Week 28 Research Review

Analysis

Week 28 produced high nominal Tc values: 302.2 K for the benchmark set, 298.0 K for La/Al-substituted Li2(Mg,Ca)BH16 and 296.3 K for the Na/Be/La variant. These numbers are undermined by a calibration problem. The benchmark run reported about 300 K, well above experimental values for H3S (~200 K at ~155 GPa), LaH10 (~250 K at ~170 GPa) and CaH6 (~215 K at ~170 GPa), so the workflow appears to overestimate Tc by roughly 50–100 K. The explicit ordered-supercell calculations on Day 2 gave only 167.1 K. That is far below the fractional-occupancy results of Days 3–4, which suggests the 296–298 K values may come from virtual-crystal or disorder-averaging artifacts rather than real chemistry. The synthesis-route study on Day 5 gave a more moderate 228 K, but only at 174 GPa, and it did not establish thermodynamic accessibility. All five external reviews raise the same points: missing crystal structures, unclear case definitions, no stated functionals or pseudopotentials, no convergence tests, no convex-hull or phonon stability analysis, and incorrect handling of fractional stoichiometry. Week 29 should therefore move from broad screening to validation: calibrate the Tc workflow against experiment, rebuild the leading candidates as integer-stoichiometry ordered supercells, and test thermodynamic, dynamic and anharmonic stability. It should also generate synthesis targets and diffraction signatures that diamond-anvil-cell experiments can check.

Next Week Preview