Magma, Magma, Quite Contaminated, How Does Your Garnet Grow?

Garnet in granitoid rocks has drawn considerable attention and discussion because of uncertainty surrounding its origins. For example, enrichment of Al, resulting in peraluminous magmas capable of crystallizing garnets, may be controlled by contamination or extreme differentiation; Mn enrichment in aplitic and pegmatitic phases suggests garnet may appear only at relatively low, near solidus temperatures. Peritectic garnet, grown by magma-wallrock reaction, may be confused with magmatic garnet, and xenocrysts of metamorphic garnet, entrained from wallrocks, further complicate interpretation. We address these uncertainties with the SIMS analysis of oxygen isotope variations in single garnet crystals and crystal populations in granitic rocks. Values of δ18O were measured on a CAMECA IMS 1280 using a 10 µm spot size and typical precision of ± 0.3 at 2 standard deviations. Analyses were corrected for instrumental mass fractionation according to the newly solved bias correction protocol for garnet (Page et al. 2010). Samples were collected from the Devonian Togus and Hallowell plutons in the south central Maine. These plutons are an ideal site for this study because they are peraluminous and contain pervasive garnet, they locally intrude pelitic, garnet-bearing wallrocks, and they have field evidence of xenolith entrainment and peritectic reaction of xenoliths and the host magmas. Garnet δ18O values of 7.5-10.5‰ show a large range of crustal input to host magmas. Crystal-to-crystal variation of δ18O in hand-samples varies up to 2‰, confirming that garnet populations have complex origins. Traverses (20-50 spots) of single crystals show that δ18O varies up to 1‰, with rims of crystals (outer 50-100µm) being up to 1‰ higher or lower than interiors. Increases of δ18O are interpreted as late-stage contamination, whereas lower δ18O rims, with correspondence to decreasing Fe/Mg ratio, suggest growth during falling magma temperature (50-100°C). Some garnet shows oscillatory variation in δ18O, recording discrete contamination events. Xenolith-hosted garnet has cores that are 1‰ higher than rims, matching values of isolated magmatic garnet in the same rock, and higher Ca cores (Xgrs ~ 0.05) than rims (Xgrs ~ 0.02) suggest crystallization at deeper crustal levels. Thus, garnet in xenoliths preserves a record initial growth of metamorphic garnet before or during entrainment of the xenolith, with subsequent and peritectic reaction causing more garnet to grow but in equilibrium with host magma. As a whole, the results from this study indicate that garnet exhibits previously unrecognized inter- and intra-crystalline δ18O heterogeneity, recording the dynamic crystallization histories; moreover, valuable complementary records of contamination can be recovered when δ18O of magmatic garnet is compared to xenocrystic garnet that survives assimilation. When garnet geochemical records are combined with geochemical and geochronologic records from zircon and other accessory minerals, a more complete record of magma origin can be obtained.