The overall goal of this video protocol is to demonstrate complete multi-step photolithography of microfluidic master molds with on-chip valves and multiple height features tunable for any application. This method is a complete overview for how to fabricate master molds with complex geometries, including on-chip membrane valves, for microfluidic devices. The main advantage of this technique is that it makes it possible to easily control flow in microfluidic devices, overcoming a major barrier to entry to microfluidics in biological applications.
Visual demonstration of this technique is critical, as the steps for photolithography are often difficult for beginners to master. Proper alignment, development, and exposure rely on visual cues and clean room experience. To begin, design your device and prepare the individual photo masks for the multilayer geometries.
Additionally, prepare about four wafers with a five-micron layer of SU-8 2050 negative photoresist, and flood expose as described in the accompanying text protocol. Place the coated wafer on a spin coater, and turn on the vacuum to affix it to the spin chuck. Use nitrogen or compressed air to blow away any dust from the surface.
Then, apply two to three milliliters of AZ 50XT positive photoresist to the center of the wafer. Spin coat the photoresist to create a 55-micron layer. Once coated, lay down the wafer carefully in a five-inch Petri dish, and let relax for 20 minutes.
Next, soft bake the wafer on a hotplate for 22 minutes while ramping the temperature from 65 degrees Celsius to 112 degrees Celsius at a rate of 450 degrees Celsius per hour. Then, remove the wafer and let it rest overnight at room temperature in a Petri dish for ambient rehydration. Tape the flow round transparency mask to the five-inch glass plate so that the print side is closest to the wafer, and load into the mask positioner of the UV mask aligner.
Expose the wager to 930 milliJoules of UV in six cycles. Develop the wafer immediately by immersing it in a stirred bath of developer for three to five minutes, or until the bath turns purple and the features emerge. Once developed, remove the wafer and rinse it well with deionized water.
Then, hard bake the wafer to melt and round valve features. Ramp the temperature from 65 to 190 degrees Celsius over the course of 15 hours at a rate of 10 degrees Celsius per hour. Once finished, turn off the hotplate and let the wafer cool to room temperature.
The features on the wafer are now rounded. This hard bake is critical to properly reflow rectangular valve features into rounded valve profiles. Shorter times may result in cracking or instability.
In order to fabricate a device with variable height features, place the cleaned wafer on a spin coater as previously shown. Apply one to two milliliters of SU-8 2050 negative photoresist to the center of the wafer, and spin the photoresist over the developed valve features. Then, carefully place the spun wafer in a five-inch Petri dish, and let it relax for 20 minutes on a flat surface, or until any streaking patterns fade.
Next, preheat two hotplates to 65 degrees Celsius and 95 degrees Celsius, and then set the wafer on the 65-degree-Celsius plate for two minutes, the 95-degree-Celsius plate for eight minutes, and the 65-degree-Celsius plate for two additional minutes to soft bake the wafer. Once the wafer cools back to room temperature, tape the flow low transparency mask to a quartz five-inch glass plate so that the print side is closest to the wafer, and load it into the mask positioner of the UV mask aligner. Then, place the wafer in a UV mask aligner chuck, and using the microscope eyepiece or camera, carefully align the new flow low layer features to the flow round valve layer features.
Begin by aligning the horizontal, vertical, and tilt axes of device borders to the device border features on the mask. Next, align crosshair features between the layers. Finally, confirm that the valve features intersect the flow low features where desired.
Next, expose the wafer to 170 milliJoule UV deposition. When finished, remove the wafer, and bake it post-exposure by switching between the two hotplates set at 65 degrees Celsius and 95 degrees Celsius. Without developing the wafer, allow it to cool to room temperature, and then sequentially add the flow high layer and then the chaotic mixer herringbone layer using SU-8 2025 as described in the accompanying text protocol.
After all layers have been completed, develop the features by immersing the wafer in a stirred bath containing 25 milliliters of SU-8 developer for 3.5 minutes, or until the features clearly emerge. Use a stereoscope to verify that the features have clear, defined feature boundaries. During development, be sure to check every 20 seconds to see that features have become fully defined and resist has washed away.
Overdevelopment may result in damaging the features, especially on complex mold designs. Then, hard bake the wafer to stabilize all of the photoresist features. Subsequently, fabricate the control layer as described in the accompanying text protocol.
Fabricate multilayer microfluidic devices in a pushup geometry on glass according to existing open-access protocols, and use visual inspection to ensure that all of the valves are properly aligned to control lines, and that all of the inlets are punched fully before proceeding. Connect Tygon tubing loaded with water to a flow control system, such as a syringe pump, fluidic controllers, or an open-source solenoid valve array with reservoirs. Next, connect metal pins to the tubing and the metal pins to the device ports at control line inlets.
Then, set the flow control system to 25 PSI for each line to pressurize the device control lines. Ensure that valves close and reopen by inspection under the microscope. In a microcenrifuge tube, suspend 3.9 milligrams of photoinitiator in 100 microliters of DI water to prepare the photoinitiator solution used to polymerize droplets into hydrogel beads.
Cover the solution to protect it from light. In a second microcentrifuge tube, add 132 microliters of deionized water, 172 microliters of PEG diacrylate, 12 microliters of the photoinitiator solution, and 85 microliters of HEPES buffer to make the hydrogel droplet solution. Transfer the hydrogel droplet solution to a customized cryogenic tube vessel.
Then, connect the tubing of the cryogenic tube vessel to a controllable pressure source, and connect the PEEK tubing to the device reagent inlet. Next, insert the PEEK tubing at the device outlet in order to collect the droplets. Remove air bubbles from the device, repressurize the system, and then depressurize the RO1 oil valve, and set the oil pressure to 10 PSI.
Next, set the PEG mixture pressure to nine PSI, depressurize the upstream valves, and adjust the pressure as necessary to produce droplets of the desired size. Determine the droplet size via microscopy using a camera with 50 FPS or higher. When the droplets have stabilized, position a UV light source over the polymerization region of the device, and apply 100 milliWatts per square centimeter of 365 nanometer light from the source onto a five millimeter spot.
Pressurize the bead sieve valve to watch polymerized beads collect and ensure that droplets have hardened into beads. Finally, depressurize the bead sieve valve and collect beads into a tube through the PEEK outlet tubing. This protocol starts by demonstrating a method for rounding flow valves.
Here, a profilometer was used to determine the typical post-reflow valve rounding profile resulting from this method, showing a height of approximately 55 microns. In the image on the left, the valve is off, and liquid can pass through the channels. Once activated by pressurizing the valves, the flow through these valves is cut off.
Here, one can see the bead synthesizer device in operation, producing hydrogel droplets in an oil emulsion at the T junction droplet generator. By partially closing a downstream flow using a sieve valve, the fluid can continue to flow, but the beads are trapped behind the valve. The resulting beads produced using this process averaged 52.6 microns in diameter with a standard deviation of only 1.6 microns.
Out of nearly 3, 000 beads, less than 1%were off by more than three standard deviations. Once mastered, this technique can be completed in three days from design to testing. This allows for rapid design iteration.
Following this procedure, even researchers with little fabrication experience can build their own complex microfluidic devices and apply them to their own biological problems. After watching this video, you should have a good understanding of how to perform the photolithography steps necessary to make microfluidic devices of any level of complexity, including devices with complex variable height features or valves.
