# Overlay 8

IOp(8/5)

Whether to pseudo-canonicalize ROHF orbitals.

-2 | Yes, and save over canonical MOs setting ILSW for UHF. |

-1 | Yes. |

0 | Default (Yes if ROHF). |

1 | No. |

IOp(8/6)

Bucket selection.

0 | Buckets for MP2: (IA/JB). |

1 | Buckets for stability: (IA/JB), (IJ/AB). |

2 | Buckets for CID or MP3: (IJ/AB), (IA/JB), (IJ/KL). |

3 | Buckets for semi-direct MP4DQ, CISD, QCISD, BD: (IJ/AB), (IA/JB), (IK/KL), (IJ/KA). |

4 | CISD or MP4SDQ or MP4SDTQ, but includes (IA/BC). |

5 | The complete set of transformed integrals. |

6 | Full transformation if this is consistent with MaxDisk, otherwise same as 3. |

7 | Full transformation if this is consistent with MaxDisk, otherwise same as 4. |

IOp(8/7)

SCF convergence test.

0 | Test that SCF has converged. |

1 | Do not test SCF convergence (mainly used for testing). |

IOp(8/8)

L811: Whether to delete MO integrals.

0 | Default (No). |

1 | Yes. |

2 | No. |

IOp(8/9)

L802: Debug control.

0 | Operate normally. |

-N | Force N orbitals per pass. |

L804: Direct Transformation Control.

0 | Operate normally. |

1 | Generate and test RInt3 array (L804). |

2 | Accumulate MP2 force constant terms in direct fashion. |

3 | Write the MO basis first derivative ERI’s to disk. |

10 | Force fully in-Core algorithm (L804 only). |

20 | Force transformed integrals in Core algorithm. |

30 | Force semi-direct transformation. |

100 | Force output bucket in Core anti-symmetrization. |

200 | Force sorting for output bucks. |

1000 | Force semi-direct mode 1. |

2000 | Force semi-direct mode 2. |

3000 | Force semi-direct mode 3 if IOp(8/6)=3. |

4000 | Force semi-direct mode 4 if IOp(8/6)=3. |

00000 | Default (10000). |

10000 | Do not symmetry compress transformed integrals. |

20000 | Do symmetry compress transformed integrals (buckets) (This will cause windowed MOs, reordered in the order of representations like occ-rep1,occ-rep2,… virt-rep1,virt-rep2,… eigenvalues and symm. assignment vectors will be put in correspondence with vectors. VGZ). |

30000 | Symmetry compress transformed integrals only if RHF. (Upper triangle of symmetry compressed integrals for IOp(8/6)=5 or 4 only! (VGZ)). |

40000 | Store buckets of single-bar integrals, not symmetry compressed. |

100000 | Reorder MOs, eigenvalues and symmetry assignment vectors according to the representations. |

IOp(8/10)

Window is selected as follows:

-N | Use the top N occupieds and lowest N virtuals. |

0 | Default, same as 4. |

N | 1 ≤ N ≤ 89 selects frozen-core type N. |

1 | The largest noble gas core is frozen. |

2 | G2 frozen-core: the largest noble gas core and main group d orbitals are frozen, except that the outer sp electrons of 3rd row and later alkali and alkali earth elements are retained. |

3 | The next to the largest noble gas core is frozen. |

4 | The largest noble gas core and main group d’s are frozen. |

5 | G3 frozen-core: the largest noble gas core is frozen, except that the outer sp electrons of 3rd row and later alkali and alkali earth elements are retained. |

6 | G4 frozen-core: the largest noble gas core is frozen, except that the outer sp electrons of 2nd row and later alkali and alkali earth elements are retained. For basis sets with double-zeta cores, core virtuals are also frozen. |

7 | CBS-Wes core: noble gas except 3sp valence K-Zn, 3d valence Ga-As. |

90 | Use all MOs. |

91 | The window is specified by IOp(8/37-38). If IOp(8/37) is 0, a card is read in indicating the start and the end. A negative value for the end deletes the top virtuals. |

92 | The window is recovered from RWF 569. |

93 | The window is recovered from file 569 on the checkpoint file. |

94 | Read a list of orbitals to freeze. |

000 | Default (200). |

10x | Use orbital energies to choose core orbitals. |

20x | Use overlap with atomic core orbitals from Harris to choose core orbitals. |

30x | Use overlap with atomic core orbitals from Core Ham to choose core orbitals. |

IOp(8/11)

MO coefficient, orbital energy, and number of electrons test.

0 | Default, same as 2 except for during BD iterations or BD=Read. |

1 | Just print a warning message. |

2 | Kill the job if any MO coefficients are greater than 1000.0 or the smallest difference between occupied and virtual orbital energies is less than 0.001. Also, kill a frozen-core job if there is significant core-valence mixing in the canonical orbitals. |

00 | Default, same as 10. |

10 | Suppress such a test (CPHF may still be done for such a case). |

20 | Kill the job if there is no corr. energy; e.g., if there is only 1 electron or 1 virtual spin-orbital. |

IOp(8/16)

L811: Maximum number of orbitals per pass (only if integral derivative file is being written). Default is as many as fit with Max Disk.

IOp(8/18)

L811: Which type of derivative transformation to do.

0 | Default, same as 3. |

1 | Non-canonical, Uij,x = -1/2 Sij,x. |

2 | Canonical, Uij,x = (Fij,x – EjSij,x) / (Ei-Ej) Note that this blows up for degenerate orbitals and is intended primarily for debugging. |

3 | Non-canonical, Uij,x = -1/2 Sij,x, except canonical in frozen-active blocks. |

4 | Non-canonical, Uij,x = -Sij,x Uji,x = 0. |

5 | Canonical occupieds, Uab,x = -Sab,x/2. |

6 | Canonical virtuals, Uij,x = -Sij,x/2. |

IOp(8/19)

L811: The nature of the perturbation(s).

0 | Default (1st order nuclear and electric field). |

IJK | Nuclear Kth order. Electric field Jth order. Magnetic field Ith order. |

IOp(8/20)

L811: Which terms to include.

0 | Default (same as 11). |

1 | MO derivative times integral term. |

10 | MO times integral derivative term. |

IOp(8/23)

L811: Algorithm control.

0 | Default (32). |

1 | Unused. |

2 | Accumulate MP2 force constant terms in direct fashion. |

3 | Write the MO basis first derivative ERI’s to disk. |

20 | Force fully direct. |

30 | Force semi-direct. |

IOp(8/24)

Whether to try to transform old amplitudes on the checkpoint file.

0 | Default: 1 if doing BD=Read and amplitudes are present; 2 otherwise. |

1 | Yes. |

2 | No. |

10 | Transform Z-amplitudes as well. |

20 | Do not transform Z-amplitudes as well. |

000 | Default, transform EOM amplitudes if transforming ground-state ones. |

100 | Transform EOM amplitudes. |

200 | Do not transform EOM amplitudes. |

IOp(8/28)

L921, L922: Hack number of occupieds for full CI.

-1 | Transform all orbitals (after freezing core) as occupieds (i.e., set NOA=NOB=NROrb in transformation). |

0 | No. |

N | Transform N orbitals (after frozen core) as occupieds (i.e., set NOA=NOB=N for purposes of transformation). |

IOp(8/29)

L811: Requested diskusage. This will determine the number of times AO integrals and derivatives are evaluated unless overridden by IOp(8/31). This only applies if the integral derivatives are not stored.

-3 | Use as much as desired, independent of MAXDISK. |

-2 | Use an amount which is similar to the maximum disk usage in other parts of the MP2 freq. code. |

-1 | Use as much as needed for maximum efficiency, subject to the limit imposed by MAXDISK. |

0 | Default (-1). |

N | N evaluations and hence N coarse tiled batches (1…6 are the currently implemented options). |

IOp(8/30)

Type of window.

0 | Default. Set up /Orb/ as indicated by IOp(8/10). |

1 | Test window. Set up for full but zero core MOs. |

-1 | Set up /Orb/ for a full window but then blank the wavefunction coefficients in L804. |

IOp(8/36)

Whether to update force constants with the MP2 product of MP2 integral derivatives term (only applies if integral derivative file is not written).

0 | Default (Yes). |

1 | Yes. |

2 | No. |

00 | Default on whether to make Poo and Pvv for MP2. (Yes if Ix is not stored, no otherwise). |

10 | Yes. |

20 | No. |

IOp(8/37)

Integer specifying first window parameter (n).

IOp(8/38)

Integer specifying second window parameter (m).

IOp(8/39)

Localized orbital method adopted in SAC/SAC-CI.

0 | Default. No localization. |

1 | Boys method. |

2 | Population method. |

3 | Boys + population method. |

IOp(8/40)

Handling of ROHF window.

0 | Default (2). |

1 | Use ROMP2 approach, forming pseudo-canonical alpha and beta orbitals and doing UHF transformation. |

2 | Treat as RHF, transforming only alpha orbitals. |

IOp(8/41)

Transformation of spin-orbitals (alpha only) within occupied and unoccupied orbital subspaces by minimum orbital-deformation (MOD) method.

0 | Default. No. |

1 | No, but save MOs. |

2 | Yes. Take reference MOs from disk if available. |

3 | No for the 1st geometry of opt, yes otherwise. |

IOp(8/42)

Whether to reorder MOs during potential surface exploration.

0 | No. |

1 | Yes. |

2 | Yes (for SAC-CI single point calculation). |

00 | Use orbital energies in ordering. |

10 | Don’t use orbital energies in ordering. |

000 | Use second moments in ordering. |

100 | Don’t use second moments in ordering. |

0000 | Use dipole moments in ordering. |

1000 | Don’t use dipole moments in ordering. |

IOp(8/46)

Indicates special case of non-HF calculation.

0 | Default – MOs are canonical HF orbitals. |

1 | Input orbitals are not canonical HF and pseudo-canonical orbitals must be generated here for the post-SCF. |

10 | Generate HF pseudo-canonical even if the original SCF method was not (i.e., Kohn-Sham). |

IOp(8/47)

Whether L804/L811 should generate results compressed over active atoms.

0 | Default (2). |

1 | Active atoms. |

2 | Full list. |

3 | Full list, but blank contributions from inactive atoms (no difference from 2 for overlay 8). |

4 | Active atoms, and store Hessian contributions over active atoms only. |

IOp(8/60-62)

Over-ride standard values of IRadAn, IRanWt, and IRanGd. For DFTCV, IRadAn defaults to 299974 rather than the global default.

IOp(8/68)

EOM-CCSD

0 | No EOM. |

1 | Do EOM with the default algorithm (right and left spaces separately). |

11 | Do EOM doing only the transition energy (right space). |

21 | Do EOM doing right and left eigenvectors using the same expansion space for both. |

31 | Do EOM doing right and left eigenvectors using biorthogonal expansion spaces. |

IOp(8/69)

EOM: Number of states per irreducible representation (largest Abelian subgroup) to do.

0 | Default (2). |

N | N per irreducible representation. |

-1 | Read the number for each irreducible representation, all from one line. |

The order of irreducible representations is the same as printed for symmetry-adapted basis functions by L301.

IOp(8/87)

Accuracy of integrals.

0 | Default (12). |

N | 10^{-N}. |

IOp(8/105)

Convergence of amplitudes for EOM iterations.

0 | Default (1.d-5). |

N | 10^{-N}. |

IOp(8/106)

Number of EOM states for LR transition densities.

0 | Default (None). |

-1 | All. |

N | First N of each symmetry. |

IOp(8/107)

EOM state of most interest.

0 | Default (1st excited state). |

N | N^{th}excited state. |

IOp(8/108)

EOM-CCSD: Total number of states to do. Guesses are taken from the checkpoint file if RdAmp was specified, with remaining states taken from the CIS guess in CIS energy order.

0 | Default (2*NIrrep) |

IOp(8/109)

IFact for Davidson in EOM-CC.

IOp(8/110)

State-to-State transition dipoles in EOM-CC:

0 | None. |

1 | From state NRoot to higher states. |

2 | From state NRoot to higher and lower states. |

IOp(8/111)

MaxIt for EOM.

IOp(8/112)

MaxMin for EOM.

IOp(8/113)

WhenSc for EOM.

IOp(8/114)

IRdLft for EOM.

IOp(8/115)

IFirst for EOM.

IOp(8/116)

Compute DFT estimate of core-core and core-valence correlation?

0 | Default. |

1 | Yes. |

2 | No. |

10 | Include empirical corrections for total energies. |

20 | Do not include empirical corrections. |

The default is not to compute the correction, and if the correction is requested, to include the total energy terms only for CBS-Wes style frozen-core (the only case for which they have been determined). Corrections are only included for elements H-Ar.

IOp(8/123)

Flag for SOS in EOM.

Last updated on: 21 October 2016. [G16 Rev. C.01]