Structure of arsenic selenide glasses by Raman and 77Se NMR with a multivariate curve resolution approach

Abstract

Solid-state Nuclear Magnetic Resonance and Raman data are usually reconstructed using devoted software. Most of the time, the experimental results must be compared to structural models obtained, for example, by Molecular Dynamics or Reverse Monte Carlo. This procedure is sophisticated, time consuming and requires specialized skills. In this paper, it is shown that chemometric unsupervised methods provide quickly interesting information from poorly resolved experimental data. In particular, it is shown that correlation maps and multivariate curve resolution enable to distinguish structural contributions and provide spectral reconstructions when applied to sets of 77Se solid-state NMR and Raman spectra of glasses with varying compositions in the AsxSe1−x glassy system, corroborating previous works. Interestingly, such an approach may provide a new way to assign unknown Raman vibration modes.