Can popular global hybrid functionals like B3LYP, M062X, PBE0 or (PBE1PBE) or pure Generalized gradient (GGA) Functionals predict accurate geometries of organic molecules ?
Why do these Functionals fail to predict accurate geometries ?
On what factors do their accuracy depend upon ?
How can we overcome these shortcomings ?
We have tried to answer all these questions in our recent publication
Article Limitations of Global Hybrids in Predicting the Geometries a...
yes but you should chose the right factors like energy cut off, wave function basis set and potential.
some times LDA give more accurate than GGA specially in metal oxide material
Dear Vikramaditya,
We normally analyze the accuracy of different DFT methods for geometry optimization by comparing DFT geometries with reported X-ray crystal structures that are similar to the molecules we want to model. This is especially important when working with relatively large organic systems (more than 50-60 atoms) since normally the kind of publication that you mentioned only include relatively small systems with no complex networks of noncovalent interactions, which can alter the geometries in a great extent.
Dear Juan V. Alegre-Requena
Although we have considered smaller/medium size systems, the conclusions are valid even for larger systems. If you go through the full paper you can understand how the contribution of HF exchange and the DFT exchange incorporated in the DFT Functional influences the delocalization patterns there by impacting the torsional energies which in turn impacts the geometries.
- Badges
- Science topic
- Similar topics
- Political Science
- International History and Politics
- Globalization