Morphology and Raman spectral parameters of Bohemian microdiamonds: implications to elastic geothermobarometry

In this work, we combine the morphology and internal structure of northwestern Bohemian microdiamonds with their Raman spectral parameters to describe and understand their relationship. We evaluate our data according to the theory of elasticity and discuss implications for elastic geothermobarometry of diamond inclusions in garnet. We conclude that microdiamonds enclosed in kyanite, garnet and zircon differ in morphology and internal structure depending on the type of the host rock and host phase. Single crystal diamond octahedra in kyanite in the acidic gneiss show predominantly Raman shift towards higher wavenumbers (upshift), while single and polycrystalline diamonds enclosed in garnet and zircon in the intermediate garnet–clinopyroxene rock yield more variable Raman shift including a shift towards lower wavenumbers (downshift). This is consistent with closed boundaries between diamond and kyanite observed using FIB-TEM, while interfaces between diamond and garnet or zircon are commonly open. Moreover, higher variability in the Raman shift in diamond hosted by garnet or zircon may be caused by complex internal structure and the presence of other phases. At the same time, a diamond in kyanite features relatively high full-width-at-half-maximum ( FWHM ) due to the anisotropy of thermal contraction, which is reflected by the plastic deformation of diamond mediated by dislocation glide at T ≥ 1000 °C. The entrapment pressure ( P trap ) for diamonds in garnet was calculated using elastic geobarometry to test its compatibility with the existing peak pressure estimated by conventional thermobarometry. The “downshifted” diamonds exhibit entrapment pressures of 4.8 ± 0.14 and 4.99 ± 0.14 GPa at an entrapment temperature of 1100 °C, using unstrained reference diamond from the literature and own measurements, respectively. This is consistent with the earlier estimates and the elastic theory and does not require any elastic resetting suggested to account for the reported upshift in garnet. Our data suggest that the upshift in diamond hosted by garnet is related to the proximity of other diamond grains. We conclude that the use of diamond inclusions in elastic barometry should be backed by careful evaluation of its internal structure and associated phases and restricted to isometric monocrystalline diamond grains not occurring in clusters as required by the method.