Comment on "Colossal Pressure-Induced Softening in Scandium Fluoride".

The results reported by Wei et al. [Phys. Rev. Lett. 124, 255502 (2020)] can be confronted with predictive, quantitative theories of negative thermal expansion (NTE) and pressure-induced softening, allowing to corroborate, or invalidate certain approaches. Motivated to corroborate the quantitative predictions of the recent Coulomb Floppy Network (CFN) microscopic theory of vibrational and thermomechanical properties of empty perovskite crystals [Tkachenko and Zaliznyak, arXiv:1908.11643 (2019)], we compared theory prediction for the mean-squared transverse displacement of the F atoms, U$_{perp}$, with that reported in Fig. 5 of Wei et al. and observed a marked discrepancy (an order-of-magnitude larger than the error bar). We then compared these results with the previously published Xray diffraction data of Greve, et al. [JACS 132, 15496 (2010)] and the neutron diffraction data of Wendt, et al. [Science Advances 5 (2019), 10.1126/sciadv.aay2748]. We found the latter two data sets to be in a good agreement with each other, as well as with the prediction of CFN theory. We thus conclude that U$_{perp}$ values reported in Fig. 5 of Wei et al. are substantially incorrect. The purpose of this Comment is twofold: (i) to caution the researchers against using the U$_{perp}$ data of Wei et al. for quantitative comparisons with theory, and (ii) to encourage Wei et al. to reconsider their analysis and obtain a reliable U$_{perp}$ data by better accounting for the beam transmission and attenuation effects.