Corrigendum: Structure and Composition of the 200 K-Superconducting Phase of H2 S at Ultrahigh Pressure: the Perovskite (SH- )(H3 S+ ).
Elijah E Gordon,Ke Xu,Hongjun Xiang,Annette Bussmann-Holder,Reinhard K Kremer,Arndt Simon,Jürgen Köhler,Myung-Hwan Whangbo
DOI: https://doi.org/10.1002/anie.201603460
2016-01-01
Abstract:where the sulfur frameworks of H3S (s) and (SH−)(H3S+) (s) have the bcc structure (space group with a=3.089 Å),1,2 the authors of this Communication made an inadvertent error of using half the energy of H2 instead of the full energy. For the P4mm, Cmmm, Cmc21, and Ima2 structures of (SH−)(H3S+) (s), the correct ΔH values are −3.200, −3.511, −3.559, and −3.582 eV, respectively. Thus, at 200 K, the correct ΔG values are −3.219, −3.529, −3.578, and −3.601 eV, respectively. In arriving at these estimates, the energy of H2 calculated for its structure at ambient pressure was employed. To estimate the ΔH values at ultrahigh pressure P, it is necessary to consider the VP term, where V is the cell volume. In the calculations, the structure of H3S (s) and the structure of (SH−)(H3S+) (s) have the same cell volume, so their relative energies are not affected by the VP term. However, one needs to include the VP term for H2 considering the above-mentioned reaction. The ΔH estimated by using Eq. (1) with the energy of H2 under ambient condition, which is substantially negative, is most likely an underestimate for the ΔH for Eq. (4) because it is energetically favorable to dissociate H2 (g) into 2 H on Re metal surface even under moderate pressure. where S (s) refers to the β-Po structure of elemental sulfur.1 Calculations by the authors show that the ΔH for Eq. (7) under the ultrahigh pressure is exothermic. For example, under 160 GPa, the ΔH values for the P4mm, Cmmm, Cmc21, and Ima2 structures of (SH−)(H3S+) are −3.630, −3.163, −3.091, and −3.056 eV, respectively. Then, on the basis of Eqs. (1) and (7), one might speculate if all H2S under ultrahigh pressure is converted to elemental S. However, this reasoning is not correct because the reaction (2) is possible only on Re metal surface because hydrogen dissolution into the bulk of Re metal does not occur. Once the surface is saturated with Re−H bonds, the surface would not be active any longer. To summarize, once the catalytic effects of Re gasket metal surface on H2S (g) and H2 (g) are taken into consideration, it is most likely that the ΔH for Eq. (4) is substantially exothermic. It is energetically more favorable for the H atoms to make Re−H bonds on the surface of the Re metal than to enter the (SH−)(SH3+) (s) lattice to form H3S (s). The authors thank Dr. Igor Mazin and Dr. Noam Bernstein for alerting them to recheck their calculations.