“…Let us now pay attention to the case of k 2 = −0.00258 (Figure 6c), which we believe to be the closest one to the real CsPbI 3 material, as aforementioned. One can first note that, for the smallest temperatures, v Rx,MF, Pnma and v Rx,HF, Pnma adopt values close to 24.0 and 34.0 cm −1 , respectively, which are very close to the first-principles results of 26.0 and 35.0 cm −1 for two low-in-frequency modes involving both iodine octahedra and displacements of Cs cations, [36] which attests of the accuracy of our numerical tools. Moreover, our results for the case of k 2 = −0.00258 is in overall qualitatively similar to that of k 2 = −0.00698 (Figure 6b) except for two noticeable features: v Mz,LF, P4/mbm and v Rx,LF, P4/mbm adopt very small values of about 8.0 and 1.0 cm −1 in the entire range of the intermediate P4/mbm phase.…”