QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: Jenny on October 14, 2014, 17:59

Title: optimized geometry problem on carbon nanotube
Post by: Jenny on October 14, 2014, 17:59
Hi, All.

I've done the optimized geometry on carbon nanotube and copper hybird system. However, after several steps the CNT went broken. As far as I know, CNT should deform instead of broke because of its good mechanical property. I've attached my initial configuration and final result here. Can someone please explain this result to me?

Thank you very much.

Jenny
Title: Re: optimized geometry problem on carbon nanotube
Post by: Anders Blom on October 14, 2014, 23:48
Hard to tell without any details, but your original model looks a bit squeezed, maybe it has too much strain/forces so when it starts it already has a strong tendency to expand. Try to relax only the Cu part, and with a longer Cu-Cu bond length. For this you can use one of the classical potential in the QuickOptimizer, it will just take seconds.
Title: Re: optimized geometry problem on carbon nanotube
Post by: Jenny on October 15, 2014, 18:07
Hard to tell without any details, but your original model looks a bit squeezed, maybe it has too much strain/forces so when it starts it already has a strong tendency to expand. Try to relax only the Cu part, and with a longer Cu-Cu bond length. For this you can use one of the classical potential in the QuickOptimizer, it will just take seconds.

Dr. Anders Blom,

Thank you for your reply. Actually I was doing the optimization of a unit cell, so I guess the QuickOptimizer cannot help here. I've attached my script here. Would you please give me some advice on the optimization without making the carbon nanotube broken? Thanks a lot.

Jenny
Title: Re: optimized geometry problem on carbon nanotube
Post by: Anders Blom on October 16, 2014, 00:09
Your script does not do any cell optimization. But even so, as far as I understand you have made a regular 5,5 CNT and tried to squeeze a lot of Cu atoms inside it. I doubt that is a relevant physical system, at least with the specific setup you have chosen. In a sense your results from the optimization proves that point.