TY - JOUR
T1 - Site-specific hydrogen diffusion rates in forsterite
AU - Padrón-Navarta, José Alberto
AU - Hermann, Joerg
AU - O'Neill, Hugh St C.
PY - 2014/4/15
Y1 - 2014/4/15
N2 - Synthetic hydroxylated titanium-bearing and undoped MgO-buffered forsterite crystals were heated at atmospheric pressure in air at temperatures from 800 to 1200°C to determine hydrogen diffusion as a function of the type of point-defect mechanism by which the hydroxyl is incorporated. OH-stretching bands, measured by infrared spectroscopy, were assigned to the four substitution mechanisms: Ti-clinohumite point defects, silicon and magnesium vacancies, and point defects associated with trivalent cations. In the experiments at 800 to 1000°C, hydrous defects related to trivalent cations and Mg-vacancies disappear first in the Ti-doped forsterite, followed by the hydrous defects related to Ti and a fraction of the hydrated Si-vacancies. Measured bulk diffusion coefficients for hydrogen exchange related to the latter process are one order of magnitude slower than previously reported and with higher activation energy (296 ± 14kJmol -1). After the elimination of the defects related to Ti, the hydroxyl concentration related to Si-vacancies remained constant for the duration of the experiments. This observation is in agreement with the very slow diffusivity of hydrogen in Si-vacancies measured in the undoped MgO-buffered dehydroxylation experiments at higher temperatures (from 1000 to 1200°C), which were ~3 orders of magnitude slower than previously reported at 1000°C, and with an activation energy of 461 ± 11kJmol -1. Hydrogen diffusion in forsterite is far more complex than previously assumed. This complexity can be used to gain more information on the timescales of the processes causing olivine to lose its structural "water", but quantitative modeling will require not only the knowledge of the intrinsic diffusivities of the different hydrous defects but also their relative proportions and the possible reactions between them.
AB - Synthetic hydroxylated titanium-bearing and undoped MgO-buffered forsterite crystals were heated at atmospheric pressure in air at temperatures from 800 to 1200°C to determine hydrogen diffusion as a function of the type of point-defect mechanism by which the hydroxyl is incorporated. OH-stretching bands, measured by infrared spectroscopy, were assigned to the four substitution mechanisms: Ti-clinohumite point defects, silicon and magnesium vacancies, and point defects associated with trivalent cations. In the experiments at 800 to 1000°C, hydrous defects related to trivalent cations and Mg-vacancies disappear first in the Ti-doped forsterite, followed by the hydrous defects related to Ti and a fraction of the hydrated Si-vacancies. Measured bulk diffusion coefficients for hydrogen exchange related to the latter process are one order of magnitude slower than previously reported and with higher activation energy (296 ± 14kJmol -1). After the elimination of the defects related to Ti, the hydroxyl concentration related to Si-vacancies remained constant for the duration of the experiments. This observation is in agreement with the very slow diffusivity of hydrogen in Si-vacancies measured in the undoped MgO-buffered dehydroxylation experiments at higher temperatures (from 1000 to 1200°C), which were ~3 orders of magnitude slower than previously reported at 1000°C, and with an activation energy of 461 ± 11kJmol -1. Hydrogen diffusion in forsterite is far more complex than previously assumed. This complexity can be used to gain more information on the timescales of the processes causing olivine to lose its structural "water", but quantitative modeling will require not only the knowledge of the intrinsic diffusivities of the different hydrous defects but also their relative proportions and the possible reactions between them.
KW - Deep Earth's water cycle
KW - Forsterite
KW - Hydrogen diffusion
KW - Hydrous point defects
KW - Silicon vacancies
KW - Titanium
UR - http://www.scopus.com/inward/record.url?scp=84897620507&partnerID=8YFLogxK
U2 - 10.1016/j.epsl.2014.01.055
DO - 10.1016/j.epsl.2014.01.055
M3 - Article
SN - 0012-821X
VL - 392
SP - 100
EP - 112
JO - Earth and Planetary Science Letters
JF - Earth and Planetary Science Letters
ER -