Accurate control of liquid–liquid interfaces is of great importance in many scientific fields. Currently, most studies on liquid–liquid interfaces are based on microfluidics in closed channels, and for open systems, it is difficult to form stable liquid–liquid interfaces in microchannels due to the interference of gas-phase molecules. Here, we introduce a new method to manipulate the motion of the liquid–liquid interface under an open microchannel. Under the effect of surface tension, the liquid–liquid interface moves in microchannels until it encounters microstructures in the microchannels, where the force equilibrates and remains stable to form a fixed shape. The shape of the interface is regulated by adjusting the dimensions of the microchannels and microstructures as well as the positions of the microstructures in the microchannels. This spontaneous liquid–liquid interface moves, stops, and remains stable in a very convenient way. The morphology and position of the interface are well manipulated using microstructures, and the liquid–liquid interface with well-defined geometry can be made to stay in different positions to play different roles.

Interfaces—spatial boundaries between two different substances, especially liquid–liquid interfaces have attracted much attention due to their unique properties that are different from bulk phases. Immiscible liquid–liquid interfaces have a unique environment different from solutions and can be used as a versatile platform for chemical reactions, adsorption, and assembly to obtain functional nanomaterials with unique properties that are not possible with many bulk syntheses. In order to better control the synthesized functional nanomaterials, it is necessary to realize the precise manipulation of the liquid–liquid interface. This is crucial for scientific fields such as chemistry,1–3 biology,4 and the environment.5 For example, Rocca et al.6 formed hydrogels with well-defined geometries in microfluidic chips by in situ interfacial polymerization, which can be useful for microfluidic applications requiring filtration or localization of analytes and reaction intermediates. Wang et al.7 and Zhang et al.8 synthesized the products by liquid–liquid interfaces, which can be used for cell behavior studies, drug delivery, and gene therapy. By controlling the position and shape of the liquid–liquid interface, it can be adapted to different application situations. Microfluidics plays an important role in the precise study of controlling liquid–liquid interfacial conditions, and there are usually two types of systems,9–12 closed microchannel systems and open microchannel systems.

Open microfluidic systems, in addition to guaranteeing the core functionality of traditional microfluidics for detecting and manipulating very small volumes of fluid, offer other unique advantages in that the device does not need to be bonded to other surfaces, is easy to fabricate and use (e.g., eliminating bubble trapping and associated equipment failures), and lowers the threshold of adoption by the end-user. The use of open channels also facilitates subsequent characterization and optimization. However, most current research on the control of liquid–liquid interfaces is based on closed microchannel systems.11,12 Because in open microfluidic systems, the liquid located in the microchannel is directly exposed to air, the system is greatly affected by evaporation, resulting in incompatible two-phase liquids failing to form an interface. Many open microfluidic systems use an oil layer to cover the channels to avoid the influence of air,10 but the introduction of the oil layer increases the complexity of the system and fails to obtain the desired and adjustable liquid–liquid interface.

In this work, we propose a new method to obtain a stable liquid–liquid interface. The solid–liquid interface is used to replace the gas–liquid interface, which reduces the interference of the motion of the gas phase with the immiscible two-phase liquid and obtains the liquid–liquid interface. Meanwhile, microstructures are incorporated into the microchannels to regulate the position and shape of the interface. The effects of microchannels and microstructures with different parameters on the final position and shape of the liquid–liquid interface are investigated.

In order to regulate the position and shape of the liquid–liquid interface in an open microchannel, this work proposes a new method to make two liquids meet and form an interface by replacing the gas–liquid interface with a solid–liquid interface. In addition, use the pinning effect of microstructures to stabilize the liquid–liquid interface. The process of regulating the interface by this method is shown in Fig. 1.

Figure 1(a) illustrates the hydrophilic treatment of the substrate surface and the outward diffusion of the aqueous phase after it falls to the bottom of the hydrophilic fluid pool. It is usually possible to make the substrate surface hydrophilic by using oxygen plasma treatment or surface chemical modification. In this work, we chose to treat the chip with oxygen plasma to make the microchannel region on the glass substrate better wettable. When the outside of the water comes into contact with the microchannels, the liquid flows forward along the microchannels, driven by capillary forces. As shown in Fig. 1(b), when the organic phase is added to the other side of the fluidic cell, the aqueous and organic phases are incompatible due to the exposure to air, and the aqueous phase that has already flowed forward along the channel flows backward into the cell, which is not conducive to the formation of a liquid–liquid interface.

It is proposed to utilize a glass cover that slowly pours from the position of the aqueous phase to the organic phase [Fig. 1(c)]. Due to the presence of the glass cover, the gas–liquid interface on the upper surface of the liquid exposed to air is gradually replaced by a solid–liquid interface. The aqueous-phase advances through the microchannel again. When the glass cover falls completely, the aqueous and organic phases come into contact, and the liquid–liquid interface is formed at the junction of the liquid pool and the microchannel [Figs. 1(d) and 1(e)]. The liquid–liquid interface formed here is unstable, and the liquid–liquid interface will continue to move until it reaches the microstructure, where the forces will reach equilibrium. The interface remains stable, as shown in Fig. 1(f).

The above-mentioned methods can be analyzed for the applied forces at the interfaces in the microchannels through the four representative cases in Fig. 1. In order to explain this principle qualitatively, the forces acting on the interfacial front (or curved liquid surface) in different cases are given in Fig. 2.

Figure 2(a) shows the forces acting on the aqueous phase just entering the microchannel, i.e., the state of Fig. 1(a). The sides of the curved liquid surface in contact with the wall are subjected to the gas–solid surface tension γsg, the liquid–solid surface tension γsl1, and the gas–liquid surface tension γlg, respectively, and the curved liquid surface is also subjected to the additional pressure ΔP0 on the left side. The direction of the combined force on the curved liquid surface is to the left, and the interface moves to the left. Mathematically, according to the Young–Laplace equation, the additional pressure on the bent liquid surface of the microchannel can be described as
ΔP =γ×(1/R1+1/R2).
(1)

Here, 1/R1 and 1/R2 are the two principal curvatures of the surface microelements cut by mutually orthogonal segments (top-view horizontal segment and depth-directed segment). Since the radius of curvature R2 in the depth direction is much larger than the feature length (width), the effect of R2 can be neglected.

Figure 2(b) shows the interference of organic phase molecules on the curved liquid surface after the dropwise addition of organic phase, i.e., the state in Fig. 1(b), where the aqueous phase in the microchannel moves to the right. At this time, the surface tension γlg' of the aqueous phase is smaller than γlg, the value of the additional pressure ΔP1 is significantly reduced, and γsg' is smaller than γsg, so the interface moves to the right. Figure 2(c) shows the force on the liquid–liquid interface after the interface between the aqueous and organic phases has been formed, i.e., the state of Figs. 1(d) and 1(f). Different from Fig. 2(a), the gas–solid interface is replaced by the organic-phase liquid–solid interface, and γsl2 is smaller than γsl1; the liquid–liquid interface replaces the gas–liquid interface, the interfacial tension γl1l2 is smaller than γlg, and the additional pressure ΔP2 becomes smaller. The liquid–liquid interface moves to the right as the bending liquid surface is subjected to a combined force to the right.

In Fig. 2(d), the liquid–liquid interface reaches equilibrium at the microstructure of the microchannel, i.e., the state in Fig. 1(f). Compared with Fig. 2(c), in the contact between the curved liquid surface and the top of the microstructure, the microstructure pegs the curved liquid surface. The two liquids coexisting in the microstructure in Fig. 1(d) generate three capillary forces τl1l2, τsl1, and τsl2 along the liquid–liquid interface and the solid–liquid interface at the contact line, which are caused entirely by cohesion between the liquids and act along the liquid–liquid and liquid–solid interfaces, and are equilibrated with τs, which is caused entirely by the attractive force between the solids and liquids.13 Therefore, the magnitude of τs in equilibrium is the magnitude of the pinning force when the three-phase contact line is pinned. τl1l2 and τsl2 attempt to push the interface to the right. The pinning force τs and the additional pressure ΔP3 work together to prevent the interface from moving to the right. The additional pressure ΔP3 is influenced by the size of the microstructure, i.e., the radius of curvature R1. Table I lists the range of structural parameters used in the mask design. The details are explored in detail in the results and discussion.

These parameters are defined in Fig. 2(d), where h is the distance between the front of the microstructure and the interface between the reservoir and the microchannel. The angle between the fabricated microstructure and the wall is 45°.

Materials and apparatus: SU-8 photoresists and supporting developers are from Micro Chem. Mask plate from Changsha Shaoguang Chromium Plate. SPIN-1200T homogenizer from MIDAS (Korea). SA series front-aligned lithography from ABM (USA). Soda-lime glass from Guluo Glass.

A series of microfluidic devices consisting of nine side-by-side parallel microchannels (length ∼7 mm) connected to reservoirs (diameter ∼6 mm) at each end were fabricated by photolithography. The fabrication of the microfluidic channel consists of the following steps: the SU-8 photoresist is first spin-coated onto a glass substrate to obtain a 10 µm thick film; after soft baking, the SU-8 photoresist is exposed to 365 nm UV light through a pre-fabricated mask plate; then, the exposed SU-8 samples are subjected to further baking, development, and hard baking. The soft baking and post-exposure baking times are 65 °C for 5 min, 95 °C for 10 min, exposure for 8 s, development for 15 s, and hard baking at 150 °C for 20 min.

Materials and apparatus: Hexanol (>99.8% purity) was from Shanghai National Pharmaceutical Reagent. AL 76 Plasma Cleaner from Alpha Plasma (Germany). Optiphot 100 optical microscope and CCD from Nikon (Japan).

Before each experiment, the prepared glass chips were cleaned with a large amount of water. For droplet liquid experiments, the glass chips and caps were first treated for 15 min in an oxygen plasma cleaner with 500 W power and 100 SCCM oxygen flow. Water (with carmine as the stain) and hexanol (with curcumin as the stain) were used as the test liquids. Adding trace amounts of the stain did not change the surface tension or contact angle of the fluid. The contact angle was nearly 0° for drops of water and ∼40° for drops of hexanol. To introduce the fluids into the microchannel, a precise quantity of droplets was carefully dispensed onto the reservoir using a pipette gun, allowing for spontaneous diffusion into the microchannel with a volume of liquid that just filled the entire reservoir. The entire procedure was completed in less than 10 min.

The glass chip was placed on a microscope stage and observed using a CCD configured on an optical microscope after the drop addition was completed. The liquid contact angle was tested by obtaining image information through a horizontally built objective lens and CCD, and ImageJ was used to process the images to obtain contact angle information. All experiments were performed at 21.0 ± 2.0 °C room temperature.

Based on the experiments conducted by the above-mentioned method, the dynamic process of the liquid–liquid interface under the action of microstructures is shown in Fig. 3.

From Figs. 3(a)3(d), we can clearly see that when the motion process of the liquid–liquid interface is from not touching the microstructures to contact the microstructures, the curvature of the curved liquid surface 1/R1 gradually increases due to the restriction of the microstructures, and the additional pressure opposite to the direction of the motion is also gradually increased according to Eq. (1), so the interface is restricted by the microstructures pinned and the tendency of the motion is slowed down. The radius of curvature of the curved liquid surface varies between w/2 and b/2. When the curved liquid surface between the microstructures grows gradually, the curvature value becomes smaller, as in Figs. 3(e)3(g). At this time, the value of the additional pressure is decreasing, and the microstructures do not play a restrictive role in this process. Therefore, we can conclude that the radius of curvature of the curved liquid surface that breaks through the restriction of the microstructures is exactly b/2, i.e., half of the spacing of the microstructures.

When the curvature changes, the pinning force also changes. As shown in Figs. 3(h) and 3(i), when the microstructures fail to pin the curved liquid surface, the curved liquid surface continues to repeat the previous process until it is pinned by the microstructures and remains in a steady state. As seen by the simulation results from Figs. 3(j)3(o), the shape changes of the liquid–liquid interface moving at the microstructure match the experimental results. Therefore, a series of microstructures can be designed to ensure that the liquid–liquid interface can be pinned by the microstructures at different locations. Maintaining the liquid–liquid interface in a stable state at the microstructures is an adaptive process, so the curvature of the liquid–liquid interface is related to the nature of the liquid, the environment, and the structural parameters, and it can be adapted to different application scenarios by adjusting the relevant factors.

In order to obtain interfaces with different curvatures, the spacing of the microstructures, the width of the microchannels, and the positions of the microstructures were regulated.

Figure 4 demonstrates the different motion states of the liquid–liquid interface at a microchannel width w of 150 µm, microstructure spacing a of 45, 75, and 90 µm, and b of 20, 50, and 65 µm, and the curvature liquids are very different when they are in the steady state. From Fig. 4(a), it is observed that when the liquid–liquid interface is in the steady state, the interface cannot break through the restriction of the first pair of microstructures, and the interface stays at the tip of the microstructures. The curvature of the front end is determined by the size of the microstructure spacing a. In Fig. 4(b), the spacing of the microstructures becomes larger, the additional pressure becomes smaller, and the interface is able to move to a farther position. The curvature of the front end at this point is mainly determined by the size of the microstructure spacing b. However, Fig. 4(c) demonstrates that when the spacing of the microstructures is too large, the microstructures fail to provide good confinement, and the liquid–liquid interface will move forward all the way through the gaps in the microstructures. The results show that the spacing of microstructures plays a crucial role in the stabilization of the liquid–liquid interface in addition to affecting the interfacial curvature.

The curvature results of the liquid–liquid interface when the microstructures are located at different positions in the channel are shown in Fig. 5.

Figure 5 shows that in a 100 µm wide microchannel, the position of the microstructures in the channel (all spacing b are 30 µm) affects the final stabilized position of the interface. Figs. 5(a)5(d) represent microstructures at distances of 5000, 4000, 3000, and 2000 µm from the microchannel opening, respectively. It can be seen that the number of microstructures that can break through the liquid–liquid interface and the shape of the interface are different at the various locations. As can be seen in Fig. 3, the radius of curvature R1 of the interface is at a minimum of b/2 and does not exceed w/2. Therefore, the radius of curvature of the bent-liquid surface in Fig. 5 ranges from 15 to 50 µm, which is correlated with the adaptation of the liquid–liquid interface at the microstructures. The number of microstructures that can be breached by the curved liquid surface varies depending on the position of the microstructures, so the position and shape of the liquid–liquid interface can be further controlled by adjusting the positions of the microstructures in the microchannels.

The results of controlling the curvature of the liquid–liquid interface by adjusting the width of the microchannel are shown in Fig. 6.

The widths of the microchannels in Fig. 6 are 100, 200, and 300 µm, and the spacing of the microstructures a is 60 µm and b is 30 µm. It is clear that each set of interfaces is well restricted to the first pair of microstructures, whose curvature 1/R1 at the tips is determined by the spacing of the microstructures b. The widths of the microchannels mainly affect the overall shape of the interfaces after stabilization. When the interfaces cannot break through the restriction of the first pair of microstructures, the curvature 1/R1 is independent of the width w of the microchannels.

Controlling liquid flow in open microchannels is important for the development of microfluidic devices and understanding the motion of the liquid–liquid interface. In this paper, a method to control the liquid–liquid interface in an open microchannel is proposed. It is proposed that the interface is formed by introducing a glass cover plate, and the pinning action of microstructures is utilized to control the interface morphology. Based on the proposed methodology, it theoretically analyzes the force situation of the liquid–liquid interface in different states. By analyzing the dynamic process of the liquid–liquid interface in microchannels containing microstructures, the effects of the width of microchannels, the spacing, and the position of microstructures on the final interface morphology are investigated, and the main influencing factors of the curvature of the interface in different types of cases are derived. This method can achieve precise control of the interface morphology by using microchannels and microstructures of different sizes, microstructures of different positions, and liquid–liquid interfaces adapted to microstructures for steady-state behavior. It is suitable for manipulation in different environments and has broad application prospects.

This work was supported in part by the National Natural Science Foundation of China (Grant Nos. 12375330, U2032157, and 11775224), supported by USTC Research Funds of the Double First-Class Initiative (Grant No. YD2310002008), the Youth Innovation Promotion Association, CAS (Grant No. 2020457).

The authors have no conflicts to disclose.

Lijuan Chen: Conceptualization (equal); Data curation (lead); Formal analysis (lead); Investigation (lead); Methodology (equal); Writing – original draft (lead). Chenfei Guo: Data curation (equal); Formal analysis (equal); Methodology (equal). Kuanqiang Zhang: Data curation (equal); Formal analysis (equal); Methodology (equal). Xu Ding: Data curation (equal); Formal analysis (equal); Methodology (equal). Ying Xiong: Formal analysis (equal); Methodology (equal). Yong Guan: Formal analysis (equal); Funding acquisition (equal); Methodology (equal). Zhao Wu: Formal analysis (equal); Funding acquisition (equal); Methodology (equal). Yangchao Tian: Funding acquisition (equal); Methodology (equal); Resources (equal); Supervision (equal). Gang Liu: Conceptualization (equal); Data curation (equal); Funding acquisition (equal); Project administration (lead); Resources (lead); Supervision (lead); Writing – review & editing (lead).

The data that support the findings of this study are available from the corresponding author upon reasonable request.

1.
F.
Zhang
,
J. B.
Fan
, and
S.
Wang
, “
Interfacial polymerization: From chemistry to functional materials
,”
Angew. Chem., Int. Ed.
59
,
21840
21856
(
2020
).
2.
K.
Piradashvili
,
E. M.
Alexandrino
,
F. R.
Wurm
, and
K.
Landfester
, “
Reactions and polymerizations at the liquid–liquid interface
,”
Chem. Rev.
116
,
2141
2169
(
2016
).
3.
N.
Puthiyottil
,
A.
Palamparambil
,
S.
Kaladi Chondath
,
S. N.
Varanakkottu
, and
M. M.
Menamparambath
, “
Interfacial tension-impelled self-assembly and morphology tuning of poly(3,4-ethylene dioxythiophene)/tellurium nanocomposites at various liquid/liquid interfaces
,”
ACS Appl. Mater.
15
,
51944
51957
(
2023
).
4.
N. M.
Oliveira
,
S.
Vilabril
,
M. B.
Oliveira
,
R. L.
Reis
, and
J. F.
Mano
, “
Recent advances on open fluidic systems for biomedical applications: A review
,”
Mater. Sci. Eng.: C
97
,
851
863
(
2019
).
5.
Z.
Wang
,
S.
Liang
,
Y.
Kang
,
W.
Zhao
,
Y.
Xia
,
J.
Yang
,
H.
Wang
, and
X.
Zhang
, “
Manipulating interfacial polymerization for polymeric nanofilms of composite separation membranes
,”
Prog. Polym. Sci.
122
,
101450
(
2021
).
6.
M.
Rocca
,
M.
Dufresne
,
M.
Salva
,
C. M.
Niemeyer
, and
E.
Delamarche
, “
Microscale interfacial polymerization on a chip
,”
Angew. Chem., Int. Ed.
60
,
24064
24069
(
2021
).
7.
W.
Wang
,
H.
Qi
,
T.
Zhou
,
S.
Mei
,
L.
Han
,
T.
Higuchi
,
H.
Jinnai
, and
C. Y.
Li
, “
Highly robust crystalsome via directed polymer crystallization at curved liquid/liquid interface
,”
Nat. Commun.
7
,
10599
(
2016
).
8.
Q.
Zhang
,
S.
Feng
,
W.
Li
,
T.
Xie
,
W.
Zhang
, and
J.-M.
Lin
, “
In situ stable generation of reactive intermediates by open microfluidic probe for subcellular free radical attack and membrane labeling
,”
Angew. Chem., Int. Ed.
60
,
8483
8487
(
2021
).
9.
G. V.
Kaigala
,
R. D.
Lovchik
, and
E.
Delamarche
, “
Microfluidics in the ‘open space’ for performing localized chemistry on biological interfaces
,”
Angew. Chem., Int. Ed.
51
,
11224
11240
(
2012
).
10.
C.
Li
,
Z.
Hite
,
J. W.
Warrick
,
J.
Li
,
S. H.
Geller
,
V. G.
Trantow
,
M. N.
McClean
, and
D. J.
Beebe
, “
Under oil open-channel microfluidics empowered by exclusive liquid repellency
,”
Sci. Adv.
6
,
eaay9919
(
2020
).
11.
P.
Jaiswal
,
U.
Kumar
, and
K. G.
Biswas
, “
Liquid–liquid flow through micro dimensional reactors: A review on hydrodynamics, mass transfer, and reaction kinetics
,”
Exp. Comput. Multiphase Flow
4
,
193
211
(
2021
).
12.
B.
Zhao
,
J. S.
Moore
, and
D. J.
Beebe
, “
Surface-directed liquid flow inside microchannels
,”
Science
291
,
1023
1026
(
2001
).
13.
J.
Fan
,
J.
De Coninck
,
H.
Wu
, and
F.
Wang
, “
Microscopic origin of capillary force balance at contact line
,”
Phys. Rev. Lett.
124
,
125502
(
2020
).