Hi, I am attempting to perform a simple PCR reaction in a PDMS chamber permanently bonded onto a glass slide, but have been having problems with bubbling and evaporation at the 95C step. Most papers I have read indicate that this can be overcome with a number of approaches, including degassing the sample, high pressure (wax) seals on lateral inlets and outlets, and modifying the PDMS with a less permeable polymer (e.g. parylene). So far, these still have not produced good results. I am just wondering if anyone has any experience-based insight into these problems that go beyond the single sentence descriptions that they garner in the journal papers. In particular, 1) Are there other techniques to try to eliminate bubbling and better seal the chambers (so the sample doesn't just escape or disappear) 2) Am I applying the current techniques properly - Degassing: 1 min at -10 " Hg, degassing the sample only Wax Sealing: High-melting point wax dripped onto lateral outlets Parylene: bonded device inserted into parylene deposition system, 1 - 4 um coating (hoping that it penetrates the holes of the device and fully coats the inside walls) 3) Is it just a matter of continuing to try and waiting for my luck to turn? In other words, have many authors left out of their papers the fact that on-chip PCR was low-yield in terms of success rate? Thanks in advance for whatever help you can give me. There are more details about fabrication and PCR processing below for your reference. -JP Hilton PDMS: Mix 1:10 sylgard curing agent to silicone mixture Degas (~30 mins) Pour onto PDMS mold on Si wafer Bake on a hotplate at 75C for 20-30mins PDMS/Glass Bonding: Clean glass in Piranha/Nanostrip Wash in DI H2O Dry at 200C for 10 mins Concurrently: place PDMS in UV/Ozone for 10 mins, glass in RIE (O2 plasma, 200 mTorr, 100 Watts for 2 mins). Bond when they both finish simultaneously Bake at ~75C for 35 mins This gives a solid PDMS/glass bond that won't break PCR: Currently I'm using a very rough setup for thermal cycling, i.e. a peltier thermoelectric chip attached to a power supply and a thermocouple. The peltier chip is placed on a counter, the surface thermocouple is placed on top but off to one side, and the glass chip with PDMS chamber is placed on top of this, with the chamber to be heated in a position symmetric to the thermocouple but also offset from the center in the opposite direction (I'm trying to get a decent measurement of temperature by balancing out the temperature gradient present on the chip and not placing the thermocouple between the thermoelectric chip and the PCR chip).