How to optimize the molecule in device.

[J.g_johnson]

Dear friends,
  I have made a two probe configuration:[Au electrode]-[molecule]-[Au electrode]. The structure of molecule(a cubic) had been optimised individually before used to construct the two probe device.
The distance between two electrode was based on experience parameter.

  The initial structure of the device is shown in figure1.
  The electrode is two layers, and the Unit Vector is:
           vector_a = [5.89465425896, 0.0, 0.0]*Angstrom
           vector_b = [0.0, 5.89465425896, 0.0]*Angstrom
           vector_c = [0.0, 0.0, 4.16814999926]*Angstrom
  Five layers of Au slab were chosen in z direction as left electrode and electrode extension just for a fast computational speed.  
  The perpendicular distance between the left electrode and the leftmost atom of the molecule as well as counterpart in the right is 2.0 Å as experience.
  More detail parameters are in the attached script.
  
  So I decided to optimized structure of the molecule in this device. We kept all gold atoms fixed and relax the molecule in the center until the forces on each atom are smaller than 0.05 eV/ Å in the
optimization.

  But the result I got was unstatisfied, because the change of molecule`s shape is too big, not cubic any more, as in figure2. So I think there must be someting wrong.
  1.The perpendicular distance between the left electrode and the leftmost atom of the molecule is 2.0 Å according to experience. But I am not sure whether it is the cause of the mistake.
  2. In this work, only five layers of Au slab are used because they are enough for such small molecule. Also I am not sure it is reasonable.
  
  The two above is just the reasons I speculated. And this problem is still making me confused.
  
 Thanks in advance!

[Umberto Martinez]

Sorry about the late reply!

1. You may try to optimize the cell along the Z direction (c vector) by unchecking the z box in the Constrain cell option in your OptimizeGeometry object.

2. Using a gold (2x2) unit cell, as you do, gives you a distance between the repeated molecules along XY of about 3.2Å which is comparable with the Pb-S distance of you molecule  (2.7Å).This means that you cannot really say that your molecule is isolated and you may instead create a periodic "2D like" configuration (which you actually get after optimization).
Try with a larger cell, (3x3).

3. I will also speculate. Might be that the Au-molecule interaction is so strong that your molecule will dissociate when placed on a Au surface. Not much to do with that.