PDOS Projection for Individual Orbital Components

Orion

How can I plot the PDOS for each individual orbital component separately?
Actually could not able to decode the last four line from the crystal23 manual :

NEWK
12 12
1 0
DOSS
3 100 3 6 1 12 0
4 1 2 3 7 s
4 4 5 8 9 px py
2 6 10 pz
END

In my system, if I need to plot px, py, and pz orbitals of Zn and Cl separately, then how to modify these last 3 lines?

CRYSTAL
0 0 0
230
19.52502047
6
30 9.713891360990E-17 -2.500000000000E-01 -3.750000000000E-01
17 9.427017681292E-02 -3.049312792878E-01 -2.362267967025E-01
7 -1.501875525291E-02 -3.310379606087E-01 -3.160580056873E-01
6 2.209194576296E-02 -3.478572492153E-01 -2.592622821244E-01
6 4.327400041674E-01 1.250000000000E-01 1.827400041674E-01
1 3.934409884276E-01 1.250000000000E-01 1.434409884276E-01

GiacomoAmbrogio

Hi Orion,
Looking at your input, I think you want to plot the orbital projected electronic DOS (in CRYSTAL this is done with the DOSS keyword, PDOS in CRYSTAL normally refers to phonon DOS).

To obtain separate px py pz projections, the correct approach is:
set the first parameter of DOSS to 2 (because you want two separate projections).
After the main DOSS line, you must then add one line per projection, specifying which atomic orbitals (AOs) belong to each projection.
Each line has the form:

n AOs    index_1    index_2    ...    index_n

The AO indices correspond to the ones printed in the CRYSTAL output. To find them, search in the SCF output for "LOCAL ATOMIC FUNCTIONS BASIS SET", you will see a table with all the basis set, that will look like this:

*******************************************************************************
   ATOM   X(AU)   Y(AU)   Z(AU)  N. TYPE  EXPONENT  S COEF   P COEF   D/F/G COEF
*******************************************************************************
  1 O   -2.793  -4.838  -4.110
                                 1 S
                                        8.589E+03 1.895E-03 0.000E+00 0.000E+00
                                        1.297E+03 1.439E-02 0.000E+00 0.000E+00
                                        2.993E+02 7.073E-02 0.000E+00 0.000E+00
                                        8.738E+01 2.400E-01 0.000E+00 0.000E+00
                                        2.568E+01 5.948E-01 0.000E+00 0.000E+00
                                        3.740E+00 2.808E-01 0.000E+00 0.000E+00
                          2-     5 SP
                                        4.212E+01 1.139E-01 3.651E-02 0.000E+00
                                        9.628E+00 9.208E-01 2.372E-01 0.000E+00
                                        2.853E+00-3.274E-03 8.197E-01 0.000E+00
                          6-     9 SP
                                       9.057E-01 1.000E+00 1.000E+00 0.000E+00
                         10-    13 SP
                                        2.556E-01 1.000E+00 1.000E+00 0.000E+00
                         14-    18 D
                                        1.292E+00 0.000E+00 0.000E+00 1.000E+00

The indeces are the numbers befor each orbital shell (1 2-5 6-9 10-13 14-18).
The notation 2-5 means indices 2 through 5, because that shell contains 4 AOs.
From these blocks you must identify the p-type orbitals for each atom (Zn and Cl). The p shells will appear in groups of three basis functions (px, py, pz).

So, in your system:

Locate the Zn atom in this printed basis list
Identify the block corresponding to its p orbitals
Extract the AO indices for px, py, pz
Do the same for Cl

Then you define one projection per line. For example, structurally something like:

NEWK
12 12
1 0
DOSS
2 100 3 6 1 12 0
3 <list of Zn p AOs>
3 <list of Cl p AOs>
END

Keep in mind that, if your basis is double- or triple-Z, each p shell appears as a separate triplet of AOs (one triplet per Z). To correctly project all p you must include the corresponding AOs from every p-shell triplet for that atom.

Let me know if you manage to do this, or if you need further help.