Anal. Chem. 1999, 71, 1712-1719
Identification of Mineral Phases on Basalt Surfaces by Imaging SIMS Jani C. Ingram,* Gary S. Groenewold, John E. Olson, and Anita K. Gianotto
Idaho National Engineering & Environmental Laboratory, P.O. Box 1625, Idaho Falls, Idaho 83415-2208 Michael O. McCurry
Department of Geology, Idaho State University, Pocatello, Idaho 83209
A method for the identification of mineral phases on basalt surfaces utilizing secondary ion mass spectrometry (SIMS) with imaging capability is described. The goal of this work is to establish the use of imaging SIMS for characterization of the surface of basalt. The basalt surfaces were examined by interrogating the intact basalt (heterogeneous mix of mineral phases) as well as mineral phases that have been separated from the basalt samples. Mineral separates from the basalt were used to establish reference spectra for the specific mineral phases. Electron microprobe and X-ray photoelectron spectroscopy were used as supplemental techniques for providing additional characterization of the basalt. Mineral phases that make up the composition of the basalt were identified from single-ion images which were statistically grouped. The statistical grouping is performed by utilizing a program that employs a generalized learning vector quantization technique. Identification of the mineral phases on the basalt surface is achieved by comparing the mass spectra from the statistically grouped regions of the basalt to the mass spectral results from the mineral separates. The results of this work illustrate the potential for using imaging SIMS to study adsorption chemistry at the top surface of heterogeneous mineral samples. The chemical reactivity at the contaminant/mineral interface greatly influences the fate and transport of many contaminants in the environment; it is at this interface that sequestration and degradation are known to occur.1 However, investigation of the interfacial chemistry at mineral surfaces is complex due to the heterogeneous nature of natural mineral surfaces. Instrumental constraints relating to sensitivity and selectivity make it difficult to interpret results from these types of surfaces. Consequently, many adsorption studies have been performed using model materials having homogeneous surfaces, which greatly simplifies data interpretation.2 An alternative approach to model studies is the use of imaging tools that can probe microregions of the surface; these tools can provide a clearer view of the heterogeneous adsorption sites (1) Davis, J. A.; Hayes, K. F. In Geochemical Process at Mineral Surfaces; Davis, J. A., Hayes, K. F., Eds.; ACS Symposium Series 323; American Chemical Society: Washington, DC, 1986; pp 2-18. (2) Henrich, V. E. In Physics and Chemistry of Mineral Surfaces; Brady, P. V., Ed.; CRC Press: Boca Raton, FL, 1996; pp 79-80.
1712 Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
present on the mineral surface. Techniques such as energy- or wavelength-dispersive spectroscopy and electron spectroscopies used with small spot excitation have proven effective for elemental imaging of minerals in the near-surface region. However, these approaches are not amenable to probing the interfacial chemistry as the depth of interrogation (on the order of nanometers to micrometers) is substantially below the top monolayer of the sample. Additionally, these techniques provide elemental and some oxidation-state information but do not provide molecular information pertinent to interrogating functional groups required for adsorption studies of organic species.3 Clearly, a requirement for investigating the chemistry of contaminant/mineral interfaces is that the imaging tool must probe the top few layers of the sample. Laser desorption mass spectrometric techniques operated with focused beams (such as laser microprobe mass analysis, LAMMA) have successfully been used to study adsorption chemistry at mineral surfaces. For example, De Waele and Adams used a laser microprobe and X-ray microanalysis to follow a leaching study of chrysotile and Chrysophosphate fibers. They were able to show the removal of phosphate upon oxalic acid treatment and were also able to detect PAHs and organic amines on the surfaces of these samples.4 However, in most cases, laser desorption instruments are microprobe instruments which probe specific microregions of the surface in contrast to imaging larger areas of the sample surface. Imaging by this approach requires high spatial control and monitoring, sophisticated software for data storage, and good reproducibility, which most laser desorption instruments do not have.5 It has been demonstrated that static secondary ion mass spectrometry (SIMS) instruments that utilize time-of-flight (TOF) mass analyzers are capable of interrogating adsorption behavior at mineral surfaces. TOF SIMS with imaging capability can provide spatial information about the composition of the sample which can be used to characterize specific regions of the sample surface. Brinen and Reich reported the use of static SIMS imaging for studying adsorption of diisobutyl dithophosphinate on galena surfaces.6 Mathez and Mogk utilized TOF SIMS to characterize carbon compounds on pyroxene surfaces in basalt.7 Pratt et al. (3) Spectroscopic Characterization of Minerals and Their Surfaces; Coyne, L. M., McKeever, S. W. S., Blake, D. F., Eds.; ACS Symposium Series 415; American Chemical Society: Washington, DC, 1990. (4) De Waele, J. K.; Adams, F. C. Scanning Microsc. 1988, 2, 209-228. (5) Kaufmann, R. L.; Rechmann, P.; Tourmann, J. L.; Schnatz, H. In Microbeam Analysis 1989; Russell, P. E., Ed.; San Francisco Press: San Francisco, CA, 1989; p 35. 10.1021/ac9811571 CCC: $18.00
© 1999 American Chemical Society Published on Web 04/03/1999
reported the use of imaging SIMS together with elastically backscattered electron and energy-dispersive X-ray images to study the extent of oxidation on nickel/copper sulfide ore surfaces.8 In a study reported by our laboratory, Cs+ adsorption on soil was investigated using TOF SIMS imaging; selective adsorption of the Cs+ at alumina sites compared to silica sites was observed.9 In the present study, the characterization of basalt surfaces is presented, and a method for identification of the mineral phases from the SIMS data is described. The purpose of this study was to develop methodology for the use of imaging SIMS for investigating the surface chemistry of basalt. Of specific interest is the adsorption and/or degradation chemistry of contaminants present in subsurface basalt as found at the Department of Energy sites in Idaho and Washington. To understand the interactions of contaminants with adsorption sites of the basalt, it is necessary to identify the mineral phases at the top surface of the basalt. Methods for identifying mineral phases from energy- or wavelengthdispersive spectroscopic (i.e., electron microprobe) data are routinely used by the geochemistry community;10 however, these methods probe below the surface (at about micrometer depths) and so do not provide information about the top surface of the samples. Identification of mineral phases from complex rock samples, such as basalt, using imaging SIMS has not been reported. It was demonstrated by Wolkenstein et al.11 that a SIMS microprobe could be used to identify different chemical phases on the surface of solder alloy. To identify the chemical phases, the SIMS data were classified using a Kohonen network. This approach was taken in the work described here; SIMS images of natural basalt surfaces were collected, and then the mineral phases were identified by using statistical grouping. Reference materials were used to verify the identified mineral phases. EXPERIMENTAL SECTION Materials. The basalt characterized in these experiments was collected at core hole 2-2A on the Idaho National Engineering & Environmental Laboratory site at a depth of 300.5-301.0 ft. Mineral separation of this basalt was performed for use as reference material in identifying mineral phases. Mineral Separation. Separation techniques used in this work included magnetic, heavy liquid, and hand separations; Hutchison12 provided a detailed description of the entire mineral separation process. The procedures used in this work are briefly described. The first phase of the mineral separation process is the reduction of the rock sample to gravel. Rock samples were typically 1-2 kg. The rock sample was first cleaned of surface debris and then crushed with a steel sledgehammer on a heavy steel plate. To avoid cross contamination between rock samples, all crushing implements were thoroughly cleaned. (6) Brinen, J. S.; Reich, F. Surf. Interface Anal. 1992, 18, 448-452. (7) Matez, E. A.; Mogk, D. M. Am. Mineral. 1998, 83, 918-924. (8) Pratt, A. R.; Franzreb, K.; McIntyre, N. S. Surf. Interface Anal. 1998, 26, 869-871. (9) Groenewold, G. S.; Ingram, J. C.; McLing, T.; Gianotto, A. K. Anal. Chem. 1998, 70, 534-539. (10) Klein, C.; Hurlbut, C. S., Jr. Manual of Mineralogy; John Wiley & Sons: New York, 1977; p 8. (11) Wolkenstein, M.; Hutter, H.; Mittermayr, Ch.; Schiesser, W.; Grasserbauer, M. Anal. Chem. 1997, 69, 777. (12) Hutchison, C. S. Laboratory Handbook of Petrographic Techniques; John Wiley & Sons: New York, 1974.
The crushed rock sample was rinsed prior to sieving as dust particles cling to the mineral grains and increase their effective diameters. The crushed rock sample was placed in a large (1000 mL) beaker which was then filled with tap water. The water was then decanted off the crushed rock, removing the finer dust particles. This process was repeated until the water added to the sample was not turbid upon shaking the beaker. After rinsing, the sample was placed in a Buehler oven and dried at 50 °C. The rinsed and dried rock sample was sieved to isolate particular ranges of particle size. Three sieve sizes used were 40, 120, and 230 mesh (0.42, 0.12, and 0.063 mm, respectively). After sieving, samples from each mesh size were weighed and placed into individual containers. The first separation technique applied to the rock samples utilized magnetic fields that attract the more magnetically susceptible mineral phases. Two magnetic techniques were used. The first was the use of a large hand magnet and the second was a Franz magnetic separator. The hand magnet was used to remove the most magnetically susceptible minerals, namely, iron/titanium oxides (magnetite and ilmenite). After the hand magnet was used, the remaining mixture was enriched in relatively nonmagnetic minerals (e.g., olivine and plagioclase). This mixture was placed in the Franz magnetic separator, which was used to discriminate the magnetic susceptibility of olivine and plagioclase. Two separates were produced in the latter step. One was enriched in olivine, and the other enriched in plagioclase. Both separates were then processed using a third technique, heavy-liquid separation. This method exploits the differing densities of each mineral phase, in this case, olivine and plagioclase. The two liquids used were bromoform and sodium metatungstate. The density of the liquid used in this work is ∼2.9 g/cm3. The density of olivine is ∼3.2 g/cm3 and plagioclase is ∼2.7 g/cm3. Thus, when a mixture containing olivine and plagioclase is placed in the heavy liquid, olivine grains will sink and plagioclase grains will float. After a sample containing olivine and plagioclase was placed in the liquid and settling had occurred, the two mineral separates were extracted, washed, and dried. The last step in mineral separation was to select individual mineral grains and pull them out of the mixture. This was accomplished by placing the mixture of minerals on a tray and viewing them through a binocular microscope. Individual mineral grains were identified and picked out of the sample using precision tweezers. This technique was used to obtain high-purity olivine and plagioclase mineral separates. It should be noted that the pyroxene is intimately intergrown with plagioclase; thus, clean separates could not be obtained. TOF-SIMS Instrumentation. Analyses were performed using a Charles Evans & Associates TOF-SIMS instrument,13 located at the Image and Chemical Analysis Laboratory at Montana State University, which is a user facility. The experimental protocol used in these experiments was very similar to those described in ref 9; a brief description of those protocols will be given here. The basalt 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 mm area was scatter rastered using the primary ion beam in a pulsed fashion. The primary ion (13) Schueler, B.; Sander, P.; Reed, D. A. Vacuum 1990, 41, 1661-1664.
Analytical Chemistry, Vol. 71, No. 9, May 1, 1999
1713
was a microfocused Ga+ gun operated at 600 pA dc, +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 ∼3 × 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 time-of-flight mass analyzer. 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 mm or less) and more modest mass resolution. The mass resolution is also compromised to some extent by the fact that the surfaces were not flat. However, in the lower mass regions (
