Calculating transfer integrals with Gaussian and the projective method.
This code ported over from the defunct Google Code. https://code.google.com/archive/p/j-from-g03/
Adapted from code by J.Kirkpatrick to deal with GAMESS-UK output; joejk2 12/09/07
-
Compile with make. This puts 'rewrite_S_Phi_E.x' and 'get_J.m' in ~/bin
-
Run three gaussian jobs:
-
single molecule 1 in the same orientation as when it is in the pair (e.g. xxx_1.com)
-
single molecule 2 -"- (e.g. xxx_2.com)
-
pair of molecules (e.g. xxx_pair.com)
N.B. Typical Gaussian input line:
-
nosymm necessary to prevent reorientation of molecule to standard orientation.
punch=mo puts orbital coefficients in fort.7. This must be moved to xxx_x.pun
iop(3/33=1) forces output of the overlap matrix in the xxx_x.log file
```
-
Run 'rewrite_S_Phi_E.x' with the prefix to gaussian calculation (1), (2) and (3) !! IN THIS ORDER !!. This program expects to find 3 log files and 3 pun files.
-
You should now have 7 output .txt files: Evls_pair.txt MOs_1.txt MOs_2.txt MOs_pair.txt S_1.txt S_2.txt S_pair.txt
-
Run 'octave ~/bin/get_J.m'. You will be prompted for the number of MOs and the HOMO of a monomer. For very small molecules (with nBasis<3) you might have to comment out some calculations attempted (on non-existant orbitals) in get_J.m
Compare examples/ethylene_b3lyp with Figure 1 in JACS 2006 128 9882-9886 http://dx.doi.org/10.1021/ja061827h