ET module


Clicked A Few Times
I'm wondering how the electronic coupling energy (V(RP)) is calculated in the ET module. I have looked at the JACS 1990, 112 4206-4214 article that is cited which states:

V(RP) = (1-S(RP)^2)^-1 * [ H(RP) - S(RP) * ( H(RR) + H(PP) ) / 2 ]

I have calculated the triplet exchange for a napthlene dimer using CDFT and get the following:

 Electronic energy of reactants     H(RR)   -2141.0488936556
 Electronic energy of products      H(PP)   -2141.0053307554
 
 Reactants/Products overlap S(RP) : -1.11D-03
 
 Reactants/Products interaction energy:
 -------------------------------------
 One-electron contribution         H1(RP)       4.1843395986
 
 Beginning calculation of 2e contribution
 Two-electron integral screening (tol2e) : 1.11D-10
 
  int_init: cando_txs set to always be   F
 Two-electron contribution         H2(RP)      -1.8070713687
 Total interaction energy           H(RP)       2.3772682299
 
 Electron Transfer Coupling Energy |V(RP)|      0.0052412415
                                                    1150.319 cm-1
                                                    0.142622 eV
                                                       3.289 kcal/mol


However, when a calculate V(RP) using the above equation I get 0.00072814 not 0.0052412415. Why the difference? Are the H(RR) and H(PP) values not normalized or somthing?

Forum Vet
Round-off issue
This is simply a round-off issue in the output. If you use the S(RP) to be -1.11D-03 you will get your result, but that this digit is a rounding of the real number. Just try -1.108D-03 for example.

I.e., the formula you use is correct and the printed data is correct, just not printed accurate enough to reproduce the number for V(RP).

Bert

Clicked A Few Times
Thanks! Is there a way to print the "real number"? Or would I have to change the source code?

Forum Vet
You would have to change the source code in src/etrans/et_calc.F around line 579.


Forum >> NWChem's corner >> General Topics