Simple DFT Software for Tutorials - Windows Platform

[FabioF]

"Simple" DFT software

Hi,

is anybody aware of a "ligthweight" DFT software to be used as a tutorial (say for undergraduate students, etc.), so very simple to configure, use, and visualize results? Installation on a Windows platform would be the best...

thanks
Fabio

 

[Epicurus]

dft code

Do you have any compliers. You will most likely need a fortranX compiler, for which you can get free for most operating systems. The other thing is that you are most likely not going to have a nice gui, more like command line stuff with some text based input file.

 

[FabioF]

Compilers won't be a big issue. Also, GUI is not the main need.
Do you have any suggestion about the most proper package for that tutorial aim?

thanks
Fabio

 

[Epicurus]

try this link

http://linux.softpedia.com/get/Science-and-Engineering/Chemistry/[/URL]

I have not personally used any of these programs but see how you go. There is a stack of free visualisation packages but good simulation packages are hard to find. If you are at a university you should have access to one of the following codes

Gaussian
dmol3
crystal0X
VASP
ADF
SIESTA
GAMESS
CASTEP

These are quite complete codes, but are not free. I have had a look around through our software and as it turns out none of it is free (shows how naive I was, I just log in and start using).

I will keep looking, may be able to dig some software written for a PhD or something but I am not as much help as I thought I would be. If you can access any of the above codes I will be able to help you get them working.

Conversely you could write your own code?

 

[Dr Transport]

http://directory.fsf.org/science/physics/octopus.html

This is a free DFT program for Linux...

 

[FabioF]

Thanks for your help.

I'm not in a university, but sometimes ago I found a free distribution of GAMESS (somewhere - can't remember anymore where it was), and I'm trying now to use it.

I'll also try with Octopus... I'll let you know.

bye

 

[FabioF]

...About writing my own one - I considered this option, but I thought it'd be a tough job, at least for a single person.

But, could you suggest a starting point, some docomentation or anything to start such a project, maybe limiting functionalities to the easier parts to implement.

Could you suggest for example how to "cut-out" such a simplified DFT simulator, from a functioanl standpoint?

I'd be definitely interesting to start a development from scratch on a Windows/Intel platform, not adapting one of the existing packages, for example with an input/output interface that is not the old-styled, FORTRAN one (even if it musn't necessarly be a GUI).

thanks
Fabio

thanks

 

[marlon]

Hi,

is anybody aware of a "ligthweight" DFT software to be used as a tutorial (say for undergraduate students, etc.), so very simple to configure, use, and visualize results? Installation on a Windows platform would be the best...

thanks
Fabio

Well all DFT software is as easy to use. I mean, if you want to do proper DFT simulations, you need to understand certain key parameters (k point mesh, energy cutoff, basis set convergence, dipole threshold, etc etc) so you should use the software of the "big DFT people".

John Robertson : CASTEP
Peter Blöchl : VASP

SIESTA is ok because it is not a plane wave code but it uses localised basis sets instead. If you use unit cells that contain more than 100 atoms, SIESTA will be quite faster than the above software (but less accurate). If you want to simulate surfaces, i would strongly advice you NOT to use the dipole correction (ie the vacuum above the unit cell) but use the symmetric unitcell with TWO interfaces in stead of one.

I mean : in stead of "metal/high k/vacuum" use "high k/metal/high k"
In the latter case, both high k layers are the same, so you have 2 identical interfaces between the metal and the high k.

If you want to know more about this, just let me know...

marlon

 

[skgupta-phy]

If you want than u can use Quantum EXPERROSO for DFT

 

[zeta]

I use and highly recommend abinit tho unfortunately this requires (as someone has said) more than just a little elementary knowledge. However if you go this route, one can go through the tutorials and happily turn out some numbers fairly quickly. I'm told (tho I haven't tried it) molden can be used with abinit for visualization, but I think you would have to create a *pdb file.

I've recently been introduced to NWCHEM + ecce through working with chemists (yes I know) and they seem to think it's the tops also

 

[zeta]

there is also batchman http://www.univ-orleans.fr/icoa/batchman/ which has GAMESS and MOPAC included