[Pw_forum] ESM Example (Al slab with bc2) and Otani/Sugino example

[Louis Fry-Bouriaux]

Hi there,


    I am trying to understand the ESM example with bc2 (metal-slab-metal) distributed with QE 6.0 which consists of an Al(001) slab between two electrodes, and at the same time I am trying to replicate the results of Otani and Sugino in their paper (Phys. Rev. B, 73, 115407, 2006) which consists of an Al/Si(111)/Al slab between two screening boundaries.


I have a few questions:


** 1. ** In the QE example the following cell params and atomic positions are specified:


CELL_PARAMETERS bohr
 10.82227686   0.00000000   0.00000000
  0.00000000  10.82227686   0.00000000
  0.00000000   0.00000000  22.67672253
ATOMIC_POSITIONS bohr
Al  0.00000000   0.00000000   0.00000000
Al  5.41113843   0.00000000   0.00000000
Al  0.00000000   5.41113843   0.00000000
Al  5.41113843   5.41113843   0.00000000


My understanding is that the cell will repeat without overlapping atoms along X and Y. What I'm not clear on is why the 3rd lattice vector is so large?


The parameter 'esm_w' defaults to zero and is not specified in the example, so does this large value mean that there is a region of vacuum? Or does it just repeat the atoms such that in the Z-direction there are 5 Al layers? I'm guessing that this 3rd cell parameter Z-value corresponds to L_z, which in the paper corresponds to z_0?


** 2. ** I have reproduced the 'surface unit cell' from Otani/Sugino's paper using VESTA.


My concern here is with the correct repetition of atoms in X and Y, and what exactly happens with the Z direction. I realize I must first rotate the unit cell such that the repeated Al atoms lie perpendicular to the Z Cartesian axis which I have not done yet.


I have tested the repetition of this cell within VESTA along vectors perpendicular to the Al plane. So my most important question is: if the height of the cell along Z is ~14.4281 Angstrom (measured from Al center plane), then the third 'CELL_PARAMETERS' entry (after rotation) should be this value exactly + the 'decay length' of the wavefunction in Z? Then the parameter 'esm_w' is used to set z_1 from the paper I suppose?


I will generate an xyz file of the surface cell, but if I repeat by 3.816 A along a surface lattice vector, it is a matter of just deleting the entries that cause an overlap of the atoms I suppose?


Thank you for your time and sorry for the very long email,

Kindest regards,

Louis

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20170223/770d37e0/attachment.html>