Apologies if I wasn't entirely clear. In terms of conjugation, I was refering to the single A1, A2, and A3mono. The addition of the lateral substituent groups (and their size) would lead to slightly smaller band gaps, which is rightly observed. This might not be the case comparing the "packed" configurations with their single counterparts, as there will be energy level splitting (this also has been observed for 2 ,3, and 4 adjacent molecules). I was interested to include the packed configuraitons in my analysis because when I move towards the "on substrate" anaylsis including boundary conditions, I try to mimic a thin film of SAMs (eventhough the nature of the overlap isn't exactly known!!).
On the substrate I am a bit puzzled by why the A3 massively decreased comparing it to the A2 and A1 (either in a single or packed configuration).
Regarding the script, I attached the one for the single A3 on substrate for the geometry optimization. Once finished I projected the MES only on the molecule