jrschmidt2 / periodic-nbo Goto Github PK

Hi Professor Schmidt,

I'm trying to do a periodic NBO analysis on CeO2 structure by VASP. I used def2-SVP basis from https://www.basissetexchange.org/.
After doing the static calculation with LNBO=T in INCAR, I can get wavefunction.dat. After that, I want to do: projection.exe basis.inp wavefunction.dat NBO.out, in which I put the def2-SVP basis in basis.inp file. However this command cannot work with the error "Unreadable basis function type in basis input file, pre-screen". I once assumed it was the error of the basis, but I've done some tests and this basis works for light elements, such as C and O. So, if I want to do a periodic NBO analysis on some heavy or rare-earth elements, may it be possible to use some basis that works for periodic NBO?

Thank you.

Best,
kr

for intel/2019.4 or newer, FFLAGS in "periodic-NBO/projection/Makefile" should also include -I/${MKLROOT}/include/intel64/lp64 when making projection.ext

When I run nbo.exe in the directory 'sample_systems/Nickel', the error was printed:

forrtl: severe (64): input conversion error, unit 10, file /home/xiazm/bin/vasp.5.4.4-ssAdNDP/sample_systems/Nickel/NBO.out
Image PC Routine Line Source
nbo.exe 000000000046F7AE Unknown Unknown Unknown
nbo.exe 000000000048E754 Unknown Unknown Unknown
nbo.exe 000000000048CAC2 Unknown Unknown Unknown
nbo.exe 0000000000459B1A Unknown Unknown Unknown
nbo.exe 000000000044D975 Unknown Unknown Unknown
nbo.exe 000000000040389E Unknown Unknown Unknown
libc-2.17.so 00002B5967B11445 __libc_start_main Unknown Unknown
nbo.exe 00000000004037A9 Unknown Unknown Unknown

The nbo.exe was compiled by the ifort compiler of intel2018 and MKL. The Makefile.template is:

#use: mkmf -t Makefile.template -p nbo.exe
#to generate a proper makefile

MODULE_PATH=/opt/intel/compilers_and_libraries_2018.1.163/linux/mkl/include/intel64/lp64
MKL_LIB=/opt/intel/compilers_and_libraries_2018.1.163/linux/mkl/lib/intel64

FC=ifort
LD=ifort
FFLAGS= -qopenmp -I$(MODULE_PATH)
LDFLAGS=-L$(MKL_LIB) -lmkl_blas95_lp64 -lmkl_lapack95_lp64 -qopenmp -lpthread -I $(MODULE_PATH) -mkl=sequential

I compiled the nbo.exe without any errors or warnnings.

Both the vasp and projection steps were successful. But the nbo.exe filed to run.

[Uploading ba

test.zip

sis.inp…]()

Dear Prof. Schmidt,

I have started using your quite lately. It runs fine on most of the systems but fails on molecules like CO or H2O. I don't get any errors until generating wavefunction.dat file. Everything finishes cleanly. Afterwards, when I use the projection code, I get the following error at the end.

Testing of projected density and Fock matrices
Bloch-space energy sum (NaN,NaN)
Bloch-space valence electrons (NaN,NaN)

could you please suggest, what could be the possible reason for the error? I have attached full output of the projection program in output.

I am waiting for your kind feedback.

Best regards,
Anurag Vohra

projection.txt

Hello,

We are currently working with VASP6.4 and are interested in using this tool to assess the types of bonds and order of bonds that nitrogen containing adsorbates make to a Pt(111) surface. We would expect based off of Lewis dot structure drawings that some of these adsorbates should be physiosorbed, but we see them sit very close to the surface and have a high binding energy. It is our hope that we could use this tool to decipher what bonds are present.

Currently it looks like this tool is only built for up to VASP5.4, would it be possible to update the Patch file to work for VASP6.4? Would that be all that would need to change, or would other files and install instructions change too? Let me know if it's feasible to update periodic NBO to work with VASP6.4.

Thank you,
Kirk