Dear Shubham, to close your microchannel I would use a silicon wafer (hydrophobized) and spincoat (or pour) a layer of PDMS on it. Cure this layer partially (ca. 10 min @ 60C)  and then place your PDMS channel on top of it. This will ensure a good sealing between the two PDMS layers. Cure the whole piece and peel the PDMS layers off the silicon. 

Thank you for the answer Daniel, but partially cured PDMS film is not working as a flat top film over the channel instead one face (top) face of the channel seems to be convex inward. And also for a higher thickness channel as 250 micron the top layer is even touching the bottom face.

Dear Shubham, I suppose you are applying oxygen plasma to activate both PDMS surfaces in order to bond them. In that case, 30 min of oxygen plasma is probably way too much. In my case I normally apply less than a minute. But it also depends on the power and the pressure of the chamber. You have to find the right parameters for your machine, but normally the time is in the range of seconds or maybe just some minutes, and the power is low, less than 100 W or around that. You could find more examples in literature. Good luck.

Thank you Rojas

My recent experiments also shows that, exposing it to longer duration is making the surface harder. In fact parameters need to be optimized yet but very less duration is useful. i kept oxygen flow as 2 sccm and power as 50 watt. but it still need to be reduced. 

Thank you all for the  support.