A hand-held, power-free microfluidic device for monodisperse droplet generation

Graphical abstract


Background
Droplet-based technologies catch a lot of interests during the past decades for its capability of screening very large numbers of samples simultaneously (usually tens of thousands). By dividing the sample and reaction reagents into separate small volume reactors, each sample is expressed individually. Such kind of platform can be used to quantify and qualify cellular genomic information at individual cell level (such as digital PCR, single cells genome profiling).
In a droplet-based system, two or more reagents are loaded in the reservoirs at the inlet of this device. The reservoirs are connected via microfluidic channels which intercept at least once at junction(s). An exit channel connects the junction and the exit port and allows fluid to exit the device. As the fluids traveling though the channels and intercept at the junction, continuous phase reagent pinches the discret phase reagent either by focusing flow (a cross-junction setting) or concurrent flow (a T-junction setting). The balance between the inertia of discret phase solution, surface tension between discret and continuous phase and viscosity enable continuous phase reagent to encapsulate the discret phase reagent and form droplets of discret phase solution. The dimension of the droplets are determined by the flow rates of all reagents (Q), surface tension (s) between these two immiscible phases, viscosity (m) of the continuous phase, density (r) of the continuous phase, the channel height (h), channel hydraulic diameter (d) at the junction and the channel width (w) at the exit of the junction are the junction. The correlation is expressed as below in Eqs. (1a) and (1b) [1] In this research, we adopted the parameters of generating nanoliter size droplets by keeping the flow rate of 3 liquids at 5, 5 and 10 mL/h (water/water/oil) [2]. However, instead of using three individual syringe pumps to drive the liquid, we develop a microfluidic device, of which all three incoming reagents share the same pressure drop (DP) through the fluidic passage. The channel length, width and height and liquid viscosity determine the resistance (R) of each channel, as showed in the equations below in Eq. (2). The channel height of this device is 120 mm. The system therefore generate flows at the flow rate ratio at 1:1:2. The design of the chip is showed in Fig. 1a. The AutoCAD file is available upon request.
A single source negative pressure is applied to the fluids to create the flow by a vacuum at the exit end of the device. The vacuum is generated on-site by pulling the plunger of a syringe. The positions of the plunger before and after pulling determine the degree of vacuum following ideal gas law P initial V initial = P final V final . A fixture is used to hold the plunger at the "pulled position" to maintain its vacuum. This microfluidic chip is designed to generate 20 mL/h flow rate when the pressure drop is 1.73 Â 10 4 Pa. As the liquids fill in and droplets are generated, the air in the syringe is compressed and vacuum is lost gradually. An analysis based on Eqs. (1a) and (1b) predicts when the change of pressure drop (DP) is smaller than 10%, the change of flow rate is therefore below 10% and the change of droplet size is not greater than 2%. The pressure drop across the chip, the flow rates of all reagents, the droplet size and generation frequency are therefore predictable, programmable, and reproducible. This low cost device is composed with a microfluidic chip, a tube, a collection vial, a syringe and a station; and is in the size of an iPad Mini. On the contrary to commercial instruments that utilize multiple pumps yet still facing challenges of fluctuating flow rate and flow rate ratio, this invention is 100% portable and power-free. Our device can be used to generate various dual phase droplets systems such as of water-in-oil, oil-in-water, water-in-water (emulsion), oil-in-oil system; or more complex, multi-layer systems, such as water-in-oil-in-water, etc. Its application includes, but not limited to, generating droplets for high throughput, parallel droplet-based mRNA or DNA profiling, digital PCR, and emulsions for pharmaceutical applications.  Pour forty grams of the mixture on the SU-8 microstructure mold and pour the remaining on another 4" Petri dish. Leave the Petri dishes in a desiccator and vacuum for 5 min until all air bubbles disappear in PDMS mixture. 4 Bring the Petri dishes to an oven and let solidify at 70 C for 30 min. 5 Score the cured PDMS along the circumference to separate from the Petri dish. Peal the PDMS off from the mold. With the featured side facing up, use hole punch to make connection ports. (inlet ports with a 3 mm hole punch and outlet port with a 1.5 mm punch). Make sure the featured PDMS is at least 3-mm thick. Double check the holes are totally clear without any PDMS stuck in. Note: Punching PDMS too rapidly will result in cracks and future leakage. It is recommended to practice on dummy PDMS until holes come out nicely. 6 PDMS specimens are washed in diluted dish detergent (1:100) using a sonicator for 10 min, followed by replacing the soapy water with DI water and sonicated twice. Make sure the specimens are totally immersed and never overlap during washing. When handling PDMS, wear powder-free gloves. 7 Blow dry the PDMS specimens using purified, compressed air. The station is made with PLA by 3D printing. The CAD file is available upon request. 2 Cut off bulb part of 3 plastic droppers with remaining height of 5 mm as reservoirs, cut off the straw part of the droppers with remaining length of 2 mm as inserts for PDMS chip as showed in Fig. 1. Make sure the inserts are cut with an angle close to but not exactly at 90 degrees angle to prevent flow blockage. Note: 200 mL pipette tips can be used in the replacement of plastic dropper. It is, however, not recommended since pipette tips tend to trap air bubbles and interfere with flow. 3 Insert all reservoirs to inlet ports and tubing B to outlet port. Note: Users can apply a thin layer of silicone glue at the reservoir insert and the tip of tubing to assure all junctions are air-tight. To verify if the system is air-tight, load 100 mL water in all three reservoirs and pull plunger to 10 mL tick quickly and then release. Water should not leak out from the junction and the plunger should re-track to its initial position within a few seconds. 4 Pull syringe plunger to 7.0 mL tick. 5 Load 100 mL aqueous solution A at reservoir 1. 6 Load 100 mL aqueous solution B at reservoir 2. 7 Load 150 mL droplet generating oil at reservoir 3. Note: To minimize droplet size variation, please follow the sequences step 6 through 8. 8 Pull plunger to 9.0 mL tick quickly and rest the syringe on the station with plunger held in position by the fixture. 9 Unscrew the sample vial cap 40 s after flow starts. Breaking the vacuum will stop the flow.
Extending the flow time will result in inconsistent droplet sizes. Note: To minimize droplet size variation, please make sure to follow this step. 10 Collect the droplets from the top layer of the collection vial (droplet generation oil is heavier than droplets). About 150,000 droplets are generated within this timeframe.

Method validation
The system was validated by two settings. First setting is designed to encapsulate aqueous solution mixed with green food grade dye for better visualization; second setting is designed to encapsulate magnetic beads (30 mm, 350 bead/mL) for single cell sequencing feasibility test. (Beads and cells have higher density than water and tend to sediment in the reservoirs and result in flow blockage. We have developed a method to mitigate this issue and will discuss in future publication) Fig. 2 shows that the droplets being formed in the chip and flow throw the tubing into the collection vial. Fig. 3 shows time-lapse images of droplet formation from the beginning of the flow untill 28 s after flow started, with 100,000 droplets generated. Fig. 4 shows solution in the collection vial contains droplets with high size consistency. 20-mL aliquots from multiple locations of the collection vial were acquired, spread on a glass slide and imaged right away. The droplet size was analyzed using ImageJ (an open source software). The average size of the droplets is 109 mm, with standard deviation of 9 mm (CV < 8%). Step-by-step workflow. a) Set up the system. Pull plunger to 7.0 mL tick. b) Load aqueous and oil solutions to the reservoirs. c) Pull the plunger to 9.0 mL tick quickly and rest the syringe on the station. d) Droplets are formed in the chip and flow to collection vial.

Appendix A. Supplementary data
Supplementary material related to this article can be found, in the online version, at doi:https://doi. org/10.1016/j.mex.2018.08.008.