CRYSTAL-CHEMICAL CONTROLS ON ARGON RETENTION OF AMPHIBOLES FROM THE ADIRONDACK MOUNTAINS, NEW YORK, USA

Argon retention properties of hornblende have been studied using diverse metamorphic lithologies in the Adirondack Lowlands and Highlands — using mass spectrometry, microprobe, and X-ray diffraction. Incremental-heating of hornblendes and biotites constrain ~1100-900 Ma uplift, exhumation, and cooling in the Lowlands following the ~1270-1170 Ma Elzevirian orogeny. Ages of hornblendes and biotites define parallel younging trends of ~3 m.y./km across the ~45 km between the St. Lawrence River and the Carthage-Colton shear zone that separates the Lowlands from the Adirondack Highlands domain. These trends, especially in hornblende, are attributable to post-900 Ma tilting of the Lowlands domain by ~9° to the NW. Hornblende ⁴⁰Ar/³⁹Ar ages range from ~1125 to ~1000 Ma from NW to SE across the Lowlands, and age-distance regression indicates that ~67% of the age variance is attributable to the regional pattern (i.e., tilting) alone. The remaining ~33% apparently embodies other factors potentially related to Ar retention per se, such as microtexture and crystal chemistry. Indeed, the apparent age spread for a given locality is ~40 m.y., which, coupled with a known cooling rate of ~1.5 K/m.y., translates into a nominal ~60 K range in hornblende Ar closure temperature. The regional age-distance trend for hornblende resolves into three subparallel compositional trends: magnesio-(hastingsitic) (oldest), pargasitic-tschermakitic, and edenitic (youngest) — listed in order of decreasing Ar retention and increasing ionic porosity. Multiple linear regressions of ⁴⁰Ar/³⁹Ar age (dependent variable) versus distance and various combinations of ~35 crystal-chemical parameters (independent variables) have been performed. The most robust of these, statistically, relates age to a combination of distance, Mg#, Na content (A-site), Si content (T-site), and the beta angle. Another highly significant regression, in which ages are first normalized by sample location, also involves several Highlands hornblendes; this regression of normalized age incorporates the compositional terms listed above, plus bulk ionic porosity and several parameters specifically related to A-site geometry. These regressions both account for ~86-90% of the ⁴⁰Ar/³⁹Ar age variance observed among Lowlands (and Highlands) hornblendes, suggesting that Ar retention reflects crystal chemistry.