International
Tables for
Crystallography
Volume I
X-ray absorption spectroscopy and related techniques
Edited by C. T. Chantler, F. Boscherini and B. Bunker

International Tables for Crystallography (2022). Vol. I. Early view chapter
https://doi.org/10.1107/S1574870720004784

Applications of XAS in earth sciences

Max Wilkea*

aInstitut für Geowissenschaften, Universität Potsdam, Karl-Liebknecht-Strasse 24-25, 14476 Potsdam, Germany
Correspondence e-mail: wilkem@uni-potsdam.de

X-ray absorption spectroscopy (XAS) is widely used in earth sciences and in this chapter an overview is given of various sample systems, experimental approaches and scientific questions. XAS is used to characterize the structural environment of elements in various minerals or glasses of natural rock samples, to determine elemental oxidation states, to understand the incorporation of trace elements in minerals and to distinguish different phases. Further, XAS is used in laboratory experiments simulating the conditions of the inaccessible Earth in order to characterize the structural, chemical and electronic properties of crystalline and liquid materials down to conditions of the Earth's core–mantle boundary. In particular, X-ray absorption near-edge structure (XANES) data have provided unprecedented insights into redox equilibria in silicate glasses and melts, whereas the local structure of major and minor components is widely determined by extended X-ray absoption fine structure (EXAFS). Finally, XAS has substantially contributed to the understanding of element dissolution and transport by aqueous fluids in the Earth's crust.

Keywords: earth sciences.

1. Introduction

X-ray absorption spectroscopy (XAS) has a long tradition in earth sciences. In general, it has been employed to complement results from techniques that are available in most laboratories, such as electron microscopy, X-ray diffraction and vibrational and optical spectroscopy. In particular, its element selectivity, its sensitivity to oxidation state and the possibility of studying noncrystalline systems and dilute concentrations have induced many applications. Since the 1980s, several review articles have provided evidence of the constant improvement in this respect as well as more detailed summaries than it is possible to provide here; for example, Brown et al. (1988[link]), Galoisy (2004[link]), Henderson, Neuville et al. (2014[link]) and Mottana (2004[link]).

2. XAS on minerals and rocks (including diamond inclusions)

In mineral sciences, XAS has often been used to improve the understanding of the crystal chemistry of elements in minerals, for example zirconium, uranium and thorium in metamict zircon and thorite (Farges & Calas, 1991[link]), titanium in various phases (Farges et al., 1997[link]) and magnesium in garnet (Quartieri et al., 2008[link]). XAS has often been used to investigate the substitution mechanism of trace elements in minerals, for example lanthanides in garnet (Quartieri, Antonioli et al., 1999[link]; Quartieri, Chaboy et al., 1999[link]; Quartieri et al., 2004[link]), zinc in phyllosilicates (Juillot et al., 2006[link]), chromium in garnet (Juhin et al., 2008[link]) and thorium in fluorapatite (Luo et al., 2011[link]). Due to the high sensitivity of X-ray absorption near-edge structure (XANES) spectra to the valence state, a vast number of studies have dealt with determining the oxidation state in minerals of iron (for example Berry et al., 2010[link]; Bourdelle et al., 2013[link]; Wilke et al., 2001[link], 2009[link]), manganese (for example Farges, 2005[link]), europium (for example Rakovan et al., 2001[link]), chromium (for example Farges, 2009[link]), vanadium (for example Simon et al., 2007[link]) and sulfur (for example Goett­licher et al., 2013[link]). Particularly for iron, the pre-edge region has turned out to be very useful. For a given oxidation state (Fe2+ or Fe3+), the pre-edge centroid position is constant (±0.05). The integrated area of the pre-edge feature, however, is a measure of the centrosymmetry of the site. For instance, square-planar (C4v) and octahedrally (Oh) coordinated iron(II) show similar areas despite their respective coordinations being four and six, respectively. In contrast, the tetrahedral configuration (Td) shows high areas because of its deviation from centrosymmetry. In the variogram (Fig. 1[link]), four end-members have been determined (Td and Oh for Fe2+ and Fe3+, respectively) and their `in-between' mixtures. These mixtures show nonlinear variations of the centroid position with the oxidation state, which is related to the fact that the pre-edges of Fe2+ and Fe3+ always differ in intensity. As a consequence, the variogram is fairly sensitive to minor amounts of Fe3+ co-existing with Fe2+, whereas for the opposite case this is much less true. Attempts to construct similar variograms for manganese (Farges, 2005[link]) and chromuium (Farges, 2009[link]) were not possible because of additional transitions that complicate the pre-edge feature. The variogram approach also works for titanium, for which it was originally designed (Farges, Brown & Rehr, 1996[link]).

[Figure 1]

Figure 1

Graph showing the pre-edge intensity versus pre-edge centroid position in a variogram after Wilke et al. (2001[link]). Points correspond to pre-edge data determined on synthetic basaltic and sodium silicate glasses. Grey fields designate pre-edge parameters for the iron coordination and oxidation state indicated. Dashed lines between fields indicate the variation of binary mixtures of the respective end-members. Filled and open circles, alkali basaltic glasses with and without phosphorus; open squares, Na2Si2O5 glass; open triangles, Na2Si3O7 glass. The grey solid curve shows the variation calculated from pre-edge parameters of completely reduced and oxidized end-members (modified after Wilke et al., 2004[link], 2005[link]).

Micro-XANES provides the possibility of obtaining spatially resolved information on the oxidation state at the same position that the chemical composition was determined (usually by an electron microprobe) within an assemblage of coexisting minerals. From the mineral composition of the assemblage, the conditions of formation may be deduced by chemical thermodynamics. Determination of the oxidation state allows additional insight into the redox conditions during formation. Furthermore, it allows differentiation between Fe2+ and Fe3+, as the two cations show different crystal-chemical behaviour. For example, the widely used geothermometers are based on Fe2+–Mg2+ partitioning between ferromagnesian silicate minerals, due to the fact that Fe2+ and Mg2+ have very similar ionic radii. Determination of the iron oxidation state enables correction for Fe3+, which may have a substantial influence on the resulting temperature estimated from the mineral assemblage (Schmid et al., 2003[link]). Muñoz et al. (2006[link]) developed a method to map out the oxidation state with a spatial resolution in the micrometre range, which they applied to determine the equilibration temperature in fine-grained chlorite aggregates (Fig. 2[link]). De Andrade et al. (2011[link]) presented an experimental protocol that is based on full-field transmission measurements with a large beam. Absorption radiographies are recorded using a CCD camera and a full XANES spectrum is recorded on each pixel by scanning the energy of the incident beam. This technique allows the acquisition of hyperspectral imaging of a millimetre-sized area within a couple of hours with submicrometre resolution. Finally, micro-XANES has been applied to study fossils, which provides insight into chemical processes during fossilization and preservation (see, for example, Egerton et al., 2015[link]). A related field that connects biological and geochemical processes are studies related to bio-mineralization, which look into variations of element concentration and speciation in, for example, mollusc shells or coral skeletons (Cuif et al., 2003[link]; Dauphin et al., 2005[link]).

[Figure 2]

Figure 2

Top: optical micrograph of a thin section of rock with chlorite in the middle, quartz at the top and phengite between the chlorite and quartz and at the bottom. Bottom: map of the edge position used as a proxy for the iron oxidation state (higher energy means more oxidized). Areas with a higher oxidation state correspond to dark brown areas in the optical micrograph (taken from Muñoz et al., 2006[link]). Copyright 2006 by the American Geophysical Union.

2.1. Diamond inclusions

In the deep Earth, diamonds form at pressures of above 4–6 GPa. Minerals trapped as inclusions thus represent the only direct samples of material in the Earth's deeper mantle. The application of hard X-rays to these inclusions has the unique advantage that data can be measured without cutting the diamond due to the low absorption of hard X-rays by diamond. The combination of micro-XANES and micro-X-ray fluorescence (micro-XRF) measurements yields information on the valence state and chemical composition, for example iron (Shiryaev et al., 2010[link]) or chromium (Kagi et al., 2013[link]). In combination with the mineral assemblage, the valence state provides information on the redox conditions prevalent in the deep mantle. The use of a confocal experimental setup was a fundamental improvement for micro-XRF and micro-XANES measurements on diamond inclusions, i.e. in addition to the highly focused X-ray beam, the XRF signal is detected through a focusing X-ray optic in front of the detector. The two foci of the optics define a sampling volume of a certain size (in a micrometre range). By scanning this sampling volume through a diamond, the three-dimensional distribution of mineral inclusions can be determined. Silversmit and coworkers investigated diamonds of ultradeep origin from Brazil at the Fe and Mn K edges (Silversmit et al., 2010[link], 2011[link]) and could identify ferropericlase and hematite close to calcium-rich inclusions (Fig. 3[link]). The presence of ferropericlase and hematite indicates the origin of the diamond to be the lower mantle (>600 km depth) with a subsequent oxidation event. The arrangement of the inclusions suggests fluid overprinting along an open crack.

[Figure 3]

Figure 3

Top left: intensity map of Fe Kα fluorescence radiation of one vertical confocal plane of diamond RS69. Top right: the corresponding confocal Fe K-edge XANES spectra for the indicated iron hotspots. Bottom left: combined 3D RGB image showing the distribution of the elements calcium (red), iron (green) and manganese (blue). Bottom right: interpolated composition image showing the proportion of ferropericlase (red) and hematite (green) in the inclusions. Reprinted with permission from Silversmit et al. (2011[link]). Copyright 2011 American Chemical Society.

3. In situ XAS in high-pressure experiments

Laboratory experiments under controlled high-pressure and high-temperature conditions are the only way to understand the properties and reactions of materials under the conditions of the deep Earth. Often, the samples can be quenched to ambient conditions without a significant change in structure and composition, so that any analysis is performed on the recovered sample. Many high-pressure compounds, however, cannot be quenched and thus have to be investigated in situ at high pressure and temperature. Diamond anvil cells are widely used for this because the diamond anvils are transparent to X-rays and are stable up to very high pressures and temperatures. Therefore, conditions down to the Earth's core can be simulated (>130 GPa and >3000 K; Boehler, 2005[link]). Typical diamond anvils are sufficiently transparent to X-rays down to ∼6 keV, i.e. XAS at the K edge of iron and elements with higher atomic numbers is feasible. In the case of studies on crystalline compounds, XAS is mostly used to provide information complementary to X-ray diffraction (XRD)methods. Kantor et al. (2006[link]) combined in situ XRD and XANES to observe the transition from cubic ferropericlase to a rhombohedrally distorted phase at 35 GPa and room temperature. Muñoz et al. (2008[link]) and Aquilanti et al. (2009[link]) studied the decomposition of ringwoodite to perovskite and ferropericlase by hyperspectral XANES mapping. Narygina et al. (2009[link]) studied the oxidation state and spin state of ferropericlase and (Mg,Fe)(Si,Al)O3 perovskite up to pressures of 85 GPa and could trace the pressure-induced spin pairing of the 3d electrons of iron. Similarly, Cerantola et al. (2015[link]) traced the spin pairing in siderite in experiments up to 51 GPa (Fig. 4[link]). Andrault et al. (2010[link]) combined XRD and Fe K-edge XANES to determine the iron partitioning between silicate perovskite and post-perovskite for (Fe,Al)-bearing MgSiO3 compositions. XANES spectra were analyzed by fitting linear combinations of spectra of perovskite and post-perovskite, which were determined by Rietveld refinements of the XRD patterns obtained from the same samples. The authors could show that iron fractionates into perovskite and were able to construct the two-phase field of coexistence of the two phases. Finally, Aquilanti et al. (2015[link]) and later Morard et al. (2018[link]) studied the melting of iron by XANES up to >4500 K at >130 GPa and used the change in the spectra to determine the melting temperature at a given pressure.

[Figure 4]

Figure 4

Fe K-edge XANES spectra of siderite (FeCO3) measured as a function of pressure. Between 40 and 44 GPa iron undergoes a high-spin to low-spin transition, which is accompanied by a reduction of the molar volume of siderite by 10%. The transition is well documented in the shift of the first EXAFS maximum (between 7160 and 7180 eV at 0 GPa) to higher energies and an intensity change at the main-edge maximum between 7120 and 7140 eV. In the pre-edge region, the doublet feature should change to a singlet feature (Lin et al., 2010[link]; Westre et al., 1997[link]), but cannot be clearly resolved in this data set. Data were taken from Cerantola et al. (2015[link]).

XAS studies in diamond anvil cells are often limited by the fact that Bragg diffraction peaks from the diamond anvils appear in the spectra that cannot be easily corrected for by normalization procedures. This problem becomes particularly severe at high energies (>10–12 keV). The best way to circumvent this problem is the use of the synthetic nano-polycrystalline diamonds developed by the group of T. Irifune (Ishimatsu et al., 2012[link]). Krstulović et al. (2020[link]) used these diamonds to investigate the local structure in NaAlGe3O8 glass up to 131 GPa in order to gain insights to the structural compression mechanism of noncrystalline matter under extreme pressures, which is of considerable interest for assessing the suggested presence of magmas as the base of the Earth's lower mantle. The quality of the background absorption and the resulting extended X-ray absoption fine structure (EXAFS) spectra is highlighted in Fig. 5[link]. This new anvil material opens up a wide field of applications for XAS using cells equipped with diamond windows. XAS studies on dilute elements in diamond anvil cells, which are performed in fluorescence mode, are often limited by a poor signal-to-background ratio, which stems from inelastic scattering of the incoming beam by the diamond anvils. A confocal fluorescence-detection setup efficiently reduces this background, as described by Wilke et al. (2010[link]), and has successfully been applied to laser-heated diamond anvil cell experiments on chemical reactions between metallic and silicate melts (Petitgirard et al., 2012[link]).

[Figure 5]

Figure 5

Left: X-ray absorption signal of an empty diamond anvil cell (DAC) equipped with nanopolycrystalline diamonds provided by T. Irifune from the Geodynamics Research Center, Ehime University, Matsuyama, Japan and a cell containing a sample of amorphous NaAlGe3O8 at the indicated pressure. Centre: k3-weighted EXAFS spectra of amorphous NaAlGe3O8 as a function of pressure. Right: the evolution of the Ge–O distance with pressure determined from these spectra. The increase in the distance up to ∼26 GPa is associated with a change in coordination from tetrahedral to octahedral. At higher pressures the coordination does not change and only the Ge–O distance decreases. Data were taken from Krstulović et al. (2020[link]).

In resonant inelastic scattering (RIXS) or resonant X-ray emission spectroscopy (RXES), the emitted or inelastically scattered radiation is analyzed using high energy resolution (Glatzel & Bergmann, 2005[link]). This technique represents an extension of XANES spectroscopy, especially for the acquisition and analysis of the pre-edge region. For 3d and 4f transition metals this region provides information on the electronic structure of the valence electrons. Lin et al. (2010[link]) applied this technique to follow the spin pairing of the 3d electrons of iron in ferropericlase. The RIXS spectra show three excitation features in the case of high-spin Fe2+ that gradually reduce to only one feature in the low-spin state (Westre et al., 1997[link]) from 48 to 80 GPa. X-ray Raman scattering (XRS) is another inelastic scattering technique. It allows the measurement of absorption edges in the soft-energy region (<1800 eV) by hard X-rays. Therefore, there is no need for vacuum conditions, as are usually required for soft X-ray spectroscopy, and samples inside a diamond anvil cell may be studied. Using this, Mao et al. (2003[link]) characterized the π and σ bonding by measurements at the C K edge in compressed graphite through a phase transition at 15–17 GPa that is accompanied by changes in the orbital hybridization. With XRS, the Fe M3,2 edge at ∼50 eV and the Fe L3,2 edge at ∼706 eV are measurable, which probe the 3p and 2p electronic states. Both edges provide information on the iron oxidation state and spin state, as documented by the exploratory studies of Nyrow and coworkers (Nyrow et al., 2016[link]; Nyrow, Sternemann et al., 2014[link]; Nyrow, Tse et al., 2014[link]). Weis et al. (2017[link]) determined the iron spin state in siderite and magnesiosiderite up to pressures of 57 GPa using the spectral changes at the the Fe M3,2 edge. Petitgirard et al. (2017[link]) considerably improved the signal-to-noise ratio for high-pressure XRS measurements by inventing a mini-anvil design. They were able to apply XRS to study the local structure of silica glass up to 108 GPa by means of the Si L edge and O K edge with unprecedented data quality (see Fig. 6[link]; Petitgirard et al., 2019[link]), providing important insights into details of the evolution of the local structure with pressure and its link to the pressure evolution of the macroscopic density of this material.

[Figure 6]

Figure 6

XRS spectra obtained by Petitgirard et al. (2019[link]) on amorphous SiO2 at the Si L2,3 edge and O K edge up to 110 GPa from experimental measurements and calculations. (a, b) Si L2,3-edge spectra. (c, d) O K-edge spectra. (e) Atomic structure of SiO2 from molecular-dynamics simulation. This figure was reproduced from Petitgirard et al. (2019[link]).

4. Oxidation state and local structure in magmas

Partial melting of the Earth's mantle and the migration of magma are major processes defining the chemical and structural evolution of the Earth throughout its existence. They are a prerequisite for the formation of the crust and the enrichment of incompatible and volatile elements from the primitive mantle of the early Earth. Structurally, silicate melts are composed of a network of corner-shared SiO4 tetrahedra with larger cations being placed in the voids in the network. The degree of polymerization of the network strongly depends on the chemical composition, i.e. the addition of a network-modifying component such as sodium to a silica melt breaks up the tetrahedral network and transforms bridging oxygens into non-bridging oxygens. In contrast, a coupled addition of sodium with charge-balancing aluminium, which enters the tetrahedral site (network former), will not change the polymerization. These basic structural features strongly influence the way that elements are incorporated into silicate melts and thus has a strong control on phase equilibria and chemical distribution in magmatic systems. XAS has a long tradition in studies on glasses and melts, addressing the valence state and local structure of cations and anions, as documented in the review article by Brown et al. (1995[link]).

Due to the high sensitivity of the XANES to the valence state of a given element, there are a large number of studies on heterovalent elements in glasses and melts. The oxidation state of iron in silicate magmas affects the stability of mineral phases and thus influences the chemical evolution during fractional crystallization processes. Ferric iron may substitute for Al3+ and Si4+ in the tetrahedral sites of the polymeric network of silicate melts. In contrast, Fe2+ is similar in size to Mg2+ and will show higher coordination. As a consequence, bulk melt properties may vary significantly as a function of the oxidation state. Analysis of the pre-edge allows a quantitative determination of the iron oxidation state, provided that a suitable calibration is available. By calibrating the pre-edge centroid position (Fig. 1[link]; Wilke et al., 2004[link], 2005[link]) to the iron oxidation state determined by an independent method (Mössbauer spectroscopy), the centroid may be used to determine the iron oxidation state in natural glasses, particularly in melt inclusions that were trapped and quenched during volcanic processes. Cottrell et al. (2009[link]) performed another calibration and applied it to micro-XANES measurements on natural glasses in order to derive information on the redox conditions during mantle melting in various geotectonic settings (Kelley & Cottrell, 2009[link]). This is certainly a valuable and elegant tool, but one has to keep in mind that the iron oxidation state may easily be altered by secondary processes, such as sub-solidus reheating or loss of volatiles (see, for example, Bucholz et al., 2013[link]; Burkhard, 2001[link]). Giuli and coworkers (Giuli et al., 2002[link], 2005[link]) studied the iron oxidation state of natural glasses formed by impact events of meteorites. The iron oxidation state may vary significantly from virtually ferrous tektites to more oxidized states in other impact glasses, which is likely to reflect differences in the formation process. The XANES at the K edge has also been used in situ at temperature on iron in melts in order to evaluate any changes that occur during the quench to a glass or to evaluate redox reaction kinetics (Cochain et al., 2008[link]; Wilke et al., 2007[link]).

Although sulfur is only present in low amounts in natural silicate melts, as a volatile component it has a major influence on magma degassing, on the atmospheric influence of volcanic eruptions and, finally, on the enrichment of ore-forming metals (Botcharnikov et al., 2011[link]; De Moor et al., 2013[link]; Holasek et al., 1996[link]). The role of sulfur in natural and technical silicate melts has been comprehensively addressed in the review volume by Behrens & Webster (2011[link]) with a chapter on the spectroscopy of sulfur in glasses (Wilke et al., 2011[link]). The solubility of sulfur strongly depends on the redox conditions. In summary, conclusive evidence from S K-edge XANES proved that sulfur in silicate glass is either dissolved as sulfide or sulfate species, with a very sharp transitional range in which both species coexist (see Fig. 7[link]). De Moor et al. (2013[link]) studied the mass-dependent fractionation of sulfur isotopes during magma degassing for two magmatic systems under very different redox conditions. By combining element concentration data, iron and sulfur speciation data by XANES and the isotopic composition of quenched pristine and degassed melt (glass in inclusions and matrix), and the isotopic composition of the gas plume of volcanoes, they derived a comprehensive picture of the processes controlling the degassing. At Erta Ale volcano in Ethiopia, equilibrium fractionation during degassing of the reduced melt best explains the observations. In contrast, kinetic fractionation during diffusive transport in the melt and degassing of the oxidized hydrous magma at Massaya volcano in Nicaragua seems to play the most important role. Fiege et al. (2014[link]) experimentally evaluated the processes controlling sulfur partitioning between the silicate melt and an aqueous fluid phase during decompression and degassing under both oxidizing and reducing conditions, where the sulfur speciation was determined on quenched run products by XANES. These results and those of De Moor et al. (2013[link]) underline the fact that the different behaviour of S2− and S6+ during kinetically controlled degassing needs to be considered in modelling the volatile release of ascending magmas.

[Figure 7]

Figure 7

Left: S K-edge XANES spectra of oxidized and reduced synthetic and natural glasses with different compositions (see Wilke et al., 2011[link]). Right, the sulfur oxidation state determined by XANES in hydrous basaltic glasses versus the logarithm of the oxygen fugacity expressed as the difference from the oxygen fugacity of the quartz–fayalite–magnetite reaction. Data were taken from Jugo et al. (2010[link]).

Further redox-sensitive elements are usually only abundant in trace amounts in natural magmas (<0.1 wt%), but they may still provide important information, particularly on redox conditions. XAS studies have addressed the relationship of the valence state to redox conditions, melt composition, temperature and pressure for vanadium (Sutton et al., 2005[link]), chromium (Berry & O'Neill, 2004[link]), europium (Cicconi et al., 2012[link]), cerium (Burnham & Berry, 2014[link]; Smythe & Brenan, 2015[link]; Smythe et al., 2013[link]), niobum and tantalum (Cartier et al., 2015[link]).

The absorption edges of the major components of naturally relevant silicate melts are mostly located at soft X-ray energies, so that vacuum conditions are necessary for the acquisition of spectra. Due to this fact, sample preparation and the acquisition of spectra are less straightforward. In addition, the measurement of EXAFS spectra is often limited by the overlap of edges, for example for aluminium and silicon, so that EXAFS spectra with the usual k-range of 12 Å−1 cannot be measured. Thus, studies at these edges focused on the XANES region. Si K-edge XANES measured on silicate glasses is not very sensitive to structural changes due to the fact that the first coordination shell varies significantly. Any structural differences in glasses are rather related to differences in the topology of the silicon tetrahedral and network-modifying cations, which only induce subtle changes in the spectra (for a compilation and further references, see Henderson, de Groot et al., 2014[link]). Analogous to silicon, aluminium in glasses has mostly been investigated by XANES. In most cases aluminium is tetrahedrally coordinated, but it may also display fivefold and sixfold coordination. However, the spectral differences are not very strong, so that quantitative assignment to aluminium species is quite difficult and not very sensitive [see Henderson, de Groot et al. (2014[link]) and references therein]. In contrast, XANES at the K edges of the network-modifying cations is much more sensitive to structural differences. Trcera and coworkers (Trcera et al., 2009[link], 2011[link]) investigated magnesium in various synthetic glasses and developed an analytical protocol that combines ab initio calculation of spectra using a plane-wave density-functional formalism with structural models obtained by classical molecular-dynamics simulations to obtain quantitative information on the local structure around magnesium. Analysis of the spectra indicate predominantly fivefold coordination for magnesium in aluminosilicate glasses and variation from fivefold to fourfold coordination in silicate glasses depending on the network-modifying cations that are present. Sodium in silicate and aluminoslicate glasses has been investigated by De Wispelaere et al. (2004[link]) and Farges et al. (2007[link]). These authors report significant sensitivity of the local structure to compositional variations. In particular, they addressed the effect of H2O dissolved in silicate melts on the coordination of sodium, which indirectly reveals the way that H2O is incorporated into silicate glasses. The effect of H2O on the coordination chemistry of iron in a simple quartzo-feldspar melt was investigated by Wilke et al. (2006[link]). XANES and XAS spectra of iron in hydrous quenched melts display considerable differences from their dry counterparts, indicating that the structure around iron in the hydrous glass is different. Using an in situ XANES experiment at pressure and temperature on hydrous haplogranitic melt at the Fe K edge, Wilke et al. (2006[link]) proved that the structure around iron does not change between dry and hydrous melts. This in situ experiment showed that the local structure around iron is difficult to quench in hydrous melts. Often, iron oxide phases quickly precipitate during cooling even if rapid-quench techniques are employed.

Trace elements (<0.1 wt%) in magmas are used to trace melting processes, the chemical evolution of magmatic complexes and the Earth's crust and mantle. They substitute for major elements in minerals and melts. Their chemical distribution between phases may be described by a simple distribution coefficient that is determined by the concentration ratio between the crystal and melt. The value of the distribution coefficient depends on the crystal and melt composition, pressure and temperature. The effect of the melt composition is directly related to the coordination chemistry of the melt, as explained above. XAS is particularly well suited to studying these elements, as it is one of the few methods that allows spectroscopy on dilute samples by measuring the fluorescence yield. In the review article by Brown et al. (1995[link]), a thorough introduction to XAS studies on these elements is provided. Almost all trace elements display ionic radii that are too large to replace silicon or aluminium in the tetrahedral network of melts, so that their incorporation strongly depends on the availability of nonbridging oxygen in the structure. The XAS data show that with increasing polymerization of the silicate melt the coordination changes from more regular symmetry to a distorted symmetry, with an increased distance to the neighbouring atoms. This is particularly evident for zirconium, uranium and thorium (Farges, 1991[link]; Farges et al., 1991[link], 1992[link]), where the coordination changes from dominantly sixfold in less polymerized melts to (partially) eightfold coordination for polymerized melts. Due to its smaller size, titanium displays a slightly different behaviour, with coordination between fourfold and sixfold, including the presence of titanium with a fivefold coordination (tetragonal pyramidal; Farges, 1997[link]; Farges, Brown, Navrotsky et al., 1996[link]). More recently, further elements in silicate glass/melts have been studied (see Henderson, de Groot et al., 2014[link]), to which platinum (0, +2 and +4 valence states), molybdenum (0, +3, +4, +5 and +6), Nb5+ and tin (2+ and 4+, compared with tantalum and tungsten) (Farges et al., 1999[link]; Farges, Siewert, Brown et al., 2006[link]; Farges, Siewert, Ponader et al., 2006[link]; Farges, Linnen et al., 2006[link]; Piilonen et al., 2006[link]) can be added, as well as yttrium (Haigis et al., 2013[link]; Simon et al., 2013[link]). A major feature of many trace elements in glasses and melts is the presence of non-Gaussian pair distributions. This becomes particularly strong for the more polymerized melt compositions and for data acquired at high temperature. As pointed out by Brown et al. (1995[link]), the analysis of EXAFS data has to account for this, otherwise the derived distances are systematically shifted to lower values and the number of neighbouring atoms is underestimated. Simon et al. (2013[link]) fitted the EXAFS using a histogram fit that is based on the Eulerian gamma function and allows asymmetric pair distributions. Fig. 8[link] shows the derived Y–O pair distribution for three glass compositions varying in the degree of polymerization. From CaSiO3 to CaAl2Si2O8 the Y–O distribution in these glasses evolves from a symmetrical Gaussian-like shape to a considerably asymmetric shape and an increased average Y–O distance. The relative changes in the distribution between the compositions are in good agreement with the Y–O pair distributions obtained from molecular-dynamics simulation (Haigis et al., 2013[link]). This approach has been extended to study the effect of CO2 on the local structure of lanthanum, yttrium and strontium in silicate melts and their structural environment in carbonate melts at high pressure and temperature (Pohlenz et al., 2018[link]).

[Figure 8]

Figure 8

Top: Y–O pair distribution for the glass compositions indicated. Glasses were doped with ∼0.6 wt% Y2O3 and synthesized at ∼2000 K by aerodynamic levitation and laser heating. The pair distribution was derived by a histogram fit to the EXAFS data based on the gamma function. Reprinted from Simon et al. (2013[link]) with permission from Elsevier. Bottom, Y–O pair distribution for the indicated melts derived from molecular dynamics at 3000 K. Reprinted from Haigis et al. (2013[link]) with permission from Elsevier..

5. Element complexation in aqueous fluids at pressure and temperature

Aqueous fluids are an important means of chemical transport in the deep Earth. The transport capacity of these fluids for a given element strongly depends on its solubility in the fluid, which in turn is strongly dependent on the complexation mechanism active in the fluid. XAS of fluids in in situ cells at pressure and temperature has provided unique evidence on the incorporation mechanism of elements in aqueous fluids. A comprehensive review of geologically relevant fluids has been published (Stefánsson et al., 2013[link]), including results from XAS studies (Pokrovski et al., 2013[link]; Sanchez-Valle, 2013[link]).

XAS at the O K edge provides insight into the structural properties of water, which have major control over its solvent properties. However, at the energy of the O K edge, in situ XAS measurements at elevated pressures and temperature are not possible. However, X-ray Raman scattering allows the detection of the O K edge with hard X-rays in an indirect way. Sahle et al. (2013[link]) used this method to follow the structural changes of water as a function of temperature and pressure up to supercritical conditions. The analysis of the spectra was based on structural models from ab initio molecular-dynamics simulation and forward modelling of XANES spectra. The results indicate a gradual change from an extended hydrogen-bonded network at ambient conditions to a distorted network with a loss of hydrogen bonding. These results complement those obtained for water under ambient conditions obtained by `regular' XAS at the O K edge (Wernet et al., 2004[link]).

The formation of ore deposits mostly involves the dissolution and transport of elements by aqueous fluids after being extracted from a source, either crystalline or magmatic. Therefore, the element speciation in the solution or vapour is key to understanding the parameters controlling the enrichment processes. The presence of chlorine, fluorine and sulfur in aqueous fluids controls the metal speciation to a large extent. This is much stronger than in magmas, where it was detected only for Mo–S directly by XAS (Farges, Siewert, Brown et al., 2006[link]) and for gold by the correlation of solubility to the sulfur oxidation state (Botcharnikov et al., 2011[link]). The speciation in the fluid at pressure and temperature cannot be quenched, so that in situ studies at pressure and temperature are mandatory. XAS is ideally suited to these experiments for most elements of interest because it provides information on the molecular structure of species as well as the oxidation state. In particular, the oxidation state is important for many ore-forming transition elements, such as copper, iron, molybdenum etc. Pokrovski et al. (2013[link]) have provided a thorough overview of the speciation of metals in fluids. Here, we would like to highlight the experimental data input by XAS that is very closely related to the development of an adequate pressure cell that works well in the most interesting pressure and temperature interval for hydrothermal systems (≤200 MPa and ≤500°C), as invented by Testemale et al. (2004[link], 2005[link]). The great success of XAS with this pressure cell, which is documented by a long list of publications, is the fact that it allows the acquisition of both XAS spectra and concentration even in low-density vapour, so that it provides a direct link between solubility and speciation (see Testemale & Brugger, 2021[link]).

Transport by fluids is also important for other trace elements that are used in geochemistry as genetic proxies to derive insight into geological processes (the rare-earth elements, zirconium, hafnium, niobium, tantalum, rubidium, strontium etc.) because fluid–rock interaction often severely alters the trace-element signature of a given rock. Here, not only are liquids and vapour of hydrothermal systems at low pressure of interest, but also those occurring under conditions of subduction-zone metamorphism up to more than 3 GPa. In order to reach the high pressure needed to study metamorphic fluids, hydrothermal diamond anvil cells as invented by Bassett et al. (1993[link]) and Schmidt & Rickers (2003[link]), which were subsequently developed further [see Wilke et al. (2010[link]) and references therein] are usually employed. Sanchez-Valle (2013[link]) provided a review of XAS studies in these systems. Studies evolved from measuring trace-element speciation in synthetic chloridic, acidic or basic fluids (see, for example, Mayanovic et al., 1999[link]; Ragnarsdottir et al., 1998[link]) to determining the element speciation in fluids in equilibrium with solid phases or silicate melt (Borchert et al., 2014[link]; Wilke et al., 2012[link]). Studying the speciation in fluids equilibrated with a realistic mineral assemblage is of major importance because all chemical components dissolved in the fluid may influence the speciation present in the fluid. In the case of zirconium, similar concentrations in equilibrium with the mineral zircon were reached for dissolution of zircon in an aqueous solution with 18 wt% Na2SiO7 or a solution with 20–24 wt% NaOH, indicating that it is not only anionic species such as Cl and F that play a role in the mobilization of elements. The complexation mechanism of zirconium for those two fluids is very different, as shown by analysis of the XANES spectra. A hydration shell with sevenfold coordination was proposed for the NaOH solution, whereas an alkali zirconosilicate complex with sixfold zirconium coordination was proposed for the Na2SiO7-bearing solution. Zirconium complexation in NaOH and HCl solutions was further investigated by combination with ab initio molecular dynamics and theoretical spectroscopy. For example, the authors were able to show that zirconium speciation in HCl solutions can largely be described by a monomeric oxychloride complex (see Fig. 9[link]; Jahn et al., 2015[link]).

[Figure 9]

Figure 9

Zr K-edge XANES measured on a 16 wt% HCl solution in equilibrium with zircon at a pressure of 1.37 GPa and a temperature of 873 K (black). Coloured lines correspond to theoretical spectra calculated with FEFF based on snapshots of ab initio molecular dynamics. Only three examples of molecular models are shown, which represent the average state for a given molecular-dynamics run of that composition. The molecular model in the box represents the most likely zirconium species present in the fluid, which corresponds to the blue spectrum. Data were taken from Jahn et al. (2015[link]).

References

First citationAndrault, D., Muñoz, M., Bolfan-Casanova, N., Guignot, N., Perrillat, J.-P., Aquilanti, G. & Pascarelli, S. (2010). Earth Planet. Sci. Lett. 293, 90–96.Google Scholar
First citationAquilanti, G., Pascarelli, S., Mathon, O., Muñoz, M., Narygina, O. & Dubrovinsky, L. (2009). J. Synchrotron Rad. 16, 376–379.Google Scholar
First citationAquilanti, G., Trapananti, A., Karandikar, A., Kantor, I., Marini, C., Mathon, O., Pascarelli, S. & Boehler, R. (2015). Proc. Natl Acad. Sci. USA, 112, 12042–12045.Google Scholar
First citationBassett, W., Shen, A., Bucknum, M. & Chou, I. (1993). Rev. Sci. Instrum. 64, 2340–2345.Google Scholar
First citationBehrens, H. & Webster, J. D. (2011). Rev. Mineral. Geochem. 73, 1–8.Google Scholar
First citationBerry, A. & O'Neill, H. (2004). Am. Mineral. 89, 790–798.Google Scholar
First citationBerry, A. J., Yaxley, G. M., Woodland, A. B. & Foran, G. J. (2010). Chem. Geol. 278, 31–37.Google Scholar
First citationBoehler, R. (2005). EMU Notes Mineral. 7, 273–280.Google Scholar
First citationBorchert, M., Wilke, M., Schmidt, C., Kvashnina, K. & Jahn, S. (2014). Geochim. Cosmochim. Acta, 142, 535–552.Google Scholar
First citationBotcharnikov, R. E., Linnen, R. L., Wilke, M., Holtz, F., Jugo, P. J. & Berndt, J. (2011). Nat. Geosci. 4, 112–115.Google Scholar
First citationBourdelle, F., Benzerara, K., Beyssac, O., Cosmidis, J., Neuville, D. R., Brown, G. E. Jr & Paineau, E. (2013). Contrib. Mineral. Petrol. 166, 423–434.Google Scholar
First citationBrown, G. E., Calas, G., Waychunas, G. A. & Petiau, J. (1988). Rev. Mineral. Geochem. 18, 431–512.Google Scholar
First citationBrown, G. E., Farges, F. & Calas, G. (1995). Rev. Mineral. 32, 317–410.Google Scholar
First citationBucholz, C. E., Gaetani, G. A., Behn, M. D. & Shimizu, N. (2013). Earth Planet. Sci. Lett. 374, 145–155.Google Scholar
First citationBurkhard, D. (2001). J. Petrol. 42, 507–527.Google Scholar
First citationBurnham, A. D. & Berry, A. J. (2014). Chem. Geol. 366, 52–60.Google Scholar
First citationCartier, A., Hammouda, T., Boyet, M., Mathon, O., Testemale, D. & Moine, B. N. (2015). Am. Mineral. 100, 2152–2158.Google Scholar
First citationCerantola, V., McCammon, C. A., Kupenko, I., Kantor, I., Marini, C., Wilke, M., Ismailova, L., Solopova, N., Chumakov, A. I., Pascarelli, S. & Dubrovinsky, L. (2015). Am. Mineral. 100, 2670–2681.Google Scholar
First citationCicconi, M. R., Giuli, G., Paris, E., Ertel-Ingrisch, W., Ulmer, P. & Dingwell, D. B. (2012). Am. Mineral. 97, 918–929.Google Scholar
First citationCochain, B., Neuville, D. R., de Ligny, D., Cormier, L., Roux, J., Baudelet, F., Pinet, O. & Richet, P. (2008). Geochim. Cosmochim. Acta, 72, A170.Google Scholar
First citationCottrell, E., Kelley, K. A., Lanzirotti, A. & Fischer, R. A. (2009). Chem. Geol. 268, 167–179.Google Scholar
First citationCuif, J.-P., Dauphin, Y., Doucet, J., Salome, M. & Susini, J. (2003). Geochim. Cosmochim. Acta, 67, 75–83.Google Scholar
First citationDauphin, Y., Cuif, J.-P., Salome, M. & Susini, J. (2005). Am. Mineral. 90, 1748–1758.Google Scholar
First citationDe Andrade, V., Susini, J., Salomé, M., Beraldin, O., Rigault, C., Heymes, T., Lewin, E. & Vidal, O. (2011). Anal. Chem. 83, 4220–4227.Google Scholar
First citationDe Wispelaere, S., Cabaret, D., Levelut, C., Rossano, S., Flank, A., Parent, P. & Farges, F. (2004). Chem. Geol. 213, 63–70.Google Scholar
First citationEgerton, V. M., Wogelius, R. A., Norell, M. A., Edwards, N. P., Sellers, W. I., Bergmann, U., Sokaras, D., Alonso-Mori, R., Ignatyev, K., van Veelen, A., Anné, J., van Dongen, B., Knoll, F. & Manning, P. L. (2015). J. Anal. At. Spectrom. 30, 627–634.Google Scholar
First citationFarges, F. (1991). Geochim. Cosmochim. Acta, 55, 3303–3319.Google Scholar
First citationFarges, F. (1997). Am. Mineral. 82, 36–43.Google Scholar
First citationFarges, F. (2005). Phys. Rev. B, 71, 155109.Google Scholar
First citationFarges, F. (2009). Phys. Chem. Miner. 36, 463–481.Google Scholar
First citationFarges, F., Brown, G. E., Navrotsky, A., Gan, H. & Rehr, J. J. (1996). Geochim. Cosmochim. Acta, 60, 3039–3053.Google Scholar
First citationFarges, F., Brown, G. E. & Rehr, J. J. (1996). Geochim. Cosmochim. Acta, 60, 3023–3038.Google Scholar
First citationFarges, F., Brown, G. E. & Rehr, J. J. (1997). Phys. Rev. B, 56, 1809–1819.Google Scholar
First citationFarges, F. & Calas, G. (1991). Am. Mineral. 76, 60–73.Google Scholar
First citationFarges, F., de Wispelaere, S., Rossano, S., Muños, M., Wilke, M., Flank, A.-M. & Lagarde, P. (2007). AIP Conf. Ser. 882, 214–216.Google Scholar
First citationFarges, F., Flank, A.-M., Lagarde, P. & Ténégal, F. (1999). J. Synchrotron Rad. 6, 193–194.Google Scholar
First citationFarges, F., Linnen, R. L. & Brown, G. E. Jr (2006). Can. Mineral. 44, 795–810.Google Scholar
First citationFarges, F., Ponader, C. W. & Brown, G. E. (1991). Geochim. Cosmochim. Acta, 55, 1563–1574.Google Scholar
First citationFarges, F., Ponader, C. W., Calas, G. & Brown, G. E. (1992). Geochim. Cosmochim. Acta, 56, 4205–4220.Google Scholar
First citationFarges, F., Siewert, R., Brown, G. E. Jr, Guesdon, A. & Morin, G. (2006). Can. Mineral. 44, 731–753.Google Scholar
First citationFarges, F., Siewert, R., Ponader, C. W., Brown, G. E. Jr, Pichavant, M. & Behrens, H. (2006). Can. Mineral. 44, 755–773.Google Scholar
First citationFiege, A., Behrens, H., Holtz, F. & Adams, F. (2014). Geochim. Cosmochim. Acta, 125, 241–264.Google Scholar
First citationGaloisy, L. (2004). EMU Notes Mineral. 6, 553–583.Google Scholar
First citationGiuli, G., Eeckhout, S. G., Paris, E., Koeberl, C. & Pratesi, G. (2005). Meteorit. Planet. Sci. 40, 1575–1580.Google Scholar
First citationGiuli, G., Pratesi, G., Cipriani, C. & Paris, E. (2002). Geochim. Cosmochim. Acta, 66, 4347–4353.Google Scholar
First citationGlatzel, P. & Bergmann, U. (2005). Coord. Chem. Rev. 249, 65–95.Google Scholar
First citationGoettlicher, J., Kotelnikov, A., Suk, N., Kovalski, A., Vitova, T. & Steininger, R. (2013). Z. Kristallogr. 228, 157–171.Google Scholar
First citationHaigis, V., Salanne, M., Simon, S., Wilke, M. & Jahn, S. (2013). Chem. Geol. 346, 14–21.Google Scholar
First citationHenderson, G. S., de Groot, F. M. F. & Moulton, B. J. A. (2014). Rev. Mineral. Geochem. 78, 75–138.Google Scholar
First citationHenderson, G. S., Neuville, D. R. & Downs, R. T. (2014). Rev. Mineral. Geochem. 78, iii–iv.Google Scholar
First citationHolasek, R., Self, S. & Woods, A. (1996). J. Geophys. Res. 101, 27635–27655.Google Scholar
First citationIshimatsu, N., Matsumoto, K., Maruyama, H., Kawamura, N., Mizumaki, M., Sumiya, H. & Irifune, T. (2012). J. Synchrotron Rad. 19, 768–772.Google Scholar
First citationJahn, S., Dubrail, J. & Wilke, M. (2015). Chem. Geol. 418, 30–39.Google Scholar
First citationJugo, P. J., Wilke, M. & Botcharnikov, R. E. (2010). Geochim. Cosmochim. Acta, 74, 5926–5938.Google Scholar
First citationJuhin, A., Calas, G., Cabaret, D., Galoisy, L. & Hazemann, J.-L. (2008). Am. Mineral. 93, 800–805.Google Scholar
First citationJuillot, F., Morin, G., Ildefonse, P., Calas, G. & Brown, G. E. Jr (2006). Am. Mineral. 91, 1432–1441.Google Scholar
First citationKagi, H., Odake, S., Ishibashi, H., Shozugawa, K., Matsuo, M., Satake, W. & Mikouchi, T. (2013). J. Mineral. Petrological Sci. 108, 172–177.Google Scholar
First citationKantor, I., Dubrovinsky, L., McCammon, C., Kantor, A., Pascarelli, S., Aquilanti, G., Crichton, W., Mattesini, M., Ahuja, R., Almeida, J. & Urusov, V. (2006). Phys. Chem. Miner. 33, 35–44.Google Scholar
First citationKelley, K. A. & Cottrell, E. (2009). Science, 325, 605–607.Google Scholar
First citationKrstulović, M., Rosa, A. D., Biedermann, N., Spiekermann, G., Irifune, T., Muñoz, M. & Wilke, M. (2020). Phys. Rev. B, 101, 214103.Google Scholar
First citationLin, J.-F., Mao, Z., Jarrige, I., Xiao, Y., Chow, P., Okuchi, T., Hiraoka, N. & Jacobsen, S. D. (2010). Am. Mineral. 95, 1125–1131.Google Scholar
First citationLuo, Y., Rakovan, J., Tang, Y., Lupulescu, M., Hughes, J. M. & Pan, Y. (2011). Am. Mineral. 96, 23–33.Google Scholar
First citationMao, W., Mao, H., Eng, P., Trainor, T., Newville, M., Kao, C., Heinz, D., Shu, J., Meng, Y. & Hemley, R. (2003). Science, 302, 425–427.Google Scholar
First citationMayanovic, R. A., Anderson, A. J., Bassett, W. A. & Chou, I. (1999). J. Synchrotron Rad. 6, 195–197.Google Scholar
First citationMoor, J. M. de, Fischer, T. P., Sharp, Z. D., King, P. L., Wilke, M., Botcharnikov, R. E., Cottrell, E., Zelenski, M., Marty, B., Klimm, K., Rivard, C., Ayalew, D., Ramirez, C. & Kelley, K. A. (2013). Geochem. Geophys. Geosyst. 14, 4076–4108.Google Scholar
First citationMorard, G., Boccato, S., Rosa, A. D., Anzellini, S., Miozzi, F., Henry, L., Garbarino, G., Mezouar, M., Harmand, M., Guyot, F., Boulard, E., Kantor, I., Irifune, T. & Torchio, R. (2018). Geophys. Res. Lett. 45, 11074–11082.Google Scholar
First citationMottana, A. (2004). EMU Notes Mineral. 6, 465–552.Google Scholar
First citationMuñoz, M., De Andrade, V., Vidal, O., Lewin, E., Pascarelli, S. & Susini, J. (2006). Geochem. Geophys. Geosyst. 7, Q11020.Google Scholar
First citationMuñoz, M., Pascarelli, S., Aquilanti, G., Narygina, O., Kurnosov, A. & Dubrovinsky, L. (2008). High Press. Res. 28, 665–673.Google Scholar
First citationNarygina, O., Mattesini, M., Kantor, I., Pascarelli, S., Wu, X., Aquilanti, G., McCammon, C. & Dubrovinsky, L. (2009). Phys. Rev. B, 79, 174115.Google Scholar
First citationNyrow, A., Sternemann, C., Tse, J. S., Weis, C., Sahle, C. J., Mende, K., Wieland, D. C. F., Cerantola, V., Gordon, R. A., Spiekermann, G., Regier, T., Wilke, M. & Tolan, M. (2016). J. Anal. At. Spectrom. 31, 815–820.Google Scholar
First citationNyrow, A., Sternemann, C., Wilke, M., Gordon, R. A., Mende, K., Yavaş, H., Simonelli, L., Hiraoka, N., Sahle, C. J., Huotari, S., Andreozzi, G. B., Woodland, A. B., Tolan, M. & Tse, J. S. (2014). Contrib. Mineral. Petrol. 167, 1012.Google Scholar
First citationNyrow, A., Tse, J. S., Hiraoka, N., Desgreniers, S., Büning, T., Mende, K., Tolan, M., Wilke, M. & Sternemann, C. (2014). Appl. Phys. Lett. 104, 262408.Google Scholar
First citationPetitgirard, S., Borchert, M., Andrault, D., Appel, K., Mezouar, M. & Liermann, H.-P. (2012). Rev. Sci. Instrum. 83, 013904.Google Scholar
First citationPetitgirard, S., Sahle, C. J., Weis, C., Gilmore, K., Spiekermann, G., Tse, J. S., Wilke, M., Cavallari, C., Cerantola, V. & Sternemann, C. (2019). Geochem. Persp. Lett. 9, 32–37.Google Scholar
First citationPetitgirard, S., Spiekermann, G., Weis, C., Sahle, C., Sternemann, C. & Wilke, M. (2017). J. Synchrotron Rad. 24, 276–282.Google Scholar
First citationPiilonen, P. C., Farges, F., Linnen, R. L., Brown, G. E. Jr, Pawlak, M. & Pratt, A. (2006). Can. Mineral. 44, 775–794.Google Scholar
First citationPohlenz, J., Rosa, A. D., Mathon, O., Pascarelli, S., Belin, S., Landrot, G., Murzin, V., Veligzhanin, A., Shiryaev, A., Irifune, T. & Wilke, M. (2018). Chem. Geol. 486, 1–15.Google Scholar
First citationPokrovski, G. S., Borisova, A. Y. & Bychkov, A. Y. (2013). Rev. Mineral. Geochem. 76, 165–218.Google Scholar
First citationQuartieri, S., Antonioli, G., Geiger, C., Artioli, G. & Lottici, P. (1999). Phys. Chem. Miner. 26, 251–256.Google Scholar
First citationQuartieri, S., Boscherini, F., Dalconi, C., Iezzi, G., Meneghini, C. & Oberti, R. (2008). Am. Mineral. 93, 495–498.Google Scholar
First citationQuartieri, S., Chaboy, J., Antonioli, G. & Geiger, C. (1999). Phys. Chem. Miner. 27, 88–94.Google Scholar
First citationQuartieri, S., Dalconi, M., Boscherini, F., Oberti, R. & D'Acapito, F. (2004). Phys. Chem. Miner. 31, 162–167.Google Scholar
First citationRakovan, J., Newville, M. & Sutton, S. (2001). Am. Mineral. 86, 697–700.Google Scholar
First citationSahle, C. J., Sternemann, C., Schmidt, C., Lehtola, S., Jahn, S., Simonelli, L., Huotari, S., Hakala, M., Pylkkänen, T., Nyrow, A., Mende, K., Tolan, M., Hämäläinen, K. & Wilke, M. (2013). Proc. Natl Acad. Sci. USA, 110, 6301–6306.Google Scholar
First citationSanchez-Valle, C. (2013). Rev. Mineral. Geochem. 76, 265–309.Google Scholar
First citationSchmid, R., Wilke, M., Oberhänsli, R., Janssens, K., Falkenberg, G., Franz, L. & Gaab, A. (2003). Lithos, 70, 381–392.Google Scholar
First citationSchmidt, C. & Rickers, K. (2003). Am. Mineral. 88, 288–292.Google Scholar
First citationShiryaev, A. A., Zubavichus, Y. V., Veligzhanin, A. A. & McCammon, C. (2010). Russ. Geol. Geophys. 51, 1262–1266.Google Scholar
First citationSilversmit, G., Vekemans, B., Appel, K., Schmitz, S., Schoonjans, T., Brenker, F. E., Kaminsky, F. & Vincze, L. (2011). Anal. Chem. 83, 6294–6299.Google Scholar
First citationSilversmit, G., Vekemans, B., Nikitenko, S., Schmitz, S., Schoonjans, T., Brenker, F. E. & Vincze, L. (2010). Phys. Chem. Chem. Phys. 12, 5653–5659.Google Scholar
First citationSimon, S., Wilke, M., Chernikov, R., Klemme, S. & Hennet, L. (2013). Chem. Geol. 346, 3–13.Google Scholar
First citationSimon, S. B., Sutton, S. R. & Grossman, L. (2007). Geochim. Cosmochim. Acta, 71, 3098–3118.Google Scholar
First citationSmythe, D. J. & Brenan, J. M. (2015). Geochim. Cosmochim. Acta, 170, 173–187.Google Scholar
First citationSmythe, D. J., Brenan, J. M., Bennett, N. R., Regier, T. & Henderson, G. S. (2013). J. Non-Cryst. Solids, 378, 258–264.Google Scholar
First citationStefánsson, A., Driesner, T. & Bénézeth, P. (2013). Rev. Mineral. Geochem. 76, 1–4.Google Scholar
First citationSutton, S., Karner, J., Papike, J., Delaney, J., Shearer, C., Newville, M., Eng, P., Rivers, M. & Dyar, M. (2005). Geochim. Cosmochim. Acta, 69, 2333–2348.Google Scholar
First citationTestemale, D., Argoud, R., Geaymond, O. & Hazemann, J. (2005). Rev. Sci. Instrum. 76, 043905.Google Scholar
First citationTestemale, D. & Brugger, J. (2022). Int. Tables Crystallogr. I. In the press.Google Scholar
First citationTestemale, D., Hazemann, J., Pokrovski, G., Joly, Y., Roux, J., Argoud, R. & Geaymond, O. (2004). J. Chem. Phys. 121, 8973–8982.Google Scholar
First citationTrcera, N., Cabaret, D., Rossano, S., Farges, F., Flank, A.-M. & Lagarde, P. (2009). Phys. Chem. Miner. 36, 241–257.Google Scholar
First citationTrcera, N., Rossano, S., Madjer, K. & Cabaret, D. (2011). J. Phys. Condens. Matter, 23, 255401.Google Scholar
First citationVala Ragnarsdottir, K., Oelkers, E., Sherman, D. & Collins, C. (1998). Chem. Geol. 151, 29–39.Google Scholar
First citationWeis, C., Sternemann, C., Cerantola, V., Sahle, C. J., Spiekermann, G., Harder, M., Forov, Y., Kononov, A., Sakrowski, R., Yavaş, H., Tolan, M. & Wilke, M. (2017). Sci. Rep. 7, 16526.Google Scholar
First citationWernet, P., Nordlund, D., Bergmann, U., Cavalleri, M., Odelius, M., Ogasawara, H., Näslund, L. A., Hirsch, T. K., Ojamäe, L., Glatzel, P., Pettersson, L. G. M. & Nilsson, A. (2004). Science, 304, 995–999.Google Scholar
First citationWestre, T. E., Kennepohl, P., DeWitt, J. G., Hedman, B., Hodgson, K. O. & Solomon, E. I. (1997). J. Am. Chem. Soc. 119, 6297–6314.Google Scholar
First citationWilke, M., Appel, K., Vincze, L., Schmidt, C., Borchert, M. & Pascarelli, S. (2010). J. Synchrotron Rad. 17, 669–675.Google Scholar
First citationWilke, M., Farges, F., Partzsch, G. M., Schmidt, C. & Behrens, H. (2007). Am. Mineral. 92, 44–56.Google Scholar
First citationWilke, M., Farges, F., Petit, P. E., Brown, G. E. & Martin, F. (2001). Am. Mineral. 86, 714–730.Google Scholar
First citationWilke, M., Hahn, O., Woodland, A. B. & Rickers, K. (2009). J. Anal. At. Spectrom. 24, 1364–1372.Google Scholar
First citationWilke, M., Klimm, K. & Kohn, S. C. (2011). Rev. Mineral. Geochem. 73, 41–78.Google Scholar
First citationWilke, M., Partzsch, G. M., Bernhardt, R. & Lattard, D. (2004). Chem. Geol. 213, 71–87.Google Scholar
First citationWilke, M., Partzsch, G. M., Bernhardt, R. & Lattard, D. (2005). Chem. Geol. 220, 143–161.Google Scholar
First citationWilke, M., Schmidt, C., Dubrail, J., Appel, K., Borchert, M., Kvashnina, K. & Manning, C. E. (2012). Earth Planet. Sci. Lett. 349–350, 15–25.Google Scholar
First citationWilke, M., Schmidt, C., Farges, F., Malavergne, V., Gautron, L., Simionovici, A., Hahn, M. & Petit, P. E. (2006). Chem. Geol. 229, 144–161.Google Scholar








































to end of page
to top of page