Abstract
The crystal structures of an albite sample from Tiburon, Marin County, California, annealed at 1080°C for 60 days, were refined from X-ray intensity data collected at 25, 500, 750, 980, and 1040°C to R factors of 0.034, 0.036. 0.037, 0.037, and 0.039 for 2443, 1993, 2013, 2016, and 2018 reflections, respectively. The crystal structures of a second albite sample from Amelia, Virginia, annealed at temperatures up to 1110°C for 133 days, were also refined to R factors of 0.040 and 0.044 based on 2018 and 430 reflections collected at 980 and 1060°C respectively.
The Tiburon sample failed to attain monoclinic symmetry with respect to either unit-cell dimensions or atomic parameters within the temperature range of study, which is in agreement with an earlier high-temperature investigation by Prewitt et al. (1976). Extrapolation of monoclinic indicators based on atomic positions and bond lengths related by pseudo-monoclinic symmetry indicates that monoclinic symmetry would be attained within the temperature range 1050-1075°C for the Tiburon sample and 1105-1110°C for the synthetic albite sample studied by Prewitt et al. (1976). The Amelia sample was found to be definitely monoclinic at 980°C. The variable inversion temperatures of these albite samples reflect differences in the degree of Al/Si disorder created by the different annealing times and temperatures. These differences in Al/Si disorder are too slight to be measured by current indirect methods.
The spatial disorder associated with the sodium atom in high albite is attributed to the response of Na to differences in the local Al/Si configurations created by Al/Si disorder. Previous interpretations restricting the number of such configurations to four and assuming isotropic Na quarter-atoms are inadequate. Changes in the electron density distribution about the Na atom as a function of temperature reflect changes in the shape of the aluminosilicate cage surrounding Na as monoclinic symmetry is gradually approached with increasing temperature.
