2005 Salt Lake City Annual Meeting (October 16–19, 2005)

Paper No. 2
Presentation Time: 3:45 PM

WHERE HAVE ALL THE RARE-EARTHS GONE? - A GRAIN-BOUNDARY TRACE-ELEMENT RESERVOIR IN METAMORPHIC ROCKS


CORRIE, Stacey and KOHN, Matthew, Geological Sciences, University of South Carolina, EWS 617, 701 Sumter St, Columbia, SC 29208, scorrie@geol.sc.edu

Accessory mineral stability is commonly assumed to depend on trace elements hosted by both accessory and major minerals. However, in graphitic rocks of Great Smoky Mountains, NC, trace element analysis of whole-rocks, in-situ laser ablation ICP-MS analysis of major silicates, and observed accessory mineral abundances indicate a major discrepancy in LREE, U, and Th budgets. The whole-rock concentrations of LREE, Th and U are each at least an order of magnitude greater than can explained by mineral compositions and abundances alone. Accessory minerals (ap, zrc) and major silicates (qtz, pl, ms, bt, grt) can account for no more than 10 ppm LREE, whereas whole-rock LREE concentrations are 200-300 ppm. Because there are no recognizable LREE accessory minerals, this observation points to a major, unrecognized reservoir of LREE. A similar discrepancy occurs for Th and U, in that measured mineral compositions can account for only ~100 ppb for both Th and U, vs. measured whole-rock concentrations of 15-20 ppm Th and 3-5 ppm U. In contrast, whole-rock HREE contents are accounted for by garnet and zircon to within a factor of ~2. Although we cannot rule out uniformly distributed sub-micron sized LREE-Th-U minerals, one intriguing possibility is that these elements are not in any specific mineral, but rather simply adsorbed onto mineral surfaces. These elements adsorb strongly onto mineral surfaces in marine sediments, and our data suggest this adsorption mechanism may carry through to amphibolite-facies conditions. Assuming crystal thicknesses of 100 nm (based on observed graphite grains), and an estimated boundary layer thickness of 1 nm, the amount of LREE that would have to be hosted on grain boundaries to account for whole-rock values is 5000-10000 ppm. If our hypothesis is correct, one important implication is that grain coarsening during deformation and/or prograde reactions could suddenly liberate LREE+Th+U, potentially stabilizing new accessory minerals (e.g., monazite) and/or producing distinct chemistries on existing grains. Thus, dissolution of earlier-formed grains or scavenging from major silicates may not be necessary to grow new accessory minerals.