Fitting experimentally measured Raman bands with theoretical model profiles is the basic operation for numerical determination of Raman peak parameters. In order to investigate the effects of peak modeling using various algorithms on peak fitting results, the representative Raman bands of mineral crystals, glass, fluids as well as the emission lines from a fluorescent lamp, some of which were measured under ambient light whereas others under elevated pressure and temperature conditions, were fitted using Gaussian, Lorentzian, Gaussian–Lorentzian, Voigtian, Pearson type IV, and beta profiles. From the fitting results of the Raman bands investigated in this study, the fitted peak position, intensity, area and full width at half-maximum (FWHM) values of the measured Raman bands can vary significantly depending upon which peak profile function is used in the fitting, and the most appropriate fitting profile should be selected depending upon the nature of the Raman bands. Specifically, the symmetric Raman bands of mineral crystals and non-aqueous fluids are best fit using Gaussian–Lorentzian or Voigtian profiles, whereas the asymmetric Raman bands are best fit using Pearson type IV profiles. The asymmetric O–H stretching vibrations of H ₂ O and the Raman bands of soda–lime glass are best fit using several Gaussian profiles, whereas the emission lines from a florescent light are best fit using beta profiles. Multiple peaks that are not clearly separated can be fit simultaneously, provided the residuals in the fitting of one peak will not affect the fitting of the remaining peaks to a significant degree. Once the resolution of the Raman spectrometer has been properly accounted for, our findings show that the precision in peak position and intensity can be improved significantly by fitting the measured Raman peaks with appropriate profiles. Nevertheless, significant errors in peak position and intensity were still observed in the results from fitting of weak and wide Raman bands having unnormalized intensity/FWHM ratios lower than 200 counts/cm ⁻¹ .

Whereas the genesis of carbonatitic rare earth element (REE) deposits has long been a focus of study, the controls on mobilization and mineralization of REEs during magmatic-hydrothermal processes still remain open to debate. Here, we present our investigation of the dissolution and crystallization of REE (fluor)carbonate minerals in alkaline carbonate brine-melts up to 850°C and 11.6 kbar. Our results show that REEs are soluble in Na ₂ CO ₃ brine-melts, achieving concentrations exceeding 8 weight % at temperatures above 650°C. The addition of calcium and/or fluoride has minimal impact on REE mobilization, whereas introduction of silica suppresses REE solubilities by half, due to britholite formation above 550°C. Upon cooling, sodium and REEs combine to crystallize in burbankite or carbocernaite in sodium-enriched brine-melts, even at fluoride saturation. However, while the brine-melts contain substantial ferro- or aluminosilicate, REE mineralization in fluorcarbonates occurs after sufficient sodium precipitation in alkaline silicate minerals, hence revealing how silicate and sodium carbonate govern REE mineralization.

Carbonatites are the most important primary sources for the rare earth elements (REEs). While fractional crystallization of carbonate minerals results in the enrichment of volatiles, alkalis, and REEs in the remaining melts, the transition from carbonatitic magmas to hydrothermal brines remains unclear. Here, we investigated the pressure-temperature-composition (P-T-X) properties of the Na₂CO₃-H₂O system up to 700°C and 11.0 kbar using a hydrothermal diamond anvil cell and a Raman spectrometer. Our results show that Na₂CO₃becomes increasingly soluble under highP-Tconditions, leading to the disappearance of melt-fluid immiscibility and the continuous transition from Na₂CO₃melts to hydrothermal brines under deep crustal conditions. Given the abundance of Na₂CO₃in highly evolved carbonatitic systems, we suggest that the continuous melt-fluid transition in deep-seated carbonatites results in REEs being sufficiently concentrated in the brine-melts to form economic ore bodies, whereas in shallow systems, REEs preferentially partition into carbonatitic magmas over synmagmatic brines and disperse in carbonatite rocks that underwent limited fractionation.

The phase transitions involving calcite (CaCO₃-I), CaCO₃-II, CaCO₃-III and CaCO₃-IIIb were investigated using a diamond anvil cell and micro-Raman spectroscopy. Based on the results obtained from in situ observations and Raman measurements made with six natural calcite crystals, the phase transition from calcite to CaCO₃-II took place between 1.56 and 1.67 GPa under ambient temperature. Under a precise pressure of 1.97 ± 0.03 GPa, three CaCO₃ samples were observed to transform from CaCO₃-II directly to CaCO₃-III, while in the other three samples both CaCO₃-III and CaCO₃-IIIb crystal structures were detected. Transformation from CaCO₃-IIIb to CaCO₃-III was completed in a short period in one sample, whereas in the other two samples coexistence of CaCO₃-III and CaCO₃-IIIb was observed over a wide pressure range from 1.97 to 3.38 GPa, with sluggish transformation from CaCO₃-IIIb to CaCO₃-III being observed after the samples were preserved under 3.38 GPa for 72 h. Hence, it can be concluded that CaCO₃-IIIb is a metastable intermediate phase occurring during the reconstructive transformation from CaCO₃-II to CaCO₃-III. Splitting of the C–O in-plane bending (ν₄) and symmetric stretching (ν₁) vibrations and appearance of new lattice vibrations in the Raman spectra of CaCO₃-III and CaCO₃-IIIb suggest a lowering in crystal symmetry during the transformation from CaCO₃-II through CaCO₃-IIIb to CaCO₃-III, which is in good agreement with the observed sequence of phase symmetries.