Crystal Mush Dynamics of Mount St. Helens and Lassen Volcanic Center: Insights from Melt Inclusions and Titanium-In-Quartz Thermobarometry

In order to better understand magma chamber dynamics at Lassen Volcanic Center (LVC) and Mount St. Helens (MSH) I examine quartz cathodoluminescent textures along with trace element abundances in quartz and compositions of quartz-hosted melt inclusions from multiple samples over a wide range of ages (LVC: 116 ka-1915; MSH: 272 ka to 84 ka). I use melt inclusion compositions to estimate quartz crystallization temperatures using zircon saturation thermometry and titanium abundances in quartz are used to estimate quartz crystallization pressures. When examined from core to rim within an individual grain, these data, along with internal CL-zoning textures reveal the dynamic P-T-X history of the magma chamber in which the quartz grew. Cathodoluminescent images of LVC and MSH quartz reveal complex zoning patterns reflecting variability in Ti concentrations in quartz. The dominant zoning patterns can be divided into five major categories: (1) CL-bright rims truncating CL-dark core/mantle, (2) CL-dark rims/mantles truncating CL-bright cores, (3) oscillatory growth zones defined by fluctuations in CL-brightness with no significant core to rim distinctions, (4) dominantly homogenous grains, that display a general absence of distinct CL-defined growth, and (5) complex textures unique to individual samples. Within individual grains, changes in CL intensity (brightness) are correlated with changes in Ti concentration. However, there is no clear distinction in Ti concentration among different quartz categories with the exception of a few CL-bright quartz cores in category 2 quartz crystals from MSH and LVC that tend to have uniquely high Ti concentration. Overall, complex textures in individual quartz grains suggest dynamic quartz growth conditions. In LVC samples, nearly every quartz grain shows evidence of multiple dissolution episodes. Quartz grains from both volcanoes are dominantly anhedral to subhedral and have embayments. Some grains from the 1915-Lassen Peak-black dacite, the 27-ka-Lassen Peak-dacite eruption of Lassen Peak and the 40-50 ka-Hat Mountain-andesite are surrounded by pyroxene coronas. MSH quartz grains are also commonly partially dissolved but are relatively more likely to be euhedral, never have pyroxene coronas and have fewer embayments, particularly those from the oldest MSH sample examined. Heterogeneous textures in quartz from individual samples suggest that quartz grains with different growth histories are commonly erupted together. Quartz crystallization temperatures and pressures were estimated based on quartz-hosted melt inclusion compositions and Ti in quartz abundances. LVC quartz crystallization pressure and temperature estimates range from 652 to 777°C at 0.09 to 3.4 kbar and trends suggest that quartz crystallization pressures and temperatures have steadily decreased over the past 116 ka. At MSH, quartz crystallization conditions range from 672°C to 766°C at 0.5 to 5.8 kbar with a significant pressure decrease occurring between 272 ka and 147 ka. These temperatures reflect relatively cool storage compared to available eruption temperature estimates. Pressure and temperature variation within individual samples and individual grains coupled with complex CL-zoning further suggests that quartz phenocrysts grew in dynamic environments. Low temperatures recorded in quartz-hosted melt inclusions, in conjunction with rounded quartz grains imply that in most cases, quartz dissolution was the last event to occur before eruption, with no quartz growth during or immediately after the events that triggered eruption. Analysis of quartz-hosted melt inclusions reveals some compositional heterogeneities in LVC and MSH quartz stable magmas. At LVC, melt inclusion compositional variability is most likely a result of localized crystal dissolution caused by low-volume mafic intrusions resulting in compositionally heterogeneities throughout the magma body. At MSH, compositional variability among melt inclusions from different rock samples may reflect a poorly organized magma system consisting of distinct magma bodies. Overall, I propose a model for quartz crystallization in the LVC system in which quartz grains grow in relatively cold crystal mush frequently recharged by small scale localized mafic intrusions. These intrusions lead to localized dissolution resulting in compositional heterogeneities throughout the residual melt of the crystal mush. This compositional heterogeneity is reflected in compositional variability in quartz-hosted melt inclusions. Quartz grains in the intruded areas are partially dissolved. Some of the magma is remobilized and erupted, some magma is remobilized but not erupted, and some magma is not affected by intrusion events. Frequent remobilization of the magma leads to crystal recycling in which quartz grains are dissolved and remobilized by mafic intrusions. This process allows grains with different growth histories to be relocated proximal to one another and eventually erupted together. Quartz crystallization at MSH during the from 272 to 84 ka during the Ape Canyon stage more likely occurred in a poorly organized system of separate magma bodies with little to no mafic input. However, changes in quartz crystallization pressures over time may reflect the early stages of organization of the MSH magma body. Additionally, the increased abundance of internal dissolution textures in quartz over time may reflect dissolution from the input of ascending relatively hotter dacitic magmas