man oeprop (1): One-Electron Property Program

DESCRIPTION

The program oeprop computes expectation values of one-electron property operators using a one-particle density matrix computed from an eigenvector in PSIF_CHKPT or read in from an external file. It is currently capable of performing Mulliken population analysis, computing electric multipole moments through octopole, electrostatic properties at atomic centers (electrostatic potential, electric field, electric field gradient, electron and spin density, dipolar anisotropic contribution to the hyperfine coupling constants), electron and spin density, electron and spin density gradient, Laplacian of electron and spin densities, electrostatic potential over an arbitrary two-dimensional (planar) rectangular grid, and molecular orbitals values over an arbitrary three-dimensional rectangular grid. Miscellaneous capabilities include computation of the relativistic first-order one-electron corrections to the energy (mass-velocity and Darwin terms), construction of natural molecular orbitals from one-particle density read from an external file (NOs can be written to PSIF_CHKPT) and computation of spatial extents - expectation values of X^2, Y^2, Z^2, and R^2 operators - of total electron density and of individual MOs (if READ_OPDM = false) or natural (if READ_OPDM = true) orbitals (MPMAX must be set to a value greater than 1 for computing these entities). Spatial extents should be used cautiously, since they depend on the reference point.

REFERENCES

Mulliken population analysis

1.: Electronic Population Analysis on LCAO-MO Molecular Wave Functions. R. S. Mulliken, J. Chem. Phys. 23, 1833 (1955), ibid. 23, 1841 (1955), ibid. 36, 3428 (1962).

Recurrence relations for one-electron integrals over Cartesian Gaussian functions.

1.: Efficient recursive computation of molecular integrals over Cartesian Gaussian functions. S. Obara and A. Saika, J. Phys. Chem. 84, 3963 (1986).

Fundamental physical constants and conversion factors.

1.: CRC Handbook of chemistry and physics. Edited by D. R. Lide. 73rd edition (1992-1993).

FILES REQUIRED

    input.dat        - Input file
    PSIF_CHKPT       - Checkpoint file

FILES UPDATED

    output.dat         
    dipmom.dat       -   Dipole moments
    esp.dat          -   Electrostatic potential on a 2D grid
    edens.dat        -   Electron density on a 2D grid
    edgrad.dat       -   Electron density gradient on a 2D grid
    edlapl.dat       -   Laplacian of the electron density on a 2D grid
    sdens.dat        -   Spin density on a 2D grid
    sdgrad.dat       -   Spin density gradient on a 2D grid
    sdlapl.dat       -   Laplacian of the spin density on a 2D grid
    mo.dat           -   Molecular orbital/Density values on a 3D grid
    mo.pov           -   MegaPov input file for rendering an image of mo.dat
    mo.cube          -   Molecular orbital(s) on a 3D grid in Gaussian94 Cube format 
    dens.cube        -   Electron/spin density(s) on a 3D grid in Gaussian94 Cube format

INPUT FORMAT

Most of the keywords are not necessary for routine tasks. The following keywords are valid:

WFN = boolean

Type of the wavefunction. This keyword is a "macro" that allows user to set most of the necessary keywords. The following values are recognized :

WFN = SCF - equivalent to READ_OPDM = false;

WFN = DETCI - equivalent to READ_OPDM = true, OPDM_FILE = 40, OPDM_BASIS = AO, OPDM_FORMAT = TRIANG;

WFN = CCSD - equivalent to EAD_OPDM = true, OPDM_FILE = 79, OPDM_BASIS = AO, OPDM_FORMAT = TRIANG;

WFN = QVCCD - equivalent to READ_OPDM = true, OPDM_FILE = 76, OPDM_BASIS = SO, OPDM_FORMAT = TRIANG;

READ_OPDM = boolean

This flag specifies if the one-particle density matrix to be read from disk. Default is false.

OPDM_FILE = integer

Specifies one-particle density matrix file number. Default is 40 (master file). To provide backward compatibility with the earlier PSI property packages (proper, ciprop, ccprop) special format of the density file is assumed when OPDM_FILE = 40 (computing properties from CI density - ciprop compatibility mode) and OPDM_FILE = 79 (computing properties from CC density - ccprop compatibility mode). As of now, in generic case onepdm must be written in the very beginning of the file. In the future PSI will have a standard onepdm file.

OPDM_BASIS = string

This option may not exist in the future. As of February 1st, 1998, a standard for the onepdm file format has not been set. This keyword should be set to either "SO" (read in onepdm matrix in SO basis) or "AO" (in AO basis). Default is "SO".

OPDM_FORMAT = string

This option may not exist in the future. This keyword should be set to either "TRIANG" (read in onepdm matrix in lower triangular form) of "SQUARE" (in square form). Default is "TRIANG"

ASYMM_OPDM = boolean

This flag specifies whether one-particle density matrix has to be symmetrized. Must be set to true if generic non-symmetric onepdm to be read (for example, from a coupled-cluster program). This keyword is for code development only. Existing PSI CC codes now in use produce symmetric onepdm, therefore there is no need to use this keyword. Default is false.

ROOT = integer

This specifies which root to do the excited state analysis for. The appropriate one particle density matrix will be read from disk. Currently implemented for DETCI and DETCAS wavefunctions.

MPMAX = integer

This integer between 1 and 3 specifies the highest electric multipole moment to be computed.

MPMAX = 1 - only electric dipole moment will be computed (default);

MPMAX = 2 - electric dipole and quadrupole moments will be computed; MPMAX = 3 - electric dipole, quadrupole, and octopole moments will be computed.

MP_REF integer

This parameter specifies the reference point for the electric multipole moments calculation.

MP_REF = 0 (default) or 1 - the center of mass;

MP_REF = 2 - the origin of the space coordinate system;

MP_REF = 3 - the center of electronic charge;

MP_REF = 4 - the center of nuclear charge;

MP_REF = 5 - the center of net charge.

CAUTION : According to classical electrodynamics, the electric 2^(n+1)-pole moment is independent of the reference point only if the electric 2^(n)-pole moment is vanishing. It means that the dipole moment will depend on the reference point if the total charge of the system is non-zero. By analogy, electric quadrupole moment will depend on the reference point if the system possesses non-zero electric dipole moment, etc.

MP_REF_XYZ = real_vector

This vector specifies the coordinates of the reference point. If this keyword is present in the input MP_REF keyword will be disregarded.

NUC_ESP = boolean

This flag specifies if electrostatic properties will be computed at the nuclei. Current list includes electrostatic potential, electric field, electric field gradient, electron and spin density, and anisotropic constribution to the hyperfine coupling constants (the latter two require setting SPIN_PROP to true). Default is true.

GRID = integer

Specifies type of property to be evaluated over a grid.

GRID = 0 (default) - compute nothing;

GRID = 1 - electrostatic potential on a two-dimensional grid;

GRID = 2 - electron density (spin density if SPIN_PROP is set to true) on a two-dimensional grid;

GRID = 3 - electron density gradient (spin density gradient if SPIN_PROP is set to true) on a two-dimensional grid;

GRID = 4 - Laplacian of the electron density (Laplacian of the spin density if SPIN_PROP is set to true) on a two-dimensional grid. According to the convention used in the field, what actually gets plotted are the Laplacians taken with negative sign.

GRID = 5 - values of molecular orbitals on a three-dimensional grid.

GRID = 6 - values of the electron density (spin density gradient if SPIN_PROP is set to true) on a three-dimensional grid.

GRID_FORMAT = string

Specifies in which format the grid output will be produced. Currently, PLOTMTV (default for 2-d grids), MEGAPOVPLUS (available for 3-d grids), and GAUSSCUBE(default for 3-d grids) are supported.

MO_TO_PLOT = vector

Specifies indices of the molecular orbitals to be computed on the 3-d grid. Indices can be specified as:

unsigned integer - index in Pitzer ordering (ordered accoring to irreps, not eigenvalues). Ranges from 1 to the number of MOs.

signed integer - index with respect to Fermi level. +1 means LUMO, +2 means second lowest virtual orbital, -1 means HOMO, etc.

All indices have to be either unsigned or signed, you can't mix and match, or you will get unpredictable results. Default is to compute HOMO and LUMO.

GRID_ORIGIN = real_vector

Specifies the origin of the grid. A rectangular grid box which envelops the entire molecule will be computed automatically if GRID_ORIGIN is missing, however, there is no default for 2-d grids.

GRID_UNIT_X = real_vector

This vector specifies the direction of the first (x) side of the grid. It doesn't have have to be of unit length. There is no default for 2-d grids.

GRID_UNIT_Y = real_vector

The same for the second (y) side. It doesn't have to be of unit length or even orthogonal to GRID_UNIT_X. There is no default for 2-d grids.

GRID_XY0 = real_2d_vector

Specifies the coordinates of the lower left corner of the grid rectangle in the 2D coordinate system defined by GRID_ORIGIN, GRID_UNIT_X, and GRID_UNIT_Y. There is no default.

GRID_XY1 = real_2d_vector

Specifies the coordinates of the upper right corner of the grid rectangle in the 2D coordinate system defined by GRID_ORIGIN, GRID_UNIT_X, and GRID_UNIT_Y. There is no default.

GRID_XYZ0 = real_3d_vector

Specifies the coordinates of the far lower left corner of the grid box in the 3D coordinate system defined by GRID_ORIGIN, GRID_UNIT_X, GRID_UNIT_Y, and the cross-product of the latter two. There is no default.

GRID_XYZ1 = real_3d_vector

Specifies the coordinates of the near upper right corner of the grid box in the 3D coordinate system defined by GRID_ORIGIN, GRID_UNIT_X, GRID_UNIT_Y, and the cross-product of the latter two. There is no default.

NIX = integer

The number of grid point along x direction. This parameter has to be greater than 1. Default is 20.

NIY = integer

The same as NIX for y direction. Default is 20.

NIZ = integer

The same as NIX for z direction. Default is 20.

GRID_ZMIN = double

Lower limit on displayed z-values for contour plots of electron density and its Laplacian. Default is 0.0

GRID_ZMAX = double

Upper limit on displayed z-values for contour plots of electron density and its Laplacian. Default is 3.0

EDGRAD_LOGSCALE = integer

Controls logarithmic scaling of the produced electron density gradient plot. Turns the scaling off if set to zero, otherwise the higher value - the stronger the gradient field will be scaled. Recommended value (default) is 5.

SPIN_PROP = boolean

Flag for computing spin properties (Mulliken population analysis of alpha and beta densities, spin densities and anisotropic contributions to the hyperfine coupling constants at atomic centers). Default is false.

PRINT = integer

This is the most important keyword - it determines amount of information printed. The following values are currently used :

PRINT = 0 - quiet mode - print out essential results only - "compact" results of Mulliken population analysis, electric multipole moments, and electrostatic properties;

PRINT = 1 (default) - all of the above plus list of tasks to be performed and list of caculation parameters;

PRINT = 2 - all of the above plus Mulliken AO population matrix and electronic and nuclear components of electric dipole moment;

PRINT = 3 - all of the above plus density matrix in AO basis and dipole moment integrals in AO (and SO) basis;

PRINT = 4 - all of the above plus basis set information, natural orbitals in terms of symmetry orbitals, overlap matrix;

PRINT >= 5 - all of the above plus coupling coefficient vectors, an occupation vector, and a modified Z-vector in MO basis.

PRINT_NOS = boolean

If WRTNOS = TRUE and this option is also TRUE, the natural orbitals will be printed to output before they are written to the checkpoint file.

WRTNOS = boolean

If TRUE, the natural orbitals will be written to the checkpoint file.

GRID OUTPUT AND PLOTTING

Currently, oeprop produces output of two-dimensional grids ready for plotting with a program PLOTMTV version 1.3.2. The program is written by Kenny Toh ([email protected]), software developer for the Technology CAD Department, Intel Corp, Santa Clara. It is a freeware package, and can be downloaded off the Internet.