The unfolded spectral weight can be visualized by an intensity map
that the weight is smeared out by a Lorentian function:
|
(1) Compilation of intensity_map.c
In the the directory 'source', please compile 'intensity_map.c' as
gcc intensity_map.c -lm -o intensity_mapand copy the executable file 'intensity_map' to your work directory.
(2) Generation of the intensity map
After finising the unfolding calculation, you can generate a file storing a mesh data
for drawing the intensity map using 'intensity_map'.
For the case of SiC () supercell in a two-dimensional honeycomb structure
with a Si vacancy discussed in the previous subsection, where the input file is 'SiC_C_SP_V.dat',
e.g., one can generate a file 'sic-intmap.txt ' storing the mesh data by
./intensity_map sic_c_sp_v.unfold_totup -c 3 -k 0.1 -e 0.1 -l -10 -u 6 > sic-intmap.txtwhere the arguments have the following meaning:
-c column of spectral weight you analyze -k degree of smearing (Bohr^{-1}) in {\bf k}-vector -e degree of smearing (eV) in energy -l lower bound of energy for drawing the map -u upper bound of energy for drawing the mapYou might be confused by the argument '-c' specifying the column number in the file. When you analyze 'System.Name.unfold_orbup(dn)', you will refer the sequence number for pseudo-atomic orbitals in 'System.Name.out'. Then, it should be noted that the number specified by '-c' is related to the sequence number for pseudo-atomic orbitals in 'System.Name.out' by .
(3) Drawing of the intensity map
Using gnuplot you can draw the intensity map.
For example, for the calculation with the input file 'SiC_C_SP_V.dat' it can be done
as follows:
gnuplot> set yrange [-10.000000:6.000000] gnuplot> set ylabel 'Energy (eV)' gnuplot> set xtics('K' 0.000000,'G' 0.722259,'M' 1.347753,'K' 1.708883) gnuplot> set xrange [0:1.708883] gnuplot> set arrow nohead from 0,0 to 1.708883,0 gnuplot> set arrow nohead from 0.722259,-10.000000 to 0.722259,6.000000 gnuplot> set arrow nohead from 1.347753,-10.000000 to 1.347753,6.000000 gnuplot> set pm3d map gnuplot> sp 'sic-intmap.txt'Then, you may obtain a figure as shown in Fig. 63(a).