SYENITE-HOSTED SAPPHIRES—PROBABLE CRUSTAL SAMPLING

The interpretation of the origin of sapphires has improved dramatically, as reported in recent reviews. However, those sapphires hosted in syenites and related igneous rocks (quartz absent), such as from Haliburton and Bancroft (Ontario, Canada), in the Kangayan syenite (Coimbatore district, Madras, India), Garba Tula (Kenya), Khibiny massif (Kola, Russia), and various locations in the Mogok Stone Tract, have remained a conundrum because the evolution of these igneous rocks is not consistent with crystallization of corundum from the melt. Models of formation include scavenging sapphires from crustal rocks through which the syenites pass during their intrusion, thus being xenocrysts, to metasomatic interactions between the hosts and fluids exsolved from the crystallization of the host and some adjacent rock depleted in silica. These latter metasomatic hosts, which are considered to be among the metamorphic sources associated with ultramafic rocks (serpentinites), have been referred to as “plumasite” (a plagioclase-corundum-mica rock, only different from syenite by the presence of corundum and lack of an igneous origin). Trace element analysis has been the method of choice to distinguish sapphire origin, although the data in the literature do not provide a unique distinction for syenite hosted sapphires from metamorphic and basalt hosted ones. New data from six different sources including three in the Mogok Stone tract provide some new insights.

Microprobe and LA-ICPMS analyses of trace elements place most of the sapphires as intermediate between basalt-hosted and metamorphic samples, whereas Cr+V values are too low in all samples to indicate an association with ultramafic rocks. Oxygen isotope values of coexisting feldspar + sapphire may or may not be in equilibrium, and spinel + sapphire do not appear to be in equilibrium, whether coeval or reaction products. Available δ¹⁸O values for sapphires for the entire suite range from +4.5 to +20.3, thus from mantle-like to O¹⁸-rich sedimentary associations, such as from marbles for the Mogok-sourced samples. This wide distribution suggests scavenging of sapphire from the crust by early syenitic magmatism, which may involve earlier metasomatic corundum formation. However, whether these processes are related to one another is unresolved.