Geochimica ~1 Cosmochimica Acfa zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Vol. $2. pp. 2537-2542 Copyright Q 1988 Pergamon Press plc. Printed in U.S.A. 0016-7037/88/$3.00 + .oo Application of Auger Electron Spectroscopy (AES) to naturally weathered hornblende DAVID W. MOGK and WILLIAMW. LOCKE, III Department of Barth Sciences, Montana State University Bozeman, MT 59717, U.S.A. zyxwvutsrqponmlkjihgfedcb (ReceivedMay 25, 1987; acceptedin revisedform August 2, 1988) Abstract-The surface chemistry of naturally weathered hornblende has been analyzed using Auger Electron Spec- troscopy. The high spatial resolution and depth profiling capabilities of this technique allow changes in the relative concentrations of cations to be determined over micron-scale areas and at depths resolvable on the sub-micron scale. These data indicate that during chemical weathering: 1) there is a systematic change in surface chemistry through a thickness up to 1200 angstroms, 2) complete cation depletion at the surface layer does not occur, 3) different components are leached to different depths to varying degrees, and 4) no new phase such as clay or smectite necessarily forms. A nonsteady state diffusion model is most consistent with these data. INTRODUCIION FIVE DECADES of observation of the products of weathering has not served to elucidate completely the processes of weathering of mineral surfaces. Specifically, we do not know, for a given environment, if chemical weathering of minerals occurs through surface reaction or volume dilfusion processes. Successive application of solution chemistry, SEM, and XPS in laboratory dissolution studies (CORRENS and VON EN- GELHARDT, 1938; LUCE et al., 1972; NICKEL, 1973; SIEVER and WOODFORD, 1979; BERNER et al., 1980, SCHO~ et al., 1981; SCHOTT and BERNER, 1983) and SEM, XPS, XRD, EDX, TEM, and EMP in the study of naturally weathered minerals (BERNER and HOLDREN, 1977; RODGERS and HOLLAND, 1979; BERNERet al., 1980; ECCLETONand Bo- LAND, 1982; NAHONand COLIN, 1982; BERNER and SCHOTT, 1982) have refined the definition of the “surface” of a material from microns to angstroms. However, these techniques have not provided direct evidence for the depth of weathering re- actions or the nature of elemental changes normal to weath- ered mineral surfaces. In studies of mafic minerals, dissolution rates often have been described as incongruent and parabolic (MCCLELLAND, 1950; LUCE et al., 1972; NICKEL, 1973; SIEVER and WOOD- FORD, 1979). The mechanisms suggested as potentially re- sponsible for dissolution rates which apparently change with time are the precipitation of amorphous aluminosilicates or iron oxides (SIEVER and WOODFORD, 1979; HOLDREN and ADAMS, 1982) and replacement by H+ (e.g. LUCEet al., 1972). The latter process includes the nonsteady state diffusion of ions within the mineral (LUCE et al., 1972), quasi-steady state diffusion through a leached shell (LUCE et al., 1972), and recrystallization of the decationated and hydrated layer (EG- GLETON and BOLAND, 1982; NAHON and COLIN, 1982). In contrast, some recent laboratory work (BERNER et al., 1980; BERNER and SCHOTT, 1982; SCHOTT and BERNER, 1985) indicates that, after the removal of fine particles gen- erated in the process of sample preparation, Si release is es- sentially linear (surfacecontrolled). In the same experiments Ca and Mg showed initial incongruency with a long-term (hours to days) return to congruent (linear) removal with Si. These results are interpreted as supporting the presence of a thin (< 10 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA A if totally depleted) layer of base replacement by H+ relative to silica and the absence (in anoxic environments) of a protective surface layer. In non-acidic, oxygenated en- vironments hydrated ferric oxide and ferric magnesium sil- icates may form and impede Si release (!&HOT-~ and BERNER, 1985). PETIT et al. (1987) directly measured hydrogen addition rather than base depletion from mineral surfaces. Resonant nuclear reaction (RNR) profiling shows hydrogen penetration to depths of 1000 A beneath the mineral surface in laboratory experiments, in contrast to X-ray photoelectron spectroscopy (XPS) results which suggest that weathering processes are restricted to the outermost few tens of angstroms. Comple- mentary secondary ion mass spectrometry (SIMS) data re- ferred to by PETIT et al. (1987) and briefly discussed by SCHOTTand PETIT (1987) show notable depletion in major cations to this depth. The apparent discrepancy between these results and earlier work requires further examination. To date, the direct observation of the inferred decationated layer of very thin layers of precipitate or recrysmlhzed material has not been possible, although numerous studies based on XPS have inferred approximate thicknesses of such layers (e.g. PETROVIC et al., 1976). This study reports the results of high-resolution Auger Electron Spectroscopy (AES) studies of the surface of naturally weathered hornblende. AES is an analytical surface chemistry technique that is routinely used to analyze elemental monolayers on solid substrates. There have been very few attempts to apply this technique to nat- urally occurring systems of geologic interest (MACKINNON and MOCK, 1985; MUCCI and MORSE, 1985; MUCCI et al., 1985; HOCHELLA et al., 1986a,b). This is somewhat surprising because it is in the near surface environment (100’s of A) that minerals interact with their geologic environment. The high lateral resolution and depth profiling capabilities of AES make it useful for fully characterizing the spatial distribution and relative concentration of elements on crystal surfaces. This study, designed to show the changes in cation ratios at shallow depths beneath the grain surface, is an outgrowth of the work cited above. It offers a new conceptual framework to understand weathering processes with results that supple- ment and complement the earlier studies. SAMPLE Naturally weathered hornblende was collected from the B horizon of a soil on Bull Lake glacial outwash (cu. 130,000 y.) from Rocky 2537