TY - JOUR
T1 - Experimental determination of cation diffusivities in aluminosilicate garnets - I. Experimental methods and interdiffusion data
AU - Elphick, Stephen C.
AU - Ganguly, Jibamitra
AU - Loomis, Timothy P.
PY - 1985/6
Y1 - 1985/6
N2 - We have carried out diffusion couple experiments using pairs of single crystals of natural garnet of dissimilar compositions in the range of 30-40 Kbar, 1,300-1,500° C, and measured the induced diffusion profiles by microprobe scanning across the interface. Significant modifications to, and experimentation with, the design of the pressure cell, furnace assembly and sample geometry were needed to obtain measurable volume diffusion at controlled P-T conditions. The diffusion profiles in the pyrope-almandine couples are short enough that retrieval of diffusion data from them must await deconvolution analysis to resolve the effect of spatial averaging of the microprobe beam. However, the profiles in the spessartine-almandine couples are sufficiently long to obviate convolution analysis. They yield interdiffusion coefficients (D) at 40 Kbar of D = 0.82×10-5 exp (-53.6±4.9 Kcal/RT) cm2/s and D=1.2×10-5 exp (-57.1±8.4 Kcal/RT) cm2/s for Fe-rich and Mn-rich compositions, respectively, and an activation volume of ∼4.7 cm3/mole. Preliminary analysis of profiles in a pyrope-almandine couple at ∼40 Kbar, 1,440°C suggests Fe-Mg interdiffusion to be an order of magnitude slower that Fe-Mn interdiffusion, and to increase with Fe/Mg ratio. The interdiffusion data reported here are in sharp disagreement with those of Freer (1979) and Duckworth and Freer (in Freer 1981) on Fe-Mn and Fe-Mg interdiffusion, respectively.
AB - We have carried out diffusion couple experiments using pairs of single crystals of natural garnet of dissimilar compositions in the range of 30-40 Kbar, 1,300-1,500° C, and measured the induced diffusion profiles by microprobe scanning across the interface. Significant modifications to, and experimentation with, the design of the pressure cell, furnace assembly and sample geometry were needed to obtain measurable volume diffusion at controlled P-T conditions. The diffusion profiles in the pyrope-almandine couples are short enough that retrieval of diffusion data from them must await deconvolution analysis to resolve the effect of spatial averaging of the microprobe beam. However, the profiles in the spessartine-almandine couples are sufficiently long to obviate convolution analysis. They yield interdiffusion coefficients (D) at 40 Kbar of D = 0.82×10-5 exp (-53.6±4.9 Kcal/RT) cm2/s and D=1.2×10-5 exp (-57.1±8.4 Kcal/RT) cm2/s for Fe-rich and Mn-rich compositions, respectively, and an activation volume of ∼4.7 cm3/mole. Preliminary analysis of profiles in a pyrope-almandine couple at ∼40 Kbar, 1,440°C suggests Fe-Mg interdiffusion to be an order of magnitude slower that Fe-Mn interdiffusion, and to increase with Fe/Mg ratio. The interdiffusion data reported here are in sharp disagreement with those of Freer (1979) and Duckworth and Freer (in Freer 1981) on Fe-Mn and Fe-Mg interdiffusion, respectively.
UR - http://www.scopus.com/inward/record.url?scp=0022202360&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0022202360&partnerID=8YFLogxK
U2 - 10.1007/BF00373039
DO - 10.1007/BF00373039
M3 - Article
AN - SCOPUS:0022202360
SN - 0010-7999
VL - 90
SP - 36
EP - 44
JO - Contributions to Mineralogy and Petrology
JF - Contributions to Mineralogy and Petrology
IS - 1
ER -