Bond critical point calculation problem

Bader charge density analysis

Moderator: moderators

Post Reply
Taha55
Posts: 8
Joined: Tue Nov 02, 2021 6:54 pm

Bond critical point calculation problem

Post by Taha55 »

Hello

''./bader NaCl_valence.cube -ref NaCl_allelec.cube -cp''

I have used this command and Why didn't the '' Unique Bond Critical Point count '' finally consist of calculation of CPFU.dat file?

Critical point is found at indices
199.750768742182 95.7334269265248 4.52264630712299
Coordinates in cartesian are
40.6828395681828 19.9596536467650 -8.81383937736660
Direct coordinates are
0.998753843710911 0.957334269265248 2.512581281734995E-002
Gradiant is
-0.243741459247190 -0.381501888886248 -0.176300307681686
Hessian is
100082.762383647 115318.295883000 195759.700209404
115318.295883000 163969.537266563 120995.289108000
195759.700209404 120995.289108000 627366.753569645
Eigenvalues are
-2818.06585625999 161484.425499556 732752.693576559
Eigenvectors are
0.865189350342933 -0.468479693437409 -0.178813212291389
0.378997627483480 0.844425193641715 -0.378558965955168
0.328341569733513 0.259755402600108 0.908140377037046
This is a ring critical point

________________________________________
Unique critical point count: 12209
Unique bond critical point count: 0
Unique ring critical point count: 11801
Unique cage critical point count: 0
Unique nucl critical point count: 408

Best regards
wc5879
Posts: 21
Joined: Thu May 22, 2014 7:49 pm

Re: Bond critical point calculation problem

Post by wc5879 »

My apologies!
I thought I replied!

In this case, the critical points (CPs) found were not self-consistent so the number of each type of CP does not match with what's in the CPFU00.dat.

This is a bug that I'm going to resolve so that stdout and CPFU00.dat are always consistent.

May I have your cube file so that I can know why a self-consistent CP set cannot be found?
Taha55
Posts: 8
Joined: Tue Nov 02, 2021 6:54 pm

Re: Bond critical point calculation problem

Post by Taha55 »

Thank you very much for your reply. The Cube files and the Quantum espresso input files used for the calculation of cube files are also attached.
Best Regard
Attachments
Quantum_espresso_input_files.zip
(2.42 KiB) Downloaded 580 times
H2O_cube_files.zip
(646.08 KiB) Downloaded 572 times
wc5879
Posts: 21
Joined: Thu May 22, 2014 7:49 pm

Re: Bond critical point calculation problem

Post by wc5879 »

This all-electron cube file contains atoms that are ~0.04 angstroms apart from each other, and the grid is relatively rough.
Theoretically, if the grid is smooth and fine enough, a self-consistent CP set can be found.
But I don't think it would be of any use since this atomic distance seems unreasonably small to me.
Taha55
Posts: 8
Joined: Tue Nov 02, 2021 6:54 pm

Re: Bond critical point calculation problem

Post by Taha55 »

Dear Professor;
Thank you very much for your reply. I made a wrong calculation using the wrong input file. I fixed this situation and handled the cube files using the correct input files, but couldn't get the bond critical point (as below). Cube files (corrected) are attached. I request you to help?

Interrogation of the soul:
Did I turn on vacuum ?
Did I tell if this is a crystall or molecule?
Did I use the CHGCAR_sum ?
Did I use the least square flag -ls ?
Critical point is like a box of chocolates. You never know what you are gonna get.

CRITICAL POINTS INFO WRITEN TO CPF.dat
CRITICAL POINTS FOUND: 33891
FINDING UNIQUE CRITICAL POINTS...
Unique critical point count: 138
Unique bond critical point count: 0
Unique ring critical point count: 0
Unique cage critical point count: 0
Unique nucl critical point count: 138
Attachments
cube_files_new.rar
(67.48 MiB) Downloaded 598 times
wc5879
Posts: 21
Joined: Thu May 22, 2014 7:49 pm

Re: Bond critical point calculation problem

Post by wc5879 »

This totoal number looks much more reasonable.
The CP set is still not self-consistent. Unfortunately, the code that does the CP type counting is bugged at this moment for non-self consistent cases.

As for why it is not self consistent, there are a few reasons.
It could be that the grid density is too low. This is a relatively large box and the charge density gradient is high. You may need to increase grid density on all axis by as much as 3 ~ 4 times.
I also noticed that there are some extra maxima where there is no nucleus. As of right now, the code is designed so that these cases cannot pass, so that we are aware of these extra maxima. We need to take a closer look to determine whether they are physical or not.

Nonetheless, the useful CPs can be found in CPFU00.dat, and the nucleus/bond/ring/cage designation should be accurate.
Taha55
Posts: 8
Joined: Tue Nov 02, 2021 6:54 pm

Re: Bond critical point calculation problem

Post by Taha55 »

Dear Professor;
Thank you very much for your reply. How do i increase the grid density or should I use a different program than the quantum espresso program for this calculation?
Taha55
Posts: 8
Joined: Tue Nov 02, 2021 6:54 pm

Re: Bond critical point calculation problem

Post by Taha55 »

wc5879 wrote:
> This totoal number looks much more reasonable.
> The CP set is still not self-consistent. Unfortunately, the code that does
> the CP type counting is bugged at this moment for non-self consistent
> cases.
>
> As for why it is not self consistent, there are a few reasons.
> It could be that the grid density is too low. This is a relatively large
> box and the charge density gradient is high. You may need to increase grid
> density on all axis by as much as 3 ~ 4 times.
> I also noticed that there are some extra maxima where there is no nucleus.
> As of right now, the code is designed so that these cases cannot pass, so
> that we are aware of these extra maxima. We need to take a closer look to
> determine whether they are physical or not.
>
> Nonetheless, the useful CPs can be found in CPFU00.dat, and the
> nucleus/bond/ring/cage designation should be accurate.


Dear profesor

I made the calculation by enlarging the grid and 40 bond critical point values were calculated. But the electron densities are not visible. What can I do for it?

Critical point is found at indices
271.510664966411 155.184412010551 49.8751526861093
Coordinates in cartesian are
54.5565384863557 31.1822901961640 10.0785214790455
Direct coordinates are
0.848470828020034 0.969902575065941 0.221667345271597
Gradiant is
-1.740892289880946E-006 9.729010810780037E-005 2.215624434143137E-005
Hessian is
-68.5334937600000 10.3300740000000 -8.98718863679978
10.3300740000000 -33.1027948800003 74.9356765344002
-8.98718863679978 74.9356765344002 -9.71342006213576
Eigenvalues are
-102.732534338394 -63.0519062860056 54.4347319222633
Eigenvectors are
0.371678701623675 -0.705079664029273 0.603918545941178
0.928360992716673 0.282916133358115 -0.241048394908172
-8.999785842834454E-004 0.650246975277902 0.759722357957489
This is a bond critical point
Attachments
cube_files.rar
(75.45 MiB) Downloaded 558 times
wc5879
Posts: 21
Joined: Thu May 22, 2014 7:49 pm

Re: Bond critical point calculation problem

Post by wc5879 »

I updated the code in the last month or two to output the charge density.
Try getting the most up-to-date code, build and run to see if you get the charge.
Also, is your newest CP set self-consistent?
Taha55
Posts: 8
Joined: Tue Nov 02, 2021 6:54 pm

Re: Bond critical point calculation problem

Post by Taha55 »

Dear Profesor;

I downloaded all the updated files (http://theory.cm.utexas.edu/svn/bader/).

I have successfully completed the update using the following command;
-------------------------------------------
asus@asus-VirtualBox:~/Desktop/$ cd bader_update
asus@asus-VirtualBox:~/Desktop/bader_update$ make -f makefile.lnx_ifort

the following error occurs;
.
.
.
ld: critpoint_mod.f90:(.text+0x98d3): undefined reference to `_gfortran_st_write'
ld: critpoint_mod.f90:(.text+0x98e7): undefined reference to `_gfortran_transfer_character_write'
ld: critpoint_mod.f90:(.text+0x98ef): undefined reference to `_gfortran_st_write_done'
ld: critpoint_mod.f90:(.text+0x9919): undefined reference to `_gfortran_st_write'
ld: critpoint_mod.f90:(.text+0x992d): undefined reference to `_gfortran_transfer_character_write'
ld: critpoint_mod.f90:(.text+0x9935): undefined reference to `_gfortran_st_write_done'
ld: critpoint_mod.f90:(.text+0x9973): undefined reference to `_gfortran_st_write'
ld: critpoint_mod.f90:(.text+0x9987): undefined reference to `_gfortran_transfer_character_write'
ld: critpoint_mod.f90:(.text+0x998f): undefined reference to `_gfortran_st_write_done'
make: *** [makefile.lnx_ifort:31: bader] Error 1

When I use the command below, the update and calculation takes place, but there are no electron densities in CPFU.dat file.

asus@asus-VirtualBox:~/Desktop/bader_update$ make -f makefile.osx_gfortran
.
.
Critical point is found at indices
271.51066496641084 155.18441201055060 49.875152686109239
Coordinates in cartesian are
54.556538486355699 31.182290196164008 10.078521479045525
Direct coordinates are
0.84847082802003393 0.96990257506594124 0.22166734527159662
Gradiant is
-1.7408922734496457E-006 9.7290108076047987E-005 2.2156244386284385E-005
Hessian is
-68.533493760000042 10.330074000000014 -8.9871886367998695
10.330074000000014 -33.102794880000253 74.935676534400017
-8.9871886367998677 74.935676534400031 -9.7134200621357643
Eigenvalues are
-102.73253433839362 -63.051906286005533 54.434731922263104
Eigenvectors are
0.37167870162367755 -0.70507966402927269 0.60391854594117744 0.92836099271667250 0.28291613335811744 -0.24104839490817351 -8.9997858428404276E-004 0.65024697527790198 0.75972235795748944
This is a bond critical point


What can be the problem? I have fortran package installed on my system.
Post Reply