Volatile Exsolution Experiments: Sampling Exsolved Magmatic Fluids

Abstract:

In magmatic arcs the conditions of volatile exsolution exert a direct control on the composition of exsolved magmatic volatiles phases (MVPs), as well as on their parental magmas. The ability to accurately assess the exchange of major and trace elements between MVPs and magmas is key to understanding the evolution of arc magmas. The trace element signatures measured in arc volcanoes, fumaroles, and hydrothermal ore deposits are greatly influenced by the role of MVPs. In order to investigate the interplay and evolution of melts and MVPs we need experimental methods to simulate MVP exsolution that impose minimal external constraints on their equilibration.

Previous experiments have focused on evaluating the exchange of elements between aqueous fluids and silicate melts under equilibrium conditions^[1,2]. However, the large mass proportion of fluid to melt in these experiment designs is unrealistic. As a result, the idealized compositions of the aqueous fluids may exert a strong control on melt compositions for which they are out of equilibrium, especially at low melt fractions. In contrast, other experiments have focused on the melt during crystallization but must calculate MVP compositions by mass balance^[3]. In order to investigate MVPs and magmas during this critical period of MVP exsolution, we present a new two-stage fluid-melt experimental design.

Stage one experiments generate super-liquidus hydrous melts using Laguna del Maule rhyolites and dactites, as analogues for ascending arc magmas. Stage two experiments allow aliquots of stage one melt/glass to crystallize and exsolve MVPs. The design then uses pressure cycling to promote infiltration of in-situ fractured quartz^[4] and traps the MVPs as synthetic fluid inclusions. We present results from trial stage 2 experiments, which produced synthetic fluid inclusions consistent with literature values of fluid-melt Cl partitioning^[5] and of sufficient size for LA-ICPMS analysis. Trace element partitioning for Li, Na, K, Cu, Mo, and Au will also be presented.

[1] Candela P.A. and Holland H.D. (1984) GCA 48, 373–380 [2] Simon et al. (2006) GCA 70, 5583–5600. [3] Clemente B., Scaillet B., Pichavant M. (2004) Jour. Pet. 45, 2171-2196 [4] Sterner S.M. and Bodnar R.J. (1991) Am. J. Sci. 291, 1–54. [5] Webster J.D. and Holloway J.R. (1988) GCA 52, 2091–2105.