$DAMP group (optional, relevant if RUNTYP=MAKEFP)
This group provides control over the screening of the
charge term in the distributed multipole expansion used by
the EFP model for electrostatic interactions, to account
for charge penetration. See
M.A.Freitag, M.S.Gordon, J.H.Jensen, W.A.Stevens
J.Chem.Phys. 112, 7300-7306(2000)
L.V.Slipchenko, M.S.Gordon
J.Comput.Chem. 28, 276-291(2007)
The screening exponents are optimized by fitting a
damped multipolar electrostatic potential to the actual
quantum mechanical potential of the wavefunction, computed
on concentric layers of united spheres (namely, "GEODESIC"
layers for WHERE=PDC in $ELPOT). See $STONE's generation
of the unscreened classical multipoles, $PDC's generation
of the true quantum potentia, and $DAMPGS.
Different multipole damping functions can be generated.
The first contains a single exponential form,
(1 - beta*exp(-alpha*r))
and the second function is a single Gaussian form,
(1 - beta*exp(-alpha*r**2))
The exponent 'alpha' values are optimized (normally with
beta=one), with starting values defined in $DAMPGS. The
exponential fit is used for fragment-fragment charge
penetration screening, while the Gaussian fit is used in ab
initio-fragment screening. See equations 28 and 4 in the
reference. These two screen only the charge-charge
interactions.
It is also possible to generate a "higher order
exponential" screening term, meaning that in addition to
the charge-charge energy, also affects charge-dipole,
charge-quadrupole, and dipole-dipole energy terms.
Words of advice:
1. Higher order screening is usually similar in accuracy to
just charge-charge screening, except in molecules without
dipole moment, such as ethylene or benzene.
2. If the bond midpoints have smaller charges, it may be
more physically reasonable to screen only the atomic
monopoles, see ISCCHG.
3. Use of the numerical Stone distributed multipole
analysis may not be fully converged with respect to the
level of highest used multipole moment (octapole) and
corresponding energy terms (quadrupole-quadrupole), which
makes screening much more problematic.
4. Accuracy of screening with the damping function of a
single exponential form depends on a region of fitting the
quantum mechanical electrostatic potential, i.e., a radius
of first sphere with grid points (parameter VDWSCL in
$PDC). A general trend is that for molecules with stronger
electrostatic interaction, and, consequently, shorter
intermolecular separations, e.g., methanol and water,
smaller values of VDWSCL are preferable, whereas for weaker
interacting molecules, e.g., dichloromethane and acetone,
bigger VDWSCL values are more acceptable. Our recommended
VDWSCL values are 0.4-0.5 for methanol, 0.5-0.8 for water,
and 0.7-0.9 for weaker bonded molecules. Note that VDWSCL
values of 1.0 and higher often result in not converged or
badly converged damping parameters, and are not
recommended. The default VDWSCL value is 0.7.
5. If the non-linear parameters alpha increase to 10, that
term is effectively removed from the screening. This
happens sometimes with buried atoms, and fairly often with
bond mid-points.
6. Double check the numerical results carefully.
ISCCHG = 0 use both atoms and bond midpoints as screening
centers (the default)
1 use only atoms as screening centers
IFTTYP = selects the type of multipole screening fit:
0 means generate a Gaussian fit, for use as
SCREEN input in $FRAGNAME.
2 means generate an exponential charge-charge fit,
for use as SCREEN2 input in $FRAGNAME.
3 means generate an exponential higher order fit,
for use as SCREEN3 input in $FRAGNAME.
If you wish to use Gaussian screening for EFP-EFP,
simply copy the SCREEN output into a SCREEN1 section.
IFTFIX = 0 means the coefficients in the fit (beta) are
free parameters
1 means the coefficients are held to unity.
In case the linear coefficients become large,
and particularly if they are negative, a fit
with unit coefficients is more reasonable.
The default is to do both fits in one run, IFTTYP(1)=2,0,
using unit coefficients, IFTFIX(1)=1,1.
The remaining parameters are seldom given:
NMAIN = the number of centers to receive a smaller alpha
initial value, 2.0, which defaults to the number
of atoms. The remaining centers, usually the
bond midpoints, receive a larger starting value,
4.0. $DAMPGS gives more control of the values.
MAXIT = maximum iterations in the fit, default=30.
THRSH = printing threshold for large deviations. The
default is 100.0 kcal/mol.
==========================================================
==========================================================
generated on 7/7/2017
