Statistical approaches in the analysis of in situ thermo-Raman spectroscopic data for gypsum as a basis for studying dehydration and phase transformations in crystalline hydrates

The study of a gypsum single crystal was carried out using in situ thermo-Raman spectroscopy in the temperature range of 80-870 K. The conventional peak fitting approach was used to qualitatively determine the temperature ranges of gypsum dehydration and phase transitions. The dependence of the intermediate products of gypsum dehydration on the measurement conditions and the state of sample was confirmed. A detailed analysis of the temperature behavior of nu(1)(SO4) in a single crystal indicated the coexistence of CaSO4 center dot 2H(2)O + gamma CaSO4 (380-440 K) and gamma-CaSO4 + beta-CaSO4 (680-740 K) phases, which was explained by the reaction front movement in a solid. For the first time, an approach based on statistical algorithms was applied for processing datasets of experimental temperature-dependent spectra of crystalline hydrates, including the calculation of the Pearson's r coefficient, skewness, increment corr parameter, based on the autocorrelation function. The use of statistical methods for individual spectral regions made it possible to reveal the temperature differences in the pretransition regions for individual modes or their groups. In particular, the region of nu(2)(SO4) symmetric bending vibrations of sulfate ions was found sensitive to dehydration and accompanying phase transition at 410 K, beginning at lower temperatures (380 K) as compared with other regions, which is associated with the lack of the hydrogen bond influence on sulfate ions after the loss of water. Thus, statistical approaches are a source of additional information to clarify the quantitative indicators of the observed effects. The use of statistics is promising in the study of multistage processes of dehydration and phase transformations in crystalline hydrates in cases where peak fitting is inconvenient.