# Molecular SiH4 in GWLS (G0W0 with Lanczos basis and Sternheimer equations) # There are 3 datasets # 1: Calculation of the Ground State density # 2: Calculation of the Ground State wavefunctions # 3: Calculation of the Self-Energy matrix element ndtset 6 ### Specific variables # Ground state calculation (density) toldfe1 1d-10 # paral_kgb1 0 # Parallel ground state calculations don't work for molecules (the minimisation algorithm becomes unstable). # Ground state calculation (wavefunctions) tolwfr2 1d-25 getden2 1 iscf2 -2 # paral_kgb2 0 # GWLS calculation for the HOMO tolwfr3 1d-25 # This is the precision to which the Sternheimer eqs. are solved in GWLS. getden3 1 getwfk3 2 optdriver3 66 # This is the optdriver triggering a GWLS calculation. gwls_band_index3 4 # This is the band for which the matrix element of the Self-energy is requested : gw_customnfreqsp3 1 # This is the number if frequencies at which the matrix element of the Self-energy is requested. gw_freqsp3 -0.5558150504 # These are the frequencies at which the matrix element of the Self-energy is requested. gwls_stern_kmax3 8 # This controls the convergence with respect to the size of the dielectric matrix. gwls_kmax_complement3 8 # This controls the convergence with respect to the size of the MODEL dielectric matrix. gw_icutcoul3 0 # For molecules, the coulomb potential needs to be spherically truncated. rcut3 9.0 # The spherical truncation has this radius. # GWLS calculation for the HOMO, without the dielectric model tolwfr4 1d-25 # This is the precision to which the Sternheimer eqs. are solved in GWLS. getden4 1 getwfk4 2 optdriver4 66 # This is the optdriver triggering a GWLS calculation. gwls_correlation4 4 # GWLS calc without the dielectric model gwls_band_index4 4 # This is the band for which the matrix element of the Self-energy is requested : gw_customnfreqsp4 1 # This is the number if frequencies at which the matrix element of the Self-energy is requested. gw_freqsp4 -0.5558150504 # These are the frequencies at which the matrix element of the Self-energy is requested. gwls_stern_kmax4 8 # This controls the convergence with respect to the size of the dielectric matrix. #gwls_kmax_complement4 8 # This controls the convergence with respect to the size of the MODEL dielectric matrix. gw_icutcoul4 0 # For molecules, the coulomb potential needs to be spherically truncated. rcut4 9.0 # The spherical truncation has this radius. # GWLS calculation of the static dielectric matrix eigenvalues (both model and exact) tolwfr5 1d-25 # This is the precision to which the Sternheimer eqs. are solved in GWLS. getden5 1 getwfk5 2 optdriver5 66 # This is the optdriver triggering a GWLS calculation. gwls_correlation5 5 # GWLS calc of dielectric matrix eigenvalues gwls_band_index5 4 # This is the band for which the matrix element of the Self-energy is requested : gwls_stern_kmax5 8 # This controls the convergence with respect to the size of the dielectric matrix. gwls_kmax_complement5 8 # This controls the convergence with respect to the size of the MODEL dielectric matrix. gw_icutcoul5 0 # For molecules, the coulomb potential needs to be spherically truncated. rcut5 9.0 # The spherical truncation has this radius. # GWLS calculation of the HOMO, shifting infinitesimally the poles away from the real axis to help convergence. EXPERIMENTAL tolwfr6 1d-25 # This is the precision to which the Sternheimer eqs. are solved in GWLS. getden6 1 getwfk6 2 optdriver6 66 # This is the optdriver triggering a GWLS calculation. zcut6 1d-6 # This is the imaginary shift applied to the poles of the dielectric matrix for the calculation of the residues. Triggers the shift if > 1d.12. Default is 0.0. gwls_band_index6 4 # This is the band for which the matrix element of the Self-energy is requested : gw_customnfreqsp6 1 # This is the number if frequencies at which the matrix element of the Self-energy is requested. gw_freqsp6 -0.5558150504 # These are the frequencies at which the matrix element of the Self-energy is requested. gwls_stern_kmax6 8 # This controls the convergence with respect to the size of the dielectric matrix. gwls_kmax_complement6 8 # This controls the convergence with respect to the size of the MODEL dielectric matrix. gw_icutcoul6 0 # For molecules, the coulomb potential needs to be spherically truncated. rcut6 9.0 # The spherical truncation has this radius. # These are mostly developer toys, kept to their default value for this test. # They should just be omitted from production runs gwls_kmax_poles 4 gwls_kmax_analytic 8 gwls_kmax_numeric 16 gwls_list_proj_freq *0.0 gwls_nseeds 1 gwls_n_proj_freq 0 gwls_recycle 2 gwls_first_seed 4 gwls_model_parameter 1.0 gwls_npt_gauss_quad 10 gwls_diel_model 2 gwls_print_debug 0 gwls_exchange 1 gwls_correlation 3 ### Common variables # Other parameters istwfk *1 # GWLS doesn't support time reversal symetry. nline 100 # GWLS linear equation solver may require many steps at problematic frequencies. # Number of bands # NOTE : The last band must be higher in energy that the higher gw_customnfreqsp requested by the user. nband 10 # Parallelisation : GWLS supports parallelism over bands and FFTs (from the ground state implementation), # but FFT parallelism tends to have poor efficiency in GWLS. # The optimum is usually to set the highest possible npband. # paral_kgb 1 # npfft 2 # npband 2 paral_kgb 0 # This is the sequential test. # Cutoff energy ecut 1.0 # Definition of the unit cell acell 3*18 rprim 1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0 # Definition of the atom types ntypat 2 znucl 14 1 # Definition of the atomic structure natom 5 typat 1 2 2 2 2 xcart 0.00000000000000 0.00000000000000 0.00000000000000 1.63281519635864 1.63281519635864 1.63281519635864 -1.63281519635864 -1.63281519635864 1.63281519635864 1.63281519635864 -1.63281519635864 -1.63281519635864 -1.63281519635864 1.63281519635864 -1.63281519635864 # BZ sampling for GW. kptopt 1 # Option for the automatic generation of k points, nkpt 1 ngkpt 1 1 1 nshiftk 1 shiftk 0.0 0.0 0.0 pp_dirpath "$ABI_PSPDIR" pseudos "PseudosGTH_pwteter/14si.pspgth, PseudosGTH_pwteter/01h.pspgth" #%% #%% [setup] #%% executable = abinit #%% [files] #%% files_to_test = #%% t15.out ,tolnlines = 20, tolabs = 1.0e-4, tolrel = 1.0e-4, fld_options = -ridiculous; #%% [paral_info] #%% max_nprocs = 1 #%% [extra_info] #%% authors = J. Laflamme Janssen #%% keywords = GW, GWLS #%% description = #%% Molecular SiH4 in GWLS (G0W0 with Lanczos basis and Sternheimer equations) #%% topics = GWls, BSE #%%