The Synthetic Chemistry Labs today remain highly non-automated, and result in low volume, costly Science. We realized that the Semiconductor Industry has perfected this problem for decades by creating mass manufacturing lines that are fully automated. These lines product electronic devices on semiconductor wafers, with millions of devices made on one 20cm wafer. As a result, the process is extremely stable, the environment around the wafer is precisely controlled, and the cost per device is correspondingly very low. Such manufacturing technology is pushing the boundaries of what is possible in the electronics industry. Why not use a similar setup in the Chemistry industry?
What it does
Our concept uses the existing equipment environment created by the semiconductor industry. Instead of semiconductor wafers, we use glass wafers. Each glass wafer is etched with microfluidic channels to build up radial miniature production lines from the centre of the wafer to the perimeter. As each production line is 10um wide, millions of theses mini production lines can fit on one wafer radially. The wafer is then spun to create a centrifugal force that drives fluids from the centre of the wafer to the outside edge. The wafers can be cut in 3D to give the option of many fluids working together along the production lines. Each mini production line can be designed to have mixing stations, separator stations, amongst many others. The glass helps to allow agressive reactions. The transparency allows us to do insitu measurements inside the equipment environment, e.g. IR/UV spectroscopy. All while in a controlled environment, and fully automated.
How we built it
Because glass was not available, we chose to test the approach using acrylic plastic. This material is hard and allows us to cut/etch microchannels in the wafer material. Complex microfluidic channels were designed and then cut/etched into the wafer using a laser machine. A self-made stepper motor system was built to control the spin rate of the wafers (many designs were created so that we can run mixing, separation stations). A sensor system was developed to measure different concentrations of drug reactions across the wafer as the device spins. A fluidic reaction system was developed to allow us to visually test the system with colored liquids that would react together, in a way that is also separable. All techniques that were used were designed to be transferrable to a glass-based system within a semiconductor environment.
Challenges we ran into
The challenges we ran in to can be separated in 4 phases: 1) Design of the spinning mechanism: Here we dismantled a hard drive with the aim of quickly having a stepper motor system that was usable. Unfortunately, this was not allowed (as the product was needed during the day). We tried to dismantle another system to use, but again had issues that we were not allowed to dismantle existing equipment. Hence, we moved to a 2nd option of building the system ourselves, which cost a lot of time. We eventually got the system up and running successsfully, and with much more control that the other systems. 2) Design of the microfluidic channels: Here we had many designs to show different approaches that could be used in the design of the wafers. For example, we had 2D designs to show mixing and separation. We have 3D designs to show that a particular radial directions can be used by multiple fluids stacked on top of one another. All this was complex and required careful alignment with the equipment experts to see what was feasible. 3) Construction of the wafers: Due to a lack of plastic, we had to be very careful about the use of plastic / recycling to ensure we had options to try prototype iterations but did not run out of plastic for the main demo! We came up with solutions to control leakage between the plates, using careful designs and rubber sealing layers (this problem would not exist in the 'real' world as you can simply bond thin layers together (standard process in semicon industry) 4) Design of the fluids to use in the prototypes Here we invested many hours trying to find basic fluids that could both react together to form a different color (to allow us to electronically sense that the reaction had taken place) AND would allow us to separate the chemicals to show that other stations on a production line were possible.
Accomplishments that we're proud of
The team contributed equally to the whole approach, which I was very proud of. We worked together on a very complex systems with many challenges (see above), and were able to create working spinning wafers in a short time
What we learned
We learnt the limitations created by having access to limited materials / tools. We learnt also that microfluidics is HARD! It takes time to optimize the design. We loved the community nature of hacking, where people were so open to share ideas with one another.
What's next for WAFER LABS
That is up to AstraZeneca ;). Ideally, the idea is discussed further with equipment manufacturers like Veeco, and tested/prototyped more seriously in a lab setting