Work flow of droplet merging
The droplet merging channel consists of three parts: the constriction channel for droplet synchronization, the spacing region for droplet separation, and the expansion chamber for droplet coalescence (Fig. 1, Fig. S1 and Movie. S1). Two sets of droplets with diameters of 35 µm and 65 µm are reinjected into two separate channels at the flow rates of 25 µL/h and 90 µL/h, respectively. When one large droplet flows through and blocks the T-shaped junction, a significant increase in the local hydrodynamic resistance is generated. This exerts pressure on the front curved surface of the small droplet, thus balancing the pressure exerted on the back curved surface by the continuous flow. As a result, the small droplet is temporarily immobilized for 0 ~ 5 ms. After the large droplet passes through, the lower downstream pressure is restored, causing the small droplet to immediately follow the large droplet and enter the channel (6 ~ 7 ms). Likewise, the small droplet transiently blocks the junction and stops the flow of the large droplet (8 ~ 9 ms). Then right after the small droplet passes through, the next large droplet enters, starting the next cycle of droplet arrangement (Fig. 2a). In this fashion, the two sets of droplets can enter the channel alternately at a frequency of around 111 Hz.
Upon synchronization, the pairs of droplets flow through another T-junction, where an oil flow is introduced from a vertical branch channel to increase the gap between adjacent droplets (Fig. 2b). Although the paired large and small droplets are spaced by the oil, due to their difference in droplet size and velocity, the small droplet can gradually catch up with the large droplet downstream. As the pairs of droplets are guided into the expansion chamber, the velocities of both droplets significantly decrease, causing them to contact with one another. When an electric field is applied, the water-oil interface stabilized by the surfactant is perturbed, triggering their coalescence (Fig. 2c). To ensure stable coalescence, the applied voltage should be within the range of 600 Vp-p to 1200 Vp-p: Insufficient voltages could not trigger droplet coalescence, while excessively high voltages might cause droplet wetting, leading to cross-contamination (Fig. S2). The merged droplets are uniform in size and color distribution, demonstrating the high efficiency of droplet synchronization, spacing, and merging (Fig. 2d).
Characterization Of Synchronization Effect
In conventional designs without any enhancement of channel blockage, the synchronization efficiency depends on the matching degree of the large- and small-droplet reinjection frequencies, f1 and f2. During a certain period, T, the number of large droplets that have entered the T-junction is given by: n1 = f1T, while the number of small droplets is given by: n2 = f2T. The ratio of one-to-one droplet pairing can thus be expressed as:
$$\text{P}\text{ = }\text{1 }-\text{ }\frac{\text{|}{\text{n}}_{\text{1}} - {\text{n}}_{\text{2}}\text{|}}{{\text{n}}_{\text{1}}}\text{ = 1 }-\text{ }\frac{\text{|}{\text{f}}_{\text{1}} - {\text{f}}_{\text{2}}\text{|}}{{\text{f}}_{\text{1}}}$$
1
In this manner, even a minor mismatch of the reinjection frequencies can result in synchronization errors and lower the one-to-one pairing ratio. In comparison, our design allows automatic tuning of mismatched frequencies: When the small droplet arrives more frequently than the large droplet, the large droplet temporarily stops the flow of small droplet and decreases its arrival frequency; and vice versa. Self-synchronization of reinjected droplets and higher efficiency of merging can thus be achieved.
To quantify the synchronization efficiency of our device, we measure the ratio of one-to-one droplet pairing at various reinjection flow rates. The synchronization ratio remains high (over 90%) across a wide range of flow rates as outlined (Fig. 3a). Moreover, for the five tested flow rates of small droplet, two to five tested large-droplet flow rates can achieve a 100% of synchronization efficiency. To further characterize the synchronization efficiency at different sets of reinjection frequencies, we measure the droplet reinjection frequencies at varied flow rates, which shows a linear relation for both the large and small droplets (Fig. S3). The flow rates in Fig. 3a can thus be transferred into droplet frequencies to indicate the tolerance in mismatching of frequencies for high-efficiency synchronization (Fig. 3b). The results show that for our device, 100% of droplets can be synchronized when f1 and f2 are within 10% of bias (|f1-f2|/f1 < 10%); whereas, for the negative control without any enhancement of blockage, the f1 and f2 must be strictly equal according to Eq. (1). Moreover, for the synchronization efficiencies of 95% and 90%, our device also allows wider ranges of mismatching (15% and 20%) than the negative control (5% and 10%).
It should be noted that, to enable self-synchronization, the droplet must be able to block the junction and stop the flow of its counterpart. Thus, the droplet size should be matched with the channel geometry within specific ranges to work properly. Specifically, for the large droplet, it enters the junction as a quasi-elliptical cylinder, which is composed of an elliptical cylinder and two identical semi-ellipsoids as shown in Fig. 3c. To block the small droplet, the length of the elliptical cylinder (top view), Le, should be larger than the width of the small droplet channel, w2. Thus, the minimum volume of the large droplet, can be expressed as:
V 1, min = Ve + 2Vs = πbcw2 + 4/3πabc (2)
Herein, Ve and Vs represent the volume of the elliptical cylinder and semi-ellipsoid, respectively; c = h/2, where h is channel height; a = b = w1/2, where w1 refers to the width of the constriction channel, since the two semi-ellipsoid can be combined and viewed as a prolate spheroid with the length of two minor axes equal to half of the channel width22. For our device, where w1 = 25 µm, w2 = 25 µm, and h = 40 µm, the minimum volume of the large droplet is calculated to be 33 pL, corresponding to a diameter of around 40 µm, which is consistent with our experimental results (Fig. S4a). For the small droplet, it enters the junction as an ellipsoid. At the critical condition when the small droplet can just block the junction, the top view of the small droplet emerges as a circle at the T-junction (Fig. 3d). Thus, the radius of the circle can be described as:
r = (lAD + lDE)/2 = (lAD + lBD2/lAD)/2 = w1/2 + w22/(8w1) (3)
For our device, the critical volume of the small droplet is calculated to be: V2, min = 2πr2h/3 = 21 pL. It corresponds to a diameter of around 34 µm, which is also consistent with our experimental results (Fig. S4b). Based upon the above results, we select 65 µm and 35 µm as the diameters of the large and small droplets for subsequent tests. Despite using these specific sizes of droplets for a demonstration, other sizes of droplets within the working ranges can also be synchronized with a high efficiency by regulating the flow rates. As shown in Table 1, for all the tested groups, a 100% of synchronization efficiency can be achieved across a wide range of large-droplet flow rate. In addition, for this blockage-based strategy, the viscous shear forces should not overcome the interfacial tension forces; otherwise, the droplet will be broken up and could not block the subsequent droplet. Thus, the flow rates should be within specific ranges. We test that for our adopted channel geometries and droplet sizes, the upper thresholds of flow rates for the small and large droplets are 80 µL/h and 400 µL/h, respectively (Fig. S5).
Table 1. Flow rate of large droplet that can keep a 100% of synchronization efficiency for different size combinations of the two sets of droplets. The flow rate of the small droplet is kept constant at 30 μL/h.
![](data:image/png;base64,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)
Droplet Synchronization At Fluctuated Droplet Sizes And Reinjection Flow Rates
As the droplet reinjection rate is proportional to the flow rate and inversely proportional to the droplet volume, polydispersity in droplet size and fluctuations in flow rates are the two most common issues that may change the droplet frequencies23. In conventional designs, these fluctuations tend to generate trains of unpaired or one-to-multiple paired droplets and lower the synchronization efficiency24; in comparison, our device, which can synchronize the frequencies of large and small droplet in case of mismatches, is well suited to overcome the fluctuations and prevent the synchronization errors. To demonstrate its superior synchronization effect, we test the efficiency of droplet pairing under fluctuated reinjection conditions, including mixing droplets with different sizes and varying the flow rates periodically, and compare the performance of our device with that of a negative-control device. The dimension of the junction is set at w1 = 75 µm and w2 = 40 µm for the negative-control device. According to Eqs. (2) and (3), the tested droplet diameters of 35 µm and 65 µm are below the thresholds (74 µm and 80 µm) to block the junction, thus presenting no synchronization effect (Fig. S6).
To demonstrate the ability of our design to accommodate size variation, we first generate four groups of small droplets with the coefficient of variation (CV) of diameter ranging from 3–9% (Fig. S7a), followed by measuring the percentage of one-to-one droplet pairing for each CV. For our device, the synchronization ratio maintains over 98% for all the tested CVs; whereas, for the negative device, the synchronization ratio decreases significantly with increasing CV (Fig. 4a and Movie. S2). We also measure the synchronization efficiency for four groups of large droplets with different CVs (Fig. S7b). Similarly, the synchronization efficiency of our device is above 98% and significantly higher than that of the negative device throughout the tests (Fig. 4b and Movie. S3). It is worth mentioning that the highest value of CV in our experiment is set at 9%, although higher values of CV can also be tuned and synchronized. This is in consideration that the droplets with CV values higher than 10% are regarded as polydisperse, which is beyond the scope of this work25.
In addition, we also assess the utility of our design in balancing the fluctuations in reinjection flow rates. The percentage of one-to-one droplet pairing is measured at periodically varied flow rates with the CV ranging from 2–8% (Movie. S4 and S5). For both the small and large droplets, the synchronization ratio of our device can remain high over 98% despite the fluctuations in flow rates. By contrast, the negative device fails to accommodate the fluctuations and shows significantly lower synchronization efficiencies (Fig. 4c and 4d).
Regulation On Droplet Pairing Ratio
Merging the droplets at one-to-one pairing ratio can meet the requirements of a number of applications. Nevertheless, for some complex assays, such as the encapsulation of controlled number of cells, higher ratios of droplet pairing and merging are frequently needed26. Here using our device, we can regulate the pairing ratio of large and small droplets from 1:1 to 1:3 on demand by simply adjusting the reinjection flow rates (Fig. 5a and Movie. S6). The flow rate of the large droplet is set constant at 100 µL/h, while the flow rate of the small droplet is adjusted from 30 µL/h to 60 µL/h. Initially, at the flow rate of 30 µL/h, all the droplets form a stable 1:1 pairing; as the flow rate increases, the portion of 1:2 droplet pairs gradually rises and reaches 100% at the flow rate of 45 µL/h; when the flow rate further increases, 1:3 droplet pairs are formed and a stable pairing is achieved at the flow rate of 60 µL/h (Fig. 5b). Hence, higher ratios of droplet pairing can be formed by further increasing the small-droplet flow rate.
In addition, in some scenarios where two types of samples are added, merging of three sets of droplets is needed27. To achieve this, we add one more injection channel in parallel to the former channel, and reinject an additional set of small droplets without fluorescent dyes. By adjusting the flow rates of the large droplet, the first small droplet, and the second small droplet at 100 µL/h, 30 µL/h, and 30 µL/h, respectively, the three sets of droplets can be synchronized to form 1:1:1 pairs. First, the large droplet and the first small droplet are alternately arranged at the first junction; then, the second small droplet gets inserted between the first small droplet and the next large droplet at the second junction; subsequently, the trains of the three droplets are spaced apart and merged in the downstream channel (Fig. 5c and Movie. S7). In this manner, more parallel channels can be introduced to increase the number of droplets in each train.
Multi-step Reaction Through Droplet Merging
The high-efficiency droplet pairing and merging allow stable addition of reagents across multiple rounds of droplet processing28,29. To demonstrate its potential for high-throughput multi-step assays, we conduct a two-step enzymatic reaction within the droplets (Fig. 6a). In this reaction, the glucose should be first oxidized by the glucose oxidase so that excessive amount of H2O2 is produced; then 4-AAP and TOPS can be steadily oxidized by HRP to produce quinoneimine, a compound with violet color. Thus, we first co-encapsulate the glucose and glucose oxidase into the large droplets. After 30 min of incubation, sufficient amount of H2O2 is produced upon oxidation of the glucose by the glucose oxidase. Then the large droplets are merged with the small droplets containing 4-AAP, TOPS, and HRP to produce quinoneimine dyes in a stable manner (Movie. S8). Upon completion of the cascade reaction, droplets with uniform size and color distribution are obtained, indicating the steady and homogeneous synthesis of quinoneimine within the droplets (Fig. 6b). Moreover, the amount of quinoneimine generated can be readily tuned by adjusting the glucose concentration. The color intensity of the resultant droplets, which is proportional to the quantity of quinoneimine, shows a linear relation with the glucose concentration, further confirming the stability and controllability of the droplet-based cascade reaction. The above results demonstrate that our system offers an ideal platform for precisely controlled multi-step reactions (Fig. 6c).
Cell-bead Pairing Through Droplet Sorting And Merging
Our droplet merging system allows high-efficiency pairing of different types of discrete targets after rounds of droplet manipulations. For instance, single cells and single barcoding beads can be paired for applications, such as high-performance single-cell sequencing30. In conventional assays, although the hydrogel beads can be closely packed and encapsulated in a controlled manner, the cell encapsulation is randomly distributed according to Poisson statistics31. As a result, only less than 10% of the droplets can encapsulate the targets of both single cells and single beads, causing a significant loss of samples and reagents. To increase the encapsulation rate, we generate the droplets with single cells and single beads separately, followed by subsequent merging in pairs (Fig. 7a). The droplets with single cells are first generated and sorted in an integrated device. The cell concentration is regulated to ensure that most droplets are empty and only 10% contain single cells. Through sorting, over 95% of the droplets can encapsulate single cells as shown in Fig. 7b. Subsequently, droplets with single acrylamide beads are generated. The hydrogel beads are closely packed and reinjected into the channel, thus beating the Poisson statistics to achieve a single-bead encapsulation rate of over 95% (Fig. 7c). Afterwards, the two sets of droplets are reinjected into our merging device for cell-bead pairing (Movie. S9). Ultimately, around 90% of the droplets encapsulate single cells and single beads, which is 9 times higher than conventional assays (Fig. 7d and 7e). The highly controlled pairing can significantly increase the droplet barcoding efficiency, thus reducing the cost of reagents and barcoding beads.