Dear all, we would like to share with you some work of ours (see https://www.materialscloud.org/discover/sssp/ and https://www.nature.com/articles/s41524-018-0127-2)
that we think it might be useful for the user community of Quantum ESPRESSO. Over the last years, we have made a substantial effort on testing pseudopotentials for precision and performance across several open-source libraries that are available for QE. In summary, we have developed a pseudopotential testing protocol that includes a verification part, based on the so-called delta-factor for elemental solids, and an efficiency part, based on plane-wave cutoff convergence tests for phonon frequencies, cohesive energies, pressures, and band structures. Adopting all these criteria, we have identified two optimal PBE pseudopotential libraries together with suggested wavefunction cutoffs for 85 elements of the periodic table, that we named "SSSP efficiency" and "SSSP precision" libraries (where SSSP stands for standard solid-state pseudopotential). The "SSSP efficiency" library is designed for practical applications that should remain affordable, and therefore pseudopotentials are chosen such that wavefunction cutoffs are as low as possible while keeping the precision reasonable. For instance this library is well suited for daily routine calculations or high-throughput materials screening. On the contrary, the "SSSP precision" library is more suited for high-precision materials modelling as it contains pseudopotentials that are the closest to all-electron calculations, with less consideration on the computational cost and the wavefunction cutoffs actually needed to converge all relevant quantities. All the results and all the calculations inputs and outputs are available on the Materials Cloud platform https://www.materialscloud.org/discover/sssp/. We stress that, apart from the plug-and-play SSSP libraries, our work provides also a useful database of verification data and convergence tests that facilitates the optimal choice of pseudopotentials and wavefunction cutoffs for custom applications. For example, some physical properties may be implemented only for some pseudopotential types or some applications may require convergence of just a subset of the quantities that we consider in the SSSP protocol. Indeed, by a look at our plots and data, a user can quickly select the optimal pseudopotential and wavefunction cutoff tailored for the specific application. We underline that the SSSP libraries are a collection of pseudopotentials generated by different authors. Please make an effort to acknowledge all original authors of libraries and methods if you use the SSSP results in your work, see here https://www.materialscloud.org/discover/sssp/table/efficiency#sssp-license for a tentative acknowledgement list. If you want to know more about our work, it is discussed in a recent article (https://www.nature.com/articles/s41524-018-0127-2) that has just been published. Thank you, The SSSP team EPFL, Lausanne
_______________________________________________ users mailing list [email protected] https://lists.quantum-espresso.org/mailman/listinfo/users
