calculation with moving atom

Total energy, geometry optimization, DFT+U, spin....

Moderator: bguster

Locked
arsyik
Posts: 14
Joined: Thu Jul 19, 2012 1:11 am

calculation with moving atom

Post by arsyik » Wed Sep 18, 2013 1:31 pm

Dear all
I apologize if my question has been discussed.
I want to perform calculations on crystals where one of the atom/molecule in it move in one direction and get the energy value of any change in his position?
How to do this?

Thanks a lot
Regard

Arsyik

ilukacevic
Posts: 271
Joined: Sat Jan 16, 2010 12:05 pm
Location: Dept. of Physics, University J. J. Strossmayer, Osijek, Croatia
Contact:

Re: calculation with moving atom

Post by ilukacevic » Thu Sep 19, 2013 8:31 am

Dear Arsyik,

If i understand correctly, "this is how you do it"

Calculate the total energy with no atoms moved and then calculate the total energy with one atom moved away from its original position. And then calculate the difference. You can do it all in one input file with 2 or more datasets.

Best regards,

Igor

arsyik
Posts: 14
Joined: Thu Jul 19, 2012 1:11 am

Re: calculation with moving atom

Post by arsyik » Thu Sep 19, 2013 8:53 am

Dear Igor

Thanks for your reply
Do i have to make calculation with molecular dynamic type (with ionmov for MD) or just normal calculation.

Thanks a lot.

arsyik

ilukacevic
Posts: 271
Joined: Sat Jan 16, 2010 12:05 pm
Location: Dept. of Physics, University J. J. Strossmayer, Osijek, Croatia
Contact:

Re: calculation with moving atom

Post by ilukacevic » Thu Sep 19, 2013 9:11 am

If you will move the atom by yourself, then just do the normal calculation of total energy.

You could use BFGS or MD, too. But the atom(s) would move automatically, according to some principle, like force gradient. Thus, you would have no control over it.

Igor

arsyik
Posts: 14
Joined: Thu Jul 19, 2012 1:11 am

Re: calculation with moving atom

Post by arsyik » Fri Sep 20, 2013 2:53 am

Ok thanks for your guidance.
About energy that we get after calculation of solid crystal ("Etotal"), what is it mean? is a lattice energy/cohesive energy/...? and how if we want to know i.e gibbs free energy, enthalpy?

Again thanks for your guidance. :D

Regard
Arsyik.

ilukacevic
Posts: 271
Joined: Sat Jan 16, 2010 12:05 pm
Location: Dept. of Physics, University J. J. Strossmayer, Osijek, Croatia
Contact:

Re: calculation with moving atom

Post by ilukacevic » Fri Sep 20, 2013 8:54 am

You're trying to get it all the easy way, eh?

Well, ok...for now. Gibbs free energy is

G = U + pV + TS

Total energy given by abinit is U. So, you can get enthalpy from it straightforward

H = U + pV

G you cannot get, since DFT is a ground state theory and, thus, T = 0 in all your calculations. You can obtain a phononic contribution to it, but that's another story.

Cohesive energy I will leave to you to figure out...it's simple.

Igor

arsyik
Posts: 14
Joined: Thu Jul 19, 2012 1:11 am

Re: calculation with moving atom

Post by arsyik » Tue Sep 24, 2013 2:57 pm

I am very sorry for my late to reply.
Yup. thanks again for your guidance.
i just confused about energy that i get from my calculation.

Now if we use MD calculation (i.e tmoldyn_01.in), we get "-kT*entropy = ..." in output file. with entropy term, but is the "k" term mean boltzmann constant?

Regard

Arsyik

Locked