IDM

go back to Main Page, Computational Resources, Chemistry & More, Computational Codes, VASP, Scripts for VASP

Instructions

Before start, please read about the Improved Dimer Method [[1]].

1. Generate your initial guess for the structure (POSCAR). To this end, you can either pre-converge with a NEB calculation, or do it by hand.

2. Be sure the metal or solid phase is frozen and the molecule is relaxed. Do a frequency calculation based on your initial guess. Specify the following in your INCAR file:

frequencies:
  IBRION =   5     #     
  EDIFF  =  1E-6   # 
  POTIM  =   0.010 #                  
  NFREE  =   2     # 

3. Based on your POSCAR and OUTCAR of the frequencies calculation, generate in a new folder the input of the IDM with the getdimer script. For example:

getdimer ../freq/POSCAR ../freq/OUTCAR POSCAR

4. Specify the following in your INCAR file:

improved dimer method:
  IBRION =  44     # IDM    
  EDIFF  =   1E-6  # 
  EDIFFG =  -0.030 # use force criteria
  POTIM  =   0.050 # Between 0.010 and 0.100.                  
  NFREE  =   2     # Between 2&5
  FINDIFF=   2     # 
  DIMER_DIST=0.010 #               
  MINROT =   0.010 # 
  STEP_SIZE= 0.010 #                
  STEP_MAX=  0.100 #                
  NSW    = 100     # Never use more than 200. Check periodically for divergence.

5. To check that it is converging well, (1) check with p4vasp that all forces are small, and (2) check that the curvature is always negative. Normal values are between -1.0 and -30.0:

grep curv OUTCAR

Script and Executable

Instructions

1. Download the program File:Dimerdiag.tgz. Extract it and put it in your ~/bin/exe/ folder

2. Copy the script in the ~/bin folder. Recommended name: getdimer.

3. Edit the script; change "rgarcia" by your username in this part:

# PART 3 - DIAGONALIZE 
/home/rgarcia/bin/exe/dimerdiag 

4. Now you can go to any folder and generate your dimer:

$ getdimer <input:POSCAR> <input:OUTCAR> <output:POSCAR>

SCRIPT

#!/bin/bash
# 
# Generates the POSCAR of a dimer calculation from the non mass-weighted Hessian matrix
# Rodrigo García-Muelas
# November 14th, 2014 - Created
# March     9th, 2015 - Debugged 1 (Thanks to Marçal & Luca)
# 
# Use with VASP 5.x or higher
# 
# INPUT
# $1 Path of the POSCAR file containing the coordinates
# $2 Path of the OUTCAR file containing the frequencies of the structure
# $3 <OUTPUT> POSCAR file
# 

# PART 0 - SECURITY CHECKS

if [ -e $3 ] ; then
echo "Warning! $3 already exist. Overwrite? (y/Y for yes)"
read  overwrite
case $overwrite in
y|Y|Yes|yes|YES) echo "$3 will be overwriten"          ;;
*)               echo "No actions taken "     ; exit 1 ;;
esac
fi

IBRION=`grep IBRION $2 | awk '{print $3}'`
if  [[  $IBRION -gt "8" || $IBRION -lt "5" ]] ; then
echo "$2 is not an OUTCAR of a frequency job."
echo "No actions taken" ; exit 1 ; fi

# PART 1 - READ POSCAR AND PREPARE INPUTS 

cp $1 diagonalizer_poscar.tmp 

# PART 2 - READ OUTCAR 

grep "Degrees of freedom DOF " $2 > temp01.tmp
freedom=`awk '{print $6}' temp01.tmp` 
echo "$freedom" > diagonalizer_matrix.tmp

grep -A $(($freedom+2)) 'SECOND DERIVATIVES' $2 > temp02.tmp
cut -b 5- temp02.tmp > temp03.tmp
tail -n $freedom temp03.tmp >> diagonalizer_matrix.tmp

# PART 3 - DIAGONALIZE 
/home/rgarcia/bin/exe/dimerdiag 

# PART 4 - CLEAN
head -n 9 $1 > $3
cat    diagonalizer_taildm.tmp >> $3
mv -f  diagonalizer_output.tmp eigenvectors.dat
rm -fv *.tmp

Alternative (old) versions

WARNING: Some of the topics discussed by David Karhánek are outdated. For instance, it is not necessary to set NWRITE=3 for the frequency calculation: IDM-IRC get-dimer-dir.sh get-irc-path.sh (by David Karhanek)

--Rgarcia 12:32, 18 November 2014 (CET)

go back to Main Page, Computational Resources, Chemistry & More, Computational Codes, VASP, Scripts for VASP