Related Experiment Videos
The behavior of density functionals with respect to basis set. I. The correlation consistent basis sets.
1Department of Chemistry, University of North Texas, Denton, TX 76203-5070, USA.
The Journal of Chemical Physics
|October 16, 2004
Summary
Density functional theory (DFT) accuracy for molecular geometries and atomization energies was tested. Hybrid functionals like B3LYP showed smooth convergence with basis set size, unlike nonlocal functionals.
Area of Science:
- Computational chemistry
- Quantum chemistry
Background:
- Density functional theory (DFT) is a widely used quantum mechanical modeling method.
- Accurate prediction of molecular geometries and atomization energies is crucial for chemical research.
Purpose of the Study:
- To assess the accuracy of six common DFT functionals for predicting molecular geometries and atomization energies.
- To evaluate the impact of basis set size on the accuracy and precision of these DFT calculations.
- To compare the performance of hybrid and nonlocal DFT functionals.
Main Methods:
- Calculated geometries and atomization energies for first-row closed-shell molecules using six DFT functionals (B3LYP, B3PW91, B3P86, BLYP, BPW91, BP86).
- Employed correlation-consistent basis sets (cc-pVxZ and aug-cc-pVxZ, x=2-5).
- Performed statistical error analysis to determine accuracy and precision.
Main Results:
- B3LYP and B3PW91 functionals demonstrated smooth convergence towards the Kohn-Sham limit as basis set size increased.
- Nonlocal functionals (BLYP, BPW91, BP86) showed relative insensitivity to the choice of basis set.
- Accuracy and precision varied among the tested functionals and basis sets.
Conclusions:
- Hybrid DFT functionals, particularly B3LYP and B3PW91, offer reliable and systematically improvable results with increasing basis set size.
- Basis set choice is less critical for nonlocal functionals, but their convergence behavior differs.
- The study provides insights into selecting appropriate DFT functionals and basis sets for molecular property predictions.