RI-MP2 Keywords in the Jaguar Input File
- Overview
- Examples
The rimp2 keyword allows you to request a resolution-of the identity Møller-Plesset perturbation theory (RI-MP2) calculation. By default, RI-MP2 is off. For more information on the RI-MP2 method, see RI-MP2 Calculations and Pseudospectral Resolution-of-the-identity MP2 Techniques. RI-MP2 keywords are given in Table 1.
|
Keyword |
Value |
Description |
|
0 |
Do not run a RI-MP2 calculation. |
|
|
|
1 |
Run a RI-MP2 calculation. |
| see below |
Specifies the auxiliary basis set to be used in a RI-MP2 calculation. By default, Jaguar automatically selects an auxiliary basis set based on the normal basis set. For example, if the normal basis set is |
|
|
−1 |
Treats the core molecular orbitals as frozen in a RI-MP2 calculation. The number of core molecular orbitals is the sum of the numbers of atomic core orbitals. The number of atomic core orbitals are listed here. |
|
|
|
0 |
No molecular orbitals are treated as frozen. |
|
|
>0 |
The number of molecular orbitals with the lowest energies to be frozen. |
|
0 |
No virtual orbitals are treated frozen by default. |
|
|
|
>0 |
The number of virtual molecular orbitals with the highest orbitals energies to be frozen. |
|
1 |
The default scaling factor for the MP2 same-spin correlation energy, which corresponds to not doing Spin Component Scaling. |
|
|
|
>0 |
The scaling factor for the MP2 same-spin correlation energy. |
|
1 |
The default scaling factor for the MP2 opposite-spin correlation energy, which corresponds to not doing Spin Component Scaling. |
|
|
|
>0 |
The scaling factor for the MP2 opposite-spin correlation energy. |
Available Auxiliary Basis Sets
The available auxiliary basis set values for the rimp2_auxbas keyword are listed below:
- cc-pvdz-rifit
- cc-pvtz-rifit
- cc-pvqz-rifit
- aug-cc-pvdz-rifit
- aug-cc-pvtz-rifit
- aug-cc-pvqz-rifit
- def2-sv(p)-rifit
- def2-svp-rifit
- def2-svpd-rifit
- def2-tzvp-rifit
- def2-tzvpd-rifit
- def2-tzvpp-rifit
- def2-tzvppd-rifit
- def2-qzvp-rifit
- def2-qzvpp-rifit
- def2-qzvppd-rifit
Default Frozen Core Settings
The following table lists the number of atomic core orbitals used by Jaguar:
| Atom | # of frozen orbitals |
| H-He | 0 |
| Li-Ne | 1 |
| Na-Ar | 5 |
| K-Zn | 9 |
| Ga-Kr | 14 |
| Rb-Cd | 18 |
| In-Xe | 23 |
| Cs-Lu | 27 |
| Hf-Hg | 34 |
| Tl-Rn | 39 |
| Fr-Lr | 43 |
By default, the RI-MP2 method uses an SCF wavefunction calculated with the pseudospectral method, which requires a basis set that is set up for the pseudospectral method—see Basis Sets for basis set information. If the SCF calculation is having convergence issues, you may want to set the scf_nops keyword to 1. This setting computes an analytic SCF wavefunction, which is generally more accurate than the pseudospectral SCF wavefunction. The pseudospectral method is still used in the calculation of RI-MP2 gradients.