Ali,
Your fragment guess is likely to get in trouble because of linear dependencies.
To avoid this you have two options
1) Ask NWChem to avoid to treat the linear dependencies (I would not suggest this option) or
2) Modify the basis set to avoid the linear dependence problem. To be more clear, I am suggesting you to
modify the iron basis set by deleting the most diffuse p function, and splitting a contraction of 2 gaussian in 2 separate (single gaussian) contractions
Fe S
20.5130280 0.2347140
9.7767920 -0.8775270
4.5735990 -0.3416360
Fe S
1.9499100 1.0000000
Fe S
0.7664880 1.0000000
Fe S
0.1007980 1.0000000
Fe S
0.0375160 1.0000000
Fe S
0.0100000 1.0000000
Fe P
63.1253260 0.0271470
11.6568550 -1.0118750
Fe P
5.1786610 0.2151680
Fe P
2.0355320 0.8250950
Fe P
0.7088180 1.0000000
Fe P
0.1101680 1.0000000
#Fe P
# 0.0294440 1.0000000
Fe D
33.6667900 0.0365290
10.2356950 0.1828250
3.4664880 0.4370690
1.1272960 0.5751780
Fe D
0.3082470 1.0000000
Fe D
0.0800000 1.0000000
Fe F
2.0000000 1.0000000
Cheers, Edo
|