『Abstract
　Prior transmission electron microscope studies showed that the surface geometry of olivine changes dramatically during natural chemical weathering. However, similar morphological evolution has not been reported in laboratory studies of olivine dissolution. In this study, we examined the development of fayalite (Fe2SiO4) surface morphology during both abiotic and biotic (using Acidithiobacillus ferrooxidans) laboratory dissolution experiments at an initial pH of 2.0. The fayalite came from Cheyenne Canyon, Colorado (Smithsonian # R 3516) and contains a few percent laihunite (olivine structure with ordered ferric iron and vacancies, 〜Fe0.8²⁺Fe0.8³⁺SiO4). High-resolution field emission low voltage scanning electron microscope (SEM) characterization of all reacted samples showed etch patterns consistent with those reported from naturally reacted olivine. High-resolution transmission electron microscope (HRTEM) data demonstrated pervasive channeling on (001), with channel spacings that range down to ＜10 nm. Formation of channel on (001) is probably initiated by preferential removal of cations from olivine M1 sites. Channeling confers at least an order of magnitude increase in surface area. Relict strips of olivine between channels contain laihumite layers that are oriented parallel to channel margins. X-ray diffraction analyses indicated that the relative abundance of laihunite is higher in reacted compared to unreacted samples. This result is consistent with prior studies of naturally weathered olivine that suggest that laihunite is far less readily dissolved than olivine.
　Samples reacted in the presence of A. ferrooxidans cells that enzymatically oxidized iron, or in solutions where ferric iron was added to simulate biological activity, dissolve at a much slower rate than samples reacted abiotically. We attribute suppression of the olivine dissolution rate to surface adsorption of Fe³⁺. It is probable that ferric iron adsorption is controlled by M2 sites in the underlying olivine structure. If this is coupled with removal of M1 cations during channel formation, then a modified laihunite-like surface will develop (vacancies in laihunite are on M1 sites). Although surface modification might only penetrate a few atomic layers, an inherently unreactive laihunite-like surface structure could explain both the pervasive channeling and the dramatic suppression of the measured dissolution rate.』

1. Introduction
2. Material and methods
3. Results
　3.1. Abiotic experiments
　3.2. Biologic experiments
　3.3. Naturally weathered fayalite from fayalite granite
4. Discussion
Acknowledgments
References