Comparison of results from exciting and wien2k

159 views
Skip to first unread message

Natalie Holzwarth

unread,
Nov 27, 2012, 7:50:32 PM11/27/12
to excit...@googlegroups.com
Dear exciting groups,
      Thanks very much for sharing your code.   So far I have found it to be easy to use compared to wien2k.   Unfortunately so far, the results using exciting are quite different from what I believe are well-converged results for the same material using wien2k.   I am not sure whether you could perhaps spot some silly errors in my use of your code.    The test case is a fake fluorite stucture with the composition of SiO2.    the input.xml file and a bashscript used to generate the attached plot are given below.  I used the species files supplied on line and mostly default parameters.  (The rmt parameter did not seem to have an effect.)   The attached plot shows the binding energy (Ry) versus lattice constant (in bohr) for exciting and for wien2k.   Thanks in advance for any suggestions.    SIncerely, Natalie Holzwarth

input.xml:<input>
    <title>SiO2 fakefluorite </title>
    <structure speciespath="/home/natalie/EL6/publiccode/exciting/excitingNov2012/species/">
        <crystal scale="8.0">
            <basevect>0.0   0.5   0.5</basevect>
            <basevect>0.5   0.0   0.5</basevect>
            <basevect>0.5   0.5   0.0</basevect>
        </crystal>
        <species speciesfile="Si.xml">
                 rmt="1.5"
                 <atom coord="0.00 0.00 0.00" />
        </species>
        <species speciesfile="O.xml">
                 rmt="1.5"
                 <atom coord="0.25 0.25 0.25" />
                 <atom coord="-0.25 -0.25 -0.25" />
        </species>
    </structure>
    <groundstate ngridk="8 8 8"
                 swidth="0.0005"
                 rgkmax="8.0"
                 xctype="LSDAPerdew-Wang">
    </groundstate>
</input>




#!/bin/bash

excitingser
cp INFO.OUT INFO.OUT.8.0
echo -n 8.0 > lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.1"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.1
echo -n 8.1 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.2"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.2
echo -n 8.2 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.3"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.3
echo -n 8.3 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.4"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.4
echo -n 8.4 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.5"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.5
echo -n 8.5 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.6"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.6
echo -n 8.6 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.7"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.7
echo -n 8.7 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.8"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.8
echo -n 8.8 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="8.9"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.8.9
echo -n 8.9 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

sed 's/crystal scale="8.0"/crystal scale="9.0"/' input.xml.save >input.xml
excitingser
cp INFO.OUT INFO.OUT.9.0
echo -n 9.0 >> lattice.dat
grep 'total energy        ' INFO.OUT |tail -1 >> lattice.dat

output -- lattice.dat


8.0 total energy                :     -438.717587265    
8.1 total energy                :     -438.723398714    
8.2 total energy                :     -438.727210530    
8.3 total energy                :     -438.729088319    
8.4 total energy                :     -438.729178866    
8.5 total energy                :     -438.727603924    
8.6 total energy                :     -438.724489210    
8.7 total energy                :     -438.720023768    
8.8 total energy                :     -438.714328367    
8.9 total energy                :     -438.707209496    
9.0 total energy                :     -438.699429264    

SnS2fakefluorite.pdf

Pablo García Risueño

unread,
Nov 28, 2012, 4:55:24 AM11/28/12
to excit...@googlegroups.com
Dear friend

I afraid I cannot properly understand your problem. In the picture you display, both curves look pretty similar. Could you explain us a little bit more about the problem.

You can increase the size of your basis by increasing the parameter rgkmax (e.g. to 9), what should result in greater accuracy.

Best regards

Pablo García Risueño

unread,
Nov 28, 2012, 9:02:34 AM11/28/12
to excit...@googlegroups.com
Hello! A mate points what follows. He says that the unit vectors of Exciting are wrong, because at your system the unit cell has a .neq. c:


Two things at first sight

Cell Dimensions for Quartz:         a = 4.9133, c = 5.4053 (Angstrom)


reference
http://www.webmineral.com/data/Quartz.shtml

No idea what kind of W2K has been done

Natalie Holzwarth

unread,
Nov 28, 2012, 12:37:37 PM11/28/12
to excit...@googlegroups.com

Sorry for including the incorrect graph with the original post -- the correct graph is now included.  (vertical axis binding energy  in Ry and horizontal axis is cubic lattice constant in  bohr.    I was trying to run a very simple structure for testing (not quartz) and compare exciting and wien2k with the same calculation.   A previous check for Si in the diamond structure was very encouraging.   However, this check with SiO2 in the fluorite structure shows very different results for exciting and wien2k.    In any case, thanks for your feedback.  
                                                                 Sincerely, Natalie Holzwarth

Pablo García Risueño

unread,
Nov 28, 2012, 12:55:25 PM11/28/12
to excit...@googlegroups.com
Some other mate recommended to send the input and species files from Exciting and Wien2k, too (so that everybody can check that your input systems are as similar as possible).

Natalie Holzwarth

unread,
Nov 28, 2012, 1:44:10 PM11/28/12
to excit...@googlegroups.com
Dear Pablo,
Thanks for your suggestion.  For exciting, I used the species file from the download tar file which I believe is the same as the one on line.   For wien2k, I have attached some of the input files.   A modified optimize script was used to run with the same lattice constants as for exciting.    thanks again for your feedback.   Natalie Holzwarth  Actually -- google would not let me attach the files so I list them below

SiO2.in0
TOT    5    (5...CA-LDA, 13...PBE-GGA, 11...WC-GGA)
NR2V      IFFT      (R2V)
  36  36  36    2.00  1    min IFFT-parameters, enhancement factor, iprint

SiO2.in1
WFFIL  EF=.56485   (WFFIL, WFPRI, ENFIL, SUPWF)
  9.00       12    6 (R-MT*K-MAX; MAX L IN WF, V-NMT
  0.30    3  0      (GLOBAL E-PARAMETER WITH n OTHER CHOICES, global APW/LAPW)
 1    0.30      0.000 CONT 1
 1   -6.73      0.001 STOP 1
 0    0.30      0.000 CONT 1
  0.30    3  0      (GLOBAL E-PARAMETER WITH n OTHER CHOICES, global APW/LAPW)
 0   -1.44      0.002 CONT 1
 0    0.30      0.000 CONT 1
 1    0.30      0.000 CONT 1
K-VECTORS FROM UNIT:4  -12.9       2.5    24   emin/emax/nband #red


SiO2.in2
TOT             (TOT,FOR,QTL,EFG,FERMI)
     -15.9      22.0 0.50 0.05                EMIN, NE, ESEPERMIN, ESEPER0
GAUSS    0.001          (GAUSS,ROOT,TEMP,TETRA,ALL      eval)
  0 0  4 0  4 4  6 0  6 4
  0 0  4 0  4 4  6 0  6 4 -3 2
 14.00          GMAX
NOFILE        FILE/NOFILE  write recprlist


SiO2_8.0.struct
SiO2-fake fluorite                                                            
F   LATTICE,NONEQUIV.ATOMS:  2225_Fm-3m                                       
MODE OF CALC=NREL unit=bohr                                                   
  8.000000  8.000000  8.000000 90.000000 90.000000 90.000000                  
ATOM   1: X=0.00000000 Y=0.00000000 Z=0.00000000
          MULT= 1          ISPLIT= 2
Si         NPT=  781  R0=0.00001000 RMT=    1.5000   Z: 14.0                  
LOCAL ROT MATRIX:    1.0000000 0.0000000 0.0000000
                     0.0000000 1.0000000 0.0000000
                     0.0000000 0.0000000 1.0000000
ATOM   2: X=0.25000000 Y=0.25000000 Z=0.25000000
          MULT= 2          ISPLIT= 2
       2: X=0.75000000 Y=0.75000000 Z=0.75000000
O          NPT=  781  R0=0.00001000 RMT=    1.5000   Z:  8.0                  
LOCAL ROT MATRIX:    1.0000000 0.0000000 0.0000000
                     0.0000000 1.0000000 0.0000000
                     0.0000000 0.0000000 1.0000000
  48      NUMBER OF SYMMETRY OPERATIONS
 1 0 0 0.00000000
 0-1 0 0.00000000
 0 0-1 0.00000000
       1
 1 0 0 0.00000000
 0 0-1 0.00000000
 0-1 0 0.00000000
       2
-1 0 0 0.00000000
 0-1 0 0.00000000
 0 0-1 0.00000000
       3
-1 0 0 0.00000000
 0 0-1 0.00000000
 0-1 0 0.00000000
       4
 0 1 0 0.00000000
-1 0 0 0.00000000
 0 0-1 0.00000000
       5
 0 0 1 0.00000000
-1 0 0 0.00000000
 0-1 0 0.00000000
       6
 0 1 0 0.00000000
 1 0 0 0.00000000
 0 0-1 0.00000000
       7
 0 0 1 0.00000000
 1 0 0 0.00000000
 0-1 0 0.00000000
       8
 0 1 0 0.00000000
 0 0-1 0.00000000
-1 0 0 0.00000000
       9
 0 0 1 0.00000000
 0-1 0 0.00000000
-1 0 0 0.00000000
      10
 0 1 0 0.00000000
 0 0-1 0.00000000
 1 0 0 0.00000000
      11
 0 0 1 0.00000000
 0-1 0 0.00000000
 1 0 0 0.00000000
      12
 0-1 0 0.00000000
-1 0 0 0.00000000
 0 0-1 0.00000000
      13
 0-1 0 0.00000000
 1 0 0 0.00000000
 0 0-1 0.00000000
      14
 0 0-1 0.00000000
-1 0 0 0.00000000
 0-1 0 0.00000000
      15
 0 0-1 0.00000000
 1 0 0 0.00000000
 0-1 0 0.00000000
      16
 1 0 0 0.00000000
 0 1 0 0.00000000
 0 0-1 0.00000000
      17
-1 0 0 0.00000000
 0 1 0 0.00000000
 0 0-1 0.00000000
      18
 1 0 0 0.00000000
 0 0 1 0.00000000
 0-1 0 0.00000000
      19
-1 0 0 0.00000000
 0 0 1 0.00000000
 0-1 0 0.00000000
      20
 0-1 0 0.00000000
 0 0-1 0.00000000
-1 0 0 0.00000000
      21
 0 0-1 0.00000000
 0-1 0 0.00000000
-1 0 0 0.00000000
      22
 0-1 0 0.00000000
 0 0-1 0.00000000
 1 0 0 0.00000000
      23
 0 0-1 0.00000000
 0-1 0 0.00000000
 1 0 0 0.00000000
      24
 0 0 1 0.00000000
 0 1 0 0.00000000
-1 0 0 0.00000000
      25
 0 1 0 0.00000000
 0 0 1 0.00000000
-1 0 0 0.00000000
      26
 0 0 1 0.00000000
 0 1 0 0.00000000
 1 0 0 0.00000000
      27
 0 1 0 0.00000000
 0 0 1 0.00000000
 1 0 0 0.00000000
      28
 1 0 0 0.00000000
 0 0-1 0.00000000
 0 1 0 0.00000000
      29
-1 0 0 0.00000000
 0 0-1 0.00000000
 0 1 0 0.00000000
      30
 1 0 0 0.00000000
 0-1 0 0.00000000
 0 0 1 0.00000000
      31
-1 0 0 0.00000000
 0-1 0 0.00000000
 0 0 1 0.00000000
      32
 0 0 1 0.00000000
-1 0 0 0.00000000
 0 1 0 0.00000000
      33
 0 0 1 0.00000000
 1 0 0 0.00000000
 0 1 0 0.00000000
      34
 0 1 0 0.00000000
-1 0 0 0.00000000
 0 0 1 0.00000000
      35
 0 1 0 0.00000000
 1 0 0 0.00000000
 0 0 1 0.00000000
      36
 0 0-1 0.00000000
 0 1 0 0.00000000
-1 0 0 0.00000000
      37
 0-1 0 0.00000000
 0 0 1 0.00000000
-1 0 0 0.00000000
      38
 0 0-1 0.00000000
 0 1 0 0.00000000
 1 0 0 0.00000000
      39
 0-1 0 0.00000000
 0 0 1 0.00000000
 1 0 0 0.00000000
      40
 0 0-1 0.00000000
-1 0 0 0.00000000
 0 1 0 0.00000000
      41
 0-1 0 0.00000000
-1 0 0 0.00000000
 0 0 1 0.00000000
      42
 0 0-1 0.00000000
 1 0 0 0.00000000
 0 1 0 0.00000000
      43
 0-1 0 0.00000000
 1 0 0 0.00000000
 0 0 1 0.00000000
      44
 1 0 0 0.00000000
 0 0 1 0.00000000
 0 1 0 0.00000000
      45
 1 0 0 0.00000000
 0 1 0 0.00000000
 0 0 1 0.00000000
      46
-1 0 0 0.00000000
 0 0 1 0.00000000
 0 1 0 0.00000000
      47
-1 0 0 0.00000000
 0 1 0 0.00000000
 0 0 1 0.00000000
      48


optimize.job
#!/bin/csh -f
 #   Modify this script according to your needs:
 #      Uncomment one of the lines ...
 #      Change run_lapw to runsp_lapw or use different convergence criterium
 #      Change save_lapw -d XXX
foreach i ( \
                                         SiO2_08.0\
                                         SiO2_08.1\
                                         SiO2_08.2\
                                         SiO2_08.3\
                                         SiO2_08.4\
                                         SiO2_08.5\
                                         SiO2_08.6\
                                         SiO2_08.7\
                                         SiO2_08.8\
                                         SiO2_08.9\
                                         SiO2_09.0\
 )
     rm SiO2.struct         # NFS-bug
     cp  $i.struct SiO2.struct
 
 # Please uncomment and adapt any of the lines below according to your needs
 #    cp  $i.clmsum C.clmsum
     x dstart
 #     x dstart -c
 #    run_lapw -ec 0.0001 -in1new 3 -in1orig -renorm
 #    runsp_lapw -ec 0.0001
 #    min -I -j "run_lapw -I -fc 1.0 -i 40 "
     run_lapw -ec 0.000001
  
     set stat = $status
     if ($stat) then
        echo "ERROR status in" $i
        exit 1
     endif
     save_lapw  $i
 #    save_lapw  -f -d XXX $i
 end
Reply all
Reply to author
Forward
0 new messages