A comparison of melt densities for several Apollo volcanic glass compositions up to 11 GPa: Implications for the role of Ti in melt compressibility

This study focuses on the density and compressibility of three lunar glass compositions encountered during the Apollo missions to the Moon. These glasses are hypothesized to have rapidly quenched as glass beads during lunar fire fountain eruptions. The lunar glass beads have distinctive colors that correspond to TiO₂ content. The glasses of interest for this study are the Apollo 15 green glass Type C (A15C) which has a low TiO₂ content of only 0.26 wt%, the Apollo 17 orange glass (74220-type), which has a moderately high TiO₂ content of 9.12 wt%, and the Apollo 14 black glass which has the highest TiO₂ content with 16.40 wt%. These glasses represent primary, unfractionated melts making them a prime candidate for experimental studies into lunar basalt density and eruptability during partial melting of the lunar mantle. Sink/float experiments were performed on both A15C and the A17 74220-type, and previously published data from the A14 black glass were used. Experiments were performed in both a piston cylinder apparatus for pressures up to 2.5 GPa and a Walker-style multi-anvil device for pressures greater than 2.5 GPa. Thus far we have estimated the density and compressibility of the orange glass successfully up to 8 GPa with future experiments planned for 10 GPa. Although these pressures exceed that of the lunar interior (~4.7 GPa), higher experimental pressures are important for determination of melt bulk modulus and identification of such factors as compositional effects on melt compressibility. Experiments have also been run on A15C for pressures relevant to the lunar interior. Our sink/float observations at 8 GPa confirm that the orange glass density curve can be approximated by a straight line at liquidus temperatures, while the black glass shows a strong decrease in slope with pressure, especially above 3 GPa. The different slopes between these two glasses creates a density crossover at approximately 6 GPa. SiO₂, TiO₂, and FeO are the largest compositional differences between the volcanic glasses investigated; however Ti⁴⁺ has the ability to exist in the silicate melt both as a network former and a network modifier (whereas Fe²⁺ is typically a network modifier and Si⁴⁺ is typically a network former under the experimental conditions we investigate). Certainly, the pronounced flattening of the molten black glass compression curve at pressures above 3 GPa may be reflecting the completion of the coordination change to 6-fold in a large number of Ti-sites. If so, then one would expect to see a compressibility decrease as melt density experiences little or no boost from coordination change with higher compressions as we see in our orange glass results. Therefore, Ti⁴⁺ has the greatest potential to be the primary driving force behind the differences in compressibility and density observed in this and previous studies of the lunar picritic glasses.