Using Non-Standard Basis Sets
The authors of the paper Emily is basing her project on (Journal of the American Chemical Society 2000, 122, 8990) took a rather creative approach in their calculations on the reaction of H2COO and SO2. They decided to put diffuse functions only on the S atoms. (When they did their calculations almost ten years ago, CCSD(T) single point calculations were extremely difficult, so they tried to keep the basis sets as small as possible. Hence their decision to put diffuse functions on only the biggest atoms in their system.)
So, how can use different basis functions for different atoms? The following is what two typical .com files for Emily’s H2COO + SO2 calculations would look like:
#N
B3LYP/gen OPT FREQ
SO2
(B3LYP)
0
1
S
O
1 B1
O
1 B2 2 A1
B1
1.4875000
B2
1.4875000
A1
109.47122
S
0
6-31+G(d,p)
****
O
0
6-31G(d,p)
****
and….
#N
B3LYP/gen OPT=(TS,tight,calcall,NoEigenTest)
SO2
TS (B3LYP)
0
1
C
O
1 B1
S
2 B2 1 A1
O
3 B3 2 A2 1 D1
O
1 B4 2 A3 3 D2
O
3 B5 4 A4 5 D3
H
1 B6 2 A5 3 D4
H
1 B7 2 A6 3 D5
A5
112.39518
A6
105.55079
B1
1.2877321
B2
2.1200000
B3
1.5240000
B4
1.3370000
B5
1.4650000
B6
1.1770000
B7
1.2752439
D1
-16.231947
D2
53.675898
A1
100.37985
D3
-124.93820
A2
93.002534
D4
-110.23403
A3
117.29470
A4
133.01522
D5
134.67090
S
0
6-31+G(d,p)
****
O
C H 0
6-31G(d,p)
****
Note that in both cases, one uses the gen keyword instead of naming a specific basis set. Then, at the end of the .com file, one specifies the basis set one wants to use for each of the atoms that appear in the geometry definition section.
(By the way, Emily, the SO2 transition structure calculation leads to an optimized geometry virtually identical to what was reported in the Aplincourt paper!)