Currently, I am optimizing YAlO₃ doped with lanthanide elements. For simpler elements like Ce and Pr, everything works fine. However, when doping with Yb, something strange happens: the calculation runs for a long time but never converges.
Initially, I set IBRION = 2, and later changed it to IBRION = 3, but neither setting produced a result.
Below are my INCAR settings and part of the OUTCAR output:
INCAR:
ALGO = Normal
EDIFF = 1E-04
ENCUT = 520
IBRION = 3
POTIM = 0.5
ISIF = 7
ISMEAR = 0
ISPIN = 2
LCHARG = True
#LORBIT = 11
LREAL = Auto
LWAVE = True
MAGMOM = 79*0.6 1*4.0
NELM = 100
NSW = 100
PREC = Accurate
SIGMA = 0.05
LDAU = .TRUE.
LDAUTYPE = 2
LDAUL = 2 -1 -1 3 # l-quatum no (-1:no U , 2:d group metals, 3:f group metals)(Y Al O Yb)
LDAUU = 3 0 0 3 # for Yb:3 and O:0 (U values (eV))
LDAUJ = 0 0 0 0 # J values (For lDAUTYPE=2, Ueff=U-J)
LMAXMIX = 6 # 4 for d-electrons, 6 for f-electrons
LDAUPRINT = 2 # 0:silent, 1:Write occupancy matrix to OUTCAR, 2: 1+potential matrix, Default : LDAUPRINT=0
LASPH = .TRUE.
OUTCAR:
EATOM=-1046.9140
kinetic energy error for atom= 0.0008 (will be added to EATOM!!)WARNING: aliasing errors must be expected set NGX to 82 to avoid them
NGY is ok and might be reduce to 82
WARNING: aliasing errors must be expected set NGZ to 58 to avoid them
aliasing errors are usually negligible using standard VASP settings
and one can safely disregard these warnings PAW_PBE Al 04Jan2001 Error EDDDAV: Call to ZHEGV failed. Returncode = 31 1 32 :
energy of atom 2 EATOM= -53.5387
kinetic energy error for atom= 0.0002 (will be added to EATOM!!)
PAW_PBE O 08Apr2002 :
energy of atom 3 EATOM= -432.3788
kinetic energy error for atom= 0.0208 (will be added to EATOM!!)
PAW_PBE Yb 23Dec2003 :
energy of atom 4 EATOM=-6750.7873
kinetic energy error for atom= 0.0018 (will be added to EATOM!!)
This is a complicate calculation with spin and LDA+U. For such system, SCF converges is indeed difficult and needs some experience and skills.
Try start with no spin and no LDA+U and then restart with spin/LDA+U from previous charge
I have no experience with the VASP package, but I hope I can help you understand what happens on the basis of the properties of the Yb ion.
The (trivalent) Ce and Pr atoms are at the beginning of the lanthanide series, and their 4f-orbitals are sufficiently extended to be represented by plane waves. Yb is at the end of the series and, due to the lanthanide contraction, its 4f-orbitals are much more localized, so that a larger number of plane waves would be necessary to represent them. But the situation is more complicated. Yb is in its 3+ state, with one hole in the 4f shell. Due to the compactness of the 4f orbitals, they barely, if at all, hybridize with those of the ligand atoms. So Yb 3+ keeps its "atomic" character, with an F_(7/2) ground state, split by the crystal electric field. This scenario seems to bne confirmed by recent experiments (see reference below), and I do not know whether VASOP can treat such a situation.
Y. Nakayama, N. Nakagawa, Y. Matsuo, T. Kaizu, Y. Harada, T. Ishihara, T. Kita; Yb-doped Y–Al–O thin films with a self-organized columnar structure and their anti-Stokes photoluminescence properties. AIP Advances 1 February 2022; 12 (2): 025110. https://doi.org/10.1063/5.0079632
The issue you are experiencing—non-convergence and failure of ZHEGV—is likely due to the inclusion of Yb, which involves f-electron states that require careful treatment in VASP. The ZHEGV error indicates an eigenvalue solver problem, which often arises due to numerical instabilities in systems with strongly correlated electrons, like those in lanthanide elements.
Here are several steps to address this issue:
Yb has partially filled f-electrons, which can create strong correlations and cause convergence issues. You may need to adjust the U value:
If LDAUU = 3 (eV) for Yb is too small or inappropriate, the self-consistent field (SCF) calculation can become unstable. Test a range of U values, e.g., 4-7 eV, which is more typical for Yb.
Modify:
LDAUU = 3 0 0 6
Since you are using LDAUTYPE = 2 (Dudarev), ensure that Ueff = U - J is meaningful. You might set:
LDAUJ = 0 0 0 0
Yb's strongly correlated states often cause SCF instabilities. Increase the maximum number of SCF cycles (NELM) to ensure convergence:
NELM = 200
Reduce mixing parameters to stabilize the SCF cycle. Add or adjust:
AMIX = 0.1
BMIX = 0.01
Switch the electronic minimization algorithm (ALGO) to a more stable setting for difficult cases involving f-electrons. Try the conjugate gradient (ALGO = All) or the block Davidson (ALGO = Fast) solver:
ALGO = All
Proper initialization of MAGMOM is crucial for systems with partially filled f-orbitals. Instead of assigning 4.0 for Yb's magnetic moment (which could be unrealistic), try:
MAGMOM = 79*0.6 1*1.0
Start with a smaller or neutral moment for Yb (1.0 or 0.6), as improper initialization can hinder convergence.
High-precision settings may cause aliasing errors, especially with large systems and f-electrons. Temporarily lower PREC to Normal and adjust the plane wave cutoff slightly:
PREC = Normal
ENCUT = 450
Explicitly set the FFT grid dimensions (NGX, NGY, NGZ) to resolve aliasing warnings:
NGX = 82
NGY = 82
NGZ = 58
Since the calculation is already unstable, smaller ionic step sizes might help:
POTIM = 0.3
Verify that the pseudopotential for Yb (PAW_PBE Yb 23Dec2003) is suitable for systems requiring f-electron treatment. Consider using Yb_sv or a GW-ready pseudopotential if available.
If the SCF calculation fails after several iterations, restart from a pre-converged charge density and wavefunction:
ICHARG = 1 # Read from CHGCAR
LWAVE = .TRUE. # Write the wavefunction file
I hope this can help
Mustapha Lasmi
Thank you for your detailed response. I tried using your method to improve my INCAR.I change my potcar to Yb_3 (because I don't have Yb_sv).It works when EDIFF = 1E-02, but it does not converge when EDIFF = 1E-04. The calculation terminates without any significant errors
The problem can be caused by several factors
Let's check one of them
Please try to use the converged WAVECAR and CHGCAR from EDIFF =1E-02 calculations for a stricter calculation
ICHARG = 1
Then set EDIFF = 1E-04
Provide me the results
This is just to confirm Mustaph Lasmi's pertinent remark on the value of U for Yb. Many years ago we performed many-body calculations on the 4f spectrum of YbP and found that the optimum value for U was 6.4 eV. The full text of the paper is available on RG. The reference is: Towards a Self-Consistent Determination of the Mixing Parameters in the Anderson Model: An Application to the 4𝑓-Excitation Spectra of YbP. Phys. Rev. Lett. 56, 2744 (1986).
Currently, the Yb calculation stops when EDIFF is set to 10⁻⁶, with part of the error message from the .e file indicating that it might be due to insufficient disk space. My WAVECAR file size has reached around 4000 MB.
In the INCAR file, I only modified ISIF and EDIFF. I also tried running the calculation without reading WAVECAR and CHGCAR, but the process disconnected from the server before completion. Additionally, reading only CHGCAR still led to a calculation failure.
Would it be better to use a more reliable server for the calculation?
forrtl: Input/output error
forrtl: severe (38): error during write, unit 17, file
...../OSZICAR
Image PC Routine Line Source
vasp-std 00000000012CDC7B for__io_return Unknown Unknown
vasp-std 000000000131E96C for_write_seq_fmt Unknown Unknown
vasp-std 000000000131C990 for_write_seq_fmt Unknown Unknown
vasp-std 0000000001291226 Unknown Unknown Unknown
vasp-std 0000000000411462 Unknown Unknown Unknown
libc-2.17.so 00002B905348F555 __libc_start_main Unknown Unknown
vasp-std 0000000000411369 Unknown Unknown Unknown
forrtl: severe (28): CLOSE error, unit 8, file "Unknown"
Image PC Routine Line Source
vasp-std 00000000012D2B02 for__exit_handler Unknown Unknown
vasp-std 00000000012CF13C for__issue_diagno Unknown Unknown
vasp-std 00000000012CDC7B for__io_return Unknown Unknown
vasp-std 000000000131E96C for_write_seq_fmt Unknown Unknown
vasp-std 000000000131C990 for_write_seq_fmt Unknown Unknown
vasp-std 0000000001291226 Unknown Unknown Unknown
vasp-std 0000000000411462 Unknown Unknown Unknown
libc-2.17.so 00002B905348F555 __libc_start_main Unknown Unknown
vasp-std 0000000000411369 Unknown Unknown Unknown
--------------------------------------------------------------------------
Primary job terminated normally, but 1 process returned
a non-zero exit code. Per user-direction, the job has been aborted.
--------------------------------------------------------------------------
forrtl: error (78): process killed (SIGTERM)
Image PC Routine Line Source
vasp-std 00000000012DAD24 for__signal_handl Unknown Unknown
libpthread-2.17.s 00002B66D000A630 Unknown Unknown Unknown
mca_btl_vader.so 00002B66E1440C1B Unknown Unknown Unknown
libopen-pal.so.40 00002B66D0CFB584 opal_progress Unknown Unknown
libmpi.so.40.10.3 00002B66CFA1B958 ompi_request_defa Unknown Unknown
libmpi.so.40.10.3 00002B66CFA6F53A ompi_coll_base_bc Unknown Unknown
libmpi.so.40.10.3 00002B66CFA6F6DE ompi_coll_base_bc Unknown Unknown
mca_coll_tuned.so 00002B66E36F9BC6 ompi_coll_tuned_b Unknown Unknown
libmpi.so.40.10.3 00002B66CFA32C77 MPI_Bcast Unknown Unknown
libmpi_mpifh.so.4 00002B66CF795CE4 pmpi_bcast Unknown Unknown
vasp-std 0000000000414534 Unknown Unknown Unknown
vasp-std 0000000001297B4C Unknown Unknown Unknown
vasp-std 0000000000411462 Unknown Unknown Unknown
libc-2.17.so 00002B66D0239555 __libc_start_main Unknown Unknown
vasp-std 0000000000411369 Unknown Unknown Unknown
forrtl: error (78): process killed (SIGTERM)
Image PC Routine Line Source
vasp-std 00000000012DAD24 for__signal_handl Unknown Unknown
libpthread-2.17.s 00002AF48E126630 Unknown Unknown Unknown
libopen-pal.so.40 00002AF48EE175D9 opal_progress Unknown Unknown
libmpi.so.40.10.3 00002AF48DB37958 ompi_request_defa Unknown Unknown
libmpi.so.40.10.3 00002AF48DB8B241 ompi_coll_base_bc Unknown Unknown
libmpi.so.40.10.3 00002AF48DB8B6DE ompi_coll_base_bc Unknown Unknown
mca_coll_tuned.so 00002AF4A186ABC6 ompi_coll_tuned_b Unknown Unknown
libmpi.so.40.10.3 00002AF48DB4EC77 MPI_Bcast Unknown Unknown
libmpi_mpifh.so.4 00002AF48D8B1CE4 pmpi_bcast Unknown Unknown
vasp-std 0000000000414534 Unknown Unknown Unknown
vasp-std 0000000001297B4C Unknown Unknown Unknown
vasp-std 0000000000411462 Unknown Unknown Unknown
libc-2.17.so 00002AF48E355555 __libc_start_main Unknown Unknown
vasp-std 0000000000411369 Unknown Unknown Unknown
forrtl: error (78): process killed (SIGTERM)
Image PC Routine Line Source
vasp-std 00000000012DAD24 for__signal_handl Unknown Unknown
libpthread-2.17.s 00002B9F6E25F630 Unknown Unknown Unknown
mca_btl_vader.so 00002B9F7B689F72 Unknown Unknown Unknown
libopen-pal.so.40 00002B9F6EF50584 opal_progress Unknown Unknown
libmpi.so.40.10.3 00002B9F6DC70958 ompi_request_defa Unknown Unknown
libmpi.so.40.10.3 00002B9F6DCC4241 ompi_coll_base_bc Unknown Unknown
libmpi.so.40.10.3 00002B9F6DCC46DE ompi_coll_base_bc Unknown Unknown
mca_coll_tuned.so 00002B9F81A98BC6 ompi_coll_tuned_b Unknown Unknown
libmpi.so.40.10.3 00002B9F6DC87C77 MPI_Bcast Unknown Unknown
libmpi_mpifh.so.4 00002B9F6D9EACE4 pmpi_bcast Unknown Unknown
vasp-std 0000000000414534 Unknown Unknown Unknown
vasp-std 0000000001297B4C Unknown Unknown Unknown
vasp-std 0000000000411462 Unknown Unknown Unknown
libc-2.17.so 00002B9F6E48E555 __libc_start_main Unknown Unknown
vasp-std 0000000000411369 Unknown Unknown Unknown
forrtl: error (78): process killed (SIGTERM)
The error is likely caused by insufficient disk space or memory.
Use LWAVE = .FALSE. and optimize resource usage by reducing file sizes, memory requirements, and parallelism.
If problems persist, switching to a more robust server may be necessary.
Modified INCAR for Resource Optimization
Here’s a modified INCAR that minimizes disk usage and improves resource management:
EDIFF = 1E-06
ISIF = 2
LWAVE = .FALSE.
LCHARG = .FALSE.
PREC = Accurate
AMIX = 0.2
BMIX = 0.001
NELM = 200
NBANDS = (default or slightly reduced)
ICHARG = 2
NWRITE = 3
I hope this can help
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Powders