New Releases by Todd Dunn

A rhyming picture book with pairs of things that go together exceptionally well, like horse and wagon and fire and dragon.

The results of investigations of the diffusion of oxygen in silicate melts, the structure of melts, and the redox equilibria in silicate melts are presented in this thesis. Five separate investigations were undertaken, all of which were directed toward gaining an understanding of the chemistry of silicate melts, and in particular the role played by oxygen in those melts. The diffusion of oxygen was investigated in six different melt compositions. Three of the compositions were in the synthetic system diopside-anorthite. Oxygen diffusion was measured in those melts at one bar pressure by means of isotopic exchange. The diffusivities determined are similar in magnitude to divalent cation diffusivities and obey the compensation law for the diffusion of divalent cations in silicate melts. Those observations suggest that there is a significant contribution to the diffusivity of oxygen from a "cation-like" diffusion mechanism. Eyring model calculations, using the oxygen diffusion data and melt viscosity data from the literature, suggest that the size of the average diffusing species is similar to that of the Si0 4 4 " anion. The diffusion of oxygen was also measured in three basaltic melts (an olivine nephelinite, an alkali basalt, and a tholeiitic basalt) at various temperatures and pressures up to 21 kilobars. The diffusivities were determined by monitoring the rate of reduction of Fe 3+ to Fe 2 + in the melts. The oxygen diffusivities measured are approximately the same as, or slightly greater than, divalent cation diffusivities in basaltic liquids. The diffusivity of oxygen shows an abrupt decrease in all three melts at approximately the same pressure as the change in the liquidus phase from olivine to pyroxene. The decreases in oxygen diffusivity are interpreted as being related to decreases in the proportion of 0 2 ' anion in the melts during reactions which decrease the proportion of olivine building units and increase the proportion of pyroxene building units in the melt. The results suggest that oxygen diffuses principally as the 0 2 ' anion in basaltic melts. The structure of lead orthosilicate melts was investigated as a function of the thermal history of the melt by means of infrared spectroscopy. The melts were observed to become increasingly polymerized with decreasing rate of cooling and with increasing duration of isothermal soaking. The presence of silicate anions larger than Si0 4 4- in the melts was demonstrated, which requires that 0 2 ' anions also be present. The proportion of 0 2 " anions in Pb 2 Si0 4 melts depends on the thermal history. Therefore, the ratio of non-bridging oxygens to tetrahedral cations (NBO/T) also depends on thermal history. The study of the structure of Pb 2 Si0 4 melts provides direct evidence of the presence of 0 2 ' anions in an orthosilicate melt. That result when combined with the implications of the diffusion studies suggests that 0 2 ' anions are present in much more polymerized melts and that the proportion of those anions in a given melt depends on the thermal history of the melt. That suggestion requires that melt models which do not incorporate 0 2 " anions be reassessed. The two investigations of iron redox equilibria in basaltic melts provide an improved predictive equation for the oxidation of iron which may account for (graphite and Fe°) in too oxidized to allow in natural melts and a model process the presence of reduced phases quenched rocks which are apparently those phases.