Partitioning of excess argon between alkali feldspars and glass in a young volcanic system

Abstract

Listen

Translate article

Argon system behavior in a young volcanic system has been investigated through analysis of alkali feldspar phenocrysts and glass separates in order to constrain the potential effects of excess Ar upon 40Ar– 39Ar ages on feldspars and the crystal/melt partition coefficient. The importance of syn-and post-emplacement processes was investigated by sampling glass and pumice clasts derived from varying post-eruptive conditions, including welded sections, vapor-phase altered sections and heavily weathered portions of the ignimbrite. Excess 40Ar was detected in all glass samples at 1–10 ppb level, yet despite this, step-heating of co-existing alkali feldspar revealed a total gas age of 0.644 ± 0.038 Ma compared with the generally accepted age of 0.668 ± 0.004 Ma for the timing of emplacement of the Arico Formation. This suggests that despite high concentrations of Ar dissolved in the glass at quenching, the alkali feldspar crystallizing from excess Ar-rich melt contain very little excess Ar. No correlation was observed between excess Ar in the glass and total gas age obtained from the co-existing alkali feldspar and a partition coefficient (K D) between feldspar and glass was constrained to be less than 10 − 3 . A correlation between the concentrations of excess Ar and atmospheric derived 36Ar in the glass (manifested as a restricted range of 40Ar/ 36Ar up to ~ 320) suggests a uniform mechanism of incorporation of Ar from these reservoirs measured in all samples prior to eruption. In an assessment of post-eruptive factors, diffusional loss of Ar was not significant during the elevated temperatures (~ up to 670 °C) associated with welding processes at the relevant timescales over which welding occurs. Since diffusive exchange occurs between crystals and melt over long periods in the magma chamber, ages determined on volcanic alkali feldspars reflect the eruption age as a result of the very low K D, but in systems containing excess Ar the ultimate accuracy of such ages will be limited by the actual value of K D. We observe that biotite, a mineral with a higher K D, yields anomalously old ages in such systems, an observation made previously for other volcanic systems. Diffusive loss of Ar from volcanic glass occurs on timescales approximating that of volcanic eruptions, suggesting that excess Ar measured in the glass separates is representative of a degassed state, and indicates even higher pre-eruptive Ar concentrations present in the magma chamber prior to eruption. Thus our measured K D is likely to be a maximum value. The significance of physical contamination of feldspar by glass is also highlighted and quantified demonstrating that less than 0.01% by weight of adhered phonolite glass was required to elevate a measured feldspar age. This is particularly significant for young samples where any such contamination effect would be magnified proportionally to the small amount of in-grown radiogenic Ar.