In addition to the fragment-specific options, there are also a number of options available in FDE calculations that will be described in the following.
FDE
{approximants to the kinetic energy dependent component of the embedding potential}
{CJCORR [rho_cutoff]}
{GGAPOTXFD exchange approximant}
{GGAPOTCFD correlation approximant}
{FULLGRID}
{RELAXCYCLES n or FREEZEANDTHAWCYCLES n}
{RELAXPOSTSCF or FREEZEANDTHAWPOSTSCF}
{EXTPRINTENERGY}
(PRINTRHO2}
end
approximants to the kinetic energy dependent component of the embedding potential
Several approximants to the kinetic-energy-dependent component of the
effective potential given in Eq. (21) of [Ref. 184] are available.
None of them is applicable if the embedded system is covalently bound to its environment.
The user is recommended to look at the numerical value of the TSNAD(LDA) parameter which is given in the units of energy and
can be considered as a measure of the overlap. The following rule of thumb should be applied:
if this parameter is smaller than the estimated interaction energy between the embedded subsystem and the environment, then
the available approximants are most probably adequate. If it exceeds this limit, the results can be less reliable.
Printing TSNAD(LDA) is not done by default, as it can be quite time-consuming. Its printing is switched on
by including "EXTPRINTENERGY", and "PRINTRHO2", and "GFULLGRID" n the FDE input block.
If no kinetic energy approximant is specified, by default the
local-density approximation (Thomas-Fermi approximant) is used.
For an assessment of approximants for weakly overlapping pairs of densities see Refs. [238,
239,
188,
241].
Based on these studies, the use of PW91k (= GGA97) is recommended.
THOMASFERMI (default)
Local-density-approximation form of vt[rhoA,rhoB] [237] derived from Thomas-Fermi expression for Ts[rho] [194, 195].
GGA97 (or PW91K)
Generalized-gradient-approximation form of vt[rhoA,rhoB] [239] derived from the Lembarki-Chermette [197] approximant to Ts[rho]. This approximant is currently the recommended one based on the numerical analysis of its accuracy [188,239] and the fact that the used enhancement factor disappears at large reduced density gradients, i.e. where the second-order gradient-expansion approximation fails [238, 241].
OBSOLETE APPROXIMANTS (can be used but GGA97 leads usually to a better embedding potential see [238,239])LLP91
Generalized-gradient-approximation form of vt[rhoA,rhoB] [238] derived from Lee-Lee-Parr [Ref. 198] approximant to Ts[rho].
PW86k
Generalized-gradient-approximation form of vt[rhoA,rhoB] [238] derived from the Fuentealba-Reyes approximant to Ts[rho] [242].
THAKKAR92
Generalized-gradient-approximation form of vt[rhoA,rhoB] [239] derived from the Thakkar approximant to Ts[rho] [201].
APPROXIMANTS WHICH MIGHT BE USEFUL ONLY FOR THEORY DEVELOPMENT
The accuracy of these approximants was investigated in detail [239, 238, 188, 241]. Each of them was shown to lead to a qualitatively incorrect embedding potential. They shouldn't be used in practical applications.
COULOMB
Neglecting completely vt[rhoA,rhoB] (vt[rhoA,rhoB] equals zero) together with the exchange-correlation component of the embedding potential introduced by Wesolowski and Warshel.
TF9W
The approximant to vt[rhoA,rhoB] [184] derived from the second-order gradient expansion [242]] for Ts[rho].
WEIZ
The approximant to vt[rhoA,rhoB] [241] derived from the von Weizsäcker approximant to Ts[rho] [186].
OL91A
Generalized-gradient-approximation form of vt[rhoA,rhoB] [238] derived from the first Ou-Yang and Levy approximant to Ts[rho] [200].
OL91B
Generalized-gradient-approximation form of vt[rhoA,rhoB] [239] derived from the second Ou-Yang and Levy approximant to Ts[rho] [200].
LONG DISTANCE CORRECTIONS TO THE EFFECTIVE POTENTIAL
CJCORR
Option to switch on a long-distance correction. By default this option is not used. As was shown in Ref. [220], with the available approximate kinetic-energy approximants, the embedding potential has the wrong form in the limit of a large separation of the subsystems. In particular, it was shown that this can have serious consequences in the case of "supermolecular expansion of electron density of each subsystem" calculations (USEBASIS option). In Ref. [220], a correction is proposed that enforces the correct long-distance limit. (See also this reference for limitations of this correction.)
CJCORR [rho_cutoff]
This option switches on the long-distance correction. This option has to be used in combination with one of the above kinetic-energy approximants. By default, a density cut-off of 0.1 is employed.
GGAPOTXFD
GGAPOTCFD
Option to specify the nonadditive exchange-correlation approximant. By default, in the construction of the effective embedding potential the exchange-correlation approximant that was specified in the XC block is used. It is possible to specify a different approximant with the GGAPOTXFD and GGAPOTCFD options. This is particularly useful in combination with the use of model potentials like SAOP, that can not be used in the embedding potential because of their orbital dependence. (For a discussion, see Ref. [189].)
GGAPOTXFD exchange approximant
The exchange approximant is used in the construction of the embedding potential. The same exchange approximants as in the XC key are available.
GGAPOTCFD correlation approximant
The correlation approximant is used in the construction of the embedding potential. The same correlation approximants as in the XC key are available.
FULLGRID
By default, FULLGRID is not used, and in FDE calculations the integration grid is generated as described in Ref. [185] by including only atoms of the frozen subsystem that are close to the embedded subsystem in the generation of the integration grid. The distance cutoff used is chosen automatically, based on the extent of the basis functions of the embedded subsystem. (It can also be chosen manually, see the option qpnear in the INTEGRATION key) This scheme results in an efficient and accurate integration grid. However, it is possible that the default integration scheme is not accurate enough. This can be the case for weakly interacting systems and when the distance between the frozen and the embedded system is large. It is therefore recommended to check the quality of the default integration grid by comparing to results obtained using the full supermolecular grid (FULLGRID option).
If the subkey FULLGRID is included, all atoms of the frozen system are included in the generation of the integration grid. This results in the same grid that would be used in a supermolecular calculation of the combined frozen and embedded system. The integration grid generated by this option might be much larger than the default grid. This option should be used to check the quality of the default integration grid.
RELAXCYCLES n or FREEZEANDTHAWCYCLES n
Specifies the maximum number n of freeze-and-thaw iterations [Ref. 240] that are performed (for frozen fragments with the RELAX) option. If a smaller number of iterations is specified as a fragment-specific option, for this fragment this smaller number is used. Furthermore, if convergence is reached earlier, no more iterations will be performed.
RELAXPOSTSCF or FREEZEANDTHAWPOSTSCF
If this option is included, several post-SCF properties will be calculated after each freeze-and-thaw cycle [Ref. 240]. These are otherwise only calculated in the last cycle.
EXTPRINTENERGY
PRINTRHO2
If the options EXTPRINTENERGY and PRINTRHO2 are included (both are needed and should be listed on separate lines), several additional quantities will be printed, including TSNAD(LDA). In order to obtain meaningful numbers, also the FULLGRID keyword (see above) has to be used.
