I recommend you yo use easy release materials as intermediate layer between PDMS and mold:

http://www.smooth-on.com/Release-Agents-and/c9_1123_1226/index.html

For cleaning, you can wash it by solvents like acetone and drying under air flow.

I like to silanize my masters with a tridecafluoro-silane. I use vapor deposition either in a desiccator or in a custom deposition chamber that I had some summer students build.

Thank you for your answers. I will try to silanize the silicon master and let you know how it worked.

Check our publications on using PVA as release material (Karlsson et al.).

(Alternatively, PDMS cannot bond to Au.)

An easy option would be soak your silicon wafer in 1~5% Pluronic F127 (available at sigma and cheap) for 30min to 1hr.  After soaking either wash wafer w/ water or directly blow dry it.  This will coat a thin layer (visible) of surfactant on your mold and that will help with PDMS releasing.  You can simply rinse your mold with water thoroughly to remove the Pluronic coating afterward

I use Trimethylchlorosilane (TMCS) to salinize the surface of Silicon model in sealed container for 15-30 min. You can hard bake the Silicone mold for about 10 min at 105 oC and cool it down before applying the TMCS to enhance the formation of molecular layer of TMCS. Be sure to use the chemicals in a chemical hood. 

You can also make the PDMS swell a little using a solvent, which will hep releasing it from the master mold. 

You can also use Triton X-100 if you have it available, but it is bio-hazard. 

Thank you very much for your answers

a few 10s of nanometer thick deposition of Parylene can provide a stable and durable release layer. It works like a charm.

Hi Sasank,

You can coat your wafers with a thin layer of sigmacote to prevent adhesion of PDMS.

In terms of PDMS residue left behind, make sure that the PDMS is totally hard before peeling. If you want to have a 'soft' PDMS, bake the system at 65C for at least 3 h. The totally hard PDMS should not leave any residue on top of the wafer.

can i de-bond the bond of PDMS PDMS created after plasma treatment by any chemical means?