PBE functional which belongs to the generalized gradient approximation functional for exchange correlation energy is quite popular these days as it provides reliable results. B3LYP approach which belongs to the hybrid approximation for the exchange-hybrid correlation functional is more famous already.
How to rationalize the behaviour of the two in the context of the geometry optimization and optical property calculations results using PBE, GGA/PBE and B3LYP functionals as implemented in Gaussian 09 and CASTEP, Quantum espresso etc. softwares.
Hi Sapna, you can read some benchmark studies about DFT. Frankly speaking, I have a few benchmark papers about the evaluation of DFT you can visit my account and have a look.
Hi Sapna. Both PBE and B3LYP provide reasonable bond lengths and angles within 0.02 A and 2 degrees, respectively. PBE (and any GGA) is known to underestimate HOMO-LUMO gaps, unlike hybrid functionals, such as B3LYP. Therefore, B3LYP is better suited for optical property calculations. The downside is that hybrid functionals tend to be ~3 orders of magnitude more expensive than GGA's in the plane wave basis set (VASP). Computational cost for the hybrid is considerably lower in the atomic basis set (Gaussian).
Thanks to all for quick respone and wonderful suggestions.
Yes Alexander, I completely agree with your statement. I have seen in my results that PBE underestimates the HOMO-LUMO gaps. But due to time constraint I preferred to use GGA/PBE for studying optical properties of my molecules. B3LYP was little time consuming. As far as qualitative results are concerned, I think PBE ( and any GGA) would also serve the purpose well.
Hi, Amrutha
I would suggest you to continue related questions on the same page where you had asked the query.
I had tried IR calculation but it did not work out. One thing to be noted is that CASTEP implements norm-conserving pseudopotentials only for IR, Raman, NMR calculation..
so. We are comparing two different methods.
B3LYP is very unusual in solid state physics (most of the time you will use HSE/PBE0), and contrarly to PBE is an hybrid functional (more expensive, inclusion of experimental fitting).
When you chose different GGA functional Wahid ullah Khan
, you are just changing some of the parameters in the definition of the XC functional.PBEsol for example, has some different parametrization that makes it more suitable to describe materials in the solid state.
There are not specific difference, depends on the functionals you consider and you need to read the paper associated to them.
Sapna Bondwal there is not a specific choice of functional that is correct. Depends what you need to calculate with respect to experimental evidence.
GGA can return completely wrong results, as well as B3LYP.
Wahid ullah Khan
I think it is depends on your system and your needs. As far as I know, TZ2P gives a double polarization calculation, but if your system does not require a large basis set, using a lower basis set would be sufficient, and of course, a reduced amount of computational cost.
Alexander Mironenko
Can you please suggest any references supporting the statement "Both PBE and B3LYP provide reasonable bond lengths and angles within 0.02 A and 2 degrees".
I have solved crystal structure of a cocrystal from powder-XRD and to confirm the proton transfer (salt or cocrystal), I have applied GGA-PBE functionals. Please help.
- Badges
- Science topic
- Similar topics
- Linguistics
- Composition Studies
- Publication