Despite recent advances in the field of N-isotope geochemistry, our understanding of the behavior of this element in the solid earth remains limited by a lack of fundamental information regarding the partitioning of ammonium and isotopic fractionation of N among coexisting mineral and fluid phases. Study of N behavior in regionally metamorphosed rocks provides the opportunity to assess intermineral NH4+ partitioning and N-isotope fractionation among coexisting micas during metamorphism and affords an application of the N system as a tracer of high-T fluid-rock interactions. Analyzed mica samples range in delta(15)N(air) from +3.3 to +11.9 parts per thousand, and contain 9 to 1820 ppm N. The outcrop at Townshend Dam, Vermont, allows examination of N behavior across-strike on a relatively small scale, and samples from western Maine demonstrate the effect of varying metamorphic conditions on N behavior in metapelites. Delta(15)N(bt-w.mica)(delta(15)N(biotite)-delta(15)N(white-mica)) ranges from -0.9 to + 2.7 parts per thousand (for all samples from both suites, mean = +0.36 parts per thousand, with 1 sigma = 0.79 parts per thousand), with samples containing a separate paragonite white-mica phase showing the greatest range (-0.12 to +1.02 parts per thousand; mean = 0.58 parts per thousand, 1 sigma = 1.03 parts per thousand). Thirteen samples containing only Na-poor muscovite (six from Townshend Dam, seven from Western Maine) have mean Delta(15)N(bt-w.mica) of 0.07 parts per thousand (1 sigma = 0.41 parts per thousand). In both suites, biotite nearly always contains more N than coexisting white mica, but N-w.mica/N-bt also shows some significant scatter (mean N-w.mica/N-bt = 0.46, with 1 sigma = 0.34). The thirteen samples, containing only a Na-poor, muscovitic white-mica phase, have mean N-w.mica/N-bt = 0.39 with 1 sigma = 0.26, similar to that reported by others for other metamorphic suites containing only muscovite as the white-mica phase. There is no obvious suggestion of equilibrium N-isotopic fractionation among coexisting micas at epidote-amphibolite to amphibolite-facies metamorphic conditions, although NH4+ appears to partition systematically among coexisting biotite and white mica. Significant scatter in both Delta(15)Nb(t-w.mica) and NH4+ partitioning (conceivably the result of differential closure to exchange during cooling or of retrograde replacements) could, however, obscure observation of small equilibrium intermica fractionations related to the characteristics of the interlayer sites in which NH4+ resides. Samples most unlike the mean in both Delta(15)N(bt-w.mica) and N-w.mica/N-bt contain abundant chlorite, some of which is likely retrograde, based on petrographic observations [chl/(chl+bt) >0.3]. Thus, retrograde replacement of biotite by chlorite may have been accompanied by fluid-mineral N-isotope exchange, perhaps involving the production of fine-grained, retrograde N-bearing white mica observed petrographically for some samples. It is also possible that sampling at scales greater than those of N-isotope equilibrium domains (e.g., across fine interlayers) results in some scatter because of varying relative modal proportions of the two micas in adjacent fine interlayers. Although further investigation of the extent of retrograde reequilibration of mica N systematics is warranted, the observed lack of systematic N-isotope fractionation among coexisting micas, and the reasonably systematic NH4+ partitioning data for these phases provide important preliminary constraints for attempts to model N-isotope behavior in fluid-rock systems. These results, and other attempts to calibrate N-isotope fractionation through field studies, point to the conspicuous lack of experimentally determined mica-fluid N-isotope fractionation factors. Copyright (C) 2000 Elsevier Science Ltd.
