Dissolution of uranophane: An AFM, XPS, SEM and ICP study

Available online at www.sciencedirect.com

Geochimica et Cosmochimica Acta 73 (2009) 2510–2533 www.elsevier.com/locate/gca

Dissolution of uranophane: An AFM, XPS, SEM and ICP study Michael Schindler a,*, Michael Freund b, Frank C. Hawthorne a, Peter C. Burns c, Patricia A. Maurice c a

Department of Geological Sciences, University of Manitoba, Winnipeg, Man., Canada R3T 2N2 b Department of Chemistry, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2 c Department of Civil Engineering and Geological Sciences, University of Notre Dame, Notre Dame, IN 46556-0767, USA Received 11 July 2007; accepted in revised form 22 January 2009; available online 14 February 2009

Abstract Dissolution experiments on single crystals of uranophane and uranophane-b, Ca(H2O)5[(UO2)(SiO3(OH)]2, from the Shinkolobwe mine of the Democratic Republic of Congo, were done in an aqueous HCl solution of pH 3.5 for 3 h, in HCl solutions of pH 2 for 5, 10 and 30 min, and in Pb2+–, Ba–, Sr–, Ca– and Mg–HCl solutions of pH 2 for 30 min. The basal surfaces of the treated uranophane crystals were examined using atomic-force microscopy (AFM), X-ray photoelectron spectroscopy (XPS) and scanning electron microscopy (SEM). Solutions after dissolution experiments on single crystals and synthetic powders were analysed with inductively coupled plasma-optical emission spectroscopy (ICP-OES) and mass spectroscopy (ICP-MS). The morphology of the observed etch pits (measured by AFM) were compared to the morphology, predicted on the basis of the bond-valence deficiency of polyhedron chains along the edges of the basal surface. Etch pits form in HCl solutions of pH 2. Their decrease in depth with the duration of the dissolution experiment is explained with the stepwave dissolution model, which describes the lowering of the surrounding area of an etch pit with continuous waves of steps emanated from the etch pit into the rest of the crystal surface. Hillocks form in an HCl solution of pH 3.5, and the chemical composition of the surface (as indicated by XPS) shows that these hillocks are the result of the precipitation of a uranyl-hydroxyhydrate phase. Well-orientated hillocks form on the surface of uranophane in a SrCl2–HCl solution of pH 2. They are part of an aged silica coating of composition Si2O2(OH)4(H2O)n. An amorphous layer forms on the surface of uranophane in a MgCl2–HCl solution of pH 2, which has a composition and structure similar to silicic acid. Small crystallites of uranyl-hydroxy-hydrate phases form on the surface of uranophane after treatment in Pb(NO3)2–HCl and BaCl2–HCl solutions of pH 2. Dissolution experiments on synthetic uranophane powders show that in the early stage of the experiments, the dissolution rate of uranophane increase in the sequence Pb(NO3)2–HCl < BaCl2–HCl < CaCl2–HCl < HCl < SrCl2–HCl < MgCl2–HCl, indicating that the dissolution of uranophane is more enhanced in solutions containing divalent cations of small ionic radii and high Lewis acidity (Mg, MgCl+). Ó 2009 Elsevier Ltd. All rights reserved.

1. INTRODUCTION Uranophane and uranophane-b, Ca(H2O)5[(UO2) (SiO3OH)]2, are the most common uranyl minerals, and oc-

* Corresponding author. Address: Department of Earth Sciences, Laurentian University, Sudbury, Ont., Canada. E-mail address: [email protected] (M. Schindler).

0016-7037/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.gca.2009.01.031

cur in a broad range of U deposits. They form as products of the oxidation of U mine- and mill-tailings and are important in understanding water–rock interactions in U deposits (Frondel, 1958; Finch and Ewing, 1992; Finch and Murakami, 1999). Uranyl silicates with polymerized sheets of uranyl pentagonal bipyramids and silicate tetrahedra such as uranophane, uranophane-b, boltwoodite, K2(H2O)3[(UO2) (SiO3OH)]2, Na-boltwoodite, (K,Na)[(H2O)3[(UO2)(SiO3 OH)]2, and sklodowskite, Mg(H2O)6 [(UO2)(SiO3OH)]2,

Dissolution of uranophane

are the final phases of the paragenetic sequence of uranyl minerals formed by corrosion of UO2 and spent nuclear fuel in laboratory experiments and in natural U deposits (Wronkiewicz et al., 1992, 1996; Pearcy et al., 1994; Finn et al., 1996; Finch and Murakami, 1999; Finch et al., 1999). The potential occurrence of uranyl silicates and uranylhydroxy-hydrates on the corrosion rind of oxidized spent nuclear fuel demand an understanding of the processes on their mineral–water interface, which includes incorporation or adsorption of radionuclides onto their surface during growth and dissolution processes (Burns et al., 1997, 2004; Chen et al., 1999, 2000; Burns and Li, 2002; Douglas et al., 2005; Schindler et al., 2006a,b, 2007a,b; Klingensmith and Burn, 2007). Many dissolution experiments on rock-forming silicate minerals show that an amorphous silica-rich layer forms on the surface of these minerals under acidic conditions but not under basic conditions (Brantley, 2005 and references therein). One of the goals of this paper is to examine dissolution processes on the basal surface of uranophane under acidic conditions. Previous dissolution experiments on single crystals of uranyl minerals show that their dissolution is strongly affected by the type of cation in solution (Schindler et al., 2006a,b, 2007a,b). These experiments were carried out in the early stage of dissolution (i.e. non steadystate conditions), where cations in aqueous solution interact directly with the surface of the uranyl minerals rather than with a potential alteration layer formed at a later stage of dissolution. Here, we examine the early stage of the dissolution of uranophane in acidic solutions containing the divalent cations Ba, Pb2+, Sr, Ca and Mg. We will compare the observed morphology of etch pits with the morphology predicted by the approach of Schindler et al. (2004a) and we will describe the formation of different types of alteration layers. 1.1. Previous dissolution studies on uranophane Bulk dissolution experiments on synthetic and natural uranophane in the pH-range of 5–10 (Nguyen et al., 1992; Casas et al., 1994; Perez et al., 2000; Prikryl, 2008) showed that their solubility is minimal in the neutral pHrange and increases with decreasing or increasing pH. The experimentally determined solubility constants for synthetic and natural uranophane are log Ks = 11.7 ± 0.6 and log Ks = 7.8 ± 0.8, and are based on the following dissolution reaction (Casas et al., 1994): CaðH2 OÞ5 ½ðUO2 ÞðSiO3 OHÞ2 þ 6Hþ $ Ca2þ þ 2ðUO2 Þ2þ þ 2H4 SiO4 þ 5H2 O

ð1Þ

Casas et al. (1994) found initial congruent dissolution of uranophane (first 2000 h) followed by subsequent precipitation of a less soluble U-bearing phase (which resulted in an excess of Ca and Si in solution). Scanning electron microscopy (SEM) examination identified soddyite, [(UO2)2(SiO4)](H2O)2, and schoepite, [(UO2)8O2(OH)12](H2O)12, on some of the altered uranophane samples.

2511

1.2. Uranophane (a + b): structural unit and crystal morphology The minerals of the uranophane group are based on [(UO2)SiO3(OH)] sheets which contain (U6+u7) pentagonal bipyramids and acidic SiO3(OH) groups (Fig. 1a). The pentagonal bipyramids share edges to form chains that are connected by (Siu4) (u = O, OH) tetrahedra. The (OH)-groups are located at the free apices of the (Siu4) tetrahedra and form hydrogen bonds to interstitial (H2O) groups (Viswanathan and Harneit, 1986; Ginderow, 1988). Uranophane has space-group symmetry P21 (mono˚ and clinic) with a = 15.909, b = 7.002, c = 6.665 A b = 97.27° (Ginderow, 1988). All minerals of the uranophane group form acicular crystals elongated parallel to the chains of polymerized pentagonal uranyl bipyramids and with prominent basal faces parallel to the sheet of polymerized uranyl polyhera (Fig. 1a and b). Steinocher and Nova´ceˇk (1939) and Branche et al. (1951) showed that the basal face of crystals of uranophane-b are bounded by the [0 0 1], [0 2 1] and [0 1 1] edges (with respect to the unit cell of uranophane). The latter two edges occur mainly on growth zones, but almost disappear in the last stages of crystallization (Steinocher and Nova´ceˇk, 1939). 2. PREDICTION OF THE MORPHOLOGY OF ETCH PITS Schindler et al. (2004a) developed a new approach to calculate the stability of edges on basal surfaces of uranyl-sheet minerals. They estimated the relative stability of an edge from (1) the bond-valence deficiency of polyhedron chains parallel to those edges, (2) the arrangement of the interstitial complexes and (3) the shift between the layers. This approach has been used to predict the morphology of basal surfaces of uranyl-sheet minerals (Schindler et al., 2004b) and to explain growth and dissolution features of schoepite, [(UO2)8O2(OH)12](H2O)12 (Schindler and Putnis, 2004), becquerelite, Ca[(UO2)3O2(OH)3]2(H2O)8 (Schindler et al., 2004c, 2006b), and wyartite, Ca[U5+(UO2)2(CO3)O4(OH)] (H2O)7 (Schindler et al., 2004c), curite, Pb3(H2O)2[(UO2)4 O4(OH)3] (Schindler et al., 2006a), billietite, Ba(H2O)8 [(UO2)6O4(OH)6] (Schindler et al., 2007a), fourmarierite, Pb1x(H2O)4[(UO2)6O32x(OH)4+2x] and synthetic Pb2 (H2O)[(UO2)10UO12(OH)6(H2O)2] (Schindler et al., 2007b). Considering the bond-valence deficiency, the shift between the layers and the arrangement of interstitial species, Schindler et al. (2004b) predicted that the probability of the occurrence of edges in the uranophane structure-type decreases in the sequence [0 1 0] > [0 0 1] > [0 1 1] > ([0 k l], k > l) > ([(0 k l], k < l). These predictions are in accordance with the observed morphology of the basal faces of minerals of this group, and explain the occurrence of [0 k l], k > l and [(0 k l] k < l) edges in early stages of growth (observed in growth zones) and the occurrence [0 0 1] and [0 1 1] edges in later growth stages of uranophane-b crystals (Steinocher and Nova´ceˇk, 1939). For etch pits forming on the basal surface of uranophane, one would expect their outline to be defined in their earlier growth stage by the edges [0 k l],

2512

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

Fig. 1. (a) The [(UO2)SiO3(OH)] structural unit in minerals of the uranophane group with the interstitial (Cau7) polyhedra shown; silicate tetrahedra are in black, uranyl polyhedra in dark grey and (Cau7) polyhedra in light grey; (b) SEM image of an untreated freshly cleaved uranophane crystal from the Shinkolobwe mine, Democratic Republic of Congo, and a schematic sketch of the most common morphology of the basal surface; (c) AFM image in height mode of parts of an untreated basal surface of uranophane, and (d) a cross-section taken perpendicular to the elongation of the crystal (dashed white line in c).

k > l [0 k l] k < l and [0 1 1], and in their later growth stages by the edges [0 1 0] and [0 0 1]. 3. EXPERIMENTAL METHODS The specimens of uranophane and uranophane-b were obtained from the William Pinch collection at the Canadian Museum of Nature and are from the Shinkolobwe mine, Democratic Republic of Congo. The chemical composition of a uranophane crystal used for AFM and XPS measurements (first set) was analysed with a Cameca SX-100 microprobe operating in the wavelength-dispersion mode with accelerating voltage of 15 kV, a specimen current of 10 nA and a beam diameter of 5 lm. The following standards were used: diopside (Ca, Si), UO2 (U) and Fayalite (Fe). Raw intensities were converted to the concentrations using the PAP (Pouchou and Pichoir, 1985) matrix correction software. On the basis of five measurements and previous crystal structure determinations, the average chemical composition of the crystal is Ca1.04Fe0.01[(UO2)2(SiO3(OH))]2(H2O)5. 3.1. Dissolution experiments for AFM and XPS examination For each dissolution experiment, we selected one elongated crystal; the size of the basal surface varied between 0.06  0.5 and 0.1  0.7 mm. Each crystal had to be mounted on sample holder with a double-sided carbon tape in order to avoid its floating on the solution surface or its destruction during remounting for AFM and XPS measurements. The samples were placed in Polyethylene cups containing 5 mL of a solution of different pH and

composition. Table 1 lists the electrolyte solutions used, their initial pH value and ionic strength, the pH value after 45 min in the experiment with synthetic uranophane powders and the saturation index with respect to amorphous silica gel; note that the electrolyte solutions of pH 2.0 have equal normality. The standard deviation of the measured pH is around ±0.2, which is based on repeated measurements of the pH in a BaCl2–HCl solution of pH 2 and an ionic strength of 1.51. The pH-meter was calibrated with a KCl–HCl buffer solution of pH 2 and an ionic strength of I = 0.2. The buffer was created by mixing 50 mL of a 0.2 M KCl aqueous solution and 13 mL of a 0.2 M HCl aqueous solution (Robinson and Stokes, 1968). All dissolution experiments were done at room temperature. The experiment in aqueous HCl solution of pH 3.5 lasted 3 h and the experiments in solutions of pH 2 for 10 and 30 min (one batch experiment lasted 1 h, see below). Furthermore, we did an in-situ AFM dissolution experiment with an HCl aqueous solution of pH 2. Here, we replaced the liquid in the fluid cell every 30 s in order to minimize any change in dissolution kinetics through increase in the activity of Ca, U and Si in solution. These experiments were carried out with fibrous and bulky uranophane crystal for 2 and 5 min, respectively, and are designated ‘‘two-minute constant-flow experiment” and ‘‘fiveminute constant-flow experiment” in order to distinguish them from the 10 and 30 min batch experiments. After each experiment, the crystals were quickly dried with a kimwipe, washed in deionized water and mounted for atomic-force microscopy (AFM), scanning electron microscopy (SEM), optical-reflection microscopy and Xray photoelectron spectroscopy (XPS). The time between

Dissolution of uranophane

2513

Table 1 Dissolution experiments on uranophane: concentration of electrolyte solutions (molarity and molality), initial pH, ionic strength, average Lewis acidity of the cationic aqueous species, pH and saturation index (log IAP  log Ks) with respect to amorphous silica gel after 45 min of the dissolution experiment with synthetic powder (after 3 hs for the experiment in HCl, pH 3.5). Electrolyte solution

Molality (mol kg1)

pH value initial

Ionic strength (molality)

hLewis acidityi (vu)

pH value 45 min

Saturation index (am. silica gel)

BaCl2 0.5 mol L1 SrCl2 0.5 mol L1 MgCl2 0.5 mol L1 CaCl2 0.5 mol L1 Pb(NO3)2 0.5 mol L1 HCl HCl

0.55 0.54 0.52

2.0 2.0 2.0

1.51 1.45 1.12

0.19 0.22 0.26

2.1 2.1 2.1

0.12 0.28 0.28

0.53 0.58

2.0 2.0

1.26 0.91

0.24 0.12

2.1 2.2

0.14 0.16

0.01 103.5

2.0 3.5

0.01 0.0003

n.a. n.a.

2.5 5.4 (3 h)

0.23 1.4

the dissolution experiment and AFM examination was a critical issue, because the surface properties of some of the samples altered with time (see below). Most of the AFM examinations were done 12 h after the dissolution experiments, but in some cases, the surfaces were examined 3 days after the dissolution experiment. The saturation indices with respect to mineral phases and amorphous silica and the ionic strength of the electrolyte solutions were calculated using Visual MINTEQ 2.53 (Gustafsson, 2007). Saturation indices with respect to uranophane (log Ks = 7.8), log Ks = 29), schoepite (log Ks = 5.38) and amorphous silica gel (log Ks = 2.71) were calculated on the basis of the solubility constant given by Casas et al. (1994) and Gustafsson (2007), respectively. The ionic strength (I) of the electrolyte solutions with I > 0.90 were calculated using the Brønsted–Guggenheim– Scatchard version of the specific ion interaction theory (SIT) (Grenthe et al., 1997) and the aqueous-species interaction coefficients from the MINTEQ database. 3.2. Dissolution experiments on synthetic uranophane powder samples Powder samples of synthetic uranophane were synthesized according to the procedure of Klingensmith and Burn (2007). The phase identity and purity were verified using diffraction patterns collected with a Scintag theta–theta diffractometer with CuKa radiation. Chemical analysis of the uranophane powders were performed by dissolving 20 mg of powder in 55 g of 6 M HCl and analysing for U, Si and Ca concentrations by inductively coupled plasma-optical emission spectroscopy (ICP-OES) with an analytical uncertainty of 3.5%. The chemical analysis of the uranophane experiments indicated U : Ca and U : Si ratios of h1.95i (±0.02) and h0.98i (±0.01), respectively. Dissolution experiments were done with the same solutions as in the AFM experiments. Here, 50 mg of uranophane were brought in contact with 25 mL of the corresponding solution in Teflon reaction vessels (2.3 mmol L1 uranophane). These vessels were constantly shaken during the dissolution experiment, producing a fine emulsion from where 5.5 mL of liquid and solid were extracted after 10, 30 and 45 min from the solutions of pH 2 and after 1, 2 and 3 h from the HCl solution of

pH 3.5. The extracted emulsions were centrifuged at 20.350g for 5 min and 1 mL of an extracted solution was diluted afterwards with a 5% HNO3 solution. Concentrations were measured with the ICP-OES and calibration for U, Si and Ca was carried out with solutions containing different ratios between the electrolyte solutions (Table 1) and 1000 ppm stock-solutions of U, Si and Ca. Matrix effects strongly affected the analysis of Ca in the CaCl2–HCl and BaCl2–HCl electrolyte solutions and the concentrations of the latter element could not be measured for the corresponding solutions. 3.3. ICP-MS analysis of solutions after dissolution experiments of single crystals We also determined the concentration of U in solutions after dissolution experiments with single crystals. Here, uranophane crystals of similar size as in the AFM experiments were mounted on double-sided carbon tape and were brought in contact with MgCl2–HCl and CaCl2–HCl solutions of pH 2. The crystals had a more fibrous character than the crystals of the AFM experiment and completely dissolved after 15 min. After the dissolution experiments, both solutions were diluted with 2% ultra-pure HNO3 and an internal standard (4.4 ppb 6Li, Rh, Re, Bi and 235U) was added to permit the correction for matrix effects and drift. The analyses were done on a ThermoFisher XSeriesII in Xt+ mode, with a sensitivity of ca. 0.25e6 cps ppm1 U. Calibration for U was done using a 1:2500 dilution of USGS reference material W-2 (U = 505 ppb). 3.4. Atomic-force microscopy and scanning electron microscopy After the dissolution experiments, the (0 0 1) surfaces of uranophane and uranophane-b crystals were examined with a Nanoscope III Atomic Force Microscope (Veeco Digital Instruments). The surfaces were scanned in contact mode and the images were subsequently analysed with the Software package of Digital Instruments. In order to verify that the scanning of the tip in contact mode did not induce erosion of the surface and alteration of the pit shape, or that drift did not substantially alter the pit shape, we always

2514

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

scanned an area at least twice in orthogonal directions. Note that the depths of etch pits given in AFM images and cross sections are the measured penetration depths of the tip into the pit before contacting the other edge. Thus the measured depth of an etch pit of small lateral dimensions should be interpreted as its minimum depth. Scanning Electron Microscopy was done using a 120 Stereoscan instrument from Cambridge Instruments and chemical analysis was done with an EDAX Genesis system 4000. 3.5. X-ray photoelectron spectroscopy The near-surface chemical composition of basal surfaces of untreated uranophane and uranophane-b crystals and of uranophane crystals treated with (a) HCl, pH 2, 30 min, (b) HCl, pH 2, two minutes constant flow, (c) HCl, pH 3.5, 3 h, (d) SrCl2–HCl, pH 2, 30 min, (e) BaCl2–HCl, pH 2, 30 min and 1 h, and (f) CaCl2–HCl, pH 2, 30 min, were characterized with a Kratos Axis Ultra X-ray photoelectron spectrometer (XPS) equipped with a magnetic-confinement charge-compensation system. The advantages of this system for insulators (e.g. uranyl minerals) have been described in detail by Nesbitt et al. (2004). For XPS measurements of untreated samples, single crystals were cleaved in air and immediately transferred to the XPS. Spectra of the U 4f, O 1s, Si 2p, Ca 2p and Ba 3d electrons were collected in high-resolution scans using monochromatic AlKa radiation (1486.6 eV) and the charge-compensation system. Spectra were recorded using 20 sweeps, scan rates per sweep of 200 ms with analyzer pass energies of 80 and 160 eV and with an aperture size of 55 lm. Resolution for the different pass energies and aperture are listed in detail in Schindler et al. (in this issue-a,b) and their effect on the FWHM values (FWHM = full width at half maximum) of the U 4f and O 1s peaks is described below. Shirley background corrections (Shirley, 1972) and Gaussian– Lorentzian peak shapes of 60 ± 10% and 30 ± 10% were used to fit the U 4f and O 1s spectra, respectively. We additionally measured Si 2p and O 1s spectra for synthetic silicic acid (Sigma) and numerous silicates: a-quartz, SiO2, forsterite Mg2SiO4 and spodumene LiAlSi2O6. The latter specimens were obtained from the mineral collection of the Department of Geological Sciences at the University of Manitoba. The electrostatic sample-charging (which was not compensated by the charge neutraliser) was corrected by setting the binding energy of the C 1s electrons of adventitious carbon on the sample surface equal to 285 eV (Wagner et al., 1979, Handbook of X-ray photoelectron spectroscopy). We use the word band to indicate a specific fitted component of the envelope of the O 1s and U 4f peak, and we use the word species to indicate U-atoms of different valence, structurally distinct O2, or U-atoms or a small cluster of atoms involving O2. 3.6. Determination of the chemical composition of surfaces After Shirley background corrections, the area of the U 4f7/2, Si 2p, Ca 2p and O 1s peaks were used to calculate the

mole proportions of U, Si, Ca and O on the surface of the uranophane samples. We tried different sets of relative sensitivity factors (RSF) for the calculation of the mole proportion of U, Ca, Si and O. The best results were obtained using the sensitivity factors 1.83 for Ca 2p, 0.78 for O 1s (Vision 2.2.6, 2006) and 8.476 for U 4f7/2 (Wagner et al., 1979). These sensitivity factors gave a relatively good agreement between the ratios of U and Ca on the untreated surfaces and in the bulk structures of uranophane and uranophane-b. Use of the sensitivity factor for Si of 0.325 (Vision 2.2.6, 2006) resulted in lower U : Si and larger Si : Ca ratios with respect to the bulk structure. The Si 2p peak is sandwiched between the U 5d5/2 and U 5d3/2 peaks at lower and higher binding-energies, respectively. The larger proportion of Si on untreated surfaces suggests that the calculated areas of the Si 2p peaks may contain contributions from the U 5d peaks. We were not able to model these contributions because the background features of the Si 2p and U 5d peaks are very different: the background of the Si 2p peak does not show a steep rise at the higher binding energy side which is observed for U 5d and many spectra involving electrons of the d- and f-shells. The contributions of the U 5d spectra to the Si 2p peak may also vary with the mole proportion of U and Si on the surface, and therefore we give higher standard deviations for the U : Si and Si : Ca ratios than for U : Ca (Table 2). The Si 2s spectrum has the advantage that it does not overlap with any other peak in the uranophane spectrum. However, reported chemical shifts for the Si 2s peak due to the degree of polymerization of silicate polyhedra are not common in the literature. Furthermore, calculations of the mole proportion of untreated uranophane surfaces using the Si 2s peak instead of the Si 2p peak resulted again in smaller U : Si and larger Si : Ca ratios for untreated crystals of uranophane and uranophane-b, indicating that the sensitivity factor for Si (from different data sets) may not be appropriate for uranophane. Hence, we decided to derive the sensitivity factor for the Si 2p spectrum by measuring the U : Si ratio of four different untreated surfaces of uranophane and uranophane-b. The resulting sensitivity factor of 0.57 was used for the calculation of mole proportions of all samples. This factor allowed us to examine in more detail the change in O : Si and O : U ratios between different uranophane samples. Table 2 lists the U : Si, U : Ca and Si : Ca ratios of the untreated surface of uranophane. Furthermore, we calculated the ratio between O and U or O and Si for surfaces that are either enriched in U or Si, respectively. The presence and proportion of U6+and U4+ were determined by peak fitting the U 4f7/2 spectra and examination of the satellite peaks of the U 4f5/2 peak. The binding energies and the proportion of the bands in the U 4f7/2 envelope are listed in Table 3. The FWHM values of the bands vary in the range 1.00–1.40 eV, but were constrained to be equal in each individual spectrum (Table 3). More details on fitting of the U 4f7/2 spectrum and standard deviations of the binding energies and proportion of the U bands can be found in Schindler et al. (in this issue-a). The O 1s spectra were fitted in order to resolve the signals from the different O bands. Schindler et al. (in this is-

Dissolution of uranophane

2515

Table 2 XPS analyses (otherwise stated) of treated basal surfaces of uranophane crystals. The composition of the surface is given in molar ratios. The cation M is either Si or U, depending on whether the corresponding surface is enriched in Si or U. Sample

U:Si molar ratio ±0.2

U:Ca molar ratio ±0.1

Si:Ca molar ratio ±0.2

O:M M = Si or U

Bulk composition

1

2

2

8.5

Untreated uranophane

1

2

2

10

Untreated uranophane-b Untreated uranophane-b Uranophane 2 mins, constant flow HCl, pH 3.5, 3 h HCl, pH 2, 30 min Ba–HCl, pH 2, 30 min Ba–HCl, pH 2, 60 min Sr–HCl, pH 2, 30 min

1 1.1 1 1.5 0.5 1.4 1.7 Si with a small fraction of U (U:Si = 2:98) 0.8 Only Si Small crystallites containing Pb (EDS)

1.9 2 1.8 3.4 1.8 1.7 2.8

1.8 1.8 1.8 2.4 3.8 1.2 1.7

11 12 11 12 17 18 9 3

2 5.2

2.5

12

Ca–HCl, pH 2, 30 min Mg–HCl, pH 2, 30 min Pb–HCl, pH 2, 30 min

Table 3 Pass energies, apertures, binding energies (eV), full width at half maximum (FWHM) and proportion (%) of bands in the U 4f7/2 peak in uranophane. Sample

Pass energy[eV]/ aperture

Uranophane (bulky crystals) Untreated 80/55 HCl, pH 3.5, 3 h

80/55

pH 2, 30 min

80/55

Ba–HCl, pH 2, 30 min

80/55

Sr–HCl, pH 2, 30 min

80/55

Uranophane (fibrous crystals) Untreated 160/55 HCl, 2 min, constant flow

160/55

Uranophane-b (fibrous crystals) Untreated 160/55 Untreated

160/55

HCl, pH 3.5, 3 h

160/55

HCl, pH 3.5, 3 h

160/55

Ba–HCl, pH 2, 60 min

160/55

Ca–HCl, pH 2, 60 min

160/55

U6+ ±0.1

% ±1

U4+ ±0.1

% ±1

382.1 1.2 382.4 1.2 382.2 1.15 382.2 1.2 382.2 1.1

80

380.6 1.2 380.7 1.2 380.7 1.15 380.6 1.2 380.6 1.1

20

71 76 78 81

382.2 1.4 382.3 1.4

92

382.2 1.3 382.1 1.2 382.6 1.3 382.4 1.3 382.2 1.2 382.2 1.2

92

93

91 92 91 92 90

29 24 22 19

380.6 1.4 380.5 1.4

8

380.5 1.3 380.4 1.2 380.6 1.3 380.7 1.3 380.3 1.2 380.5 1.2

8

7

9 8 9 8 10

Additional elements

Contains Ba (4%) Contains Ba (6%)

sue-b) showed that one can distinguish the following Obands in the O 1s spectra of uranyl-oxy-salt minerals, including uranophane and uranophane-b (Fig. 2 and Table 4): (1) 529.0–530.0 eV: O2 in the equatorial plane of the uranyl polyhedra and bonded exclusively to U (indicated as UAOAU) (2) 530.5–531.4 eV: O2 of the uranyl group (which are involved in bonds of higher p-bonding character), indicated as O@U@O; (3) 531.2–531.4 eV: O2 in the equatorial plane (which are part of the silicate tetrahedra, indicated as TAO); (4) 531.8–533.0 eV: OH groups in the equatorial plane (indicated as OH); (5) 533.0–534.0 eV: H2O groups in the interstitial complex (indicated as H2Ointerst); (6) 534.0 eV: H2O groups physisorbed on the basal surface (indicated as H2Oadsorb). Schindler et al. (in this issue-a) showed that the O@U@O band in uranyl-hydroxy-hydrate minerals occur at 530.6– 531.4 eV, which would overlap with the TAO band for uranophane. In this way, proportions of TAO bands in spectra for treated uranophane surfaces may contain contributions from O@U@O species of uranyl-hydroxy-hydrate phases that occur on the surface of uranophane (see below). Schindler et al., in this issue-a,b discussed the resolution of the Ag 3d5/2, U 4f and O 1s line spectra for silver (standard) and uranyl minerals, respectively. They showed that the differences in FWHM values for different pass energies and apertures are smaller for U 4f and O 1s than for Ag 3d5/2. These small differences for different pass energies can be recognized in the U 4f spectra of untreated and treated uranophane samples measured with pass energies of 80 and 160 eV (Fig. 3a and c). Here, the peak fits of the corre-

2516

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

Fig. 2. The O 1s spectra for untreated and treated surfaces of uranophane; the location of the bands O@U@O, TAO, OH and H2Ointerst are indicated with vertical grey-shaded bars.

sponding spectra resulted in average FWHM values of h1.17i and h1.27i eV for pass energies of 80 and 160 eV (Table 3), respectively. We considered these variations in the FWHM values in the fitting procedures of the O 1s spectra and used two different types of constraints (Table 4):

(1) For spectra measured with a pass-energy of 160 eV, the UAOAU, O@U@O, TAO, OH and H2Ointerst were fitted with the FWHM values 1.2, 1.2, 1.2, 1.4 and 1.6 eV, respectively. In some cases, it was necessary to fit the spectrum with two H2O bands (H2Ointerst

Dissolution of uranophane

2517

Table 4 Binding energies (eV), full width at half maximum (FWHM) and proportion (%) of bands in O 1s spectra of selected uranophane and uranophane-b samples. Fitted proportions of the bands OH and H2Ointerst in the spectra are the sum of the proportions OH and C@O and H2Ointerst and CAOAC, respectively; the proportions of C@O and CAOAC are contributions from O-bearing components of the sample holder and adventitious species to the O 1s peak. Sample

UAOAU ±0.1

Bulk composition

% ±2 0

Analyzer resolution/aperture: 80/55 Untreated 529.8 2 uranophane pH 2, 1.1 30 min Sr–HCl, pH 2, 30 min

–

Pass energy/aperture: 160/55 Untreated 529.7 uranophane-b 1.2 Uranophane 2 min, constant flow HCl, pH 3.5, 3 h 529.5 1.2

1

1

% ±2

TAO/ O@U@O ±0.1

24 530.6 1.1 530.6 1.1 530.7 1.1 530.6 1.1

Ba–HCl, pH 2, 30 min

Ba–HCl, pH 2, 60 min Ca–HCl, pH 2, 30 min Mg–HCl, pH 2, 30 min

O@U@O ±0.1

530.6 1.2 530.7 1.2 530.9 1.2 530.7 1.2 530.7 1.2

27

% ±2 35

531.2 1.1 531.5 1.1 SiAOH/ SiAOASi

26

15

531.3 1.1

15

23

531.2 1.2 531.3 1.2 531.4 1.2 531.3 1.2 531.2 1.2

21

7 2

21 14 33 18

OH ±0.1

and H2Oadsorb) between 533 and 535.1 eV (see below). In these cases, the FWHM values of the H2O bands were constrained to be 1.4 eV. (2) For spectra measured with a pass-energy of 80 eV, the FWHM values of bands were constrained to be 0.1 eV less than the corresponding values for the spectra measured with a pass-energy of 160 eV; e.g. the UAOAU, O@U@O, TAO, OH and H2Ointerst were fit with the FWHM values 1.1, 1.1, 1.1, 1.3 and 1.5(1) eV, respectively. The binding energies of the O@U@O and TAO bands were constrained to be equal at 530.65(5) and 531.25(5) eV, respectively. The initial binding energies of the bands UAOAU, OH, and H2Ointerst were set to be 529.6, 532.4 and 533.4 eV, but were not constrained in the final leastsquare cycles. Schindler et al. (in this issue-b) showed that contributions from O-bearing species of the sample holder and adventitious species to the O 1s spectrum are on average 11% for the spectra of untreated uranophane and uranophane-b. These O-bearing species are C@O and CAOAC and they overlap with the OH and H2Ointerst bands in the O 1s spectrum of a uranyl mineral; i.e. their contribution does not affect the proportions of the bands UAOAU,

11

24 21 26 16

% ±2

H2Ointerst ±0.1

12 532.3 1.3 532.4 1.3 532.4 1.2 532.4 1.2 532.3 1.4 532.1 1.4 532.3 1.4 532.4 1.4 532.4 1.4

24 42 83

32

33 33 32 21 39 Si–OH2

% ±2

H2Oadsorb ±0.1

% ±2

C@O/ CAOAC

29 533.1 1.5 533.7 1.3 533.7 1.2 533.6 1.3 533.4 1.6 533.2 1.6 533.5 1.6 533.3 1.6 533.5 1.6 533.2 1.4

10

6/5

28

6/6

7

4/4

22

534.6 1.3

2

6/7

13

2/7

18

0/4

20

7/6

6

6/7

16

6/5

82

534.0 1.4 536.1 1.4

12

0/5

1

O@U@O and TAO. The contributions of C@O and CAOAC to OH and H2Ointerst can be calculated from their proportions in the C 1s peak and from the C : O ratio on the surface (for details see Schindler et al., in this issue-b). Subtraction of the calculated contributions of the CAO bands from the fitted proportions of OH and H2Ointerst gives the actual proportions of the latter species on the surface of the mineral. Table 4 lists the contributions of C@O and CAOAC and the corrected proportions of OH and H2Ointerst. We also considered the contributions of C@O and CAOAC to the O 1s spectrum when calculating the relative proportions of O versus Si and U. The corrected values for the ratios of O : Si and O : U are given in Table 2. 3.7. Error estimation for the relative peak areas of bands in the O 1s spectrum Fitting O 1s spectra can give a number of solutions in which the relative peak areas of the different O bands differ considerably. The amount of variation between the different fits depends on the number of bands, their degree of overlap, their FWHM values and on the constraints of their binding energies. In order to test the variation in the relative peak areas of different O bands in the O 1s spectrum

2518

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

Fig. 3. The U 4f spectra for untreated and treated surfaces of uranophane; the location of the bands U6+ and U4+ are indicated with vertical grey-shaded bars.

of untreated uranophane (Fig. 2a), we fitted the corresponding spectrum with no constraints on the FWHM values or binding energies. The relative peak areas of the UAOAU, O@U@O, TAO, OH, H2Ointerst bands varied be-

tween 0% and 4%, 20% and 40%, 14% and 34%, 26% and 36% and 7% and 20%, indicating large variations in the relative peak areas of all bands. Ten fits for the O 1s spectrum with the above listed constraints on FWHM values and

Dissolution of uranophane

binding energies gave the following variations in relative areas: 0–3%, 26–31%, 23–28%, 28–32% and 14–16% for the UAOAU, O@U@O, TAO, OH, H2Ointerst bands, respectively. These smaller ranges show that fitting O 1s spectra can produce reproducible results, if binding energies and FWHM values for different O-bands are defined prior to the fitting and constrained accordingly. The average standard deviation for a relative peak area, derived from the latter set of fits is ±2%, which is considered as the standard deviation for all relative peak areas of O bands listed in Table 4. 4. RESULTS 4.1. Relief of an untreated surface Uranophane crystals have good cleavage parallel to (1 0 0) and (0 0 1). Cleaving uranophane crystals parallel to (1 0 0) results in two fragments with flat basal surfaces (Fig. 1b). AFM images in height mode (Fig. 1c) show ripples parallel to the elongation of the crystals. These ripples have a lateral width of 1 lm and a height of 20 nm (Fig. 1d), and may have formed while cleaving the crystal. 4.2. XPS spectra of an untreated surface The uranophane-b crystals used in this work have a more fibrous character than the crystals of uranophane, resulting in a rougher surface after cleaving parallel to the basal surface. In general, a rougher surface has a higher number of underbonded equatorial O-atoms, which protonate in air, forming terminal (OH) or (H2O) groups. Tables 2 and 4 show that this is indeed the case: the surfaces of uranophane-b have higher mole proportions of O and higher and lower proportions of OH and TAO species, respectively.

2519

Assuming identical sensitivity factors for all species of O, the ratio between the apical O-atoms and the equatorial O-atoms (O@U@O) : (TAO + OH + UAOAU) on a perfectly flat basal surface should be identical to the ratio in the bulk structure: 24 / (35 + 12 + 0) = 0.51. We know from AFM (Fig. 1c) and optical images that this is not the case for the basal surfaces of uranophane and uranophane-b. The untreated surfaces contain ripples which are defined by edges with underbonded equatorial O-atoms. Hence, one expects lower ratios of (O@U@O) : (TAO + OH + UAOAU + MAOH2 (M = Si or U)) for the untreated surfaces of both minerals. This is indeed the case: the ratios for uranophane and uranophane-b are 27 / (2 + 26 + 22 + 8) = 0.44 and 23 / (1 + 21 + 31 + 11) = 0.34; lower values but still reasonably close to the ratios in the bulk structures. Table 4 shows that the mole proportions of H2Ointerst are significantly lower on the untreated surfaces than in the bulk structure, which Schindler et al. (in this issue-b) explained by invoking dehydration of the uranophane crystals during exposure to the ultra-high-vacuum in the spectrometer. The untreated surfaces of uranophane contain significant amounts of U4+ (Fig. 3c and Table 3). Schindler et al. (in this issue-a) showed that the occurrence of U4+ is not a result of exposure to ultra-high vacuum, or to the X-ray beam. They discussed the potential occurrence of nano-size crystals of U4+ phases in uranyl minerals and possible charge-balance mechanisms for a surface structure containing U4+. 4.3. Surface features formed in HCl solutions of pH 2 Fig. 4a shows typical etch pits on the basal (1 0 0) surface of uranophane after treatment with an HCl solution of pH

Fig. 4. AFM images in height mode of the basal surface of uranophane treated with an HCl solution of pH 2 for 10 min; (a) deep and shallow etch pits appear in black and dark grey, respectively; (b)–(e) different observed morphologies of etch pits and a schematic sketch of the orientation of the pits and the edges defining their outline.

2520

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

2 for 10 min. Two different types of etch pits can be recognized with similar lateral dimensions but different morphology: (1) deep etch pits with irregular to rounded outlines (black-coloured bottom), and (2) shallow etch pits with rectilinear outlines (grey-coloured bottom). The deeper etch pits vary from triangular-to-rectilinear (Fig. 4b–e), are bounded by the edges [0 0 1], [0 1 0], [0 2 1] and [0 1 2], whereas the rectilinear etch pits are elongated parallel to [0 1 0]. Fig. 5a and b shows typical etch pits on the basal (1 0 0) surface of uranophane after treatment with an HCl solution of pH 2 for 30 min. All etch pits are elongated parallel to [0 1 0] and their outlines often display plane-group symmetry m. The etch pits are often bounded by edges [0 1 0] and [0 0 1], and less frequently by [0 1 1] and [0 1 2] (Fig. 5b). Almost all etch pits that formed after 10 and 30 min retain the shape of their outline at greater depths (Figs. 4b–e, and 5b). Furthermore, small particles occur in and around the etch pits after 30 min of treatment. Fig. 6a (left) shows a typical AFM image of the basal (1 0 0) surface of uranophane after the ‘‘five-minute constant-flow experiment” with an HCl solution of pH 2. The surface appears very rough and is characterized by irregular grooves and elongated etch pits parallel to [0 1 0]. 4.4. Cross sections of etch pits The cross sections of the etch pits were taken parallel to the [0 0 1] edge, which is perpendicular to the elongation of both the etch pits and the uranophane crystals. The etch pit, formed in the five-minute constant-flow experiment, has nearly symmetrical sides with convex or planar surfaces, which reach a pronounced maximum depth at the centre

of the pit (Fig. 6a). After the 10-min batch experiment, the sides of the triangular etch pits are very asymmetric. There is a steep edge at one side of the pit and the maximum depth occurs close to that edge (Fig. 6b). After the 30-min batch experiment, the pit is shallower than those that formed after 10 min where the bottom of the pit is fairly flat and almost extends across the complete width of the pit (Fig. 6c). Lastly, the sides of the etch pits are nearly symmetric with steep edges. 4.5. Composition of the surface after treatment with a HCl solution of pH 2 The surface of uranophane, after treatment with an HCl solution of pH 2 for 30 min is enriched in Si, suggesting that the particles on the surface are part of a Si-rich phase. The rougher surface and the presence of particles, resulted in a higher mole proportion of O and larger proportions of OH and H2Ointerst (Table 4). Fig. 3c–e shows U 4f spectra of the untreated surface and the surface treated with an HCl solution of pH 2. It is apparent that the latter spectrum contains a U4+ band with a peak area similar to that in the spectrum of the untreated uranophane crystal, indicating that no significant oxidation of U4+ or reduction of U6+ occurred during treatment with a HCl solution of pH 2. As described above, constant-flow experiments with aqueous HCl solutions of pH 2 were done with fibrous and platy crystals for 2 and 5 min, respectively. The fibrous uranophane crystals almost completely dissolved during the two-minute constant-flow experiment, whereas the platy uranophane crystals had similar basal dimensions before and after the five-minute constant-flow experiment. The

Fig. 5. (a) AFM image in height mode of parts of the basal surface treated with an HCl solution of pH 2 for 30 min; (b) AFM images in height mode, showing typical morphologies of etch pits formed after 30 min, and a sketch showing the orientation and outline of the etch pits; (c) the average lateral area as a function of the average depth of etch pits formed in HCl solutions of pH 2.

Dissolution of uranophane

2521

Fig. 6. AFM images in height mode (left) and cross-section (right) of etch pits formed in an HCl solution of pH 2 after (a) the five-minute constant-flow experiment, (b) after the 10-min batch experiment, and (c) after the 30-min batch experiment; the orientations of the cross sections are indicated by solid lines.

surface chemical composition was only determined for the crystal of the two-minute constant-flow experiment (see above). Inspection of Tables 2 and 4 show that the surface of the latter crystal has similar ratios of U : Si, U : Ca and Si : Ca as the untreated surface (Table 2), but due to its higher roughness, a higher mole proportion of O and larger proportions of OH and H2Ointerst occurs on its surface (Fig. 2b). 4.6. Surface features formed in an HCl solution of pH 3.5 Fig. 7a shows typical hillocks on the basal (0 0 1) surface of uranophane after dissolution in an aqueous HCl solution of pH 3.5. The rounded hillocks have similar lateral dimensions of 0.5  0.5 lm and individual hillocks have a maximum height of 70 nm. The basal surface contains a small number of etch pits that are completely embedded in a matrix of hillocks. The outlines of these pits are so poorly defined that their edges could not be indexed. The treated surface after dissolution in an HCl solution of pH 3.5 for 3 h is enriched in U, indicating that the hillocks are part of a U-rich phase. The higher relief of the surface results in a higher mole proportion of O and higher proportions of OH and H2Ointerst. The U 4f spectra of uranophane and uranophane-b treated with a HCl solution of pH 3.5 are shown in Fig. 3b–d. The spectra show similar U6+ : U4+ ratios as the corresponding spectra of the untreated surfaces (Fig. 3a–c), indicating that no significant oxidation or reduction occurred on the surface of the crystals during the experiment.

4.7. Surface features formed in CaCl2–HCl solution of pH 2 Fig. 7b is an AFM image of the basal surface after treatment in a CaCl2–HCl solution of pH 2 for 30 min and shows a surface coating with an average thickness of approximately 2 nm. The coating consists of small particles of maximum size 255  200  20 nm. Twelve hours after the dissolution experiment, these particles attached to the AFM tip during scanning, indicated the soft adhesive nature of the coating. The coating itself shows a variable thickness on the surface. Fig. 9b shows an SEM image of a crystal with a thick coating on the left side and a surface that does not seem to be covered by a coating on the right side. However, the AFM image shown in Fig. 7b was taken from the right side, suggesting that the whole crystal is covered at least by a thin coating. The treated surface is enriched in Si which is consistent with the coating being a Si-rich phase. Similar to the surfaces treated with HCl solutions of pH 2 and 3.5, the treated surface contains higher proportions of OH and H2Ointerst than the untreated surface (Table 3). 4.8. Surface features formed in BaCl2–HCl solution of pH 2 Fig. 8a shows typical etch pits and growth features on the basal (0 0 1) surface of uranophane after dissolution in an aqueous BaCl2–HCl solution of pH 2 for 30 min. The etch pits are lens shaped and elongated parallel to [0 0 1] (Fig. 8b) (perpendicular to the elongation of the uranophane crystals). They are shallow with minimum depths

2522

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

Fig. 7. (a) AFM image in deflection mode of hillocks formed in an HCl solution of pH 3.5 after 3 h; (b) AFM image in height mode of a coating formed after 30 min in a Ca–HCl solution of pH 2; (c) AFM image in deflection mode of a coating formed after 30 min in a MgCl2– HCl solution of pH 2; AFM images in deflection mode (d) and height mode (e), showing the occurrence of well-orientated hillocks formed after 30 min in a SrCl2–HCl solution of pH 2.

Fig. 8. (a) AFM images in height (left) and deflection (right) mode, showing etch pits and crystallites formed after 30 min in a BaCl2–HCl solution of pH 2; (b) sketch showing the morphology and orientation of etch pits and steps in (a); (c) AFM image in deflection mode of small crystallites formed after 30 min in a BaCl2–HCl solution of pH 2; (d) AFM image in deflection mode showing small crystallites embedded in a soft coating formed after 1 h in a BaCl2–HCl solution of pH 2; (e) AFM image in deflection mode of small crystallites formed after 30 min in a PbCl2–HCl solution of pH 2.

of 3.2 nm, corresponding to two unit cells (or four layers along [1 0 0]). Further distinctive features of the basal surface are the occurrence of curved steps, a fine surface coat-

ing and small crystallites. The curvature of the steps is the result of a gradual change in the orientation of the steps from [0 1 0] to [0 0 1] (Fig. 8a and b). The fine surface coating

Dissolution of uranophane

2523

Fig. 9. SEM images of the basal surfaces after treatment for 30 min in (a) an MgCl2–HCl solution of pH 2, (b) a CaCl2–HCl solution of pH 2, (c) a SrCl2–HCl solution of pH 2 (back-scattering image) and (d) a BaCl2–HCl solution of pH 2.

has a similar appearance and thickness (1–2 nm) as the coating observed on the basal surface treated with a CaCl2–HCl solution (see above). However, the latter coating covers large parts of the basal surface (Fig. 7b), whereas the coating formed in a BaCl2–HCl solution occurs only at a few places on the basal surface (Fig. 8a). The surface was scanned 3 days after the dissolution experiment, and the particles of the fine coating did not adhere to the AFM tip. Small crystallites formed on the basal (0 0 1) surface of uranophane after treatment in an aqueous Ba–HCl solution of pH 2 (Fig. 8c). These crystallites contain Ba, are of various sizes and have well-defined faces and edges. In order to test if the amount of coating and the small crystallites increases with the duration of the dissolution experiment, we did an experiment in a BaCl2–HCl solution for 1 h. Fig. 8d shows the occurrence of small crystallites embedded in a rough layer. The surface was scanned 12 h after the dissolution experiment, and particles of this layer adhered to the AFM tip (as in the CaCl2–HCl experiment). The basal surfaces after the 30 and 60 min experiments appeared very similar in SEM images. Fig. 9d shows an SEM image taken of the surface treated with a Ba–HCl solution for 30 min. The surface has a low relief and does not contain any significant large dissolution or growth features. The XPS spectra show that the treated surfaces are enriched in U whereas the mole proportions of U increased from 30 to 60 min of treatment. Calculation of the surface composition, including the Ba 3d peak at 770.4 eV, shows that the proportion of Ba increased from 4% to 6% during the experiment. Fig. 2a, c, and d shows the O 1s spectra of the untreated surface and the treated surfaces with BaCl2–HCl solutions of pH 2. Arrows in Fig. 2c indicate the occurrence of smaller shoulders around the prominent maxima and shoulders,

suggesting the presence of H2Ointerst, OH, TAO, and O@U@O species on the surfaces after treatment with a BaCl2–HCl solution for 30 and 60 min. The O 1s spectra of the untreated and the treated surfaces indicate an increase in the proportions of OH, H2Ointerst and H2Oadsorb after treatment for 30 min, but then a decrease in these species after the 60 min treatment. The proportion of these species on the surface treated for 60 min are actually similar to those of the untreated surface (Fig. 2 and Table 4). 4.9. Surface features formed in Pb(NO3)2–HCl solution of pH 2 Small crystallites formed on the basal (0 0 1) surface of uranophane after treatment in an aqueous Pb(NO3)2–HCl solution of pH 2 (Fig. 8e). The small crystallites contain Pb and display features (shape, edges and variations in size) similar to those of the crystallites formed in aqueous BaCl2– HCl solution of pH 2 (Fig. 8b). There were no coating or etch pits visible in the AFM and SEM images. The absence of etch pits provides no information on the dissolution mechanism of uranophane in the presence of a solution containing Pb2+ cations. SEM images (not shown) of the treated uranophane crystal show a basal surface with low relief and no large dissolution or growth features. The relief on the basal surface is nearly identical to the relief of the basal surface treated with a BaCl2–HCl solution of pH 2 (Fig. 9d). 4.10. Surface features formed in SrCl2–HCl of pH 2 The surface treated with a SrCl2–HCl solution was examined 3 days after the dissolution experiment where none of the observed surface features adhered to the

2524

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

AFM tip during scanning. Fig. 7d and e show hillocks on the basal (0 0 1) surface of uranophane which formed after treatment in an aqueous SrCl2–HCl solution of pH 2. The hillocks are elongated parallel to [0 0 1], and are bounded by rounded edges along their elongation. They have a maximum height of 93 nm and maximum lateral dimensions of 1150  320 nm. Their variation in size and pronounced elongation parallel to [0 0 1] distinguish them from the hillocks formed in an HCl solution of pH 3.5. Fig. 9a shows an SEM image taken in backscatter mode of the corresponding basal surface. The surface has low relief but apparently is not chemically homogenous. The major part of the SEM image appears white, however darker parts are visible around edges and steps. The XPS spectra show that (a) the treated surface is extremely depleted in U and Ca, (b) the dominant O species is either OH or O2 bridging two silica tetrahedra (Table 4) and (c) it has the lowest O : Si ratio (3 : 1) of all treated surfaces (Table 2).

the AFM tip during scanning. XPS spectra show that the surface does not contain any U and that the major O species on the surface is H2Ointerst or more likely, the positive species (Si–OH2)+, (Tables 2 and 4), (Duval et al., 2002). 4.12. Chemical composition of the solutions after the dissolution experiments Fig. 10 shows the results of the chemical analyses of the solutions used in the dissolution experiment with synthetic powders of uranophane. Fig. 10a and b indicate the stoichiometric ratios between U : Si and U : Ca in all of the analysed solutions, respectively. The size of the data points in the latter two plots corresponds to the experimental uncertainty of the largest measured concentration of U with 4.3 mmol  3.5% = ±0.15 mmol. Lines in the plot indicate the average atomic ratios between U and Si (0.98 : 1) and U and Ca (1.95 : 1) of the untreated powder sample. The plots show that almost all solutions are depleted in Si with respect to U (Fig. 10a) and that their U : Ca ratios are close to mineral stoichiometry (Fig. 10b). The former observation indicates the formation of a secondary phase enriched in Si. Solid phases after centrifugation were identified as uranophane (XRD) and an amorphous Si-rich phase with minor U (SEM). Uranyl-hydroxy-hydrate phases could not be identified in the XRD-pattern. Fig. 10c shows the concentration of U after the 10, 30 and 45 min experiments. Here, the concentration of U increases with time and approaches the maximum possible concentration for a completely dissolved uranophane sample with 2  2.3 mmol L1 = 4.6 mmol L1. Closer inspection of Fig. 10c shows that

4.11. Surface features formed in a MgCl2–HCl solution of pH 2

Ca concentration [mmol]

Si concentration [mmol]

Fig. 9a shows a SEM image in backscatter mode of the basal surface of uranophane 3 days after the treatment with a MgCl2–HCl solution of pH 2. The surface of uranophane is characterized by a Si-rich coating of high roughness, and attempts to produce AFM images of reasonable quality were unsuccessful. Fig. 7c shows an AFM image from the surface of another crystal taken 1 day after treatment with an MgCl2–HCl solution of pH 2. The (presumably) dry surface of the Si-enriched coating is characterized by cracks and spallings and the coating material did not adhere to

a

4 3 2 1 0 0

1

2

3

2.5

b

2.0 1.5 1.0 0.5 0.0 0

4

1

3 Mg

2

Sr HCl Ca Ba

1

Pb

0

d

-1

c

4

10

20

30

40

Duration of experiment [min]

2

3

4

U concentration [mmol] U conc. / time [mmol min ]

Concentration U [mmol]

Uranium concentration [mmol]

50

0.20

Mg Sr HCl

0.15

Ca

0.10

Pb

Ba

0.05 0.00 0

10

20

30

40

50

Duration of experiment

Fig. 10. Results of the chemical analysis by ICP-OES after dissolution experiments with powders of synthetic uranophane: plots of the concentrations [mmol1] of (a) U versus Si and (b) U versus Ca; lines are indicating the U:Si and U:Ca atomic ratios in the untreated uranophane sample; (c) the concentration of U [mmol L1] after 10, 30 and 45 min for the different dissolution experiment; (d) change in the concentration of U per minute ([U]time) between 0 and 10, 20 and 30 and 30 and 45 min; different symbols in (c) and (d) indicate data from different dissolution experiments.

Dissolution of uranophane

(1) During the treatment with MgCl2–HCl, SrCl2–HCl and HCl (pH 2) solutions, the concentrations of U are always higher than during the treatment with CaCl2–HCl, BaCl2–HCl and Pb(NO3)2–HCl solutions. (2) During the treatment with CaCl2–HCl solutions, the concentration of U is higher in the early stage (after 10 min) but lower in the later stage (45 min) than during the treatment in the Pb(NO3)–HCl solution. Fig. 10d shows the change in concentration of U per minute ([U]time) versus the duration of the experiment. The data points at the 10 min marker corresponds to [U]time in the first 10 min of the experiments, whereas the data points at 30 and 45 min correspond to [U]time between the 10th and 30th and 30th and 45th minute of the experiments, respectively. Closer inspection of Fig. 10d shows that at earlier stages of the experiment, [U]time increases in the different solutions in the sequence Pb(NO3)2 < BaCl2 < CaCl2 < HCl < SrCl2 < MgCl2. In the latter stage of the experiment, where the uranophane sample in the MgCl2 solution has been completely dissolved, [U]time has the largest value in the Pb(NO3)2–HCl solution and similar values in all of the other solutions. AFM, XPS and SEM examinations of single crystals and powders indicate the formation of U- and Si-bearing phases during the dissolution experiments. Hence, [U]time is not only controlled by the dissolution rate of uranophane but also by the stability of the U-bearing phases in acidic electrolyte solutions. Thus, we cannot use [U]time to calculate exact dissolution rates for uranophane. The concentration of Ca is not affected by the formation of these phases but were only measured in the solutions Pb(NO3)2–HCl, HCl (pH 2), SrCl2–HCl and MgCl2–HCl. Here, the change in the concentration of Ca with time ([Ca]time) increases in the sequence Pb(NO3)2 < HCl < SrCl2 < MgCl2. This ranking agrees with the ranking for [U]time and suggests that the latter ranking of [U]time expresses the relative dissolution rates of uranophane in the early stages of the different dissolution experiments. The concentration of U after the complete dissolution of single crystals of uranophane in MgCl2–HCl and CaCl2– HCl solution is in the range of 0.265 mmol L1. This value indicates the maximum concentration of U and Si during the dissolution experiments prior to the AFM studies, where single crystals of similar size did not completely dissolve. 4.13. Calculated saturation indices with respect to uranophane, schoepite and amorphous silica Table 1 lists calculated saturation indices with respect to amorphous silica-gel for solutions of pH 2 and 3.5 after treatment of uranophane powder samples for 45 min and 3 h, respectively. The indices indicate that all solutions of pH 2 were supersaturated with respect to amorphous silica-gel, which agrees with the observed precipitation of the amorphous Si-rich phase during the dissolution experiment. All solutions were undersaturated with respect to schoepite, even in the HCl solution with an initial pH of 3.5 and a final pH of 5.4 (log IAP  log Ks = 0.6). In all solutions, the de-

2525

gree of saturation with respect to uranophane increases with time. For example in the experiment with the HCl solution of pH 2, the degree of saturation increases from log IAPlog K2 = 9.4 after 10 min, to 6.9 after 30 min and to 6.1 after 45 min of treatment. Note that the degree of saturation with respect to uranophane is higher in the CaCl2–HCl solution (log IAP  log Ks = 7.9) than in all other electrolyte solutions ðlog IAP  log K s ¼ 11:8PbðNO3 Þ2 –HCl –  9:6MgCl2 –HCl Þ. The concentration of U after the complete dissolution of a single crystal is 0.265 mmol L1. If we assume a similar concentration for Si, the saturation index with respect to the amorphous silica-gel would be log IAP  log Ks = 1.84, indicating that all solutions during the dissolution experiments of single crystals were undersaturated with respect to amorphous silica-gel. 5. DISCUSSION Uranophane and uranophane-b commonly form fibrous crystals with small basal faces, which do not allow AFM examination of dissolution and precipitation processes on their surfaces. The use of platy crystals with a sufficiently large basal surface made it possible to record for the first time dissolution–precipitation phenomena on the basal surface of uranophane. 5.1. Morphology of etch pits as a function of the state of saturation of the solution The bulk-dissolution experiments with powder samples of uranophane showed a decrease in the degree of undersaturation with the duration of the dissolution experiment. These results can be used to explain the change in the morphology of etch pits which formed in the experiments with HCl solutions of pH 2 (in the absence of divalent cations). Figs. 4 and 5 show that the outlines of the etch pits change with an increase in saturation of the aqueous solution. Elongated etch pits defined mainly by [0 1 0] edges form in HCl solutions of pH 2 that have an extremely low saturation with respect to uranophane. Triangular etch pits defined by the [0 1 2] and [0 1 0] edges form in a more highly saturated solution (Fig. 4b–e) and rectilinear etch pits defined by [0 1 0] and [0 0 1] edges form in solutions with an even higher saturation state with respect to uranophane (Fig. 5a and b). Fig. 4b–e shows the possible growth sequences of etch pits that formed during the 10-min batch experiments with an HCl solution of pH 2. Etch pits with a triangular outline may represent the earliest stage of growth. The triangular etch pits transform into rectilinear etch pits during growth parallel to the basal surface (Fig. 4b–e). The rectilinear etch pits may represent the latest stage of growth because their outline is similar to those of the etch pits observed after 30 min of treatment. The occurrence of edges at different degrees of saturation agrees with our predictions (Schindler et al., 2004b) and is similar to the observations on growth zones of uranophane: the edge with the lowest bond-valence deficiency ([0 1 0]) is the only edge that forms at high dissolution rates of uranophane. Continuous growth of

2526

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

etch pits and the contemporaneous increase in saturation of the solutions, lead to the elongation of etch pits terminated by [0 0 1] and [0 1 1] edges and the disappearance of the [0 2 1] and [0 1 2] edges. 5.2. Depth and lateral dimensions of etch pits as a function of the saturation state of the solution The average area and the depth of etch pits may be used to express the stability of the edges and the change in relief of a surface. Here, higher relief of an etch pit is not the result of a faster dissolution rate perpendicular to the sheets, rather it is an expression of the difference between the lowering of the surrounding surface through dissolution and growth of the etch pit perpendicular to the sheet. Fig. 5c shows the area of an etch pit as a function of the average depth observed in the 10 and 30 min batch experiments. The triangular-to-rectilinear etch pits (observed in the 10 min batch experiment, Fig. 4b–e) have a larger ratio between depth and lateral area than the rectilinear etch pits observed in the 10 and 30 min batch experiments. Hence, the ratio between the lowering of the surface and the growth of the etch pit perpendicular to the sheet, decreases with increasing saturation state. This conclusion is in accordance with the observed depth of etch pits that formed during the five-minute constant-flow experiment. The depth of these etch pits is far greater than the depth of etch pits that formed in the batch experiments (Fig. 6a and b), however, the number of etch pits in the five-minute constant-flow experiment is too small to form a representative set of data for Fig. 5c. 5.3. Etch-pit nucleation as a function of saturation state of the solution There are numerous studies on the formation of etch pits and their dependence on the saturation state in solutions (e.g. Burton et al., 1951; Frank, 1951; Carbera et al., 1954; Lasaga, 1983; Lasaga and Blum, 1986; Teng, 2004). The formation of etch pits generally depends on the character of the dislocations (i.e. strain energy and the size of the dislocation cores) and the saturation state of the solution (e.g. Frank, 1951; Carbera et al., 1954). The magnitude of etch-pit formation correlates with the degree of undersaturation in solution, and at a critical saturation state, no etch pits can nucleate and step dissolution becomes the dominant dissolution mechanism (e.g. Lasaga, 1983; Teng, 2004). On the other hand, two-dimensional surface nucleation of etch pits becomes operative at extremely low saturation states (Burton et al., 1951), (i.e. a state where pits start to form in defect-free areas). The structure of uranophane is characterized by sheets of polymerised uranyl polyhedra. Dislocation effects in structures based on sheets of polymerized polyhedra are characterized by point defects, which results in the gliding of layers of polyhedra parallel to the basal surface (e.g. Noe and Veblen, 1999). Nucleation of an etch pit and its growth perpendicular to the basal surface of uranophane is controlled by the occurrence and number of point defects between the layers of polymerized uranyl polyhedra. An etch pit may grow perpendicular to the basal surface until

the saturation of the solution is not able to overcome the free-energy barrier to etch-pit formation on the underlying sheet of polymerized uranyl polyhedra. The results of the dissolution experiments on uranophane show that the ratio between the growth-rate perpendicular to the surface and the lowering of the surrounding surface increases with decreasing saturation state. One reason for this phenomenon must be a change in the growthrate of the pit perpendicular to the basal surface and the lowering of the surface peripheral to the etch pits (see below). For example at a higher saturation state, the growth of a pit perpendicular to the basal face may have been limited by the occurrence of point defects in the underlying sheets. At lower saturation states such as, in the five-minute constant-flow experiment, the etch pits may have grown perpendicular to the basal face through continuous etchpit nucleation on defect-free areas at the bottom of the pits. 5.4. Lowering of the surface and the dissolution stepwave model The dissolution stepwave model describes the role of etch pits in terms of their ability to generate a continual sequence of steps (Lasaga and Lu¨ttge, 2001; Lassga and Lu¨ttge, 2003; Arvidson et al., 2004). Here, the peripheries of the etch pits can be considered as locations at which the mineral surface drops significantly through the formation of an etch pit. These peripheries always contain a number of steps that move into the rest of the mineral surface during their dissolution. The continuous movement of these steps into the mineral surface produces dissolution stepwaves. With the increasing duration of the experiment, these stepwaves control the overall dissolution rate of the crystal. Formation of stepwaves elegantly explain the lowering of the surface around an etch pit on the surface of uranophane. After the initial formation of an etch pit (typically after 5– 10 min of the dissolution experiment), stepwaves move into the rest of the crystal and continuously lower the surface of the crystal. After 30 min of the experiment, the surface surrounding the etch pit has been flattened out and the etch pit is more shallow than those of the first generation. 5.5. Possible chemical composition of the hillocks formed at pH 3.5 The XPS spectra show that the hillocks formed in an HCl solution of pH 3.5 are part of a U-rich phase with a high proportion of OH and H2Ointerst. Similar hillocks of a U-rich phase were found on the surface of calcite after the interaction with a uranyl-acetate solution of pH 4.5 (Schindler and Putnis, 2004). The authors identified the corresponding precipitate as schoepite, [(UO2)8O2(OH)12](H2O)12, together with a large amount of X-ray- amorphous material. The AFM and XPS observations indicate that the hillocks on the surface of uranophane also represents a UO3(H2O)n phase. We showed above, that solutions after dissolution experiments with HCl solutions of pH 3.5 were undersaturated with respect to schoepite, which suggests that the hillocks are part of an amorphous phase rather than that of schoepite.

Dissolution of uranophane

5.6. Possible chemical composition of the small crystallites The chemical and morphological features of the Pb- and Ba-containing crystallites clearly show that the corresponding phases differ in structure and composition from uranophane. The XPS spectra display that the surface treated with a Ba–HCl solution of pH 2 is enriched in U, which indicates the precipitation of a Ba-uranyl-hydroxy-hydrate phase on the surface of uranophane. This conclusion is supported by the following facts: (1) The mole proportion of Ba and the U : Si ratio, increases with the duration of the Ba–HCl experiment. (2) The surface treated with a BaCl2 solution for 1 h has a higher proportion of O@U@O and a similar proportion of TAO than the untreated surface. As described above, the latter band overlaps with the O@U@O band that occurs in spectra of uranylhydroxy-hydrate minerals. Ba-uranyl-hydroxy-hydrate minerals include protasite, Ba[(UO2)3O3(OH)2](H2O)3, and billietite, Ba(H2O)8[(UO2)6 O4(OH)6]. Schindler et al. (2004a,b,c) showed that the latter mineral forms easily at room temperature by the interaction of a weakly acidic Ba solution with dehydrated schoepite, an observation that suggests that billietite also formed on the surface of uranophane. The Pb-bearing crystallites may also represent a uranyl-hydroxy-hydrate phase, since Pb-bearing uranyl-hydroxy-hydrate minerals have a lower solubility than other alkaline-earth-bearing uranyl-hydroxy-hydrate phases (Finch and Murakami, 1999). 5.7. Si-enriched coatings and particles: morphological features Information on the composition of the Si-enriched coating and the orientation of the Si-enriched particles was obtained from XPS spectra and AFM and SEM images taken from the surfaces that were treated with CaCl2–HCl, SrCl2– HCl and MgCl2–HCl solutions of pH 2. The Si-enriched coatings and particles show hereby different morphological and physical features: (I) Hillocks as part of a thin silica-rich coating occur on the basal surface of uranophane 3 days after treatment with a SrCl2–HCl solution of pH 2 (Fig. 7d) and material did not attach to the AFM tip during scanning. (II) A thick Si-rich layer with cracks and spallings occurs on the surface of uranophane 1 day after treatment with an MgCl2–HCl solution of pH 2 (Fig. 7c and d) and material did not attach to the AFM tip during scanning. (III) A thin Si-rich layer occurs 12 h after treatment with a CaCl2–HCl solution of pH 2 (Fig. 7b) and material adhered to the AFM tip. The latter two observations are in accordance with AFM observations of wet and dried Si-enriched layers on the surface of feldspar. In the first case, Hellmann et al. (1992), Jordan et al. (1999) and Teng et al. (2001) observed

2527

a wet and soft Si-enriched layer with particles attached to the AFM tip during scanning. In the second case, Teng et al. (2001) and Casey et al. (1989) reported the occurrence of a dry Si-enriched layer 6 h after the dissolution experiment, which were characterized by cracks and spallings. These observations suggest that the Si-rich hillocks formed on the surface of uranophane 3 days after treatment with a SrCl2–HCl solution of pH 2, differ structurally and chemically from Si-rich coatings observed on the surface of uranophane and feldspar after the treatment in acidic solutions. 5.8. Si-enriched coatings and particles: structural characterization and chemical composition Structural differences on the surface of silicates can be explored via the examination of the binding energies of the Si 2p and O 1s peaks. The binding energy of the Si 2p peak (Si 2p3/2 and Si 2p1/2) shifts to higher binding energies with an increase in the number of polymerized silica tetrahedra. We additionally examined the Si 2p and O 1s spectra of a-quartz (framework), spodumene (chain) and forsterite (isolated polyhedron). Here, the binding energy of the Si 2p peak increase from 102.0 to 102.2 eV for uranophane (Fig. 11e) and forsterite, to 102.6 for spodumene, and 103.2 eV for quartz (Fig. 11c). Furthermore, the binding energy increases with the degree of protonation of the Oatoms of the silica tetrahedra. For example, the Si 2p spectrum of synthetic silicic acid, H2SiO3(H2O)n (Sigma) occurs at 103.9 eV (Fig. 11a). Fig. 11b–d shows the Si 2p spectra of the silica coatings formed on uranophane after treatment with a MgCl2–HCl and SrCl2–HCl solution of pH 2, respectively. Here, the Si 2p spectrum of the latter coating occurs at slightly higher binding energies than the spectrum for untreated uranophane, whereas the Si 2p peak of the coating formed in a MgCl2–HCl solution occurs at similar binding energies as the Si 2p peak for silicic acid. Fig. 12 shows the O 1s spectra for uranophane, aquartz, silicic acid and the Si-rich coatings. The O 1s spectra indicate that the major species on the surface of silicic acid and the coating formed in a MgCl2–HCl solution is H2Ointerst or most likely the positive termination (Si– OH2)+ (Duval et al., 2002), whereas the surfaces of a-quartz and the coating formed in a SrCl2–HCl solution contain mainly the species SiAOASi and SiAOH, respectively. The Si 2p and O 1s spectra show that the coating formed in a MgCl2–HCl solution is structurally and chemically similar to silicic acid. This agrees with the examinations of Sirich coatings on feldspars, which showed that the latter coatings have a composition and structure similar to that of silica gel (Casey et al., 1989). The coating observed 3 days after treatment in a SrCl2–HCl solution of pH 2, has a lower degree of hydration than the coating formed in a MgCl2–HCl solution and its structure contains silica tetrahedra in a higher and lower degree of polymerization than the structures of uranophane and quartz, respectively. The above observations can now be compared with the measured O : Si ratio for silicic acid and the coatings on uranophane (Table 2). The measured O : Si ratios of silicic acid and the silica coating formed on the surface of urano-

2528

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

Fig. 11. The Si 2p spectra for various silica phases; maximum and minimum binding energies for the Si 2p peaks are indicated with lines.

Fig. 12. The O 1s spectra for various silicate phases. The location of the bands H2Oadsorb, SiAOH2, SiAOH/SiAOASi and SiAO are indicated with vertical grey-shaded bars.

phane after treatment with a MgCl2 solution of pH 2 are 5.3 : 1 and 5.2 : 1, respectively. These ratios were calculated with the sensitivity factor (RSF = 0.57) derived from the spectra for the untreated uranophane crystals (see above).1 The calculated O : Si ratio with 3 : 1 (using RSF = 0.57 for Si 2p) for the coating on the uranophane surface treated with a SrCl2–HCl solution of pH 2, indicates an average

polymerization degree of a chain (H2SiO3 or Si2O2(OH)4), where the non-bridging O groups must be protonated to obtain electro-neutrality. This chemical composition agrees with the above conclusions of the polymerization degree and chemical composition of the coating. Chemical compositions and the degree of polymerization of the silica tetrahedra indicate that the examined coating on the surface of uranophane 3 days after treatment with a SrCl2–HCl solution has undergone an aging process, where H2Ointerst species were released and Si-tetrahedra started to polymerize. The occurrence of such an aged coating on the surface would explain its different morphological

1 Note, the calculated O:Si ratios would be 1.8 : 1 and 1.7 : 1 for a sensitivity factor of Si with RSF = 0.328 from the software Vision 2.2.6 (2006).

Dissolution of uranophane

features as compared to other Si-rich coatings on the surface of uranophane and feldspar. 5.9. Si-enriched coatings and particles: possible mechanisms of formation There is some controversy regarding the mechanism of formation of the Si-enriched coating on rock-forming silicates treated with acidic solutions; i.e. whether the coating is a product of leaching or dissolution-re-precipitation process (see Brantley, 2005 and references therein). The major argument for a leaching process is that many of the acidic solutions used in the dissolution experiments were undersaturated with respect to amorphous silica. Observations that support the idea of a dissolution reprecipitation process on the surface–water interface are (1) The decrease in thickness of the amorphous silica layer with the flow rate of the acidic solution above the crystal, which should have no effect on the thickness of a coating formed by a leaching process (Teng et al., 2001). (2) The sharp interface between the amorphous layer and the structure of the mineral, as examined on the surface of feldspar with high-resolution transmission electron microscopy (Hellmann et al., 2003). We observe similar complex results for the dissolution experiments on uranophane: (a) During the dissolution of single crystals, all solutions were undersaturated with respect to amorphous silica-gel. (b) A silica-rich coating did not form during the constant-flow experiment for 2 min. A further argument for a dissolution–re-precipitation process at the interface of uranophane is the structure of the mineral, where the silicate polyhedron is isolated and shares a common edge with a uranyl polyhedron (short: SiAUO2ASi). Removal of a uranyl polyhedron would require breaking SiAOAU bonds along the common edge of the SiAUO2ASi groups and the common corners between two adjacent SiAUO2ASi groups. Nevertheless, we cannot be definitive on the formation of the silica-rich layer on the surface of uranophane and further TEM studies of the interface between the mineral structure and the amorphous silica layer may give more insights to the formation of this coating. 5.10. Si-enriched coating and Ba–U-bearing crystallites Let us examine the change in the XPS spectra with the duration of the Ba–HCl experiment at pH 2 (Fig. 2c and d). As mentioned above, the U : Si ratio and the mole proportion of Ba increases continuously, whereas the ratios Si:Ca, O@U@O : OH and O@U@O : H2Ointerst decreases in the first 30 min but then increases between 30 and 60 min (see above and Table 2). The change in these ratios between 30 and 60 min of the experiment suggests either

2529

precipitation of a BaAUASiAOAOH phase [with (O@U@O) > (Si + OH) and a similar proportion of TAO as the untreated surface uranophane] or simultaneous formation of a BaAUAOAOH phase [with O@U@O > OH] and a H4SiO4(H2O)n coating [with BaAUAOAOH > H4SiO4(H2O)n]. There are two observations indicating that the latter process is more likely: (1) The AFM image (Fig. 8d) clearly shows the occurrence of small crystallites embedded in a soft coating on the surface of uranophane. (2) No mineral or synthetic BaAUASiAOAOH phase contains a higher molar proportion of U than Si. The presence of a BaAUAOAOH phase would also explain the high proportion of the TAO band in the O 1s spectrum of the surface treated for 1 h (Fig. 2a), because the latter band overlaps with the O@U@O band in O 1s spectra of uranyl-hydroxy-hydrate minerals (see above). 5.11. Composition of the surface versus proportions of the O species We have now constrained all potential secondary phases that formed during the dissolution of uranophane in different solutions. Now we will consider how the change in the mole proportions of U and Si affect the proportions of the different O species on the corresponding surfaces. On the surface of uranophane, the species O@U@O and TAO (or O@U@O for the uranyl-hydroxy-hydrates formed on uranophane) can occur only in uranyl polyhedra and in silicate polyhedra that are linked to uranyl polyhedra. The latter case assumes that there are no unprotonated SiAO groups (very likely at pH 2) and a limited number of SiAOASi bridges, which is reasonable because there are no SiAOASi bridges in uranophane. Thus, one would expect a correlation between the proportions of O@U@O and TAO (or TAO + O@U@O) and the mole proportions of U. Fig. 13a and b shows that this is indeed the case: the proportion of O@U@O and TAO (TAO + O@U@O) correlates strongly with the mole proportion of U. The composition of the surfaces treated with HCl solutions of pH 3.5 deviate slightly from the correlation between O@U@O and the mole proportion of U. However Schindler et al. (in this issue-b) showed that the O 1s spectra of non-freshly cleaved UO3(H2O)n phases are characterized by extremely high proportions of OH and H2Ointerst (indicated with s in Fig. 13a). We showed above that the composition of a surface of uranophane with regard to the U : Si ratio and the mole proportion of O depends on the roughness and the type of precipitate, suggesting that there is a relation between the ratio of (O@U@O + TAO) : OH and the U : Si ratio. The corresponding plots (Fig. 13c) show distinct groups of data, which can be distinguished by the type of surface or surface precipitates: (a) Untreated surfaces have similar U : Si ratios but, depending on their roughness, have different ratios of (O@U@O + TAO) : OH; (marked with a and b

2530

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

Proportion of O=U=O

35

5.12. The effect of aqueous cations on dissolution in HCl solutions of pH 2

a

30 25 20 15

s

10 5 0 0

2

4

6

8

10

mol proportion of U

Proportion of T-O

30

b

25 20 15 10 5 0 0

2

4

6

8

10

(O=U=O + T-O) : OH ratio

mol proportion of U 3.0 2.5

c

d

a

2.0

b

1.5 1.0

c

0.5

12d

0.0 0.0

0.5

1.0

1.5

2.0

U : Si ratio

Fig. 13. The proportions of (a) O@U@O and (b) TAO as functions of the mole proportions of U and the ratio of (c) (O@U@O + TAO) : OH are plotted against the ratios of U : Si; the small letter s refers to the formation of a schoepite-type phase and the small letters a, b, c, d and 12d refer to the corresponding O 1s spectra in Figs. 2 and 12, and data points corresponding to surfaces with similar types of precipitates are indicated with fields of different tones of grey (see text for details).

in Fig. 13c as a reference to Fig. 2a and b), which show typical O 1s spectra of this group; (b) Surfaces with U-rich precipitates have higher U : Si ratios and lower (O@U@O + TAO) : OH ratios than the untreated surfaces; marked with c in Fig. 13c as a reference to Fig. 2c, which shows a typical O 1s spectrum of this group; (c) Surfaces with Si-rich coatings have lower ratios of U : Si, O@U@O : OH and TAO : OH with respect to an untreated surface; marked with 12d in Fig. 13c, as a reference to Fig. 12d, which shows the O 1s spectra for this group. (d) The surface treated with a BaCl2–HCl solution for 60 min contains a U-rich precipitate and a Si-coating (marked with d in the Fig. 13c, as a reference to Fig. 2d, which shows a typical O 1s spectrum of the surface).

It is apparent from the above observations that the type of cation in solution has a major effect on the occurrence and morphology of etch pits, the relief and chemical composition of the basal surface and the chemical composition of the solutions after dissolution experiments on powder samples. One can distinguish three different types of phenomena: (1) Interaction of uranophane with Pb2+(NO3)2–HCl and BaCl2–HCl solutions of pH 2 for 30 min results in basal surfaces of low relief, the formation of small Ba- or Pb-bearing crystallites and, in the case of the BaCl2–HCl experiment in the formation of small amounts of a hydrated-silica coating. (2) Interaction of uranophane with SrCl2–, CaCl2– and MgCl2–HCl solutions of pH 2, results in basal surfaces of medium and high relief and in the formation of larger amounts of a hydrated-silica coating. (3) Dissolution experiments with synthetic uranophane powders showed that the dissolution rate of uranophane in the early stage of dissolution experiments increased with the different solutions in the sequence Pb(NO3) < BaCl2 < CaCl2 < HCl < SrCl2 < MgCl2. (4) The degree of saturation with respect to uranophane is always higher in the CaCl2–HCl solution than in all other solutions (see above). Hence, the results of its dissolution experiment can not be compared with the other experiments in terms of the dissolutionpromoting or -inhibiting ability of the cations in solution. These observations suggest that the dissolution of uranophane is more enhanced in solutions containing smaller cations such as Mg and Sr than Ba or Pb. However, this conclusion is only valid for the early dissolution stage at which time, the interaction between the surface and cations in solution is mainly controlled by the surface features of uranophane rather than by the surface of the alteration layer. At a later stage of dissolution, the properties of alteration layers such as porosity, surface area and surface charge may control the dissolution rate of uranophane: their examination must await further experiments. 5.13. Lewis acids and bases on the surface of uranophane The interaction between a cation in solution and the surface of a mineral is controlled by the speciation of the cation in solution and the type and number of surface terminations. For example in the MgCl2–HCl solution with a pH of 2, the aqueous species [[6]Mg(H2O)6]2+ and [[6]MgCl(H2O)5]+ occur in a ratio of 56 : 44. Philips et al. (1997) showed that F ions in aqueous {AlFx(H2O)6x}3x species affect the rate of exchange of remaining water molecules between the bulk solution and the Al coordination sphere. Similarly, the stability of a bond between a Mg and an O-atom on the surface of uranophane should be then affected by the presence of a Cl ion in the corresponding aqueous species; i.e. the inner-sphere complexes

Dissolution of uranophane

[Osurface–[6]Mg(H2O)5]2+ and [Osurface–[6]MgCl(H2O)4]+ should have different stabilities. The stability of the inner-sphere complexes between the cationic aqueous species and O-atoms of the surface of uranophane can be explained with their Lewis-acid and base strengths. The Lewis-acid strength can be defined as the valence of a cation Mz+ divided by its average coordination number (Brown, 2002) and the Lewis-acid strength of a [[n]Mz+Clx]z1 complex can be defined as the charge of the complex divided by the number of emanated bonds (Schindler and Hawthorne, 2001). For example, the Lewis acidity of Mg is 0.36 vu, which corresponds to an average coordination number of 2/0.36 = 5.55 (Brown, 2002). For the same coordination number of Mg, the Lewis acidity of the [[5.5]MgCl]+ complex would be 1/5.5 = 0.18 vu. Hence, the average Lewis acidity of Mg-bearing aqueous species in a MgCl2–HCl solution of pH 2 would be 0.33  56% + 0.18  44% = 0.26. Using the Lewis acidities listed by Brown (2002), and Schindler and Hawthorne (2001), we calculated the average Lewis acidity for aqueous species in the electrolyte solutions used in the dissolution experiments (Table 1). It is apparent that the Lewis acidity increases for the cationic aqueous species in the sequence Pb < Ba < Sr < Ca < Mg. The Lewis base strength of an anionic surface group can be defined as the missing bond-valence of an anion divided by the number of accepted bonds emanated from cationic species in solution (Schindler et al., 2004a). For example, an O-atom of a surface termination SiAO accepts 1 vu from the SiAO bond and requires additional 1.0 vu. Assuming a [4]-coordinated O-atom, the O-atom accepts three more bonds from species in solution. Hence, the Lewis basicity of the SiAO termination is 1/3 = 0.33 vu. The surface of uranophane contains three different types of Lewis bases on its surface: the strong Lewis bases UAO (0.49 vu) and SiAO (0.33 vu) and the weak Lewis base [7]U = O (0.12 vu). Strong Lewis acids and bases show strong affinities for each other (Brown, 2002). This affinity can be applied to the selectivity of divalent cations for mineral-oxide surfaces. Here, strong Lewis bases on the surfaces of Al-, Fe- and Sioxides have a higher affinity for inner-sphere complexes with strong Lewis acids in solution. This affinity increases for divalent cations in the sequence Ba < Sr < Ca < Mg (Hayes and Katz, 1996, p. 155). Contrary to this, weak Lewis bases on the surface of smectites show a higher selectivity versus weaker Lewis acids and the affinity to adsorb cations increases in the sequence Mg < Ca < Sr < Ba (Sposito, 1984, p. 129; Hayes and Katz, 1996, p. 155). In terms of the affinity of the Lewis bases on the surface of uranophane, the equatorial UAO and SiAO terminations (strong Lewis bases) have stronger affinities versus the stronger Lewis acids Mg and Ca, whereas the apical U = O terminations (weak Lewis bases) have stronger affinities versus the weaker Lewis acids Ba and Pb. 5.14. Degree of dissolution in acidic solutions The different degrees of dissolution in the presence of specific cations in solution may be explained by the role of the different Lewis acids in solution during the two prin-

2531

ciple mechanisms involved in a surface-controlled dissolution process: (i) formation of an activated complex, and (ii) the detachment of the species from the structure (Stumm, 1992; Schindler et al., 2006b): (1) If a cation in the solution forms an inner-sphere complex with an O-atom of the uranyl ion, it can inhibit protonation of the equatorial O-atoms that link polyhedra or polyhedron clusters; in turn, this will reduce the number of activated sites and hence the degree of dissolution of the basal surface. Following the discussion above, the weak Lewis bases U@O have a higher affinity to form inner-sphere complexes with the weaker Lewis acids (larger cations) and therefore the latter cations can inhibit more effectively the protonation of the equatorial O-atoms. (2) Once equatorial O-atoms are protonated and the bonds between polyhedra or polyhedron groups are broken, the cations of the solution can form innersphere complexes with the equatorial surface termination UAO and SiAO. These surface terminations are strong Lewis bases, which show a higher affinity for the strong Lewis-acid Mg (small cation). The formation of inner-sphere complexes with smaller cations will increase the detachment rate of the polyhedra or polyhedron groups and thus increase the degree of dissolution of the basal surface. 5.15. Comparison with surface examinations of dissolution processes on other uranyl-phases Schindler et al. (2006a,b, 2007a,b) examined dissolution processes for becquerelite, billietite (Schindler et al., 2007b), fourmarierite, Pb1x(H2O)4[(UO2)6O32x(OH)4+2x] and synthetic Pb2(H2O)[(UO2)10UO12 (OH)6(H2O)2] in different electrolyte solutions of pH 2. They found that the relief of dissolution features (e.g. etch pits and grooves) on the basal surface increases with decreasing ionic radius of the cation in solution (increasing Lewis-acid strength). In the case of billietite and the Pb-uranyl-hydroxy-hydrates, they found that the relief of the dissolution features increases in the following sequence: KCl < Pb(NO3)2 < SrCl2 < NaCl2 < CaCl2 = MgCl2 = LiCl and (BaCl2, KCl) < (SrCl2, NaCl) < CaCl2 < MgCl2, respectively. Schindler et al. (2006a,b, 2007a,b) proposed that the degree of dissolution of a single crystal can be related to the relief of its basal surface. Similar to the discussion in this paper, they argued that a larger cation adsorbed on the surface can inhibit the protonation of equatorial O-atoms that link polyhedra or clusters of polyhedra and thus reduces the degree of dissolution on the basal surface. 6. CONCLUSIONS Examination of the dissolution processes of uranophane under acidic conditions shows that the surface processes vary strongly with: (1) The pH of the solution: etch pits form at pH 2 and a uranyl-hydroxy-hydrate phase precipitates at pH 3.5.

2532

M. Schindler et al. / Geochimica et Cosmochimica Acta 73 (2009) 2510–2533

(2) The valence of the cations in solution: etch pits form dominantly in the absence of divalent cations in solution. (3) The duration of the dissolution experiment: deep etch pits form in constant-flow experiments and shallow etch pits occur after the 30-min batch experiments. (4) The type of cation in solution: small crystallites form in the presence of Ba and Pb2+, and hydrated-silica coatings form in the presence of Ba, Sr, Ca and Mg. (5) Dissolution experiments on synthetic powders indicate that the dissolution rate at the early stage of the experiment increases in the sequence Pb(NO3)– HCl < BaCl2–HCl < CaCl2–HCl < HCl < SrCl2– HCl < MgCl2–HCl; indicating that the dissolution of uranophane is more enhanced in electrolyte solutions containing divalent cations of small ionic radii and high Lewis acidity (Mg, MgCl+). The latter observation agrees with previous dissolution experiments on single crystals of billietite, Ba(H2O)8 [(UO2)6O4(OH)6], fourmarierite, Pb1x(H2O)4[(UO2)6 O32x(OH)4+2x] and synthetic Pb2(H2O)[(UO2)10UO12 (OH)6(H2O)2], where the degree of dissolution inversely correlates with the radius of the cation in solution (Schindler et al., 2007a,b). ACKNOWLEDGMENTS This work was supported by a Canada Research Chair in Crystallography and Mineralogy and by a Discovery Grant to F.C.H. from the Natural Sciences and Engineering Research Council of Canada. M.S.F. was supported by a Canada Research Chair in conducting polymers. M.S.F. and F.C.H. are supported by a CFI grant for surface science at the University of Manitoba. P.C.B. and P.A.M. were supported by the National Science Foundation Environmental Molecular Sciences Institute at the University of Notre Dame (EAR02-21966). We thank Jennifer Szymanowski and Balz Kamber for the ICP-OES and ICP-MS analyses, Dwayne D. Crush for assistance on the AFM, Jennifer Durocher for editing the paper and Sergio R. Mejia and Panseok Yang for help on the SEM and Electron Microprobe, respectively. We also thank Associate Editor Jacques Schott, Jordi Bruno and two pseudo-anonymous reviewers for their comments on an earlier version of the manuscript.

REFERENCES Arvidson R. S., Beig M. S. and Lu¨ttge A. (2004) Single-crystal plagioclase feldspar dissolution rates measured by vertical scanning interferometry. Am. Mineral. 89, 51–56. Branche G., Chervet J. and Guillemin C. (1951) Nouvelles espe`ce uranife`res francßaises. Bull. Soc. Francß. Mine´ral. Cristall. 74, 457–488. Brantley S. L. (2005) Reaction Kinetics of Primary Rock-Forming Minerals Under Ambient Conditions. In Treat. Gheochem. (ed. J. I. Dreyer). Elsevier, 644 p.. Burns P. C., Ewing R. C. and Miller M. L. (1997) Incorporation mechanism of actinide elements into the structures of U6+ phases formed during the oxidation of spent nuclear fuel. J. Nucl. Mater. 245, 1–9. Burns P. C. and Li Y. (2002) The structures of becquerelite and Srexchanged becquerelite. Am. Mineral. 87, 550–557.

Burns P. C., Deely K. M. and Skanthakumar S. (2004) Neptunium incorporation into uranyl compounds that form as alteration products of spent nuclear fuel: Implications for geologic repository performance. Radiochim. Acta 92, 151–159. Brown I. D. (2002) The Chemical Bond in Inorganic Chemistry. The Bond-Valence Model. Oxford University Press. Burton W. K., Carbera N. and Frank F. C. (1951) The growth of crystals and the equilibrium structure of their surfaces. R. Soc. Philos. Trans. A 243, 299–358. Carbera N., Levine M. M. and Plaskett J. S. (1954) Hollow dislocations and etch pits. Phys. Rev. 96, 1153. Casas, I., Bruno, J., Cera, E., Finch, R. J., Ewing, R. C. (1994) Kinetic and thermodynamic studies of uranium minerals; assessment of the long-term evolution of spent nuclear fuel. SKB Tech. Report 94-16, 73 p. Casey W. H., Westrich H. R., Arnold G. W. and Banfield J. F. (1989) The surface chemistry of labradorite feldspar. Geochim. Cosmochim. 53, 821–832. Chen F., Burns P. C. and Ewing R. C. (1999) 79Se: geochemical and crystallo-chemical retardation mechanisms. J. Nucl. Mater. 275, 81–94. Chen F., Burns P. C. and Ewing R. C. (2000) Near-field behavior of 99Tc during the oxidative alteration of spent nuclear fuel. J. Nucl. Mater. 278, 225–232. Douglas M., Clark S. B., Friese J. I., Arey B. W., Buck E. C., Hanson B. D., Utsunomiya S. and Ewing R. C. (2005) Microscale characterization of uranium(VI) silicate solids and associated neptunium(V). Radiochim. Acta 93, 265– 272. Duval Y., Mielszarski J. A., Pokrovsky O. S., Mielczarski E. and Ehrhardt J. J. (2002) Evidence of the existence of three types of species at the quartz-aqueous solution interface at pH 0–10: XPS surface group quantification and surface complexation modeling. J. Phys. Chem. B 106, 2937–2945. Finch R. J. and Ewing R. C. (1992) The corrosion of uraninite under oxidizing conditions. J. Nucl. Mater. 190, 133–156. Finch R. J., Buck E. C., Finn P. A. and Bates J. K. (1999) Oxidative Corrosion of Spent UO2 Fuel in Vapor and Dripping Groundwater at 90 °C. In Scientific Basis for Nuclear Waste Management XXII, Mater. Res. Soc. Symp. Proc., vol. 556 (eds. D. J. Wronkiewicz and J. H. Lee). Materials Research Society, Warrendale, PA, pp. 431–439. Finch R. J. and Murakami T. (1999) Systematics and paragenesis of uranium minerals. Rev. Mineral. 38, 91–180. Finn P. A., Hoh J. C., Wolf S. F., Slater S. A. and Bates J. K. (1996) The release of uranium, plutonium, cesium, strontium, technetium and iodine from spent fuel under unsaturated conditions. Radiochim. Acta 74, 65–71. Frank F. C. (1951) Capillary equilibria of dislocated crystals. Acta Crystallogr. 4, 497–501. Frondel C. (1958) Systematic mineralogy of uranium and thorium. US Geol. Surv. Bull. 1064, 400 p.. Ginderow D. (1988) Structure de l’uranophane alpha, Ca(UO2)2 (SiO3OH)25H2O. Acta Crystallogr. C 44, 421–424. Grenthe I., Plyasunov A. and Spahiu K. (1997) Estimations of Medium Effects on Thermodynamic Data. In Modelling in Aquatic Chemistry (eds. I. Grenthe and I. Puigdomenech). OECD Nuclear Energy Agency, pp. 325–426. Gustafsson J. P. (2007) Visual MINTEQ, Version 2.53. KTH. Department of Land and Water Resources Engineering, Stockholm, Sweden. Hayes K. F. and Katz L. E. (1996) Application of X-Ray Absorption Spectroscopy for Surface Complexation Modeling of Metal Ion Sorption. In Physics and Chemistry of Material (ed. P. V. Brady). CRC Press.

Dissolution of uranophane Hellmann R., Drake B. and Kjoller K. (1992) Using Atomic Force Microscopy to Study the Structure, Topography and Dissolution of Albite Surfaces. In Water–Rock Interaction WRI-7 (eds. Y. K. Kharaka and A. S. Maest). A.A. Balkema, Rotterdam, pp. 149–152. Hellmann R., Penisson J. M., Hervig R. L., Thomassin J.-H. and Abrioux M.-F. (2003) An EFTEM/HRTEM high-resolution study of the near surface of labradorite feldfspar altered at acid pH: evidence for interfacial dissolution–reprecipitation. Phys. Chem. Miner. 30, 192–197. Jordan G., Higgins S. R., Eggleston C. M., Swapp S. M., Janney D. E. and Knauss K. G. (1999) Acidic dissolution of plagioclase: in situ observations by hydrothermal atomic force microscopy. Geochim. Cosmochim. Acta 63, 3183–3191. Klingensmith A. L. and Burn P. C. (2007) Neptunium substation in synthetic uranophane and soddyite. Am. Mineral. 92, 1946– 1951. Lasaga A. C. (1983) Kinetics of silicate dissolution. 4th International Symposium on Water–Rock Interactions. Misasa, Japan, pp. 269–274. Lasaga A. C. and Blum A. E. (1986) Surface chemistry, etch pits and mineral–water interactions. Geochim. Cosmochim. Acta 50, 2363–2379. Lasaga A. C. and Lu¨ttge A. (2001) Variation of crystal dissolution rate based on a dissolution stepwave model. Science 291, 2400– 2404. Lassga A. C. and Lu¨ttge A. (2003) A model for crystal dissolution. Eur. J. Mineral. 15, 603–615. Nesbitt H. W., Bancroft G. M., Davidson R., McIntyre N. S. and Pratt A. R. (2004) Minimum XPS core-level line widths of insulators, including silicate minerals. Am. Mineral. 89, 878– 882. Nguyen S. N., Silva R. J., Weed H. C. and Andrews J. E. (1992) Standard Gibbs free-energy of formation at the temperature 303.14 K of 4 uranyl silicates-soddyite, uranophane, sodium boltwoodite and sodium weeksite. J. Chem. Thermo. 24, 359– 376. Noe D. C. and Veblen D. R. (1999) HRTEM analysis of dislocation cores and stacking faults in naturally deformed biotite crystals. Am. Mineral. 84, 1925–1931. Pearcy E. C., Prikryl J. D., Murphy W. M. and Leslie B. W. (1994) Alteration of uraninite from the Nopal I deposit, Pen˜a Blanca district, Chihuahua, Mexico, compared to degradation of spent nuclear fuel in the proposed US high-level nuclear waste repository at Yucca Mountain, Nevada. Appl. Geochem. 9, 713– 732. Perez I., Casas I., Martin M. and Bruno J. (2000) The thermodynamics and kinetics of uranophane dissolution in bicarbonate test solutions. Geochim. Cosmochim. Acta 64, 603–608. Philips B. L., Casey W. H. and Crawford S. N. (1997) Solvent exchange in AlFx(H2O)6x 3x(aq) complexes: ligand-directed labialization of water as an analogue for ligand-induced dissolution of oxide minerals. Geochim. Cosmochim. 61, 3041– 3049. Pouchou J. L. and Pichoir F. (1985) ‘‘PAP” u(rZ) procedure for improved quantitative microanalysis. Microbeam Anal. 10, 104– 106. Prikryl J. D. (2008) Uranophane dissolution and growth in CaCl2– SiO2 (aq) test solutions. Geochim. Cosmochim. 72, 4508–4520. Robinson R. A. and Stokes R. H. (1968) Electrolyte Solutions, the Measurement and Interpretation of Conductance, Chemical Potential, and Diffusion in Solutions of Simple Electrolytes, second ed., rev. Butterworths, London. Schindler M. and Hawthorne F. C. (2001) A bond-valence approach to the structure, chemistry and paragenesis of

2533

hydroxyl-hydrated oxysalt minerals. I. Theory. Can. Mineral. 39, 1225–1242. Schindler M. and Putnis A. (2004) Crystal growth of schoepite on the (104) surface of calcite. Can. Mineral. 42, 1629–1649. Schindler M., Mutter A., Hawthorne F. C. and Putnis A. (2004a) Prediction of crystal morphology of complex uranyl-sheet minerals. I. Theory. Can. Mineral. 42, 1629–1649. Schindler M., Mutter A., Hawthorne F. C. and Putnis A. (2004b) Prediction of crystal morphology of complex uranyl-sheet minerals. II. Observation. Can. Mineral. 42, 1651–1666. Schindler M., Hawthorne F. C., Putnis C. and Putnis A. (2004c) Growth of uranyl-hydroxy-hydrate and uranyl-carbonate minerals on the (104) calcite surface. Can. Mineral. 42, 1683–1697. Schindler M., Mandaliev P., Hawthorne F. C. and Putnis A. (2006a) Dissolution of uranyl-hydroxy-hydrate minerals. I. Curite. Can. Mineral. 44, 415–431. Schindler M., Hawthorne F. C., Burns P. C. and Maurice P. A. (2006b) Dissolution of uranyl-hydroxy-hydrate minerals. II. Becquerelite. Can. Mineral. 44, 1207–1225. Schindler M., Hawthorne F. C., Halden N. M., Burns P. C. and Maurice P. A. (2007a) Dissolution of uranyl-hydroxy-hydrate minerals. III. Billietite. Can. Mineral. 45, 945–962. Schindler M., Hawthorne F. C., Burns P. C. and Maurice P. A. (2007b) Dissolution of uranyl-oxide-hydroxy-hydrate minerals. IV. Fourmarierite and synthetic Pb2(H2O)[(UO2)10UO12(OH)6(H2O)2]. Can. Mineral. 45, 963–981. Schindler, M., Freund, M., Hawthorne, F. C., Burns, P. C. (in this issue). XPS spectra of uranyl minerals and synthetics. I. The U 4f spectrum. Geochim. Cosmochim. Schindler, M., Freund, M., Hawthorne, F. C., Burns, P. C. (in this issue). XPS spectra of uranyl minerals and synthetics. II. The O 1s spectrum. Geochim. Cosmochim. Shirley D. A. (1972) High-resolution X-ray photoemission spectrum of the valence bands of gold. Phys. Rev. B 5, 4709–4714. Sposito G. (1984) The Surface Chemistry of Soils. Oxford University Press. Steinocher V. and Nova´ceˇk R. (1939) On b-uranotile. Am. Mineral. 24, 324–338. Stumm W. (1992) Chemistry of the Solid–Water Interface. John Wiley & Sons, New York, p. 428. Teng H. H. (2004) Controls by saturation state on etch pit formation during calcite dissolution. Geochim. Cosmochim. Acta 68, 253–262. Teng H. H., Fenter P., Cheng L. W. and Sturchio N. C. (2001) Resolving orthoclase dissolution processes with atomic force microscopy and X-ray reflectivity. Geochim. Cosmochim. Acta 65, 3459–3474. Vision 2.2.6. (2006) Kratos Analytical Ltd. Viswanathan K. and Harneit O. (1986) Refined crystal structure of b-uranophane Ca(UO2)2(SiO3OH)25H2O. Am. Mineral. 71, 1489–1493. Wagner C. D., Riggs W. M., Davis L. E., Moulder J. F. and Mailenberg G. M. (1979) Handbook of X-Ray Photoelectron Spectroscopy. Perkin-Elmer. Wronkiewicz D. J., Bates J. K., Gerding T. J., Veleckis E. and Tani B. S. (1992) Uranium release and secondary phase formation during unsaturated testing of UO2 at 90 °C. J. Nucl. Mater. 190, 107–127. Wronkiewicz D. J., Bates J. K., Wolf S. F. and Buck E. C. (1996) Ten year results from unsaturated drip tests with UO2 at 90 °C: implications for the corrosion of spent nuclear fuel. J. Nucl. Mater. 238, 78–95. Associate editor: Jacques Schott