modified becke-johnson mBJ

Message

user224 · #1 Post by **user224** » Sat Aug 10, 2013 8:05 pm

Hi all,

I'm trying to investigate semiconductor band gaps using the mBJ potential (http://prl.aps.org/abstract/PRL/v102/i22/e226401).

I relax my structures using LDA/PBE before selecting mBJ using:

METAGGA=MBJ
LASPH=.TRUE.

However, my calculation isn't converging, regardless of my choice of ALGO. What are the options available for this type of potential?

If anybody has any experience or insight into using mBJ I'd be very appreciative.

Thank you

<span class='smallblacktext'>[ Edited ]</span>

magnon963 · #2 Post by **magnon963** » Mon Aug 26, 2013 1:38 pm

I'm facing the same problem. Did you find some way for this problem.

user224 · #3 Post by **user224** » Wed Sep 04, 2013 12:48 pm

unfortunately not. let me know if you have any luck!

user224 · #4 Post by **user224** » Wed Sep 04, 2013 12:59 pm

unfortunately not. let me know if you have any luck!

jianghpku · #5 Post by **jianghpku** » Fri Sep 13, 2013 3:23 am

Maybe you will be intertested in having a look at the following paper,
http://jcp.aip.org/resource/1/jcpsa6/v1 ... p134115_s1

wyxsdu · #6 Post by **wyxsdu** » Tue Dec 17, 2013 11:30 am

I meet the same problem and then inreased NLM=200, it converged. Compared with PBE or LDA, it converged very slow.