Excess argon in melt inclusions in zero-age anorthoclase feldspar from Mt. Erebus, Antarctica, as revealed by the [formula omitted] method

Abstract

Historically erupted (1984) anorthoclase phenocrysts from Mt. Erebus yield K Ar and 40 Ar 40 Ar apparent ages as old as 700 ka indicating the presence of excess argon. 40 Ar 39 Ar furnace step heating results from anorthoclase reveal a positive correlation between the Cl K ratio and apparent age. Because chlorine (up to 1700 ppm) is present in melt inclusions but not in the anorthoclase crystal lattice, this correlation suggests that the excess argon is associated with melt inclusions trapped within the anorthoclase during rapid crystal growth. Confirmation of the source of excess argon comes from step-heating experiments on multiple anorthoclase aliquots separated from two phenocrysts and one glass aliquot prepared from the matrix of a volcanic bomb. The anorthoclase phenocrysts were crushed and HF etched to yield aliquots containing 30%, 10%, and 1 % melt inclusions. The step-heated anorthoclase with 30% and 10% melt inclusions yielded the highest Cl K ratios and apparent integrated ages of 640 ± 30 ka and 179 ± 16 ka, respectively. The anorthoclase with I% melt inclusions yielded significantly lower Cl K ratios and apparent integrated ages of 48 ± 8 ka and 50 ± 30 ka. The step-heated volcanic glass yielded the least variable Cl K ratios and a total gas age of 101 ± 16 ka. Argon released from the anorthoclase and the trapped melt inclusions can be distinguished by differences in their degassing behavior, allowing geologically more reasonable ages to be obtained. Melt inclusions exposed on the exterior of anorthoclase grains principally degas during furnace extraction at temperatures less than 1200°C. Inclusions entirely within anorthoclase grains principally degas at temperatures greater than 1200°C when incongruent melting of the anorthoclase allows melt inclusion hosted excess argon to escape. Anorthoclase aliquots prepared with less than 1% inclusions can be fitted with a plateau for heating steps below 1200°C to yield ages as young as 8 ± 2 ka, whereas steps above 1200°C yield ages in excess of 100 ka. However, anorthoclase aliquots containing 10–30% melt inclusions yield ages in excess of 200 ka for heating steps below 1200°C. Minimizing the effects of the excess argon from melt inclusions relies on sample preparation and step-heating. Fine crushing and treatment with hydrofluoric acid removes many of the larger melt inclusions. Small melt inclusions which remain within the anorthoclase degas primarily at temperatures above 1200°C. Temperatures below 1200°C yield the most accurate ages. Attempts at post-analytically correcting for the chlorine-correlated excess argon are hindered by the variations in 40 Ar E Cl within and between samples. Elevated 40 Ar E Cl ratios in bubbles within the melt inclusions, as deduced from in vacuo crushing experiments, are the most likely cause for some or all of a sample's total 40 Ar E Cl variation. In addition, relative solubilities of argon and chlorine within phonolitic melts may be partly responsible for variations in 40 Ar E Cl .