Cs+ Speciation on Soil Particles by TOF-SIMS Imaging - Analytical

Morphologic and spectral investigation of exceptionally well-preserved bacterial biofilms from the Oligocene Enspel formation, Germany. J.K.W Toporski...
0 downloads 0 Views 101KB Size
Anal. Chem. 1998, 70, 534-539

Cs+ Speciation on Soil Particles by TOF-SIMS Imaging Gary S. Groenewold,* Jani C. Ingram, Travis McLing, and Anita K. Gianotto

Idaho National Engineering and Environmental Laboratory, Idaho Falls, Idaho 83415-2208 Recep Avci

Image and Chemical Analysis Laboratory, EPS 259, Montana State University, Bozeman, Montana 59717

Soil particles exposed to CsI solutions were analyzed by imaging time-of-flight secondary ion mass spectrometry and also by scanning electron microscopy/energy-dispersive X-ray spectroscopy (SEM/EDS). The results showed that Cs+ could be detected and imaged on the surface of the soil particles readily at concentrations down to 160 ppm, which corresponds to 0.04 monolayer. Imaging revealed that most of the soil surface consisted of aluminosilicate material. However, some of the surface was more quartzic in composition, primarily SiO2 with little Al. It was observed that adsorbed Cs+ was associated with the presence of Al on the surface of the soil particles. In contrast, in high SiO2 areas of the soil particle where little Al was observed, little adsorbed Cs+ was observed on the surface of the soil particle. Using EDS, Cs+ was observed only in the most concentrated Cs+-soil system, and Cs+ was clearly correlated with the presence of Al and I. These results are interpreted in terms of multiple layers of CsI forming over areas of the soil surface that contain substantial Al. These observations are consistent with the hypothesis that the insertion of Al into the SiO2 lattice results in the formation of anionic sites, which are then capable of binding cations. The characterization of natural or anthropomorphic chemicals on mineral surfaces is an important topic, because the surface chemistry exerts a large influence on chemical mobility and eventual fate. The surface chemistry can be highly variable on a microscopic scale, and this inhomogeneity can confound characterization efforts. Highly irregular morphological features and mineral phases that are less than 1 µm across can defeat surface characterization techniques like reflectance infrared spectroscopy, and X-ray photoelectron spectroscopy. Consequently, spectroscopic imaging investigations of contaminants on natural mineral surfaces have been few. Cesium contamination on soil is one system for which spectroscopic information would be of high interest.1,2 The 134 and 137 isotopes decay by γ emission and are formed in high (1) Evans, D. W.; Alberts, J. J.; Clark, R. A. Geochim. Cosmochim. Acta 1983, 47, 1041-9. (2) Heong, C.-H.; Kim, C.-S.; Kim, S.-J., Park, S.-W. J. Environ. Sci. Health 1996, A31, 2173-92.

534 Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

fission yield; the 137 isotope has a moderately long half-life (30 years). The Cs isotopes comprise one of the lasting health problems from the Chernobyl accident.3,4 Cs can be highly mobile in some environments, and geochemically it has many of the same characteristics as potassium as a consequence of similar ionic radii in solution.2 Hence, there is motivation for understanding the interaction of Cs+ with naturally occurring mineral surfaces at the molecular level. Cs sorption has been extensively investigated, primarily by using sequential extractions together with γ spectroscopy (for radioisotopes) or atomic absorption for detection.1,3-6 This approach has been applied to the study of Cs contamination on basalts and smectites,7 clays (illite, kaolinite),6,8 and soils.9 Cs was shown to prefer the mineral soil horizons in high-organic soils.10 From these studies, it has been possible to infer mechanistic details: Cs will tenaciously adhere to adsorption sites and can be supplanted only by K+ and NH4+. It appears to prefer surface “defects”, which have been termed frayed edge, and wedge sites.11-13 However, understanding of Cs-soil systems would benefit from direct spectroscopic information. The spectroscopic approach in this study utilized an imaging SIMS instrument14 for characterization of soil particles that had been exposed to Cs+. SIMS is well suited to the analysis of Cs+ because it readily forms gas-phase secondary ions. The SIMS instrument utilizes microfocused primary ion guns, achieving spatial resolutions of less than 1 µm. The ion optics transmit the secondary ions through three electrostatic sectors to a channelplate detector, such that the spatial information is preserved. The (3) Fawaris, B. H.; Johanson, K. J. Sci. Total Environ. 1995, 170, 221-8. (4) Carbol, P.; Skarnemark, G.; Skalberg, M. Sci. Total Environ. 1993, 130/ 131, 129-37. (5) Williams, T. M. Environ. Geol. 1993, 21, 62-9. (6) Von Gunten, H. R.; Benes, P. Radiochim. Acta 1995, 69, 1-29. (7) Ames, L. L.; McGarrah, J. E.; Walker, B. A.; Salter, P. F. Chem. Geol. 1982, 35, 205-225. (8) Desmet, G. M.; Van Loon, L. R.; Howard, B. J. Sci. Total Environ. 1991, 100, 105-24. (9) Essington, E. H.; Fowler, E. B.; Polzer, W. L. Soil Sci. 1981, 132, 13-8. (10) Bunzl, K.; Schimmack, W. Chemosphere 1989, 18, 2109-20. (11) Wauters, J.; Vidal, M.; Elsen, A.; Cremers, A. Appl. Geochem. 1996, 11, 595-9. (12) Vidal, M.; Roig, M.; Rigol, A.; Llaurado, M.; Rauret, G.; Wauters, J.; Elsen, A.; Cremers, A. Analyst 1995, 120, 1785-91. (13) Thiry, Y.; Myttenaere, C. J. Environ. Radioact. 1993, 18, 247-57. (14) Schueler, B.; Sander, P.; Reed, D. A. Vacuum 1990, 41, 1661-4. S0003-2700(97)00517-9 CCC: $15.00

© 1998 American Chemical Society Published on Web 02/01/1998

optics are robust to the extent that images of samples having inhomogeneous morphology can be fairly easily acquired.15 Timeof-flight (TOF) mass spectrometers are now capable of high mass resolution (m/∆m ) 10 000 on flat surfaces). The present study shows that imaging TOF-secondary ion mass spectrometry (SIMS) is very readily applied to the imaging characterization of Cs+ on soil particles and demonstrates a strong correlation between Cs+ adsorption and areas of the soil surface where Al is prevalent. In contrast, it is shown that portions of the surface that are dominated by Si have a much lower affinity for Cs+. These conclusions are strongly supported by energydispersive X-ray spectroscopy (EDS) analyses. EXPERIMENTAL SECTION Characterization of Unexposed Soil Samples. The soil sample chosen for the present study was collected from a formerly used defense site at Raritan, NJ. This soil was selected because it had been investigated as a substrate for previous surface contamination studies; hence, benchmark information existed for this sample.16 The soil was sieved, and the largest soil fraction that passed the 0.0049-in. (0.012-cm) mesh but not the 0.0098-in. (0.025-cm) mesh was selected for study. The surface area of this fraction of the soil was measured at 2.2 m2/g, using a Micromeritics Flowsorb II 2300 instrument (determination based on the Brunauer, Emmet, and Teller (BET) N2 adsorption method17). Cation-exchange capacity was measured by acid washing the soil sample and then saturating the surface with Na+.18 The soil surface was then washed with distilled deionized water. Finally, the Na+ was displaced with NH4+, and the Na+ concentration was measured at 2.5 mequiv/100 g. The soil samples were also characterized by X-ray diffraction using a Phillips Electronic Corp. PW1710 instrument. The copper anode was operated at 20 mA, 40 keV. The D spacing was calculated using the Cu KR1 wavelength of 1.540 60 Å. This analysis showed that the soil samples were predominantly SiO2, with minor amounts of aluminosilicate: there was a high level of agreement between the diffraction pattern of the soil particles and that of benchmark SiO2. Exposure of Soil Samples. The soil samples were exposed to CsI/H2O solutions by adding 100 µL to a 50-mg soil sample and then allowing the solvent water to evaporate. Soil concentrations corresponding to 13 000, 1500, and 160 ppm were generated in this fashion. Particles from these samples were then analyzed using TOF-SIMS and scanning electron microscopy (SEM)/EDS (see below). Subsequently, a fraction of the exposed soil samples were washed three times by covering them with distilled, deionized water (18 MΩ), decanting, and then allowed to dry again. Particles from the washed samples were then reanalyzed. Time-of-Flight Secondary Ion Mass Spectrometry Analysis. TOF-SIMS analyses were performed using a Charles Evans (15) Brinen, J. S.; Reich, F. Surf. Interface Anal. 1992, 18, 448-52. (16) (a) Groenewold, G. S.; Ingram, J. S.; Appelhans, A. D.; Delmore, J. E.; Dahl, D. A. Environ. Sci. Technol. 1995, 29, 2107-11. (b) Groenewold, G. S.; Ingram, J. S.; Delmore, J. E.; Appelhans, A. D.; Dahl, D. A. J. Hazard Mater. 1995, 41, 359-70. (17) Adamson, A. W. Physical Chemistry of Surfaces; John Wiley & Sons: New York, 1990; p 609. (18) Chapman, H. D. In Methods of Soil Analysis. Part 2: Chemical and Microbiological Properties Black, C. A., Ed.; American Society of Agronomy: Madison, WI, 1965.

& Associates TOF-SIMS instrument,14,19 located at the Image and Chemical Analysis Laboratory at Montana State University. In the present study, soil particles were pressed into indium foil and mounted in the sample holder of the instrument. The sample was then admitted to the source and analyzed at a base pressure of ∼5 × 10-9 Torr. When the particle was analyzed, an 80 × 80 µm area was scatter rastered using the primary ion beam in a pulsed fashion. The scatter raster samples the target surface in a random fashion and is used because it is easier to mitigate charge buildup compared to pattern rastering, in which neighboring regions are sequentially interrogated. The temporal width of a single pulse was 16.28 ns, and the repetition rate was 10 kHz. The primary ion was a microfocused Ga+ gun which operated at 600 pA dc, at + 15 keV relative to ground. The loaded target stage biased at +3 keV for analysis; thus, primary ion impact energy was 12 keV. This information was used to calculate a flux density of 9.5 × 109 ions/(s cm2). Particles were typically analyzed for 5 min, and hence, the total dose imparted to the samples was ∼2.8 × 1012 ions/cm2. The instrument is equipped with a secondary ion immersion lens, which will extract and transmit ions having a wide range of kinetic energy from the sample region to the TOF. Hence, the instrument is well suited to the task of analyzing samples having irregular morphology, which can lead to variable kinetic energy. The primary ion beam was operated in an unbunched mode for the acquisition of the imaged data. The compromise of this acquisition mode is high spatial resolution (1 µm or less) and more modest mass resolution (m/∆m). The mass resolution is also compromised to some extent by the fact that the surfaces were not flat. Resolution increases with increasing mass: m/z 28 (Si+), 600; m/z 56 (Fe+), 800; m/z 133 (Cs+), 1200; and m/z 301 (Cs2Cl+), 9700. In the lower mass regions, the mass resolution was sufficient to distinguish inorganic from organic ions. Scanning Electron Microscopy/Energy-Dispersive X-ray Spectroscopy. The mineral compositions of the soil particles were investigated using SEM and EDS. The particles were not polished prior to SEM and EDS (or SIMS) analysis, because an analysis of the undisturbed particles was desired. For this reason, the error in the EDS data may be augmented because the surfaces analyzed were not flat. SEM images were obtained using an Amray model 1830 instrument, which was operated with a 20keV electron beam. EDS analyses were performed using a FisonsKevex Delta 5 instrument. As in the case of the SEM analyses, 20-keV incident electrons were used. The spatial resolution for this analysis is approximately 1-2 µm, which corresponds to the diameter of the approximate volume excited by the incident electron beam. The EDS analyses were standardless, and ZAF (Z #, absorption, fluorescence) corrections were accomplished by using the extended PHIRHOZ (XPP, Quantex+ version 6 software, Fisons Instrument Manufacturers, Inc. (Valencia, CA)) quantification routine. For elements that are >5 atom %, the accuracy of this method is on the order of 4-8% (relative) when flat, polished specimens were analyzed. Accuracy for unpolished samples is on the order of 2-12% (relative), using conventional stands.20 (19) Fister, T. F.; Strossman, G. S.; Willett, K. L.; Odom, R. W.; Linton, R. W. Int. J. Mass Spectrom. Ion Processes 1995 14, 387. (20) Small, J. A.; Armstrong, J. T. In Proceedings Microscopy and Microanalysis 1996; Bailey, G. W., Corbett, J. M., Dimlich, R. V. W., Michael, J. R., Zaluzec, N. J., Eds.; San Francisco Press: San Francisco, CA, 1996; pp 496-7.

Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

535

Figure 1. Total ion image of a soil particle that had been exposed to 13 000 ppm CsI, allowed to dry, and then washed. The area imaged is 80 × 80 µm. The image is typical of total ion images of both exposed and unexposed particles. The numbers next to the color scale represent secondary ion counts. Lighter colors represent areas of higher counts.

RESULTS AND DISCUSSION Analysis of Unexposed Soil Particles. The total ion image of a particle (Figure 1) from the TOF-SIMS analysis shows irregular bright and dark areas. The bright areas result from more intense emission of secondary ions and can be attributed to either of two factors. The first is the existence of areas that have favorable chemistry for secondary ion emission, e.g., high concentrations of group I cations on the surface. This factor implies that the surface chemistry is inhomogeneous. The inhomogeneity is unlikely to influence an assessment of Cs+ distribution, because group I elements are unique in SIMS in that they display no substantial matrix effects. The second factor is that the bright regions are at optimum focus relative to the secondary ion extraction lens, and hence, contrast with areas that are not in focus. This second factor relates to the topography of the sample surface and must be considered when imaging TOFSIMS data are being interpreted. The assumption that we have made is that when the spectra from two areas are the same (i.e., have the same relative abundances), they are considered to have the same chemical composition, even though they may have different absolute intensities. Changes in chemical composition on the surface are accompanied by changes in the SIMS spectrum. Single-ion images for Al+ and Si+ were generated retrospectively using the inorganic component of the ion signal at nominal m/z 27 and 28. At the mass resolution achievable using the instrument in the imaging mode, Si+ could not be resolved from AlH+. However, we assign the signal at m/z 28 primarily to Si+ for the following reasons. The excellent peak shape of this ion permitted accurate mass measurement. The measured value of m/z 28 was 2 millimass units (mmu) lower than the theoretical for Si+. In contrast, this value was 15 mmu lower than theoretical for AlH+. Additionally, the metal hydride cations are normally observed together with the bare metal cations. In the present analyses, this is not the case: m/z 28 in certain regions is observed nearly to the exclusion of Al+ (m/z 27). The single-ion images of Al+ and Si+ (Figures 2 and 3) clearly show that the chemical composition of the surface is not homogeneous. Most of the areas are high in Al+ (yellow box on the images) and have substantial Si+, but some areas which are 536 Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

Figure 2. Al+ image of soil particle (same particle as in Figure 1.) The numbers next to the color scale represent secondary ion counts. Lighter colors represent areas of higher counts.

Figure 3. Si+ image of the soil particle (same particle as in Figure 1). The numbers next to the color scale represent secondary ion counts. Lighter colors represent areas of higher counts.

very high in Si+ (blue box on the images) are practically depleted in Al+. The SIMS spectra from the high Al+ areas were very similar: in addition to Al+ (m/z 27), abundant ions corresponding to Na+ (m/z 23) and Si+ (m/z 28) were observed (Figure 4). Fe+ (m/z 56), K+ (m/z 39), Ca+ (m/z 40), and Mg+ (m/z 24) also contribute to the inorganic signature, but these ions are for the most part low in abundance. Numerous organic ions are also observed. The most abundant are m/z 29+, 39+, 41+, 43+, 55+, and 57+, and they and most likely correspond to hydrocarbon ions. The origin of these ions is adsorption of contaminants, either from the ambient atmosphere or from the instrument vacuum. They are ubiquitous in “static” SIMS spectroscopy.21 The areas that were depleted in Al (Figure 5) turned out to be significant in the context of Cs adsorption. The SIMS spectra generated from these areas contained abundant Si+ and SiOH+ ions; in contrast, the abundance of Al+ and Na+ was very low. The EDS analyses of the unexposed particles were largely consistent with the TOF-SIMS analysis. They showed that the soil particles were predominantly SiO2, with some Al and minor amounts of Fe and Ti (Table 1). The SEM backscattered electron images revealed the presence of some microscopic zircon (zirconium oxide) and ilmenite (iron titanate) particles, but there were few of these, and they were not representative of the bulk of the sample. (21) Benninghoven, A. Surf. Sci. 1973, 35, 427-57.

Figure 4. SIMS spectrum from high-Al+ area (yellow box in the SIMS images). Abundant Cs+ and Na+ are observed.

Figure 5. SIMS spectrum from high-Si+ area (pale blue box in the SIMS images). Only low-abundance Cs+ was observed.

Table 1. EDS Analysis of Unexposed Soil Sample and Soil Exposed to CsI (13 000 ppm) atom % exposed soil element O Al Si K Ti Fe I Cs

unexposed soil

dull gray area

bright area

58 7.6 33 0.6 0.2 1.2

56 1.4 43

18 21 24 2.2 5.0 13 17

Analysis of Soil Particles Exposed to CsI, Not Washed. Soil samples were exposed to three CsI solutions having different concentrations. The CsI concentrations of the resulting soil samples were estimated at 13 000, 1500, and 160 ppm (mass/ mass) by assuming that all of the CsI was adsorbed to the soil surface. The samples were generated by allowing the solvent to evaporate, leaving the salt on the surface of the soil particles.

These samples were not washed prior to analysis. The highest concentration sample had a Cs+ concentration approximately double the cation-exchange capacity of the soil, which was measured at 2.5 mequiv/100 g of soil (this would correspond to a Cs+ concentration of 3300 ppm). Hence this sample represents the case where excess Cs+ is present relative to the total number of cation exchange sites, and surface precipitation would be expected in addition to sorption. The exposed samples could also be described in terms of monolayer coverage, by assuming a molecular area of 25 Å2 for CsI and an even distribution on the surface. Using this approach, the three samples had surface coverages of 3, 0.4, and 0.04 monolayers. For the most heavily exposed samples, similar TOF-SIMS spectra were generated from the majority of the areas within the images (e.g., Figure 4). The spectra from these areas were characterized by abundant Cs+ (m/z 133), Al+, and Na+. Cs+ in these regions was reproducible on the order of 100-150% of the Al+ peak, for the 1300 ppm samples. Si+ was also observed, but at a lower abundance than Al+. In addition to the atomic species, molecular Cs-bearing adduct ions could also be observed which are indicative of the chemical environment of Cs on the surface. Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

537

Figure 6. SEM image of soil particle exposed to CsI solutions and then dried but not washed. Concentration of CsI is estimated at 13 000 ppm. Particles are several hundred micrometers across.

Multiple layers of CsI on the surface of the soil particles are most likely responsible for the formation CsI+ (m/z 260), Cs2+ (m/z 266), Cs2I+ (m/z 393), and much lower abundant ions corresponding to Cs3I2+ (m/z 653) and Cs4I3+ (m/z 913). Molecular adduct ions corresponding to Cs2O+ (m/z 282), and Cs2OH+ (m/z 283) were also observed and are indicative of Cs+ interaction with the soil matrix (sorption). A reasonably abundant ion corresponding to Cs2Cl+ was also observed. This could originate from the contaminant solution or from Cl- present on the surface of the soil particles. In contrast to the high-Al+/high-Cs+ regions, smaller areas could be observed where the abundance of the Cs+ peak was markedly reduced. Significantly, these areas were also depleted in Al+, evidenced by the low-abundance m/z 27+ (e.g., Figure 5). Analysis of different Al+ depleted regions showed Cs+ abundance at 1-10% of the Si+ ion for the 1300 ppm unwashed sample. This observation indicates that Cs+ has an affinity for regions that are higher in Al and may be less attracted to surface regions where only SiO2 is present. This is consistent with the concept that isomorphic substitution of Al into the SiO2 matrix induces anionic charge centers on the surface that can serve to complex with Cs+ and other cations.2 The conclusion is also in accord with the work of MacKenzie, who noted a correlation between 137Cs and Al in vertical profiles of sediments.22 SEM images (electron backscattering) of the most heavily exposed samples (13 000 ppm) showed that the previously featureless soil particles now contained substantial bright regions, which EDS revealed to contain abundant CsI (Figure 6). A composite EDS analysis of many particles showed Cs present at 1 atom %, which is in reasonable agreement with what the sample was exposed to (0.7 atom %). An analysis of the dull gray areas (22) MacKenzie, A. B. In Techniques for Identifying Transuranic Speciation in Aquatic Environments; International Atomic Energy Agency: Vienna, 1981; pp 257-262. (23) Ingram, J. C.; Groenewold, G. S.; Appelhans, A. D.; Dahl, D. A.; Delmore, J. E. Anal. Chem. 1996, 68, 1309-16.

538

Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

showed that the particles were almost entirely Si and O, with a minor amount of Al (Table 1). No Cs was detectable in these areas. In contrast, analysis of the bright regions revealed substantial amounts of Cs, I, Si, O, and significantly, Al. A second difference between the TOF-SIMS and the SEM/ EDS analysis is that Al is much more prevalent in the TOF-SIMS analysis. An experiment was performed in which an unexposed soil sample was sputter-cleaned, to sample the underlying layers of the soil particle. A depletion of Na+ and the organic ions was observed, but the ratio of Al+/Si+ was unchanged. It may be that at the greater depths sampled in the SEM/EDS experiments, Al is less prevalent than it is closer to the surface (0-10 nm sampled by the TOF-SIMS). This would also be consistent with the X-ray diffraction data, which indicated that the soil was primarily SiO2. The TOF-SIMS and EDS analyses both indicated a correlation between Al and Cs on the soil particle surfaces. However, the two analytical techniques appeared to oppose each other regarding the distribution of Cs on the particles. In the TOF-SIMS analysis, Cs appears on most particles, covering most of those particles. In contrast, Cs appears only in selected regions in the EDS analyses. These results can be reconciled by considering that EDS is sampling the surface at depths up to 2 µm and is insensitive to monolayer coverings. We believe that EDS is only detecting Cs in those areas where abundant Al-bearing sites are adsorbing substantial Cs and that this creates nucleation sites where CsI surface precipitates can and do crystallize. In areas where insufficient Al is present, CsI is not observed by EDS. This explanation is supported by the observation that neither Cs nor I was observed in the EDS analyses of the samples of lower concentration. TOF-SIMS analyses of the 1500 ppm unwashed sample resulted in the generation of similar SIMS spectra from different areas of the particles imaged. These spectra contained significant Si+, Al+, and adsorbed Cs+; the latter ion was not as abundant as in the previous case. As in the case of the more concentrated sample, a small fraction of the area produced SIMS spectra that contained high Si+, but low Al+ and, consequently, very little adsorbed Cs. Cs+ was clearly observable in the TOF-SIMS analyses of the 160 ppm soil samples (80 ppm Cs), although by comparison to the major ions in the spectrum, its abundance was small. A typical signal to noise ratio was estimated at 180 for this ion, based on the ion background in the mass range 132.7-133.0 observed in an unexposed soil particle analysis. This suggests that it should be possible to observe Cs in concentrations as low as 4 ppm with a signal to noise ratio of 10, which corresponds to a surface concentration of ∼4 × 10-4 monolayers (Cs+ ionic radii 1.7 Å) in this system. This is ∼1 order of magnitude lower than the detection limit reported for organics on soil;23 the lower detection limit in the present case is due to a higher ionization probability for Cs and to improved selectivity stemming from the high mass resolution capability of the TOF analyzer. The lowest concentration regime from which meaningful imaging data can be acquired is somewhat higher, ∼80 ppm in our estimation. For Cs+, this corresponds to ∼1.5 × 10-2 monolayer. For this soil sample, it corresponds to ∼2% of the sites available for cation exchange. Analysis of Soil Particles Exposed to CsI, Washed with Deionized Water. The soil particles were then repeatedly washed using deionized water to remove any soluble Cs species,

viz., crystallized CsI. The washing appeared to have had the desired effect because no Cs or I was observed in the EDS analyses of even the most concentrated sample. The total ion images of these particles revealed numerous bright features, and the TOF-SIMS analyses of these areas revealed Cs+ ions which were as abundant as in the spectra from the unwashed samples (Figure 4). Abundant Al+ was also abundant in these spectra. However, no Cs2+ or CsxIy+ species were observable, which is consistent with the notion that the CsI salts had been removed from the surface of the soil sample. However, Cs2OH+ and (CsOH)2H+ were observed, which is consistent with what would be expected from Cs bound to the surface of the aluminosilicate lattice. The total ion image also revealed several dark areas, which suggested that these corresponded to topographical holes in the particle. However, when the Si+ image was reconstructed, abundant Si+ could be clearly observed in several of the dark areas. The SIMS spectra from these regions contained very little Cs+ (Figure 5). Similar data was recorded for the washed soil samples that (originally) contained 1500 and 160 ppm CsI. The analyses of the washed samples supports the idea that the preference of Cs+ for the Al-bearing regions is a mineral surface phenomenon and is not related to multiple layers of CsI on the soil. CONCLUSIONS These results show clearly that Cs+ spatially correlates with Al+, which strongly suggests that Cs+ prefers to adsorb to silicate (24) Wolkenstein, M.; Hutter, H.; Mittermayr, Ch.; Schiesser, W.; Grasserbauer, M. Anal. Chem. 1997, 69, 777-82.

minerals contain Al and tends to eschew minerals that are predominantly SiO2. This demonstrates the facility with which imaging TOF-SIMS can be used for characterization of adsorbed contaminants on naturally occurring mineral surfaces. This is particularly significant because phase separation is not required for analysis, which could be an important feature for investigation of highly heterogeneous systems such as basalts. The current approach to the study of microphase chemistry of these minerals is to crush and then laboriously separate microscopic phases using a needle and a forceps. Only after this is completed can investigation of the chemistry begin, and even then the chemical result will be a composite of weathered and freshly exposed mineral faces. One drawback to the imaging TOF-SIMS approach is that the volume of data generated can make data reduction an intractable problem. For example, the volume of data for this study was 197 megabytes in 21 files. This problem has motivated research in a neural network approach to phase identification24 and is currently being pursued in our laboratories. ACKNOWLEDGMENT We gratefully acknowledge the support of the U.S. Department of Energy, Office of Health and Environmental Research, under Contract DE-AC07-94ID13223. We also acknowledge the assistance of Mike Hankins and Arnie Erickson and helpful discussions with Keith Daum. Received for review May 19, 1997. Accepted November 10, 1997.X AC9705172 X

Abstract published in Advance ACS Abstracts, December 15, 1997.

Analytical Chemistry, Vol. 70, No. 3, February 1, 1998

539