All-core PAW datasets for high-energy stopping simulations

Develop projector augmented-wave datasets for time-dependent density functional theory stopping-power simulations that explicitly treat all core states, thereby addressing the frozen-core approximation and capturing deep-core excitations that become important at high projectile energies.

Background

The simulations use a frozen-core approximation in which deeply bound core states are not included in the valence partition and therefore cannot be excited. The paper identifies this approximation as a primary source of the underestimation of stopping power at projectile energies above the Bragg peak, because core excitations and their coupling to the valence response can become significant.

Including the deepest core states would require extremely small partial-wave cut-off radii and correspondingly demanding spatial resolution within the GPAW framework. The paper therefore leaves the development of PAW datasets that explicitly treat all core states as future work to improve the accuracy of high-energy stopping-power calculations.