Experimental investigation of the kinetics of Ostwald ripening of quartz in silicic melts

Abstract.

To characterise the kinetics of Ostwald ripening of quartz, we conducted four series of experiments in systems consisting of quartz, with an initial grain size of ≈1.3 to 6 µm, in equilibrium with hydrous silicic liquids. Two series were performed with a haplogranitic liquid containing 6.5 wt% H₂O at 900 °C and 1 GPa. The third series was made in the quartz–anorthite system at water saturation, 900 °C, and 1 GPa. The last series was made in the quartz–albite system at water saturation, 800 °C, and 0.2 GPa. In all series, we observed a relatively small but systematic increase of the mean grain size of quartz, % MathType!MTEF!2!1!+- % feaaeaart1ev0aaatCvAUfKttLearuavTnhis1MBaeXatLxBI9gBae % bbnrfifHhDYfgasaacH8YjY-vipgYlH8Gipec8Eeeu0xXdbba9frFj % 0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs0dXdbPYxe9 % vr0-vr0-vqpWqaaeaabiGaciaacaqabeaabaqaamaaaOqaaiqbdsga % Kzaaraaaaa!29AA! $\bar d$ , with increasing run duration, t. The largest increase was in the quartz–anorthite system: % MathType!MTEF!2!1!+- % feaaeaart1ev0aaatCvAUfKttLearuavTnhis1MBaeXatLxBI9gBae % bbnrfifHhDYfgasaacH8YjY-vipgYlH8Gipec8Eeeu0xXdbba9frFj % 0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs0dXdbPYxe9 % vr0-vr0-vqpWqaaeaabiGaciaacaqabeaabaqaamaaaOqaaiqbdsga % Kzaaraaaaa!29AA! $\bar d$ increased by a factor of 4.3 after 326 h; the smallest increase was measured in the quartz–albite system: % MathType!MTEF!2!1!+- % feaaeaart1ev0aaatCvAUfKttLearuavTnhis1MBaeXatLxBI9gBae % bbnrfifHhDYfgasaacH8YjY-vipgYlH8Gipec8Eeeu0xXdbba9frFj % 0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs0dXdbPYxe9 % vr0-vr0-vqpWqaaeaabiGaciaacaqabeaabaqaamaaaOqaaiqbdsga % Kzaaraaaaa!29AA! $\bar d$ increased by a factor of only ≈1.6 after 1,173 h. The experimental data yield very good linear fits in both ln t vs ln% MathType!MTEF!2!1!+- % feaaeaart1ev0aaatCvAUfKttLearuavTnhis1MBaeXatLxBI9gBae % bbnrfifHhDYfgasaacH8YjY-vipgYlH8Gipec8Eeeu0xXdbba9frFj % 0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs0dXdbPYxe9 % vr0-vr0-vqpWqaaeaabiGaciaacaqabeaabaqaamaaaOqaaiqbdsga % Kzaaraaaaa!29AA! $\bar d$ and ln t vs % MathType!MTEF!2!1!+- % feaaeaart1ev0aaatCvAUfKttLearuavTnhis1MBaeXatLxBI9gBae % bbnrfifHhDYfgasaacH8YjY-vipgYlH8Gipec8Eeeu0xXdbba9frFj % 0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs0dXdbPYxe9 % vr0-vr0-vqpWqaaeaabiGaciaacaqabeaabaqaamaaaOqaaiqbdsga % Kzaaraaaaa!29AA! $\bar d$ diagrams. The slopes in the ln t vs ln% MathType!MTEF!2!1!+- % feaaeaart1ev0aaatCvAUfKttLearuavTnhis1MBaeXatLxBI9gBae % bbnrfifHhDYfgasaacH8YjY-vipgYlH8Gipec8Eeeu0xXdbba9frFj % 0-OqFfea0dXdd9vqaq-JfrVkFHe9pgea0dXdar-Jb9hs0dXdbPYxe9 % vr0-vr0-vqpWqaaeaabiGaciaacaqabeaabaqaamaaaOqaaiqbdsga % Kzaaraaaaa!29AA! $\bar d$ diagrams, between 1/5 and 1/7, are, however, much smaller than the value of 1/3 predicted by the Lifshitz–Slyozov–Wagner theory for diffusion-controlled Ostwald ripening. A possible explanation for this discrepancy is that the diffusion-controlled regime in our experiments is only attained after a transient regime lasting from 20 h in series III to 100 h in series IV. A more straightforward explanation of the experimental results is that the rate-limiting mechanism for Ostwald ripening is quartz growth by surface nucleation not diffusion in the silicic liquid. Finally, we extrapolated our data to geological time scales to evaluate the importance of Ostwald ripening in natural quartz-bearing magmatic systems: (1) quartz cannot coarsen measurably by Ostwald ripening over reasonable time scales if the initial grain size is ≈1 mm or more; and (2) Ostwald ripening may be very active at the end of nucleation events and result in the consumption of a significant proportion of crystalline nuclei.