O3LYP and LC-wPBE give different results from G09

From NWChem

Viewed 1979 times, With a total of 3 Posts
Jump to: navigation, search
Forum >> NWChem's corner >> General Topics

Clicked A Few Times
Threads 2
Posts 6
10:51:45 AM PDT - Mon, Jun 4th 2012
Hi, I am using NWCHEM 6.1 to get a single point energies for Ni(CO)4. I am using a Huzinaga-Dunning DZP basis set. To check my definitions, I compared the single point energies with Gaussian 09. They give different results. For LC-omegaPBE, I am using the same definition as in NWCHEM documentaion. I used same basis sets, same convergence criterion and fine grid for NWCHEM calculations whereas ultra fine grid in G09. I made sure that the geometry and units are same for both programs. For other functionals that I compared, these conditions give similar single point energies in NWCHEM and G09 (within 0.5 kcal). Can someone point to if I am defining the following functionals correctly? (they seem right to me ). 
I defined O3LYP and LC-omegaPBE functionals as follows:

LC-omegaPBE:
xc xwpbe 1.00 cpbe96 1.0 hfexch 1.00
cam 0.3 cam_alpha 0.00 cam_beta 1.00
E(NWCHEM6.1) = -1961.195574745674
E(G09) = -1961.27046296

O3LYP:
XC vwn_5 0.19 lyp 0.81 Hfexch 0.1161 slater 0.9262 OPTX 0.8133
E(NWCHEM6.1) = -1960.95894758
E(G09) = -1961.91519808

Gaussian used the following scaling constants for O3LYP:
ScaHFX= 0.116100
ScaDFX= 1.000000 (local, exchange) 0.813300 (non-local, exchange) 1.000000 (local, correlation) 0.810000 (nonlocal, correlation)
I think G09 also uses slater and vwn_5 local exchange end local correlation functionals. LYP treats some local correlation, rest of it is treated by vwn_5, so I am not sure why the scaling is 1.00 for local correlation part (VWN_5 ??) in G09


I would appreciate if somebody can help.

Thanks.

Cheers,
BSN.

Forum Vet
Threads 3
Posts 831
4:27:10 PM PDT - Mon, Jun 4th 2012
Change of OPTX coefficient
Beulahsu,
Have a look at the following discussion from CCL
http://www.ccl.net/chemistry/resources/messages/2008/10/09.007-dir/index.html

As a result of that discussion, you might want to use the following input line that uses a coefficient
of 1.164393477 (=0.8133*1.43169) for OPTX

XC vwn_5 0.19 lyp 0.81 Hfexch 0.1161 slater 0.9262 OPTX 1.164393477

Cheers, Edo
Edited On 4:28:39 PM PDT - Mon, Jun 4th 2012 by Edoapra

  • Niri Forum:Admin, Forum:Mod, NWChemDeveloper, bureaucrat, sysop
    Profile
    Send PM
Gets Around
Threads 4
Posts 152
7:17:51 AM PDT - Tue, Jun 5th 2012
Hi Beulashu,

For the lc-wpbe, can you try it with cam=0.4 ? From papers it looks like the G09 default is w = 0.4.

Best,
-Niri

Clicked A Few Times
Threads 2
Posts 6
1:42:26 AM PST - Thu, Jan 17th 2013
Thanks for the suggestions. It has been ages since I got back to the question.


O3LYP worked but not LC-wPBE (with cam 0.4). My recent post on this is here: http://www.nwchem-sw.org/index.php/Special:AWCforum/st/id721/Are_these_definitions_correct...
Edited On 1:47:14 AM PST - Thu, Jan 17th 2013 by Beulahsu


Forum >> NWChem's corner >> General Topics



Who's here now Members 0 Guests 1 Bots/Crawler 0

AWC's: 2.5.10 MediaWiki - Stand Alone Forum Extension
Forum theme style by: AWC
Retrieved from "http://www.nwchem-sw.org/index.php/Special:AWCforum/st/id472"