V21A-3009
Time and temperature dependence of the re-equilibration processes in plagioclase-hosted melt inclusions

Tuesday, 15 December 2015
Poster Hall (Moscone South)
Melissa J Drignon1, Roger L Nielsen2 and Frank J Tepley III2, (1)Oregon State University, CEOAS, Corvallis, OR, United States, (2)Oregon State University, Corvallis, OR, United States
Abstract:
Interpretation of compositional data from melt inclusions and their host requires that we understand the effects of the two main post-entrapment processes. First, the crystals grow from the melt inclusion at the time the host lava erupts and cools. Second, the melt inclusion and its host undergo diffusive re-equilibration. These processes have been well established for olivine in olivine-hosted melt inclusions (Danyushevsky et al., 2000; 2004). No similar study has been done for plagioclase-hosted inclusions. To better understand post entrapment crystallization in these crystals, we have conducted a set of time-series experiments on plagioclases in Plagioclase Ultraphyric Basalts (PUBs, Nielsen et al., 1995; 1998; 2011; Sours-Page et al., 2000; Lange et al., 2012; 2013). Plagioclase crystals were heated near the entrapment temperature (1200-1230°C) for 30 min, 1h, 4 h and 4 days. Based on the S content of the melt inclusions, ~80% of the melt inclusions retain their integrity with respect to S after homogenization. It is based on the assumption that S will degas if the inclusion has ruptured, that all S is present as S=, and that the melts are initially sulfide saturated. This is not true for the major element composition of the melt inclusions. For run times of 30 min to 4 hours, the melt inclusions do not show significant compositional changes. However, after 4 days, the average composition of the re-homogenized melt inclusions is shifted toward higher MgO values and lower Al2O3 concentrations regardless of run temperature. These results suggest that re-homogenization techniques held at high temperature for less than 4 hours will return information that is most relevant to the re-homogenization of the post-entrapment crystals. The long term “drift” of the major elements suggests that diffusive re-equilibration with the host crystal takes place at a rate that is too slow to be seen during typical re-homogenization times used (e.g. 20 min-2 hrs.).

Key words: Plagioclase-hosted melt inclusions, PUBs, Time-series experiment