[Pw_forum] How can I define a Supercell in my simulation?

[Giovanni Cantele]

> Thank you so much for your attention. I have seen Marzari's group 
> simulations on carbon nanotube which were done with Quantum-Espresso 
> but all examples of Quantum-Espresso on its official website are for 
> periodic mater so I don't have a sample input file for a non-periodic 
> simulation.
A truly non-periodic calculation CANNOT be performed within 
Quantum-Espresso because the code will always assume a three-dimensional 
lattice (according to the value of ibrav you specify in the input file), 
yet the definition of a supercell allows you to calculate all the 
properties (DOS, band structure, phonons, ...) of a non-periodic system 
(molecules, surfaces, nanowires, defects, ....).

The "supercell" input file has the SAME structure of a "standard" input 
file. Some "examples" can be found in the Quantum-Espresso "examples"!   
: example09 (SiH4 molecule), example11 (isolated atom), WorkFct_example 
(surface), and others.

The supercell definition allows you to perform numerical simulations of 
two-, one- and zero-dimensional systems, EVEN IF in all those cases the 
system will be replicated in all directions. See (pag. 18-19) 
http://www.fisica.uniud.it/~giannozz/Corsi/metnum.pdf and (p. 1054) 
Payne et al, Rev. Mod. Phys. 64, just to cite two references.
The basic issue is that if, for example, you want to calculate 
properties of a zero-dimensional system (e.g. a molecule), the unit cell 
parameters must be defined in such a way that vacuum space must separate 
periodic replica. Such space must be enough to prevent such replica from 
interacting. How much depends on your system and the convergence of the 
calculations must be checked. Therefore: supercell size in the i-th 
direction (assumed "non-periodic") = system size in that direction + 
vacuum space

Also consider that, if the i-th direction is a non-periodic one, you 
don't need to use k-point sampling along that direction (so, for a 
molecule, you could use only gamma point).
> Where can I find an input sample file (pw.x input file) for a 
> non-periodic simulation?
see above

> How should I define ibrav 
> <file:///E:/PhDThesis/Simulation/QUANTUM-ESPRESSO/Doc/INPUT_PW.html#id2982558> 
> | celldm 
> <file:///E:/PhDThesis/Simulation/QUANTUM-ESPRESSO/Doc/INPUT_PW.html#id2982665> 
> | assume_isolated? 
> <file:///E:/PhDThesis/Simulation/QUANTUM-ESPRESSO/Doc/INPUT_PW.html#id2983511>
ibrav: it depends on the system, in the case of a molecule one can 
usually specify a cubic lattice (ibrav=1), but some other choices can be 
more efficient if for example the molecule size is not nearly the same 
in all directions. For a nanowire a choice can be a tetragonal lattice 
(ibrav=6), where the "c" axis is the periodicity direction.
celldm: you must define your supercell and choose it accordingly
assume_isolated: not mandatory


> Should I define CELL_PARAMETERS 
> <file:///E:/PhDThesis/Simulation/QUANTUM-ESPRESSO/Doc/INPUT_PW.html#id2985370> 
> ?
As for the "periodic" system, you can either use ibrav /= 0 and 
celldm(1:6) or ibrav=0 and CELL_PARAMETERS
> I try to calculate electronic structure of a carbon nanotube when 
> another biomolecule dock with it. have seen ever such as this by 
> Quantum-Espresso? I want to know that it's a conceptual barrier to 
> this purpose  with Quantum-Espresso? 
No conceptual barrier, just a matter on how large (size, number of atoms 
and electrons) is you system, because the larger it is, the more 
(computationally) expensive is the calculation!

Giovanni


-- 



Dr. Giovanni Cantele
Coherentia CNR-INFM and Dipartimento di Scienze Fisiche
Universita' di Napoli "Federico II"
Complesso Universitario di Monte S. Angelo - Ed. 6
Via Cintia, I-80126, Napoli, Italy
Phone: +39 081 676910
Fax:   +39 081 676346
E-mail: giovanni.cantele at cnr.it
        giovanni.cantele at na.infn.it
Web: http://people.na.infn.it/~cantele
Research Group: http://www.nanomat.unina.it