User Tools

Site Tools


howto:running_qe_computation

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision
Previous revision
howto:running_qe_computation [2019/09/02 07:55] mtaillefumierhowto:running_qe_computation [2021/12/08 10:18] (current) mtaillefumier
Line 1: Line 1:
 +====== How to run calculations like Quantum ESPRESSO ======
 +
 +
 ===== Introduction ===== ===== Introduction =====
  
Line 8: Line 11:
 An other popular program for plane wave DFT is [[http://www.quantum-espresso.org|Quantum Expresso]] (QE). A full description of QE functionalities is outside the scope of this tutorial, but one of the main functionalities is ground state minimization and geometry optimization of crystals. In this tutorial, we will show how to do QE calculations with CP2K + An other popular program for plane wave DFT is [[http://www.quantum-espresso.org|Quantum Expresso]] (QE). A full description of QE functionalities is outside the scope of this tutorial, but one of the main functionalities is ground state minimization and geometry optimization of crystals. In this tutorial, we will show how to do QE calculations with CP2K +
 SIRIUS. We will consider the simple case of silicium doped with germanium and SIRIUS. We will consider the simple case of silicium doped with germanium and
-show how to convert a QE input file to cp2k input file. Files for this tutorial can be found [[https://www.cp2k.org/?image=potentials-sige.tar.gz&ns=&tab_details=view&do=media&tab_files=files|here]] while the cp2k input files can be found in the cp2k tests directory.+show how to convert a QE input file to cp2k input file. Files for this tutorial can be found [[https://www.cp2k.org/_media/howto:si7ge.tar.gz|here]] while the cp2k input files can be found in the cp2k tests directory.
  
 === Word of caution with the pseudo-potential files === === Word of caution with the pseudo-potential files ===
Line 199: Line 202:
 for cp2k input file. Note the presence of the keyword ''SCALED'' in the section ''COORD'' for cp2k input file. Note the presence of the keyword ''SCALED'' in the section ''COORD''
 which indicates that cp2k should treat the coordinates as given in the lattice which indicates that cp2k should treat the coordinates as given in the lattice
-basis. Putting these two sections together, we have+basis. **The coordinates do not have to be given in the lattice basis, any format supported by cp2k will work**. Putting these two sections together, we have
  
 <code> <code>
Line 417: Line 420:
 ^ QE keyword ^ CP2K (SIRIUS) keyword  ^ section ^ desciption ^ ^ QE keyword ^ CP2K (SIRIUS) keyword  ^ section ^ desciption ^
 | nspins = 1, 2  | NUM_MAG_DIMS = 0, 1 | PW_DFT/PARAMETERS | computation in LDA and SLDA | | nspins = 1, 2  | NUM_MAG_DIMS = 0, 1 | PW_DFT/PARAMETERS | computation in LDA and SLDA |
-| nocolin = true | NUM_MAG_DIMS =    | PW_DFT/PARAMETERS | non colinear magnetism |+| nocolin = true | NUM_MAG_DIMS =    | PW_DFT/PARAMETERS | non colinear magnetism |
 | electron_maxstep | NUM_DFT_ITER | PW_DFT/PARAMETERS | number of scf iterations | | electron_maxstep | NUM_DFT_ITER | PW_DFT/PARAMETERS | number of scf iterations |
 | mixing_beta | beta| PW_DFT/MIXER| mixing parameter for the broyden mixer| | mixing_beta | beta| PW_DFT/MIXER| mixing parameter for the broyden mixer|
Line 431: Line 434:
 Energy Energy
 -------------------------------------------------------------------------------- --------------------------------------------------------------------------------
-valence_eval_sum          :        -4.94444720 +valence_eval_sum          :        -5.08815701 
-<rho|V^{XC}>              :       -30.28985204 +<rho|V^{XC}>              :       -30.23739234 
-<rho|E^{XC}>              :       -43.77347053+<rho|E^{XC}>              :       -43.72437220
 <mag|B^{XC}>              :         0.00000000 <mag|B^{XC}>              :         0.00000000
-<rho|V^{H}>               :       106.74147275 +<rho|V^{H}>               :       106.32392644 
-one-electron contribution :       -81.39606791 (Ha),      -162.79213582 (Ry) +one-electron contribution :       -81.17469111 (Ha),      -162.34938222 (Ry) 
-hartree contribution      :        53.37073638 +hartree contribution      :        53.16196322 
-xc contribution           :       -43.77347053 +xc contribution           :       -43.72437220 
-ewald contribution        :       -68.41309923+ewald contribution        :       -68.41309922
 PAW contribution          :         0.00000000 PAW contribution          :         0.00000000
-Total energy              :      -140.21190129 (Ha),      -280.42380259 (Ry)+Total energy              :      -140.15019932 (Ha),      -280.30039863 (Ry)
  
-band gap (eV) :         0.36618436 +band gap (eV) :         0.34564591 
-Efermi        :         0.24797319+Efermi        :         0.24495087
  
-iteration :  10, RMS 3.843002363419E-11, energy difference : 0.000000000000E+00+iteration :  11, RMS 1.265254419560E-11, energy difference : 0.000000000000E+00
  
-converged after 11 SCF iterations! +converged after 12 SCF iterations!
- +
- ENERGY| Total FORCE_EVAL ( SIRIUS ) energy (a.u.):         -140.211901294509545+
  
 + ENERGY| Total FORCE_EVAL ( SIRIUS ) energy (a.u.):         -140.150199315127225
 </code> </code>
  
Line 551: Line 553:
  
 you should obtain a file ''Si7Ge-pos-1.xyz'' that contains the positions of the atoms and ''Si7Ge-1.ener'' that contains the different energies and temperature as a function of time. The calculations are done in the micro-canonical ensemble which means that the total energy (the sum of the potential which is the total energy given by SIRIUS and the kinetic energy) is conserved during the integration process. It is possible to compare these results with QE, but instead of doing the molecular dynamics with QE, we take the positions of the atoms at each time step and compute the potential energy for this atom position with QE.  you should obtain a file ''Si7Ge-pos-1.xyz'' that contains the positions of the atoms and ''Si7Ge-1.ener'' that contains the different energies and temperature as a function of time. The calculations are done in the micro-canonical ensemble which means that the total energy (the sum of the potential which is the total energy given by SIRIUS and the kinetic energy) is conserved during the integration process. It is possible to compare these results with QE, but instead of doing the molecular dynamics with QE, we take the positions of the atoms at each time step and compute the potential energy for this atom position with QE. 
- +{{:howto:cp2k-sirius-md.png|}} 
-{{:howto:qe-cp2k-sirius.png|}}<imgcaption image1|>Computation of the potential energy with cp2k/sirius and quantum expresso</imgcaption> +The total energy given by QE and CP2K/SIRIUS are shifted by a small offset. More generally, converting the input file from QE to CP2K is not a warranty to obtain the same results for the total energy. The reasons for this multiple : 
 +  * Indicating the same cutoff does not warranty the fft grid size. 
 +  * QE and SIRIUS treat the radial integrals interpolation differently 
 +  * The functional in QE do not use libxc by default, while CP2K/SIRIUS does. 
 +  * The minimization method is sensitive to the initial states. 
 +For all these reasons, we should not try to compare results at the binary level. 
howto/running_qe_computation.1567410943.txt.gz · Last modified: 2020/08/21 10:15 (external edit)