Format of the pseudopotential files (norm-conserving): ---top of the file--- **Identifier or 'dft', SEE BELOW 'Name', Zval, lmax, nlc, nnl, core_corr, lloc, bhstype If nlc=0 and nnl=0 (pseudopotentials in numerical form) Zmesh, xmin, dx, nmesh, number_of_wavefunctions For l=0,lmax: "Comment Line" (Vnl(i,l), i=1,mesh) End do Else if nlc.ne.0 .and. nlc.ne.0 (pseudopotentials in analytical form) alpha_loc(1), alpha_loc(2), c_loc(1), c_loc(2) For l=0,lmax: (alpha_l(i),i=1,3), (a_l(i),i=1,3), (b_l(i),i=1,3) End do If core_correction: a_nlcc, b_nlcc, alpha_nlcc End if Zmesh, xmin, dx, nmesh, number_of_wavefunctions End if For n=0,number_of_wavefunctions: "Comment Line" l, occupation number (chi(i,n), i=1,mesh) End do ---bottom of the file--- New format for gradient-corrected pseudopotentials: the first line contains 'dft' where dft is a string containing any of the following keywords (in any nonconflicting arrangement): "sla" Slater Exchange (default) "noexc" No Exchange "nocor" No Correlation "pz" Perdew-Zunger formula for LDA (default) "gl" Gunnarson-Lunqvist "hl" Hedin-Lunqvist "pw" Perdew-Wang "vwn" Vosko-Wilk-Nusair "wig" Wigner "lyp" Lee-Yang-Parr (correl. only) "obz" Ortiz-Ballone (for Perdew-Zunger formula) "obw" Ortiz-Ballone (for Perdew-Wang formula) "nogcx" No Gradient Correction on exchange (default) "nogcc" No Gradient Correction on correlation (default) "b88" Becke88 grad.corr. on exchange "p86" Perdew86 grad.corr. on correlation "ggx" GGA on exchange "ggc " GGA on correlation "bp" Becke88+Perdew86 Grad.Corr. "pw91" Perdew-Wang correl. + GGA "blyp" Lee-Yang-Parr functional + Becke88 If **Identifier is found, the Perdew-Zunger LDA is assumed