# CH system, PBE0 and HSE06, interatomic distance optimization ndtset 3 #DATASET 1 PBE (the pseudopotentials for C and H are with PBE XC) #DATASET 2 PBE0 ixc2 41 getwfk2 1 getxred2 1 #DATASET 3 HSE06 ixc3 -428 getwfk3 2 getxred3 2 !For testing purpose, HSE XC functional ! has to be converged very precisely ! (because of libXC v3->v4 changes) nstep3 15 tolvrs3 1.d-20 #Common acell 5 2*4 Angstrom diemac 1.0d0 diemix 0.5d0 ecut 13 pawecutdg 35 ionmov 2 ntime 15 tolrff 0.005 tolmxf 5.0d-5 kptopt 1 ngkpt 1 1 1 shiftk 3*0.0 istwfk 1 natom 2 nsppol 2 spinmagntarget 1 nband 4 nbdbuf 0 nstep 10 ntypat 2 typat 1 2 xcart -0.562 0 0 0.562 0 0 Angstrom znucl 6 1 pp_dirpath "$ABI_PSPDIR" pseudos "Pseudodojo_paw_pbe_standard/C.xml, H4.GGA_X_PBE+GGA_C_PBE-paw.xml" #%% #%% [setup] #%% executable = abinit #%% [files] #%% files_to_test = #%% t69.out, tolnlines = 70, tolabs = 1.100e-03, tolrel = 3.000e-01, fld_options = -easy #%% [paral_info] #%% max_nprocs = 1 #%% [extra_info] #%% authors = X. Gonze #%% keywords = #%% description = #%% CH molecule spin-polarized, in a big box. #%% PAW : first PBE, then PBE0, then HSE06. Interatomic distance optimization. #%% Test the geometry convergence in the hybrid functional case #%% The results obtained with slightly better converged ecut (ecut 15) and box size (6 5 5 Angstrom) #%% compare favourably with those mentioned in Arnardi's report, as well as with those of VASP. #%% Values for the distance d, in Angstrom, for PBE, PBE0 and HSE06 respectively : #%% Present calculation (ecut 13, 5 4 4 box) : 1.130, 1.121, 1.122 Angstrom #%% Better calculation (ecut 15, 6 5 5 box) : 1.136, 1.125, 1.126 Angstrom #%% Even better calculation (ecut 18, 7 6 6 box) : 1.138, 1.126, 1.127 Angstrom #%% ABINIT Arnardi report (ecut acell unknown) : 1.136, 1.124, 1.124 Angstrom #%% VASP Arnardi report (ecut acell unknown) : 1.136, 1.124, NA Angstrom #%% Experimental : 1.120 Angstrom. #%% topics = Hybrids #%%