Dear friends
When I run geometry optimization, the input file can run normally, but there is a warning prompt. I change the ecut according to the warning prompt, but no matter I increase or decrease, the warning prompt is in. Please help me solve this problem, with the highest respect.
log:— !WARNING
src_file: m_drivexc.F90
src_line: 711
message: |
Density went too small (lower than xc_denpos) at 558648 points
and was set to xc_denpos = 1.00E-14. Lowest was -0.54E-02.
This might be due to (1) too low boxcut or (2) too low ecut for
pseudopotential core charge, or (3) too low ecut for estimated initial density.
Possible workarounds : increase ecut, or define the input variable densty,
with a value larger than the guess for the decay length, or initialize your,
density with a preliminary LDA or GGA-PBE if you are using a more exotic xc functional.
The list of pseudopotentials you’re using and the XC functional
The geometry of your system. Do you have lot of vacuum?
Also note that this warning is usually harmless if only the initial SCF iterations are affected. You can ignore this message If the warning disappears in the last iterations and the SCF cycle converges.
Dear gamtteo:
This is ZHAO Yongsheng.
I’m glad to hear from you about my problem.
I use Norm-conserving pseudopotella(NC);
The pseudopotential I used is GGA-PBE and XC is 11;
My system is 3*3 supercells, 5 layers, and the vacuum layer is 15 angstroms;
This warning is not only for the initial SCF loop, but also after at least four cycles.And I found that when IonMov =22 was used,the diamond structure was optimized into a graphite layered structure and the total energy decreased.
Thank you for taking the time to answer my questions.
Dear ZHAO Yongsheng,
If the warning is there only for the first SCF steps and disappear for the latest SCF cycles, then there is no major problem.
For the diamond to graphite, I guess this is because you put vacuum and the code end up to the graphite (local?) minimum during the relaxation. You said to be the case for ionmov=22, does it means it happens for this ionmov value but not with other ionmov values?
Best wishes,
Eric
The current warning (density went too small (lower than XC_ Depos at 279756 points ）has been appearing, and the value is constantly changing. I did set up a 15 angstrom vacuum layer and optimized it with ionmov = 2,3,22. It’s not partial graphitization, it’s structural integrity. I’m thinking about whether it’s because I cut the slab structure after I optimized the geometry of the unit cell, or because I chose the wrong value of ionmov.
Bset wish for you.
Yongsheng
Do you know whether your pseudo has non-linear core correction (NLCC)?
The C pseudo from the pseudodojo is generated with NLCC but the other legacy pseudos may not have NLCC.
To understand if the pseudo has NLCC, look at this section of the header:
0.00000 0.00000 0.00000 rchrg,fchrg,qchrg
If NLCC is not used, rchrg is set to 0 (as in this case).
Dear Gmatteo:
I’m glad to see your reply. I checked my pseudo pseudopotential file (c.psp8) and its ncll doesn’t seem to be 0, 5.99, 4.0 , 0.0 rchrg fchrg qchrg. At present, I have debugged my program. After adjusting the value of ionmov to 2, the program can run, and the warning still exists, but the SCF calculation can converge. Only in the calculation process, the diamond structure will be optimized to graphite structure, and the total energy changes greatly.
Best wishes
This is my input file：
Dia111-2x2-5lay-15A-slab : computation of the surface total energy
#Definition of the unit cell
acell 9.550676 9.550676 44.916899 # Lattice parameters of slab