Dear Mahdi, The reason why the atomic charges do not converge is that a 3d grid is lousy when it comes to representing an all-electron density like the one you get from plot_num=21 near the atomic nuclei. In addition to plot_num=21, you need to generate a cube file containing the valence pseudo-density (plot_num=0), which is what is actually used in the pw.x calculation. Then, use plot_num=21 to generate the Bader basins but integrate plot_num=0 in them. A simple example input for critic2 is:
crystal bleh_21.cube load bleh_21.cube id rhoae load bleh_0.cube id rhops integrable rhops yt This is necessary because, while plot_num=0 integrates to the correct number of valence electrons, it doesn't give the correct Bader basins. In fact, it is perfectly possible that, in some cases, the plot_num=0 density does not have maxima at the nuclear positions at all. Best, Alberto --- Dr. Alberto Otero de la Roza Department of Chemistry, University of British Columbia, Okanagan Kelowna, British Columbia, Canada V1V 1V7 * Mahdi Faghihnasiri <[email protected]> [2017-12-02 23:33:51 +0330]: > Dear all, > > I am trying to calculate the atomic charge of BaMoO3 with Bader analysis. I > am using Quantum ESPRESSO for post-processing and Critic2 for Bader > analyses. the exchange-correlation function is PBEsol and the type of > pseudopotentials are PAW. so I used the plot_num=21 for pp.x calculation. > I have read rest of the Pw_forum for bader analysis but this statement has > a lot of ambiguity. as I found out, for Bader analysis I have to use denser > FFT (but how much?). for this purpose I have tried different nr1, nr2, nr3: > > ==================================== > SCF.in: > nr1=600, nr2=600, nr3=600 > > SCF.out: > Dense grid: 307531 G-vectors FFT dimensions: ( 600, 600, 600) > Smooth grid: 59085 G-vectors FFT dimensions: ( 576, 576, 576) > > Critic2 Out: > * Critical point list, final report (non-equivalent cps) > Topological class (n|b|r|c): 3(5) 3(42) 0(0) 2(11) > Morse sum: -48 > > * Integrated atomic properties > # Integrable properties 1 to 4 > # Id cp ncp Name Z mult Volume Pop > Lap Charge > 1 1 1 Ba 56 -- 1.37246087E+02 6.70321481E+01 > 7.09104580E-02 6.70321481E+01 > 2 2 2 Mo 42 -- 6.97318032E+01 4.13474024E+01 > 2.91573989E+00 4.13474024E+01 > 3 3 3 O 8 -- 7.70440931E+01 9.21734416E+00 > -9.95550129E-01 9.21734416E+00 > 4 4 3 O 8 -- 7.70440931E+01 9.21734416E+00 > -9.95550129E-01 9.21734416E+00 > 5 5 3 O 8 -- 7.70440931E+01 9.21734416E+00 > -9.95550128E-01 9.21734416E+00 > ------------------------------------------------------------------------------------------------ > Sum 4.38110170E+02 1.36031583E+02 > -4.21448737E-08 1.36031583E+02 > ==================================== > > > ==================================== > SCF.in: > nr1=800, nr2=800, nr3=800 > > SCF.out: > Dense grid: 429697 G-vectors FFT dimensions: ( 800, 800, 800) > Smooth grid: 108767 G-vectors FFT dimensions: ( 750, 750, 750) > > Critic2 Out: > * Critical point list, final report (non-equivalent cps) > Topological class (n|b|r|c): 3(5) 0(0) 3(60) 3(17) > Morse sum: 48 > > > * Integrated atomic properties > # Integrable properties 1 to 4 > # Id cp ncp Name Z mult Volume Pop > Lap Charge > 1 1 1 Ba 56 -- 1.37215953E+02 5.97476664E+01 > 2.80099625E-02 5.97476664E+01 > 2 2 2 Mo 42 -- 6.97266839E+01 4.04985130E+01 > 1.20777162E+00 4.04985130E+01 > 3 3 3 O 8 -- 7.70558444E+01 9.21662377E+00 > -4.11927139E-01 9.21662377E+00 > 4 4 3 O 8 -- 7.70558444E+01 9.21662377E+00 > -4.11927139E-01 9.21662377E+00 > 5 5 3 O 8 -- 7.70558444E+01 9.21662377E+00 > -4.11927139E-01 9.21662377E+00 > ------------------------------------------------------------------------------------------------ > Sum 4.38110170E+02 1.27896051E+02 > 1.70404130E-07 1.27896051E+02 > ==================================== > > > > as you can see, for different "nr" I found out different atomic charge. > even with a Dense grid around 429697 G-vectors, the charge does not get > converged (Ba = 59.7). > > > > > Please if somebody has a clue about this let me know. > > > > > *Mahdi FaghihNasiri* > Department of Physics > Shahrood University of Technology > Shahrood, Iran > _______________________________________________ > Pw_forum mailing list > [email protected] > http://pwscf.org/mailman/listinfo/pw_forum _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
