I am a new user of ABINIT and have recently been studying the non-linear tutorial, following the tutorial using pead-lda to calculate the AlP structure can work properly. However, whPreformatted texten I calculate the nlo of two crystal structures, Ba(BO2)2 (hereinafter referred to as BBO) and BaTiO3(hereinafter referred to as BTO), both structures are calculated using pead-lda, and tnlo_2o_DS4_DDB and tnlo_2o_DS5_DDB can be obtained in tnlo_2. However, when tnlo_3 is calculated, the BBO structure cannot be combined with mrgddb to generate TNLO_3.DDB.OUT file, and the data of 2nd and 3rd derivatives are always missing. The BTO can successfully generate tnlo_3.ddb.out. However, in the calculation of tnlo_4, there will be an error “Cannot find block corresponding to non-linear optical susceptibilities in DDB file”. I do not understand what the problem is. Is there any problem with the parameters or structure I set? I hope you can help me. The calculation input files and corresponding output files of BTO have been attached. Please give me some guidance. Thank you so much.
in addition,I also installed ABINIT-9.10 and ABINIT-10.3 on my small workstation, but both versions reported the same error.
Your problem is probably atpol and related variables. See its definition in the online doc: it takes 2 integer values only, which are 1st and last atom. Usually you will want
rfatpol 1
This is why you are missing some 2nd and 3rd derivatives involving atoms 3 to natom
Thank you very much, I finally made it through this, but there is a new error “The lowest mode appears to be a “true” negative mode,
and not an acoustic mode. This precludes the computation
of the frequency-dependent dielectric tensor.
Action : likely there is no action to be taken, although you,
could try to raise your convergence parameters (ecut and k-points).
For your information, here are the four lowest frequencies :
-1.31528628E-03 -1.16649197E-03 -1.16649197E-03 -9.11083074E-04” when tnlo_4.
Therefore, I re-sorted out the calculation process of this tutorial, first of all, we should select appropriate ecut and nkpt values, and carry out their appropriate values in accordance with the base3 tutorial. I wonder whether tnlo_2.abi can be calculated directly after selecting appropriate nkpt and ecut in accordance with the base3 tutorial. Is there anything missing in the process?
Well, if the structure is dynamically unstable then you have to break the symmetry and find the ground state. For BTO at low T this is certainly not the cubic phase. I don’t know the diagram of BBO but it’s probably similar with tetragonal or orthorhombic phases at lower T.
Once the harmonic frequencies are all stabilized you can run the NLO, but of course the number of atoms will be higher and it will be heavier numerically.
In passing, for high T phases you can have a look at anharmonic dynamics which renormalize the harmonic frequencies, this is done in the TDEP code (TDEP developers · GitHub) and also in a module of abinit called aTDEP (aTDEP1 - abinit). Disclaimer: this is quite a bit more work, not a simple post-processing step.