Environ. Sci. Technol. 2004, 38, 6032-6036
Spectroscopic Characterization of the Uranium Carbonate Andersonite Na2Ca[UO2(CO3)3]·6H2O S A M E R A M A Y R I , * ,† T H U R O A R N O L D , † TOBIAS REICH,‡ HARALD FOERSTENDORF,† GERHARD GEIPEL,† GERT BERNHARD,† AND ANDREAS MASSANEK§ Institut fu ¨ r Radiochemie, Forschungszentrum Rossendorf, P.O. Box 510119, 01314 Dresden, Germany, Institut fu ¨r Kernchemie, Johannes Gutenberg-Universita¨t Mainz, Fritz-Strassmann-Weg 2, 55128 Mainz, Germany, and Mineralogische Sammlung, Technische Universita¨t Bergakademie Freiberg, A.-G.-Werner-Bau Brennhausgasse 14, 09596 Freiberg, Germany
The uranium carbonate andersonite Na2Ca[UO2(CO3)3]‚ 6H2O was synthesized and identified with classical analytical and spectroscopic methods. The classical methods applied were powder X-ray diffraction (XRD), nitric acid digestion, and scanning electron microcopy combined with energy-dispersive spectroscopy (SEM/EDS). To characterize andersonite spectroscopically, time-resolved laserinduced fluorescence spectroscopy (TRLFS), X-ray photoelectron spectroscopy (XPS), and Fourier transform infrared spectroscopy (FT-IR) were used. Natural and synthetic andersonite samples were characterized with the nondestructive TRLFS by six fluorescence emission bands at 470.6, 486.1, 505.4, 526.7, 549.6, and 573.9 nm. In addition, andersonite was characterized by FT-IR measurements by the appearance of the asymmetric stretching vibration of the uranyl cation [ν3(UO22+)] at 902 cm-1 with a shoulder at 913 cm-1. XPS measurements verified the composition of the synthetic andersonite sample. The measured intensity ratios of the XPS lines agree with the stoichiometry of Na2Ca[UO2(CO3)3]‚6H2O. The XPS features of the inner valence molecular orbitals are characteristic of the [UO2(CO3)3]4- structural moiety. These spectroscopic methods can be used to identify in a fingerprinting procedure secondary U(VI) phases in mixtures with other phases or as thin coatings on mineral and rock surfaces.
Introduction Secondary uranium phases encountered in the vicinity of uranium contaminated sites have been identified predominantly by methods such as powder X-ray diffraction (XRD), X-ray fluorescence analysis (XFA), scanning electron microscopy/energy-dispersive spectrometry (SEM/EDS), and acid digestion with subsequent chemical analysis. However, many of these secondary uranyl(VI) phases occur in diminutive quantities or in coexistence with other minerals, so their * Correspondingauthorphone: +4913139-25317/307;fax: +496131 39-24510; e-mail:
[email protected]. † Forschungszentrum Rossendorf. ‡ Johannes Gutenberg-Universita ¨ t Mainz. § Technische Universita ¨ t Bergakademie Freiberg. 6032
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004
isolation as pure phase from the natural substrate is very difficult to achieve or even impossible. Spectroscopic techniques [e.g., X-ray photoelectron spectroscopy (XPS), Fourier transform infrared spectroscopy (FT-IR), and time-resolved laser-induced fluorescence spectroscopy (TRLFS)] provide alternatives to the above listed classical methods. For qualitative and quantitative analysis, FT-IR spectroscopy requires only micrograms of sample, the nondestructive TRLFS only a few square micrometers of sample area, and XPS surface layers with a thickness of a few nanometers. The characteristic spectra obtained with these methods are interpreted by matching them with spectra of known uranium minerals in a fingerprinting procedure. In doing so the identification of even minor amounts of secondary uranyl(VI) phases as thin coatings on minerals and rocks is possible (1). In this study, a common secondary uranyl(VI) mineral, sodium calcium uranyl carbonate andersonite [Na2Ca[UO2(CO3)3]‚6H2O], was investigated with IR-spectroscopy and for the first time with XPS and TRLFS. In addition, a natural andersonite sample was analyzed with TRLFS and compared to the TRLFS spectra of the synthesized andersonite. Andersonite belongs to the liebigite group, which is represented by the general formula M2[UO2(CO3)3]‚nH2O, where M stands for Mg, Ca, Na, K, and certain combinations thereof, and n is a number ranging from 5 to 18. Andersonite is found in paragenesis with gypsum, bayleyite, and schrockingerite (2) and has a narrow field of formation in the system Na+-Ca2+-[UO2(CO3)3]4--H2O (3). It may form under oxidizing conditions as a result of continued interactions between groundwater and uranyl oxyhydroxides, which eventually leads to their replacement by uranyl carbonates and uranyl silicates (4). Andersonite is then formed at pH > 7 and a high Na+/Ca2+ concentration ratio (5). Andersonite has been established as one of the end products in alkaline-carbonate environments, such as in the Hillside Mine, Yavapai County, Arizona (2). It has also been found at Myrthengraben, Semmering (Lower Austria) (6); in Stripa, Va¨stmanland, Sweden (7); and in Jachymov, Czech Republic (8). Liebigite and andersonite may coexist on the basis of the stability diagrams of Alwan and Williams (5). Andersonite is a yellow mineral, but it will turn to very pale yellow upon dehydration. It shows intense blue-green fluorescence. Uranyl(VI) carbonates are an important group of minerals since they seem to play a role in uranium geochemical migration and in uranium fixation in limestone (9). It is wellknown that at pH < 9.5 and > 6.5 in carbonate-rich subsurface waters (higher partial pressure of CO2), uranyl ions are transported as uranyl-carbonates, which are the dominant aquatic uranyl(VI) species in natural waters. Recent studies of seepage waters of a mine tailing pile and of mine waters of a uranium mine in Saxony, Germany, led to the discovery of the important aqueous uranyl(VI) species [Ca2UO2(CO3)3](aq) (10, 11). Together with adsorption reactions, the solubility of secondary uranium(VI) phases is likely to play an important role in controlling the source term for the uranium migration (12, 13) through geological materials (e.g., rocks, unconsolidated tailings, and soils). In this context, the identification of even minor amounts or thin coatings of secondary uranyl(VI) phases will contribute to a better understanding of the uranium chemistry in natural environments, leading eventually to advanced restoration strategies for uraniumcontaminated sites and may be also to improved risk assessments. 10.1021/es049417l CCC: $27.50
2004 American Chemical Society Published on Web 10/12/2004
The aim of this study is the characterization of andersonite, Na2Ca[UO2(CO3)3]‚6H2O, with the spectroscopic methods TRLFS, FT-IR, and XPS. In particular, the characterization of the fluorescence properties of andersonite is the focus of this paper.
Experimental Section Synthesis of Andersonite. Andersonite was synthesized by slowly adding under constant stirring 10 mL of uranyl nitrate solution (2.0 mol/L, p.A., Merck) to 100 mL of potassium carbonate solution (0.6 mol/L, p.a., Merck), followed by a stoichiometric addition of 10 mL calcium nitrate solution (2.0 mol/L, p.a., Merck) and 10 mL sodium nitrate solution (4.0 mol/L, p.a., Merck). The pH of the prepared solution was 6.45. Then the pH was adjusted to 8.0 by addition of diluted potassium carbonate solution (0.1 mol/L). The solution was allowed to crystallize by evaporation at room temperature for a period of 3 weeks. The precipitated cubelike yellow crystals were detached from the wall and bottom of the beaker, subsequently washed by decantation with MilliQwater, and allowed to air-dry. The formation of andersonite can be explained by a twostage process. First, the aquatic [UO2(CO3)3]4- complex is formed. Subsequently, sufficient amounts of Ca2+ and Na+ ions were added, which then initiated the precipitation of andersonite. Following the above-described method, andersonite crystals with a high-phase purity were synthesized. Characterization. Chemical Analysis (Nitric Acid Digestion with Subsequent Analysis). The air-dried synthetic compound was dissolved in 10% M HNO3 (Suprapur, Merck). The uranium content was determined using an inductively coupled plasma (Ar-plasma) mass spectrometer (Elan 500, Perkin-Elmer, U ¨ berlingen, Germany). The concentration of Na and Ca was determined by flame-atomic absorption spectrometry (AAS 4100, Perkin-Elmer, U ¨ berlingen, Germany). Powder X-ray Diffraction. The powder X-ray diffraction diagram of the synthesized andersonite was recorded with a Universal-Ro¨ntgen-Diffraktometer (URD 6, Freiberger Pra¨zisionsmechanik, Freiberg, Germany) using Cu KR, (λ ) 0.1542 nm) at 40 kV and 30 mA. Silicon powder was used as an external standard. The diffractometer was operated in Bragg-Brentano geometry in step mode with a step width of 0.05°. The aperture was 0.2 nm, a Soller slit of 20 mrad was used. The X-ray diffraction pattern was recorded in the 2θ range from 5° to 60°. The Diffracplus Evaluation program, version 2.2 (Siemens AG), was used for data processing, reflection indexing, and refining. The d-lattice spacing and the lattice constants (a, b, c) were calculated after peak fitting of a measurement with a silicon standard using the program Win-Metric, version 3.04 (SIGMAC GmbH). Scanning Electron Microscopy/Energy-Dispersive Spectrometry. The synthesized andersonite samples were investigated by scanning electron microscopy (SEM, model Zeiss DSM 962). The andersonite crystals were crushed in a mortar prior to SEM sample preparation. The powder specimen was then fixed on a SEM sample holder and either gold-coated for microscopic characterization in secondary electron detection mode (SE) or sputtered with carbon for EDS analysis. The applied acceleration voltage was in the range of 15-30 keV. Time-Resolved Laser-Induced Fluorescence Spectroscopy. A Nd:YAG laser system (model GCR 190, Spectra Physics) was used to study the solid samples of andersonite. A detailed description of the experimental setup can be found in ref 14. An excitation wavelength of 266 nm was used. The spectra were recorded in the range from 450 to 600 nm with delay times from 0.1 to 100 µs after the application of the laser pulse. The gate time was 0.2 µs. The average laser energy was 2 mJ. For every delay time, the fluorescence signal was
averaged by sampling three single spectra over 100 laser shots. All functions (time controlling, device settings, reading the spectra, data storage) of the spectrometer were computercontrolled. The computer software Grams/386TM (Galactic Ind. Corp.) was used for deconvolution of the spectra. The time dependencies of the spectra were calculated with the Origin 6.1G Client (Microcal Software Inc.) and Excel 2000 (Microsoft Software Inc.) programs. The sample was measured at ambient temperature and at a relative humidity of 60 ( 2%. Fourier Transform Infrared Spectroscopy. Diffuse reflectance infrared Fourier transform (DRIFT) measurements were carried out using a Perkin-Elmer GX-2000 instrument equipped with a MCT (mercury-cadmium telluride) detector. The spectral resolution was 4 cm-1 in the frequency range from 4000 to 600 cm-1. The synthetic powder (0.7 mg) was mixed with solid KBr (300 mg). X-ray Photoelectron Spectroscopy. The andersonite sample was prepared for X-ray photoelectron spectroscopy measurement from finely dispersed powders milled in an agate mortar. A small amount of this powder was pressed into an indium foil and transferred into the spectrometer. The same procedure was used to prepare a CaCO3 (p.a., Merck) sample, which was studied by XPS for comparison. The photoelectron spectra were measured at room temperature under a vacuum of 7 × 10-9 mbar using a custom-built XPS system (SPECS GmbH, Berlin, Germany). The photoelectron spectra were excited using the nonmonochromatic KR radiation from a high-intensity twin anode (Al/Mg) X-ray source XR-50. Due to interference of the Na and O KLL Auger lines with photoelectron lines, the photoelectron spectra were recorded both with Al KR (1486.6 eV) and Mg KR (1253.6 eV) radiation. The spectra were recorded with a constant analyzer pass energy of 13 eV using the hemispherical energy analyzer PHOIBOS 100. The spectrometer resolution measured as the full width at halfmaximum of the Ag 3d5/2 line was 1.0 eV. The electrostatic sample charging was corrected for by setting the binding energy (Eb) of C 1s electrons of adventitious carbon on the sample surface equal to 285.0 eV. The error of the determined binding energies is (0.1 eV. The intensity of the photoelectron lines was determined using the software SpecsLab 1. The error in the relative line intensities was less than 10%.
Results and Discussion Chemical Composition. Andersonite, Na2Ca[UO2(CO3)3]‚ 6H2O, contains theoretically one uranium atom, two sodium cations, one calcium cation, three carbonate groups, and six water molecules. The results of the nitric acid digestion of the synthesized andersonite and its subsequent chemical analysis with ICP-MS and AAS show that it is composed of 37.05 ( 0.27 wt % U, 6.33 ( 0.40 wt % Ca, and 7.04 ( 0.20 wt % Na. Within the analytical errors, these measured valued are in good agreement with theoretical values of 36.95 wt % U, 6.22 wt % Ca, and 7.14 wt % Na. The analytical results clearly indicate that the synthesis of andersonite as pure phase was successful. Powder X-ray Diffraction. Andersonite has trigonal/ rhombohedral symmetry and systematic extinction according to space group R-3mH (6, 15, 16). The XRD pattern (not shown) of synthetic andersonite shows that the reflections are in good agreement with the corresponding PDF file of Mereiter (15) and Coda et al. (16). The lattice parameters of the synthesized andersonite sample were calculated from the measured and clearly identifiable reflections. The values obtained, a ) b ) 17.889 ( 0.007 nm, c ) 23.739 ( 0.005 nm, V ) 6585.396 ( 4.996 nm3, are in accordance with those of Mereiter (15) and Coda et al. (16). VOL. 38, NO. 22, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6033
FIGURE 2. FT-IR (DRIFT) spectrum of synthetic andersonite.
FIGURE 1. Fluorescence emission bands of synthetic andersonite (A). (The dotted lines show the curve fit of the spectra (Gaussian shapes).) Fluorescence spectra of synthetic andersonite as a function of delay time (B). To study the mineral stability of andersonite, the sample was stored in air at normal humidity (60%) for 1 year and measured again with XRD. The same X-ray diffraction pattern as before was recorded, indicating that the sample was well preserved under the specified storage conditions. Scanning Electron Microscopy/Energy-Dispersive Spectrometry. SEM investigations of synthetic andersonite showed that the crystal exhibits pseudocubic morphologies, which is in accordance with observations made by Axelrod et al. (2) and Vochten et al. (3). The major elements of andersonite, i.e., U, Ca, and Na, were detected by EDS in the expected ranges and imply that a pure andersonite sample without contamination was synthesized. Time-Resolved Laser-Induced Fluorescence Spectroscopy. So far, detailed spectral information concerning uranium minerals are rare and, if available, are usually incomplete (17) However, recently Geipel et al. (1) reported TRLFS investigations of some uranium phosphates and uranium arsenates. Amayri et al. (18-20) applied TRLFS to characterize several uranium carbonates, e.g., bayleyite, liebigite, swartzite. It is known that the free uranyl ions show good fluorescence spectra with six characteristic fluorescence emission bands and a characteristic lifetime of 1.80 ( 0.8 µs (1). It is also well-known that the uranyl triscarbonato ion, UO2(CO3)34-, shows no fluorescence at 25 °C at all (10, 21). However, in the case where alkaline earth metals, such as Ca, are constituents of the solid or aqueous uranium(VI) carbonato species, an intensive fluorescence signal is obtained (10, 11, 18, 20, 21). In general, TRLFS investigations on solid uranium(VI) phases yield characteristic information concerning the fluorescence emission bands as well as the fluorescence lifetime. In this context it has to be pointed out that the fluorescence emission bands are the primary features of TRLFS spectra and the fluorescence lifetime a secondary one. The fluorescence lifetime may vary, depending on the number of neighboring water molecules surrounding the uranium(VI) atom. Such characteristic spectral information is useful for identifying fluorescent uranium minerals in a “fingerprint” procedure (1). The obtained fluorescence data may also provide a link to solution spectra of known and so far unknown aqueous uranium(VI) species in uranium6034
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004
contaminated environmental waters (18). Figure 1A shows the detected fluorescence emissions bands of synthetic andersonite. The deconvoluted fluorescence spectrum reveals six characteristic intensive fluorescence emission bands at 470.6, 486.1, 505.4, 526.7, 549.6, and 573.9 nm. The fluorescence emission bands were determined with an error of (1 nm. The TRLFS spectra displayed in Figure 1B show that the fluorescence intensity of the above listed emission bands decreases with time. A characteristic fluorescence lifetime for synthetic andersonite of 65.2 ( 0.6 µs was calculated from these TRLFS spectra. To compare the obtained fluorescence data of synthetic andersonite with a natural andersonite sample, we conducted an additional TRLFS study. A natural andersonite sample as a thin coating on a rock surface was obtained from the Mineralogical Collection of the Technical University Mining Academy Freiberg as specimen number 45540 (Honeybee No. 2 Mine, Cane Springs Canyon, San Juan Co., Utah). Again six intensive fluorescence emission bands for the natural sample at 468.4, 485.2, 504.8, 526.2, 549.6, and 575.4 nm were detected, which are in excellent agreement with the above listed fluorescence emission bands of the synthesized andersonite sample. The respective lifetime of 33.1 ( 3.9 µs, which was calculated from the TRLFS spectra of the natural andersonite sample, was slightly smaller than the one obtained for the synthesized sample. This small discrepancy in lifetime between the synthesized and the natural andersonite samples is attributed to a different number of water molecules around the uranium(VI) atoms or to a minor contamination in the natural andersonite sample. For example, minor amounts of Fe decrease the fluorescence lifetime (22). The results shown above clearly indicate that TRLFS is a very useful tool for identifying in a nondestructive way secondary uranium minerals even as a thin coating on rock surfaces. Fourier Transform Infrared Spectroscopy. The DRIFT spectrum of andersonite is shown in Figure 2. An excellent agreement with the spectrum measured in transmission (KBr pellet) was observed and is, therefore, not explicitly shown. Additionally, the DRIFT spectrum is in very good agreement with the spectrum of natural and synthetic andersonite published earlier (8, 23). This verifies again the high efficiency of the synthesis procedure introduced by this work. The bands at 3408 and 3547 cm-1 represent OH stretching vibrations, which can be mainly assigned to water molecules of the mineral phase. The sharp band at 3547 cm-1 possibly indicates the presence of a weakly bound species of water molecules in the crystal structure. The asymmetric stretching vibration of the uranyl cation [ν3 UO22+] appears at 902 cm-1,
TABLE 1. Comparison of Infrared Spectra of Synthetic (This Work) and Natural Andersonite from Refs 8 and 23a wavenumber (cm-1) natural andersonite synth. andersonite DRIFT (this work) 3547 3408 1571 1526 1383 1092 1080 913 sh 902 851 sh 847 832 800 727 700
from ref 23 3510 1580 1535 1385 1080 913 sh 901 848 727 697
from ref 8
tentative assignment
3563 3430 1578 1522 1378 1091 1081 915 sh 903 854 849
ν(OH) ν(OH) ν3(CO32-) ν3(CO32-) ν3(CO32-) ν1(CO32-) ν1(CO32-) ν3(UO22+) ν2(CO32-) ν1(UO22+) (?)
795 728 701
ν4(CO32-) ν4(CO32-)
FIGURE 3. X-ray photoelectron spectrum of U 4f electrons of andersonite, Na2Ca[UO2(CO3)3]‚6H2O, after satellite subtraction due to nonmonochromatic Mg Kr excitation.
a sh ) shoulder. Tentative assignments were adopted from C ˇ ejka (24).
TABLE 2. Electron Binding Energies (Mg Kr Excitation) in eV of Andersonite, Na2Ca[UO2(CO3)3]‚6H2O, and Comparison with Literature Values for K4[UO2(CO3)3] (25)a
b
line
this work
ref 25
line
this work
Na 1s U 4d5/2 U 4d3/2 O 1s U 4f7/2 U 4f5/2
1071.6 740.5b 783.0b 531.5 382.1 392.9
740.6 782.7 531.4 382.1 392.9
Ca 2p3/2 Ca 2p1/2 C 1s U 5d5/2 U 5d3/2 Na 2s
347.2 350.8 289.5 98.1 106.5 63.1
ref 25
98.1 106.5
a Binding energy C 1s of adventitious carbon equals 285.0 eV. Determined with Al KR excitation.
showing a shoulder at 913 cm-1. This is in accordance with spectra of other [UO2(CO3)3]4- minerals containing divalent metal cations and water (19, 24). The symmetric stretching vibration of the uranyl cation [ν1(UO22+)] is normally inactive in infrared spectroscopy. With regard to the crystal structure of andersonite (16), a distortion of this mode can be expected, which potentially makes this mode apparent with low intensity in infrared spectra. The very weak band at 832 cm-1 is possibly a good candidate for representing this mode. However, an exact assignment cannot be given since coincidental overlapping bands representing δ UOH vibrations cannot be ruled out. The most significant bands of the spectrum represent the carbonate anions bidentately bound to the uranyl cation. The bands at 1571 and 1383 cm-1 can be assigned to the doubly degenerated ν3 mode of the CO32- anion showing C2v symmetry. The strong splitting of this mode (∆ν ≈ 190 cm-1) is characteristic for the bidentate binding of the anion to the UO22+ ion. For symmetry reasons, a splitting of the ν4 mode is also expected, and the bands at 727 and 700 cm-1 have to be assigned to this mode. The other vibrational modes (ν1 and ν2) of the carbonates are showing up at 1080 and 847 cm-1, respectively (see Table 1). The refinement of the crystal structure shows two different species of CO32- present in andersonite (16). In the infrared spectrum these structurally nonequivalent functional groups are obviously reflected by the appearance of additional bands and shoulders for the vibrational modes (ν1-ν3) of the CO32- anion as they are given in Table 1. X-ray Photoelectron Spectroscopy. Table 2 shows the measured electron binding energies of the elements U, Na, Ca, C, and O present in andersonite. As can be seen, the
FIGURE 4. X-ray photoelectron spectrum of valence electrons of andersonite, Na2Ca[UO2(CO3)3]‚6H2O, and CaCO3 after satellite subtraction due to nonmonochromatic Mg Kr excitation. measured values agree with the corresponding binding energy values for K4[UO2(CO3)3] (25). Within the experimental uncertainties, the measured C 1s binding energy of 289.5 eV is identical to the corresponding value of 289.7 eV for Na2CO3 (26). The measured binding energies are indicative of a compound consisting of hexavalent uranium, carbonate, Na+, and Ca2+. Figure 3 shows the XPS spectrum in the region of the U 4f electrons. Intense shake-up satellite lines at the higher binding energy side accompany the U 4f7/2 and U 4f5/2 main lines. The satellites have a 3.4 eV higher binding energy than the corresponding 4f electrons. The relative intensity of these satellite lines is approximately 16% of the corresponding 4f line. Similar parameters have been observed for other uranyl compounds (25). Therefore, the chemical shift of the U 4f lines and their satellite structure can be used to identify the oxidation state of uranium. Another clear identification of the presence of hexavalent uranium is the absence of a narrow and intense U 5f line at approximately 4 eV binding energy (Figure 4). The outer valence electrons, the Ca 3p, and Na 2p electrons centered at 5, 23, and 30 eV binding energy, respectively, dominate the XPS spectrum shown in Figure 4. The two small peaks centered at 10 and 12 eV binding energy originate from the molecular orbital of the CO32- group, as can be seen by comparing the andersonite spectrum with that of CaCO3 (see Figure 4). The same features were also observed in the XPS VOL. 38, NO. 22, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6035
TABLE 3. Relative Intensities of the U 4f, Ca 2p, Na 2s, C 1s, and O 1s Linesa XPS lines
experiment
sensitivity factor
atomic ratio
U 4f/Ca 2p U 4f/Na 2s U 4f/C 1s U 4f/O 1s
8.6 59.3 16.0 1.5
9.7 115.0 48.8 16.6
U:Ca ) 1:1.1 U:Na ) 1:1.9 U:C ) 1:3.0 U:O ) 1:11.1
a The spectra were recorded with Al KR excitation. The atomic ratios were calculated from the relative line intensities and corresponding sensitivity factors. These factors were derived from theoretical photoionization cross sections (28).
spectra of the Th(IV) compounds with ligands of D3h symmetry, i.e., CO32- and NO3- (25). The weak structure in the region of 15-20 eV is due to the electrons of the 2a2u and 2e1u IVMO (inner valence molecular orbital) connected with U 6p3/2 electrons. The energy difference ∆E between the valence electron levels that relate to the 2a2u and 2e1u IVMO levels in the [UO22+O6]10- cluster with the point symmetry D6h is 4.6 eV (Figure 4). Using the relationship between ∆E and the bond distances between uranium and its axial and equatorial oxygen atoms, r1 ) (U-Oax) and r2 ) (U-Oeq) (27), the calculated bond distances are r1 ) 1.71 and r2 ) 2.42 Å. These results can be compared to the values of r1 ) 1.77 and r2 ) 2.44 Å determined by single-crystal X-ray diffraction (15). The intensity of the measured XPS lines was used to verify the composition of the synthetic andersonite sample. Table 3 summarizes the measured intensity ratios and the calculated atomic ratios. Except for oxygen, all atomic ratios agree with the stoichiometry of Na2Ca[UO2(CO3)3]‚6H2O. The somewhat lower content of oxygen atoms is due to the loss of hydrated water in the vacuum chamber of the XPS spectrometer. In summary, the results of both the qualitative and quantitative XPS analysis confirmed the formation of andersonite. The spectral features of the U 4f electrons and the IVMO 2a2u and 2e1u electrons can be used to identify the [UO2(CO3)3]4- structural moiety. The presented spectroscopic methods, i.e., TRLFS, XPS, and FT-IR spectroscopy, are suitable methods for identifying in a fingerprinting procedure secondary U(VI) phases in mixtures with other phases or as thin coatings on mineral and rock surfaces. The obtained results are also of fundamental significance for uranyl(VI) adsorption studies, since the queried spectroscopic information, obtained for wellcharacterized uranium minerals, can be used to identify the coordination environment and structure of unknown adsorbed uranium(VI) surface species on minerals.
Muschter for conducting FT-IR measurements, and E. Cristalle for SEM/EDS investigations.
Literature Cited (1) Geipel, G.; Bernhard, G.; Rutsch, M.; Brendler, V.; Nitsche, H. Radiochim. Acta 2000, 88, 757. (2) Axelrod, J. M.; Grimaldi, F. S.; Milton, C.; Murata, K. J. Am. Mineral. 1951, 36, 1. (3) Vochten, R.; Van Haverbeke, L.; Van Springel, K. Can. Mineral. 1993, 31, 167. (4) Finch, R.; Murakami, T. UraniumsMineralogy, Geochemistry and the Environment. Review of Mineralogy; Burns, P. C., Finch, R., Eds.; The Mineralogical Society of America: Washington DC, 1999; Vol. 38, p 91. (5) Alwan, A. K.; Williams, P. A. Mineral. Mag. 1980, 43, 665. (6) Tufar, W. N. Jb. Mineral. Abh. 1967, 106, 191. (7) Wellin, E. Arkiv. Mineral. Geol. 1958, 2 (27), 373. (8) C ˇ ejka, J.; Urbanec, Z. J. Thermal Anal. 1988, 33, 389. (9) C ˇ ejka, J. Sb.; Vys, Sˇ k. Chem.sTechnol. Praze, Mineral. 1965, 7, 75. (10) Bernhard, G.; Geipel, G.; Brendler, V.; Nitsche, H. Radiochim. Acta 1996, 74, 87. (11) Bernhard, G.; Geipel, G.; Brendler, V.; Nitsche, H. J. Alloys Compd. 1998, 271, 201. (12) Murphy, W. M. Mater. Res. Soc. Symp. Proc. 1997, 465, 713. (13) Sandino, M. C.; Grambow, B. Radiochim. Acta 1994, 66/67, 37. (14) Geipel, G.; Brachmann, A.; Brendler, V.; Bernhard, G.; Nitsche, H. Radiochim. Acta 1996, 75 (4), 199. (15) Mereiter, K. Anz. Oesterr. Akad. Wiss. Math. Naturwiss. 1986, 3, 39. (16) Coda, A.; Della Giusta, A.; Tazzoli, V. Acta Crystallogr. 1981, B37, 1496. (17) Robbins, M. The Collector’s Book of Fluorescent Minerals; Van Nostrand Reinhold Company: New York, 1983; p 228. (18) Amayri, S. Ph.D. Thesis, TU Dresden, 2002, Dresden, Germany; FZR Science Technical Report, 2002, FZR-359, ISSN 1437-322X. (19) Amayri, S.; Arnold, T.; Foerstendorf, H.; Geipel, G.; Bernhard, G. Can. Mineral. 2004, 42 (4), 963. (20) Amayri, S.; Reich, T.; Arnold, T.; Geipel, G.; Bernhard, G. J. Solid State Chem. (in press). (21) Bernhard, G.; Geipel, G.; Brendler, V.; Reich, T.; Amayri, S.; Nitsche, H. Radiochim. Acta 2001, 89, 511. (22) Hoffman, M. Z.; Bolletta, F.; Moggi, L.; Hug, G. L. J. Phys. Chem. Ref. Data 1989, 18, 219. (23) Huang, C. K.; Kerr, P. F. Am. Mineral. 1960, 45, 311-324. (24) C ˇ ejka, J. Mineralogy, Geochemistry and the Environment. Review of Mineralogy; Burns, P. C., Finch, R., Eds.; The Mineralogical Society of America, Washington, DC, 1999; Vol. 38, p 521. (25) Teterin, Yu. A.; Baev, A. S. TsNIIatominform 1986, 192 (in Russian). (26) Morgan, W. E.; Van Wazer, J. R. J. Chem. Phys. 1973, 77 (7), 964. (27) Nefedov, V. I.; Teterin, Yu. A.; Reich, T.; Nitsche, H. Dokl. Akad. Nauk 1996, 348, 634 (in Russian). (28) Band, I. M.; Kharitonov, Yu. I.; Trzhaskovskaya, M. B. At. Data Nucl. Data Tables 1979, 23, 443.
Acknowledgments
Received for review April 19, 2004. Revised manuscript received August 4, 2004. Accepted August 30, 2004.
The authors thank J. Drebert for XPS analyses, U. Schaefer for ICP-MS and AAS analyses, A. Scholz for XRD analyses, K.
ES049417L
6036
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004
