I would suggest a different approach than just running a huge device optimization, which can be quite time-consuming (and also because I still have no information about your settings etc, I can't really advice in more detail).
From basic physics you can usually argue that the main effects (not all, but main, probably) controlling the electron flow through a molecule are related to the bonding between the molecular and the surface, and the properties of the molecule itself. So if you feel that that molecule is well described by the previous optimization, then just focus on how it's connected to the surface - for instance just manually vary the distance until you find the lowest energy (make 3-7 configurations with varying distance, compute the total energy in a slab configuration, and the look the minimum of the energy curve).
Of course there are other subtle effects, but at least this will give you some additional experience in ATK and get some initial results to work further with. Obtaining the optimal geometry of a complex system is a difficult problem which should be approached in a systematic fashion, - it rarely helps to apply a brute force "black box" method and just run a huge optimization calculation. However, no matter your choice of method it will take some time, and reducing accuracy dramatically to make it faster is not really a solution, it's cheating, since you can't really trust the results anyway. Sure, you may not always need DoubleZetaPolarized, but that needs to be checked also, separately, to ensure that whatever basis set you do choose actually describes your material reasonably.
Also, in the end: if you expect detailed assistance with your particular problem (such as suggestions on choice of parameters etc) you must share your input file or at least the relevant parts of it, otherwise we can - at best - just guess, and that's really of no help to you.
