Murnaghan equation for lattice parameters?

[new_986]

Murnaghan equation for lattice parameters?

Hi everybody...
I'm trying to determine lattice constant (a) for BN by calculating the total energy for different values of a using ABINIT, then chose the one which correspond the minimum value of energy, Is this convenient way to determine a ??

thanks...
with regards

 

[Useful nucleus]

To get the lattice constant of a presumably cubic material, hence the only degree of freedom is "a" , one has variety of options.

1- Running multiple self-consistent calculations at different values of "a" ( Values of "a" that sufficiently bracket the experimental lattice constant" and choose the one that minimizes the energy. This method is not very accurate as you may miss a lot of if you do not run enough calculations.

2- Performing multiple self-consistent calculations and then fitting to Birch-Murnghan equation, form which you can obtain a more reliable lattice constant compared to method 1.

3- Some codes can minimize the energy by varying the atomic coordinates and the simulation cell volume and shape. So, in a single (slightly expensive) calculation you can obtain the lattice constant. However, since volume of the simulation cell vary, special care must be taken.


To me the second method is easy to perform and gives also more information such as the bulk modulus. Furthermore you can use the fitted Birch-Murnghan equation in computing the pressure, enthalpy, etc...

 

[new_986]

To get the lattice constant of a presumably cubic material, hence the only degree of freedom is "a" , one has variety of options.

1- Running multiple self-consistent calculations at different values of "a" ( Values of "a" that sufficiently bracket the experimental lattice constant" and choose the one that minimizes the energy. This method is not very accurate as you may miss a lot of if you do not run enough calculations.

2- Performing multiple self-consistent calculations and then fitting to Birch-Murnghan equation, form which you can obtain a more reliable lattice constant compared to method 1.

3- Some codes can minimize the energy by varying the atomic coordinates and the simulation cell volume and shape. So, in a single (slightly expensive) calculation you can obtain the lattice constant. However, since volume of the simulation cell vary, special care must be taken.


To me the second method is easy to perform and gives also more information such as the bulk modulus. Furthermore you can use the fitted Birch-Murnghan equation in computing the pressure, enthalpy, etc...

Thanks Mr. useful nucleus..
There is no implementation for Murnghan equation in ABINIT, anyway, could you please recommend me a reference about any three above ways
thanks a lot and best regards

 

[Useful nucleus]

I would not expect any code to have Birch-Murnghan readily implemented for the user. However, if you now how to perform curve fitting , it is an easy task to fit your simulation results.

A good introductory reference for such calculations is:
Density functional theory:a practical introduction by David S. Sholl, Janice A. Steckel
You may consult Chapter 2 &3

Another resource would be MIT Open course ware ; in particular the lab sessions of course 3.320

http://ocw.mit.edu/courses/material...eling-of-materials-sma-5107-spring-2005/labs/

 

[daveyrocket]

Abinit does do Useful nucleus's solution #3. Look at the variables ionmov, optcell, ecutsm and dilatmx.

 

[new_986]

Thanks dears a lot