REGIONAL GEOCHEMICAL CONSTRAINTS ON THE PETROGENESIS OF KYANITE QUARTZITES IN THE PIEDMONT PROVINCE OF VIRGINIA AND AN EVALUATION OF ANOMALOUS GA/AL VALUES

Kyanite quartzite is a minor but distinctive rock type that occurs throughout the Piedmont Province of central Virginia. The origin of these quartzites has been debated, with suggested protoliths ranging from aluminous sandstone to hydrothermally altered volcanic rock. As a follow-up to initial results for the economically important Willis and Baker Mt. deposits (Owens & Dickerson 2001), we present new observations and geochemical data bearing on this problem from all significant occurrences of kyanite quartzite in Virginia. Our whole-rock dataset consists of 23 samples from 9 separate localities, including Willis and Woods Mts. (Buckingham Co.), Baker and Leigh Mts. (Prince Edward Co.), near Madisonville (Charlotte Co.), and near Plato (Halifax Co.). All quartzites are mineralogically similar, and contain variable amounts of kyanite (10-50%), white mica (<10%; green Cr-muscovite at Madisonville), and rutile (<2%). Pyrite occurs sporadically, but is locally abundant. Topaz is a notable accessory in some samples. Major element compositions are dominated by SiO₂ and Al₂O₃, and show low, but variable TiO₂ (<1.7 wt%), Fe₂O₃-T (typically <2.5 wt%), and K₂O (<0.5 wt%). All other major element oxides occur in negligible amounts. Concentrations of many trace elements (Ni, Co, Zn, Ga, Nb, Y, Rb) are near or below detection limits in most samples. Levels of V and Sn correlate crudely with TiO₂. Samples from Madisonville contain extremely high Cr (~1200 ppm), a factor of 2 higher than Cr-rich samples from nearby Baker Mt. All other samples contain negligible Cr (<50 ppm). High-Cr rocks contain lower Zr and Zr/TiO₂ than low-Cr rocks, consistent with contrasting (mafic vs. felsic) igneous protoliths for each type. We further suggest that the current chemical compositions of these rocks reflect severe leaching of many elements in acidic, hydrothermal fluids. In support of this interpretation, we note that Ga/Al values in many samples are unusually low compared to most rocks, suggesting extreme Ga leaching. We hypothesize that hydrothermal fluid composition is an important factor leading to this Ga depletion, with H₂SO₄ and HF (as indicated by topaz) playing key roles. A few samples with normal Ga/Al indicate less-severe leaching, either because of different fluid compositions or relative location in the hydrothermal system.