Simultaneous Mineralogical Quantification and Chemical Characterization of Soil Clays

-A new chemical mass balance technique has been developed for simultaneous mineralogical quantification and chemical characterization of soil clays. The procedure includes separation of the whole clay (<2 ~tm fraction) into six particle size fractions (<0.02, <0.06, <0.2, 0.02-0.06, 0.06-0.2, and 0.22 um fractions), chemical analysis of the whole clay and each of the six fractions, and fitting of a nonlinear chemical mass balance model to the chemical analyses. As written, the chemical mass balance model is valid only for samples containing mixtures of quartz, kaolinite, illite, and mixed-layered smectiteillite. Samples containing carbonates and free iron compounds may be analyzed using the technique if these phases are chemically removed prior to particle size fractionation. Accuracy of the new technique was tested using synthetic data and found to depend on the quality of the input data; however, clay phase quantification within three percentage points of known values was readily achieved. Precision of the technique was evaluated by independently preparing and analyzing five samples of the same soil clay. Standard deviations for clay phase percentages (w:w) in the <2/~m fraction were all less than one percent. The new technique yields accurate determinations of chemistry for the smectitic and illitic phases in mixed-layered smectite-illite, and qualitative estimates for the chemistry of l0 A-illite. The elemental compositions of quartz and kaolinite are assumed a priori and treated as constants within the non-linear chemical mass balance model. Key Words--Chemistry, Illite, Mass-balance, Protoillite, Quantify, Smectite. I N T R O D U C T I O N Quantitative clay mineralogy by most x-ray diffraction (XRD) techniques requires mineral standards with XRD properties similar to those of the mineral phases in unknown samples (Brindley 1980). Unfortunately, the heterogeneous nature of clays in soils and sediments inherently limits the reliability of reference clays (e.g., Wyoming bentonite, Georgia kaolinite, etc.) for use as standards in XRD analyses. The Rietveld technique does not require mineral standards, but does require that the structure factor be constant for a given reflection and that no non-Bragg diffraction effects be exhibited (Bish 1993). For most clay-size layer silicates, the structure factor is not constant and stacking disorders cause two-dimensional diffraction effects; hence applicability of the Rietveld technique for quantification of clays in soils and sediments is severely limited (Bish 1993). The mult i-component approach to quantitative clay mineralogy (Alexiades and Jackson 1966; Johnson et al 1985; Engler and Iyengar 1987) integrates a variety of independent measures. For example, illite is quantified based on the K20 content of an unknown sample after accounting for K associated with feldspars and by assuming a value for the K20 content of illite in the sample. Vermiculite is quantified by assuming that vermiculite is the only mineral phase capable of "fixing K" (Alexiades and Jackson 1965). Gibbsite and kaCopyright 9 1994, The Clay Minerals Society olinite are quantified using differential thermal analyses based on the magnitude of specific endotherms (Dixon 1966). Modem mass balance techniques for quantitative mineralogy incorporate aspects of the multi-component approach, but they typically rely heavily on chemical mass balance. Pearson (1978) considered the range of published chemical compositions to define upper and lower limits for compositions of mineral phases in unknown samples. A graphical solution to sets of simultaneous linear equations depicted regions of selfconsistent solutions, and thus defined upper and lower limits for proportions of mineral phases in unknown samples. Other researchers (Gold et al 1983; Hodgson and Dudeney 1984; Braun 1986; Slaughter 1989; Calvert et al 1989) have developed computer programs for quantitative mineralogy based on constrained chemical mass balance techniques similar to that used by Pearson (1978). Hodgson and Dudeney (1984) considered stoichiometries and water mass balance to constrain solutions to linear chemical mass balance equations. Slaughter (1989) constrained linear chemical mass balance equations using semi-quantitative XRD estimates of mineralogy, measured d-values, as well as stoichiometries and ranges in chemical compositions for mineral phases. Calvert et al (1989) used a nonlinear mass balance model to quantify clay mineralogy; however, they relied on assumed ranges for chemical