Hi, I am interesting in doing DFT calculations of some elements at high-pressure, e.g. carbon.
Some pseudopotential (PP) generating schemes achieve superior total energy convergence versus planewave cutoff but at the "cost" of increasing r_c. The paper by Troullier and J. L. Martins PRB 43, 1993 (1991) states that if one matches the transferrabiliy of the pseudopotentials (I assuming through the logarithmic derivatives) then the radii used in the TM scheme are double that in the HSC scheme. The result is that the PP generated in the TM scheme is softer than the HSC. My question is: Will the PP give meaning results once the atomic seperation is less than r_c. In general, I would think not. So the next question is that for a fixed r_c what PP will be softess: TM, RRKJ, Ultrasoft-PP, etc.? I would appreciate comments and any relevant citations that I may have overlooked. 1613 Denise Dr. Apt. D Forest Hill, MD 21050 (217) 417-5210
