Full List of INPUT Keywords
System variables
These variables are used to control general system parameters.
suffix
Type: String
Description: In each run, ABACUS will generate a subdirectory in the working directory. This subdirectory contains all the information of the run. The subdirectory name has the format: OUT.suffix, where the
suffix
is the name you can pick up for your convenience.Default: ABACUS
calculation
Type: String
Description: Specify the type of calculation.
scf: perform self-consistent electronic structure calculations
nscf: perform non-self-consistent electronic structure calculations. A charge density file is required
relax: perform structure relaxation calculations, the
relax_nmax
parameter depicts the maximal number of ionic iterationscell-relax: perform cell relaxation calculations
md: perform molecular dynamics simulations
get_pchg: obtain partial (band-decomposed) charge densities (for LCAO basis only). See
nbands_istate
andbands_to_print
for more informationget_wf: obtain wave functions (for LCAO basis only). See
nbands_istate
andbands_to_print
for more informationget_S : obtain the overlap matrix formed by localized orbitals (for LCAO basis with multiple k points). the file name is
SR.csr
with file format being the same as that generated by out_mat_hs2gen_bessel : generates projectors, i.e., a series of Bessel functions, for the DeePKS method (for LCAO basis only); see also keywords
bessel_descriptor_lmax
,bessel_descriptor_rcut
andbessel_descriptor_tolerence
. A file namedjle.orb
will be generated which contains the projectors. An example is provided in examples/H2O-deepks-pwtest_memory : obtain a rough estimation of memory consuption for the calculation
test_neighbour : obtain information of neighboring atoms (for LCAO basis only), please specify a positive search_radius manually
Default: scf
esolver_type
Type: String
Description: choose the energy solver.
ksdft: Kohn-Sham density functional theory
ofdft: orbital-free density functional theory
tddft: real-time time-dependent density functional theory (TDDFT)
lj: Leonard Jones potential
dp: DeeP potential, see details in md.md
ks-lr: Kohn-Sham density functional theory + LR-TDDFT
lr: LR-TDDFT with given KS orbitals
Default: ksdft
symmetry
Type: Integer
Description: takes value 1, 0 or -1.
-1: No symmetry will be considered. It is recommended to set -1 for non-colinear + soc calculations, where time reversal symmetry is broken sometimes.
0: Only time reversal symmetry would be considered in symmetry operations, which implied k point and -k point would be treated as a single k point with twice the weight.
1: Symmetry analysis will be performed to determine the type of Bravais lattice and associated symmetry operations. (point groups, space groups, primitive cells, and irreducible k-points)
Default:
0:
if calculation==md/nscf/get_pchg/get_wf/get_S or gamma_only==True;
If (dft_fuctional==hse/hf/pbe0/scan0/opt_orb or rpa==True). Currently symmetry==1 is not supported in EXX (exact exchange) calculation.
If efield_flag==1
1: else
symmetry_prec
Type: Real
Description: The accuracy for symmetry judgment. Usually the default value is good enough, but if the lattice parameters or atom positions in STRU file is not accurate enough, this value should be enlarged.
Note: if calculation==cell_relax, this value can be dynamically changed corresponding to the variation of accuracy of the lattice parameters and atom positions during the relaxation. The new value will be printed in
OUT.${suffix}/running_cell-relax.log
in that case.Default: 1.0e-6
Unit: Bohr
symmetry_autoclose
Type: Boolean
Availability: symmetry==1
Description: Control how to deal with error in symmetry analysis due to inaccurate lattice parameters or atom positions in STRU file, especially useful when calculation==cell-relax
False: quit with an error message
True: automatically set symmetry to 0 and continue running without symmetry analysis
Default: True
kpar
Type: Integer
Description: divide all processors into kpar groups, and k points will be distributed among each group. The value taken should be less than or equal to the number of k points as well as the number of MPI processes.
Default: 1
bndpar
Type: Integer
Description: divide all processors into bndpar groups, and bands (only stochastic orbitals now) will be distributed among each group. It should be larger than 0.
Default: 1
latname
Type: String
Description: Specifies the type of Bravias lattice. When set to
none
, the three lattice vectors are supplied explicitly in STRU file. When set to a certain Bravais lattice type, there is no need to provide lattice vector, but a few lattice parameters might be required. For more information regarding this parameter, consult the page on STRU file.Available options are (correspondence with ibrav in QE(Quantum Espresso) is given in parenthesis):
none: free structure
sc: simple cubic (1)
fcc: face-centered cubic (2)
bcc: body-centered cubic (3)
hexagonal: hexagonal (4)
trigonal: trigonal (5)
st: simple tetragonal (6)
bct: body-centered tetragonal (7)
so: orthorhombic (8)
baco: base-centered orthorhombic (9)
fco: face-centered orthorhombic (10)
bco: body-centered orthorhombic (11)
sm: simple monoclinic (12)
bacm: base-centered monoclinic (13)
triclinic: triclinic (14)
Default: none
psi_initializer
Type: Integer
Description: enable the experimental feature psi_initializer, to support use numerical atomic orbitals initialize wavefunction (
basis_type pw
case).NOTE: this feature is not well-implemented for
nspin 4
case (closed presently), and cannot use withcalculation nscf
/esolver_type sdft
cases. Available options are:0: disable psi_initializer
1: enable psi_initializer
Default: 0
init_wfc
Type: String
Description: Only useful for plane wave basis only now. It is the name of the starting wave functions. In the future. we should also make this variable available for localized orbitals set.
Available options are:
atomic: from atomic pseudo wave functions. If they are not enough, other wave functions are initialized with random numbers.
atomic+random: add small random numbers on atomic pseudo-wavefunctions
file: from binary files
WAVEFUNC*.dat
, which are output by setting out_wfc_pw to2
.random: random numbers
with
psi_initializer 1
, two more options are supported:nao: from numerical atomic orbitals. If they are not enough, other wave functions are initialized with random numbers.
nao+random: add small random numbers on numerical atomic orbitals
Default: atomic
init_chg
Type: String
Description: This variable is used for both plane wave set and localized orbitals set. It indicates the type of starting density.
atomic: the density is starting from the summation of the atomic density of single atoms.
file: the density will be read in from a binary file
charge-density.dat
first. If it does not exist, the charge density will be read in from cube files. Besides, when you donspin=1
calculation, you only need the density file SPIN1_CHG.cube. However, if you donspin=2
calculation, you also need the density file SPIN2_CHG.cube. The density file should be output with these names if you set out_chg = 1 in INPUT file.auto: Abacus first attempts to read the density from a file; if not found, it defaults to using atomic density.
Default: atomic
init_vel
Type: Boolean
Description:
Default: False
mem_saver
Type: Boolean
Description: Used only for nscf calculations.
0: no memory saving techniques are used.
1: a memory saving technique will be used for many k point calculations.
Default: 0
diago_proc
Type: Integer
Availability: pw base
Description:
0: it will be set to the number of MPI processes. Normally, it is fine just leave it to the default value.
>0
: it specifies the number of processes used for carrying out diagonalization. Must be less than or equal to total number of MPI processes. Also, when cg diagonalization is used, diago_proc must be the same as the total number of MPI processes.
Default: 0
nbspline
Type: Integer
Description: If set to a natural number, a Cardinal B-spline interpolation will be used to calculate Structure Factor.
nbspline
represents the order of B-spline basis and a larger one can get more accurate results but cost more. It is turned off by default.Default: -1
kspacing
Type: Real
Description: Set the smallest allowed spacing between k points, unit in 1/bohr. It should be larger than 0.0, and suggest smaller than 0.25. When you have set this value > 0.0, then the KPT file is unnecessary, and the number of K points nk_i = max(1, int(|b_i|/KSPACING_i)+1), where b_i is the reciprocal lattice vector. The default value 0.0 means that ABACUS will read the applied KPT file. If only one value is set (such as
kspacing 0.5
), then kspacing values of a/b/c direction are all set to it; and one can also set 3 values to set the kspacing value for a/b/c direction separately (such as:kspacing 0.5 0.6 0.7
).Note: if gamma_only is set to be true, kspacing is invalid.
Default: 0.0
min_dist_coef
Type: Real
Description: a factor related to the allowed minimum distance between two atoms. At the beginning, ABACUS will check the structure, and if the distance of two atoms is shorter than min_dist_coef*(standard covalent bond length), we think this structure is unreasonable. If you want to calculate some structures in extreme conditions like high pressure, you should set this parameter as a smaller value or even 0.
Default: 0.2
device
Type: String
Description: Specifies the computing device for ABACUS.
Available options are:
cpu: for CPUs via Intel, AMD, or Other supported CPU devices
gpu: for GPUs via CUDA or ROCm.
Known limitations:
ks_solver
must also be set to the algorithms supported. lcao_in_pw currently does not supportgpu
.Default: cpu
precision
Type: String
Description: Specifies the precision of the PW_SCF calculation.
Available options are:
single: single precision
double: double precision
Known limitations:
pw basis: required by the
single
precision optionscg/bpcg/dav ks_solver: required by the
single
precision options
Default: double
Electronic structure
These variables are used to control the electronic structure and geometry relaxation calculations.
basis_type
Type: String
Description: Choose the basis set.
pw: Using plane-wave basis set only.
lcao: Using localized atomic orbital sets.
lcao_in_pw: Expand the localized atomic set in plane-wave basis, non-self-consistent field calculation not tested.
Default: pw
ks_solver
Type: String
Description: Choose the diagonalization methods for the Hamiltonian matrix expanded in a certain basis set.
For plane-wave basis,
cg: cg method.
bpcg: bpcg method, which is a block-parallel Conjugate Gradient (CG) method, typically exhibits higher acceleration in a GPU environment.
dav: the Davidson algorithm.
dav_subspace: subspace Davidson algorithm
For atomic orbitals basis,
lapack: This method is only avaliable for serial version. For parallel version please use scalapack_gvx.
genelpa: This method should be used if you choose localized orbitals.
scalapack_gvx: Scalapack can also be used for localized orbitals.
cusolver: This method needs building with CUDA and at least one gpu is available.
If you set ks_solver=
genelpa
for basis_type=pw
, the program will be stopped with an error message:genelpa can not be used with plane wave basis.
Then the user has to correct the input file and restart the calculation.
Default: cg (plane-wave basis), or genelpa (localized atomic orbital basis, if compiling option
USE_ELPA
has been set),lapack (localized atomic orbital basis, if compiling optionENABLE_MPI
has not been set), scalapack_gvx, (localized atomic orbital basis, if compiling optionUSE_ELPA
has not been set and if compiling optionENABLE_MPI
has been set)
nbands
Type: Integer
Description: The number of Kohn-Sham orbitals to calculate. It is recommended to setup this value, especially when smearing techniques are utilized, more bands should be included.
Default:
nspin=1: max(1.2*occupied_bands, occupied_bands + 10)
nspin=2: max(1.2*nelec_spin, nelec_spin + 10), in which nelec_spin = max(nelec_spin_up, nelec_spin_down)
nspin=4: max(1.2*nelec, nelec + 20)
nelec
Type: Real
Description:
0.0: the total number of electrons will be calculated by the sum of valence electrons (i.e. assuming neutral system).
>0.0
: this denotes the total number of electrons in the system. Must be less than 2*nbands.
Default: 0.0
nelec_delta
Type: Real
Description: the total number of electrons will be calculated by
nelec
+nelec_delta
.Default: 0.0
nupdown
Type: Real
Description:
0.0: no constrain apply to system.
>0.0
: this denotes the difference number of electrons between spin-up and spin-down in the system. The range of value must in [-nelec ~ nelec]. It is one method of constraint DFT, the fermi energy level will separate to E_Fermi_up and E_Fermi_down.
Default: 0.0
dft_functional
Type: String
Description: In our package, the XC functional can either be set explicitly using the
dft_functional
keyword inINPUT
file. Ifdft_functional
is not specified, ABACUS will use the xc functional indicated in the pseudopotential file. On the other hand, if dft_functional is specified, it will overwrite the functional from pseudopotentials and performs calculation with whichever functional the user prefers. We further offer two ways of supplying exchange-correlation functional. The first is using ‘short-hand’ names such as ‘LDA’, ‘PBE’, ‘SCAN’. A complete list of ‘short-hand’ expressions can be found in the source code. The other way is only available when compiling with LIBXC, and it allows for supplying exchange-correlation functionals as combinations of LIBXC keywords for functional components, joined by a plus sign, for example, ‘dft_functional=‘LDA_X_1D_EXPONENTIAL+LDA_C_1D_CSC’. The list of LIBXC keywords can be found on its website. In this way, we support all the LDA,GGA and mGGA functionals provided by LIBXC.Furthermore, the old INPUT parameter exx_hybrid_type for hybrid functionals has been absorbed into dft_functional. Options are
hf
(pure Hartree-Fock),pbe0
(PBE0),hse
(Note: in order to use HSE functional, LIBXC is required). Note also that HSE has been tested while PBE0 has NOT been fully tested yet, and the maximum CPU cores for running exx in parallel is \(N(N+1)/2\), with N being the number of atoms. And forces for hybrid functionals are not supported yet.If set to
opt_orb
, the program will not perform hybrid functional calculation. Instead, it is going to generate opt-ABFs as discussed in this article.Default: same as UPF file.
xc_temperature
Type: Real
Description: specifies temperature when using temperature-dependent XC functionals (KSDT and so on).
Default : 0.0
Unit: Ry
pseudo_rcut
Type: Real
Description: Cut-off of radial integration for pseudopotentials
Default: 15
Unit: Bohr
pseudo_mesh
Type: Integer
Description:
0: use our own mesh for radial integration of pseudopotentials
1: use the mesh that is consistent with quantum espresso
Default: 0
nspin
Type: Integer
Description: The number of spin components of wave functions.
1: Spin degeneracy
2: Collinear spin polarized.
4: For the case of noncollinear polarized, nspin will be automatically set to 4 without being specified by the user.
Default: 1
smearing_method
Type: String
Description: It indicates which occupation and smearing method is used in the calculation.
fixed: fixed occupations (available for non-coductors only)
gauss or gaussian: Gaussian smearing method.
mp: methfessel-paxton smearing method; recommended for metals.
mp2: 2-nd methfessel-paxton smearing method; recommended for metals.
mv or cold: marzari-vanderbilt smearing method.
fd: Fermi-Dirac smearing method: \(f=1/\{1+\exp[(E-\mu)/kT]\}\) and smearing_sigma below is the temperature \(T\) (in Ry).
Default: gauss
smearing_sigma
Type: Real
Description: Energy range for smearing.
Default: 0.015
Unit: Ry
smearing_sigma_temp
Type: Real
Description: Energy range for smearing,
smearing_sigma
= 1/2 kBsmearing_sigma_temp
.Default: 2 *
smearing_sigma
/ kB.Unit: K
mixing_type
Type: String
Description: Charge mixing methods.
plain: Just simple mixing.
pulay: Standard Pulay method. P. Pulay Chemical Physics Letters, (1980)
broyden: Simplified modified Broyden method. D.D. Johnson Physical Review B (1988)
In general, the convergence of the Broyden method is slightly faster than that of the Pulay method.
Default: broyden
mixing_beta
Type: Real
Description: In general, the formula of charge mixing can be written as \(\rho_{new} = \rho_{old} + \beta * \rho_{update}\), where \(\rho_{new}\) represents the new charge density after charge mixing, \(\rho_{old}\) represents the charge density in previous step, \(\rho_{update}\) is obtained through various mixing methods, and \(\beta\) is set by the parameter
mixing_beta
. A lower value of ‘mixing_beta’ results in less influence of \(\rho_{update}\) on \(\rho_{new}\), making the self-consistent field (SCF) calculation more stable. However, it may require more steps to achieve convergence. We recommend the following options:0.8:
nspin=1
0.4:
nspin=2
andnspin=4
0: keep charge density unchanged, usually used for restarting with
init_chg=file
or testing.0.1 or less: if convergence of SCF calculation is difficult to reach, please try
0 < mixing_beta < 0.1
.
Note: For low-dimensional large systems, the setup of
mixing_beta=0.1
,mixing_ndim=20
, andmixing_gg0=1.0
usually works well.Default: 0.8 for
nspin=1
, 0.4 fornspin=2
andnspin=4
.
mixing_beta_mag
Type: Real
Description: Mixing parameter of magnetic density.
Default:
4*mixing_beta
, but the maximum value is 1.6.
Note that mixing_beta_mag
is not euqal to mixing_beta
means that \(\rho_{up}\) and \(\rho_{down}\) mix independently from each other. This setting will fail for one case where the \(\rho_{up}\) and \(\rho_{down}\) of the ground state refers to different Kohn-Sham orbitals. For an atomic system, the \(\rho_{up}\) and \(\rho_{down}\) of the ground state refers to different Kohn-Sham orbitals. We all know Kohn-Sham orbitals are orthogonal to each other. So the mixture of \(\rho_{up}\) and \(\rho_{down}\) should be exactly independent, otherwise SCF cannot find the ground state forever. To sum up, please make sure mixing_beta_mag
and mixing_gg0_mag
exactly euqal to mixing_beta
and mixing_gg0
if you calculate an atomic system.
mixing_ndim
Type: Integer
Description: It indicates the mixing dimensions in Pulay or Broyden. Pulay and Broyden method use the density from previous mixing_ndim steps and do a charge mixing based on this density.
For systems that are difficult to converge, one could try increasing the value of ‘mixing_ndim’ to enhance the stability of the self-consistent field (SCF) calculation.
Default: 8
mixing_restart
Type: double
Description: If the density difference between input and output
drho
is smaller thanmixing_restart
, SCF will restart at next step which means SCF will restart by using output charge density from perivos iteration as input charge density directly, and start a new mixing. Notice thatmixing_restart
will only take effect once in one SCF.Default: 0
mixing_dmr
Type: bool
Availability: Only for
mixing_restart>=0.0
Description: At n-th iteration which is calculated by
drho<mixing_restart
, SCF will start a mixing for real-space density matrix by using the same coefficiences as the mixing of charge density.Default: false
mixing_gg0
Type: Real
Description: Whether to perfom Kerker scaling for charge density.
>0: The high frequency wave vectors will be suppressed by multiplying a scaling factor \(\frac{k^2}{k^2+gg0^2}\). Setting
mixing_gg0 = 1.0
is normally a good starting point. Kerker preconditioner will be automatically turned off ifmixing_beta <= 0.1
.0: No Kerker scaling is performed.
For systems that are difficult to converge, particularly metallic systems, enabling Kerker scaling may aid in achieving convergence.
Default: 1.0
mixing_gg0_mag
Type: Real
Description: Whether to perfom Kerker preconditioner of magnetic density. Note: we do not recommand to open Kerker preconditioner of magnetic density unless the system is too hard to converge.
Default: 0.0
mixing_gg0_min
Type: Real
Description: the minimum kerker coefficient
Default: 0.1
mixing_angle
Type: Real
Availability: Only relevant for non-colinear calculations
nspin=4
.Description: Normal broyden mixing can give the converged result for a given magnetic configuration. If one is not interested in the energies of a given magnetic configuration but wants to determine the ground state by relaxing the magnetic moments’ directions, one cannot rely on the standard Broyden mixing algorithm. To enhance the ability to find correct magnetic configuration for non-colinear calculations, ABACUS implements a promising mixing method proposed by J. Phys. Soc. Jpn. 82 (2013) 114706. Here,
mixing_angle
is the angle mixing parameter. In fact, onlymixing_angle=1.0
is implemented currently.<=0: Normal broyden mixing for \(m_{x}, m_{y}, m_{z}\)
>0: Angle mixing for the modulus \(|m|\) with
mixing_angle=1.0
Default: -10.0
Note: In new angle mixing, you should set mixing_beta_mag >> mixing_beta
. The setup of mixing_beta=0.2
, mixing_beta_mag=1.0
usually works well.
mixing_tau
Type: Boolean
Availability: Only relevant for meta-GGA calculations.
Description: Whether to mix the kinetic energy density.
True: The kinetic energy density will also be mixed. It seems for general cases, SCF converges fine even without this mixing. However, if there is difficulty in converging SCF for meta-GGA, it might be helpful to turn this on.
False: The kinetic energy density will not be mixed.
Default: False
mixing_dftu
Type: Boolean
Availability: Only relevant for DFT+U calculations.
Description: Whether to mix the occupation matrices.
True: The occupation matrices will also be mixed by plain mixing. From experience this is not very helpful if the +U calculation does not converge.
False: The occupation matrices will not be mixed.
Default: False
gamma_only
Type: Integer
Availability: Only used in localized orbitals set
Description: Whether to use gamma_only algorithm.
0: more than one k-point is used and the ABACUS is slower compared to the gamma only algorithm.
1: ABACUS uses gamma only, the algorithm is faster and you don’t need to specify the k-points file.
Note: If gamma_only is set to 1, the KPT file will be overwritten. So make sure to turn off gamma_only for multi-k calculations.
Default: 0
printe
Type: Integer
Description: Print out energy for each band for every printe step
Default: 100
scf_nmax
Type: Integer
Description: This variable indicates the maximal iteration number for electronic iterations.
Default: 100
scf_thr
Type: Real
Description: It’s the density threshold for electronic iteration. It represents the charge density error between two sequential densities from electronic iterations. Usually for local orbitals, usually 1e-6 may be accurate enough.
Default: 1.0e-9 (plane-wave basis), or 1.0e-7 (localized atomic orbital basis).
scf_ene_thr
Type: Real
Description: It’s the energy threshold for electronic iteration. It represents the total energy error between two sequential densities from electronic iterations.
Default: -1.0. If the user does not set this parameter, it will not take effect.
Unit: eV
scf_thr_type
Type: Integer
Description: Choose the calculation method of convergence criterion.
1: the criterion is defined as \(\Delta\rho_G = \frac{1}{2}\iint{\frac{\Delta\rho(r)\Delta\rho(r')}{|r-r'|}d^3r d^3r'}\).
2: the criterion is defined as \(\Delta\rho_R = \frac{1}{N_e}\int{|\Delta\rho(r)|d^3r}\), where \(N_e\) is the number of electron.
Note: This parameter is still under testing and the default setting is usually sufficient.
Default: 1 (plane-wave basis), or 2 (localized atomic orbital basis).
chg_extrap
Type: String
Description: Methods to do extrapolation of density when ABACUS is doing geometry relaxations or molecular dynamics.
atomic: atomic extrapolation.
first-order: first-order extrapolation.
second-order: second-order extrapolation.
Default: first-order (geometry relaxations), second-order (molecular dynamics), else atomic
lspinorb
Type: Boolean
Description: Whether to consider spin-orbital coupling effect in the calculation.
True: Consider spin-orbital coupling effect, and
nspin
is also automatically set to 4.False: Do not consider spin-orbital coupling effect.
Default: False
noncolin
Type: Boolean
Description: Whether to allow non-collinear polarization, in which case the coupling between spin up and spin down will be taken into account.
True: Allow non-collinear polarization, and
nspin
is also automatically set to 4.False: Do not allow non-collinear polarization.
Default: False
soc_lambda
Type: Real
Availability: Relevant for soc calculations.
Description: Sometimes, for some real materials, both scalar-relativistic and full-relativistic can not describe the exact spin-orbit coupling. Artificial modulation may help in such cases.
soc_lambda
, which has value range [0.0, 1.0] , is used for modulate SOC effect.In particular,
soc_lambda 0.0
refers to scalar-relativistic case andsoc_lambda 1.0
refers to full-relativistic case.Default: 1.0
Electronic structure (SDFT)
These variables are used to control the parameters of stochastic DFT (SDFT), mix stochastic-deterministic DFT (MDFT), or complete-basis Chebyshev method (CT). In the following text, stochastic DFT is used to refer to these three methods. We suggest using SDFT to calculate high-temperature systems and we only support smearing_method “fd”. Both “scf” and “nscf” calculation are supported.
method_sto
Type: Integer
Availability: esolver_type =
sdft
Description: Different methods to do stochastic DFT
1: Calculate \(T_n(\hat{h})\ket{\chi}\) twice, where \(T_n(x)\) is the n-th order Chebyshev polynomial and \(\hat{h}=\frac{\hat{H}-\bar{E}}{\Delta E}\) owning eigenvalues \(\in(-1,1)\). This method cost less memory but is slower.
2: Calculate \(T_n(\hat{h})\ket{\chi}\) once but needs much more memory. This method is much faster. Besides, it calculates \(N_e\) with \(\bra{\chi}\sqrt{\hat f}\sqrt{\hat f}\ket{\chi}\), which needs a smaller nche_sto. However, when the memory is not enough, only method 1 can be used.
other: use 2
Default: 2
nbands_sto
Type: Integer or string
Availability: esolver_type =
sdft
Description: The number of stochastic orbitals
> 0: Perform stochastic DFT. Increasing the number of bands improves accuracy and reduces stochastic errors, which scale as \(1/\sqrt{N_{\chi}}\); To perform mixed stochastic-deterministic DFT, you should set nbands, which represents the number of KS orbitals.
0: Perform Kohn-Sham DFT.
all: All complete basis sets are used to replace stochastic orbitals with the Chebyshev method (CT), resulting in the same results as KSDFT without stochastic errors.
Default: 256
nche_sto
Type: Integer
Availability: esolver_type =
sdft
Description: Chebyshev expansion orders for stochastic DFT.
Default: 100
emin_sto
Type: Real
Availability: esolver_type =
sdft
Description: Trial energy to guess the lower bound of eigen energies of the Hamiltonian Operator \(\hat{H}\).
Default: 0.0
Unit: Ry
emax_sto
Type: Real
Availability: esolver_type =
sdft
Description: Trial energy to guess the upper bound of eigen energies of the Hamiltonian Operator \(\hat{H}\).
Default: 0.0
Unit: Ry
seed_sto
Type: Integer
Availability: esolver_type =
sdft
Description: The random seed to generate stochastic orbitals.
>= 0: Stochastic orbitals have the form of \(\exp(i2\pi\theta(G))\), where \(\theta\) is a uniform distribution in \((0,1)\).
0: the seed is decided by time(NULL).
<= -1: Stochastic orbitals have the form of \(\pm1\) with equal probability.
-1: the seed is decided by time(NULL).
Default: 0
initsto_ecut
Type: Real
Availability: esolver_type =
sdft
Description: Stochastic wave functions are initialized in a large box generated by “4*
initsto_ecut
”.initsto_ecut
should be larger than ecutwfc. In this method, SDFT results are the same when using different cores. Besides, coefficients of the same G are the same when ecutwfc is rising to initsto_ecut. If it is smaller than ecutwfc, it will be turned off.Default: 0.0
Unit: Ry
initsto_freq
Type: Integer
Availability: esolver_type =
sdft
Description: Frequency (once each initsto_freq steps) to generate new stochastic orbitals when running md.
positive integer: Update stochastic orbitals
0: Never change stochastic orbitals.
Default: 0
npart_sto
Type: Integer
Availability: method_sto =
2
and out_dos = 1 or cal_cond =True
Description: Make memory cost to 1/npart_sto times of the previous one when running the post process of SDFT like DOS or conductivities.
Default: 1
Geometry relaxation
These variables are used to control the geometry relaxation.
relax_method
Type: String
Description: The methods to do geometry optimization.
cg: using the conjugate gradient (CG) algorithm. Note that there are two implementations of the conjugate gradient (CG) method, see relax_new.
bfgs: using the Broyden–Fletcher–Goldfarb–Shanno (BFGS) algorithm.
cg_bfgs: using the CG method for the initial steps, and switching to BFGS method when the force convergence is smaller than relax_cg_thr.
sd: using the steepest descent (SD) algorithm.
fire: the Fast Inertial Relaxation Engine method (FIRE), a kind of molecular-dynamics-based relaxation algorithm, is implemented in the molecular dynamics (MD) module. The algorithm can be used by setting calculation to
md
and md_type tofire
. Also ionic velocities should be set in this case. See fire for more details.
Default: cg
relax_new
Type: Boolean
Description: At around the end of 2022 we made a new implementation of the Conjugate Gradient (CG) method for
relax
andcell-relax
calculations. But the old implementation was also kept.True: use the new implementation of CG method for
relax
andcell-relax
calculations.False: use the old implementation of CG method for
relax
andcell-relax
calculations.
Default: True
relax_scale_force
Type: Real
Availability: only used when
relax_new
set toTrue
Description: The paramether controls the size of the first conjugate gradient step. A smaller value means the first step along a new CG direction is smaller. This might be helpful for large systems, where it is safer to take a smaller initial step to prevent the collapse of the whole configuration.
Default: 0.5
relax_nmax
Type: Integer
Description: The maximal number of ionic iteration steps, the minimum value is 1.
Default: 1
relax_cg_thr
Type: Real
Description: When move-method is set to
cg_bfgs
, a mixed algorithm of conjugate gradient (CG) method and Broyden–Fletcher–Goldfarb–Shanno (BFGS) method is used. The ions first move according to CG method, then switched to BFGS method when the maximum of force on atoms is reduced below the CG force threshold, which is set by this parameter.Default: 0.5
Unit: eV/Angstrom
cal_force
Type: Boolean
Description:
True calculate the force at the end of the electronic iteration
False no force calculation at the end of the electronic iteration
Default: False if
calculation
is set toscf
, True ifcalculation
is set tocell-relax
,relax
, ormd
.
force_thr
Type: Real
Description: Threshold of the force convergence in Ry/Bohr. The threshold is compared with the largest force among all of the atoms. The recommended value for using atomic orbitals is 0.04 eV/Angstrom (0.0016 Ry/Bohr). The parameter is equivalent to force_thr_ev except for the unit. You may choose either you like.
Default: 0.001
Unit: Ry/Bohr (25.7112 eV/Angstrom)
force_thr_ev
Type: Real
Description: Threshold of the force convergence in eV/Angstrom. The threshold is compared with the largest force among all of the atoms. The recommended value for using atomic orbitals is 0.04 eV/Angstrom (0.0016 Ry/Bohr). The parameter is equivalent to force_thr except for the unit. You may choose either you like.
Default: 0.0257112
Unit: eV/Angstrom (0.03889 Ry/Bohr)
force_thr_ev2
Type: Real
Description: The calculated force will be set to 0 when it is smaller than the parameter
force_thr_ev2
.Default: 0.0
Unit: eV/Angstrom
relax_bfgs_w1
Type: Real
Description: This variable controls the Wolfe condition for Broyden–Fletcher–Goldfarb–Shanno (BFGS) algorithm used in geometry relaxation. You can look into the paper Phys.Chem.Chem.Phys.,2000,2,2177 for more information.
Default: 0.01
relax_bfgs_w2
Type: Real
Description: This variable controls the Wolfe condition for Broyden–Fletcher–Goldfarb–Shanno (BFGS) algorithm used in geometry relaxation. You can look into the paper Phys.Chem.Chem.Phys.,2000,2,2177 for more information.
Default: 0.5
relax_bfgs_rmax
Type: Real
Description: This variable is for geometry optimization. It stands for the maximal movement of all the atoms. The sum of the movements from all atoms can be increased during the optimization steps. However, it can not be larger than
relax_bfgs_rmax
Unit: Bohr
Default: 0.8
relax_bfgs_rmin
Type: Real
Description: This variable is for geometry optimization. It indicates the minimal movement of all the atoms. When the movement of all the atoms is smaller than relax_bfgs_rmin Bohr, and the force convergence is still not achieved, the calculation will break down.
Default: 1e-5
Unit: Bohr
relax_bfgs_init
Type: Real
Description: This variable is for geometry optimization. It stands for the sum of initial movements of all of the atoms.
Default: 0.5
Unit: Bohr
cal_stress
Type: Boolean
Description:
True: calculate the stress at the end of the electronic iteration
False: no calculation of the stress at the end of the electronic iteration
Default: True if
calculation
iscell-relax
, False otherwise.
stress_thr
Type: Real
Description: The threshold of the stress convergence. The threshold is compared with the largest component of the stress tensor.
Default: 0.5
Unit: kbar
press1, press2, press3
Type: Real
Description: The external pressures along three axes. Positive input value is taken as compressive stress.
Default: 0
Unit: kbar
fixed_axes
Type: String
Availability: only used when
calculation
set tocell-relax
Description: Axes that are fixed during cell relaxation. Possible choices are:
None: default; all of the axes can relax
volume: relaxation with fixed volume
shape: fix shape but change volume (i.e. only lattice constant changes)
a: fix a axis during relaxation
b: fix b axis during relaxation
c: fix c axis during relaxation
ab: fix both a and b axes during relaxation
ac: fix both a and c axes during relaxation
bc: fix both b and c axes during relaxation
Note : fixed_axes = “shape” and “volume” are only available for relax_new = True
Default: None
fixed_ibrav
Type: Boolean
Availability: Must be used along with relax_new set to True, and a specific latname must be provided
Description:
True: the lattice type will be preserved during relaxation
False: No restrictions are exerted during relaxation in terms of lattice type
Note: it is possible to use
fixed_ibrav
withfixed_axes
, but please make sure you know what you are doing. For example, if we are doing relaxation of a simple cubic lattice (latname
= “sc”), and we usefixed_ibrav
along withfixed_axes
= “volume”, then the cell is never allowed to move and as a result, the relaxation never converges.
Default: False
fixed_atoms
Type: Boolean
Description:
True: The direct coordinates of atoms will be preserved during variable-cell relaxation.
False: No restrictions are exerted on positions of all atoms. However, users can still fix certain components of certain atoms by using the
m
keyword inSTRU
file. For the latter option, check the end of this instruction.
Default: False
cell_factor
Type: Real
Description: Used in the construction of the pseudopotential tables. It should exceed the maximum linear contraction of the cell during a simulation.
Default: 1.2
Density of states
These variables are used to control the calculation of DOS. Detailed introduction
dos_edelta_ev
Type: Real
Description: the step size in writing Density of States (DOS)
Default: 0.01
Unit: eV
dos_sigma
Type: Real
Description: the width of the Gaussian factor when obtaining smeared Density of States (DOS)
Default: 0.07
Unit: eV
dos_scale
Type: Real
Description: Defines the energy range of DOS output as (emax-emin)*(1+dos_scale), centered at (emax+emin)/2. This parameter will be used when dos_emin and dos_emax are not set.
Default: 0.01
Unit: eV
dos_emin_ev
Type: Real
Description: the minimal range for Density of States (DOS)
If set, “dos_scale” will be ignored.
Default: Minimal eigenenergy of \(\hat{H}\)
Unit: eV
dos_emax_ev
Type: Real
Description: the maximal range for Density of States (DOS)
If set, “dos_scale” will be ignored.
Default: Maximal eigenenergy of \(\hat{H}\)
Unit: eV
dos_nche
Type: Integer The order of Chebyshev expansions when using Stochastic Density Functional Theory (SDFT) to calculate DOS.
Default: 100
NAOs
These variables are used to control the generation of numerical atomic orbitals (NAOs), whose radial parts are linear combinations of spherical Bessel functions with a node (i.e., evaluate to zero) at the cutoff radius. In plane-wave-based calculations, necessary information will be printed into OUT.${suffix}/orb_matrix.${i}.dat
, which serves as an input file for the generation of NAOs. Please check SIAB package for more information.
bessel_nao_ecut
Type: Real
Description: “Energy cutoff” (in Ry) of spherical Bessel functions. The number of spherical Bessel functions that constitute the radial parts of NAOs is determined by sqrt(
bessel_nao_ecut
)\(\times\)bessel_nao_rcut
/\(\pi\).Default:
ecutwfc
bessel_nao_tolerence
Type: Real
Description: tolerance when searching for the zeros of spherical Bessel functions.
Default: 1.0e-12
bessel_nao_rcut
Type: Real
Description: Cutoff radius (in Bohr) and the common node of spherical Bessel functions used to construct the NAOs.
Default: 6.0
bessel_nao_smooth
Type: Boolean
Description: if True, NAOs will be smoothed near the cutoff radius by \(1-\exp\left(-\frac{(r-r_{cut})^2}{2\sigma^2}\right)\). See
bessel_nao_rcut
for \(r_{cut}\) andbessel_nao_sigma
for \(\sigma\).Default: True
bessel_nao_sigma
Type: Real
Description: Smoothing range (in Bohr). See also
bessel_nao_smooth
.Default: 0.1
DeePKS
These variables are used to control the usage of DeePKS method (a comprehensive data-driven approach to improve the accuracy of DFT). Warning: this function is not robust enough for the current version. Please try the following variables at your own risk:
deepks_out_labels
Type: Boolean
Availability: numerical atomic orbital basis
Description: Print labels and descriptors for DeePKS training in OUT.${suffix}. The names of these files start with “deepks”.
Note: In
LCAO
calculation, the path of a numerical descriptor (anorb
file) is needed to be specified under theNUMERICAL_DESCRIPTOR
tag in theSTRU
file. For example:NUMERICAL_ORBITAL H_gga_8au_60Ry_2s1p.orb O_gga_7au_60Ry_2s2p1d.orb NUMERICAL_DESCRIPTOR jle.orb
Default: False
deepks_scf
Type: Boolean
Availability: numerical atomic orbital basis
Description: perform self-consistent field iteration in DeePKS method
Note: A trained, traced model file is needed.
Default: False
deepks_equiv
Type: Boolean
Availability: numerical atomic orbital basis
Description: whether to use equivariant version of DeePKS
Note: the equivariant version of DeePKS-kit is still under development, so this feature is currently only intended for internal usage.
Default: False
deepks_model
Type: String
Availability: numerical atomic orbital basis and
deepks_scf
is trueDescription: the path of the trained, traced neural network model file generated by deepks-kit
Default: None
bessel_descriptor_lmax
Type: Integer
Availability:
gen_bessel
calculationDescription: the maximum angular momentum of the Bessel functions generated as the projectors in DeePKS
NOte: To generate such projectors, set calculation type to
gen_bessel
in ABACUS. See also calculation.Default: 2
bessel_descriptor_ecut
Type: Real
Availability:
gen_bessel
calculationDescription: energy cutoff of Bessel functions
Default: same as ecutwfc
Unit: Ry
bessel_descriptor_tolerence
Type: Real
Availability:
gen_bessel
calculationDescription: tolerance for searching the zeros of Bessel functions
Default: 1.0e-12
bessel_descriptor_rcut
Type: Real
Availability:
gen_bessel
calculationDescription: cutoff radius of Bessel functions
Default: 6.0
Unit: Bohr
bessel_descriptor_smooth
Type: Boolean
Availability:
gen_bessel
calculationDescription: smooth the Bessel functions at radius cutoff
Default: False
bessel_descriptor_sigma
Type: Real
Availability:
gen_bessel
calculationDescription: smooth parameter at the cutoff radius of projectors
Default: 0.1
Unit: Bohr
deepks_bandgap
Type: Boolean
Availability: numerical atomic orbital basis and
deepks_scf
is trueDescription: include bandgap label for DeePKS training
Default: False
deepks_v_delta
Type: int
Availability: numerical atomic orbital basis
Description: Include V_delta label for DeePKS training. When
deepks_out_labels
is true anddeepks_v_delta
> 0, ABACUS will output h_base.npy, v_delta.npy and h_tot.npy(h_tot=h_base+v_delta). Meanwhile, whendeepks_v_delta
equals 1, ABACUS will also output v_delta_precalc.npy, which is used to calculate V_delta during DeePKS training. However, when the number of atoms grows, the size of v_delta_precalc.npy will be very large. In this case, it’s recommended to setdeepks_v_delta
as 2, and ABACUS will output psialpha.npy and grad_evdm.npy but not v_delta_precalc.npy. These two files are small and can be used to calculate v_delta_precalc in the procedure of training DeePKS.Default: 0
deepks_out_unittest
Type: Boolean
Description: generate files for constructing DeePKS unit test
Note: Not relevant when running actual calculations. When set to 1, ABACUS needs to be run with only 1 process.
Default: False
OFDFT: orbital free density functional theory
of_kinetic
Type: String
Availability: OFDFT
Description: The type of KEDF (kinetic energy density functional).
wt: Wang-Teter.
tf: Thomas-Fermi.
vw: von Weizsäcker.
tf+: TF\(\rm{\lambda}\)vW, the parameter \(\rm{\lambda}\) can be set by
of_vw_weight
.lkt: Luo-Karasiev-Trickey.
Default: wt
of_method
Type: String
Availability: OFDFT
Description: The optimization method used in OFDFT.
cg1: Polak-Ribiere. Standard CG algorithm.
cg2: Hager-Zhang (generally faster than cg1).
tn: Truncated Newton algorithm.
Default:tn
of_conv
Type: String
Availability: OFDFT
Description: Criterion used to check the convergence of OFDFT.
energy: Ttotal energy changes less than
of_tole
.potential: The norm of potential is less than
of_tolp
.both: Both energy and potential must satisfy the convergence criterion.
Default: energy
of_tole
Type: Real
Availability: OFDFT
Description: Tolerance of the energy change for determining the convergence.
Default: 2e-6
Unit: Ry
of_tolp
Type: Real
Availability: OFDFT
Description: Tolerance of potential for determining the convergence.
Default: 1e-5
Unit: Ry
of_tf_weight
Type: Real
Availability: OFDFT with
of_kinetic=tf, tf+, wt
Description: Weight of TF KEDF (kinetic energy density functional).
Default: 1.0
of_vw_weight
Type: Real
Availability: OFDFT with
of_kinetic=vw, tf+, wt, lkt
Description: Weight of vW KEDF (kinetic energy density functional).
Default: 1.0
of_wt_alpha
Type: Real
Availability: OFDFT with
of_kinetic=wt
Description: Parameter alpha of WT KEDF (kinetic energy density functional).
Default: \(5/6\)
of_wt_beta
Type: Real
Availability: OFDFT with
of_kinetic=wt
Description: Parameter beta of WT KEDF (kinetic energy density functional).
Default: \(5/6\)
of_wt_rho0
Type: Real
Availability: OFDFT with
of_kinetic=wt
Description: The average density of system.
Default: 0.0
Unit: Bohr^-3
of_hold_rho0
Type: Boolean
Availability: OFDFT with
of_kinetic=wt
Description: Whether to fix the average density rho0.
True: rho0 will be fixed even if the volume of system has changed, it will be set to True automatically if
of_wt_rho0
is not zero.False: rho0 will change if volume of system has changed.
Default: False
of_lkt_a
Type: Real
Availability: OFDFT with
of_kinetic=lkt
Description: Parameter a of LKT KEDF (kinetic energy density functional).
Default: 1.3
of_read_kernel
Type: Boolean
Availability: OFDFT with
of_kinetic=wt
Description: Whether to read in the kernel file.
True: The kernel of WT KEDF (kinetic energy density functional) will be filled from the file specified by
of_kernel_file
.False: The kernel of WT KEDF (kinetic energy density functional) will be filled from formula.
Default: False
of_kernel_file
Type: String
Availability: OFDFT with
of_read_kernel=True
Description: The name of WT kernel file.
Default: WTkernel.txt
of_full_pw
Type: Boolean
Availability: OFDFT
Description: Whether to use full planewaves.
True: Ecut will be ignored while collecting planewaves, so that all planewaves will be used in FFT.
False: Only use the planewaves inside ecut, the same as KSDFT.
Default: True
of_full_pw_dim
Type: Integer
Availability: OFDFT with
of_full_pw = True
Description: Specify the parity of FFT dimensions.
0: either odd or even.
1: odd only.
2: even only.
Note: Even dimensions may cause slight errors in FFT. It should be ignorable in ofdft calculation, but it may make Cardinal B-spline interpolation unstable, so please set
of_full_pw_dim = 1
ifnbspline != -1
.Default: 0
Electric field and dipole correction
These variables are relevant to electric field and dipole correction
efield_flag
Type: Boolean
Description: added the electric field.
True: A saw-like potential simulating an electric field is added to the bare ionic potential.
False: Not added the electric field.
Default: False
dip_cor_flag
Type: Boolean
Availability: with dip_cor_flag = True and efield_flag = True.
Description: Added a dipole correction to the bare ionic potential.
True:A dipole correction is also added to the bare ionic potential.
False: A dipole correction is not added to the bare ionic potential.
Note: If you want no electric field, parameter efield_amp should be zero. Must be used ONLY in a slab geometry for surface alculations, with the discontinuity FALLING IN THE EMPTY SPACE.
Default: False
efield_dir
Type: Integer
Availability: with efield_flag = True.
Description: The direction of the electric field or dipole correction is parallel to the reciprocal lattice vector, so the potential is constant in planes defined by FFT grid points, efield_dir can set to 0, 1 or 2.
0: parallel to \(b_1=\frac{2\pi(a_2\times a_3)}{a_1\cdot(a_2\times a_3)}\)
1: parallel to \(b_2=\frac{2\pi(a_3\times a_1)}{a_1\cdot(a_2\times a_3)}\)
2: parallel to \(b_3=\frac{2\pi(a_1\times a_2)}{a_1\cdot(a_2\times a_3)}\)
Default: 2
efield_pos_max
Type: Real
Availability: with efield_flag = True.
Description: Position of the maximum of the saw-like potential along crystal axis efield_dir, within the unit cell, 0 <= efield_pos_max < 1.
Default: Autoset to
center of vacuum - width of vacuum / 20
efield_pos_dec
Type: Real
Availability: with efield_flag = True.
Description: Zone in the unit cell where the saw-like potential decreases, 0 < efield_pos_dec < 1.
Default: Autoset to
width of vacuum / 10
efield_amp
Type: Real
Availability: with efield_flag = True.
Description: Amplitude of the electric field. The saw-like potential increases with slope efield_amp in the region from efield_pos_max+efield_pos_dec-1) to (efield_pos_max), then decreases until (efield_pos_max+efield_pos_dec), in units of the crystal vector efield_dir.
Note: The change of slope of this potential must be located in the empty region, or else unphysical forces will result.
Default: 0.0
Unit: a.u., 1 a.u. = 51.4220632*10^10 V/m.
Gate field (compensating charge)
These variables are relevant to gate field (compensating charge) Detailed introduction
gate_flag
Type: Boolean
Description: Controls the addition of compensating charge by a charged plate for charged cells.
true: A charged plate is placed at the zgate position to add compensating charge. The direction is determined by efield_dir.
false: No compensating charge is added.
Default: false
zgate
Type: Real
Description: position of the charged plate in the unit cell
Unit: Unit cell size
Default: 0.5
Constraints: 0 <= zgate < 1
block
Type: Boolean
Description: Controls the addition of a potential barrier to prevent electron spillover.
true: A potential barrier is added from block_down to block_up with a height of block_height. If dip_cor_flag is set to true, efield_pos_dec is used to smoothly increase and decrease the potential barrier.
false: No potential barrier is added.
Default: false
block_down
Type: Real
Description: lower beginning of the potential barrier
Unit: Unit cell size
Default: 0.45
Constraints: 0 <= block_down < block_up < 1
block_up
Type: Real
Description: upper beginning of the potential barrier
Unit: Unit cell size
Default: 0.55
Constraints: 0 <= block_down < block_up < 1
block_height
Type: Real
Description: height of the potential barrier
Unit: Rydberg
Default: 0.1
Exact Exchange
These variables are relevant when using hybrid functionals.
Availablity: dft_functional==hse/hf/pbe0/scan0/opt_orb or rpa==True, and basis_type==lcao/lcao_in_pw
exx_hybrid_alpha
Type: Real
Description: fraction of Fock exchange in hybrid functionals, so that \(E_{X}=\alpha E_{X}+(1-\alpha)E_{X,\text{LDA/GGA}}\)
Default:
1: if dft_functional==hf
0.25: else
exx_hse_omega
Type: Real
Description: range-separation parameter in HSE functional, such that \(1/r=\text{erfc}(\omega r)/r+\text{erf}(\omega r)/r\)
Default: 0.11
exx_separate_loop
Type: Boolean
Description: There are two types of iterative approaches provided by ABACUS to evaluate Fock exchange.
False: Start with a GGA-Loop, and then Hybrid-Loop, in which EXX Hamiltonian \(H_{exx}\) is updated with electronic iterations.
True: A two-step method is employed, i.e. in the inner iterations, density matrix is updated, while in the outer iterations, \(H_{exx}\) is calculated based on density matrix that converges in the inner iteration.
Default: True
exx_hybrid_step
Type: Integer
Availability: exx_separate_loop==1
Description: the maximal iteration number of the outer-loop, where the Fock exchange is calculated
Default: 100
exx_mixing_beta
Type: Real
Availability: exx_separate_loop==1
Description: mixing_beta for densty matrix in each iteration of the outer-loop
Default: 1.0
exx_lambda
Type: Real
Availability: basis_type==lcao_in_pw
Description: It is used to compensate for divergence points at G=0 in the evaluation of Fock exchange using lcao_in_pw method.
Default: 0.3
exx_pca_threshold
Type: Real
Description: To accelerate the evaluation of four-center integrals (\(ik|jl\)), the product of atomic orbitals are expanded in the basis of auxiliary basis functions (ABF): \(\Phi_{i}\Phi_{j}\sim C^{k}_{ij}P_{k}\). The size of the ABF (i.e. number of \(P_{k}\)) is reduced using principal component analysis. When a large PCA threshold is used, the number of ABF will be reduced, hence the calculation becomes faster. However, this comes at the cost of computational accuracy. A relatively safe choice of the value is 1e-4.
Default: 1E-4
exx_c_threshold
Type: Real
Description: See also the entry exx_pca_threshold. Smaller components (less than exx_c_threshold) of the \(C^{k}_{ij}\) matrix are neglected to accelerate calculation. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 1e-4.
Default: 1E-4
exx_v_threshold
Type: Real
Description: See also the entry exx_pca_threshold. With the approximation \(\Phi_{i}\Phi_{j}\sim C^{k}_{ij}P_{k}\), the four-center integral in Fock exchange is expressed as \((ik|jl)=\Sigma_{a,b}C^{a}_{ij}V_{ab}C^{b}_{kl}\), where \(V_{ab}=(P_{a}|P_{b})\) is a double-center integral. Smaller values of the V matrix can be truncated to accelerate calculation. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 0, i.e. no truncation.
Default: 1E-1
exx_dm_threshold
Type: Real
Description: The Fock exchange can be expressed as \(\Sigma_{k,l}(ik|jl)D_{kl}\) where D is the density matrix. Smaller values of the density matrix can be truncated to accelerate calculation. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 1e-4.
Default: 1E-4
exx_c_grad_threshold
Type: Real
Description: See also the entry exx_pca_threshold. \(\nabla C^{k}_{ij}\) is used in force and stress. Smaller components (less than exx_c_grad_threshold) of the \(\nabla C^{k}_{ij}\) matrix are neglected to accelerate calculation. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 1e-4.
Default: 1E-4
exx_v_grad_threshold
Type: Real
Description: See also the entry exx_pca_threshold. With the approximation \(\Phi_{i}\Phi_{j}\sim C^{k}_{ij}P_{k}\), the four-center integral in Fock exchange is expressed as \((ik|jl)=\Sigma_{a,b}C^{a}_{ij}V_{ab}C^{b}_{kl}\), where \(V_{ab}=(P_{a}|P_{b})\) is a double-center integral. \(\nabla V_{ab}\) is used in force and stress. Smaller values of the V matrix can be truncated to accelerate calculation. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 0, i.e. no truncation.
Default: 1E-1
exx_schwarz_threshold
Type: Real
Description: In practice the four-center integrals are sparse, and using Cauchy-Schwartz inequality, we can find an upper bound of each integral before carrying out explicit evaluations. Those that are smaller than exx_schwarz_threshold will be truncated. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 1e-5. (Currently not used)
Default: 0
exx_cauchy_threshold
Type: Real
Description: In practice the Fock exchange matrix is sparse, and using Cauchy-Schwartz inequality, we can find an upper bound of each matrix element before carrying out explicit evaluations. Those that are smaller than exx_cauchy_threshold will be truncated. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 1e-7.
Default: 1E-7
exx_cauchy_force_threshold
Type: Real
Description: In practice the Fock exchange matrix in force is sparse, and using Cauchy-Schwartz inequality, we can find an upper bound of each matrix element before carrying out explicit evaluations. Those that are smaller than exx_cauchy_force_threshold will be truncated. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 1e-7.
Default: 1E-7
exx_cauchy_stress_threshold
Type: Real
Description: In practice the Fock exchange matrix in stress is sparse, and using Cauchy-Schwartz inequality, we can find an upper bound of each matrix element before carrying out explicit evaluations. Those that are smaller than exx_cauchy_stress_threshold will be truncated. The larger the threshold is, the faster the calculation and the lower the accuracy. A relatively safe choice of the value is 1e-7.
Default: 1E-7
exx_ccp_threshold
Type: Real
Description: It is related to the cutoff of on-site Coulomb potentials. (Currently not used)
Default: 1e-8
exx_ccp_rmesh_times
Type: Real
Description: This parameter determines how many times larger the radial mesh required for calculating Columb potential is to that of atomic orbitals. For HSE, setting it to 1 is enough. But for PBE0, a much larger number must be used.
Default:
1.5: if dft_functional==hse
5: else
exx_distribute_type
Type: String
Description: When running in parallel, the evaluation of Fock exchange is done by distributing atom pairs on different processes, then gather the results. exx_distribute_type governs the mechanism of distribution. Available options are
htime
,order
,kmean1
andkmeans2
.order
: Atom pairs are simply distributed by their orders.htime
: The balance in time is achieved on each processor, hence if the memory is sufficient, this is the recommended method.kmeans1
,kmeans2
: Two methods where the k-means clustering method is used to reduce memory requirement. They might be necessary for very large systems. (Currently not used)
Default:
htime
exx_opt_orb_lmax
Type: Integer
Availability: dft_functional==opt_orb
Description: The maximum l of the spherical Bessel functions, when the radial part of opt-ABFs are generated as linear combinations of spherical Bessel functions. A reasonable choice is 2.
Default: 0
exx_opt_orb_ecut
Type: Real
Availability: dft_functional==opt_orb
Description: The cut-off of plane wave expansion, when the plane wave basis is used to optimize the radial ABFs. A reasonable choice is 60.
Default: 0
Unit: Ry
exx_opt_orb_tolerence
Type: Real
Availability: dft_functional==opt_orb
Description: The threshold when solving for the zeros of spherical Bessel functions. A reasonable choice is 1e-12.
Default: 0
exx_real_number
Type: Boolean
Description:
True: Enforce LibRI to use
double
data type.False: Enforce LibRI to use
complex
data type.
Default: depends on the gamma_only option
True: if gamma_only
False: else
rpa_ccp_rmesh_times
Type: Real
Description: How many times larger the radial mesh required is to that of atomic orbitals in the postprocess calculation of the bare Coulomb matrix for RPA, GW, etc.
Default: 10
Molecular dynamics
These variables are used to control molecular dynamics calculations. For more information, please refer to md.md in detail.
md_type
Type: String
Description: Control the algorithm to integrate the equation of motion for molecular dynamics (MD), see md.md in detail.
fire: a MD-based relaxation algorithm, named fast inertial relaxation engine.
nve: NVE ensemble with velocity Verlet algorithm.
nvt: NVT ensemble, see md_thermostat in detail.
npt: Nose-Hoover style NPT ensemble, see md_pmode in detail.
langevin: NVT ensemble with Langevin thermostat, see md_damp in detail.
msst: MSST method, see msst_direction, msst_vel, msst_qmass, msst_vis, msst_tscale in detail.
Default: nvt
md_nstep
Type: Integer
Description: The total number of molecular dynamics steps.
Default: 10
md_dt
Type: Real
Description: The time step used in molecular dynamics calculations.
Default: 1.0
Unit: fs
md_thermostat
Type: String
Description: Specify the temperature control method used in NVT ensemble.
nhc: Nose-Hoover chain, see md_tfreq and md_tchain in detail.
anderson: Anderson thermostat, see md_nraise in detail.
berendsen: Berendsen thermostat, see md_nraise in detail.
rescaling: velocity Rescaling method 1, see md_tolerance in detail.
rescale_v: velocity Rescaling method 2, see md_nraise in detail.
Default: nhc
md_tfirst, md_tlast
Type: Real
Description: The temperature used in molecular dynamics calculations.
If
md_tfirst
is unset or less than zero, init_vel is autoset to betrue
. If init_vel istrue
, the initial temperature will be determined by the velocities read fromSTRU
. In this case, if velocities are unspecified inSTRU
, the initial temperature is set to zero.If
md_tfirst
is set to a positive value and init_vel istrue
simultaneously, please make sure they are consistent, otherwise abacus will exit immediately.Note that
md_tlast
is only used in NVT/NPT simulations. Ifmd_tlast
is unset or less than zero,md_tlast
is set tomd_tfirst
. Ifmd_tlast
is set to be different frommd_tfirst
, ABACUS will automatically change the temperature frommd_tfirst
tomd_tlast
.Default: No default
Unit: K
md_restart
Type: Boolean
Description: Control whether to restart molecular dynamics calculations and time-dependent density functional theory calculations.
True: ABACUS will read in
${read_file_dir}/Restart_md.dat
to determine the current step${md_step}
, then read in the correspondingSTRU_MD_${md_step}
in the folderOUT.$suffix/STRU/
automatically. For tddft, ABACUS will also read inWFC_NAO_K${kpoint}
of the last step (You need to set out_wfc_lcao=1 and out_app_flag=0 to obtain this file).False: ABACUS will start molecular dynamics calculations normally from the first step.
Default: False
md_restartfreq
Type: Integer
Description: The output frequency of
OUT.${suffix}/Restart_md.dat
and structural files in the directoryOUT.${suffix}/STRIU/
, which are used to restart molecular dynamics calculations, see md_restart in detail.Default: 5
md_dumpfreq
Type: Integer
Description: The output frequency of
OUT.${suffix}/MD_dump
in molecular dynamics calculations, which including the information of lattices and atoms.Default: 1
dump_force
Type: Boolean
Description: Whether to output atomic forces into the file
OUT.${suffix}/MD_dump
.Default: True
dump_vel
Type: Boolean
Description: Whether to output atomic velocities into the file
OUT.${suffix}/MD_dump
.Default: True
dump_virial
Type: Boolean
Description: Whether to output lattice virials into the file
OUT.${suffix}/MD_dump
.Default: True
md_seed
Type: Integer
Description: The random seed to initialize random numbers used in molecular dynamics calculations.
< 0: No srand() function is called.
>= 0: The function srand(md_seed) is called.
Default: -1
md_tfreq
Type: Real
Description: Control the frequency of temperature oscillations during the simulation. If it is too large, the temperature will fluctuate violently; if it is too small, the temperature will take a very long time to equilibrate with the atomic system.
Note: It is a system-dependent empirical parameter, ranging from 1/(40*md_dt) to 1/(100*md_dt). An improper choice might lead to the failure of jobs.
Default: 1/40/md_dt
Unit: \(\mathrm{fs^{-1}}\)
md_tchain
Type: Integer
Description: Number of thermostats coupled with the particles in the NVT/NPT ensemble based on the Nose-Hoover style non-Hamiltonian equations of motion.
Default: 1
md_pmode
Type: String
Description: Specify the cell fluctuation mode in NPT ensemble based on the Nose-Hoover style non-Hamiltonian equations of motion.
iso: The three diagonal elements of the lattice are fluctuated isotropically.
aniso: The three diagonal elements of the lattice are fluctuated anisotropically.
tri: The lattice must be a lower-triangular matrix, and all six freedoms are fluctuated.
Default: iso
Relavent: md_tfreq, md_tchain, md_pcouple, md_pfreq, and md_pchain.
ref_cell_factor
Type: Real
Description: Construct a reference cell bigger than the initial cell. The reference cell has to be large enough so that the lattice vectors of the fluctuating cell do not exceed the reference lattice vectors during MD. Typically, 1.02 ~ 1.10 is sufficient. However, the cell fluctuations depend on the specific system and thermodynamic conditions. So users must test for a proper choice. This parameters should be used in conjunction with erf_ecut, erf_height, and erf_sigma.
Default: 1.0
md_pcouple
Type: String
Description: The coupled lattice vectors will scale proportionally in NPT ensemble based on the Nose-Hoover style non-Hamiltonian equations of motion.
none: Three lattice vectors scale independently.
xyz: Lattice vectors x, y, and z scale proportionally.
xy: Lattice vectors x and y scale proportionally.
xz: Lattice vectors x and z scale proportionally.
yz: Lattice vectors y and z scale proportionally.
Default: none
md_pfirst, md_plast
Type: Real
Description: The target pressure used in NPT ensemble simulations, the default value of
md_plast
ismd_pfirst
. Ifmd_plast
is set to be different frommd_pfirst
, ABACUS will automatically change the target pressure frommd_pfirst
tomd_plast
.Default: -1.0
Unit: kbar
md_pfreq
Type: Real
Description: The frequency of pressure oscillations during the NPT ensemble simulation. If it is too large, the pressure will fluctuate violently; if it is too small, the pressure will take a very long time to equilibrate with the atomic system.
Note: It is a system-dependent empirical parameter. An improper choice might lead to the failure of jobs.
Default: 1/400/md_dt
Unit: \(\mathrm{kbar^{-1}}\)
md_pchain
Type: Integer
Description: The number of thermostats coupled with the barostat in the NPT ensemble based on the Nose-Hoover style non-Hamiltonian equations of motion.
Default: 1
lj_rule
Type: Integer
Description: The Lennard-Jones potential between two atoms equals: $\(V_{LJ}(r_{ij})=4\epsilon_{ij}\left(\left(\frac{\sigma_{ij}}{r_{ij}}\right)^{12}-\left(\frac{\sigma_{ij}}{r_{ij}}\right)^{6}\right)=\frac{C_{ij}^{(12)}}{{r_{ij}}^{12}}-\frac{C_{ij}^{(6)}}{{r_{ij}}^{6}}.\)$
The parameters lj_epsilon and lj_sigma should be multiple-component vectors. For example, there are two choices in the calculations of 3 atom species:
Supply six-component vectors that describe the interactions between all possible atom pairs. The six-component vectors represent lower triangular symmetric matrixs, and the correspondence between the vector component \(\sigma _k\) and the matrix element \(\sigma (i,j)\) is $\(k= i(i+1)/2 +j\)$
Supply three-component vectors that describe the interactions between atoms of the same species. In this case, two types of combination rules can be used to construct non-diagonal elements in the parameter matrix.
1: geometric average: $\(\begin{array}{rcl}C_{ij}^{(6)}&=&\left(C_{ii}^{(6)}C_{jj}^{(6)}\right)^{1/2}\\C_{ij}^{(12)}&=&\left(C_{ii}^{(12)}C_{jj}^{(12)}\right)^{1/2}\end{array}\)$
2: arithmetic average: $\(\begin{array}{rcl}\sigma_{ij}&=&\frac{1}{2}\left(\sigma_{ii}+\sigma_{jj}\right)\\ \epsilon_{ij}&=&\left(\epsilon_{ii}\epsilon_{jj}\right)^{1/2}\end{array}\)$
Default: 2
lj_eshift
Type: Boolean
Description: It True, the LJ potential is shifted by a constant such that it is zero at the cut-off distance.
Default: False
lj_rcut
Type: Real
Description: Cut-off radius for Leonard Jones potential, beyond which the interaction will be neglected. It can be a single value, which means that all pairs of atoms types share the same cut-off radius. Otherwise, it should be a multiple-component vector, containing \(N(N+1)/2\) values, see details in lj_rule.
Default: No default
Unit: Angstrom
lj_epsilon
Type: Real
Description: The vector representing the \(\epsilon\) matrix for Leonard Jones potential. See details in lj_rule.
Default: No default
Unit: eV
lj_sigma
Type: Real
Description: The vector representing the \(\sigma\) matrix for Leonard Jones potential. See details in lj_rule.
Default: No default
Unit: Angstrom
pot_file
Type: String
Description: The filename of DP potential files, see md.md in detail.
Default: graph.pb
msst_direction
Type: Integer
Description: The direction of the shock wave in the MSST method.
0: x direction
1: y direction
2: z direction
Default: 2
msst_vel
Type: Real
Description: The velocity of the shock wave in the MSST method.
Default: 0.0
Unit: Angstrom/fs
msst_vis
Type: Real
Description: Artificial viscosity in the MSST method.
Default: 0.0
Unit: g/(mol*Angstrom*fs)
msst_tscale
Type: Real
Description: The reduction percentage of the initial temperature used to compress volume in the MSST method.
Default: 0.01
msst_qmass
Type: Real
Description: Inertia of the extended system variable. You should set a number larger than 0.
Default: No default
Unit: \(\mathrm{g^{2}/(mol^{2}*Angstrom^{4})}\)
md_damp
Type: Real
Description: The damping parameter used to add fictitious force in the Langevin method.
Default: 1.0
Unit: fs
md_tolerance
Type: Real
Description: Thr temperature tolerance for velocity rescaling. Velocities are rescaled if the current and target temperature differ more than
md_tolerance
.Default: 100.0
Unit: K
md_nraise
Type: Integer
Description:
Anderson: The “collision frequency” parameter is given as 1/
md_nraise
.Berendsen: The “rise time” parameter is given in units of the time step: tau =
md_nraise
*md_dt
, somd_dt
/tau = 1/md_nraise
.Rescale_v: Every
md_nraise
steps the current temperature is rescaled to the target temperature.
Default: 1
cal_syns
Type: Boolean
Description: Whether the asynchronous overlap matrix is calculated for Hefei-NAMD.
Default: False
dmax
Type: Real
Description: The maximum displacement of all atoms in one step. This parameter is useful when cal_syns = True.
Default: 0.01
Unit: bohr
DFT+U correction
These variables are used to control DFT+U correlated parameters
dft_plus_u
Type: Integer
Description: Determines whether to calculate the plus U correction, which is especially important for correlated electrons.
1: Calculate plus U correction with radius-adjustable localized projections (with parameter
onsite_radius
).2: Calculate plus U correction using first zeta of NAOs as projections (this is old method for testing).
0: Do not calculate plus U correction.
Default: 0
orbital_corr
Type: Integer
Description: Specifies which orbits need plus U correction for each atom type (\(l_1,l_2,l_3,\ldots\) for atom type 1, 2, 3, respectively).
-1: The plus U correction will not be calculated for this atom.
1: For p-electron orbits, the plus U correction is needed.
2: For d-electron orbits, the plus U correction is needed.
3: For f-electron orbits, the plus U correction is needed.
Default: -1
hubbard_u
Type: Real
Description: Specifies the Hubbard Coulomb interaction parameter U (eV) in plus U correction, which should be specified for each atom unless the Yukawa potential is used.
Note: Since only the simplified scheme by Duradev is implemented, the ‘U’ here is actually U-effective, which is given by Hubbard U minus Hund J.
Default: 0.0
yukawa_potential
Type: Boolean
Description: Determines whether to use the local screen Coulomb potential method to calculate the values of U and J.
True:
hubbard_u
does not need to be specified.False:
hubbard_u
does need to be specified.
Default: False
yukawa_lambda
Type: Real
Availability: DFT+U with
yukawa_potential
= True.Description: The screen length of Yukawa potential. If left to default, the screen length will be calculated as an average of the entire system. It’s better to stick to the default setting unless there is a very good reason.
Default: Calculated on the fly.
uramping
Type: Real
Unit: eV
Availability: DFT+U calculations with
mixing_restart > 0
.Description: Once
uramping
> 0.15 eV. DFT+U calculations will start SCF with U = 0 eV, namely normal LDA/PBE calculations. Once SCF restarts whendrho<mixing_restart
, U value will increase byuramping
eV. SCF will repeat above calcuations until U values reach target defined inhubbard_u
. As foruramping=1.0 eV
, the recommendations ofmixing_restart
is around5e-4
.Default: -1.0.
omc
Type: Integer
Description: The parameter controls the form of occupation matrix control used.
0: No occupation matrix control is performed, and the onsite density matrix will be calculated from wavefunctions in each SCF step.
1: The first SCF step will use an initial density matrix read from a file named
[initial_onsite.dm](http://initial_onsite.dm/)
, but for later steps, the onsite density matrix will be updated.2: The same onsite density matrix from
initial_onsite.dm
will be used throughout the entire calculation.
Note : The easiest way to create
initial_onsite.dm
is to run a DFT+U calculation, look for a file namedonsite.dm
in the OUT.prefix directory, and make replacements there. The format of the file is rather straight-forward.
Default: 0
onsite_radius
Type: Real
Availability:
dft_plus_u
is set to 1Description:
The
Onsite-radius
parameter facilitates modulation of the single-zeta portion of numerical atomic orbitals for projections for DFT+U.The modulation algorithm includes a smooth truncation applied directly to the tail of the original orbital, followed by normalization. Consider the function: $\( g(r;\sigma)=\begin{cases} 1-\exp\left(-\frac{(r-r_c)^2}{2\sigma^2}\right), & r < r_c\\ 0, & r \geq r_c \end{cases} \)$
where \(\sigma\) is a parameter that controls the smoothing interval. A normalized function truncated smoothly at \(r_c\) can be represented as:
\[ \alpha(r) = \frac{\chi(r)g(r;\sigma)}{\langle\chi(r)g(r;\sigma), \chi(r)g(r;\sigma)\rangle} \]To find an appropriate \(\sigma\), the optimization process is as follows:
Maximizing the overlap integral under a normalization constraint is equivalent to minimizing an error function:
\[ \min \langle \chi(r)-\alpha(r), \chi(r)-\alpha(r)\rangle \quad \text{subject to} \quad \langle \alpha(r),\alpha(r)\rangle=1 \]Similar to the process of generating numerical atomic orbitals, this optimization choice often induces additional oscillations in the outcome. To suppress these oscillations, we may include a derivative term in the objective function (\(f'(r)\equiv \mathrm{d}f(r)/\mathrm{d}r\)):
\[ \min \left[\gamma\langle \chi(r)-\alpha(r), \chi(r)-\alpha(r)\rangle + \langle \chi'(r)-\alpha'(r), \chi'(r)-\alpha'(r)\rangle\right] \quad \text{subject to} \quad \langle \alpha(r),\alpha(r)\rangle=1 \]where \(\gamma\) is a parameter that adjusts the relative weight of the error function to the derivative error function.
Unit: Bohr
Default: 5.0
vdW correction
These variables are used to control vdW-corrected related parameters.
vdw_method
Type: String
Description: Specifies the method used for Van der Waals (VdW) correction. Available options are:
d2
: Grimme’s D2 dispersion correction methodd3_0
: Grimme’s DFT-D3(0) dispersion correction methodd3_bj
: Grimme’s DFTD3(BJ) dispersion correction methodnone
: no vdW correction
Default: none
vdw_s6
Type: Real
Availability:
vdw_method
is set tod2
,d3_0
, ord3_bj
Description: This scale factor is used to optimize the interaction energy deviations in van der Waals (vdW) corrected calculations. The recommended values of this parameter are dependent on the chosen vdW correction method and the DFT functional being used. For DFT-D2, the recommended values are 0.75 (PBE), 1.2 (BLYP), 1.05 (B-P86), 1.0 (TPSS), and 1.05 (B3LYP). For DFT-D3, recommended values with different DFT functionals can be found on the here. The default value of this parameter in ABACUS is set to be the recommended value for PBE.
Default:
0.75: if
vdw_method
is set tod2
1.0: if
vdw_method
is set tod3_0
ord3_bj
vdw_s8
Type: Real
Availability:
vdw_method
is set tod3_0
ord3_bj
Description: This scale factor is relevant for D3(0) and D3(BJ) van der Waals (vdW) correction methods. The recommended values of this parameter with different DFT functionals can be found on the webpage. The default value of this parameter in ABACUS is set to be the recommended value for PBE.
Default:
0.722: if
vdw_method
is set tod3_0
0.7875: if
vdw_method
is set tod3_bj
vdw_a1
Type: Real
Availability:
vdw_method
is set tod3_0
ord3_bj
Description: This damping function parameter is relevant for D3(0) and D3(BJ) van der Waals (vdW) correction methods. The recommended values of this parameter with different DFT functionals can be found on the webpage. The default value of this parameter in ABACUS is set to be the recommended value for PBE.
Default:
1.217: if
vdw_method
is set tod3_0
0.4289: if
vdw_method
is set tod3_bj
vdw_a2
Type: Real
Availability:
vdw_method
is set tod3_0
ord3_bj
Description: This damping function parameter is only relevant for D3(0) and D3(BJ) van der Waals (vdW) correction methods. The recommended values of this parameter with different DFT functionals can be found on the webpage. The default value of this parameter in ABACUS is set to be the recommended value for PBE.
Default:
1.0: if
vdw_method
is set tod3_0
4.4407: if
vdw_method
is set tod3_bj
vdw_d
Type: Real
Availability:
vdw_method
is set tod2
Description: Controls the damping rate of the damping function in the DFT-D2 method.
Default: 20
vdw_abc
Type: Integer
Availability:
vdw_method
is set tod3_0
ord3_bj
Description: Determines whether three-body terms are calculated for DFT-D3 methods.
True: ABACUS will calculate the three-body term.
False: The three-body term is not included.
Default: False
vdw_C6_file
Type: String
Availability:
vdw_method
is set tod2
Description: Specifies the name of the file containing \(C_6\) parameters for each element when using the D2 method. If not set, ABACUS uses the default \(C_6\) parameters (Jnm6/mol) stored in the program. To manually set the \(C_6\) parameters, provide a file containing the parameters. An example is given by:
H 0.1 Si 9.0
Namely, each line contains the element name and the corresponding \(C_6\) parameter.
Default: default
vdw_C6_unit
Type: String
Availability:
vdw_C6_file
is not defaultDescription: Specifies the unit of the provided \(C_6\) parameters in the D2 method. Available options are:
Jnm6/mol
(J·nm^6/mol)eVA
(eV·Angstrom)
Default: Jnm6/mol
vdw_R0_file
Type: String
Availability:
vdw_method
is set tod2
Description: Specifies the name of the file containing \(R_0\) parameters for each element when using the D2 method. If not set, ABACUS uses the default \(R_0\) parameters (Angstrom) stored in the program. To manually set the \(R_0\) parameters, provide a file containing the parameters. An example is given by:
Li 1.0 Cl 2.0
Namely, each line contains the element name and the corresponding \(R_0\) parameter.
Default: default
vdw_R0_unit
Type: String
Availability:
vdw_R0_file
is not defaultDescription: Specifies the unit for the \(R_0\) parameters in the D2 method when manually set by the user. Available options are:
A
(Angstrom)Bohr
Default: A
vdw_cutoff_type
Type: String
Description: Determines the method used for specifying the cutoff radius in periodic systems when applying Van der Waals correction. Available options are:
radius
: The supercell is selected within a sphere centered at the origin with a radius defined byvdw_cutoff_radius
.period
: The extent of the supercell is explicitly specified using thevdw_cutoff_period
keyword.
Default: radius
vdw_cutoff_radius
Type: Real
Availability:
vdw_cutoff_type
is set toradius
Description: Defines the radius of the cutoff sphere when
vdw_cutoff_type
is set toradius
. The default values depend on the chosenvdw_method
.Default:
56.6918 if
vdw_method
is set tod2
95 if
vdw_method
is set tod3_0
ord3_bj
Unit: defined by
vdw_radius_unit
(defaultBohr
)
vdw_radius_unit
Type: String
Availability:
vdw_cutoff_type
is set toradius
Description: specify the unit of
vdw_cutoff_radius
. Available options are:A
(Angstrom)Bohr
Default: Bohr
vdw_cutoff_period
Type: Integer Integer Integer
Availability:
vdw_cutoff_type
is set toperiod
Description: The three integers supplied here explicitly specify the extent of the supercell in the directions of the three basis lattice vectors.
Default: 3 3 3
vdw_cn_thr
Type: Real
Availability:
vdw_method
is set tod3_0
ord3_bj
Description: The cutoff radius when calculating coordination numbers.
Default: 40
Unit: defined by
vdw_cn_thr_unit
(default:Bohr
)
vdw_cn_thr_unit
Type: String
Description: Unit of the coordination number cutoff (
vdw_cn_thr
). Available options are:A
(Angstrom)Bohr
Default: Bohr
Berry phase and wannier90 interface
These variables are used to control berry phase and wannier90 interface parameters. Detail introduce
berry_phase
Type: Boolean
Description: controls the calculation of Berry phase
true: Calculate Berry phase.
false: Do not calculate Berry phase.
Default: false
gdir
Type: Integer
Description: the direction of the polarization in the lattice vector for Berry phase calculation
1: Calculate the polarization in the direction of the lattice vector a_1 defined in the STRU file.
2: Calculate the polarization in the direction of the lattice vector a_2 defined in the STRU file.
3: Calculate the polarization in the direction of the lattice vector a_3 defined in the STRU file.
Default: 3
towannier90
Type: Integer
Description: Controls the generation of files for the Wannier90 code.
1: Generate files for the Wannier90 code.
0: Do not generate files for the Wannier90 code.
Default: 0
nnkpfile
Type: String
Description: the file name generated when running “wannier90 -pp …” command
Default: seedname.nnkp
wannier_method
Type: Integer
Description: Only available on LCAO basis, using different methods to generate “*.mmn” file and “*.amn” file.
1: Calculated using the
lcao_in_pw
method, the calculation accuracy can be improved by increasingecutwfc
to maintain consistency with the pw basis set results.2: The overlap between atomic orbitals is calculated using grid integration. The radial grid points are generated using the Gauss-Legendre method, while the spherical grid points are generated using the Lebedev-Laikov method.
Default: 1
wannier_spin
Type: String
Description: the spin direction for the Wannier function calculation when nspin is set to 2
up
: Calculate spin up for the Wannier function.down
: Calculate spin down for the Wannier function.
Default:
up
out_wannier_mmn
Type: Bool
Description: write the “*.mmn” file or not.
0: don’t write the “*.mmn” file.
1: write the “*.mmn” file.
Default: 1
out_wannier_amn
Type: Bool
Description: write the “*.amn” file or not.
0: don’t write the “*.amn” file.
1: write the “*.amn” file.
Default: 1
out_wannier_eig
Type: Bool
Description: write the “*.eig” file or not.
0: don’t write the “*.eig” file.
1: write the “*.eig” file.
Default: 1
out_wannier_unk
Type: Bool
Description: write the “UNK.*” file or not.
0: don’t write the “UNK.*” file.
1: write the “UNK.*” file.
Default: 0
out_wannier_wvfn_formatted
Type: Bool
Description: write the “UNK.*” file in ASCII format or binary format.
0: write the “UNK.*” file in binary format.
1: write the “UNK.*” file in ASCII format (text file format).
Default: 1
TDDFT: time dependent density functional theory
td_edm
Type: Integer
Description: the method to calculate the energy density matrix
0: new method (use the original formula).
1: old method (use the formula for ground state).
Default: 0
td_print_eij
Type: Real
Description:
<0: don’t print \(E_{ij}\).
>=0: print the \(E_{ij}\ (<\psi_i|H|\psi_j>\)) elements which are larger than td_print_eij.
Default: -1
td_propagator
Type: Integer
Description: method of propagator
0: Crank-Nicolson.
1: 4th Taylor expansions of exponential.
2: enforced time-reversal symmetry (ETRS).
Default: 0
td_vext
Type: Boolean
Description:
True: add a laser material interaction (extern laser field).
False: no extern laser field.
Default: False
td_vext_dire
Type: String
Description: If
td_vext
is True, the td_vext_dire is a string to set the number of electric fields, liketd_vext_dire 1 2
representing external electric field is added to the x and y axis at the same time. Parameters of electric field can also be written as a string, liketd_gauss_phase 0 1.5707963267948966
representing the Gauss field in the x and y directions has a phase delay of Pi/2. See below for more parameters of electric field.1: the direction of external light field is along x axis.
2: the direction of external light field is along y axis.
3: the direction of external light field is along z axis.
Default: 1
td_stype
Type: Integer
Description: type of electric field in space domain
0: length gauge.
1: velocity gauge.
Default: 0
td_ttype
Type: Integer
Description: type of electric field in time domain
0: Gaussian type function.
1: Trapezoid function.
2: Trigonometric function.
3: Heaviside function.
4: HHG function.
Default: 0
td_tstart
Type: Integer
Description: number of steps where electric field starts
Default: 1
td_tend
Type: Integer
Description: number of steps where electric field ends
Default: 100
td_lcut1
Type: Real
Description: cut1 of interval in length gauge
E = E0 , cut1<x<cut2
E = -E0/(cut1+1-cut2) , x<cut1 or cut2<x<1Default: 0.05
td_lcut2
Type: Real
Description: cut2 of interval in length gauge
E = E0 , cut1<x<cut2
E = -E0/(cut1+1-cut2) , x<cut1 or cut2<x<1Default: 0.05
td_gauss_freq
Type: Real
Description: frequency (freq) of Gauss type electric field (fs^-1)
amp*cos(2pi*freq(t-t0)+phase)exp(-(t-t0)^2/2sigma^2)Default: 22.13
td_gauss_phase
Type: Real
Description: phase of Gauss type electric field
amp*(2pi*freq(t-t0)+phase)exp(-(t-t0)^2/2sigma^2)Default: 0.0
td_gauss_sigma
Type: Real
Description: sigma of Gauss type electric field (fs)
amp*cos(2pi*freq(t-t0)+phase)exp(-(t-t0)^2/2sigma^2)Default: 30.0
td_gauss_t0
Type: Real
Description: step number of time center (t0) of Gauss type electric field
amp*cos(2pi*freq(t-t0)+phase)exp(-(t-t0)^2/2sigma^2)Default: 100
td_gauss_amp
Type: Real
Description: amplitude (amp) of Gauss type electric field (V/Angstrom)
amp*cos(2pi*freq(t-t0)+phase)exp(-(t-t0)^2/2sigma^2)Default: 0.25
td_trape_freq
Type: Real
Description: frequency (freq) of Trapezoid type electric field (fs^-1)
E = amp*cos(2pi*freq*t+phase) t/t1 , t<t1
E = amp*cos(2pi*freq*t+phase) , t1<t<t2
E = amp*cos(2pi*freq*t+phase) (1-(t-t2)/(t3-t2)) , t2<t<t3
E = 0 , t>t3Default: 1.60
td_trape_phase
Type: Real
Description: phase of Trapezoid type electric field
E = amp*cos(2pi*freq*t+phase) t/t1 , t<t1
E = amp*cos(2pi*freq*t+phase) , t1<t<t2
E = amp*cos(2pi*freq*t+phase) (1-(t-t2)/(t3-t2)) , t2<t<t3
E = 0 , t>t3Default: 0.0
td_trape_t1
Type: Real
Description: step number of time interval 1 (t1) of Trapezoid type electric field
E = amp*cos(2pi*freq*t+phase) t/t1 , t<t1
E = amp*cos(2pi*freq*t+phase) , t1<t<t2
E = amp*cos(2pi*freq*t+phase) (1-(t-t2)/(t3-t2)) , t2<t<t3
E = 0 , t>t3Default: 1875
td_trape_t2
Type: Real
Description: step number of time interval 2 (t2) of Trapezoid type electric field
E = amp*cos(2pi*freq*t+phase) t/t1 , t<t1
E = amp*cos(2pi*freq*t+phase) , t1<t<t2
E = amp*cos(2pi*freq*t+phase) (1-(t-t2)/(t3-t2)) , t2<t<t3
E = 0 , t>t3Default: 5625
td_trape_t3
Type: Real
Description: step number of time interval 3 (t3) of Trapezoid type electric field
E = amp*cos(2pi*freq*t+phase) t/t1 , t<t1
E = amp*cos(2pi*freq*t+phase) , t1<t<t2
E = amp*cos(2pi*freq*t+phase) (1-(t-t2)/(t3-t2)) , t2<t<t3
E = 0 , t>t3Default: 7500
td_trape_amp
Type: Real
Description: amplitude (amp) of Trapezoid type electric field (V/Angstrom)
E = amp*cos(2pi*freq*t+phase) t/t1 , t<t1
E = amp*cos(2pi*freq*t+phase) , t1<t<t2
E = amp*cos(2pi*freq*t+phase) (1-(t-t2)/(t3-t2)) , t2<t<t3
E = 0 , t>t3Default: 2.74
td_trigo_freq1
Type: Real
Description: frequency 1 (freq1) of Trigonometric type electric field (fs^-1)
amp*cos(2*pi*freq1*t+phase1)*sin(2*pi*freq2*t+phase2)^2Default: 1.164656
td_trigo_freq2
Type: Real
Description: frequency 2 (freq2) of Trigonometric type electric field (fs^-1)
amp*cos(2*pi*freq1*t+phase1)*sin(2*pi*freq2*t+phase2)^2Default: 0.029116
td_trigo_phase1
Type:Real
Description: phase 1 (phase1) of Trigonometric type electric field
amp*cos(2*pi*freq1*t+phase1)*sin(2*pi*freq2*t+phase2)^2Default: 0.0
td_trigo_phase2
Type: Real
Description: phase 2 (phase2) of Trigonometric type electric field
amp*cos(2*pi*freq1*t+phase1)*sin(2*pi*freq2*t+phase2)^2Default: 0.0
td_trigo_amp
Type: Real
Description: amplitude (amp) of Trigonometric type electric field (V/Angstrom)
amp*cos(2*pi*freq1*t+phase1)*sin(2*pi*freq2*t+phase2)^2Default: 2.74
td_heavi_t0
Type: Real
Description: step number of switch time (t0) of Heaviside type electric field
E = amp , t<t0
E = 0.0 , t>t0Default: 100
td_heavi_amp
Type: Real
Description: amplitude (amp) of Heaviside type electric field (V/Angstrom)
E = amp , t<t0
E = 0.0 , t>t0Default: 2.74
out_dipole
Type: Boolean
Description:
True: output dipole.
False: do not output dipole.
Default: False
out_current
Type: Boolean
Description:output current in real time TDDFT simulations with the velocity gauge
True: output current.
False: do not output current.
Default: False
out_current_k
Type: Boolean
Description:output tddft current for all k points.
True: output tddft current for all k points.
False: output current in total.
Default: False
out_efield
Type: Boolean
Description: output TDDFT Efield or not(V/Angstrom)
True: output efield.
False: do not output efield.
Default: False
out_vecpot
Type: Boolean
Description: output TDDFT Vector potential or not(a.u.)
True: output Vector potential in file “OUT.suffix/At.dat”
False: do not output Vector potential.
Default: False
init_vecpot_file
Type: Boolean
Description: Init vector potential through file or not
True: init vector potential from file “At.dat”.(a.u.) It consists of four columns, representing istep and vector potential on each direction.
False: calculate vector potential by integral of Efield
Default: False
ocp
Type: Boolean
Availability:
For PW and LCAO codes. if set to 1, occupations of bands will be setting of “ocp_set”.
For TDDFT in LCAO codes. if set to 1, occupations will be constrained since second ionic step.
For OFDFT, this feature can’t be used.
Description:
True: fix the occupations of bands.
False: do not fix the occupations of bands.
Default: False
ocp_set
Type: String
Description: If ocp is True, the ocp_set is a string to set the number of occupancy, like ‘1 10 * 1 0 1’ representing the 13 band occupancy, 12th band occupancy 0 and the rest 1, the code is parsing this string into an array through a regular expression.
Default: none
Variables useful for debugging
t_in_h
Type: Boolean
Description: Specify whether to include kinetic term in obtaining the Hamiltonian matrix.
0: No.
1: Yes.
Default: 1
vl_in_h
Type: Boolean
Description: Specify whether to include local pseudopotential term in obtaining the Hamiltonian matrix.
0: No.
1: Yes.
Default: 1
vnl_in_h
Type: Boolean
Description: Specify whether to include non-local pseudopotential term in obtaining the Hamiltonian matrix.
0: No.
1: Yes.
Default: 1
vh_in_h
Type: Boolean
Description: Specify whether to include Hartree potential term in obtaining the Hamiltonian matrix.
0: No.
1: Yes.
Default: 1
vion_in_h
Type: Boolean
Description: Specify whether to include local ionic potential term in obtaining the Hamiltonian matrix.
0: No.
1: Yes.
Default: 1
test_force
Type: Boolean
Description: Specify whether to output the detailed components in forces.
0: No.
1: Yes.
Default: 0
test_stress
Type: Boolean
Description: Specify whether to output the detailed components in stress.
0: No.
1: Yes.
Default: 0
colour
Type: Boolean
Description: Specify whether to set the colorful output in terminal.
0: No.
1: Yes.
Default: 0
test_skip_ewald
Type: Boolean
Description: Specify whether to skip the calculation of the ewald energy.
0: No.
1: Yes.
Default: 0
Electronic conductivities
Frequency-dependent electronic conductivities can be calculated with Kubo-Greenwood formula [Phys. Rev. B 83, 235120 (2011)].
Onsager coefficients:
\(L_{mn}(\omega)=(-1)^{m+n}\frac{2\pi e^2\hbar^2}{3m_e^2\omega\Omega}\)
\(\times\sum_{ij\alpha\mathbf{k}}W(\mathbf{k})\left(\frac{\epsilon_{i\mathbf{k}}+\epsilon_{j\mathbf{k}}}{2}-\mu\right)^{m+n-2} \times |\langle\Psi_{i\mathbf{k}}|\nabla_\alpha|\Psi_{j\mathbf{k}}\rangle|^2\)
\(\times[f(\epsilon_{i\mathbf{k}})-f(\epsilon_{j\mathbf{k}})]\delta(\epsilon_{j\mathbf{k}}-\epsilon_{i\mathbf{k}}-\hbar\omega).\)
They can also be computed by \(j\)-\(j\) correlation function.
\(L_{mn}=\frac{2e^{m+n-2}}{3\Omega\hbar\omega}\Im[\tilde{C}_{mn}(\omega)]\) Guassian smearing: \(\tilde{C}_{mn}=\int_0^\infty C_{mn}(t)e^{-i\omega t}e^{-\frac{1}{2}s^2t^2}dt\) Lorentzian smearing: \(\tilde{C}_{mn}=\int_0^\infty C_{mn}(t)e^{-i\omega t}e^{-\gamma t}dt\)
\(C_{mn}(t)=-2\theta(t)\Im\left\{Tr\left[\sqrt{\hat f}\hat{j}_m(1-\hat{f})e^{i\frac{\hat{H}}{\hbar}t}\hat{j}_ne^{-i\frac{\hat{H}}{\hbar}t}\sqrt{\hat f}\right]\right\}\),
where \(j_1\) is electric flux and \(j_2\) is thermal flux.
Frequency-dependent electric conductivities: \(\sigma(\omega)=L_{11}(\omega)\).
Frequency-dependent thermal conductivities: \(\kappa(\omega)=\frac{1}{e^2T}\left(L_{22}-\frac{L_{12}^2}{L_{11}}\right)\).
DC electric conductivities: \(\sigma = \lim_{\omega\to 0}\sigma(\omega)\).
Thermal conductivities: \(\kappa = \lim_{\omega\to 0}\kappa(\omega)\).
cal_cond
Type: Boolean
Availability: basis_type =
pw
Description: Whether to calculate electronic conductivities.
Default: False
cond_che_thr
Type: Real
Availability: esolver_type =
sdft
Description: Control the error of Chebyshev expansions for conductivities.
Default: 1e-8
cond_dw
Type: Real
Availability: basis_type =
pw
Description: Frequency interval (\(\mathrm{d}\omega\)) for frequency-dependent conductivities.
Default: 0.1
Unit: eV
cond_wcut
Type: Real
Availability: basis_type =
pw
Description: Cutoff frequency for frequency-dependent conductivities.
Default: 10.0
Unit: eV
cond_dt
Type: Real
Availability: basis_type =
pw
Description: Time interval (\(\mathrm{d}t\)) to integrate Onsager coefficients.
Default: 0.02
Unit: a.u.
cond_dtbatch
Type: Integer
Availability: esolver_type =
sdft
Description: exp(iH*dt*cond_dtbatch) is expanded with Chebyshev expansion to calculate conductivities. It is faster but costs more memory.
If
cond_dtbatch = 0
: Autoset this parameter to make expansion orders larger than 100.
Default: 0
cond_smear
Type: Integer
Description: Smearing method for conductivities
1: Gaussian smearing
2: Lorentzian smearing
Default: 1
cond_fwhm
Type: Real
Availability: basis_type =
pw
Description: FWHM for conductivities. For Gaussian smearing, \(\mathrm{FWHM}=2\sqrt{2\ln2}s\); for Lorentzian smearing, \(\mathrm{FWHM}=2\gamma\).
Default: 0.4
Unit: eV
cond_nonlocal
Type: Boolean
Availability: basis_type =
pw
Description: Whether to consider nonlocal potential correction when calculating velocity matrix \(\bra{\psi_i}\hat{v}\ket{\psi_j}\).
True: \(m\hat{v}=\hat{p}+\frac{im}{\hbar}[\hat{V}_{NL},\hat{r}]\).
False: \(m\hat{v}\approx\hat{p}\).
Default: True
Implicit solvation model
These variables are used to control the usage of implicit solvation model. This approach treats the solvent as a continuous medium instead of individual “explicit” solvent molecules, which means that the solute is embedded in an implicit solvent and the average over the solvent degrees of freedom becomes implicit in the properties of the solvent bath.
imp_sol
Type: Boolean
Description: calculate implicit solvation correction
Default: False
eb_k
Type: Real
Availability:
imp_sol
is true.Description: the relative permittivity of the bulk solvent, 80 for water
Default: 80
tau
Type: Real
Description: The effective surface tension parameter that describes the cavitation, the dispersion, and the repulsion interaction between the solute and the solvent which are not captured by the electrostatic terms
Default: 1.0798e-05
Unit: \(Ry/Bohr^{2}\)
sigma_k
Type: Real
Description: the width of the diffuse cavity that is implicitly determined by the electronic structure of the solute
Default: 0.6
nc_k
Type: Real
Description: the value of the electron density at which the dielectric cavity forms
Default: 0.00037
Unit: \(Bohr^{-3}\)
Quasiatomic Orbital (QO) analysis
These variables are used to control the usage of QO analysis. QO further compress information from LCAO: usually PW basis has dimension in million, LCAO basis has dimension below thousand, and QO basis has dimension below hundred.
qo_switch
Type: Boolean
Description: whether to let ABACUS output QO analysis required files
Default: 0
qo_basis
Type: String
Description: specify the type of atomic basis
pswfc
: use the pseudowavefunction in pseudopotential files as atomic basis. To use this option, please make sure in pseudopotential file there is pswfc in it.hydrogen
: generate hydrogen-like atomic basis (or with Slater screening).szv
: use the first set of zeta for each angular momentum from numerical atomic orbitals as atomic basis.
warning: to use
pswfc
, please use norm-conserving pseudopotentials with pseudowavefunctions, SG15 pseudopotentials cannot support this option. Developer notes: for ABACUS-lcao calculation, it is the most recommend to useszv
instead ofpswfc
which is originally put forward in work of QO implementation on PW basis. The information loss always happens ifpswfc
orhydrogen
orbitals are not well tuned, although making kpoints sampling more dense will mitigate this problem, but orbital-adjust parameters are needed to test system-by-system in this case.Default:
szv
qo_strategy
Type: String [String…](optional)
Description: specify the strategy to generate radial orbitals for each atom type. If one parameter is given, will apply to all atom types. If more than one parameters are given but fewer than number of atom type, those unspecified atom type will use default value.
For
qo_basis hydrogen
minimal-nodeless
: according to principle quantum number of the highest occupied state, generate only nodeless orbitals, for example Cu, only generate 1s, 2p, 3d and 4f orbitals (for Cu, 4s is occupied, thus \(n_{max} = 4\))minimal-valence
: according to principle quantum number of the highest occupied state, generate only orbitals with highest principle quantum number, for example Cu, only generate 4s, 4p, 4d and 4f orbitals.full
: similarly according to the maximal principle quantum number, generate all possible orbitals, therefore for Cu, for example, will generate 1s, 2s, 2p, 3s, 3p, 3d, 4s, 4p, 4d, 4f.energy-full
: will generate hydrogen-like orbitals according to Aufbau principle. For example the Cu (1s2 2s2 2p6 3s2 3p6 3d10 4s1), will generate these orbitals.energy-valence
: from the highest n (principal quantum number) layer and n-1 layer, generate all occupied and possible ls (angular momentum quantum number) for only once, for example Cu, will generate 4s, 3d and 3p orbitals.
For
qo_basis pswfc
andqo_basis szv
all
: use all possible pseudowavefunctions/numerical atomic orbital (of first zeta) in pseudopotential/numerical atomic orbital file.s
/p
/d
/…: only use s/p/d/f/…-orbital(s).spd
: use s, p and d orbital(s). Any unordered combination is acceptable.
warning: for
qo_basis hydrogen
to usefull
, generation strategy may cause the space spanned larger than the one spanned by numerical atomic orbitals, in this case, must filter out orbitals in some wayDefault: for
hydrogen
:energy-valence
, forpswfc
andszv
:all
qo_screening_coeff
Type: Real [Real…](optional)
Description: rescale the shape of radial orbitals, available for both
qo_basis hydrogen
andqo_basis pswfc
. cases but has different meaning.For
qo_basis pswfc
For each atom type, screening factor \(e^{-\eta|\mathbf{r}|}\) is multiplied to the pswfc to mimic the behavior of some kind of electron. \(\eta\) is the screening coefficient. If only one value is given, then will apply to each atom type. If not enough values are given, will apply default value to rest of atom types. This parameter plays important role in controlling the spread of QO orbitals together withqo_thr
.For
qo_basis hydrogen
If any float number is given, will apply Slater screening to all atom types. Slater screening is a classic and empirical method roughly taking many-electron effect into account for obtaining more accurate results when evaluating electron affinity and ionization energy. The Coulomb potential then becomes \(V(r) = -\frac{Z-\sigma}{r}\). For example the effective nuclear charge for Cu 3d electrons now reduces from 29 to 7.85, 4s from 29 to 3.70, which means Slater screening will bring about longer tailing effect. If no value is given, will not apply Slater screening.Default: 0.1
Unit: Bohr^-1
qo_thr
Type: Real
Description: the convergence threshold determining the cutoff of generated orbital. Lower threshold will yield orbital with larger cutoff radius.
Default: 1.0e-6
PEXSI
These variables are used to control the usage of PEXSI (Pole Expansion and Selected Inversion) method in calculations.
pexsi_npole
Type: Integer
Description: the number of poles used in the pole expansion method, should be a even number.
Default: 40
pexsi_inertia
Type: Boolean
Description: whether inertia counting is used at the very beginning.
Default: True
pexsi_nmax
Type: Integer
Description: maximum number of PEXSI iterations after each inertia counting procedure.
Default: 80
pexsi_comm
Type: Boolean
Description: whether to construct PSelInv communication pattern.
Default: True
pexsi_storage
Type: Boolean
Description: whether to use symmetric storage space used by the Selected Inversion algorithm for symmetric matrices.
Default: True
pexsi_ordering
Type: Integer
Description: ordering strategy for factorization and selected inversion. 0: Parallel ordering using ParMETIS, 1: Sequential ordering using METIS, 2: Multiple minimum degree ordering
Default: 0
pexsi_row_ordering
Type: Integer
Description: row permutation strategy for factorization and selected inversion, 0: No row permutation, 1: Make the diagonal entry of the matrix larger than the off-diagonal entries.
Default: 1
pexsi_nproc
Type: Integer
Description: number of processors for PARMETIS. Only used if pexsi_ordering == 0.
Default: 1
pexsi_symm
Type: Boolean
Description: whether the matrix is symmetric.
Default: True
pexsi_trans
Type: Boolean
Description: whether to factorize the transpose of the matrix.
Default: False
pexsi_method
Type: Integer
Description: the pole expansion method to be used. 1 for Cauchy Contour Integral method, 2 for Moussa optimized method.
Default: 1
pexsi_nproc_pole
Type: Integer
Description: the point parallelizaion of PEXSI. Recommend two points parallelization.
Default: 1
pexsi_temp
Type: Real
Description: temperature in Fermi-Dirac distribution, in Ry, should have the same effect as the smearing sigma when smearing method is set to Fermi-Dirac.
Default: 0.015
pexsi_gap
Type: Real
Description: spectral gap, this can be set to be 0 in most cases.
Default: 0
pexsi_delta_e
Type: Real
Description: an upper bound for the spectral radius of \(S^{-1} H\).
Default: 20
pexsi_mu_lower
Type: Real
Description: initial guess of lower bound for mu.
Default: -10
pexsi_mu_upper
Type: Real
Description: initial guess of upper bound for mu.
Default: 10
pexsi_mu
Type: Real
Description: initial guess for mu (for the solver).
Default: 0
pexsi_mu_thr
Type: Real
Description: stopping criterion in terms of the chemical potential for the inertia counting procedure.
Default: 0.05
pexsi_mu_expand
Type: Real
Description: if the chemical potential is not in the initial interval, the interval is expanded by this value.
Default: 0.3
pexsi_mu_guard
Type: Real
Description: safe guard criterion in terms of the chemical potential to reinvoke the inertia counting procedure.
Default: 0.2
pexsi_elec_thr
Type: Real
Description: stopping criterion of the PEXSI iteration in terms of the number of electrons compared to numElectronExact.
Default: 0.001
pexsi_zero_thr
Type: Real
Description: if the absolute value of CCS matrix element is less than this value, it will be considered as zero.
Default: 1e-10
Linear Response TDDFT
These parameters are used to solve the excited states using. e.g. LR-TDDFT.
xc_kernel
Type: String
Description: The exchange-correlation kernel used in the calculation. Currently supported:
RPA
,LDA
,PBE
,HSE
,HF
.Default: LDA
lr_solver
Type: String
Description: The method to solve the Casida equation \(AX=\Omega X\) in LR-TDDFT under Tamm-Dancoff approximation (TDA), where \(A_{ai,bj}=(\epsilon_a-\epsilon_i)\delta_{ij}\delta_{ab}+(ai|f_{Hxc}|bj)+\alpha_{EX}(ab|ij)\) is the particle-hole excitation matrix and \(X\) is the transition amplitude.
dav
: Construct \(AX\) and diagonalize the Hamiltonian matrix iteratively with Davidson algorithm.lapack
: Construct the full \(A\) matrix and directly diagonalize with LAPACK.spectrum
: Calculate absorption spectrum only without solving Casida equation. TheOUT.${suffix}/
directory should contain the files for LR-TDDFT eigenstates and eigenvalues, i.e.Excitation_Energy.dat
andExcitation_Amplitude_${processor_rank}.dat
output by settingout_wfc_lr
to true.
Default: dav
lr_thr
Type: Real
Description: The convergence threshold of iterative diagonalization solver fo LR-TDDFT. It is a pure-math number with the same as pw_diag_thr, but since the Casida equation is a one-shot eigenvalue problem, it is also the convergence threshold of LR-TDDFT.
Default: 1e-2
nocc
nvirt
Type: Integer
Description: The number of virtual orbitals (staring from LUMO) used in the LR-TDDFT calculation.
Default: 1
lr_nstates
Type: Integer
Description: The number of 2-particle states to be solved
Default: 0
abs_wavelen_range
Type: Real Real
Description: The range of the wavelength for the absorption spectrum calculation.
Default: 0.0 0.0
out_wfc_lr
Type: Boolean
Description: Whether to output the eigenstates (excitation energy) and eigenvectors (excitation amplitude) of the LR-TDDFT calculation. The output files are
OUT.${suffix}/Excitation_Energy.dat
andOUT.${suffix}/Excitation_Amplitude_${processor_rank}.dat
.Default: False
abs_broadening
Type: Real
Description: The broadening factor \(\eta\) for the absorption spectrum calculation.
Default: 0.01