UT theoretical chemistry code forum

dear all,
Acualy I work with the vasp.4.6 code. I'im interssing in the diffusion impurities in transition metals, so I would like to use the neb method to finf the saddle point. I have seen in the manual of vasp, I find this : Vasp.4.X supports the elastic band method to calculate energy barriers. The INCAR, KPOINTS, and POTCAR files must be located in the directory in which VASP is started. In addition, a set of subdirectories (numbered 00,01,02...) must be created, and each subdirectory must contain one POSCAR file. The tag

IMAGES= number of images

(specified in the INCAR file) forces VASP to run the elastic band method. The number of nodes must be dividable by the number of images (the NPAR switch can still be used as described above). VASP divides the nodes in groups, and each group then works on one ``image''. The first group of nodes reads the POSCAR file from the directory 01, the second group from 02 etc. In the elastic band method, the endpoints are kept fixed, and the position of the end points must be supplied in the files 00/POSCAR and XX/POSCAR, where XX is

XX=number of images+1

My question : is the poscar located in the each subderactory the same, if not how tfind the poscar more appropriate for each subderactory.

I wait any typical example for the neb method in vasp.
Thanks for your time and your help
Warda

To set up a NEB calculation, using our toolbox, you begin with your initial and final images (say POSCAR_1 and POSCAR_2). Then you use the nebmake.pl script available as a part of vtstsscripts (http://theory.cm.utexas.edu/vtsttools/downloads/). See http://theory.cm.utexas.edu/vtsttools/scripts/ for clarifications.
That will generate a set of POSCARs that are a linear interpolation between the initial and final images. Then you set the auxiliary tags in your INCAR file as listed on http://theory.cm.utexas.edu/vtsttools/neb/ . It is usually good to start with 8 images (running on 8, 16, etc. processors). When everything is setup then VASP is launched from the parent directory, i.e. the one containing the 00, 01, 02 etc. subdirectories. INCAR, KPOINTS and POTCAR files should be in the parent directory and only POSCARs in the 01,02,etc. subdirectories. The end images (e.g. if using 8 images then it would be 00 and 09) should also contain the OUTCAR from the run used to minimize those structures.

I think that's about it.

Note that for this to work you MUST download our vtstscripts and place in your $PATH as well as recompiling VASP with our additional .F files ... see http://theory.cm.utexas.edu/vtsttools/downloads/.

Hopefully this helps - happy VASP-ing

Thank you andri, I will try with following your suggestion.
I have just test the nebmake.pl, it works but Idon't know if the other moduls located vtstcode.tar.gz are compiled in the parallel VASP (vasp.4.6.3) because all my calculation are done in central calculation in parallel version. I can't access to recompile Vasp with the other modul located vtstcode.tar.gz.
Thanks andri

The vtstscripts packages contains many scripts that rely on output generated by our additions to VASP in vtstcode. But the nebmake.pl should work fine for you (as you have tried) even though you don't have the files in vtstcode compiled with VASP. Most of the scripts though you will have to use at your own risk - probably all or most of them that deal with extracting results from the OUTCAR will not work properly for you, without adding vtstcode to VASP and recompile.

Andri,
I Have just tried to run the nebmake.pl ( POSCAR1 POSCAR2 6) where vasp.pm located. As output, I haved 6 subderectory created wich contains POSCAR files. but I don't inderstand the importance of chain.F dimer.F dynmat.F lanczos.F neb.F located vtstcode.tar.gz.

Thanks andri and sorry for my slow comprehension

What you have done is fine. You can generate an neb using nebmake.pl and run the standard vasp code and find a converged minimum energy path.

The source files in vtstcode.tar.gz provide transition state extensions to the vasp code. These source files can be built and linked into the vasp program. These extensions allow for saddle point finding using the min-mode following methods, and recent modifications to the NEB, such as the climbing image, a better tangent, and double nudging. In many cases, these extensions write information to the OUTCAR, or other output files which are used by the vtst scripts. If you do not rebuild vasp with the vtstcode source files, the extensions will not be available and any related scripts will not function properly.

thanks for all clarification, but Istill anbiguous in the find of parameter wich we must introduce in the incar file in order the neb method work fine