Of course it will be ideal to have the source and drain, the connects to the environment, as thin as the channel but there are two difficulties. The first is practical, i.e, if you try to deposit Au thinner than 10 nm (forget about Ni which becomes highly reactive in the nanoscale) you would in all probability land up with Au nanoislands or droplets, not a continuous strip. The second point is more fundamental. Around 2 nm you have the Mott transition where metals have a bandgap opening up, so they are no longer metals at all.

Thanks @Alokmay Datta, for answering and your great knowledge.

What can be minimum thickness of Au, to have continuous film.

what about other metals like Ti, Cr, and Ag, Can i deposit thin films of these materials. for deposition i am using E-beam Evaporator, 

Don't worry about that equation. All it says is that what is the expected short channel effect (DIBL, subthreshod swing, Vt roll-off) at a given channel thickness. At the end of the day, having a thick enough S/D - both semiconductor part to spread the resistance and the metal part to make a reliable contact - matters far more than a possibly few nanometer of better scaling of the gate length. Use the dimensions that the state-of-the-art "experimental" work use as a starting point.