This thread is locked.Only browsing is available.
Top Page > Browsing
The convergence of RFC5
Date: 2020/01/02 12:23
Name: Alisa   <614327091@qq.com>

Dear developers,

Sorry for bothering. We have some questions about using RFC5 for geometry optimization and tested it with an interface system. When we used the system with large mismatch (a=11.53, b=3.95, c=29.456; alpha=90, beta=90, gamma=90, mismatch=4.5%) to test, we find that the original structure remains basically unchanged (a=11.72, b=4.05, c=28.95,alpha=89.9 beta=89.9 gamma=90)。But when we used the inteface with small mismatch(a=27.82, b=3.95 ,=29.456; alpha=90 beta=90 gamma=90, mismatch=2.29%), the optimized structure varies greatly(a=28.31, b=4.08, c=28.61; alpha=85.9 beta=89.9 gamma=89.9).
We don't understand why the smaller the lattice mismatch the bigger the distortion of the geometry.

Best regards,
Alisa
メンテ
Page: [1]

Re: The convergence of RFC5 ( No.1 )
Date: 2020/01/03 00:31
Name: T. Ozaki

Hi,

The adjustment of lattice parameters at interfaces is caused by a several factors
such as lattice mismatch, interaction strength, layer thickness, and relative hardness
between the two materials of the interface. The interplay among those factors determines
the adjustment of lattice parameters.

I wonder that the latter case may have much stronger interaction, or one of the material
in the latter case is much harder (nearly equal to having larger cohesive energy) than
the other. Or the layer thickness might be different from the former case due to the large
in-plane supercell.

Of course, you need to make sure that both the calculations were performed properly
before you discuss the physics.

Regards,

TO
メンテ

Page: [1]