For the remainder of the PRF-funded period, we divided our work into two aspects; the microfluidic membrane array device and creating a lipid membrane that may be conjugated to an in situ polymerized hydrogel.

            Work on the microfluidic array device began with a brief redesign. As shown in Figure 1, the microfluidic device was originally planned to consist of five layers, with Glass, Si, PDMS, Si, Glass. Concerned about the effects of capacitance of the Si as well as some processing issues with the Si, we experimented successfully with glass-glass thermal bonding. We were able to obtain good bonds between top two and bottom two layers of the structure. Also problematic was the creation of the center layer. We molded the center layer of PDMS over patterned photoresist on a Si wafer, as is normally done. However, typically the PDMS is only bonded on the wafer side surface. For our device design, we wish to bond to both sides of the PDMS. We find that the exterior surface is significantly un-flat, as a result of the meniscus formed where the PDMS wet the photoresist post. This flatness caused the PDMS-glass bonding process to fail frequently, since the PDMS did not make good contact to the glass in the absence of applied pressure, or made good contact but strained in the presence of an applied bonding pressure. At this point we are attempting to solve these problems and test the device.

            The work with the lipid derivative was much more successful. We created a cross-linkable lipid from N-acryloxysuccinimide and diphytanoylphosphatidylethanolamine (Figure 2). We created free-standing membranes from this lipid mixed with conventional lipid surrounded by the hydrogel precursor we have used previously and polymerized the gel and lipid. Using applied pressure, we were able to confirm that the lipid did bond to the gel successfully (Figure 3), mechanically strengthening it as we had hoped and extending its lifetime to over 2 weeks, a substantial improvement over conventional lifetimes (< 1 day).

Figure 3: Capacitance (a proxy for membrane area) as a function of applied pressure for membranes bonded and unbonded to the gel on one side. The negative pressure data show the difference between the conjugated and unconjugated membranes.