A sub-volcanic chemical pump: experimental and theoretical investigations of volatile element transport beneath volcanoes

Explosive volcanic eruptions are driven by the release of volcanic gases stored in underground magma bodies. These gases can carry an abundance of trace elements, some of which are released into the atmosphere, some are redistributed through the sub-volcanic magma body and others are concentrated and precipitated as hydrothermal ore deposits. Understanding how volcanic gases transport trace elements and where they go is therefore of central importance in our understanding of both volcano dynamics and hydrothermal ore formation. This proposal builds on considerable previous work on Mount St. Helens volcano in the USA. This work has documented a number of lines of evidence that certain volatile trace elements, such as Li and Pb, are efficiently moved from one part of the magma body to another. Why they are released from some parts of the magma body and accumulate in others is not well understood. The reason for this is that the ability of trace elements to dissolve in volcanic gases is not well constrained. A key unknown is the partitioning of trace elements between coexisting silicate melts and volcanic gas as a function of pressure, temperature and gas composition. A compounding problem is that some gases are homogeneous when released from magma at depth but condense to two separate phases at shallower pressure. Again, how trace elements partition between these separate phases is not well constrained, although it has been proposed that this type of vapour condensation is key to the formation of hydrothermal ore bodies. Based on our studies at Mount St. Helens we have hypothesised that the pressure drop in a magma body following major eruptions may lead first to condensation of vapour and then the redissolving of this vapour as the magma becomes repressurised at the end of the eruption. This constitutes a novel sort of 'chemical pump', which we want to test out by means of this proposal. To do this we plan to use high pressure and temperature experiments to determine the partitioning of selected trace elements between coexisting melt and vapour and to combine this with further studies of trace elements trapped in tiny globules of glass, known as melt inclusions, in volcanic crystals. The virtue of studying melt inclusions is that they effectively trap the pre-eruptive volatile inventory, including trace element, prior to eruption and therefore hold clues to the conditions under which magma was stored and how gases move around within this magma. Our experimental studies will be allied to dynamical models of volcanic processes using a new 2D computer code developed by the Visiting Researcher. This code has had considerable success in explaining many enigmatic features of volcano behaviour but has never before been applied to the problems of gas escape, gas redistribution and gas retention that occur during and after an eruption. We propose that the concentration of volatile trace elements in melt inclusions is a valuable archive of sub-volcanic degassing processes. Unlocking this archive requires the kind of integrated geochemical, petrological and theoretical study proposed here.