TABLE OF CONTENTS
INTRODUCTION
&ENVIRON
atomicspread | cion | cionmax | electrolyte_entropy | electrolyte_linearized | env_confine | env_dielectric_regions | env_electrolyte_ntyp | env_electrostatic | env_external_charges | env_nrep | env_optical_permittivity | env_pressure | env_static_permittivity | env_surface_tension | environ_nskip | environ_restart | environ_thr | environ_type | rion | sc_carrier_density | sc_permittivity | system_axis | system_dim | system_ntyp | temperature | verbose | zion
&BOUNDARY
alpha | corespread | derivatives | electrolyte_alpha | electrolyte_distance | electrolyte_mode | electrolyte_rhomax | electrolyte_rhomin | electrolyte_softness | electrolyte_spread | electrolyte_tbeta | filling_spread | filling_threshold | ifdtype | nfdpoint | radial_scale | radial_spread | radius_mode | rhomax | rhomin | sc_distance | sc_spread | softness | solvationrad | solvent_distance | solvent_mode | solvent_radius | solvent_spread | stype | tbeta
&ELECTROSTATIC
auxiliary | core | inner_maxstep | inner_mix | inner_solver | inner_tol | maxstep | mix | mix_type | ndiis | pbc_axis | pbc_correction | pbc_dim | preconditioner | problem | screening | screening_type | solver | step | step_type | tol
EXTERNAL_CHARGES
Q | x | y | z | spread | dim | axis
DIELECTRIC_REGIONS
EpsSt | EpsOpt | x | y | z | width | spread | dim | axis
INTRODUCTION
Input data format: { } = optional, [ ] = it depends, | = or
All quantities whose dimensions are not explicitly specified are in
RYDBERG ATOMIC UNITS. Charge is "number" charge (i.e. not multiplied
by e); potentials are in energy units (i.e. they are multiplied by e).
BEWARE: TABS, DOS <CR><LF> CHARACTERS ARE POTENTIAL SOURCES OF TROUBLE
Comment lines in namelists can be introduced by a "!", exactly as in fortran
code. Comments lines in cards can be introduced by either a "!" or a "#"
character in the first position of a line.
Do not start any line in cards with a "/" character.
Structure of the environ.in input file:
===============================================================================
&ENVIRON
...
/
[ &BOUNDARY
...
/ ]
[ &ELECTROSTATIC
...
/ ]
[ EXTERNAL_CHARGES { bohr | angstrom }
Q1 0.0 0.0 0.0 { spread_Q1 dim_Q1 axis_Q1 }
Q2 0.5 0.0 0.0 { spread_Q2 dim_Q2 axis_Q2 } ]
[ DIELECTRIC_REGIONS { bohr | angstrom }
EPSst_E1 EPSopt_E1 0.0 0.0 0.0 width_E1 { spread_E1 dim_E1 axis_E1 }
EPSst_E2 EPSopt_E2 0.0 0.0 0.0 width_E2 { spread_E2 dim_E2 axis_E2 } ]
Namelist: &ENVIRON
|
This namelist is always needed !
atomicspread(i), i=1,ntyp |
REAL |
Default: |
-0.5D0
|
Status: |
OPTIONAL
|
In the calculation of electrostatic contributions, ionic charge densities are modelled
as gaussians of fixed spread, as specified by atomicspread(ityp) for each atomic type.
Results are identical to using point-like charges (as is usually done in PW), unless the gaussian
spreads are too large. The default value of 0.5 a.u. was derived to be safe enough in most
common atom types. Too small of a value may require larger density cutoffs (ecutrho).
|
cion(i), i=1,env_electrolyte_ntyp |
REAL |
Default: |
1.D0
|
Status: |
OPTIONAL
|
Molar concentration of ionic countercharge (in mol/L).
|
cionmax |
REAL |
Default: |
1.D3
|
Status: |
OPTIONAL
|
Maximum molar concentration of ionic countercharge (in mol/L).
|
electrolyte_entropy |
CHARACTER |
Default: |
'full'
|
Status: |
OPTIONAL
|
Keyword to set the electrolyte entropy terms that are affected by the Stern-layer correction.
'ions' : only ionic terms (Ringe et al. J. Chem. Theory Comput. 12, 4052).
'full' : all terms (Dabo et al. arXiv 0901.0096).
|
electrolyte_linearized |
LOGICAL |
Default: |
.FALSE.
|
Status: |
OPTIONAL
|
Controls if the linear-regime approximation is employed for the electrolyte model.
It can be used in combination with the numerical Poisson-Boltzmann (PB) model as well as with the
Gouy-Chapman-Stern correction that exploits the 1D-solution of the PB equation (see pbc_correction).
|
env_confine |
REAL |
Default: |
0.D0
|
Status: |
REQUIRED
|
Confinement potential barrier in Rydberg atomic units.
This keyword controls the activation of the confinement potential contribution
to the solute's Hamiltonian. A value of 0.D0 means no confinement contribution
from the environment.
The confinement potential may be used to favour electron localization: a potential that
smoothly switches between zero and env_confine, following the complementary of
the interface function, provides a destabilizing contribution that penalizes
delocalized states.
|
env_dielectric_regions |
INTEGER |
Default: |
0
|
Status: |
REQUIRED
|
Number of fixed dielectric regions. This keyword controls how many regions of fixed
permittivities (static and optical) need to be included in the simulation box. Shape,
position, and permittivities of each region need to be specified in the DIELECTRIC_REGIONS card.
|
env_electrolyte_ntyp |
INTEGER |
Default: |
0
|
Status: |
REQUIRED
|
Number of countercharge species to be added in the continuum region modelling
the electrolyte. If different from zero, must be greater or equal to 2 in order
for the electrolyte to be charge neutral.
|
env_electrostatic |
LOGICAL |
Default: |
.FALSE.
|
Status: |
REQUIRED
|
Generic keyword that flags the need to read the &ELECTROSTATIC namelist. Any electrostatic
embedding keyword (env_static_permittivity, env_electrolyte_ntyp) will turn this keyword on.
One needs to turn on this keyword explicitly to activate electrostatic embedding effects
that do not have a specific activation keyword, such as for PBC correction schemes.
|
env_external_charges |
INTEGER |
Default: |
0
|
Status: |
REQUIRED
|
Number of fixed external charges. This keyword controls how many fixed external densities
of charge need to be included in the simulation box. Shape, position, and amount of charge
of each external density need to be specified in the EXTERNAL_CHARGES card.
|
env_nrep(i), i=1,3 |
INTEGER |
Default: |
0
|
Status: |
OPTIONAL
|
Number of additional replicas of the system cell on each side along the three axis.
nrep = 1 means there is one more cell on the left and on the right of the cell.
The environment cell is (2*nrep+1) times the system cell along the three axis.
|
env_optical_permittivity |
REAL |
Default: |
1.D0
|
Status: |
REQUIRED
|
Optical dielectric permittivity of the electrostatic continuum embedding model,
only needed for TDDFPT calculations. A value of 1 (=vacuum) means no dielectric
effects in linear response calculations.
|
env_pressure |
REAL |
Default: |
see below
|
Status: |
REQUIRED
|
External pressure (P) of the environment in GPa.
This keyword controls the activation of the volume-dependent contribution to
the solute's Hamiltonian (P*V). A value of 0.D0 means no volume contribution
from the environment.
This contribution may be straightforwardly used to compute the electronic enthalpy,
i.e. to model finite systems under pressure, as proposed by M. Cococcioni et al. in
Phys. Rev. Lett. 94, 145501 (2005). This contribution can also be used as a simplified
approach to more complex and general non-electrostatic contributions to solvation,
as in the SCCS approach. In this second case, env_pressure needs not correspond to the
real external pressure of the environment, but is used as a fitting parameter
(and can assume negative values).
See O. Andreussi, I. Dabo and N. Marzari, J. Chem. Phys. 136 064102 (2012)
DEFAULT:
solvent_mode .EQ. electronic | full :
environ_type .EQ. input | vacuum : 0.D0
environ_type .EQ. water : -0.35D0
environ_type .EQ. water-cation : 0.125D0
environ_type .EQ. water-anion : 0.450D0
solvent_mode .EQ. ionic : -0.35D0
|
env_static_permittivity |
REAL |
Default: |
see below
|
Status: |
REQUIRED
|
Static dielectric permittivity of the electrostatic continuum embedding model.
This keyword (like all the env_* keywords) is also the flag which controls the
activation of the specific contribution. A value of 1 (=vacuum) means no dielectric
effects from the environment.
DEFAULT:
environ_type .EQ. water | water-cation | water-anion : 78.3D0
environ_type .EQ. input | vacuum : 1.D0
|
env_surface_tension |
REAL |
Default: |
see below
|
Status: |
REQUIRED
|
Surface tension (gamma) of the environment in CGS units dyn/cm.
This keyword controls the activation of the surface-dependent contribution to
the solute's Hamiltonian (gamma*S). A value of 0.D0 means no surface
contribution from the environment.
This contribution may be straightforwardly used to compute cavitation free energies,
as proposed by Scherlis et al. in J. Chem. Phys. 124, 074103 (2006).
NOTE that the current implementation uses an improved definition of the quantum-surface.
This contribution can also be used as a simplified approach to the more general
non-electrostatic contributions to solvation, as in the SCCS approach.
In this second case, env_surface_tension does not need to correspond to the
real surface tension of the solvent, but is used as a fitting parameter.
See O. Andreussi, I. Dabo and N. Marzari, J. Chem. Phys. 136 064102 (2012)
DEFAULT:
solvent_mode .EQ. electronic | full :
environ_type .EQ. input | vacuum | water-anion : 0.D0
environ_type .EQ. water : 50.D0
environ_type .EQ. water-cation : 5.D0
solvent_mode .EQ. ionic : 50.D0
|
environ_nskip |
INTEGER |
Default: |
1
|
Only include/update environment contributions after the first environ_nskip steps of CP dynamics.
|
environ_restart |
LOGICAL |
Default: |
.FALSE.
|
Status: |
OPTIONAL
|
Compute environ contributions during the initialization step, useful for 'restart' calculations
and systems with a good-enough initial guess.
Must be set to activate environ in nscf calculations.
|
environ_thr |
REAL |
Default: |
1.D-1
|
Status: |
IMPORTANT - the default value is only valid for small systems - too conservative for larger systems
|
Only include/update environment contributions when SCF accuracy is below this threshold.
Since the environment region is defined in terms of the electronic density, the test is done
in order to avoid computing unphysical environment contributions, usually to skip the
environ calculation during the first couple of SCF step.
IMPORTANT: as the SCF accuracy is an extensive property (increases with the number of electrons in
the system), the optimal environ_thr will also vary with system size.
|
environ_type |
CHARACTER |
Default: |
'input'
|
Status: |
OPTIONAL
|
Set up all of the environment embedding flags and interface
parameters according to predefined types:
'input':
Do not use predefined types, read the flags of the different contribution from input
or use the defaults values (which correspond to no contributions).
'vacuum':
All environment embeddings are turned off, but pbc corrections, external charges and
electrolyte charges can be present.
'water':
Set up the main physical constants to the experimental values for water at room temperature
(env_static_permittivity = 78.3). Set the tunable embedding flags to the values optimized
to reproduce aqueous solvation of small neutral organic compounds, corresponding to
non-electrostatic contributions modelled via pressure (env_pressure = -0.35 GPa)
and surface tension (env_surface_tension = 50 dyn/cm) effects (see references below).
For electrons-dependent interfaces (self-consistent continuum solvation, SCCS), set soft
interface parameters to the optimal values (rhomax = 0.005; rhomin = 0.0001)
derived in O. Andreussi, I. Dabo and N. Marzari, J. Chem. Phys. 136 064102 (2012).
For ions-dependent interfaces (soft-sphere continuum solvation, SSCS), set rigid
interface parameter to the optimal value (alpha = 1.12) derived in
G. Fisicaro et al. J. Chem. Theory Comput. 13, 8, 3829 (2017).
'water-cation':
Set up the main physical constants to the experimental values for water at room temperature
(env_static_permittivity = 78.3). Set the tunable embedding flags to the values optimized
to reproduce aqueous solvation of small organic cations.
SCCS parameters (env_pressure = 0.125 GPa; env_surface_tension = 5.0 dyn/cm;
rhomax = 0.0035; rhomin = 0.0002) were derived in C. Dupont, O. Andreussi and
N. Marzari, J. Chem. Phys. 139, 214110 (2013)
SSCS parameters (env_pressure = -0.35 GPa; env_surface_tension = 50.0 dyn/cm;
alpha = 1.10) were derived in G. Fisicaro et al. J. Chem. Theory Comput.
13, 8, 3829 (2017).
'water-anion':
Set up the main physical constants to the experimental values for water at room temperature
(env_static_permittivity = 78.3). Set the tunable embedding flags to the values optimized
to reproduce aqueous solvation of small organic cations.
SCCS parameters (env_pressure = 0.45 GPa; env_surface_tension = 0.0 dyn/cm;
rhomax = 0.0155; rhomin = 0.0024) were derived in C. Dupont, O. Andreussi and
N. Marzari, J. Chem. Phys. 139, 214110 (2013)
SSCS parameters (env_pressure = -0.35 GPa; env_surface_tension = 50.0 dyn/cm;
alpha = 0.98) were derived in G. Fisicaro et al. J. Chem. Theory Comput.
13, 8, 3829 (2017).
|
rion |
REAL |
Default: |
0.D0
|
Status: |
OPTIONAL
|
Mean atomic radius of ionic countercharge (a.u.).
|
sc_carrier_density |
REAL |
Default: |
0.D0
|
Status: |
OPTIONAL
|
Concentration of charge carriers within the semiconductor (cm^-3).
|
sc_permittivity |
REAL |
Default: |
1.D0
|
Status: |
OPTIONAL
|
Dielectric permittivity of the semiconductor.
|
system_axis |
INTEGER |
Default: |
3
|
Main axis of the embedded system, only necessary for partially periodic systems
(1 = x, 2 = y, 3 = z).
|
system_dim |
INTEGER |
Default: |
0
|
Dimensionality of the embedded system, used to determine size (only orthogonally periodic
dimensions) and position (0 = 0D, 1 = 1D, 2 = 2D).
|
system_ntyp |
INTEGER |
Default: |
0
|
Specify the atom types that are used to determine the origin and size of the embedded system,
for system-dependent properties in Environ (e.g. see later system-dependent boundary).
Atom types up to system_ntyp are used, all atoms are used by default or if system_ntyp == 0.
|
temperature |
REAL |
Default: |
300.D0
|
Status: |
OPTIONAL
|
Temperature of electrolyte solution, needed by Poisson-Boltzmann equation. Also used in
semiconductor simulations.
|
verbose |
INTEGER |
Default: |
0
|
Status: |
OPTIONAL
|
Control the amount of output written to specific output files, mostly useful for debugging purposes
verbose .EQ. 0 : minimal information written to standard output
verbose .EQ. 1 : additional information written to environ.debug file
verbose .EQ. 2 : dumping of main physical quantities on the real-space grid in the form of
*.cube files (only at the end of each SCF cycle)
verbose .GE. 3 : dumping of several intermediate physical quantities on the real-space grid
as this is done at every SCF step it will slow down the calculation significantly
|
zion(i), i=1,env_electrolyte_ntyp |
REAL |
Default: |
1.D0
|
Status: |
OPTIONAL
|
Valence of ionic countercharge.
|
|
|
Namelist: &BOUNDARY
|
This namelist is not needed if there are no embedding schemes requiring a continuum interface
alpha |
REAL |
Default: |
see below
|
Status: |
OPTIONAL
|
Main parameter of ions-dependent interface function, corresponding to the homogeneous
scaling factor of ionic radii (specified by the radius_mode keyword or by solvationrad
(ntyp)).
DEFAULT:
solvent_mode .EQ. electronic | full :
environ_type .EQ. input | vacuum : 1.D0
environ_type .EQ. water : 1.12D0
environ_type .EQ. water-cation : 1.10D0
environ_type .EQ. water-anion : 0.98D0
solvent_mode .EQ. ionic : 1.D0
|
corespread(i), i=1,ntyp |
REAL |
Default: |
-0.5D0
|
Status: |
OPTIONAL
|
Numerical spread of the atomic-centered Gaussian functions used to model core electrons and
ions for interfaces defined on the full set of degrees of freedom of the QM system
(interface mode = full).
NOTE: this numerical parameter should only be used to avoid artefacts coming from the missing
core electrons, if this parameter affects the final results, a WARNING will be issued and a smaller
(even null) value should be used instead.
|
derivatives |
CHARACTER |
Default: |
see below
|
Status: |
OPTIONAL
|
Specify the numerical approach used to compute derivatives of the continuum interface with respect
to real-space coordinates or QM degrees of freedom.
'fft' : fast Fourier transform
'chain' : chain rule derivatives for as much as possible (and FFTs for the rest)
'highmem' : analytic derivatives for soft-sphere computed by storing all spherical functions
and derivatives
'lowmem' : more efficient analytic derivatives
DEFAULT:
solvent_mode .EQ. electronic | full : 'chain'
solvent_mode .EQ. ionic : 'lowmem'
|
electrolyte_alpha |
REAL |
Default: |
1.D0
|
Status: |
OPTIONAL
|
Main parameter of ions-dependent interface function, corresponding to the homogeneous
scaling factor of ionic radii (specified by the radius_mode keyword or by solvationrad(ntyp)).
Used for the electrolyte interface.
|
electrolyte_distance |
REAL |
Default: |
0.D0
|
Status: |
OPTIONAL
|
Distance of the system-dependent interface from the center of the system, computed as the
center of ionic charge of the atomic types entering the definition of the system
(as specified in the system_ntyp keyword of the &ENVIRON namelist).
Used for the electrolyte interface.
|
electrolyte_mode |
CHARACTER |
Default: |
'electronic'
|
Status: |
REQUIRED
|
Choice of the interface function representing the boundary between the QM region and the
electrolyte embedding environment. Please note that a separate set of keywords is used to
specify the parameters of the electrolyte interface function (all exploiting the
electrolyte_ prefix).
'electronic':
interface depends self-consist. on electronic density, corresponds
to the Self-consistent Continuum Solvation approach (SCCS).
'ionic':
interface defined on atomic positions, corresponds to the
Soft-sphere Continuum Solvation approach (SSCS).
'full':
similar to electronic, but an extra density is added to
represent the core electrons and the nuclei. This extra
density is defined as the sum of gaussian functions centered
on atomic positions of width specified by the corespread(ityp) keyword.
'system':
interface defined as a simple analytical function of system position and
dimensionality.
|
electrolyte_rhomax |
REAL |
Default: |
0.005D0
|
Status: |
OPTIONAL
|
First parameter of the density-dependent interface function, roughly corresponding to the
density threshold of the continuum model. Used for the electrolyte interface.
|
electrolyte_rhomin |
REAL |
Default: |
0.0001D0
|
Status: |
OPTIONAL
|
Second parameter of the density-dependent interface function, when stype=1. Used for the
electrolyte interface.
|
electrolyte_softness |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
Numerical spread of the soft-sphere functions used for the ions-dependent interfaces.
Used for the electrolyte interface.
|
electrolyte_spread |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
Numerical spread of the analytical function used for the system-dependent interface.
Used for the electrolyte interface.
|
electrolyte_tbeta |
REAL |
Default: |
4.8D0
|
Status: |
OPTIONAL
|
Second parameter of the density-dependent interface function, when stype=0. Used for the
electrolyte interface.
|
filling_spread |
REAL |
Default: |
0.02D0
|
Status: |
OPTIONAL
|
For solvent-aware interfaces, numerical spread of the switching function used to decide whether
the continuum void should be filled or not.
|
filling_threshold |
REAL |
Default: |
0.825D0
|
Status: |
OPTIONAL
|
For solvent-aware interfaces, threshold of filled-fraction used to decide whether to fill or not
a continuum void. The default value of 7/8 correspond to the geometrical condition required to
homogenously fill a spherical void of radius smaller than the solvent radius.
|
ifdtype |
INTEGER |
Default: |
1
|
Status: |
OPTIONAL
|
The gradient of the dielectric function is computed in real-space using finite differences.
Different finite differences schemes have been implemented following:
P. Holoborodko, Smooth noise robust differentiators, 2008
http://www.holoborodko.com/pavel/numerical-methods/numerical-derivative/smooth-low-noise-differentiators
Each scheme can exploit different numbers of points of the real-space grid (as defined by nfdpoint).
ifdtype .EQ. 1 : Central differences
ifdtype .EQ. 2 : Low-noise Lanczos (m=2)
ifdtype .EQ. 3 : Low-noise Lanczos (m=4)
ifdtype .EQ. 4 : Smooth noise-robust (n=2)
ifdtype .EQ. 5 : Smooth noise-robust (n=4)
Central differences are used by default and have been tested more deeply. The other schemes
work fine, but are not deeply tested in terms of performances.
|
nfdpoint |
INTEGER |
Default: |
2
|
Status: |
IMPORTANT
|
Number of point from the real-space grid, to be used by the different finite-difference schemes
to compute gradients.
Number of points = 2 * nfdpoint + 1
e.g. ifdtype.EQ.1 .AND. nfdpoint.EQ.1 correspond to central differences with three points
IMPORTANT: nfdpoint .EQ. 1 seems to be enough for most applications, but more refined
finite-difference schemes are needed (nfdpoint.EQ.2 is enough) for energy conservation
in MD simulations in continuum dielectric.
See test case reported in O. Andreussi, I. Dabo and N. Marzari, J. Chem. Phys. 136 064102 (2012)
|
radial_scale |
REAL |
Default: |
2.D0
|
Status: |
OPTIONAL
|
For solvent-aware interfaces, compute the filled fraction on a spherical volume whose radius is
radial_scale * solvent_radius. The default value of 2.0 guarantees that spherical voids of the size
of a solvent molecule will be homogenously filled. Larger values will let the algorithm sample a
larger volume of space, increasing the non-locality of the algorithm.
|
radial_spread |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
For solvent-aware interfaces, numerical spread of the step function used to compute the filled fraction.
|
radius_mode |
CHARACTER |
Default: |
'uff'
|
Status: |
OPTIONAL
|
Specify the predefined set of atomic radii to be used when building an ions-dependent interface.
'pauling' : R.C. Weast, ed., Handbook of chemistry and physics (CRC Press, Cleveland, 1981)
'bondi' : A. Bondi, J. Phys. Chem. 68, 441 (1964)
'uff' : A.K. Rapp/'{e} et al. J. Am. Chem. Soc. 114(25) pp.10024-10035 (1992)
'muff' : Fisicaro JCTC (2017)
|
rhomax |
REAL |
Default: |
see below
|
Status: |
OPTIONAL
|
First parameter of the density-dependent interface function, roughly corresponding to
the density threshold of the continuum model.
DEFAULT:
solvent_mode .EQ. electronic | full :
environ_type .EQ. input | vacuum | water : 0.005D0
environ_type .EQ. water-cation : 0.0035D0
environ_type .EQ. water-anion : 0.0155D0
solvent_mode .EQ. ionic : 0.005D0
|
rhomin |
REAL |
Default: |
see below
|
Status: |
OPTIONAL
|
Second parameter of the density-dependent interface function, when stype=1.
DEFAULT:
solvent_mode .EQ. electronic | full :
environ_type .EQ. input | vacuum | water : 0.0001D0
environ_type .EQ. water-cation : 0.0002D0
environ_type .EQ. water-anion : 0.0024D0
solvent_mode .EQ. ionic : 0.0001D0
|
sc_distance |
REAL |
Default: |
0.D0
|
Status: |
OPTIONAL
|
Distance from the system where the Mott-Schottky boundary starts.
|
sc_spread |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
Spread of the interfaces for the Mott-Schottky boundary.
|
softness |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
Numerical spread of the soft-sphere functions used for the ions-dependent interfaces.
|
solvationrad(i), i=1,ntyp |
REAL |
Default: |
-3.D0
|
Status: |
OPTIONAL
|
Atomic radii used for the ions-dependent interface as introduced by G. Fisicaro et al.
J. Chem. Theory Comput. 13, 8, 3829 (2017). These values will overwrite the defaults
set by the radius_mode keyword.
|
solvent_distance |
REAL |
Default: |
1.D0
|
Status: |
OPTIONAL
|
Distance of the system-dependent interface from the center of the system, computed as the
center of ionic charge of the atomic types entering the definition of the system
(as specified in the system_ntyp keyword of the &ENVIRON namelist).
|
solvent_mode |
CHARACTER |
Default: |
'electronic'
|
Status: |
REQUIRED
|
Choice of the interface function representing the boundary between the
QM region and the solvent embedding environment. Dielectric, surface and
volume embedding will all act on the solvent interface (as opposed to
the electrolyte embedding that will act on the electrolyte interface).
'electronic':
interface depends self-consist. on electronic density, corresponds
to the Self-consistent Continuum Solvation approach (SCCS).
'ionic':
interface defined on atomic positions, corresponds to the
Soft-sphere Continuum Solvation approach (SSCS).
'full':
similar to electronic, but an extra density is added to
represent the core electrons and the nuclei. This extra
density is defined as the sum of gaussian functions centered
on atomic positions of width specified by the corespread(ityp) keyword.
'system':
inteface defined as a simple analytical function of system position and
dimensionality.
|
solvent_radius |
REAL |
Default: |
0.D0
|
Status: |
OPTIONAL
|
Size of the solvent molecules, used in the solvent-aware algorithm to decide whether to fill
a continuum void or not. If set equal to 0.D0, use the standard local algorithm.
|
solvent_spread |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
Numerical spread of the analytical function used for the system-dependent interface.
|
stype |
INTEGER |
Default: |
2
|
Status: |
OPTIONAL
|
The shape of the environment region is defined according to a specific switching function
of the electronic density:
stype .EQ. 0 : Original switching function from Fattebert and Gygi.
Requires two parameters: rhomax and tbeta
stype .EQ. 1 : Optimally smooth switching function from the SCCS method, redefined for the
non-electrostatic part.
Requires two parameters: rhomax and rhomin
stype .EQ. 2 : Optimally smooth switching function from the SCCS method of Andreussi et al.
Requires two parameters: rhomax and rhomin
|
tbeta |
REAL |
Default: |
4.8D0
|
Status: |
OPTIONAL
|
Second parameter of the density-dependent interface function, when stype=0.
|
|
|
Namelist: &ELECTROSTATIC
|
This namelist is not needed if there are no electrostatic embedding effects
auxiliary |
CHARACTER |
Default: |
see below
|
Status: |
OPTIONAL
|
The electrostatic problem is defined and solved in terms of the electrostatic potential or of an auxiliary quantity.
'none' : solve for the electrostatic potential.
'full' : solve for the auxiliary charge (e.g. for 'generalized' problems solve for
the polarization charge, used to be the default in Environ v0.2).
'ioncc': solve for the electrolyte charge.
DEFAULT:
IF (solver .EQ. 'fixed-point') THEN
auxiliary = 'full'
ELSE
auxiliary = 'none'
|
core |
CHARACTER |
Default: |
'fft'
|
Status: |
OPTIONAL
|
The choice of the core numerical methods to be exploited for the different operations.
'fft' : fast Fourier transforms
|
inner_maxstep |
INTEGER |
Default: |
200
|
Status: |
OPTIONAL
|
Same as maxstep, but refers to the inner loop in the electrostatic calculation
(nested schemes only).
|
inner_mix |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
Same as mix, but refers to the inner loop in the electrostatic calculation
(nested schemes only).
|
inner_solver |
CHARACTER |
Default: |
see below
|
Status: |
OPTIONAL
|
Same as solver, but refers to the algorithm used to get the solution of the inner loop
in the electrostatic problem (nested schemes only).
'cg' :
'sd' :
'fixed-point' :
'direct' :
DEFAULT:
IF (env_electrolyte_ntyp .GT. 0 .AND. pbc_correction .NE. 'gcs') THEN
IF (.NOT. electrolyte_linearized) THEN
inner_solver = 'cg'
ELSE
inner_solver = 'none'
|
inner_tol |
REAL |
Default: |
1.D-10
|
Status: |
IMPORTANT
|
Accuracy of the inner loop in the electrostatic calculation (nested schemes only).
As for the tol parameter, higher accuracies will require more cycles, but will ensure
smoother convergence of the outer electrostatic loop (thresholds smaller than 1.D-15
are recommended).
|
maxstep |
INTEGER |
Default: |
200
|
Status: |
OPTIONAL
|
Maximum number of steps to get the solution of the electrostatic problem.
|
mix |
REAL |
Default: |
0.5D0
|
Status: |
OPTIONAL
|
Linear mixing parameter for iterative solver. Usually does not affect results
(and it shouldn't) and does not affect performances, large values work fine in
most common applications.
|
mix_type |
CHARACTER |
Default: |
'linear'
|
Status: |
OPTIONAL
|
The mixing method for iterative calculations.
'linear' :
'anderson' :
'diis' :
'broyden' :
|
ndiis |
INTEGER |
Default: |
1
|
Status: |
OPTIONAL
|
The order of DIIS interpolation of iterative calculation.
|
pbc_axis |
INTEGER |
Default: |
3
|
Status: |
OPTIONAL
|
For partially periodic simulation cells (1D or 2D), choice of the sides with periodic
boundary conditions (1 = x, 2 = y, 3 = z)
pbc_dim .EQ. 1 : specified axis is along the 1D direction.
pbc_dim .EQ. 2 : specified axis is orthogonal to 2D plane.
|
pbc_correction |
CHARACTER |
Default: |
'none'
|
Status: |
OPTIONAL
|
Type of pbc correction scheme to be used.
'none' : no correction
'parabolic' : parabolic point-counter-charge (PCC) correction, only implemented for 0D
and 2D systems
'gcs' : Gouy-Chapman-Stern correction, only implemented for 2D systems. It
exploits the one-dimensional analytic solution of the Poisson-Boltzmann
equation. The analytic potential replaces the numerical one where the
distance from the system is larger than electrolyte_distance (all charge
distributions, including the polarization charge, must be fully included
within this distance)
'ms' : Mott-Schottky correction, only implemented for 2D systems, assumes the
system is embedded in a semiconductor with sc_permittivity and
sc_carrier_density. It exploits the one-dimensional analytic solution of
the Mott-Schottky equation. The analytic potential replaces the
numerical one where the distance from the system is larger than sc_distance
(all charge distributions, including the polarization charge, must be fully
included within this distance)
|
pbc_dim |
INTEGER |
Default: |
-3
|
Status: |
OPTIONAL
|
Dimensionality of the simulation cell, i.e. periodic boundary conditions are applied
on 3/2/1/0 sides of the cell.
|
preconditioner |
CHARACTER |
Default: |
'sqrt'
|
Status: |
OPTIONAL
|
The type of preconditioner.
'none' : no preconditioner
'left' : left linear preconditioner eps nabla v = r
'sqrt' : sqrt preconditioner sqrt(eps) nabla (sqrt(eps) * v) = r
|
problem |
CHARACTER |
Default: |
see below
|
Status: |
OPTIONAL
|
Type of electrostatic problem that need to be solved. This keyword is usually set
automatically by the ENVIRON namelist, but it may be used to specify a linearized
or modified version of an electrostatic problem.
'poisson' : standard Poisson equation, with or without periodic boundary conditions
'generalized' : generalized Poisson equation, a dielectric embedding must be present
'pb' : non-linear Poisson-Boltzmann equation, an electrolyte embedding must
be present, a dielectric embedding is optional
'modpb' : non-linear size-modified Poisson-Boltzmann equation, an electrolyte
embedding must be present, a dielectric embedding is optional
'linpb' : linearized version of the Poisson-Boltzmann problem
'linmodpb' : linearized version of the modified Poisson-Boltzmann problem
DEFAULT:
IF (env_electrolyte_ntyp .GT. 0 .AND. pbc_correction .NE. 'gcs') THEN
IF (electrolyte_linearized) THEN
IF (cionmax > 0.D0 .OR. rion > 0.D0) THEN
problem = 'linmodpb'
ELSE
problem = 'linpb'
ELSE IF (env_static_permittivity .GT. 1.D0 .OR. env_dielectric_regions .GT. 0) THEN
problem = 'generalized'
ELSE
problem = 'poisson'
|
screening |
REAL |
Default: |
0.D0
|
Status: |
OPTIONAL
|
Same as mix, but refers to the inner loop in the electrostatic calculation
(nested schemes only).
|
screening_type |
CHARACTER |
Default: |
'none'
|
Status: |
OPTIONAL
|
Uses the screened coulomb Green's function instead of the vacuum one.
'none' : unscreened coulomb
'input' : screened coulomb with screening length provided in input
'linear' : screened coulomb with screening length from linear component
of the problem
'optimal' : screened coulomb with screening length optimized (to be defined)
|
solver |
CHARACTER |
Default: |
see below
|
Status: |
OPTIONAL
|
Type of numerical solver used to get the solution of the electrostatic problem.
'direct' : Get the direct solution of the problem as provided by the adopted
numerical core. At this time, the FFT numerical core of Environ only
allows the 'poisson' problem to be solved directly
'fixed-point' : Use a fixed-point search algorithm (used to be the default for the
'generalized' problem in Environ v0.2)
'sd' : Use a steepest-descent algorithm, possibly preconditioned
'cg' : Use a conjugate-gradient algorithm, possibly preconditioned
'newton' : Use Newton's root-finding method
DEFAULT:
IF (env_electrolyte_ntyp .GT. 0 .AND. pbc_correction .NE. 'gcs') THEN
IF (electrolyte_linearized) THEN
solver = 'cg'
ELSE
solver = 'newton'
ELSE IF (env_static_permittivity .GT. 1.D0 .OR. env_dielectric_regions .GT. 0) THEN
IF (pbc_correction .NE. 'gcs') THEN
solver = 'cg'
ELSE
solver = 'fixed-point'
ELSE
solver = 'direct'
|
step |
REAL |
Default: |
0.3D0
|
Status: |
OPTIONAL
|
Step size to be used if step_type = 'input' (inherits the tasks of the old mixrhopol).
|
step_type |
CHARACTER |
Default: |
'optimal'
|
Status: |
OPTIONAL
|
The step size in gradient descent algorithms or iterative mixing.
'optimal' : step size that minimize the cost function on the descent direction
'input' : fixed step size as defined in input (step keyword)
'random' : random step size within zero and twice the optima value
|
tol |
REAL |
Default: |
1.D-5
|
Status: |
IMPORTANT
|
Accuracy of the electrostatic calculation: higher accuracies will require more cycles,
but will ensure smoother SCF convergence. As electrostatic solver iterations are
usually cheaper than SCF cycles, it is recommended to increase this accuracy whenever
the SCF has problems converging.
The units depend on the actual quantity that is computed by the algorithm, either the
potential (auxiliary = 'none') or an auxiliary charge (auxiliary = 'full'). In the
latter case, much smaller thresholds should be adopted (e.g. lower than 1.D-8).
|
|
|
Card: EXTERNAL_CHARGES { bohr | angstrom } |
Syntax:
EXTERNAL_CHARGES { bohr | angstrom
}
|
Description of items:
Card's options: |
bohr | angstrom
|
Default: |
(DEPRECATED) bohr
|
bohr : atomic positions are in cartesian coordinate, in atomic units
(i.e. Bohr radii)
angstrom : atomic positions are in cartesian coordinates, in Angstrom
|
Q |
REAL |
Total charge of external object
|
x, y, z
|
REAL |
Positions
NOTE: each atomic coordinate can also be specified as a simple algebraic expression,
see the description in the input file of PW.
|
spread |
REAL |
Default: |
0.5
|
Gaussian spread of the external charge density in atomic units
|
dim |
INTEGER |
Default: |
0
|
Dimensionality of the charge density
dim .EQ. 0 : point-like (gaussian shaped) charge density
dim .EQ. 1 : linear (gaussian shaped) charge density
dim .EQ. 2 : planar (gaussian shaped) charge density
|
axis |
INTEGER |
Default: |
3
|
Axis of the external charge density
if dim.EQ.0 : axis has no meaning/use
if dim.EQ.1 : axis identifies the direction of the linear charge density.
axis .EQ. 1|2|3 means lines along x|y|z respectively
if dim.EQ.2 : axis identifies the direction orthogonal to the planar charge density.
axis .EQ. 1|2|3 means planes orthogonal to x|y|z
|
|
|
Card: DIELECTRIC_REGIONS { bohr | angstrom } |
Syntax:
DIELECTRIC_REGIONS { bohr | angstrom
}
|
Description of items:
Card's options: |
bohr | angstrom
|
Default: |
(DEPRECATED) bohr
|
bohr : positions are in cartesian coordinate, in atomic units
(i.e. Bohr radii)
angstrom: positions are in cartesian coordinates,
in Angstrom
|
EpsSt |
REAL |
Static permittivity inside of dielectric region
|
EpsOpt |
REAL |
Optical permittivity inside of dielectric region
|
x, y, z
|
REAL |
Positions
NOTE: each coordinate can also be specified as a simple algebraic expression, see the description in the input file of PW.
|
width |
REAL |
Width of the dielectric region in atomic units
|
spread |
REAL |
Default: |
0.5
|
Spread of the dielectric region in atomic units
|
dim |
INTEGER |
Default: |
0
|
Dimensionality of the dielectric region
dim .EQ. 0 : sphere-like (erfc shaped) region
dim .EQ. 1 : cylinder-like (erfc shaped) region
dim .EQ. 2 : planar (erfc shaped) region
|
axis |
INTEGER |
Default: |
3
|
Axis of the dielectric region.
dim .EQ. 0 : axis has no meaning/use
dim .EQ. 1 : axis identifies the direction of the linear charge
density.
axis .EQ. 1|2|3 means lines along x|y|z respectively
dim .EQ. 2 : axis identifies the direction orthogonal to the planar
charge density.
axis .EQ. 1|2|3 means planes orthogonal to x|y|z
|
|
|
|