One of the most important thing in DFT calculation is using an appropriate functional/basis function. Depending on this choice, the result may be reversed.
In general, it is safe to follow prior research. In addition, if you are an experimental chemist and not familiar with the computational chemistry, use “B3LYP/6-31G(d)” for the first choice, which is not too bad option.
However, when computing for a new system, how do you choose a functional?
reference the benchmark study
Benchmark papers are regularly posted in Journal of chemical theories and computation (JCTC) published by ACS. Benchmark papers are database-based papers on the results of calculations using a combination of plural functionals and basis sets for a specific reaction or chemical events.
By reading the benchmark paper, you can roughly grasp the characteristics of each functional. For example, A functional is not appropriate hydrogen transfer reaction, B functional can not estimate weak interaction such as hydrogen bonding, C functional overestimate the enthalpy, D functional is good for Rh/Ru catalized -bond activation etc.
Comparing with the experimental data
In the calculation of finding the transition state structure, it is possible to narrow down the functionals based on whether the actual experiment result agrees with the calculation result. For example, very rough comparison, if a very high activation energy is obtained as a calculation result although it is a room temperature reaction, it might not be an inappropriate functional…
Comparing with the Crystal Structure
Also, if the compound has a X-ray crystal structure, it may be possible to ascertain the validity of the functional by comparing it with the structure optimized by DFT calculation.
This method is widely used why even if the calculation result in the gas phase and the crystal structure are compared, it shuold not match. . .
Comparing the UV or NMR Spectra
In the case you are working on the photochemistry, it is better to compare the computed UV spectra with the experimental data. You can use the functional that has the smallest error between the computational and the experimental results.
Read a guid book
A book called “Which functional should I chose? (pdf)” is available.
- Why is the calculation results of Gaussian and GAMESS different?
- Gaussian Poem
- Gaussian16 Benchmark
- Gaussian09 Secret Code ? Morokuma Energy Decomposition
- Transition State? Imaginary Frequency?
- Ryzen Threadripper Gaussian16 Benchmark
- When the IRC calculation doesn’t work…
- How many kcal/mol ?
- Why atomic units (a.u.) are used in QM calculation?
- Building a PC with Ryzen Threadripper 1950x for Scientific Computation