A Revised Low-Al Clinopyroxene-Liquid Geothermometer for High-Silica Igneous Systems

Abstract:

Increasingly we are able to investigate the pressure-temperature-composition histories of large, high-silica systems in order to understand their evolution and the timescales leading to eruption. A variety of minerals have been used to aid in this task, including feldspar, quartz, zircon, and pyroxene. The development of petrologic tools based on pyroxene is appealing due to the mineral’s widespread occurance in both felsic and mafic systems, as well as in meteorites. Pyroxene, which is an important mafic phase in many high-silica systems, can be used to probe the higher-temperature portion of the system’s history through geothermometry as well as diffusion chronometry, which greatly depends on the calculation of an accurate temperature.

In order to determine crystallization temperatures for specific minerals, it is essential to use a geothermometer that is appropriately calibrated for the compositional range of the mineral of interest. However, clinopyroxene (cpx) from many well-known high-silica systems are not included in existing cpx-opx and cpx-liquid geothermometer calibration datasets, which tend to focus on cpx from mafic systems. Also, these high-silica system cpx are of a composition that is not well represented in historical experimental data; they are often high in Fe and very low in Al₂O₃ (e.g., Yellowstone cpx Mg# = ~56, 0.53–0.73 wt% Al₂O₃; Bandelier Tuff cpx Mg# = ~27, 0.28–0.91 wt% Al₂O₃; Pantelleria cpx Mg# = ~43, 0.25–0.72 wt% Al₂O₃; Sierra la Primavera cpx Mg# = 0.22–0.29 wt% Al₂O₃; and Paektu Millenium cpx Mg# = ~26, 0.14–1.78 wt% Al₂O₃). We find that existing cpx-opx and cpx-liquid geothermometers do not work well for these cpx, and predict temperatures >50°C in error of low-Al cpx-saturated experiments.

A new regression of Putirka’s [RiMG, 2008] cpx-liquid geothermometer calibrated with 75 experimentally-derived cpx of a similar composition to that of Yellowstone post-caldera rhyolite cpx increases the geothermometer’s dependence on the Mg# and Na+K component of the liquid and decreases its dependence on the Ca+Si component of the liquid. This revised geothermometer shows a 5x improvement in uncertainty for experiments conducted at <850°C, and overall reproduces experimental conditions to ±20°C. Future experiments will add additional calibration data in the <850°C range.