王昱喬-第五週作業
Use Spartan or WebMO to calculate the electron affinity of atoms with atomic number 1~10, by using CCSD(T)、B3LYP and MP2 theories with 6-31G(d,p)、6-31++G(d,p) 、6-311G(2df,2p) and aug-cc-pVTZ basis sets, and do some simple statistical analysis.
Data link:Google Drive
electron affinity & mean absolute error ( unit: kcal/mol )
1, use basis set 6-31G(d,p) for different theories
|
|
H |
He |
Li |
Be |
B |
C |
N |
O |
F |
Ne |
|
MP2/6-31G(d,p) |
39.4 | 860.6 | 4.4 | 45.3 | 33.2 | 5.4 | 54.2 | 14.0 | -24.7 | 1067.9 |
|
B3LYP/6-31G(d,p) |
24.1 | 829.9 | -3.7 | 34.3 | 25.0 | 1.3 | 35.0 | 5.0 | -24.2 | 1009.1 |
|
CCSD(T)/6-31G(d,p) |
37.6 | 862.8 | -5.3 | 47.1 | 35.8 | 9.0 | 54.2 | 16.4 | -19.9 | 1063.3 |
|
experimental value |
-17.4 | 5.0 | -14.3 | 4.3 | -5.5 | -29.1 | 1.7 | -33.7 | -77.0 | 6.9 |
| mean absolute error | 51.1 | 846.1 | 12.8 | 37.9 | 36.8 | 34.3 | 46.1 | 45.5 | 54.1 | 1039.9 |
2, use basis set 6-31++G(d,p) for different theories
|
|
H |
He |
Li |
Be |
B |
C |
N |
O |
F |
Ne |
|
MP2/6-31++G(d,p) |
-3.0 | 127.9 | -5.7 | 15.3 | 1.2 | -22.7 | 20.1 | -25.1 | -78.5 | 172.0 |
|
B3LYP/6-31++G(d,p) |
-20.5 | 117.4 | -12.6 | 5.3 | -9.6 | -31.4 | -3.2 | -37.4 | -81.0 | 157.6 |
|
CCSD(T)/6-31++G(d,p) |
-16.8 | 68.1 | -14.2 | 8.5 | -5.3 | -28.1 | 6.9 | -30.7 | -76.4 | 140.6 |
|
experimental value |
-17.4 | 5.0 | -14.3 | 4.3 | -5.5 | -29.1 | 1.7 | -33.7 | -77.0 | 6.9 |
| mean absolute error | 6.0 | 99.5 | 3.5 | 5.4 | 3.7 | 3.2 | 9.5 | 5.1 | 2.0 | 149.8 |
3, use basis set 6-311G(2df,2p) for different theories
|
|
H |
He |
Li |
Be |
B |
C |
N |
O |
F |
Ne |
|
MP2/6-311G(2df,2p) |
7.3 | 478.8 | -4.7 | 22.6 | 9.7 | -12.8 | 41.1 | 3.0 | -42.8 | 859.8 |
|
B3LYP/6-311G(2df,2p) |
-4.8 | 457.9 | -9.8 | 16.8 | 6.4 | -12.6 | 26.9 | -3.5 | -41.7 | 813.0 |
|
CCSD(T)/6-311G(2df,2p) |
-16.8 | 68.1 | -14.2 | 8.5 | -5.3 | -28.1 | 6.9 | -30.7 | -76.4 | 140.6 |
|
experimental value |
-17.4 | 5.0 | -14.3 | 4.3 | -5.5 | -29.1 | 1.7 | -33.7 | -77.0 | 6.9 |
| mean absolute error | 12.6 | 329.9 | 4.7 | 11.7 | 9.1 | 11.3 | 23.3 | 23.3 | 23.4 | 597.6 |
4, use basis set aug-cc-pVTZ for different theories
|
|
H |
He |
Li |
Be |
B |
C |
N |
O |
F |
Ne |
|
MP2/aug-cc-pVTZ |
-10.1 | 68.3 | -8.0 | 9.1 | -4.8 | -29.2 | 14.3 | -32.7 | -83.9 | 141.3 |
|
B3LYP/aug-cc-pVTZ |
-21.0 | 61.9 | -12.8 | 2.2 | -10.6 | -31.6 | -4.9 | -38.7 | -81.4 | 130.2 |
|
CCSD(T)/aug-cc-pVTZ |
-16.8 | 68.1 | -14.2 | 8.5 | -5.3 | -28.1 | 6.9 | -30.7 | -76.4 | 140.6 |
|
experimental value |
-17.4 | 5.0 | -14.3 | 4.3 | -5.5 | -29.1 | 1.7 | -33.7 | -77.0 | 6.9 |
| mean absolute error | 3.8 | 61.1 | 2.6 | 3.7 | 2.0 | 1.2 | 8.2 | 3.0 | 4.0 | 130.5 |
experimental value website: Computational Chemistry Comparison and Benchmark DataBase
Organize the calculated data of ionization energy, and compare it with the experimental value
- Conclusion
1,When we pluse the diffuse function to noble gas is more useful to reduce the error than pluse the polarization GTO.
2,Compare to last week homework, we can know that the noble gas error might come for the calculation from the anion.