Quantification of phyllosilicates in soils often depends on a combination of methods, including x-ray diffraction (XRD), cation exchange and fixation capacity, and microscopy. We examined a Rb-fixation (Rb_110°C) method to quantify vermiculite content in soils derived from granitic and non-granitic alluvium of the San Joaquin Valley (SJV), California. In this method a sample is saturated with Rb and heated to 110°C followed by Rb displacement with NH_4. Any Rb remaining after NH₄ treatment was determined directly by instrumental neutron activation analysis (INAA). In the INAA analysis the reaction ⁸⁵Rb + n → ⁸⁶Rb + g was quantified by measurement of the gamma radiation. The amount of vermiculite (g kg^-1) was calculated as fixed Rb content (cmol_c kg^-1)/0.154. For comparison, Rb fixation was also estimated after a 1-hr incubation (Rb_1hr) with 2 mmol RbCl followed by a 30-min extraction with NH₄Cl. Fixed Rb was calculated as the difference between the initial Rb added and the extracted Rb. 

While the amount of Rb fixation in the 1 h technique was correlated with Rb fixed by heating (R ² = 0.48**), the more rigorous heating treatment fixed almost 35 times more Rb. Data for Rb_110°C had a highly significant correlation with soil clay content (R ² = 0.85***). However, XRD data showed that clay fractions were dominated by smectite, chlorite, and mica while vermiculite, the major K- and Rb-fixing mineral, was a dominant mineral in most silt fractions. Silt content correlated with Rb_1hr fixation (R ² = 0.48**), but did not correlate with Rb_110°C fixation (R ² = 0.21^ns).

For soils from the SJV the Rb_110°C fixation method did not provide a reliable estimate of vermiculite content, relative to XRD results.  Our data suggest that Rb fixation with heating to 110°C method overestimates the amount of vermiculite if high-charge smectites are present.