The overall goal of this procedure is to demonstrate a microfluidic platform that combines resistive pulse sensing with code division multiple access to multiplex the detection and sizing of particles in multiple microfluidic channels. This technology, named microfluidic CODES can help realize fully integrated and truly portable labware chip devices that are well suited for the point of care testing of biological samples in resources limited settings. The main advantage of this technique is that it can electronically track spatial, temporal manipulation of particles on the microfluidic chip, eliminating the need for an external instrument such as a microscope.
Our technology is compatible with soft lithography and can easily be integrated into a microphotic device where particles are fractionated to provide a direct electronic readout similar to Coulter counter. To begin construction of the microfluidic device generate a set of four, seven bit gold codes. Then design four unique electrode layouts based on the gold codes using a computer aided design, or CAD software such as AutoCAD.
Finally, have the Photomasked with the designed electrode layout manufactured by a Photomask supplier. Next, soak a four inch borosilicate glass wafer in a five to one piranha solution at 120 degrees Celsius for 20 minutes. After cleaning, heat the wafer on a hot plate at 200 degrees Celsius for 20 minutes to evaporate residual water.
Place the clean, dry wafer in a spin coater. Apply 2 milliliters of negative photoresist solution to the wafer and spin coat at 3000 revolutions per minute for 40 seconds. Dry the spin coated wafer on a hot plate at 150 degrees Celsius for one minute.
The cover the wafer with a chrome mask in the desired electrode pattern. Expose the masked photoresist surface to 365 nanometer UV light to attain 225 millijoule per centimeter squared. The bake the exposed photoresist on a hot plate at 100 degrees Celsius for one minute.
Immerse the wafer in photoresist developer for 15 seconds, then wash the pattern wafer in a gentle spray of deionized water and dry the wafer under a stream of nitrogen gas. Next, place the patterned wafer in an electron beam metal evaporator. Deposit a 20 nanometer thick chrome layer and a 80 nanometer thick gold layer onto the wafer at a rate of one Angstrom per second.
Then etch the underlying photoresist by ultrasonic heating the metal coated wafer in acetone for 30 minutes at 40 kilohertz and 100%amplitude. Use a dicing saw to cut the wafer into smaller pieces as needed. To begin fabricating the microfluidic channel mold, clean and dry a four inch silicon wafer in the same way as the borosilicate wafer described previously.
Place the silicon wafer in a spin coater and apply four milliliters of negative photoresist solution. Spin coat the wafer at 500 rpm for 15 seconds, then at 1, 000 rpm for 15 seconds, and finally at 3, 000 rpm for 60 seconds. Set the wafer face up on an acetone soaked clean room wipe to remove residual photoresist from the back and the edges of the wafer.
Bake the wafer at 65 degrees Celsius for one minute and then at 95 degrees Celsius for two minutes. Place a chrome mask pattern for the microfluidic channels on the dry wafer. Expose the photoresist to 365 nanometer UV light at 180 millijoules per centimeter squared and then bake the wafer again at 65 and 95 degrees Celsius for one and two minutes respectively.
Placed the patterned wafer in a container of photoresist developer and gently shake the container for three minutes. Rinse the developed wafer in isopropanol and dry the wafer under a stream of nitrogen gas. Bake the wafer at 200 degrees Celsius for 30 minutes, then use a profilometer to check that the pattern photoresist is uniformly thick across the wafer.
Place the wafer in a vacuum desiccator along with 200 microliters of trichlorosilane in an uncovered Petri Dish. Allow the wafer to sit in the desiccator with the trichlorosilane for eight hours to silanize the wafer surface. To begin assembling the device, use general purpose clean room tape to affix the silicon wafer mold in a 150 millimeter diameter Petri Dish.
Add 50 grams of a 10 to one mixture of polydimethylsiloxane prepolymer to the Petri Dish and degas the mixture in a vacuum desiccator for one hour. Cure the degassed mixture at 65 degrees Celsius for at least four hours. Use a scalpel to cut out the cured PDMS layer and then peel the cured layer off the mold with tweezers.
Cut the PDMS into small pieces. Punch the inlet and outlet microfluidic channel holes with a biopsy puncher. Place the PDMS layer pattern face down on clear room tape to clean the micromachined surface.
Rinse the previously prepared electrode bearing glass substrate with acetone, isopropanol, and deionized water. Dry the substrate under a stream of nitrogen gas. Place the PDMS layer and the substrate in an RF plasma generator set to 100 milliwatts with the micromachine sides facing up.
Activate the micromachine surfaces in oxygen plasma for 30 seconds. Next use an optical microscope to align the pattern PDMS layer with the surface electrodes. Once aligned, allow the surfaces to come into physical contact to seal the PDMS layer onto the glass substrate.
It is crucial that the coating electrode pattern on glass substrate is properly aligned with the PDMS microfluidic channels. Once aligned properly, the particle interaction with the surface electrode will generate a desired code waveform for multiplexing. Bake the assembled device at 70 degrees Celsius for five minutes, glass side down.
Finally solder wires to the electrode contact pads to finish device assembly. To begin the experiment, place the microfluidic device on an optical microscope stage. Connect the device reference electrode to the signal output port of a lock-in amplifier and apply a 400 kilohertz sine wave.
Connect the positive and negative sensor electrodes to two independent transimpedance amplifiers. Connect both transimpedance amplifiers to the differential voltage inputs of the lock-in amplifier with the positive sensor signal to be subtracted from the negative sensor signal. Connect the demodulator output of the lock-in amplifier to a data acquisition unit.
In the data acquisition software, set a sampling rate for the lock-in amplifier output of 1 Megahertz. Set up a high speed camera to optically record the operation of the device as seen under the microscope. Draw a prepared cell suspension into a syringe.
Secure the sample syringe in a syringe pump and connect the syringe to the inlet channel. Direct the outlet channel to a waste container. Use the syringe pump to drive the cell suspension through the device at a constant flow rate while recording the impedance modulation signal.
After experiment's completion, process the electrical data with analysis software. Compare the processed electrical signal with images from the high speed camera to create a calibration curve for cell size. A cell suspension was flowed through a microfluidic sensor device with four unique electrode patterns derived from orthogonal sensor codes.
All four sensor signals were recorded from a single electrical output. The individual sensor associated with each recorded signal was identified by correlation of the recorded sensor signals with all possible codes, which produced clearly distinguishable, autocorrelation peaks. Waveforms produced by the interfering signals from simultaneous detection of cells in all four channels were resolved with an iterative algorithm.
A recorded waveform was correlated with all possible codes and the largest autocorrelation peak was identified. The corresponding individual sensor signal was reconstructed and subtracted from the input waveform. The residual signal was passed to the next iteration as the input and the process continued until the residual signal produced no autocorrelation peaks.
Estimated signals were refined based on an optimization algorithm seeking the best fit between the reconstructed and the original recorded waveforms using the least squares approximation. Cell location, size, and time to cross the sensor were then determined from the channel number, amplitude, duration, and relative timing of the estimated sensor signals. The procedure was validated by comparison of the electrical signals with optical measurements from the high speed camera.
Once mastered, this technique is very easy to implement, because it is very simple from a hardware perspective. It has no active on-chip component. It is directly compatible with soft lithography and the signal processing relies on a simple computational algorithm.
Following this protocol you can fabricate microfluidic chips with code based multiplex electrical sensors and decode electrical signals for bioanalytical measurement. This versatile, scalable, electronic sensing technology, can be readily integrated into various microfluidic devices to realize quantitative assays by spatial temporally tracking particles as they processed on the chip. After watching this vide, you should have a good understanding of how to design, fabricate, and implement a microfluidic CODES technology.
