QuantumATK Forum

QuantumATK => General Questions and Answers => Topic started by: Dhirendra on November 15, 2015, 21:15

Title: Surface Shift in Geometry Optimization
Post by: Dhirendra on November 15, 2015, 21:15

In creating the interface geometry, how do I decide the x,y,z shifts (under 'Shift Surfaces')? There is not well experimental data on that. How do I approach global minima. Is the 'geometryoptimization' strongly depend on the initial structure?

Title: Re: Surface Shift in Geometry Optimization
Post by: zh on November 16, 2015, 01:27

This kind of question is quite general and there is no well-defined answer to it. The search of global minima of interface is quite challenging because of large numbers of degree of freedom in the energy landscape. The simply use of geometry optimization would fail or take very long simulation time.  In literature some groups tried to apply some evolution algorithms to search the interface structures. It seems to look promising.

"how do I decide the x,y,z shifts (under 'Shift Surfaces')?"
You may choose an initial values depending on the covalent or ionic radii of atoms in the interface regions.

" Is the 'geometryoptimization' strongly depend on the initial structure?"
Of course, it does.

Title: Re: Surface Shift in Geometry Optimization
Post by: Subhban15 on November 20, 2015, 20:48

Quote from: zh on November 16, 2015, 01:27

In literature some groups tried to apply some evolution algorithms to search the interface structures. It seems to look promising.

Could you please suggest some such algorithms or provide some links, so that we can Optimize in shorter time?

Thanks in advance! :)

Title: Re: Surface Shift in Geometry Optimization
Post by: Anders Blom on November 22, 2015, 20:53

http://docs.quantumwise.com/tutorials/device_relaxation.html

Title: Re: Surface Shift in Geometry Optimization
Post by: Subhban15 on November 25, 2015, 21:10

Thank you very much! :)