Final Report Summary - DEGASSING METALS (The role of diffusion in volcanic metal emissions)
The role of diffusion in volcanic metal emissions
Background:
Volcanic eruptions are driven by the exsolution of volcanic gases. Volatile components such as H2O, CO2, SO2, H2S and others are dissolved in magma at depth, but their solubility depends on pressure and thus these volatiles form gas bubbles when magma ascends. Gas bubbles lend buoyancy to magma, speeding its ascent to the surface, culminating in eruption. However when the gas escapes from magma, its ascent slows due to loss of buoyancy compounded by the resulting increase in viscosity and crystallization. Thus gas escape can lead to a diminished explosivity or even an end to magma ascent.
Gas escape requires permeability of magma to gases. In its simplest form, individual bubbles can rise through the magma, but more often this requires stress to deform bubbles and even fracture the magma. Gas escape routes in viscous magma thus often consist of transient bubble networks and fractures. All magma will experience some gas escape, but it is the efficiency of that escape that determines whether and how magma arrives at the surface. This study has explored the use of trace metals to study the kinetics of gas escape.
Many trace metals are naturally enriched in volcanic gases (e.g. Nriagu 1989; Hinkley et al., 1999; Pyle and Mather, 2003), it is the reason why ancient volcanic systems are often associated with ore deposits, but is also a source of local and global natural pollution of the surface environment. During gas escape the gas will continue to chemically interact with the melt in a game of catch-up. Eruption quenches the melt, freezing the chemical diversity resulting from this interaction. This study aimed to unlock the potential of this archive and provide constraints on the efficiency of gas escape.
Research activity:
The main aim of this study was to assess the extent of gas-melt interaction and how such traces can be used to study the kinetics of magma degassing. The methodology included experimental determination of diffusion coefficients for a range of metals and observation of diffusion profiles in gas pathways in natural volcanic rocks. The determination of gas-melt partition coefficients is the third step, which will be done in a follow-up project.
Diffusion coefficients: Diffusion coefficients were determined using a diffusion couple approach (e.g. Koepke and Behrens, 2001), whereby two glass cylinders, one of which is doped with metals, are juxtaposed and heated for a fixed amount of time to allow diffusion of the metals from one cylinder to the other. Experiments were conducted in rapid quench cold-seal pressure vessels on hydrous rhyolitic glasses at 850, 1000 and 1150oC and 150-200 MPa. The experimental run products were sectioned perpendicular to the interface of the two cylinders and analysed by electron microprobe, synchrotron XRF and LA-ICP-MS to determine the diffusivities of Br, As, Zn, Pb, Bi, Tl, Sb, Cd, Mo, and In. The Arrhenius parameters are indicated below. Experiments were also conducted with different concentrations of ligands, H2O, S, Cl, to assess the effects of the presence of these ligands on the mobility of the metals with which they form complexes in the gas. Preliminary data confirm the well-known (e.g. Zhang et al., 2010) increase in diffusivity with H2O concentration through its effect on the melt viscosity, but suggest that there is only a minor effect for Cl and a small effect for S. More experiments are needed to explore this further.
Arrhenius parameters for a variety of metals and Br. These are the first As, In, Bi, Mo and Br diffusivities for hydrous rhyolite.
Observations of metals along natural gas pathways: Four examples of gas pathways were selected for analysis: a bomb with tubular vesicles from Lipari, tuffisite veins from Torfajokull and Chaiten and a banded obsidian from Chaiten. All of these features have seen interaction between gas and melt and one of the tuffisite veins and the banded obsidian have already gone some way to welding and erasing the gas pathway. All of the studied examples showed disturbances in the metal concentrations. The tubular vesicles from Lipari are enriched in Cu and depleted in Zn and others along the interface. However unfortunately these samples proved to be hydrated and the extent of hydration coincides with the extent of the enrichments and depletions, suggesting that the same post-magmatic process caused the disturbance. This observation indicates that this process was not governed by diffusion and more importantly it provides a way to distinguish primary and secondary elemental profiles. Interestingly, this sample did retain a magma degassing profile, which extends beyond the later overprint, but only for H2O, the fastest diffusing component.
The most spectacular result is that of the tuffisite veins. These proved to be either enriched or depleted in metals, but both show a very clear signal of magma degassing and gas transport. These enrichments can be used to estimate gas fluxes through the veins and lifespans of individual veins. These calculations suggest that tuffisite veins are highly efficient gas pathways that transport gas from previously exsolved gas pockets.
Fractures and veins are transient features with lifespans of minutes to days; eventually welding causes impermeability and gas transport ceases. Texturally the gas pathway lingers in the form of banding and crystal orientations, such as in banded obsidians. This study has shown that a chemical signal lingers as well and beyond the textural legacy of magma degassing and has shown a way to extract timescales from the chemical signals.
Outcome and impact:
Gas pathways are associated with significant disturbances in metal concentrations: magma degassing causes chemical heterogeneity. Traces of this gas-melt interaction can be used to provide information on the kinetics of magma degassing. Diffusivities of a range of metals As, Zn, Pb, Bi, Tl, Sb, Cd, Mo, and In cover several orders of magnitude providing a large window into past degassing processes. To calculate how these differences in diffusivities affect gas compositions, or vice versa provide information on magma ascent, better partition coefficient data is needed. A proposal based on these results has been submitted for an ERC starting grant to continue this work. The results of this study have no immediate socio-economic impact, however the results of follow-up work will be highly relevant to volcanic hazard assessment and economic geology. Additionally there may be comparisons with industrial applications such as waste incineration, nuclear waste storage and ore processing.
References:
Hinkley et al (1999) EPSL 170:315-325; Koepke and Behrens (2001) GCA 65:1481-1498; Nriagu (1989) Nature 338:47-49; Pyle and Mather (2003) Atm. Environ. 37:5115-5124; Zhang et al (2010) Rev in Mineral and Geochem 72:311-408.
Background:
Volcanic eruptions are driven by the exsolution of volcanic gases. Volatile components such as H2O, CO2, SO2, H2S and others are dissolved in magma at depth, but their solubility depends on pressure and thus these volatiles form gas bubbles when magma ascends. Gas bubbles lend buoyancy to magma, speeding its ascent to the surface, culminating in eruption. However when the gas escapes from magma, its ascent slows due to loss of buoyancy compounded by the resulting increase in viscosity and crystallization. Thus gas escape can lead to a diminished explosivity or even an end to magma ascent.
Gas escape requires permeability of magma to gases. In its simplest form, individual bubbles can rise through the magma, but more often this requires stress to deform bubbles and even fracture the magma. Gas escape routes in viscous magma thus often consist of transient bubble networks and fractures. All magma will experience some gas escape, but it is the efficiency of that escape that determines whether and how magma arrives at the surface. This study has explored the use of trace metals to study the kinetics of gas escape.
Many trace metals are naturally enriched in volcanic gases (e.g. Nriagu 1989; Hinkley et al., 1999; Pyle and Mather, 2003), it is the reason why ancient volcanic systems are often associated with ore deposits, but is also a source of local and global natural pollution of the surface environment. During gas escape the gas will continue to chemically interact with the melt in a game of catch-up. Eruption quenches the melt, freezing the chemical diversity resulting from this interaction. This study aimed to unlock the potential of this archive and provide constraints on the efficiency of gas escape.
Research activity:
The main aim of this study was to assess the extent of gas-melt interaction and how such traces can be used to study the kinetics of magma degassing. The methodology included experimental determination of diffusion coefficients for a range of metals and observation of diffusion profiles in gas pathways in natural volcanic rocks. The determination of gas-melt partition coefficients is the third step, which will be done in a follow-up project.
Diffusion coefficients: Diffusion coefficients were determined using a diffusion couple approach (e.g. Koepke and Behrens, 2001), whereby two glass cylinders, one of which is doped with metals, are juxtaposed and heated for a fixed amount of time to allow diffusion of the metals from one cylinder to the other. Experiments were conducted in rapid quench cold-seal pressure vessels on hydrous rhyolitic glasses at 850, 1000 and 1150oC and 150-200 MPa. The experimental run products were sectioned perpendicular to the interface of the two cylinders and analysed by electron microprobe, synchrotron XRF and LA-ICP-MS to determine the diffusivities of Br, As, Zn, Pb, Bi, Tl, Sb, Cd, Mo, and In. The Arrhenius parameters are indicated below. Experiments were also conducted with different concentrations of ligands, H2O, S, Cl, to assess the effects of the presence of these ligands on the mobility of the metals with which they form complexes in the gas. Preliminary data confirm the well-known (e.g. Zhang et al., 2010) increase in diffusivity with H2O concentration through its effect on the melt viscosity, but suggest that there is only a minor effect for Cl and a small effect for S. More experiments are needed to explore this further.
Arrhenius parameters for a variety of metals and Br. These are the first As, In, Bi, Mo and Br diffusivities for hydrous rhyolite.
Observations of metals along natural gas pathways: Four examples of gas pathways were selected for analysis: a bomb with tubular vesicles from Lipari, tuffisite veins from Torfajokull and Chaiten and a banded obsidian from Chaiten. All of these features have seen interaction between gas and melt and one of the tuffisite veins and the banded obsidian have already gone some way to welding and erasing the gas pathway. All of the studied examples showed disturbances in the metal concentrations. The tubular vesicles from Lipari are enriched in Cu and depleted in Zn and others along the interface. However unfortunately these samples proved to be hydrated and the extent of hydration coincides with the extent of the enrichments and depletions, suggesting that the same post-magmatic process caused the disturbance. This observation indicates that this process was not governed by diffusion and more importantly it provides a way to distinguish primary and secondary elemental profiles. Interestingly, this sample did retain a magma degassing profile, which extends beyond the later overprint, but only for H2O, the fastest diffusing component.
The most spectacular result is that of the tuffisite veins. These proved to be either enriched or depleted in metals, but both show a very clear signal of magma degassing and gas transport. These enrichments can be used to estimate gas fluxes through the veins and lifespans of individual veins. These calculations suggest that tuffisite veins are highly efficient gas pathways that transport gas from previously exsolved gas pockets.
Fractures and veins are transient features with lifespans of minutes to days; eventually welding causes impermeability and gas transport ceases. Texturally the gas pathway lingers in the form of banding and crystal orientations, such as in banded obsidians. This study has shown that a chemical signal lingers as well and beyond the textural legacy of magma degassing and has shown a way to extract timescales from the chemical signals.
Outcome and impact:
Gas pathways are associated with significant disturbances in metal concentrations: magma degassing causes chemical heterogeneity. Traces of this gas-melt interaction can be used to provide information on the kinetics of magma degassing. Diffusivities of a range of metals As, Zn, Pb, Bi, Tl, Sb, Cd, Mo, and In cover several orders of magnitude providing a large window into past degassing processes. To calculate how these differences in diffusivities affect gas compositions, or vice versa provide information on magma ascent, better partition coefficient data is needed. A proposal based on these results has been submitted for an ERC starting grant to continue this work. The results of this study have no immediate socio-economic impact, however the results of follow-up work will be highly relevant to volcanic hazard assessment and economic geology. Additionally there may be comparisons with industrial applications such as waste incineration, nuclear waste storage and ore processing.
References:
Hinkley et al (1999) EPSL 170:315-325; Koepke and Behrens (2001) GCA 65:1481-1498; Nriagu (1989) Nature 338:47-49; Pyle and Mather (2003) Atm. Environ. 37:5115-5124; Zhang et al (2010) Rev in Mineral and Geochem 72:311-408.