Fay, M. J.; Proctor, A.; Hoffmann, D. P.; Houalla, M.; Hercules, D. M. Determination of the Mo surface environment of Mo/TiO2 catalysts by EXAFS, XANES and PCA. Microchim. Acta 1992, 109, 281–293. (20) Ressler, T.; Wong, J.; Roos, J.; Smith, I. Quantitative speciation of Mn-bearing particulates emitted from autos burning (methylcyclopentadienyl)manganese tricarbonyl-added gasolines using XANES spectroscopy. Environ. Sci. Technol. 2000, 34, 950–958. (21) Wasserman, S. R. The Analysis of Mixtures: Application of Principal Component Analysis to XAS Spectra. J. Phys. IV France 1997, 7, C2-203–C2-205. (22) Malinowski, E. Determination Of Number Of Factors And Experimental Error In A Data Matrix. Anal. Chem. 1977, 49, 612–617. (23) Manceau, A.; Marcus, M.; Tamura, N. Quantitative speciation of heavy metals in soils and sediments by synchrotron X-ray techniques. Appl. Synchrotron Radiat. Low-Temp. Geochem. 2002, 49, 341–428. (24) Manceau, A.; Matynia, A. The nature of Cu bonding to natural organic matter. Geochim. Cosmochim. Acta 2010, 74, 2556–2580. (25) Bateson, A.; Curtiss, B. A method for manual endmember selection and spectral unmixing. Remote Sens. Environ. 1996, 55, 229–243. (26) Sutton, S.; Bajt, S.; Delaney, J.; Schulze, D.; Tokunaga, T. Synchrotron X-Ray-Fluorescence Microprobe - Quantification And Mapping Of Mixed-Valence State Samples Using Micro-Xanes. Rev. Sci. Instrum. 1995, 66, 1464–1467.

ARTICLE

(27) Templeton, A. S.; Knowles, E. J.; Eldridge, D. L.; Arey, B. W.; Dohnalkova, A. C.; Webb, S. M.; Bailey, B. E.; Tebo, B. M.; Staudigel, H. A seafloor microbial biome hosted within incipient ferromanganese crusts. Nat. Geosci. 2009, 2, 872–876. (28) Vekemans, B.; Janssens, K.; Vincze, L.; Aerts, A.; Adams, F.; Hertogen, J. Automated segmentation of mu-XRF image sets. X-Ray Spectrom. 1997, 26, 333–346. (29) Hansel, C. M.; Benner, S. G.; Neiss, J.; Dohnalkova, A.; Kukkadapu, R. K.;Secondary mineralization pathways induced by dissimilatory iron reduction of ferrihydrite under advective flow. Geochim. Cosmochim. Acta 2003, 67, 2977–2992. (30) McCollom, T. M.; Bach, W. Thermodynamic constraints on hydrogen generation during serpentinization of ultramafic rocks. Geochim. Cosmochim. Acta 2009, 73, 856–875. (31) Klein, F.; Bach, W.; Jons, N.; McCollom, T.; Moskowitz, B.; Berquo, T. Iron partitioning and hydrogen generation during serpentinization of abyssal peridotites from 15 N on the Mid-Atlantic Ridge. Geochim. Cosmochim. Acta 2009, 73, 6868–6893. (32) Klein, C. Some Precambrian banded iron-formations (BIFs) from around the world: Their age, geologic setting, mineralogy, metamorphism, geochemistry, and origins. Am. Mineral. 2005, 90, 1473–1499. (33) Iyer, K.; Austrheim, H.; John, T.; Jamtveit, B. Serpentinization of the oceanic lithosphere and some geochemical consequences: Constraints from the Leka Ophiolite Complex, Norway. Chem. Geol. 2008, 249, 66–90. (34) Seyfried, W. E.; Pester, N. J.; Ding, K.; Rough, M. Vent fluid chemistry of the Rainbow Hydrothermal System (36° N, MAR): Phase equilibria and in-situ pH controls on subseafloor alteration processes. Geochim. Cosmochim. Acta 2011, 75, 1574–1593. (35) Aja, S. U.; Small, J. S. The solubility of a low-Fe clinochlore between 25 and 175 degrees C and P v= P H2O. Eur. J. Miner. 1999, 11, 829. (36) Voegelin, A.; Kaegi, R.; Frommer, J.; Vantelon, D.; Hug, S. J. Effect of phosphate, silicate, and Ca on Fe (III)-precipitates formed in aerated Fe (II)-and As (III)-containing water studied by X-ray absorption spectroscopy. Geochim. Cosmochim. Acta 2010, 74, 164–186. (37) Ryan, C. G.; Siddons, D. P.; Moorhead, G.; Kirkham, R.; Dunn, P. A.; Dragone, A.; De Geronimo, G. Large detector array and real-time processing and elemental image projection of X-ray and proton microprobe fluorescence data. Nucl. Instrum. Methods Phys. Res., Sect. B 2007, 260, 1–7.

4474

dx.doi.org/10.1021/es104292n |Environ. Sci. Technol. 2011, 45, 4468–4474