Rotational invariance? and help with data deck
----------
X-Sun-Data-Type: text
X-Sun-Data-Description: text
X-Sun-Data-Name: text
X-Sun-Content-Lines: 108
Regarding the earlier discussion on discrepancy between MOPAC and G90:
I responded with a query that perhaps rotational invariance in the MNDO-family
of methods may play a role (note, we saw that different values of physical
constants was the cause of that specific example). However, my questions about
invariance remain unresolved.
Mike Frisch writes:
> Mark Thompson suggest that a lack of rotational invariance might
> account for the descrepancy. Nice try, but that's not it. The
> difference between using the standard and z-matrix orientations in
> Gaussian 92 is 5x10^-7 kcal/mole. Similarly, using two different
> orientations in MOPAC 6.0 produces no change in the 5 decimal places
> printed in the energy.
ok...so I tried two different orientations in MOPAC 6.0 using formaldehyde.
No matter how I place the x,y,z coordinates, the program rotates the molecule
into some "standard" orientation before performing the calculation.
Putting two different orientations into MOPAC actually generates the
exact same set of cartesian coordinates used internally for the calculation.
I would like to run these calcs using the absolute coordinates I type in.
Is this possible?...can anyone help me with this? Am I building the data
deck incorrectly?
Anyway, whatever help I could get with these data decks would be
helpful. Thanks in advance.
Mark Thompson
*********************** case 1 *************************
Here is my data deck form.dat:
PRECISE FOCK 1SCF GEO-OK ITRY=100 1ELECTRON DENSITY VECTORS LARGE
formaldehyde
DATE: Thu May 14 16:03:43 1992
C 0.000000 0 0.578379 0 0.000000 0
O 0.000000 0 -0.629421 0 0.000000 0
H -0.920165 0 1.146823 0 0.000000 0
H 0.920165 0 1.146823 0 0.000000 0
Here is the geometry that MOPAC generates to perform the calc.
(from the output file)
CARTESIAN COORDINATES
NO. ATOM X Y Z
1 C 0.0000 0.0000 0.0000
2 O 1.2078 0.0000 0.0000
3 H -0.5684 0.9202 0.0000
4 H -0.5684 -0.9202 0.0000
(molecule has been rotated onto the xy plane along the x-axis)
********************** case 2 *****************************
Here is the same geometry in another orientation:
(rotated 90 deg. about y-axis and displaced 5 angs. along x)
PRECISE FOCK 1SCF GEO-OK ITRY=100 1ELECTRON DENSITY VECTORS LARGE
formaldehyde
DATE: Thu May 14 16:03:43 1992
C 5.000000 0 0.578379 0 0.000000 0
O 5.000000 0 -0.629421 0 0.000000 0
H 5.000000 0 1.146823 0 -0.920165 0
H 5.000000 0 1.146823 0 0.920165 0
Here is the geometry that MOPAC generates to perform the calc.
CARTESIAN COORDINATES
NO. ATOM X Y Z
1 C 0.0000 0.0000 0.0000
2 O 1.2078 0.0000 0.0000
3 H -0.5684 0.9202 0.0000
4 H -0.5684 -0.9202 0.0000
(same as case 1)
**********************************************************
Reordering the atoms in the data deck does give different
coordinates in MOPAC. However, since the molecule is still lined
up on the x-axis and in the xy plane, the rotational invariance
of the two-electron integrals method will not be observed.
----------
X-Sun-Data-Type: default
X-Sun-Data-Description: default
X-Sun-Data-Name: .signature
X-Sun-Content-Lines: 14
----------------------------------------------------------------
Mark A. Thompson email: d3f012[ AT ]pnlg.pnl.gov
Sr. Research Scientist voice: 509-375-6734
Molecular Science Research Center FAX : 509-375-6631
Pacific Northwest Laboratory
PO Box 999, Mail Stop K1-90
Richland, WA. 99352
----------------------------------------------------------------
Disclaimer: The views expressed in this message are solely my own and
do not represent Battelle Memorial Institute, Pacific Northwest
Laboratory, or any of its clients.