Petrogenesis of Montana, USA Sapphires Inferred from Oxygen Isotopes and Zircon Inclusions

Abstract Montana hosts the largest sapphire deposits in the US, but the genesis of and connection among the various secondary and primary sapphire occurrences remains cryptic. In situ SIMS measurements of oxygen isotopes in sapphires and zircon inclusions in sapphires provide an opportunity to study the isotope and trace element geochemistry in order to understand sapphire-forming protoliths (i.e. crustal setting and alteration). Sapphire from Montana was transported as xenocrysts in carrier (host) magmas that resorbed sapphire exteriors during transport. The timing and nature of sapphire genesis is elucidated by SIMS measurements of trace elements and U–Pb from discrete zones in zircon inclusions with rims that are interpreted to be syngenetic with host sapphire. Montana sapphires exhibit a large range of δ18O values, from −3‰ to +12‰ VSMOW. However, all but two anomalous crystals fall in the range of 0‰ to 8‰. There is significant crystal-to-crystal variability yet averages at most deposits are consistent with high-temperature equilibration with the mantle (δ18O(Crn) = 4.4‰ to 5.7‰), with the exception of the commercial sapphire deposits at Rock Creek that average 2.7‰. Ruby analyses are limited, but typically have lower δ18O values compared to sapphires from the same detrital localities. Homogeneity within individual crystals (avg. 2 s = ±0.2‰) indicates the absence of isotopically distinct fluid or melt during crystallization. But intercrystalline δ18O ranges by up to 7‰ at a single locality, suggesting sapphire variability at a deposit reflects heterogeneity in the original protolith. Oxygen isotope fractionations between zircon rims and surrounding sapphire suggest comagmatic zircon inclusions and corundum equilibrated at high temperature. No correlation is seen for the degree of radiation damage and alteration of δ18O(Zrc) when zircon inclusions are surrounded and armored by sapphire. U–Pb ages and trace elements were measured in a small subset of syngenetic zircon inclusions in Dry Cottonwood Creek sapphires, revealing a Proterozoic (1778 ± 9 Ma) age for the protolith of sapphires at this locality and a likely polygenetic history. Previous work has suggested formation of these sapphires through partial melting of anorthosites and several anorthosites occur locally and match the age of zircon inclusion cores—the Boehls Butte anorthosite (~180 km NW of Rock Creek) and the Bitterroot anorthosite (~55 km W of Rock Creek) could correlate with Al-rich protoliths at depth. Proterozoic U–Pb ages of zircon from the Boehls Butte anorthosite (1787 ± 2 Ma) match well with the age of zircon inclusion cores in Dry Cottonwood Creek sapphires and suggest genesis in these or similar protoliths. Zircon rims with Tera-Wasserburg lower intercept ages of 110 ± 9 Ma are consistent with previous observations of a xenocrystic relationship to the ~50 Ma Eocene volcanic rocks. Corundum that formed over 50 Ma prior to being scavenged by Eocene magmas likely originated by the anatexis of Precambrian anorthosites and possibly other aluminum-rich rocks at depth.