Mechanism of Hydrogen Isotope Exchange Between Hydrous Minerals and Molecular Hydrogen:
Ion Microprobe Study of D/H Exchange and
Calculations of Hydrogen Self-Diffusion Rates

Torsten W. Vennemann Institut für Mineralogie, Petrologie und Geochemie, Universität Tübingen,

Wilhelmstr. 56, 72074 Tübingen, Germany

James R. O'Neil Dept. of Geological Sciences, University of Michigan, Ann Arbor, MI 48109, USA

Etienne Deloule CRPG-CNRS, BP 20, 54501 Vandoeuvre les Nancy Cedex, France

Marc Chaussidon CRPG-CNRS, BP 20, 54501 Vandoeuvre les Nancy Cedex, France

Results of recent experiments have demonstrated that hydrogen isotope exchange between hydrous minerals and molecular hydrogen provides a viable method of determining hydrogen isotope fractionation factors at moderate temperatures of 150 to 400°C (Vennemann and O'Neil, 1996). To further our understanding of the mechanisms and rates of exchange in this system, we have made ion-probe measurements of the D/H ratios of exchanged and unexchanged minerals on a small scale (ª 30µm) and have calculated self-diffusion rates of hydrogen in hydrous minerals from bulk exchange experiments.

For the ion-probe measurements, hydrous minerals
were exchanged with molecular hydrogen enriched in HD (dD ª +5600”) for several days. The results for minerals containing Fe3+ (biotite and hornblende) on both orientated crystals and crystals analyzed at random orientation indicate that even though exchange occurs well into the interior of the crystals (4 µm or more), the dD values are extremely heterogeneous with some spots retaining dD values close to those of the starting composition, while adjacent spots have dD values of several thousand permil. No such heterogeneity is observed in unexchanged crystals. Subsequent electron microprobe analyses on the scale of about 10µm indicate that the chemical composition of the minerals remains constant. However, changes in the Fe2+/Fe3+ ratios were measured for bulk biotite and hornblende samples exchanged at 300°C and 400°C. Reduction of iron is also accompanied by a small but significant increase in the water content of the minerals (0.1 to 0.2wt%). No such reduction was observed in epidote, the only other major Fe-bearing species analyzed. Furthermore, strong crystallographic effects are present, with hydrogen diffusion parallel to the c-axis in hornblende being significantly faster than diffusion parallel to the a- or b-axis, a behaviour similar to that observed for oxygen diffusion in hornblende (Farver and Giletti, 1985).

Diffusion rates (D) and activation energies (Q) calculated for hydrogen self-diffusion in kaolinite, epidote and muscovite are remarkably similar to previous estimates (or extrapolations thereof) made on the basis of hydrogen isotope exchange experiments using water as the exchange medium. For kaolinite values of Q and the pre-exponential factor (D0) are 80 kJ/mol and 9.60¥10-6 cm2/sec, respectively, at calibration temperatures between 150 and 275°C, while for epidote Q = 77 kJ/mol and D0 = 1.99¥10-6 cm2/sec (150 to 400°C). For muscovite values of D vary from 6.10¥10-15 cm2/sec at 300°C to 2.95¥10-14 cm2/sec at 400°C (two points only). Rates of diffusion in biotite and hornblende (D ª 4¥10-14 cm2/sec at 300°C and 5¥10-13 cm2/sec at 400°C, for both), however, are about 2 orders of magnitude faster than previous estimates for systems with water as the exchange medium. In all cases the quoted values are given for diffusion in an infinite cylinder. An ion-probe depth profile parallel to the c-axis gives a diffusion rate of 7.37¥10-14 cm2/sec for hornblende exchanged at 400°C. The relatively fast rates of exchange in biotite and hornblende are likely controlled by chemical reactions, specifically the reduction of iron. The similarity of the rates of hydrogen self-diffusion in kaolinite, epidote and muscovite may be taken to suggest that the diffusing species in these experiments, as well as in experiments with water as the exchange medium, is the same. Present results, together with the observed decoupled hydrogen-oxygen exchange in kaolinite, illite, and montmorillonite as noted by O'Neil and Kharaka (1976), may imply that atomic hydrogen is the common diffusing species.


Farver, J. & Giletti B.J., Geochim. Cosmochim. Acta 49, 1403-1411 (1985).

O'Neil, J.R. & Kharaka, Geochim. Cosmochim. Acta 40, 241-246 (1976).

Vennemann, T.W. & O'Neil, J.R., Geochim. Cosmochim. Acta (submitted) (1996).