TY - JOUR

T1 - Electron probe microanalysis of Fe 2+ /ΣFe ratios in calcic and sodic-calcic amphibole and biotite using the flank method

AU - Li, Xiaoyan

AU - Zhang, Chao

AU - Almeev, Renat R.

AU - Zhang, Xue Chun

AU - Zhao, Xin Fu

AU - Wang, Lian Xun

AU - Koepke, Jürgen

AU - Holtz, Francois

N1 - Funding information: We thank Florian Kabacinski and Lukas Messerschmidt for their help with the wet chemistry colorimetric method, and Julian Feige for sample preparation. Heidi E. Höfer (Goethe-Universität Frankfurt) is appreciated for donating garnet references. Timothy Rose (Smithsonian National Museum of Natural History) is thanked for providing us extraordinary amounts of amphibole reference NMNH 143965 and NMNH 111365. We thank Filippo Ridolfi for helpful discussion on calculation of amphibole formula using AMFORM. We appreciate the insightful comments from Heidi E. Höfer and another anonymous reviewer, as well as the editorial work of Balz Kamber. This study was supported by DFG (German Research Foundation) project BE 1720/40 . We thank Florian Kabacinski and Lukas Messerschmidt for their help with the wet chemistry colorimetric method, and Julian Feige for sample preparation. Heidi E. Höfer (Goethe-Universität Frankfurt) is appreciated for donating garnet references. Timothy Rose (Smithsonian National Museum of Natural History) is thanked for providing us extraordinary amounts of amphibole reference NMNH 143965 and NMNH 111365. We thank Filippo Ridolfi for helpful discussion on calculation of amphibole formula using AMFORM. We appreciate the insightful comments from Heidi E. Höfer and another anonymous reviewer, as well as the editorial work of Balz Kamber. This study was supported by DFG (German Research Foundation) project BE 1720/40.

PY - 2019/3/30

Y1 - 2019/3/30

N2 - We present an analytical approach to determine the Fe 2+ /ΣFe ratio in amphiboles and biotites using electron probe microanalysis (EPMA). The “flank method” applied in this study is based on the fact that FeL line spectra show different peak position and intensity for Fe 2+ and Fe 3+ , which is associated with a resonant process involving both X-ray absorption and emission. The flank positions for Lα and Lβ, which correspond to sites with minimal Fe 3+ L 3 absorption and maximal Fe 2+ L 2 absorption respectively, are determined based on the L-line spectra difference of two Fe-rich ferric and ferrous garnet endmembers, i.e. andradite and almandine. Spectra intensities measured at the Lα and Lβ flank positions for selected amphibole and biotite references can be quantitatively correlated with their Fe 2+ weight contents by a linear relation: Lβ/Lα = a + b*Fe 2+ , in which a and b are constants calibrated upon reference materials. It is emphasized that the constants a and b may vary significantly for different mineral groups and analytical conditions. Thus, a calibration is necessary for each analytical session. Our tests show that the potential beam damage during EPMA that may induce oxidation of the measured material can be minimized by using a large beam size (e.g., 20 μm). The Fe 2+ /ΣFe ratios determined using the EPMA flank method in collected natural calcic and sodic-calcic amphiboles (FeO T = 2.4–12.7 wt%, Fe 2+ /ΣFe = 0.45–75) and biotites (FeO T = 4.3-28.2 wt%, Fe 2+ /ΣFe = 0.67–0.87) are consistent within an error of ±0.1 with the data obtained from the colorimetric wet chemistry method. The EPMA flank method has the potential to be widely applied for accurate in-situ determination of Fe oxidation states in a wide range of minerals, provided that references are available.

AB - We present an analytical approach to determine the Fe 2+ /ΣFe ratio in amphiboles and biotites using electron probe microanalysis (EPMA). The “flank method” applied in this study is based on the fact that FeL line spectra show different peak position and intensity for Fe 2+ and Fe 3+ , which is associated with a resonant process involving both X-ray absorption and emission. The flank positions for Lα and Lβ, which correspond to sites with minimal Fe 3+ L 3 absorption and maximal Fe 2+ L 2 absorption respectively, are determined based on the L-line spectra difference of two Fe-rich ferric and ferrous garnet endmembers, i.e. andradite and almandine. Spectra intensities measured at the Lα and Lβ flank positions for selected amphibole and biotite references can be quantitatively correlated with their Fe 2+ weight contents by a linear relation: Lβ/Lα = a + b*Fe 2+ , in which a and b are constants calibrated upon reference materials. It is emphasized that the constants a and b may vary significantly for different mineral groups and analytical conditions. Thus, a calibration is necessary for each analytical session. Our tests show that the potential beam damage during EPMA that may induce oxidation of the measured material can be minimized by using a large beam size (e.g., 20 μm). The Fe 2+ /ΣFe ratios determined using the EPMA flank method in collected natural calcic and sodic-calcic amphiboles (FeO T = 2.4–12.7 wt%, Fe 2+ /ΣFe = 0.45–75) and biotites (FeO T = 4.3-28.2 wt%, Fe 2+ /ΣFe = 0.67–0.87) are consistent within an error of ±0.1 with the data obtained from the colorimetric wet chemistry method. The EPMA flank method has the potential to be widely applied for accurate in-situ determination of Fe oxidation states in a wide range of minerals, provided that references are available.

KW - Amphibole

KW - Biotite

KW - EPMA

KW - Fe oxidation state

KW - Fe /ΣFe ratio

KW - Flank method

UR - http://www.scopus.com/inward/record.url?scp=85061317364&partnerID=8YFLogxK

U2 - 10.1016/j.chemgeo.2019.01.009

DO - 10.1016/j.chemgeo.2019.01.009

M3 - Article

AN - SCOPUS:85061317364

VL - 509

SP - 152

EP - 162

JO - Chemical Geology

JF - Chemical Geology

SN - 0009-2541

ER -