A Users Guide to ZINDO  a Comprehensive SemiEmpirical SCF/CI Package.
This is one of the most versatile SCF/CI programs available and as
a consequence, it has a great number of options. It can perform PPP,
EHT, IEHT, CNDO and INDO type calculations. This handout is a
(hopefully) straightforward guide to using this program. If you have
any suggestions on how to make this handout more understandable, please
feel free to contact any of the authors listed below.
Closed shell CNDO program
M. Zerner and C. Warren Uppsala 1969
Closed shell INDO program
M.C. Zerner and J.E. Ridley Guelph 1972
Open shell UHF INDO program
M.C. Zerner and A.D. Bacon Guelph 1975
Transition metal parameterization
M.C. Zerner and R.F. Kirchner Stanford 1977
Open shell restricted HartreeFock Guelph 1980
M.C Zerner and W. D. Edwards
Rumer diagram CI 1980
M.C. Zerner , J. McKelvey and W.D. Edwards Guelph
Taken in part from D.D. Shillady VCU
Localization 1984
J.C. Culberson and M.C. Zerner Florida
Taken in part from W. Luken and J.C. Culberson Duke
Geometry Optimization 1985
J.D. Head, M.C. Zerner, B. Weiner Guelph
and Florida
Fragment Orbitals, Namelist Input 1986
A.D. Cameron, M.C.Zerner Florida
Spin Orbit Treatment
M. Kotzian, N. Roesch and M C. Zerner Munich 1991
Rumer Treatment
Polarizabilities
Bill Parkinson, Jenwei Yu and M. C. Zerner Florida 19861996
SpinOrbit treatment
M.Kotzian, N.Roesch,R.Pitzer,M.C.Zerner Munich
Double group treatment and Florida 1989
Projected UHF Florida 1995
M. Cory and M. C. Zerner
Solvent Effects Tartu and Florida 19911996
Toomas Tamm, Mati Karelson and M.C. Zerner

 Contents 

1. References
2. Introduction
3. Section I : Closed shell ground state scf
Section II : Open shell (uhf) scf
Section III : Configuration Interaction
Section IV : Geometry optimization
Section V : Electron assignment
Section VI : Calculation of polarizabilities
Section VII : Spinorbit calculation
Section VIII: Power users
 Print options
 Vectors
 Electrostatic potentials
 Triplet parametrization
 Configuration mixing for metals
 Point charges
 Polarizabilities
 Partial or projected density of states
 Self consistent reaction field
 Resonance Integrals
 Localization
 Fragment orbitals
 Memory Management
 Historical switches
 Self Consistent Field Convergence
4. Some program arrays
5. File structure
6. Control words and data blocks


 References: 

Pople,Santry,Segal.............J. Chem. Phys., 43, S129,(1965)
Pople,Segal....................J. Chem. Phys., 43, S136,(1965)
Pople,Segal....................J. Chem. Phys., 44, 3289,(1966)
Santry,Segal ...............J. Chem. Phys., 47, 158,(1967)
Santry.........................J. Amer. Chem. Soc., 90,3309,(1968)
Ridley,Zerner ................Theoret. Chim. Acta, 32,111,(1973)
Ridley,Zerner .................Theoret. Chim. Acta, 42,223(1976)
Bacon,Zerner .................Theoret. Chim. Acta, 53,21 (1979)
Zerner,Loew,Kirchner,MuellerWesterhoff
.................J. Amer. Chem. Soc., 102,589(1980)
Head,Zerner .................Chem. Phys. Lett., 122,264(1985)
Head,Zerner .................Chem. Phys. Lett., 131,359(1986)
Anderson,Edwards,Zerner........Inorg. Chem., 25,2728(1986)
Edwards,Zerner.................Theoret. Chim. Acta, 72,347(1987)
Kotzian, Roesch,Zerner.........Theoret. Chim. Acta, 81,201(1992)
.........Intern. J. Quant. Chem, S25,545(1991)

 INTRODUCTION: 

The ZINDO program has, at its heart, two different semiempirical
procedures: a method for calculating spectroscopic properties
(spectroscopic gammas) and a method for calculating geometries
(theoretical gammas). Each has a different set of options, and
consequently it is important that you decide what sort of calculation
you want to perform. For example, if you want charge distributions,
dipole moments or orbital energies, you should use spectroscopic gammas.
If you want relative energies of ground state molecules, you should use
theoretical gammas. Geometry optimization MUST be done with theoretical
gammas. Configuration Interaction SHOULD be done with spectroscopic
gammas in order to obtain the best electronic spectroscopic predictions.
In all cases, the input file will consist of at least 3 sections:
$TITLEI, $CONTRL and $DATAIN. The data fields for the $CONTRL
section is free format, namelist type input. This means that the
value of variables, switches and arrays needed to run the program can be
read in, by name, in random order and in a nearly arbitrary format. All
that is required is a valid name, followed by an equal sign (may be omitted
for backwards compatibility, but such omission may not be supported
in the future), then followed by one or more values. Comments begin with
an ! (exclaimation point) and continue until the end of the line. Comments
may be included anywhere in the file. Some data blocks require FORMATted
data in a specific order, but the sections themselves may appear in any
order. The section names must contain a $ sign in column 2, immediately
followed by the name of the section (upper or lower case). Each section
ends with a similar line containing $END .
Arrays can be assigned on elementbyelement basis, e.g. A(1)=4
a(4)=10 a(2)=9, or in longer sections. The index of the first element
must always be provided: a(1)=4 9 0 10 . If the first method is used,
the skipped index numbers will remain at default values.
Some sections, $CONTRL in particular, allow the actual input to be
placed on the same line, but since this is not universal (yet). For
example, the following input would be valid:
$contrl scftyp=uhf $end
You may have several similar sections in the same file, and you will be
able to choose between them by shifting the section name out of column 2.
In the following example, the first input is ignored, but the second one is
processed:
$contrl scftyp=rhf $end
$contrl scftyp=uhf $end
The following examples illustrate the input required for a number of
common cases.

Special Comments** Set IDD2 = 0 for routine cases that scf converge
easily to save memory
Set N1 = 0 for runs with spectroscopic parameters
if no CI is done to save memory
For closedshell or doublet CI, NCONF .LE.4 is
faster if the CI can be done this way (CIS).


 Section I: Closed shell ground state scf 

$TITLEI
Makebelieve water calculation used as an illustrative example
20 June, 1988. The title can be as long as you wish.
Updated, TT, 28 May 1996.
$END
$CONTRL
SCFTYP = RHF RUNTYP = ENERGY ENTTYP = COORD UNITS = ANGS
ASYM = C2V
! ***** Base name for temporary and output files *****
ONAME = water
NEL = 8 MULT = 1 ITMAX = 20 SCFTOL = 0.000010
APX = INDO/1 INTTYP = 1
INTFA(1) = 1.000000 1.267000 0.585000 1.000000 1.000000
$END
$DATAIN
0.000000 0.000000 0.000000 8
0.000000 1.000000 1.000000 1
0.000000 1.000000 1.000000 1
$END
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Input for CLOSED SHELL GROUND STATE WATER Explained
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
$TITLEI
The title can be anything you wish and as long as you wish.
$END
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
$CONTRL section
The first five variables determine the type of calculation and
the nature of the input.
SCFTYP = RHF < a standard closed shell RHF type calculation
= UHF  an Unrestricted HF type calculation
= UHFA  a UHF calculation followed by annihilation of
the next higher spin component
= PUHF  a UHF calculation followed by projecting out
the pure S=Ms densities. An analysis of the
higher spin components, i.e. S > Ms, contributions
to the unprojected wavefunction is also performed.
TO calculate the Mulliken Population for
each component, place PUHF_MULPOP in the $OUTPUT
block. The number of multiplets that are considered
can be limited by PUHFST in the $CONTROL block.
PUHFST = 0 is the default, and all components
of the UHF wavefunction are analyzed from M = Sz to
M = NEL/2, or until the weight of a multiplet is
below an internal threshold, now 10**(8). Else:
PUHFST=ABCD where ABCD is a four character integer,
it should be viewed as two contiguous I2 words.
Depending what PUHFST is yields varying results, e.g.
0103 = 103 = 0301 = 301 would report the energies
for the first three spincomponents. (There is no
case for which the weights of all components are not
reported.) In addition, if PUHF_MULPOP is also set
then the Mulliken populations of the three spin states will also be reported. If PHUFST=2=02 is given, then
only the energy of the second component is given, and
only it's Mulliken population analysis if PUHF_MULPOP
is also set. If PUHFST=1 then only the weights are
reported, and if PUHF_MULPOP is set then the mull.
pop.s for all reported weights.
See PUHF below for other options.
= ROHF  a spin restricted open shell RHF calculation.
This option requires additional input!
= SUHF  This is a UHF calculation with laGrange constraint
that the multiplicity be MULT, M > Sz. The
Langrange constraint LAMBDA = X may be given in the
$CONTRL block. The default is LAMBDA = 0.2 . See
N. Handy. For large enough LAMBDA this approaches
the ROHF solution, but convergence is slow for
large LAMBDA.
= GVB  General valence bond a la John Cullen.
If GVB, then in the control block is
needed GVBFN(1) = N1, N2, N3
with N1 = number of canonical mo's doubly
occupied, N2= number of gvb pairs, N3 = number of
open shell orbitals. The default is
GVBFN(1) = 0, (NELMULT+1)/2, MULT1
Note that GVBFN(3) = N3 must equal MULT1
GVBFN(4) = 2 scales the gvb iterations with the
inverse Hessian in the line search. This is
recommended for openshell calcs for convergence
' and MAY always be better.
RUNTYP = ENERGY < a simple SCF calculation
= GEOM  a geometry optimization calculation
= CI  a configuration interaction calculation
= CIF  if not a Rumer CI, see below, this
keyword will save considerable memory.
= RPA  an RPA calculation will be set up as a
separate job. See ZINDORPA.HLP. The
operation requires pgm rpago with input
ONAME_rpa.DAT and generates output
ONAME_rpa.out, i.e., rpago < ONAME_rpa.DAT.
= SELFEN  Calculates the ionisation spectrum from
the electron propagator Formalism. Several
files are created ONAME_sel.xxx. Although
these are kept, they can be deleted unless
a restart option is used.
= CIF  Calculation is to be followed by a general
configuration interaction using fullci pgm
Three files are created, oname_CIF.DAT which
is in clear text. Then ther is oname_oneel.file
and oname_twoel.file that contain the one and
two electron integrals blocked as given in
oname_CIF.DAT
ENTTYP = COORD < atomic positions in cartesian coordinates
= ZMAT < atomic positions in internal coordinates
= PDB < protein data base input
= CAR < MSI CAR input file
= SCOORD < atomic symbol, X, Y, Z and, perhaps, TAB
UNITS = ANGS < cartesian coordinates in units of angstrom
= BOHR  cartesian coordinates in units of bohrs
= CANGS < cartesian coordinates in units of angstrom
center the coordinates, and rotate to principle
moments of inertia
= CBOHR < cartesian coordinates in units of bohrs
center the coordinates, and rotate to principle
moments of inertia
ASYM = C2V < Point group is C2V. Other possibilities include
= C2 the following Abelian groups. In all cases, the
= CI C2 axis must lie along the z axis of the molecul
= CH
= C2H
= D2
= D2H
NOTE. If in the $CONTRL block DETSYM = 1, the program will determine
molecular symmetry
Output File Name
ONAME = ANYNAME  This 8 character name will be used to generate
filenames used by the program.
Switches
NAT = 3  Number of atoms (usually omitted). This must be equal
to the actual number of atoms in $DATAIN.
NEL = 8  Number of valence electrons
(or CHARGE = 0.0)  will generate 8 electrons
MULT = 1  Multiplicity (number of unpaired electrons + 1)
ITMAX = 20  Maximum number of SCF cycles. This depends on the
size and complexity of the problem. 20 is usually
enough for most moderate size problems.
SCFTOL = 0.0001  Convergence criteria for the diagonal of the
density (bond orders). If differences between
densities from one cycle to the next are all
less than this, convergence is assumed.
Program will print the message:
SENSE LIGHT 2 IS ON
APX = EHT  Extended Huckel ITMAX > 0 iterative Extended Huckel
CNDO/1
CNDO/2
INDO/1 <<<<< (default) Method of choice (author's favorite)
INDO/2
NDDO/1
NDDO/2
PPP theory
AM1
PM3
MNDO
The following is supported for compatibility:
IAPX = 0 : Extended Huckel 1 : CNDO/1 2 : CNDO/2
3 : INDO/1 (default) 4 : INDO/2 5 : NDDO/1
6 : NDDO/2 7 : PPP 8 : XINDO
51 : MNDO 52 : AM1 53: PM3
INTTYP = 0 <<<<<< Theoretical (Fo) Gammas (GEOMETRY theory)
The higher Slater Condon Factors are semiempirical
or they are scaled calculated
= 1 <<<<<< MatagaNishimoto Gammas (SPECTROSCOPIC theory)
= 2  Not presently used
= 3  OhnoKlopman Gammas (CNDO/INDO/PPP only)
= 4  PariserParr Gammas
= 5  Modified Warshel Gammas
= 1  All Integrals abinitio
GFAC = 0  default
= 1  calculate all single excitations for a doublet ground
state and create the file ONAME_gfac.DAT. This
file is for pickup by the esr gfactor program.
For this option the DYNAL(7) = the number of active
mo's under the $CIINPU block.(same as ACTSP) Note
that N1 can be 0 or 2 for this to work properly.
= 2  as above, but calculate only those excitations needed
for the gfactor program (i>m, and m>a, where i is
doubly occupied, m is singly occupied and a is virtual.)
Interaction factors
INTFA(1) = Ssigma, Psigma, Ppi, Dsigma, Dpi, Ddelta
For spectroscopic singletrecommended for nearly ALL cases
(especially for calculations for UV/visible spectra)
INTFA(1) = 1.0 1.267 0.585 1.0 1.0 1.0
For ordinary CNDO or INDOrecommended for geometry optimisation
INTFA(1) = 1.0 1.0 1.0 1.0 1.0 1.0
For EHT (Hoffmann recommends factors of 1.75)
INTFA(1) = 1.89 1.89 1.89 1.89 1.89 1.89
For PPP
INTFA(1) = 0.585 0.585 0.585 0.585 0.585 0.585
$END  Other switches and parameters can go in this section
See power users section (section VI) for more information
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
$DATAIN section: Atomic coordinates and atomic numbers
One line for each atom or centre. Comments are allowed anywhere; they
start with an exclamantion mark and continue through end of line.
The first line which is blank (not a comment and no coordinates),
or which contains the string '$END', signals the end of input.
There are several options in which to enter the atomic information:
METHOD 1: The traditional XYZ coordinates. Free format. The first four
 fields (X,Y,Z,atomic number) are required, others are
optional, but quantum numbers and exponents must appear
in pairs as explained below:
X Y Z Atomic Number Net Valence Charge Basis Set Exponent
X Y Z  Cartesian coordinates for each atom in units specified
in variable UNITS
Atomic Number  Integer
Net Valence Charge = 4.0 for neutral carbon
= 8.0 for neutral Fe
= 0.0 for DEFAULT values (omit if default and no
exponents are input)
For PPP calculations *** INPUT IS REQUIRED ***
= 1.0 for C
= 1.0 for N (PYRIDINE)
= 2.0 for N (PYRROLE)
Basis set definition = 1 for S
= 2 for S,P
= 3 for nS,nP,nD
= 4 for nS,nP,(n1)D
= 5 for ns, np,(n1)d,(n2)f
= 6 to be continued
= 0 for DEFAULT
Exponents can be overridden (omit for defaults):
n, Exponent(S); n, Exponent(P); n, Exponent(D)
n is the principle quantum number for the orbital
EXAMPLES:

Oxygen atom, all defaults
0.000000 0.000000 0.000000 8
This is the same as:
0.000000 0.000000 0.000000 8 6.000000 2
Carbon atom, reset exponents
0.000000 0.000000 0.000000 6 4.000000 2 2 1.7000 2 1.7000
Point charge
0.000000 0.000000 0.000000 0 0.500000
$END
METHOD 2: Redefining atomic types using TABS:

Input X, Y, Z, atomic number as above, then add END or TABxx to the end
of the line. IF END, all information is defaulted as in the very first
example above. If TAB, you will need to add a $TABS input block, which
defines atomic types, distinhuished by the two characters immediately
following 'TAB':
1.11111111 1.23456789 2.0 6 END
In this case all other information about the Carbon atom (=6) is
defaulted. Point charges cannot be inputted this way.
1.11111111 1.23456789 2.0 6 TABC1
if this is the case than there must be a $TABS section defining 'C1':
$TABS
C1 ZCORE = 4.00 NTYP = 3 NS = 2 ZETAS = 1.625 NP = 2 ZETAP = 1.595 #
ND = 2 ZETAD = 1.595 IPD = 2.0
COULS = (in eV) COULP = (in eV) COULD = (in eV) and COULF = (in eV)
C2 ..........
$END
In the above all atoms with a tab of C1 will have the specified
attributes:
ZCORE is the core charge, and together with the atomic number = 0
in the $DATAIN section could specify a point charge.
The value of ZCORE must be given if TABS are used <=====**
NTYP as above. In this case a s,p,d basis set, with NS=NP=ND=2
A d polarization function has been added with exponent 1.595.
An IP for the D orbital IPD has been added, as there is no
default value for a 3d function on carbon.
USS, UPP, UDD, UFF allows one to enter core integrals directly
bypassing the use of the IPS, etc., in the calculation of USS,etc.
These values are in Hartrees and are generally NEGATIVE.
A # marks a continuation line follows for this tab.
METHOD 3: Zmatrix input. If in the control block there is ENTYP = ZMAT
 zmatrix input can be used. The format is identical to
the one used in MOPAC and AMPAC, except that default
connectivity for atoms 1, 2, and 3 is not supported.
You will need to specify the full connectivity of all
atoms.
Example:
$TITLEI
Fe2fdx  Zmatrix model  PDBXray of Cyanobacteria Amobaena
$END
$CONTRL
SCFTYP = ROHF RUNTYP = ENERGY
NEL = 106
APX = INDO/1 INTTYP = 1 MULT = 11
UNITS = ANGS
SCFTOL = 0.000001 ITMAX= 10 IPRINT = 0
ONAME = fe_zmat
.
.
.
ENTTYP= ZMAT
$END
$DATAIN
X 0.000000 0 0.000000 0 0.000000 0 0 0 0
X 2.000000 0 0.000000 0 0.000000 0 1 0 0
Fe 1.350000 0 90.000000 0 0.000000 0 1 2 0
Fe 1.350000 0 90.000000 0 180.000000 0 1 2 3
S 2.190000 0 51.943492 0 90.000000 0 4 1 2
S 2.190000 0 51.943492 0 270.000000 0 4 1 2
S 2.290000 0 106.230000 0 90.000000 0 3 1 6
S 2.290000 0 106.230000 0 270.000000 0 3 1 6
S 2.290000 0 106.230000 0 90.000000 0 4 1 6
S 2.290000 0 106.230000 0 270.000000 0 4 1 6
C 1.820000 0 109.471220 1 180.000000 0 7 3 1 TABC1
.
.
.
$END
Again you may add the word TABXY, changing the defaults. For the above
example there must be a C1 in a $TABS section.
Atoms marked with X are dummy atoms, used for convenience if desired.
For atom number 'a' :
ATSYMB R(AM) M1 THETA(AMN) M2 PHI(AMNO) M3 ATOMM ATOMN ATOMO
R(AM) is the distance between atom numbers A (this atom) and M, etc.
in Angstroms unless UNITS = BOHRS in the $CONTRL section.
M1, M2 M3 are for the moment ignored. If coordinates are to be
frozen in a geometry optimisation thay must be frozen in the $INTCOR
section.
METHOD 4: PDB input:
$CONTRL
ENTYP = PDB
.
.
$END
This is strictly formatted:
Column: Content:
1 16 'ATOM' or 'HETATM'
2 711 Atom serial number (may have gaps)
3 1316 Atom name in IUPAC standard format
4 17 Alternate location indicator(chain) A,B,C,etc.
5 1820 Residue name in IUPAC standard format
6 22 RES LETTER
7 2326 Residue sequence number (order as below)
8 27 LET2
9 2830 Code for insertions of residues (ie 66A & 66B)
10 3138 X coordinate
11 3946 Y coordinate
12 4754 Z coordinate
13 5560 Occupancy
14 6166 Temperature Factor
15 6870 Footnote number
SAMPLE INPUT
1671b36780b2367801896745016b80
ATOM 15 SG CYS A 112 20.479 13.319 10.861 1.00 18.56
ATOM 16 1HB CYS A 112 19.270 11.709 9.494 1.00 0.00
ATOM 17 2HB CYS A 112 19.072 13.366 8.877 1.00 0.00
ATOM 37 CU CU A 130 0.739 0.425 3.680 1.00 0.00
NOTE THAT CHEMICAL SYBMOLS WILL BE RECOGNIZED. The above will be
interpreted as Cu not a carbon atom on unit U
If there is a chance for confusion, as CUB not Cu but C on a fragment
UB, a warning will be issued. These warnings are important and very
one of them needs to be checked*******
If HETATM is used this will force the first two symbols to represent
the atom
METHOD 5 MSI CAR FILE
$CONTRL
ENTYP = PDB
.
.
$END
$DATAIN
CL44 2.776314625 0.955053888 1.178162957 Z1 0 Cl Cl 0.233
CL45 0.838342261 2.158296441 1.518279685 Z1 0 Cl Cl 0.205
CL47 2.737601206 1.233552863 1.574915779 Z1 0 Cl Cl 0.233
C53 1.316782944 0.609552934 0.798633465 Z1 0 c C 0.133
C531 0.272834566 1.048458177 1.626236132 Z1 0 c C 0.133
H5311 0.098288626 0.286630199 2.348989090 Z1 0 h H 0.062
H5312 0.102758098 2.021418857 1.510777376 Z1 0 h H 0.062
H531 1.660075960 0.393999938 0.947691724 Z1 0 h H 0.062
H532 1.650766299 1.306030235 0.037701254 Z1 0 h H 0.062
CL55 0.881809615 2.383359128 1.247763094 Z1 0 Cl Cl 0.205
TI22 0.771202504 0.130491180 0.141784459 Z1 0 Ti046 Ti 0.769
C56 0.672848279 0.948648042 1.452684755 Z1 0 c C 0.277
H561 1.289017484 0.149035105 1.886452730 Z1 0 h H 0.042
H562 1.262906973 1.682075104 0.880662876 Z1 0 h H 0.042
H563 0.081084428 1.464279370 2.241174807 Z1 0 h H 0.042
$END
No format is required  the lines are parsed.
The information used in the CAR input is
 X Y Z    Symbol 
Method 6: Simply (ENTTYP = SCOORD)
atomic symbol X Y Z
C 0.000 0.000 1.342000
or atomic symbol X Y Z TAB
Fe 0.000 0.000 2.000 TABFE
X Y Z can be in Angstroms (default or UNITS = ANGS), or in Bohrs
(UNIT = BOHRS )

 Section II: Open shell (uhf) scf 

$TITLEI
Water Positive Ion. An openshell UHF calculation
approximating the doublet state.
20 June, 1988. The title can be as long as you wish.
$END
$CONTRL
SCFTYP = UHF RUNTYP = ENERGY ENTTYP = COORD UNITS = ANGS
ASYM = C2V
! ***** Output file name *****
ONAME = waterp
NEL = 7 MULT = 2 ITMAX = 20 SCFTOL = 0.000010
APX = INDO/1 INTTYP = 1
INTFA(1) = 1.000000 1.267000 0.585000 1.000000 1.000000
$END
$DATAIN
0.000000 0.000000 0.000000 8
0.000000 1.000000 1.000000 1
0.000000 1.000000 1.000000 1
$END
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Input for OPEN SHELL UNRESTRICTED HARTREE FOCK SCF Explained
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
For open shell UHF calculations, the only changes in $CONTRL are:
RUNTYP = UHF
MULT = 2
NEL = 17
making the molecule open shell. No other changes are needed for UHF
calculations.
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Data for an OPEN SHELL RESTRICTED HARTREE FOCK SCF calculation
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
$TITLEI
Makebelieve open shell water doublet cation; an illustrative
example. 5 July, 1988.
$END
$CONTRL
SCFTYP = ROHF
RUNTYP = ENERGY ENTTYP = COORD UNITS = ANGS
ASYM = C2V
NOP = 1
NDT = 1
FOP(1) = 6.0 1.0
! ***** Output file name *****
ONAME = water
NEL = 7
MULT = 2
ITMAX = 20 SCFTOL = 0.000010
INTTYP = 1 APX = INDO/1
INTFA(1) = 1.000000 1.267000 0.585000 1.000000 1.000000
$END
$DATAIN
0.000000 0.000000 0.000000 8
0.000000 1.000000 1.000000 1
0.000000 1.000000 1.000000 1
$END
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Input for OPEN SHELL RESTRICTED HARTREE FOCK SCF Explained
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
For open shell RHF calculations additional data is required
SCFTYP = ROHF
MULT = 2
NEL = 7
NOP = 1  The number of open shell orbitals
NDT = 1  The class of the open shell state
see below for classes available
FOP(1) = 6.0 1.0  The number of electrons in each shell
First number is number of electrons in
closed shell. For default cases, second
number is total number of electrons in
all open shells.
Default cases: Let FF = Number of closed shell electrons
A) One electron, one open orbital  DOUBLET
NOP = 1
NDT = 1
FOP(1) = FF.000000 1.000000
B) N electrons, N open orbitals  Highest spin case
(ie 4 electrons, 4 open orbitals  QUINTET)
NOP = 4
NDT = 1
FOP(1) = FF.000000 4.000000
C) One electron, N open orbitals (ie N=4  DOUBLET)
NOP = 4
NDT = 1
FOP(1) = FF.000000 1.000000
D) Two orbitals, one electron  DOUBLET See C
Two orbitals, two electrons  TRIPLET See B
Two orbitals, two electrons  SINGLET Two types
One electron in each orbital
NOP = 2
NDT = 1
FOP(1) = FF.000000 2.000000
Two electrons in one orbital, none in the other XaXb+YaYb
NOP = 2
NDT = 2
FOP(1) = FF.000000 2.000000
Two electrons in one orbital, none in the other XaXbYaYb
NOP = 2
NDT = 3
FOP(1) = FF.000000 2.000000
XaXb+YaYb and XaXbYaYb are degenerate in certain groups
E) Three electrons, two open orbitals  DOUBLET
NOP = 2
NDT = 1
FOP(1) = FF.000000 3.000000
F) Averaged multiplet, M open orbitals with N electrons
This is the CAHF option.
NOP = M
NDT = 9
FOP(1) = FF.000000 N
From orbitals obtained from the Configuration Averaged Hartree
Fock calculations, a Rumer CI (see below) can be performed to
project out states of a specific multiplicity or symmetry. In
such a case it is best to store the vectors (VEC=9) and restart
the CAHF calculation and Rumer CI after examining the nature of
the orbitals obtained.
G) Average energy for a given N electrons in M orbital of given S
This is the MAHF = multiplet averaged hartreefock
NOP = M
NDT = 8
FOP(1) = FF.0000 N
This option is best for the higher multiplicities, but has
often convergence problems for the lower multiplets
Open shell RHF  General case
Requires input of vector coupling coefficients
MIM(2) = 3 2  Total number of orbitals per open shell
AR(1) = 1.0 1.0 0.0  The alpha vector coupling coefficients
read in as a linearized matrix
NOTE alpha = 1  A
A is from the energy expression
BR(1) = 1.0 1.0 2.0  The beta vector coupling matrix
NOTE beta = 1  B
B is from the energy expression.
EXAMPLE  3 degenerate orbitals, 4 electrons, TRIPLET
NOP = 3
NDT = 0
FOP(1) = FF.000000 4.000000
MIM(2) = 3 0
AR(1) = 0.062500
BR(1) = 0.125000

 Section III: Configuration interaction 

$TITLEI
HC=CCN For G. Diercksen Max Planck Institute for Astrophysics
$END
$CONTRL
SCFTYP = RHF RUNTYP = CI ENTTYP = COORD UNITS = ANGS
ASYM = C2V
INTTYP = 1 APX = INDO/1 NEL = 18
MULT = 1 ITMAX = 20
INTFA(1) = 1.000000 1.267000 0.585000 1.000000 1.000000
! ***** C.I. size information *****
CISIZE=100 ACTSPC=17
! ***** Output file name *****
ONAME = HCCCN
$END
$DATAIN
0.000000 0.000000 2.568000 7
0.000000 0.000000 1.378000 6
0.000000 0.000000 0.000000 6
0.000000 0.000000 1.205000 6
0.000000 0.000000 2.263000 1
$END
$CIINPU
2 1 10 1 0 0 0 1 1 2 10
60000.00 0.000000
0 0 0 0 0 0 0 0 0
1 9 9
21 1 9 10 17
$END
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Input for CONFIGURATION INTERACTION Explained
::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Changes to the $CONTRL section consist of the following:
RUNTYP = CI
CISIZE=101 ACTSPC=17
CISIZE = 101  The CI is expected to generate no more than 100
configurations in EACH irrep of the group (C2V)
plus the generating state
If no value for CISIZE is given in the input, an
estimate for this value and for ACTSPC is made.
This does not work for SPNORB.GT.0.
If this estimate is incorrect, you can override
it by specifying CISIZE and ACTSPC in the input.
ACTSPC = 17  The last active orbital in this CI is the
(virtual) orbital 17
For special CI use value of (MOMAXMOMIN+1)
For general CI use value of (MOMAX)
For MP/2 use value of (MOMAX)
Some old data sets may contain an array DYNAL. The 6th element
of DYNAL is the same as CISIZE, and the 7th element is ACTSPC. Other
elements are ignored.
An additional section, $CIINPU has been added which details the type
of configuration interaction calculation to be done
Briefly, the CI input consists of:
(A) Initial CI switches (FORMAT 11I5)
(B) Thresholds for reducing the number of configurations (2F10.6)
(C) Point group and irrep specification (9I5)
(D) Symmetry information (OMIT if symmetry is used in SCF) (40I2)
(E) Orbital occupancy of reference determinant(s) (16I5)
(F) Excitation information (16I5)
*** NOTE *** THIS IS FORMATTED INPUT!!!
(A) Initial CI switches; eleven switches in 11I5 FORMAT
This is known below as the "CI LINE"
N1 N2 N3 N4 N5 N6 N7 N8 N9 N10 N11
N1 = 1  Calculate only the diagonal elements of CI matrix
= 2  CI for SINGLETS from closed shell reference
or SIngles only for a ground state doublet
= 3  CI for TRIPLETS from closed shell reference
= 4  CI for TRIPLETS and SINGLETS
or SIngles only for a ground state doublet for the
DOUBLETS and QUARTETS
= 5  General CI using Rumer diagrams, any multiplicity
Use for ROHF SCF ground states
= 9  GUGA CI. This generates input for the DRT program,
then TRN program then SDG program, see appropriate
help files.
=20  MP/2 perturbation theory
=21  NesbetEpstein second order perturbation theory
=50  Double Group CI (with SpinOrbit)
N2 = 1  The number of reference determinants to be used
to generate the excited configurations.
If this is an MP/2 calculation on a UHF wavefunctio
then N2 = 2, one for alpha and one for beta orbital
Only for Rumer diagram CI (N1 = 5) can there be
more than ONE reference determinant.
N3 = 0  The number of roots of the CI matrix to calculate
DEFAULT is the 10 lowest
N4 = 1  The multiplicity of the CI. It need not be the
the same as that of the reference SCF
Natural Orbital SectionAvailable only for Rumer CI (N1 = 5)
N5 = N  Perform Natural Orbital Optimization on state N
DEFAULT = 1. This option works only if N1 = 5
N6 = M  Average the first M densities for NO'S
DEFAULT=1.
N7 = 0  Do not perform natural orbital optimization
= 1  Perform noniterative NO analysis
= N  Perform N iterations in the NO analysis
This option works only for the Rumer diagram CI
Transition Moment Section
N8 = 0  Do not calculate transition moments
> 0  Calculate the transition moments between
N9 = states N8 through N9 into N10 through N11
N10 = DEFAULT is N9 = 1 N10 = 2 N11 = 10
N11 =
(B) Criteria to reduce the number of configurations included in the CI
ECUT COMP (FORMAT 2F10.6)
ECUT = 0 or a negative number for normal calculation
= Positive threshold value (in cm1) below which all
configurations are included. For example a value of
60000. means any configuration with a DIAGONAL CI
element less than 60,000 cm1 is included
COMP = A value indicating the extent of interaction to be included
For example a value of 500.0 (cm1) means that ANY generated
configuration having an OFF DIAGONAL CI element larger than
500 cm1 with any of the states accepted by ECUT (above) is
included. A value of 0.0 includes all configurations
(C) Point group symmetry information (FORMAT 9I5)
ISYM IREP IREP IREP IREP IREP IREP IREP IREP
ISYM = N  the order of the Abelian group followed by UP TO
eight IREPPs to be considered, in turn, for the CI.
IF NO SYMMETRY IS TO BE CONSIDERED LEAVE THIS LINE BLANK. or
ISYM = 0.
If symmetry is assigned, eg., ASYM = C2V, or DETSYM = 1, and the program
determines symmetry, and if all the CI's for the different irreducible
representations are the same, then a single master codetor and generating
line can be used if you do not specify a different master codetor and
generating line. This also works if DETSYM = 1 and ISYM = 1
EXAMPLES:
4 1 2 3 4  do all four irreps for C2V
2 2 1  do irreps 2 and then 1 for C2
8 6 7 8  do the last 3 irreps for D2H
Character tables for Abelian groups
ISYM=2: IRREP CI C2 CS
1 1 1 Ag A A'
2 1 1 Au B A''
ISYM=4 IRREP C2H C2V V=D2
1 1 1 1 1 Ag A1(Z) A1
2 1 1 1 1 Au(Z) A2 B1(Z)
3 1 1 1 1 Bu(X/Y) B1(X) B2(Y)
4 1 1 1 1 Bg B2(Y) B3(X)
ISYM=8 IRREP D2H
1 1 1 1 1 1 1 1 1 Ag
2 1 1 1 1 1 1 1 1 Au
3 1 1 1 1 1 1 1 1 B1g(XY)
4 1 1 1 1 1 1 1 1 B1u(Z)
5 1 1 1 1 1 1 1 1 B2g(XZ)
6 1 1 1 1 1 1 1 1 B2u(Y)
7 1 1 1 1 1 1 1 1 B3u(X)
8 1 1 1 1 1 1 1 1 B3g(YZ)
(D) Orbital symmetry specifications >>>>> USUALLY OMITTED <<<<<
CI Specification(s)
*** NOTE: *** If the CI is done on more than one irrep, then the CI
specification lines must be repeated for each irrep considered
THERE MUST BE A REFERENCE DETERMINANT AND GENERATING LINES FOR
EACH IRREP CONSIDERED.
(E) First line  The number of generating lines and the reference
determinant in FORMAT nI5
EXAMPLE:
2 4 4 5
Means there are two lines specifying excitations from the
reference determinant:
(1a,1b,2a,2b,...4a,4b,5a)
The list must start with a closed shell alpha/beta pair
followed by open shell coupled alpha/beta pairs and
ending with unpaired alpha orbitals.
For closed shell cases specify only the outermost occupied pair
For ordinary CI (N1 = 2) only one reference determinant is
allowed.
For Rumer CI (N1 = 5) several reference determinants
are allowed and these do NOT need to be the SCF determinant.
In addition, for many open shells it is convenient to use the
words DOC(doubly occuppied), SOC (singly occuppied) and VIR
(virtual or empty). For example,
2 4 4 5 > 2 4*DOC SOC
1 3 3 5 5 4 > 1 3*DOC SOC DOC
1 3 3 5 5 > 1 3*DOC VIR 1*DOC
NOTE: REFERENCE DETERMINANT(S) AND GENERATING LINES ARE
REQUIRED FOR EACH IRREP CALCULATED
(F) Second line(s)  Generating lines in FORMAT 16I5
COLUMNS 15 = 1 Individually specified
singly excited configurations
= 2 Individually specified
doubly excited configurations
= 21 Multiply specified
singly excited configurations
= 22 Multiply specified
doubly excited configurations
= 32 Multiply specified TYPE I
doubly excited configurations
COLUMNS 678 = Occupied orbitals > Virtual orbitals
Examples of CI specification lines
1 4 5  One singly excited configuration with electron
removed from MO 4 and placed into MO 5
(4 INTO 5)
or
21 2 4 5 6
 All possible singly excited configurations
generated by removing electrons from MO 2
through 4 and placing them into MO 5 to 6,
one electron at a time. (24 INTO 56)
or
This is equivalent to specifying the following
1 2 5
1 3 5
1 4 5
1 2 6
1 3 6
1 4 6
2 2 5 4 8
 One doubly excited configuration (2 INTO 5, 4 INTO 8)
2 I A J B
 I into A, J into B, requires that I .LE. J and A .LE. B
22 2 4 5 8 1 3 5 8
 Multiply specified double excitations
(24 into 58 and 13 into 58)
*** NOTE ***
All ' 2X' lines (multiconfigurations) before ' X' lines
(single configurations) ALL INDIVIDUALLY SPECIFIED CONFIGS AND
REFERENCE DETERMINANTS ARE INCLUDED, REGARDLESS OF SYMMETRY!
TRIPLES AND QUANDRUPLES BEFORE SINGLES AND DOUBLES
SPECIFIED FULL CI AFTER ALL OTHERS.
ie
23 2 5 6 9
21 1 5 6 10
22 1 5 6 10 1 5 6 10
29 4 5 6 7
*** If N1 = 5 (Rumer CI) Additional generating lines are allowed
21.... SINGLES.
22.... DOUBLES
23.... TRIPLES
24.... QUADS
29 3 5 6 10 FULL CI between orbitals 35 into 610
For the Rumer CI, the multiplicity of the CI may be different
from the multiplicity of the reference determinant. Under those
circumstances, the reference determinant used for generating the
list of configurations may not, itself, be included in the list of
configurations
*** If N1 = 20 (MP/2)
Only one reference determinant is allowed. In the case
of UHF, specify BETA symmetry, reference determinant, and
ONLY ONE GENERATING LINE. Then repeat for ALPHA set.
EXAMPLE:
(blank line)  (No cutoff or threshold)
4 1  (4 irreps, do only irrep No. 1)
1 2 4 3 1 2 3 1
 (Sym of BETA MO.'s)
1 1 2 3 4  (One Gen line, 1234 are BETA)
22 1 4 5 8 1 4 5 8
 (Multiple doubles, 14 into 58)
(now repeat for alpha spin)
1 2 4 3 1 3 2 1
 (ALPHA and BETA sym not the same)
1 1 2 3 4 5
 (ALPHA occupied in aufbau order)
22 1 5 6 8 1 5 6 8
 (Note extra alpha electron)
ADDITIONAL CONFIGURATION INTERACTION SPECIFICATIONS
21 2 4 5 8 3 4
 All excitations (24 TO 58) but only
those with MOs (24) with large AO
character of type 03 (PY) and with MOs (58)
with large AO character of type 04 (PZ)
This option does not work for singly
specified configurations (0001, etc)
1=S, 2=PX, 3=PY, 4=PZ,
5=D(Z2), 6=D(X2Y2), 7=D(XY),
8=D(XZ), 9=D(YZ).
or
22 2 4 5 8 1 3 5 8 4 4
 All doubly excited configurations
(24 TO 58, 13 TO 58)
but only when MOs 24 and 13 have
SMALL type 04 contributions and MOs
58 and 58 have SMALL 04 contribs.
or
32 1 5 6 9
 Singlet configurations, all possible double
excitations removing 2 electrons from MO 1 and
placing both in MO 6. Then 2 electrons from 1
into 7,... 2 electrons from MO 5 into MO 9
(Type I doubles 15 into 69)
42 1 6 2 8 3 7 4 10 11 12
 Double excitations
1 into 6 with 2 into 8 and
1 into 7 with 2 into 10 and
1 into 11 with 2 into 12 and
3 into 6 with 4 into 8 and
3 into 7 with 4 into 10 and
3 into 11 with 4 into 12
$END
The I420(I2) format is no longer supported.
The use of symmetry in the CI will generate a Summary table at the end
which eleminates reduntant states, and orders the states in increasing
energy. Reduntant states are those that are specified singly, or those
that are used to generate other states (The Master Codetor's). The
really belong only to their own irreps, but serve as markers for the
calculation. On occasion the CI will change the lowest energy state
to be other than that used as the marker in each irrep. If this should
happen, then the oscillator strength will not be printed for that
irrep in the summary table, as it will have been calculated from the
lowest configuration in each irrep, and not neccesarily from the
lowest energy (ground) state of the molecule.
To correct this situation, the lowest energy state of the molecule
should be specified in each irrep singly, so this state is the lowest
energy in each irrep.
***AUTOMATIC CI
There is an automatic option for generating closed shell CI
This will generate a 10 up 10 down CI.
$CIINPU
N1 N2 N3 N4 N5 N6 N7 N8 N9 N10 N11 as before
EPS ECUT as before
0 = ISYM or 1 see below
1 generate the 10 up 10 down
$END
If ISYM = 1 above, and DETSYM = 1 in the $CONTRL block, the CI will
automatically be symmetry blocked.
If N10 is a number less than N9 then the excited state spectrum between
states will also result. For example, if N8 = 1, N9 = 10, N10 = 1
and N11 = 50, all states from 1 thru 10 into 1 thru 50 will be calculated.
In addition each of the first 10 states will be analyzed and the
population analysis printed (in the ZDO basis). If bond orders are desired for each of the states N10 through N9, then CIBONDS should appear as a keyword in
the $OUTPUT section. Spin densities for excited states after CI can also
be obtained using AISO = NNMM in the $CONTRL block, where states NN to MM
will be examined. This is a four digit number as 0130 states 1 to 30.
***

 Section IV: Geometry optimization 

$TITLEI
NH3 BFGS UPDATE, April 6, 1988
$END
$CONTRL
SCFTYP = RHF ENTTYP = COORD UNITS = ANGS
RUNTYP = GEOM
INTTYP = 0 APX = INDO/1 III = 2001
NEL = 8 MULT = 1 ITMAX = 12
SCFTOL = 0.000000
! ***** Interaction factors *****
INTFA(1) = 1.0 1.0 1.0 1.0 1.0 1.0
! ***** Output file name *****
ONAME = g1
$END
$DATAIN
0.000000 0.000000 0.000000 7
1.026604 0.000000 0.411484 1
0.513302 0.889066 0.411484 1
0.513302 0.889066 0.411484 1
$END
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Input for GEOMETRY OPTIMIZATION Explained
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
The following switches should be set to run a geometry optimisation
RUNTYP = GEOM  Geometry Optimization
INTTYP = 0  Theoretical Gammas <<<<< MOST IMPORTANT
SCFTOL = 0.00  SCF states must be well converged
INTFA(1) = 1.0 1.0 1.0 1.0 1.0 1.0
In addition, a number of switches may be set in order to control the
details of the geometry optimization procedure.
The most important switch for geometry optimisation is III which is
a four digit integer made up of 4 parts: IA:IB:IC:ID
Choose a value for each of IA, IB, IC, ID and concatenate to form a
4 digit integer from these values. Enter this value for III.
For normal calculations the following switches are recommended.
III = 0000  NO GEOMETRY OPTIMISATION
= 2000  for ground state structures
= 3120  for transition state calculations
Additional possibilities (for experienced users) are:
 Search for minimum, transition state, etc.?
IA = 0 energy calculation only (No optimization)
= 1 energy and gradient calculated (No optimization)
= 2 for stable geometry
= 3 for transition state
= 4 for transition state restart (not operational)
 Type of Optimisation 
IB = 0 Newton Raphson
= 1 Augmented hessian
if transition state followed by f+f
= 2 Minimize norm squared of gradient (full treatment)
= 3 Homotopic path continuation (with IB = 2)
= 4 Mimimize norm squared of gradient (approx. treatment)
= 5 Homotopic path continuation (with IB = 4)
(** = 6 Gradient extremal for transition states. For example
III = 3620. Note you must include a target set of
coordinates even though they are not used as is needed
for all transition set calculations.**)
= 7 LTP option for transition states.
(** = 8 use NewtonRaphson with trust radius to control step
length after either Augmented hessian or gradient extremal
have found 1 negative root of the hessian.( see note below
regarding use of IP and IQ ) **)
 Type of search 
IC = 0 Updated Hessian
= 1 Updated Hessian with last iteration with analytic Hessian
= 2 Analytic Hessian
= 3 gradient only, no Hessian
= 4 Analytic Hessian calculated with first order approximation (V1)
= 5 Analytic Hessian calculated with second order approximation
 Threshold for convergence 
ID = 0 Optimisation converged when component of grads < 1.D3
= 1 converged when component of grads < 5.D4
= 2 converged when component of grads < 1.D4
= 3 converged when component of grads < 1.D5
= 4 converged when component of grads < 1.D6
= 5 converged when step components < 1.D4
= 6 converged when energy change < 5.D7
= 7 converged when grad norm < 1.D3
= 8 converged when grad norm < 5.D4
= 9 converged when grad norm < 1.D4
(** NOTE:
For special transition state calculations III can be used as a six
digit integer of the form IP:IQ:IA:IB:IC:ID. Here IP and IQ can be
used to override the default options built in for the Augmented
hessian and gradient extremal methods, viz. Allowing the use of
NewtonRaphson with step length controlled by a trust region to move
to the nearest extreme point, after the hessian attains 1 negative
eigenvalue (IP=8).
This option is useful to override the default minimization of norm
squared of the gradient when using Augmented hessian. This feature
may also be useful when using gradient extremal. IP can take values
4 and 8 corresponding to the values of IB and IQ can take values 2,
4 and 5 corresponding to values of IC.
If the V1 approximation to the exact analytic hessian is chosen
(IB=4), then care must be exercised in the choice of a mode since
the eigenvalues of the V1 hessian may be depressed in comparison
with those of the exact analytic hessian. A good choice of MODE in
such cases is 3. **)
** Other Geometry Switches **
MODE refers to the initial vibrational mode to be followed by the
transition state search in a "breakup" or rearrangement
reaction.
ITYPE controls the type of update procedure used
ITYPE = 1 <<<<< BFGS UPDATE (DEFAULT)
= 2  MURTAGHSARGEANT (OLD FNMIN)
= 3  DFP
= 4  GREENSTADT
LMASS = .TRUE. <<<<< if mass scaled coordinates are used
= .FALSE.  if not
RHO = 0.03 <<<<< DEFAULT
0.50  an exact line search, DON'T DO IT!
 0.01 < RHO < 0.45 is a typical range
SIGMA = 0.9 <<<<< DEFAULT (a loose line search)
= 1.0  any decrease in energy
= 0.5  exact line search, DON'T DO IT!
 0.55 < SIGMA < 0.95 is a typical range
PHASE is a parameter used for transition state calculations
PHASE < 4.5  Energy is maximized along the direction
vector between reactants and products or
guessed transition state
PHASE < 3.5  First move is along mode of Hessian that
most strongly overlaps with the vector
between reactants and products or guessed
transition state
PHASE < 2.5  First move is along mode of Hessian that
most strongly overlaps with input direction
vector
PHASE < 1.5  phase of move along chosen mode is
determined by the vector between reactants
and products or guessed transition state
(For stable state calculations, phase of
move along chosen mode is determined by
the vector between transition state and
reactants)
PHASE = 1  phase of chosen mode of Hessian is reversed
PHASE = +1  phase of chosen mode of Hessian is unchanged
LINV = .TRUE.  Inverse Hessian is used
= .FALSE.  Direct Hessian is used
(This depends on ITYPE)
If one wants an Updated Augmented Hessian then LINV must be
set FALSE on input.
LEXACT = .TRUE.  Once negative eigenvalue is found in a
transition state calculation with subroutine
SADDLE, then exact method for minimizing the
norm of the gradient is used
ISRCH = 1  Normal line search method (subroutine SEARCH)
= 2  Alternate method (subroutine STPSIZ)
STPTOL = 0.4 au  Maximum allowed stepsize norm (for IC = 4)
DEFAULT = 0.4 au
** NOTE **
The program will accept internal coordinates and constrained
optimisation is allowed in internal coordinates.
Specification of internal coordinates are described below and
are completely general.
The internal (or valence bond) coordinate option is an alternative
to the more conventional approach of minimizing molecular energies
with respect to cartesian coordinates. Internal coordinates form a
natural approach to constrained optimization problems. In a large
molecule or cluster one may be interested in optimizing a limited
number of bond lengths and/or bond angles. Experience, however, has
indicated that the initial complete optimization of large systems
may be better performed using the cartesian coordinates approach.
The cartesian approach usually does not produce large relative
atom movements, whereas the internal procedure links together atom
movements and can result in large displacements, particularly of
the end atom in a long chain of atoms.
Basic outline of the internal coordinate procedure:
As in any standard molecular calculation a set of cartesian
coordinates is read in by the program, the atom labels correspond
to the order in which the atomic coordinates are read in (i.e.,
atom 1 corresponds to the first line of coordinates, 2 the second
line etc.). Then a separate set of internal coordinates, which are
specilfied by the relevant atom labels, are read in from the group
$INTCOR
! An example of some internal coordinates for water.
INTERNALS
102
103 20103
$END
The different internal coordinates are: 
a) Bond Lengths  designated by two integer atom labels i and j
b) Bond angles  denoted by three atom labels i, j, k where j is
the atom at the vertex of the angle.
c) Torsion and wag angles  these require 4 atom labels i, j, k, l
as well as an integer indicating the
appropriate coordinate type.
ITYPE = 1 : ijkl  the torsional coordinate is the angle
between the planes ijk and ijl.
l
/
ji
\
k
ITYPE = 2 : ijkl  this type is similar to type 1 however
atom l is joined to atom j. Again the
coordinate is the angle between the
planes ijk and ijl.
l
/
ij
/
k
ITYPE = 3 : ijkl  this is a wag coordinate, atoms ijk
define a plane and the angle gives the
line il out of the plane ijk.
\
l j
t \ /
 i
\
k
ITYPE = 4 : ijkl  same as 3 but the angle is
180(angle from 3).
j
 l
 /
/ t
i l
\
\
k
Input of Internal coordinates:
Each input line with internl coordinate data specifies the position
of one new atom relative to atoms with known positions.
Line 1: A bond length coordinate. This gives the labels for the
initial two atoms from which the complete system is built up.
Line 2: Bond length coordinate and either an angle coordinate or
another bond length. These two coordinates relate the
third atom to the first two.
Remaining lines:
Each remaining atom is specified in turn by three internal
coordinates. Each of these coordinates should only include
one atom label not previously specified. The various
combinations of allowed coordinates are: 
1) Three bond lengths ril, rjl, rkl ; positions atom l
2) One bond length ril and two bond angles Ojil, Okil.
3) One bond length, one bond angle, and one torsion angle.
Common combinations are
a) ril, Ojil, tijkl (ITYPE=1)
b) rjl, Oijl, tijkl (ITYPE=2)
c) ril, Ojil (or Okil), tijkl (ITYPE=3 or 4)
4) Two bond lengths and one torsion angle. These
coordinates enable the closure of rings of atoms, e.g.
in the figure, l is the only previously unspecified
atom. Ring closure can be effected by: 
h
\
l
/
ij
/ \
k k'
a) rjl, rhl, tijkl (ITYPE=2)
b) rjl, rhl, tjik'l (ITYPE=1)
A total of (N1) data lines are required for input, where N is the
number of atoms in the system. Each coordinate is expressed as a single
integer using the rule: 
a) Bond length rij = I*100 + J: i.e., 102 for r12
b) Bond angle Oijk = I*1000 + J*100 + K
c) Torsion tijkl = I*10000000 + J*100000 + K*1000 + L*10 + ITYPE
Constrained Optimization
A coordinate is not optimized when it is input as a negative integer,
i.e., 010203 fixes angle O123.
The program allows scale factors to be associated with each
coordinate. However, when no scale factor is given the default is one.
Data Format:
Line 1: bond length (and scale factor) I5,F10.6
Line 2: 2nd and 3rd coords. (and factors) I5,I7,2F10.6
.
Line N1: 3 coords (and factors) I5,I7,I10,3F10.6
NOTE Bond lengths are listed before bond angles, which in turn are
listed before torsion angles on each line.
Error checks and redundant coordinates.
The subroutines in the internal coordinate package make several
simple checks for errors in the input coordinates and prints out
selfexplanatory error messages.
A more difficult error is associated with redundant coordinates. For
example consider NH3:
H

Ob  Oa
N
/ \
/Oc \
H H
The figure above indicates a suitable set of coordinates when the
molecule is nonplanar. However if NH3 becomes planar then one of
the bond angles is redundant since
Oa + Ob + Oc = 360
and there is no coordinate present describibg the atomic
coordinates perpendicular to the plane of the molecule. An error in
the program occurs because the redundant coordinates produce a
singular coordinate transformation matrix which needs to be inverted.
To diagnose the redundant coordinate the transformation matrix is
diagonalized, and the eigenvector (or vectors) associated with zero
eigenvalue groups together with the redundant coordinates. If a
coordinate is entered twice, one of the coordinates will be found
as a redundant coordinate.
Constrained optimisation can also be done in cartesian coordinates.
This also requires an INTCOR Block.
$INTCOR
! An example of a four atom system in which the second and third atoms
! have there z coordinates frozen during the geometry optimisation.
CARTESIANS
1.000000 1.000000 1.000000 < this is the x, y and z scaling
1.000000 1.000000 0.000000 factors of atom 1 in free format
1.000000 1.000000 0.000000 1.0=full optimisation
1.000000 1.000000 1.000000 0.0=completely frozen
$END
An example of a calculation for a transition state
$TITLEI
NH3 Transition state
April 6, 1988
$$$$$GEOMETRY OPTIMIZATION SUMMARY$$$$$
IT ENERGY SCF GRAD MAX ALPHA LEFT
CONVERGENCE NORM GRAD FAILS FAILS
0 12.522849 0.000013 0.001337 0.000700 0.000000 0 0
1 12.521285 0.000002 0.013850 0.005653 0.000000 0 0
2 12.516034 0.000005 0.021692 0.009547 0.000000 0 0
3 12.509819 0.000006 0.025393 0.012428 0.000000 0 0
0 12.509819 0.000006 0.025393 0.012428 0.000000 0 0
1 12.507360 0.000008 0.023395 0.012907 0.325770 0 0
2 12.507139 0.000004 0.008693 0.004908 0.967428 0 0
3 12.507084 0.000000 0.001000 0.000493 1.000000 0 0
$END
$CONTRL
SCFTYP RHF RUNTYP GEOM ENTTYP COORD UNITS ANGS
INTTYP 0 IAPX 3 III 3120 NAT 4
NEL 8 MULT 1 ITMAX 10 MODE 1
ITYPE 1 LMASS .TRUE. LINV .TRUE. LEXACT .FALSE.
ISRCH 1
SCFTOL 0.000000 RHO 0.01 SIGMA 1.0 PHASE 4.0
STPTLD 0.4
! ***** Basis set and C.I. size information *****
DYNAL(1) = 0 3 1 0 0 0 0
! ***** Interaction factors *****
INTFA(1) = 1.0 1.267 0.585 1.0 1.0 1.0
! ***** Output file name *****
ONAME = g6
$END
$DATAIN
0.000000 0.000000 0.000000 7 5.000000 2
0.966500 0.000000 0.410672 1 1.000000 1
0.483250 0.837018 0.410672 1 1.000000 1
0.483250 0.837018 0.410672 1 1.000000 1
0.000000 0.000000 0.400000 7 5.000000 2
0.966500 0.000000 0.310672 1 1.000000 1
0.483250 0.837018 0.310672 1 1.000000 1
0.483250 0.837018 0.310672 1 1.000000 1
$END
$INTCOR
CARTESIANS
1.000000 1.000000 1.00000
1.000000 1.000000 1.000000
1.000000 1.000000 1.000000
1.000000 1.000000 1.000000
$END

 Section V: Electron assignment 

$TITLEI
Hexaminocobalt(II)  quartet C.I.
from doublet ROHF  assign electrons
April 6, 1988
$END
$CONTRL
SCFTYP = ROHF RUNTYP = CI ENTTYP = COORD UNITS = ANGS
ASYM = CS
INTTYP = 1 APX = INDO/1
NEL = 55 MULT = 2 ITMAX = 10
SCFTOL = 0.000110
! ***** C.I. size information *****
CISIZE=200 ACTSPC=45
! ***** ROHF information *****
NOP = 2 NDT = 1
FOP(1) = 54.000000 1.000000
! ***** Interaction factors *****
INTFA(1) = 1.00000 1.26700 0.64000 1.00000 1.00000 1.00000
! ***** Output file name *****
ONAME = ci1
$END
$DATAIN
0.000000 0.000000 0.000000 27
0.000000 0.000000 2.200000 7
0.000000 0.980520 2.546670 1
0.849160 0.490260 2.546670 1
0.849160 0.490260 2.546670 1
0.000000 0.000000 2.200000 7
0.000000 0.980520 2.546670 1
0.849160 0.490260 2.546670 1
0.849160 0.490260 2.546670 1
2.200000 0.000000 0.000000 7
2.546670 0.980520 0.000000 1
2.546670 0.490260 0.849160 1
2.546670 0.490260 0.849160 1
2.200000 0.000000 0.000000 7
2.546670 0.980520 0.000000 1
2.546670 0.490260 0.849160 1
2.546670 0.490260 0.849160 1
0.000000 2.200000 0.000000 7
0.980520 2.546670 0.000000 1
0.490260 2.546670 0.849160 1
0.490260 2.546670 0.849160 1
0.000000 2.200000 0.000000 7
0.980520 2.546670 0.000000 1
0.490260 2.546670 0.849160 1
0.490260 2.546670 0.849160 1
$END
$CIINPU
! ***** Configuration Interaction specification *****
5 3 10 4 0 0 0 0 0 0 0
60000.0 0.000000
2 1 2
2 26 26 27 27 28
21 20 28 28 34
22 23 28 28 29 23 28 28 29
2 26 26 27 27 29
21 20 29 28 34
22 23 29 28 29 23 29 28 29
2 25 25 28 28 29 29 26
21 25 29 26 27
22 25 29 26 27 25 29 26 27
2 26 26 27 27 28
21 20 28 28 34
22 23 28 28 29 23 28 28 29
2 26 26 27 27 29
21 20 29 28 34
22 23 29 28 29 23 29 28 29
2 25 25 28 28 29 29 26
21 25 29 26 27
22 25 29 26 27 25 29 26 27
$END
$ASSINP
! ***** Electron assignment specification *****
10 7 5 5 0 0 0 0
0.000000 5 0.400000 0 0.000000 0 0.000000
0.000000 6 0.400000 0 0.000000 0 0.000000
2.000000 7 0.400000 0 0.000000 0 0.000000
2.000000 8 0.400000 0 0.000000 0 0.000000
2.000000 9 0.400000 0 0.000000 0 0.000000
0.500000 5 0.400000 0 0.000000 0 0.000000
0.500000 6 0.400000 0 0.000000 0 0.000000
0.000000 7 0.400000 0 0.000000 0 0.000000
0.000000 8 0.400000 0 0.000000 0 0.000000
0.000000 9 0.400000 0 0.000000 0 0.000000
$END
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
Data for ELECTRON ASSIGNMENT explained
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
$ASSINP
Switches for assignment of electrons to MOs are in FORMAT 8I5
Several options exist for assigning orbital occupations
and are determined by the value of NASS(1)
NASS(1) = 0 <<<<< Molecular orbitals are filled in order of
increasing orbital energy (aufbau)
NASS(1) =1  For ROHF calculations: MOs are rearranged and
the lowest NEL/2 are filled.
NASS(1) =2  Molecular orbitals are filled on the basis
of symmetry information
NASS(1) > 0  Orbital occupations are assigned on the basis
of the atomic orbital character of the MOs.
 For NASS(1) > 0  First Line FORMAT 8I5
NASS(1) = J  J is a positive number specifying the the number of
lines of data to be read in
NASS(2) = K  K is the total number of ELECTRONS to be assigned by
this method
NASS(3) = 0  For closed shell RHF
= L  Assign L ALPHA orbital occupancies (UHF)
Assign L CLOSED shell orbitals (ROHF)
NASS(4) = M  Assign M BETA orbital occupancies (UHF)
Assign M orbitals to the first open shell (ROHF)
NASS(5) = N  Assign N orbital occupations in subsequent
NASS(6) open shell(s) (ROHF).
NASS(7)
NASS(8)
 For NASS(1) > 0  Subsequent line(s)
Orbital occupancies (by shells) as listed above. FORMAT for
specifying occupancy is given by the values of
ELECTRONS AOSPECS THRESHOLD (F10.6,3(I5,F10.6))
ELECTRONS  The number of electrons to be assigned to a
particular MO. This is a FLOATING POINT number
ranging from 0.0 (a hole) to 2.0 (F10.6)
AOSPECS  The number of the AO basis function as it appears
in the MO. AO 3 would be the third basis function.
AOSPEC is an I5 integer that may specify up to two
different AOs. A value of 05025 specifies both AO
5 and AO 25. (I5)
THRESHOLD  The test used to determine the character of the MO
As many as three pair of (AO/THRESH) can be
specified on each line. (F10.6)
EXAMPLES:
10 7 5 5  There are 10 lines assigning 7 electrons,
5 for the first shell, and 5 for the second
0.000000 5 0.4  Assign ZERO electrons to the MO which has
a coefficient for AO number 5 that when
SQUARED is larger than 0.4
Nine more lines are required for this example
3 2 0 2 1  3 lines assigning TWO electrons, none
for the closed shell, two for the first
open shell, one for the second.
0.00000 3 5 0.200000  Assign zero electrons to the MO in which
the product of the coefficients for AO 3
and AO 5 is a larger negative number
than 0.2000000 This line is for the
first open shell
1.00000 22 1.00000  Assign two electrons to the MO with a
node on AO 22 for second open shell
2.00000 20 0.400000 25 0.400000  Assign two electrons to
the MO with coefficients
for AO 20 > 0.4 and for
for AO 25 > 0.4
** NOTE ** MOs are scanned from the HOMO/LUMO gap going up 1, then
going down 1, then up 2, down 2,... until all assignments
are satisfied.
 NASS(1) = 1 
Next line specifies rearranged order of starting vectors
in (16I5). First orbital and last orbital specified must
not be rearranged. Note that this requires the use of
starting vectors and the rearrangment only takes place on
the first cycle. Subsequent cycles revert to NASS(1) = 0
BECAUSE OF ORBITAL RELAXATION, THIS IS A DANGEROUS OPTION!
For UHF, two such lines must be given, first for beta spin
orbitals, then alpha.
EXAMPLE(RHF):
1
3 4 5 7 6 8  Specifies transposing MO 6 and MO 7
before filling by aufbau. This procedure is only done on the
first cycle. All other cycles are done by aufbau.
 For NASS(1) = 2 
NASS(2) = 0
NASS(37) = 0 or 1. A zero specifies that the orbitals of
this operator are to be filled by the aufbau principal,
a one specifies that this shell will be symmetry assigned
by the following lines. For this case the number
of mo's in each symmetry type are given in order.
The format is 8I5.
i.e:
2 0 0 1 1 1 0 NASS.
4 3 2 1 second shell
0 1 1 1 third shell
0 0 1 2 fourth shell
This is an example of an open shell ROHF calculation
with five shells. The first shell is filled by aufbau,
the second shell has 4 mo's of irrep 1, 3 of irrep 2,
2 of irrep 3, and 1 of irrep 4. The next open shell
has no mo's of irrep 1 filled, 1 of irrep 2, one of
irrep 3, and one of irrep 4.

 Section VI: Calculation of Polarizabilities 

IN the $CONTRL block add
RUNTYP = CI
POLAR = 1 < This is Bill PArkinson's TDA and RPA for
alpha, TDA for beta and gamma. and approx.
RPA for beta. This is fast for alpha and
beta, and the size of the CI should be
set equal to the particle*hole space + 1
in DYNAL (sixth entry)
CAUTION: THE CI LINE DETERMINES THE HOLE AND
PARTICLE SPACE INCLUDED IN THE CALCULATION
SEE THE GENERATING LINE
= 2 < This is Kanis for frequency dependent beta.
CAUTION: THE NUMBER OF STATES INCLUDED WILL
NOT EXCEED N3 ON THE CI LINE. A selection
of states based on the energy differences
between ground and excited states is made.
If these states do not contain the oscillator
strength, the results will not be good.
This code has undergone major restructuring, and
it is unlikely problems here are the fault of
Dave Kanis! This code knows no symmetry. If symmetry
is to be used, use POLAR = 4
= 3 < This is both of the above
= 4 < This is the Yu  Parkinson  Zerner frequancy
code. This is the fastest code for polarizability
and includes the solvent effect if ISCRF . NE.0
= 0 < This is the default: do not calculate pol
arizabilities.
If using the frequency dependent Kanis code, these frequencies
can be specified in the $CONTRL block.
$CONTRL
.
.
FREQ(1) = 0.000 0.650 1.12 1.30 < up to 6 in ev.
FREQ(1) = 0.000 0.000 < only static polarizabilities.
FREQ(1) = 0.000 2.000 1.000 < six frequencies will
be calculated from 0.0ev
to 2.00ev in equal steps.
If no frequencies are given, the 3 defaults are 0.0, 0.6491 and 1.17 ev

 Section VII: Spinorbit calculations

$TITLEI
Cerium Oxide Doublegroup C.I. + spinorbit
triplet CI from Triplet ground state
$END
$CONTRL
SCFTYP = ROHF RUNTYP = CI ENTTYP = COORD UNITS = ANGS
ASYM = C2V
INTTYP = 1 APX = INDO/1
NEL = 10 MULT = 3 ITMAX = 40
SCFTOL = .000010 SPNORB = 1 VEC = 9
! THE ALLOCATION FOR THE CISIZE BELOW = 14, IS AN ESTIMATE OF THE NUMBER
! OF DETERMINANTS NEEDED FOR CALCULATING OSCILLATOR STRENGTH
! IF THIS IS A PROBLEM KEEP UNIT = MOMENTS EXTERNAL
DYNAL(1) = 0 0 1 0 1 14 20
! ***** C.I. size information *****
CISIZE=10 ACTSPC=20
ONAME = CeO1
! ***** ROHF information *****
NOP = 8 NDT = 0
FOP(1) = 8.000000 1.000000 1.000000
MIM(2) = 7 1
AR(1) = 1.000000 0.000000 1.000000
BR(1) = 1.000000 1.000000 1.000000
! ***** Interaction factors *****
INTFA(1) = 1.00000 1.26700 0.58500 1.00000 1.00000 1.00000
! ***** Beta Values *****
LBETA(1) = 58 8 0 58 58
BETA(1) = 1.00000 21.00000 00.00000 6.00000 12.00000
$END
$DATAIN
0.000000 0.000000 0.000000 58
0.000000 0.000000 1.821000 8
$END
$CIINPU
! ***** Configuration Interaction specification *****
! N1 REF ROOTS MULT NATO DEN NOPT FROM>TO FROM>TO
50 1 50 3 0 0 0 1 4 1 50
! 50=DGCI 1 = 1 REF, 50 states, 3=triplet, 0,0,0 then transition moments
! 1 t0 4 into 1 to 50 (or max number in this case)
0
! 0 NO THREHOLDS ARE OPERATIVE HERE
4 1 2 3 4
! THIS IS 4 IRREPS, THE NEXT VALUES ARE ORDER OF COMPONENTS OF ANGULAR
! MOMENT INTEGRALS, AND LEAVE THIS ALONE. THE LAST VALUE IS THE NO OF
! IRREPS CONSIDERED.
58 0.126217 08 0.018846 00 00.000000 58 0.045998 58 0.076995
! THE ABOVE IS THE RESET OF SPIN ORBIT FACTORS "ZETA" IF DESIRED, THREE P,
! THEN D, THEN F: ATOMIC NUMBER THEN ZETA, IN ATOMIC UNITS
! SYMMETRY SET 1
!KSYM NBCFG NCFGIN NCFGDE MXOP KORBOC SEE POWER USERS SECTION
1 1 0 0 2 0
! SYMMETRY 1, NO. OF BASIC CONFIGS = 1, NO. OF CONFIGS INSERTED,
! No. of CONFIGS DELETED, 2=MAX No. OF OPEN_SHELLS, KORBOC,
! SEE CIGNLS IN POWER USERS SECTION
2222100000010000000011
!<MO's><EXCITATION LEVELS
! FIRST BASIC CONFIGURATION HAS TWO ELECTRONS IN THE FIRSTFOURTH
! MO, THEN 1 IN THE FIFTH AND 1 IN THE 12TH MO. THERE ARE 20 MO'S.
! THERE ARE TWO GROUPS OF EXCITATIONS GIVEN AFTER THIS. ONLY SINGLES
! IN THE FIRST GROUP, AND ONLY SINGLES IN THE SECOND GROUP. THE NEXT
! LINE SPECIFIES THE ACTIVE ORBITAL IN EACH GROUP.
! SEE SUB GENCFG IN POWER USERS SECTION
5 6 7 8 9 10 11 1
! THE FIRST GROUP SINGLY EXCITES ALL ORBITALS IN 511. NO EXCITATIONS
! IN THE SECOND GROUP.
! 5 6 7 8 9 10 11 12 2 WOULD BE ALL DOUBLE EXCITATIONS IN THIS GROUP
! OF ORBITALS
! SEE SUB GENCFG IN POWER USERS SECTION
!NROOTS ICIWRT
50 1 0 0 0 0 0
! FOR THE ABOVE SEE CIDG: 50 = MAX NO. OF STATES, THEN ANALYSE THE CI (=1).
! THE OTHER OPTIONS ARE EXPLAINED IN THE POWER USERS SECTION BUT ARE EASILY
! DEFAULTED TO ZERO.
1 0
! SYMMETRY SET 2
2 1 0 0 2 0
2222100000010000000011
5 6 7 8 9 10 11 1
50 1 0 0 0 0 0
1 0
! SYMMETRY SET 3
3 1 0 0 2 0
2222100000010000000011
5 6 7 8 9 10 11 1
50 1 0 0 0 0 0
1 0
! SYMMETRY SET 4
4 1 0 0 2 0
2222100000010000000011
5 6 7 8 9 10 11 1
50 1 0 0 0 0 0
1 0
$END
Some comments about output:
Some knowledge about the double group will be needed to interpret
these results and to assign state symmetries. Recall that for states
of halfintegral spin (or J) the results that are meaningful occur
in irrep #5 of the double group. For integral values of spin (J)
the meaningful results occur in the first four irreps; ie, those
that would appear in the normal group.

There are two ways to do spinorbit calculations. The above uses the
double group formalism of Pitzer (see Pitzer, Roesch and Zerner). It
is effective for small molecules and works over determinants, thus
including all possible spin mixings. The second method uses the Rumer
formalism of Manne and Zerner as developed by Kotzian, Roesch and Zerner
and mixes the spin states of interest with the states of the next highest
multiplicity. In the example that follows singlets are mixed with
triplets. This is a much faster proceedure.

$TITLEI
Rumer CI with spinorbit on water.
****** NOT FOR EXPORT ********
Configuration Interaction roots
E( 1)= 12.34235069 DELTA ENERGY(CM1)= 0.0
E( 2)= 12.28911576 DELTA ENERGY(CM1)= 11683.8
E( 3)= 12.27087987 DELTA ENERGY(CM1)= 15686.1
E( 4)= 12.19520501 DELTA ENERGY(CM1)= 32294.9
E( 5)= 12.18698504 DELTA ENERGY(CM1)= 34099.0
E( 6)= 12.16225904 DELTA ENERGY(CM1)= 39525.8
E( 7)= 12.16175970 DELTA ENERGY(CM1)= 39635.4
E( 8)= 12.15182784 DELTA ENERGY(CM1)= 41815.2
E( 9)= 12.14167006 DELTA ENERGY(CM1)= 44044.6
E( 10)= 12.08022492 DELTA ENERGY(CM1)= 57530.3
E( 11)= 12.02513743 DELTA ENERGY(CM1)= 69620.7
E( 12)= 12.01461330 DELTA ENERGY(CM1)= 71930.4
$END
$CONTRL
SCFTYP = RHF RUNTYP = CI
DYNAL(1) = 0 2 1 0 0 200 80
IAPX = 3 INTTYP = 1
! SPNORB = 1 FOR RUMER SPIN ORBIT TREATMENT. N1 IN CIINPU = 5 FOR RUMER
! VALENCE BOND CONSTRUCTION, = 105 FOR BRANCHING DIAGRAM CONSTRUCTION.
! (THIS USES A SCHMIDT ORTHOGONALIZATION ON THE RUMER STRUCTURES IN THE
! GENEOLOGICAL ORDER)
SPNORB = 1
NAT = 3 NEL = 8 MULT 1
SCFTOL = 0.00001 ITMAX=20
UNITS = BOHR
$END
$DATAIN
0.000000 0.000000 0.000000 8
0.000000 1.513900 1.171700 1
0.000000 1.513900 1.171700 1
$END
$CIINPU
5 1 20 3 0 0 0 1 1 1 5
60000.000 0.000000
1
! THESE RESET SPIN ORBIT ZETA PARAMETERS FIRST THREE ARE P, THEN D THEN F.
! UNITS ARE EV. NOTE O value IS USED HERE BUT THOSE FOR Ce ARE AS EXAMPLES
! LEAVE A LINE OF:
! 0 0.000000 FOR DEFAULT.
58 0.126217 06 0.018846 00 00.000000 58 0.045998 58 0.076995
! THE NEXT TWO LINES ARE FOR THE TRIPLET CI
1 4 4
21 1 4 5 6
! THE NEXT TWO LINES ARE FOR THE SINGLET CI
1 4 4
21 1 4 5 6
$END

 Section VIII: Power users section 

Additional switches for $CONTRL section
 Print options 
IPRINT = 0 <<<<< for normal operations
= 1  Prints H, F, COULOMB, and other calculated matrices
=1  for reduced output.
 Molecular volume and area 
VOLTYPE = GEPOL uses tesserae and gepol93 to calculate surface
accessible area and volume.
RADTYPE = MASS radius from mass density, default
MAXDIM radius from maximum distance/2 + 0.5Angs
RAVE radius from average effective distances from
center of solute to surface of solvent.
If a radius is = 0.0 in the $SCRFIN block, it RADTYPE = MASS
is > 0.2 in this block, it overwrites the mass density
option and the MAXDIM option, but will be used in RAVE
 Vectors 
VEC = 0 <<<<< DEFAULT Normal calculation
= 8  Read MO coef from file with extnsion .vec8 ,
first BETA, Then ALPHA
= 9  Store MO coef on file with extension .vec9 ,
Every 5th cycle
New vectors will be written in format suitable
with reading later using option 8
= 10  Read MO coef from file ONAME.vec8 to start and store MO
coef every 5th cycle on file ONAME.vec9
 Effective core potential
VCORE = 1 for the Zerner effective core potential
 Moessbauer section 
MOSSB = 1 calculate Moessbauer spectra, Nuclear Quadrupole
Moment of Fe set to 0.187
2 same with 0.210
 Memory Management 
DEBUG0 = .TRUE. Print names of subroutines being called to unit
0 = screen. Default = .FALSE.
DEBUGM = .TRUE. Print memory allocation and deallocation information
Default = .FALSE.
MEMINI = .TRUE. Initialize dynamic memory to standard patterns
(all reals are 1.23456789012345) Default = .FALSE.
TRACELBL = unit as AOINTS, MOMENTS, etc, will track all io
on this unit.
 Electrostatic potentials 
IELEC = 0 <<<<< DEFAULT
= 1  FILE 23 XXXX.inb is cataloged for calculation
of electrostatic potentials using ELESTA program
 Molecular Orbital Plot
A molecular orbital plot appears below the MO eigenvector printout.
This plot can be changed
NPLOT = Number of lines in the plot. The default is 50, and this
fits on one page of printout. This can be extended up to
200 lines, which is often useful for transition metal
complexes which have low lying d orbitals
NBELOW = Number of orbital to be shown below HOMOLUMO gap. The
number of virtual orbitals plotted is always 10. The default
value is NBELOW = 20; ie, 10 virual and 20 occuppied
mo's are plotted on 50 lines.
$CONTRL
.
.
NPLOT = 50 NBELOW = 20
.
.
$END
 Triplet parameterization 
ITRIP = 0 <<<<< DEFAULT
= 1  Triplet parameterization. This is an alternate
parameterization. See J.E. Ridley & M. Zerner
Theoret. Chim. Acta, 32,111,(1973)
Theoret. Chim. Acta, 42,223(1976)
 Configuration mixing for metals 
ISW2 = 0 <<<<< Valence bond mixing of configurations DEFAULT
= 1  D(n1)S(1) configuration of the metal
= 2  D(n2)S(2) configuration of the metal
= 3  D(N) configuration of the metal. Only available
for the second transition series.
= 4  Input extent of configuration mixing
First give atomic no. than amount of D(N2)S(2),
then D(N1)S(1), then D(N) if any.
Two sets max in 2(I5,3F10.6)
This option has its own group in the data set 
The format is I5,%F10.6, and if one line is given, two will be
expected. Atomic number, then the five mixing values, and these
must add up to 1.000000.
For the main transition series
COFSQ2(ATOMIC NO., TYPE) TYPE = 1, 2, 3, 4 FOR
D(N1)S(1), D(N2)S(2), D(N), AND D(N1)SP RESPECTIVELY.
For the lanthanides:
TYPE 1 = F(N3)D S(2) : TYPE 2 = F(N2) S(2)
FOr the actinides:
TYPE =1, 5F**(N3) 6d 7S**2: TYPE=2, 5F**(N3) 7P 7S**2:
TYPE =3, 5F**(N4) 6d**2 7S**2: TYPE=4, 5F**(N4) 6d**3 7S:
TYPE =5 5F**(N2) 7S**2
$MIXCOF
Z 0.421000 0.220000 0.220000 0.100000 0.040000 <Atomic number,
0 0.000000 Then the mixing
coef. Two lines
. are required.
$END
 Output control 
Add a section $OUTPUT, which contains keywords. Various parts of the
program print additional information depending on these keywords.
Recognized keywords are:
ALL Changes the default value for IPRINT to 1.
ALL_VARIABLES Prints all variables that the user could have assigned
in $CONTRL, no matter whether they actually where input.
This is a good way to study the default values.
CIMATRICES Prints the CI matrix and the Overlap Matrix between
CSF's.
CIBONDS Analyzes the bond orders of excited states from N9
to N8 of the CI section, and will also estmate the
width of transitions between N1 and N10 thru N11.
CIVECTORS Print the entire C.I. vector array. CAREFUL! This may be
huge!
CONFILE Creates an output file with .con extension and places
atomic connection informatino on it.
DIHEDRAL Prints all dihedral angles. For a large molecule this
could be quite lengthy.
DISTANCES Prints the enetire Bond length matrix for each cycle
of a geometry search.
DIPOLEMAP Prints the dipole matrix elements between CI states.
GEOM_ITERATIONS Prints the detailed information on a geometry search.
HMATRICES Prints all oneelectron matrices
MIN Changes the default value for IPRINT to 1.
MOS Prints the detailed MO's. This could be lengthy
The default prints only the largest components of
each MO.
NO_CORE Do not calculate the innershell core mo's
NO_OPTFILE By default, Zindo creates an output file with .opt
extension and puts geometry optimisation summary in it.
By specifying this keyword, the .opt file is deleted
at the end of run.
OVERLAP Prints the Overlap matrix
PCM_DEBUG Prints debug information for the PCM solvent model
PCM_VISUAL Prints information from the PCM solvent model, which
can be used to visualize the cavity surface and point
charges
PRNT_SPINDN Calculate and print the spin densities for an ROHF,
ROHF/CI or UHF function
PUHF_MULPOP Calculate and print the Mullikan population for
projected UHF functions.
PUHF_NOAISO ?
RPA_OUTPUT For an RPA calculation, this will cause extended
output over the states.
SCF_ITERATIONS Prints additional information about SCF iterations.
SUMMARY Prints an oneline summary at the end of run, including
SCF energy and dipole moment. Can be useful for processing
many output files with 'grep'.
SUMRYFILE Creates a file with .smry extension and puts geometry
optimisation summary into it.
 Point charges 
$CONTRL
.
PTCG = N
...
$END
If point charges are used to modify the one electron matrix,
the number of these point charges is N. The
coordinates are then given under group as:
$PTCHGI
A title line of 70 or less symbols  you MUST give a title!
The Screening constant (aa) and a bulk dielectric (eps).
in 2F10.6. Leave zero if no dielectric screening.
A negative value of aa sets e = abs(aa)*R. R in a.u.
A zero value of aa, and a value of eps, sets e = eps.
Else e = eps (eps1)*exp(aa*R**2)
The coordinates in angstroms, X, Y, Z, and the
'atomic no.' of the pt. charge,and the charge,
in 3F10.6,I5,F10.6. one pt per line until all
pts are given. If the 'atomic no.' is zero a true pt.
charge is considered. if not a potential will be
generated as if the pt. were an atom of given charge
and atomic no.
$END
For example:
$PTCHGI
This is a test of point charge option
0.00000 0.00000 < aa and eps
0.000000 0.000000 3.000000 0 1.000000 < a positive pt.
0.000000 0.000000 3.000000 0 1.000000 < a negative pt.
$END
Note that a limited no. of pt. charges can also be added as atoms.
In such a case they cannot be screened.
 Other options 
AISO = N The magnetic isotropic hyperfine splitting constants
will be calculated for state N for ROHFCI wavefunctions,
(The authors favorite method.), or for the HFSCF reference
for the UHF, SUHF and PUHF methods. For the PUHF case the
pure S=Ms multiplet is project out of the UHF reference and
thus the densities are of those for a proper spin state.
(They are variational, but this wave function is no longer
stationary with respect to the SCF variational parameters.)
(KEYWORD ESR or EPR)
NMR = 0  NMR option This is FLORG/LORG and MCZ doesn't
like the implementation
SPNORB = 0 <<<<< DEFAULT for normal calculations
= 1  Spin Orbit calculation will be performed.
Both a SINGLET and a TRIPLET CI are required
 Self Consistent Reaction Field: 
ISCRF = 0 <<<<< DEFAULT for normal calculations
= 1  Self Consistent Reaction Field. Theory A for the
Ground state, theory C for the relaxation of
the excited state.
= 2  Self Consistent Reaction Field. Theory A for the
ground state, but C1 (mean field) for the relaxation
of the excited state.
= 3  Static electric field or FF = 1
= 4  Our version of the PCM Solvent model.
This option enables processing the data block $GEPOL
(optional).
= 5  SCRF on excited states (Absorbtion spectra)
fully relaxed theory, as in ISCRF =2 ,A
= 6  SCRF on excited states (Absorbtion spectra)
approximate theory, as in ISCRF =1 , B.
= 7  SCRF on excited states, mean field, A1
= 8  SCRF on excited states, mean field, B1
= 9  SCRF multicavity, theory A
= 10  SCRF multicavity, theory B
= 1116  Infinite relaxation for absorption or emmision
 see below.
= 19  SCRF multicavity/ellips, theory C
= 20  SCRF multicavity/ellips, theory C1
= 30  COSMO  NOT FUNCTIONAL
= 40  reserved for Piet van Duijnen's Direct Reaction Field
(not part of current Zindo)
= 50  Amsol model
If a self consistent reaction field calculation is to be
performed then additional information is required, in FORMAT
3I5,6F10.6.
ISCRF = As above, repeat this value.
NMU = The number of terms in the multipole expansion.
Default is 1, the dipole term.
IDISP = Is this a calculation for dispersion? If not
= 0 no dispersion
= 1 the solvent calculation, save the manifold of
excited states under SOLNAM_solvent.file.
= 2 use the above file in calculating the spectrum
of the present solute in the above solvent, including
dispersion.
A0 = Cavity radius in angstroms: if this is set to
zero, mass density will be used to determine A0.
EPS = Static dielectric constant for SOLVENT
Default = 78.5(water)
XND = Refractive index for SOLVENT
Default = 1.33287(water)
DENSV = Density of the solvent (g/cc), Can be ignored, unless
dispersion is calculated. Default = 1.0(water)
AMUSV = Molar mass of solvent (amu), Can be ignored, unless
dispersion is calculated. No default.
DIPSV = Dipole moment of solvent, Can be ignored
FNODEN = Solvent reduced number density, from Piorrotti scaled
particle theory. This is defaulted to 0.44 (spherical)
unless the solvent is water, than 0.371. This value
is used to calculated the cavitation energy.
If ISCRF = 3 then an addition line is required that contains the
field strength in au., with the X, Y, Z components.
1au = 5.14193*10**11 Volts/meter.
Formatted as 3F10.6
For example:
$CONTRL
...
ISCRF = 1
..
$END
$SCRFIN
1 2 0 2.000000 78.500000 1.332870<The solvent is water
The cavity of the solute
$END molecule has radius 2.0A
or
$CONTRL
...
ISCRF = 3
or FF = 1
..
$END
$SCRFIN
3
0.000000 0.000000 0.020000 <A field of 0.02 a.u. in
$END the Z direction
or
$CONTRL
....
ISCRF = 10
....
$END
$SCRFIN
THEORY = B < Theory B for excited states
EPS = 3.0
N_D = 1.414
CAVITY=ELLIPS RHO=1.0 < an ellipse, generate shape
automatically, assume density
equals 1. for the volume.
The default cavity type is SPHERE.
$END
$CONTRL
...
ISCRF = 16 <C1 relaxed theory
..
$END
$SCRFIN
16 0 0 2.000000 78.500000 1.332870<The solvent is water
1 2 < The state 'n'
fully relaxed, state 'm'
only electronically
relaxed. In the example
n =1 the grd state, and
m = 2, corresponding to
absoprtion.
$END molecule has radius 2.0A
$CONTRL
....
ISCRF = 50 < AMSOLCM2 THEORY
....
$END
$SCRFIN
JSCRF=1 < Refers to parameters for a given model
IDOSOL=0 < = 0 first gas phase, then solvent calculation
= 1 only the solvent calculation
ICMD=13 < Can be defaulted 11 = AM1, 12 = PM3, 13 = INDO
EOPT=1.80 < Optical dielectric, also can be given as N_D
that is, EOPT = N_D**2 (N_D = 1.34 or EOPT=1.80)
DIELEC=78.3 < Bulk dielectric
ICREAD=0 < { power users options to reset parameters,
ICSAVE=0 { see setsol.f
$END
In the $SCRFIN block, Key words may also be used to replace formatted input:
a) see above  summarized
First line; in (format 3I5,6F10.6,F5.3)
iscrf,nmu,idisp,a0,eps,xnd,densv,amusv,dipsv,fnoden
if iscrf=3 then the next line contains as 3f10.6: dipgr(1:3)
if iscrf = 1116, the next line contains as 2I5: iexstate, and
jexstate
if idisp =1 or 2, the next line contains an assignment in the form:
SOLNAM=XXXXX where XXXXX is the name of the solvent file for
dispersion calculation.
$SCRFIN
2 2 2 0.000000 2.280000 1.500000 0.876500 78.1100
SOLNAM = benzene
$END
The above is an example of a calculation in which ISCRF = 2,
using dipole and quadropole, default radius of the solute to
mass density, dielectric constant, index of refraction, density
of the solvent(benzene) = 0.8765, and the molar mass of benzene
is 78.11 amu. Note that the benzene_solvent.file is already
present, calculated from a previous benzene CI calculation using
See RADTYPE below.
$SCRFIN
2 2 1 0.000000 2.280000 1.500000 0.876500 78.1100
SOLNAM = benzene
$END
b) new, namesbased version. The general format is the same as
for $CONTRL block (Namevalue pairs). If the first
nonblank character in $SCRFIN is numeric, old format is
assumed, else new format is assumed.
PLEASE NOTE THAT NOT ALL COMBINATIONS OF KEYWORDS WORK. THE KEY
WORDS MARKED WITH (*) ARE GOOD FOR THE MULTICAVITY/ELLIPSOIDAL
(THE SCRF2 PACKAGE, ISCRF=9,10,19,20) ONLY!
FF = 0 DEFAULT
FF = 1 static field as above. With this option a static
electric field and the self consitent reaction field
can be simultaneously applied. The ISCRF = anything
$SCRFIN
2 2 0 2.000000 78.500000 1.332870<The solvent is water
0.000000 0.000000 0.001000 The field is 0.001 along z
$END
$SCRFIN
THEORY = A
CAVITY = MULTI
EPS = 80.0
N_D = 1.34
NMOLEC = 2
A0(1) = 0.37 1.00
FRAG(1) = 1 2
OVERLAP = .F.
$END
The following variable names are recognized in the new format:
ISCRF = sets the default value for ISCRF. Please note that:
1) ISCRF in $SCRFIN overrides any value assigned in $CONTRL
2) THEORY, CAVITY, and RELAX are mutually exclusive with ISCRF
The recommended practice is to use only ISCRF, _or_ use the
combination of the modelspecific keywords as described next:
THEORY= A B C A1 B1 C1 This sets the basic theory. Default=A
AMSOL, DRF, COSMO
CAVITY= SPHERE ELLIPS MULTI PCM This sets the shape of the cavity.
Default=SPHERE
RELAX= NONE INFABS INFEMI Type of solvent relaxation for solvent
calculation. NONE is the default; INFABS is for infinite
relaxation, absorbtion; INFEMI is for infinite relaxation,
emission.
NMOLEC= Number of molecules (fragments, cavities) for the multicavity
theory; ignored otherwise.
NMU= Highest multipole moment to be considered in reaction field.
0=charge only, 1=dipole, 2=quadrupole, 3=octupole, 4=hexa
decapole.
default=1
EPS= The dielectric permittivity of the medium. For multicavity,
this may be input as an array, e.g. EPS(1)=78.5 EPS(2)=30.0
N_D= The index of refraction of the medium. For multicavity, this
may be input as an array, just like EPS.
AMUSV= Molar mass of solvent in a.m.u.
RHO= Density of the solute (g/cm**3)
DENSV= Density of the solvent (g/cm**3)
DIPSV= Dipole moment of the solvent
FNODEN= The number density of the solvent, defaults to 0.44
(hard spheres)
A0= Radius or radii for the cavity(es). For a single spherical
cavity, just the radius. For multicavity, input as an array
the radii for each cavity: a0(1)=2.5 3.4 6.7 . For ellipsoidal,
input the three halfaxes in the order (x,y,z) as an array.
See also RADTYPE below.
FRAG(1)= Array designating atoms to specific fragments. For example,
if atoms 1 and 3 belong to fragment 1, and atom 2 belongs
to fragment 2, input as FRAG(1)=1 2 1
ALTDEN= A logical value, .T. if an alternate massdensity formula
is to be used in cavity size calculations. Default=.F.
OVERLAP= A logical value, .T. if cavities are allowed to overlap in
multicavity model. Default=.F.
MULTIPLY= A factor used to multiply all cavity sizes by. This can be used
for studies of cavity size; it is applied both to user inputted
and to autocalculated cavities, including the PCM model.
Default=1.0 .
CENTER= MASS CORE VOLUME  which factor is used to locate the centres
of cavities for the multicavity & ellispoidal approcahes.
Dipole integrals are recalculated relative to these centers.
default=MASS .
IDISP= Flag for dispersion calculations:
0  ignore dispersion (default)
1  estimate dispersion (calculating solvent)
2  use dispersion results from a previous run (calcu
lating solute)
For options 1 and/or 2 the following data are necessary:
SOLNAM= name of the 'solvent file' where information about the states
is stored.
IEXSTATE= The number of the selected excited state.
This is needed for the infinite realaxed models on
state IEXSTATE.
 Polarizable Continuum Model for Solvation 
If the Polarizable Continuum Model for solvation is used, an additional
set of keywords is recognized in $SCRFIN.
Reminder: ISCRF=4 for PCM.
If the PCM model is used, the electrostatic potential will be calculated
from the one center charges and dipoles as default. If only the charges
are used the keyword
ZDO_CHARGES
must be used. The default is
POINT_DIPOLES.
Mulliken charges can be used in stead of the ZDO moment expansion, and this
is
MULLIKEN
either ZDO_CHARGES, or MULLIKEN or PINT_DIPOLES
Keywords:
LOOP=INSIDE (def)  the surface charges are converged as part
if the SCF procedure
LOOP=OUTSIDE  looping around calls to SCF procedure in
order to converge the charges
SELFPOL=.T. (def)  selfpolarization is performed
SELFPOL=.F.  selfpolarization is omitted
SELFLOOP=.T.  inside each PCM call, internal loops are
used to selfpolarize charges
SELFLOOP=.F. (def)  the selfpolarization of the charges is
assumed to take place as part of the overall
procedure
FIELD=MULLIKEN  use Mulliken charges to generate field
This forces LOOP=OUTSIDE
FIELD=ZDO  use ZDO charges to generate field
FIELD=ZDO+DIP  use ZDO charges + point dipoles to generate
field (default)
NORMALIZ=BEFORE  charges are renormalized before the self
polarization procedure is started.
NORMALIZ=AFTER(def)charges are renormalized after being
selfpolarized.
NORMALIZ=NEVER  charges are never renormalized
Please note that the default is not to loop around neither
inside, use ZDO+dipole field, include selfpolarization, and
normalize after looping (on every cycle in case of inloop PCM)
Please note that Mulliken charges are available only with loop=outside
approach at this time.
The additional input block, $GEPOL, may be present to directly interact
with the GEPOL93 cavitygeneration algorithm. The options are documented
in the GEPOL93 manual and source code. No blank lines or free format;
the block is scanned starting from the very first line by GEPOL itself.
A typical set may include:
$GEPOL
LPRIN
TESSE
$END
 COSMO solvent Model 
The COSMO Solvent model was implemenmted into ZINDO by
Greg Pearl, Andras Klamt, and Michael Zerner... This Solvent model
is still being tested and should not be used for Reseach... 062197
The COSMO solvent model is set with ISCRF = 30 or theory = COSMO
Then using the normal solvent model switchs allowed in the SCRFIN, in
addition to these switches there are 2 new keyword used specifically
by COSMO.
NSPA  Number of Segments per Atom
DEFAULT (NSPA = 42)
NPPA  Number of Basic grid points per atom
DEFAULT (NPPA = 1082)
COSMO Currently is not prop[erly implemented into the memory managment
of ZINDO and therefore requires a minimum of 210 Meg of RAM under SunOS2.6
The current implementation of COSMO is also extremely slow using the
defualt value for NSPA and NPPA...
 Projected UHF 
there are 7 additional input switches that append what RUNTYP=PUHF
calculates/reports. these are PUHFST and AISO found in the $CONTROL
section of the input deck and, PUHF_MULPOP, PRNT_SPINDN, PUHF_NOAISO
and PUHF_OVLAP and PUHF_BONDOR found in the $OUTPUT section of the
input deck.
a straight RUNTYP=PUHF run will output the weights and energies of all
multiplets whose weight > 10**(8), as well as for the unprojected
UHF wavefunction. (regardless of the values of the 4 output control
switches the weights and are always reported.)
PUHFST:
this switch allows the selection of which multiplet/s to analyze by
energy. i.e. PUHFST=ABCD will select multiplets AB through CD to be
analyzed by energy. e.g. PUHFST=0103=103=0301=301 will cause multiplets
13 to be analyzed by energy, or PUHFST=02=2 will cause multiplet 2 to
be so analyzed. if PUHFST=1 then only the weights and are
reported. the program default is PUHFST=0 (for larger systems the
determination of the energy can be very time consuming, PUHFST is used
to eliminate the needless computation of unwanted energies)
PRNT_SPINDN:
this switch is dependent on AISO and IPRINT, if AISO is set then the
spin density matrices for the components requested by AISO will be
printed out if PRNT_SPINDN is found in the $OUTPUT section of the input
deck. If(IPRINT.GE.1) then the full matrix is written out,
if (IPRINT.LE.0) then only the diagonal elements are written out.
PUHF_OVLAP:
this switch will cause the dominant UHF overlaps to be printed out.
i.e. the overlap between each occupied alpha MO and whichever
occupied beta MO that has the maximum overlap with it is outputted.
PUHF_BNDODR:
this switch is dependent on AISO , if AISO is set then the
bond index matrices for the components requested by AISO will be
printed out if PUHF_BDNODR is found in the $OUTPUT section of the input
deck.

if AISO.NE.0 in the $CONTROL input section then logical AISO is set to
.true. and turns on calculation of the isotropic hf splittings:
if PUHF_MULPOP and PUHF_NOAISO are found in the $outout section
of the input deck then MULISO is set to .false.
 Resonance Integrals: 
It is possible to alter Beta values for a calculation by
using the keywords 
LBETA(1) = a1 a2 a3 a4 a5
BETA(1) = sp1 sp2 sp3 d4 f5
The ai are the atomic numbers of the centers to be
changed and the spi, di, and fi are the new beta values
to be used. Note that the first three positions are
reserved for beta(s,p), the fourth for
beta(d) and the fifth for beta(f). This option is for ease
in reparameterisation of the resonance integral. If the values
for a1, a2 or a3 are greater than 1000, the the beta(p) of
atomic number (a3  1000) is set.
i.e.
LBETA(1) = 6 1006 0 26 0
BETA(1) = 17.0 12.0 0.00 23.0 0.00
or if only one value is changed, can use:
LBETA(4) = 26 < Iron
BETA(4) = 23.0 < Beta value for iron in ev.
For a more systematic study of parameters there is the
For IBETA = 0 (Default) or IBETA = 1
$BETAIN
! s s s s p p p p d d f f
! atomic numbers  THERE MUST BE 12
6 7 0 0 6 7 0 0 0 0 0 0
! THe 12 beta values corresponing to the above atomic numbers (eV)
$END
For IBETA = 5
$BETAIN
! s s s s p p p p d d f f
! atomic numbers  THERE MUST BE 12
6 7 0 0 6 7 0 0 0 0 0 0
! THe 12 beta values corresponing to the above atomic numbers (eV)
17.0 26.0 0.00 0.00 17.0 26.0 0.00 0.00 0.00 0.00 0.00 0.00
! The fkappa values fk(1,1) fk(1,2) fk(2,2): If one value is given, all
! three must be given, else all three are defaulted.
0.000000 0.000000 25.00000
! the fkappa values fk(1,3) fk(2,3) fk(3,3). If one value is given, all
! three must be given, else all are defaulted. If these values are
! given the previous fk(1,1) fk(1,2) fk(2,2) must have been given.
10.00000 10.00000 25.00000
$END
If IBETA = 6, then the exponents and the power can also be added
$BETAIN
! First the beta values, B(i)
! ssig ssig ssig psig psig psig ppi ppi ppi dsig dpi ddel
! atomic numbers  there must be 12
6 7 0 6 7 0 6 7 0 0 0 0
! The 12 beta values corresponding to the above atomic numbers (eV)
17.0 26.0 0.00 17.0 26.0 0.00 17.0 26.0 0.00 0.00 0.00 0.00
! Now the exponents,a(i)
! atomic numbers  there must be 12
6 7 0 6 7 0 6 7 0 0 0 0
! The 12 exponents corresponding to the above atomic numbers (Bohrs**1)
0.10 0.10 0.00 0.10 0.10 0.00 0.10 0.10 0.00 0.00 0.00 0.00
! The M value in H(IJ) = [R**M*(EXP(a(i)+a(j))*R**2)*(B(i)+B(j))/2]* S(ij)
1.00 < a floating point number
! If no value above is given it is defaulted to 0.5
!
! The fkappa values fk(1,1) fk(1,2) fk(2,2): If one value is given, all
! three must be given, else all three are defaulted.
0.000000 0.000000 25.00000
! the fkappa values fk(1,3) fk(2,3) fk(3,3). If one value is given, all
! three must be given, else all are defaulted. If these values are
! given the previous fk(1,1) fk(1,2) fk(2,2) must have been given.
10.00000 10.00000 25.00000
! the fkappa values fk(1,4) fk(2,4) fk(3,4) and fk(4,4). If one value is
! given, all four must be given, else all are defaulted. If these values are
! given the previous fk(1,3) fk(2,3) fk(3,3) must have been given.
5.000000 5.000000 5.000000 15.00000
$END
For testing additional empirical parameters, then the program will read
LBETB(1) = a1 a2 a3 a4 a5
BETB(1) = sp1 sp2 p3 d4 f5
The same as it has read LBETA and BETA in the $CONTROL block. See
subroutines paramd.f and paramx.f (in file paramd.f)
**For the regular CNDO and INDO models this adds a term proportional
to the weighted overlap to the oneelectron matrix, but not the
starting matrix.**
 Interaction Factors 
Generally the interaction factors are defaulted, as given above.
The defaults can be overwritten with:
INTFA(1) = 1.00 1.267 0.585 1.00 1.00 1.00, etc. as given above.
If this is done the order is f, f, f,
f, f, f, with all other
interactions taken = 1: i.e., f =1.0, etc.
For greater fleibility this can be overwritten as FACTR(1) =
$CONTRL
.
.
! INTFA(1) = 1.000 1.2670 0.5850 1.000 1.000 1.000 1.000
FACTR(1) = 1.0 1.00 1.267 1.0 1.0 1.0 1.0 1.0 1.0 1.00
FACTR(11) = 0.0 0.0 1.585 0.0 1.0 1.0 0.0 1.0 1.0 1.0
FACTR(21) = 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 1.0 1.0
FACTR(31) = 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0
.
.
$END
The first ten entries are the sigma interactions
ss spsig psigpsig, sdsig, psigdsig, dsigdsig, sfsig, etc.
The next ten are the pi
spispi, spippi, ppippi, etc. the first two are dummies
The next ten are the delta, the first five are dummies
The last ten are phi interactions, only the last is meaningful.
 Localization: 
In the Control Section:
LOC = 0 <<<<< DEFAULT for normal calculations
> 0  Orbital Localization to be performed (See Below)
The occupied orbitals of a molecule may be localized using a Boys, Fermi,
or double projector localization. A four digit number controls the
localization.
DIGIT1: First occupied localization  (1) BOYS
(2) FERMI
(3) DOUBLE PROJECTOR
(5) BOYS WITH ORBITAL CULL
(6) FERMI WITH ORBITAL CULL
(7) PROJ. WITH ORBITAL CULL
DIGIT2: Second occ. localization  (1) BOYS
(2) FERMI
DIGIT3: First virtual localization  (1) BOYS
(2) FERMI
(3) DOUBLE PROJECTOR
(4) I.V.O.
DIGIT4: Second virt. localization  (1) BOYS
(2) FERMI
(4) I.V.O.
Example: LOC = 3130 = double projector on occ. orbs followed by boys on
occ. orbs then double projector on virtual orbs.
Orbital cull is set only once, orbitals will be culled for the first
and second localization, and any virtual orbital localization.
for the boys localization the ci switch (line #1 kci) must be set to 66
if more than 12 orbitals are to be localized using the boys method.
both the fermi, double projector localization and orbital culls(removal)
need additional data. this data is stored on unit four (name.int),
the same unit used FOR internal coordinate data.
CULL
for an orbital cull
Line (1) : ********CLOCAL (*******CLOCALV for the virtual orbitals)
Line (2) : # of culling orbitals (I5)
Line (3) : orbital range (2I5)
Line (4) : # of basis function types in culling orbital, that number
OF basis function type numbers. (13I5)
Lines (3) and (4) should be repeated as many times as needed.
Example:
$LOCINP
********CLOCAL <Line(1)
2 <Line(2)
1 15 <Line(3)
3 1 2 3 <Line(4)
16 17 <Line(3)
2 1 2 <Line(4)
$END
NOTE: This orbital cull will remove orbital(s) OF px,py and pz
character from orbitals 1 thru 15, and remove orbital(s)
OF px, and py character from orbitals 16 and 17.
FERMI
for a Fermi localization:
Line (1) : ********LOCAL (*******LOCALV for the virtual orbitals)
Line (2N) : X,Y,Z (3F10.6) probe electron coordinates (# OF occupied
orbitals  # projected)
Example:
$LOCINP
********LOCAL <Line(1)
1.200000 1.000000 2.300000 <Line(2)
Line(2) is repeated as specified above.
$END
DOUBLE PROJECTOR
for a Double Projector localization:
Line (1) : ********PLOCAL (*******PLOCALV for the virtual orbitals)
Line (2) : # of projected orbitals (I5)
Line (3) : # of nonzero basis functions, b.f. #, b.f. coef
(I5,4(I5,F10.6))
Line (3A): continue line3 (first field blank) until # of nonzerO
basis functions is reached.
continue lines (3) and (3a) until # of projected orbitals is reached.
Example:
$LOCINP
********PLOCAL <Line(1)
3 <Line(2)
2 1 1.0 2 1.0 <Line(3)
5 3 0.3 4 0.4 6 0.6 9 0.9 Line(3)
10 0.8 <Line(3a)
1 11 1.0 <Line(3)
$END
In the above example three mo's are projected from those that are
occuppied. The first sought is that which is most like a 50%50%
mix of a.o.'s 1 and 2, the next is an m.o. that is 0.3 a.o. #2,
0.4 a.o. #4, etc., and the last m.o. is as pure a.o. 11 as can be
projected. The remaining orbitals not projected are as much like the
original canonical m.o.'s as possible, with those specified projected
out.
If the double projector is used, there is the additioal option
of specifying a file produced by ZINDO from a previous run. The
file must be suffixed with .vec8, as is with the usual restart from
input vecors. Below the file "test.vec8" supplies the vectors to be
projected from the current run.
$LOCINP
********PLOCAL < the occuppied set 2A8
FILE = test < the vectors are in "test.vec8"
4 < the homo to homo3 mo in test.vec8, I5
8 5 1 < 8 basis fns in the projecting set, the,
********PLOCALV homo is the 5th, ao #1 of this set is
FILE = test the #1 ao of the current calculation, 3I5.
3 < In the virtual space the lumo to
8 5 1 lumo + 3 will be used. Again 8 basis fns
$END in the.vec8 file, MO 5 is the homo, etc.
After specifying which ao of the projecting orbitals match the ao
of the present run, the orbitals of the projecting and present runs
must be in the same order.
Current Run Ao.number
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Saved on test.vec8
1 2 3 4 5 6 7 8
After the 8th ao in this case, no projecting orbitals are specified.
 Dipole integrals 
In the $CONTRL block
DIPOLE = 1, is the default, and the dipole moments are calculated
with the one center charge and polarization integrals.
DIPOLE = 2, The dipole integrals are calculated including the two
center bond terms, and these integrals are then sym.
orthogonalized if a ZDO model is used.
DIPOLE = 3, Only the one center diagonal (or charge term) is used.
DIPOLE =11,12,13 as above, but quadropole moment is also calculated
DIPOLE =21,22,23 as above, but octopole moment is also calculated
 Fragment Orbital Analysis: 
First run the composite (super) molecule. This will require
In the $CONTRL block:
FRAG = 0 < If none. This is the default.
= 1 this is a fragment orbital calculation
= 2 the fragment being calculated is an atom.
Additional input required.
Then a FRAGIN block is required.
$FRAGIN
free ethylene < a title
ethyl < The name of the data for the composite molecule that
will be analized into fragments. This is the ONAME
of this job, and will produce a file called "ethyl.frag"
methyl1 < the name of the first fragment, and this must be the
ONAME of the ZINDO job that runs this fragment. The
output for the fragment analysis will be methyl1.frag.
methyl2 < the name of the second fragment: Give the names of
all fragments
$END
NOTE: The composite (larger) molecule must be the first parameter
and the fragments must follow in the same order as they do in the
composite molecule.
This first job will output a ONAME.PMO (or, as above, ethyl.PMO).
Then each fragment must be calculated, with the same coordinates as
in the composite molecule. In the control block, FRAG =1 or 2 (see below)
and the FRAGIN block needs only a title. Each run will produce a
ONAME.frag file. These names must agree with those supplied in the
FRAGIN block of the initial run.
If a fragment is a single atom then the switch must be 2,
this will ensure that the aos are in the correct order.
Otherwise use 1.
The actual fragment analysis is done by a separate program
called frganl.f. It uses some files generated by the ZINDO
program.
Once you have obtained .frag files for the molecule and all
of its fragments you can start the analysis by entering
frganl < (moleculename).PMO
Output will appear in a file ONAME.fmout
where ONAME is that of the compound molecule.
example: Job 1.
$TITLEI
ethylene fragment test
$END
$CONTRL
SCFTYP RHF RUNTYP ENERGY ENTTYP COORD UNITS ANGS
INTTYP 1 IAPX 3 NAT 6 NEL 12
MULT 1 ITMAX 30 FRAG 1
SCFTOL 0.000200
ONAME = ethyl
$END
$DATAIN
2.000000 0.645000 0.000000 6
2.000000 1.190000 0.943968 1
2.000000 1.190000 0.943968 1
2.000000 0.685000 0.000000 6
2.000000 1.230000 0.943968 1
2.000000 1.230000 0.943968 1
$END
$FRAGIN
! ***** Molecular fragmentation specification *****
free ethylene
ethyl
methyl1
methyl2
$END
Job 2.
$TITLEI
meth1 first fragment
$END
$CONTRL
SCFTYP RHF RUNTYP ENERGY ENTTYP COORD UNITS ANGS
INTTYP 1 IAPX 3 NAT 3 NEL 6
MULT 1 ITMAX 30 FRAG 1
SCFTOL 0.000200
ONAME = methyl1
$END
$DATAIN
2.000000 0.645000 0.000000 6
2.000000 1.190000 0.943968 1
2.000000 1.190000 0.943968 1
$END
$FRAGIN
! ***** Molecular fragmentation specification *****
free ethylene
methylene
$END
Job 3.
$TITLEI
meth2 second fragment
$END
$CONTRL
SCFTYP RHF RUNTYP ENERGY ENTTYP COORD UNITS ANGS
INTTYP 1 IAPX 3 NAT 3 NEL 6
MULT 1 ITMAX 30 FRAG 1
SCFTOL 0.000200
DYNAL(1) = 0 2 1 0 0 0 0
ONAME = methyl2
$END
$DATAIN
2.000000 0.685000 0.000000 6
2.000000 1.230000 0.943968 1
2.000000 1.230000 0.943968 1
$END
$FRAGIN
! ***** Molecular fragmentation specification *****
free ethylene
methylene
$END
Job 4.
frganl < ethyl.PMO &
 Partial, local or Projected Density of States: 
In the $CONTRL block the keyword LDOS = 1 will cause a density of staes
analysis to be performed directly after the Mulliken Population Analysis.
The output contains the number of molecular orbitals within a mo energy
interval, and how much s. p d and f character this band of mo's has.
 Memory Management 
By default, the program attempts to grab as much memory as possible for
various arrays, and internal I/O files. In many situations this may not
be desirable. In such situations the user can limit the available
memory by specifying MEMORY= in the $CONTRL block. This puts an upper
limit (in bytes) on the memory the program can use. Please note that
the executable itself, and statically allocated arrays take about 12
megabytes (possibly more in the future), which comes in addition to
the amount allowed by MEMORY switch.
Various I/O units can also be either stored in the memory or on
disk. The default is always in memory, if there is memory
available. If there is a shortage of memory, the user can make
possibly better choices than the primitive prioritybased algorithm in
fileop.f . For such purpose, add the block $EXTERN into the input
file, and list the names of the units (omitting the initial IO_ and
possibly some characters from the end  see program output!) which can
be made external. The keyword 'ALL' makes all units external, thus
conserving the most of memory. Please note that the latter may also
make the program very slow.
For more information about the I/O units, see programmers manual,
units.doc, and study the relevant sections of output.
 Self Consistent Field Convergence 
In general, the SCF is accelerated by using the dynamic damping
of Zerner and Bacon, and Zerner and Hehenberger. If the program
detects a problem with convergence, the DIIS method of Pulay
is switched on. The default value for the DIIS to be put into
operation is when the DIFF in the energy from cycle to cycle
is less than 10**(7) Hartrees. The default is equivalent
to
DIIS = 7
in the $CONTRL section. This value can be changed: ie,
DIIS = 6 < when DIFF = 10**(6) start DIIS procedure.
etc.
IDD2 = 0 will shut the DIIS option off. This is often a good
idea if the self consistent field calculation does not
converge.
IDD2 = 1 is the default. See above for DIIS control.
IDD2 = 2 scales the density matrix elements for atoms further
than 5 Angstroms for the first 14 SCF cycles.
IDD2 = 20 is level shifting, 1.0 Hartree/(cycle number) until
the 10th cycle, then the level shift is cut off.
The level shifting will not be followed by DIIS.
= 21 is level shifting as above. If appropriate, DIIS
will be invoked to accelerate convergence.
 ESR COUPLING 
CLOPPA J coupling manual
(version 1)
1) Line 1
Molecule name (oname). It comes from ZINDO.
2) Line 2
a) To get the total coupling constants values write: 981
and go to 11) and do n=m=0.
b) To localize write: II1 where II is the total number of
atomic orbitals (AOs).
3) Line 3 and the following n
Atoms with their corresponding AOs. Given by ZINDO.
4) Line n+1
Write IIJJKK where II is the number of occupied molecular
orbitals (MOs), JJ the number of virtual MOs and KK the
number of localizations. II and JJ can be chosen belonging
to a fragment to calculate their contribution to total J orB
they can be the total number (IPPP).
5) Line n+2
Write LLIIJJ where LL is the number of AOs used in this
localization step and II and JJ the number asigned to the
occupied and virtual MOs of this step. It is used a decreasing
order.
B
6) Line n+3,......,n+m
Atom numbers and its corresponding AO.
7) Line n+m+1
Blank
8) Line n+m+2
Repeat from 5) to 7) to complete the KK steps in line n+mm (last
blank line).
9) Line n+mm+1
Occupied MOs from 01 to total number.
10) Line n+m+2
Virtual MOs from 01 to total number.
11) Line n+m+3
Keyword for propagator: propa3 for triplet and propa2 for singlet.
12) Line n+m+4
Write: 10001 AABB, where AA and BB are the atom numbers to get
J(AA BB) CLOPPA.
13) Line n+m+5
Blank
14) Line n+m+6
Write 99.
Example
a)To calculate only total J values
ch4 line 1
981
propa3
99 line 6
b)To get J CLOPPA values
ch4 line 1 (oname)
081 line 2
1 1
1 2
1 3
1 4
2 1
3 1
4 1
5 1
040404
050404
1 1
1 2
1 3
1 4
2 1
050303
1 1
1 2
1 3
1 4
3 1
050202
1 1
1 2
1 3
1 4
4 1
050101
1 1
1 2
1 3
1 4
5 1
01 02 03 04 05
01 02 03 04 05
propa3 propa2
10001 0201
000000
99
 Spin Orbit Double Group 
These are the switches defined in subroutine CIDG see above example
of double group CI.
NROOTS NUMBER OF ROOTS OF THE CI MATRIX
ICIWRT CI WRITE OPTION
ICIWRT = 0 PRODUCTION RUN, MINIMUM OUTPUT
= 1 PRINTS ALL OF CI VECTORS
= 2 PRINTS THE CI MATRIX
= 3 PRINTS THE COMPLETE LIST OF CI DATA
ICIPUN CI PUNCH OPTION
ICIPUN = 0 DO NOT WRITE THE CI VECTORS
= 1 WRITE THE CI VECTORS ON IUNT2B
INTWRT INTEGRAL WRITE OPTION
INTWRT = 0 PRODUCTION RUN, MINIMUM OUTPUT
= 1 WRITES OUT THE ONEELECTRON
INTEGRALS
= 2 CALLS ANALYZ TO PERFORM
PERTURBATIVE ENERGY ANALYZATION
 ORIGINALLY:
= 2 WRITES OUT THE ONE AND TWO
ELECTRON INTEGRALS
IHAMRD HAMILTONIAN READ OPTION
IHAMRD = 0 THE HAMILTONIAN MATRIX IS COMPUTED
= 1 THE HAMILTONIAN MATRIX IS READ
IN FROM IUNT2A
IANALZ VECTOR AND ENERGY STATISTICS OPTION
IANALZ = 0 NO VECTOR OR ENERGY BREAKDOWN
= 1 PROVIDE VECTOR AND ENERGY BREAKDOWN
MAXNE0 MAXIMUM NUMBER OF NONZERO HAMILTONIAN MATRIX ELEMENTS
PER ROW; IGNORED UNLESS IHAMRD = 1, IN WHICH CASE THE
VALUE IS OBTAINED FROM THE PREVIOUS RUN'S OUTPUT
This ends the data on this line.
KSYM DOUBLEGROUP SYMMETRY OF STATE
KSYM = 1 TO 5 FOR C2V, D2
KSYM = 1 TO 10 FOR D2H
NTOTFG NUMBER OF SPATIAL CONFIGURATIONS
 Historical switches (Probably not functional) 
ISAVE = 0 <<<<< DEFAULT Normal calculation
= 1  Save ground state on tape unit 14
= 2  Recall ground state calc from tape for CI, unit 14
(Unit 14 and Unit 16 must be cataloged data sets!)
= 3  Recall ground state calc. as above, and recover
transformed integrals from Unit 2 (direct access)
= 4  Recall 1 AND 2 Electron files and calculate electroni
energy. Unit 11 and Unit 16 Must be saved.
= 5  AB INITIO...........
Recall 1 and 2 electron files from MOLECULE program
and calculate electronic energy.
Unit 8 (MOLECULE integrals) must be saved.
IPUN = 0 <<<<< DEFAULT Normal calculation
= 1  Punch charge and bond order matrix (Unit 7)
= 2  Read previous density matrix to start iteration
= 3  Read previous density matrix to start iteration and
punch the calculated density matrix (Unit 7)

 Some program arrays 

NEL = Number of electrons in the molecule
NB = Number of basis orbitals
EIG = Eigenvalue vector
C(I,J) = Matrix of eigenvectors, the Ith AO in the Jth MO
is stored as C(L) where L=I+J(J1)*NB
NU(I) = Atom to which the Ith orbital belongs
NW(I) = Type of atomic basis orbital of Ith Atomic orbital
CO(J,I) = Coordinates of atom I
NTYP(I) the type of Ith atom
TYPE=1 for nS basis
=2 for nS,nP basis
=3 for nS,nP,nD basis
=4 for nS,nP,(n1)D basis.
=5 for nS,nP,(n1)D,(n2)F
NP(I) = Principal quantum number of Ith atomic orbital
ALPHA(I,J)=Jth orbital exponent of Ith AO
J=46 expansion coefficient for Ith AO
NA=Number of atoms
KAN(I) = Atomic number of Ith atom

 Control words and data blocks 

The following describes the structure of the blocks in the data
set. Note that most of them require the data in format once the
data begins. That is, you can not put comments in the middle of
the cartesian coordinates for example.
The INPUT deck.
col 12345678
BLOCKS: $XXXXXX

NAME'separator'VALUE(S)

 valid 'separator' is an equal (=) sign
 or a space ( ).

ARRAYS are NAME(IND)'separator'a b c ... where
 NAME(IND) = a,
 NAME(IND+1) = b, etc

LOGICAL VALUES are .TRUE. .FALSE.

COMMENTS begin with "!", go to the
 end of the line and can occur
 anywhere in the INPUT deck, except
 inside of a bunch of formatted data.

$END
The BLOCK descriptor, $XXXXXX, must begin in column 2, NAMES can
occur anywhere followed by VALUES.
Here is a list of all WORDS which are recognized by the NAME directed
input routine of ZINDO and should be in the $CONTRL group.
(THE FOLLOWING LIST IS OUTDATED AND SHOULD BE CONSIDERED HISTORICAL)
Logicals: LMASS 
LEXACT  Geometry Optimization
LINV 
Integers: DYNAL( )  Basis set definition + C.I. spec.
PUN
VEC  Read or save vectors
IDD1 =9 Calculates the kinetic energy
AND TRANSFROM IT TO ZDO
=1 Calculates the electronnuclear attraction
explicitly.
IDD2 .ne. 0 => DIIS is possible.
=2 Dynamic damping, of use in large systems
IBETA =0 Sum form for beta,default.
=1, product form for beta
=2, sum form with a+b/R
=3, product form with a+b/R
=5, beta is of pair type.
=6, beta = r**n(exp(a*r**2)*betave
INTTYP  Integral type
IAPX  Approximate theory
III 
MODE  Geometry Optimization
ITYPE 
ISRCH 
IPRINT  Print switch
NAT  number of atoms
NEL  number of electrons
IELEC  electrostatic potentials
MULT  multiplicity
ITRIP  triplet parametrization
ISW2  type of configuration mixing
ICHARG
NMR  nmr option
PTCG  number of point charges
SPNORB  spin orbit option
ITMAX  maximum number of iterations
DIIS  scf convergence needed before DIIS is used
ISCRF  self consistent reaction field
NMU  scrf
LOC  localization
FRAG  molecular fragmentation
NOP  number of open shell orbitals.
NDT  default coupling constants for ROHF or not
MIM( )  total number of orbitals per open shell
LBETA( )  atomic number of atom for new betas
POLAR  the switch for calculating polarizabilities.
DIPOLE  1=Default=1center integrals,2=twocenter
integrals.
Reals: SCFTOL  convergence criteria
RHO 
SIGMA  Geometry optimization
PHASE 
STPTOL 
STPTLD 
A0 
EPS 
XND  Self consistent reaction field
RHO 
AMUSV 
DIPSV 
DIGPR( )
INTFA( )  interaction factors
FOP( )  number of closed shell electrons, # e/open shell
AR( )  alpha coupling contants
BR( )  beta coupling constants
BETA( )  new beta parameters
Characters: SCFTYP  RHF etc
RUNTYP  ENERGY etc
ENTTYP  atomic coordinates or zmatrix
UNITS  angstroms or bohr
ASYM  molecular symmetry
ONAME  output file name
FNAME  fragment name
Recognized BLOCKS in the data set are 
$TITLEI
$CONTRL
$DATAIN  formatted
$CIINPU  formatted
$ASSINP  formatted
$LOCINP  formatted
$FIXCOR  formatted
$MIXCOF  formatted
$PTCHGI  formatted
$FRAGIN  formatted
$SCRFIN  formatted

