This proposal is for renewed funding for a program of research on atomic-scale structure and dynam-ics of crystalline, glassy, and molten silicates and oxides of interest to the Earth Sciences. This program emphasizes the application of techniques of Nuclear Magnetic Resonance (NMR) to obtain fundamental data to help better understand and predict processes of interest in geochemistry, petrology, and geophys-ics. NMR is an element-specific experimental technique that can provide quantitative data on local struc-ture around atoms such as H, Li, B, C, O, F, Na, Mg, Al, Si, P, K (etc.) in crys-talline and amorphous sol-ids and liquids, out to distances of second atom neighbors and some-times beyond. It is also uniquely sen-sitive to atomic motions at relatively slow time scales (seconds to nanoseconds), that are particularly in-teresting in processes such as diffusion, phase transitions, and viscous flow. An unusual capability of this research effort is in situ, high tempera-ture NMR to temperatures up to 1500 degrees C. Under previous funding, this project has produced a wide range of results on coordination en-vironments of both cations and anions in glasses and melts, how composition, temperature and pressure change the structure, and how these changes affect thermodynamic and transport prop-erties. In particular, quantifying the extent of structural disorder in liquids, and relating this to configurational entropy and viscosity, have been emphasized; related questions of disorder in crystalline silicates and aluminosilicates have also been of high priority. During the current funding period, significant progress has been made in determining the degree of Al-Si ordering in aluminosilicate glasses and melts, including for the first time the use of 17O "triple quantum" NMR ("3QMAS") to directly count the proportions of Al-O-Al oxygens. A statistical thermodynamic model of the ordering, formulated independently from 29Si NMR data, is well supported by these results and allows prediction and comparison with data on properties. Other studies using this method have made new contributions to our understanding of the complex process of water dis-solution in aluminosilicate melts and to detect variations from strict "aluminum avoidance" in zeolites. For the latter, new NMR methods including "five quantum" techniques have shown dramatically en-hanced spectral resolution. A newly installed 14.1 Tesla spectrometer (600 MHz 1H frequency), and ac-cess to an 18.8 T instrument at Stanford (800 MHz), have allowed dramatic improvements in resolution and sensitivity, which have been especially useful in studying tiny (1 to 10 mg) high pressure samples to determine structural changes in melts and in mantle minerals such as Al-bearing MgSiO3 perovskite. Many of these projects will continue and be extended if this proposal is funded. A particular emphasis will be on quantifying the effects of temperature on melt structure, a critical and poorly understood issue that is at the heart of understanding melt properties. A major tool in this effort will be a recently con-structed fast quench apparatus that will allow sampling of the melt structure from a wide range of "fictive temperatures", even for small, expensive, isotopically enriched samples. Further in situ, high temperature NMR will be used to characterize dynamics in aluminosilicate melts directly. Studies will continue of disorder in crystalline aluminosilicates (including framework and sheet silicate minerals), and its effects on the site-specific rates of exchange among framework oxygens and H2O. Collaborations on high pres-sure glasses will continue, to explore simultaneous changes in both cation and anion coordination and dif-ferent P/T pathways for recording pressure effects on melt structure; collaborations on site occupancy and order/disorder in upper and lower mantle minerals will also be extended. New NMR methods and tech-nologies will continue to be developed which should have a wide range of applicability both within and outside of the Earth Sciences.
