Search:

# Format Of ECPDATA File

The parameters of the effective core potentials are specified in the file ECPDATA in the same way as the basis sets are gathered in GENBAS. As an example, the entry for Cu is given below.

ECP parameter sets for Cu

  *
CU:ECP-10-SK
# ECP BY STEVENS/KRAUSS FOR CU - 10 CORE ELECTRONS - LMAX = 2
*
NCORE = 10    LMAX = 2
d
-10.00000000    1  511.9951763
-72.55482820    2   93.2801074
-12.74502310    2   23.2206669
s-d
3.00000000    0  173.1180854
23.83518250    1  185.2419886
473.89304880    2   73.1517847
157.63458230    2   14.6884157
p-d
5.00000000    0  100.7191369
6.49909360    1  130.8345665
351.46053950    2   53.8683720
85.50160360    2   14.0989469
*
CU:ECP-18-SK
# ECP BY STEVENS/KRAUSS FOR CU - 18 CORE ELECTRONS - LMAX = 3
*
NCORE = 18    LMAX = 3
f
-18.00000000    1  359.2137111
-119.92593970    2   67.5347369
-29.55328670    2   14.7222923
-10.28924330    2    3.9975558
-.78363630    2    1.1889410
s-f
3.00000000    0   19.6202650
20.15792750    1    5.1604389
34.50019060    2    1.2306099
-18.98120030    2    1.0850105
p-f
5.00000000    0   31.9385762
20.60853280    1   14.9202125
56.00168880    2   15.6835232
57.21701070    2    4.9311614
7.71778780    2    1.0622167
d-f
.25986160    2    5.1159991
-.46216800    2     .7396784
*


The first line contains a single star. The name of the data group starts with the element symbol followed by the ECP nickname. It is useful to give a comment introduced by the #--symbol in the line below in order to indicate the origin of the ECP. The actual ECP data is given in between two lines with a *-symbol. The first line specifies the number of core electrons described by the ECP (NCORE) and the maximum angular momentum number of the projection operators (LMAX) in integers i.e. s=0, p=1, d=2, etc. These are followed by the description of the effective core potential which consists of the angular momentum numbers and by the analytical representation of the operator. The latter includes the coefficients {$c_m$}, the exponents {$N_m$} of {$r$} and the exponents {$\alpha_m$} of the gaussians in the expression for {$U_l(r)$}:

{$U_l(r) = \sum_m{c_m e^{\alpha_m r^2} r^{N_m}}.$}