2006 Philadelphia Annual Meeting (22–25 October 2006)

Paper No. 24
Presentation Time: 1:30 PM-5:30 PM


WADOSKI-MCCLELLAND, Eva R. and CLARK, Christine M., Geography and Geology, Eastern Michigan University, 205 Strong Hall, Ypsilanti, MI 48197, ewados@gmail.com

The tourmaline-group minerals are boron-rich cyclosilicates with the general formula of XY3 Z6 T6O18 (BO3)3 V3W. Normalization of the tourmaline-group is riddled with problems due to its complex and varying chemical end-members. Tourmaline formulae are often based on a variety of normalization methods and stoichiometric assumptions, due to the difficulty in measuring the lighter elements (such as B, H and Li) as well as identifying the valence state of iron. Because of these potentially erroneous assumptions, tourmaline samples may be misidentified, such as buergerite for “fluor-schorl”. Most analyses of the chemistry-crystal structure link in tourmaline have been based on sites that have a great deal of variability (e.g. the octahedral sites). Our approach is based on the variations in bond-length related to the B-site. The B-site is unique in tourmalines as it has been observed to only host boron, thus any variation in site-geometry must be in response to chemical variations at other sites. Previous research of this issue has shown that while the bond-length is fixed, variations do occur in B-O2 and B-O8 bond-lengths with chemistry at the cation sites. In order to authenticate this relationship, electron-microprobe analysis and single-crystal X-ray diffraction data were collected from the published literature. Data from the literature were combined with our own suite of tourmalines and analyzed statistically using step-wise linear regression. Statistical analysis of the stereochemistry of tourmaline shows primary influence by the Z'-site, but secondary influence by the X- and Z-sites. Therefore, we suggest that tourmaline-normalization assumptions can be validated by examining the bond-lengths of the B-site.