Pseudo

This keyword requests that a model potential be substituted for the core electrons. The Cards option is by far its most-used mode. Gaussian supports a new effective core potential (ECP) input format (similar to that used by ExtraBasis) which is described in the Format tab. When reading-in pseudopotentials, do not give them the same names as any internally-stored pseudopotentials: CEP, CHF, LANL1, LANL2, LP-31, SDD, and SHC.

If used with ONIOM, the Pseudo and keyword applies to all layer of the ONIOM. If you want to read in ECPs only for one ONIOM layer, then use the GenECP keyword instead.

Without any options, this keyword defaults to Pseudo=Read.

Read

Read pseudopotential data from the input stream. Input is described in the next subsection below. Cards is a synonym for Read.

SOScal

When reading pseudopotentials from input using Pseudo=Read, scale the spin-orbit coefficients (if present) by 2/L as appropriate for published CRENBL potentials. The default is not to scale, which is appropriate for Dolg (Stuttgart) potentials.

CHF

Requests the Coreless Hartree-Fock potentials. This option is normally used with the LP-31G basis sets.

SHC

Requests the SHC potentials.

LANL1

Requests the LANL1 potentials.

LANL2

Requests the LANL2 potentials.

Old

Read pseudopotential data using the old format (used by Gaussian 92 and earlier versions).

Full ECP Input Format

Effective Core Potential operators are sums of products of polynomial radial functions, Gaussian radial functions and angular momentum projection operators. ECP input therefore specifies which potential to use on each atomic center, and then includes a collection of triplets of:

(coefficient, power of R, exponent)

for each potential for each term in each angular momentum of the ECP. Since only the first few angular momentum components have different terms, the potential is expressed as (1) terms for the general case, typically d or f and higher projection, and (2) the extra terms for each special angular momentum. Thus for an LP-31G potential, which includes special s and p projected terms, the input includes the general (d and higher) term, the s-d term (i.e., what to add to the general term to make the s component) and the p-d term.

All ECP input is free-format. Each block is introduced by a line containing the center numbers (from the molecule specification) and/or atomic symbols, specifying the atoms and/or atoms types to which it applies (just as for general basis set input-see the discussion of the Gen keyword). The list ends with a value of 0.

The pseudopotential for those centers/atoms follows:

Name,Max,ICore

Name of the potential, maximum angular momentum of the potential (i.e., 2 if there are special s and p projections, 3 if there are s, p, and d projections), and number of core electrons replaced by the potential. If Name matches the name of a previous potential, that potential is reused and no further input other than the terminator line is required.

For each component (I=1 to Max) of the current potential, a group of terms is read, containing the following information:

Title

A description of the block, not otherwise used.

NTerm

Number of terms in the block.

NPower,Expon,Coeff[,SO]

Power of R, exponent, and coefficient for each of the NTerm terms. NPower includes the R² Jacobian factor. The optional SO coefficient is for use with ECP basis sets which include this term.

Simplified ECP Input Format

Gaussian adds flexibility to ECP input by allowing it to include pre-defined basis sets names. An ECP definition may be replaced by a line containing the standard keyword for a pre-defined basis set. In this case, the ECPs within the specified basis set corresponding to the specified atom type(s) will be used for that atom (see the examples).

In Pseudo input, keywords for these ECPs are of the form XYn where n is the number of core electrons which are replaced by the pseudopotential and X denotes the reference system used for generating the pseudopotential (S for a single-valence-electron ion or M for a neutral atom).

Y specifies the theoretical level of the reference data: HF for Hartree-Fock, WB for Wood-Boring quasi-relativistic and DF for Dirac-Fock relativistic. For one- or two-valence electron atoms SDF is a good choice; otherwise MWB or MDF is recommended (although for small atoms or for the consideration of relativistic effects, the corresponding SHF and MHF pseudopotentials may be useful).

The Stuttgart/Dresden ECPs are not uniformly available across the periodic table. The following table shows the availability of the various XY combinations, along with valid values for n. The Defaults columns list the equivalencies for the SDD keyword (which selects an all electron basis set through Cl and ECPs thereafter) and when IOp(3/6) is set to 6 (which selects ECPs for all elements).

Note: These ECPs are not available for elements 87 (Fr), 88 (Ra), and 105 and higher.

The basis set data follows the molecule specification section. The first line of the ECP data requests that a potential be read in (type 7) for atoms number 1 and 3 (the oxygen atoms). No potential is to be used for atoms 2 and 4 (the hydrogen atoms).

The second line of ECP data begins the input for the centers requiring a read-in potential: in this case, oxygen atoms. The potential on these centers is named OLP, it is a general term and applies to angular momentum 2 (d) and higher, and the potential replaces two electrons. Next comes a title for the general term (D component), and the number of components of that term (3); the individual components follow on the next 3 lines. Next come the corrections for the projected terms in two sections, lowest angular momentum first. Each section again consists of a title line, the number of terms to follow, and then the terms themselves.

Using Standard Basis Set Keywords to Specify ECPs. The following input file illustrates the use of the simplified ECP input format: