Maybe try increasing your convergence criteria both for the geometry optimization and the actual SCF. Sometimes when atoms are mildly shifted along a direction in which the potential surface is flat, a too lax convergence criterium may give you a final point just mildly deviated enough from the actual optimum in order to create something appearing as a fake transition state.

Jürgen Weippert Thanks for the reply.

I already used very strict convergence criteria for scf and vc relax, also for phonon calculation very strict threshold has been employed.

Please find the convergence criteria for all below:

1. Vc-relax : conv_thr = 1.0d-14

2. scf : conv_thr = 1.0d-15

3. phonon : tr2_ph = 1e-15

Also please find the input files for above .

--

Thanks

Ambasta Vipulkumar Sanjaykumar

Undergraduate Researcher

Vipul Kumar Ambasta By increasing q point can solve (fast forwarding solution). Let's say nq1 = 5/6, nq2 = 5/6, nq3 = 5/6

Further

1- I have doubt in smearing factor what you used (Gaussian).

2- I don't think so your digits of KPOINTS considerable. It should be more dense.

3- I never show ecutwfc = 35, please re verify.

4- Please once choose separate cell parameter and ibrav = 0 for doing vc relax.

https://lists.quantum-espresso.org/pipermail/users/2016-April/035046.html

Shivam Kansara Thanks for your detailed reply. I have edited the input files as per your suggestions and restarted the calculation with smearing 'cold' as suggested by materials project.

Will update how it goes.

--

Thanks

Shivam Kansara Sorry for the late reply, had caught up in some work , so was unable to reply back to you.

So for the readers,

this is what helped me

increasing kpoint density to 9x8x8

smearing ='cold'

qpoints in phonon calculation 5x5x5

Despite, the fact that it was quite computationally expensive, it worked out for me.

Very thankful to the community

--

Regards

Ambasta Vipulkumar Sanjaykumar

Undergraduate researcher