Ferric/ferrous iron ratios in sodic amphiboles: Mossbauer analysis, stoichiometry-based model calculations and the high-resolution microanalytical flank method

Citation
M. Enders et al., Ferric/ferrous iron ratios in sodic amphiboles: Mossbauer analysis, stoichiometry-based model calculations and the high-resolution microanalytical flank method, CONTR MIN P, 140(2), 2000, pp. 135-147
Citations number
41
Categorie Soggetti
Earth Sciences
Journal title
CONTRIBUTIONS TO MINERALOGY AND PETROLOGY
ISSN journal
00107999 → ACNP
Volume
140
Issue
2
Year of publication
2000
Pages
135 - 147
Database
ISI
SICI code
0010-7999(200012)140:2<135:FIRISA>2.0.ZU;2-4
Abstract
Although the electron microprobe has become the standard microanalytical to ol in modern geosciences, conventional electron microprobe analysis does no t allow determination of the valence states of elements such as Fe. The cor rect classification of minerals and interpretation of reaction microfabrics and grain zonation require high-quality information on ferric/ferrous rati os on a scale of micrometers. The flank method developed by Hofer et al. (1 994, Eur J Mineral 6:407-418) has revived new interest in electron-induced X-ray-spectroscopy to resolve oxidation states in minerals with high spatia l resolution. We have recharacterized well-documented sodic amphiboles of t he glaucophane-ferroglaucophane-riebeckite-magnesioriebeckite series by ele ctron probe microanalysis and combined the microanalytical data with ferric /ferrous ratios from Mossbauer spectroscopy, Li data from bulk ICP-AES anal ysis and H2O data from bulk Karl-Fischer titration. The combination of micr oanalysis and high-quality analysis on the bulk materials results in a data set that allows comparison of model-based stoichiometric calculations and the calibration of the high-resolution flank method. The calibration obtain ed allows ferric/ferrous ratios to be determined within an error of +/- 5%. We have found it necessary to apply an empirical correction for absorption phenomena. The advantages of the method must be weighed against the comple x calibration procedures necessary and thus the flank method will probably not find use as a routine method. However, in cases where high-resolution d ata in terms of valence state are needed, the flank method will provide use ful data on ferric/ferrous ratios down to minimum FeOtotal content of 6-8 w t%.