Depinning of Multiphase Fluid Using Light and Photo-Responsive Surfactants.

The development of noninvasive and robust strategies for manipulation of droplets and bubbles is crucial in applications such as boiling and condensation, electrocatalysis, and microfluidics. In this work, we realize the swift departure of droplets and bubbles from solid substrates by introducing photoresponsive surfactants and applying asymmetric illumination, thereby inducing a "photo-Marangoni" lift force. Experiments show that a pinned toluene droplet can depart the substrate in only 0.38 s upon illumination, and the volume of an air bubble at departure is reduced by 20%, indicating significantly faster departure. These benefits can be achieved with moderate light intensities and dilute surfactant concentrations, without specially fabricated substrates, which greatly facilitates practical applications. Simulations suggest that the net departure force includes contributions from viscous stresses directly caused by the Marangoni flow, as well as from pressure buildup due to flow stagnation at the contact line. The manipulation scheme proposed here shows potential for applications requiring droplet and bubble removal from working surfaces.

Introduction

Dynamic control of fluid motion in a multiphase system offers exciting and transformative manipulation capabilities for a wide spectrum of laboratory and industrial applications, such as phase-change heat transfer,[1] water harvesting,[2] and electrocatalysis.[3] For example, the critical heat flux (CHF) in boiling heat transfer, which is the maximum heat flux before the nucleate boiling regime breaks down, can be enhanced by over 100% by surfaces that facilitate bubble formation and departure.[4,5] Similarly, the fabrication and manufacturing of multiscale and multiphase materials[6] necessitate the precise and rapid control of precursors, such as prescribing the alignment of fibers, to achieve desired material properties with high throughput.[7] Past research efforts explored the use of coatings and surface structures to passively control fluid behaviors,[8−11] whereas recent advances have enabled dynamic manipulation of bubble and droplet motions by introducing materials that respond to thermal,[12] chemical,[13] electrical,[14] or magnetic stimuli.[15] Notably, electrowetting-on-dielectrics systems can precisely move, merge, and split discrete droplets.[16] However, thermally and chemically actuated systems typically have long response times and reversibility issues. Electrowetting and magnetically controlled systems generally require high voltages[17] or strong magnetic fields,[18] with implications for reliability and safety. In addition, sophisticatedly patterned electrodes are needed, complicating practical implementations.

Recently, stimuli-responsive surfactants have emerged as an attractive alternative for dynamic manipulation. Recent work has utilized ionic surfactants to dynamically change the friction coefficient,[19] wettability,[20] and the onset of nucleate boiling,[21] by controlling the attraction to a surface using an electric field requiring only a few volts. However, to realize spatial control, such an approach still requires prepatterned electrodes, which limits spatial resolution and real-time reconfigurability. An alternative approach leverages photoresponsive surfactants, which can reversibly switch molecular conformation and thereby change interfacial tension when illuminated at specific wavelengths (Figure a). They are exciting candidates for dynamic fluid manipulation, because light can be easily reconfigured with a high spatial resolution (up to the diffraction limit) and instantaneous response. Experimental studies have successfully demonstrated manipulation of micro/nanoparticles,[22] liquid drops,[23,24] Janus drops,[25] and solid swimmers[26] in liquids; theoretical work has elucidated the link between light-induced surface tension changes and the resulting flow fields, for applications including translating drops and large interfaces.[27,28] However, many practical multiphase applications such as boiling, condensation, microfluidic-based synthesis, and electrocatalysis require removing bubbles and droplets from a solid surface, which has not been investigated using photoresponsive surfactants, to the best of our knowledge.

Figure 1

Mechanism of photo-Marangoni effect for light-activated bubble/droplet departure. (a) Photoresponsive surfactants reversibly switching molecular conformation upon illumination of light with appropriate wavelengths. In this example, forward switching is activated by UV light. The reverse reaction occurs once UV is removed and is accelerated by green light or white light. The surface tension is changed reversibly by UV/green illumination. (b) A surface tension gradient can be established along the interface by applying light with an intensity gradient, or only to one region on the interface. The subsequent Marangoni flow drives the bubble/droplet to depart. (c) Reversible photoconversion of the spiropyran (SP) isomer to the merocyanine (MC) isomer form. (d) 0.1 mM spiropyran water solution changes from transparent to light pink under 365 nm UV illumination. (e) 0.1 mM spiropyran water solution switches from MC form back to SP form, as shown by the color change from light pink to transparent.

In this study, we introduce the use of photoresponsive surfactants to promote bubble and droplet departure from a solid surface. We employ a photoresponsive surfactant that we recently designed and synthesized,[29] which is capable of achieving fast, reversible, and significant interfacial changes of different solvents, thanks to its large polarity change under illumination. To highlight the potential of this approach, we demonstrate the removal of droplets and bubbles pinned on a solid substrate using dilute (0.1 mM) photoresponsive surfactants and low-intensity (0.25 mW/mm2) lights. This is achieved by creating a net force on the bubble or droplet due to the Marangoni effect induced by the nonuniform distribution of these photoresponsive surfactants (Figure b). Since light is used to induce Marangoni stresses, we refer to this effect as “photo-Marangoni”. In particular, we demonstrate that the departure of an otherwise stably pinned droplet can occur in less than 0.4 s. For pinned air bubbles, the departure volume can be reduced by 20%, indicating that departure is significantly accelerated. Numerical simulations are performed to understand the dominating factors contributing to the bubble and droplet departure. The physical insights on fundamental fluid manipulation gained in this study have broad potential applications, such as drug delivery, boiling heat transfer, and electrocatalysis.

Scaling Estimates of Photo-Marangoni Effects in Drop/Bubble Departure

To estimate the potential significance of photo-Marangoni forces in the depinning of drops and bubbles, we perform order-of-magnitude estimates. For this purpose, it is useful to briefly review the underlying molecular mechanisms.

Photoresponsive surfactants are synthesized by covalently attaching a photoswitchable molecule to a surfactant carrier.[30] The photoswitchable moiety can exist as two different isomers. The first isomer is thermodynamically stable in the dark, but will undergo a conformational switch to the metastable isomer initiated by illumination of a specific wavelength (Figure a). If illumination is stopped or a different wavelength of light is applied, the isomer can revert back to the thermodynamically stable form. The reversible photoconversion is associated with a change of surface tension or interfacial tension γ of the surfactant-containing solution (Figure a).[31,32]

Applying light to only one side of a single bubble or droplet whose surface is covered with photoresponsive surfactants (Figure b) causes a localized photoreaction and induces a change of γ at the interface between the two fluids, which drives a local Marangoni flow toward the region of higher γ. The Marangoni flow pushes surrounding fluid (downward in Figure b), such that the bubble/droplet experiences a Marangoni-induced force FM in the opposite direction, easing the depinning and departure from the substrate (the fluid mechanics of generic Marangoni-induced droplet propulsion are reviewed in ref (13)). We propose here that, if a gradient in γ can be tuned and sustained by maintaining a constant gradient of two photoswitched isomers, bubbles and droplets can be continuously removed from the solid substrate on-demand.

To estimate the relative change of surface tension Δγ/γ that is required to depin an air bubble immersed in an aqueous solution (Figure b), we consider the relative order of magnitude of the relevant forces. For a static bubble pinned to a substrate, the bubble diameter db at departure from a substrate is obtained by balancing buoyancy Fb, which is of the order gΔρdb3 (where Δρ is the density difference), and the vertical component of the pinning surface force Fpin,vertical ∼ γdb sin2 ρ (where φ is the macroscopic or apparent contact angle in the liquid, and “ ∼ ” denotes order of magnitude), yielding db ∼ sin φ(γ/gΔρ)1/2. Here, the pinning force (Figure b) originates from the surface tension γ multiplied by the contact line length πdb sin φ, which results in a vertical component of Fpin,vertical ∼ γdb sin2 φ. For photo-Marangoni forces to assist departure, they must be non-negligible by comparison to buoyancy for a bubble of size db or smaller. If a surface tension change of the order Δγ exists across the height of the bubble and can be maintained to drive a flow, the resulting photo-Marangoni force that can assist depinning is of the order Δγdb; requiring this to be at least 10% of the buoyancy force sets Δγ ≳ 0.1gΔρdb. In the extreme case of microgravity applications or of immiscible fluids with matched densities, the photo-Marangoni force alone must overcome the totality of the pinning surface force Fpin,vertical, which requires Δγ ≳ γ sin2 φ. These order-of-magnitude estimates inform our choice of photoresponsive surfactant, as explained below.

Characterization of SP-DA-PEG

An ideal surfactant to achieve bubble departure in a multiphase system should display a large surface tension change, good reversibility, and fast switching kinetics activated by a low-intensity light. As such, we designed and synthesized[29,33] a spiropyran-based photoisomerizable molecule (SP-DA-PEG) using a recently developed, norbornadiene enabled Diels–Alder (DA) click chemistry (Supporting Information).[33,34] As shown in Figure c, SP-DA-PEG comprises the photoswitchable spiropyran unit at the surfactant tail and the poly(ethylene glycol) monomethyl ether (PEG, Mn ≈ 550 g/mol) chain as the hydrophilic headgroup on which spiropyran is immobilized. The surfactant design enables the largest change in polarity upon photoswitching as the initially amphiphilic molecule loses its hydrophobicity upon irradiation. The isomerization of the spiropyran (SP) isomer to the merocyanine (MC) isomer is activated by UV light (300–375 nm), whereas the reverse photoisomerization occurs upon the removal of UV irradiation, which can be further accelerated by application of visible light (500–600 nm)[23,31] (Figure c). In Figure d, the color of a 0.1 mM spiropyran water solution changes from transparent to light pink upon irradiation of 365 nm UV light, using a low-power fiber-coupled UV LED. The region not illuminated by the UV light remains transparent. After illuminating the entire solution with a high-intensity UV lamp, which turns the solution pink (MC isomer), we illuminated the solution using a fiber-coupled white LED (Figure e). This quickly generated a clear region, which indicates the fast reverse reaction of MC to SP under white light. All MC isomers are completely transformed back to SP in the right-hand image of Figure e. In previous work, we explored the photowetting behaviors of SP-DA-PEG and revealed the excellent reversibility and stability of SP-DA-PEG at oil/water interfaces by examining both the kinetics and the interfacial tensions.[29] Compared with many previously reported photoresponsive surfactants, the ease of synthesis and well-studied properties of SP-DA-PEG make it an ideal candidate for the manipulation of fluid systems explored in this work.

We measured the surface tension response of SP-DA-PEG solutions during photoisomerization using the pendant drop method[35] on a commercial tensiometer (Theta Flex, Biolin Scientific). Although an accurate measurement of interfacial/surface tension of surfactant solutions can be challenging, as the time scales for surfactants to diffuse to the interface and for the droplet to relax to full equilibrium can take tens of minutes,[36] these pendant drop measurements are directly relevant to depinning applications, where time scales are inherently much shorter than those associated with molecular diffusion.

First, we applied alternating irradiation of 365 nm UV and white light to an aqueous SP-DA-PEG solution (0.1 mM) and observed the response of an air bubble. In Figure a, γw0 quickly rises to 49.5 from 46 mN/m after 1 s of UV irradiation. Once UV is removed and the white light is applied, γw0 undergoes a rapid decrease. After 16 cycles of irradiation, γw0 can recover its original value, which indicates that spyropiran could potentially maintain its performance through large numbers of depinning cycles.

Figure 2

Reaction kinetics of SP-DA-PEG isomerization. (a) Surface tension of 0.1 mM SP-DA-PEG–water solution under repetitive pulsed UV light. (b) Surface tension of 0.1 mM SP-DA-PEG water solution under UV irradiation with different light intensity. (c) Reversible change of interfacial tension between 0.1 mM SP-DA-PEG–toluene solution and DI water.

Figure b shows the time evolution of γw0 of 0.1 mM aqueous SP-DA-PEG solution under different light intensities. At the smallest intensity of 0.02 mW/mm2, γw0 changes from 46.5 to around 50.4 mN/m in 38 s. When light intensity is increased, more SP molecules are transformed into the open-ring conformation, MC, at a faster rate, and the resultant surface tension is increased to 54.3 mN/m within 8 s. In addition, most SP molecules can be converted into the open-ring form and the photoconversion process saturates once the light intensity exceeds 0.25 mW/mm2, which is still a relatively low intensity in most optical applications. After the UV light is removed, γw0 reverts to its original value for all cases.

We also measured the dynamic response of the interfacial tension γtw between deionized (DI) water and a 0.1 mM SP-DA-PEG–toluene solution, as shown in Figure c. Under 365 nm UV light with intensity of 0.59 mW/mm2, γtw quickly changes from 12.33 ± 0.05 mN/m (SP) to 8.59 ± 0.10 mN/m (MC), corresponding to a 30% change.

Droplet Departure Experiments

We carried out experiments on the departure of a liquid droplet, consisting of toluene, in water. The experimental setup is shown in Figure a. A toluene droplet is injected into a water bath through a round hole (d = 0.61 mm) using a syringe pump to control the injection rate. The droplet contains 0.1 mM SP-DA-PEG; this concentration showed the most significant interfacial tension change through our pendant drop experiments. The droplet is stabilized at ∼15 μL (corresponding to a diameter of approximately 3 mm) and remains pinned to the hole (Figure b). Later, another equal-sized droplet is generated and equilibrated for 10 s. Then, the output from a multiline argon ion UV laser (351.1 + 363.8 nm, beam diameter of 1.3 mm) is directed to the top of the droplet (height of 3.5 mm). The incident light intensity on the droplet is 44 mW/mm2, as measured in the empty container. The laser enables precise targeting of the photoswitch, and the higher intensity accelerates the isomerization to MC.

Figure 3

Droplet departure activated by UV laser and LED. (a) Schematic of the experimental setup of laser-activated droplet depinning. A ∼15 μL droplet is generated by pushing the toluene solution with 0.1 mM SP-DA-PEG through a hole (0.61 mm in diameter). A collimated UV laser is directed at the top part of the droplet to trigger the photo-Marangoni effect. Experimental results show that (b) without UV laser illumination, a ∼15 μL droplet stays static, and (c) with UV laser illumination, a ∼15 μL droplet departs in only 0.38 s. (d) The photo-Marangoni effect facilitates the depinning of a toluene droplet (∼11.5 μL) when the droplet top is irradiated by a low-intensity UV LED. The depinning time td is 3.10 s. (e) When the UV LED illuminates the bottom of a same-sized toluene droplet, the photo-Marangoni effect suppresses the departure of the droplet by creating an upward Marangoni flow. The depinning time td increases to 40.00 s. (f) Depinning times td of droplets when illuminated with a UV LED as a function of droplet volumes. td becomes orders of magnitude higher when UV irradiation is applied at the droplet bottom.

Figure c demonstrates that the initially pinned toluene droplet departs the substrate in only 0.38 s after UV irradiation (SI Movie 1). We define the instance when the laser is turned on as t = 0 s. A pink region rapidly forms within the top part of the droplet where the UV beam passes through, indicative of the onset of the photoconversion. The widening of the pink region observed at t = 0.18 s suggests fluid flow from the pink MC-rich lower interfacial tension region to the clear SP-rich higher interfacial tension region. The fluid motion inside the toluene droplet is demonstrated by visual inspection of tracer particles (SI Movies 2 and 3). The detailed flow field and force analysis will be discussed in the following sections. To ensure that droplet departure is activated by photo-Marangoni effect rather than the possible thermal-Marangoni effect induced by laser heating within the droplet, we simulated the temperature rise ΔT within the toluene droplet (see Supporting Information). It is found that ΔT is limited to only 0.25 °C before droplet departure (t = 0.38 s). In addition, the droplet temperature is experimentally measured to have increased from 21.7 ± 0.1 to 23.2 ± 0.8 °C during 17 s of laser heating (see Supporting Information). Previous work demonstrated that γtw for water/toluene decreases by 1.8% when the temperature is increased from 20 to 25 °C.[37] Therefore, thermal Marangoni effects are negligible in this experiment. This estimate also illustrates the advantage of the photo-Marangoni effect over thermal Marangoni, in that a much smaller light intensity can manipulate multiphase flow without the need to generate a substantial temperature gradient in the droplet. Furthermore, the light intensity used for photo-Marangoni experiments (44 mW/mm2) is 2–4 orders of magnitude lower than those previously reported for thermal Marangoni examples.[38,39]

We further demonstrate that droplet departure can be activated with even lower light intensity, and that the time for droplet departure td shows a strong dependence on droplet size. Here, td is defined as the time of droplet departure after UV irradiation. In this second set of experiments, we use a fiber-coupled UV LED with a light intensity of only 0.59 mW/mm2 (this is nearly 75 times smaller than in our experiments with the UV laser). A computer-controlled droplet dispenser is used to generate a toluene droplet with precisely controlled volume. As before, the droplet contains 0.1 mM SP-DA-PEG and is immersed in DI water. The UV LED illuminates the top part of a 11.5 μL droplet, as shown in Figure d. As a result, γtw locally decreases, and a downward Marangoni flow drives the droplet to depart after 3.10 s. By contrast, a reverse-direction photo-Marangoni flow can be generated by illuminating the bottom part of the droplet with the UV LED (Figure e, SI Movie 4). This reverse photo-Marangoni flow hinders droplet departure, as shown by the fact that td increases by an order of magnitude (td = 40.0 s). The droplet eventually still departs the surface; a contributing factor of this departure may be the reduced pinning force Fpin as a result of the decreased interfacial tension γtw after the photoconversion. As shown in Figure c, photoconversion will cause γtw to decrease by 30%. As the pinning force is proportional to γtw, the buoyancy may eventually be able to overcome the pinning force and the downward photo-Marangoni force. In Figure f, we measured the depinning time td for toluene droplets with different volumes. As the droplet volume increases, td decreases exponentially due to increased buoyancy, yet the departure time is significantly shorter for droplets with UV applied at the top than at the bottom for all droplets tested.

Bubble Departure Experiments

In this section, we show that the photo-Marangoni effect can also be used to assist the departure of air bubbles. The assisted departure is demonstrated by a reduction in the maximum bubble volume Vd before departure. Figure a shows the experimental setup. A syringe pump is used to generate an air bubble in a chamber filled with 0.1 mM SP-DA-PEG in toluene solution. The same UV LED used for the droplet experiments illuminates from 2.5 cm above the bubble. The highly diverging feature of its light naturally creates a light intensity gradient in the vertical direction along the bubble. Furthermore, a 532 nm green laser shines a beam horizontally at the bottom of the bubble. The green light irradiation accelerates reverse isomerization from MC to SP and subsequently enhances the composition gradient of surfactants along the interface. In turn, this increases the surface tension difference across the bubble. In our previous work,[24] it was found that the surface tension of SP-DA-PEG toluene solutions (γt0) decreases after the photoisomerization from SP to MC. Therefore, the bubble top is rich in MC and has a lower surface tension, leading to a downward photo-Marangoni flow which drives the bubble to depart. Using the pendant drop method, γt0 is measured to reduce slightly from 23.25 to 22.75 mN/m upon UV irradiation, a modest change (only 2%) compared to that of γtw.

Figure 4

Departure of an air bubble incurred by the photo-Marangoni effect. (a) Schematic of the experimental setup in the bubble experiment. The photo-Marangoni effect is driven by a UV LED illuminating downward from 2.5 cm above the bubble, and it is further assisted by a collimated green laser beam shining near the bottom of the bubble. Together, these enhance the difference in isomer composition across the bubble, thereby increasing the surface tension difference. (b) A pinned air bubble (∼3.36 μL) naturally shrinks in toluene solution and fails to depart. (c) An air bubble of the same initial volume leaves the surface under the photo-Marangoni effect. (d) Volume change of an air bubble before and after UV irradiation; departure occurs at a volume that is smaller than the initial one. (e) Bubble volume at departure Vd in SP solution, MC solution, SP solution when UV LED is on and green laser is off, and SP solution when UV LED and green laser are both on.

We compared Vd for bubbles injected into SP solutions, MC solutions, and solutions with a mixture of SP and MC isomers by controlling the illumination conditions. Time-lapse images of the injected bubble immersed in 0.1 mM SP-DA-PEG–toluene solution when no light is applied are shown in Figure b. At t = −10 s, the syringe pump is stopped and the bubble remains pinned to the surface. A slow shrinkage is observed over a period of 60 s, which is likely due to a small imperfection in the seal of the gas supply line. In contrast, when UV LED is turned on (at t = 0 s) for an equal-sized bubble generated with the same method (Figure c), the originally shrinking bubble starts to inflate as the photo-Marangoni lift force is applied and departs from the surface at t = 58 s. The time-dependent volume of the bubble during this process is shown in Figure d. The thermal Marangoni effect is negligible, as the solution temperature is measured to increase by only 0.5 °C during 1 min of UV irradiation (see Supporting Information). Bubble departure volumes for SP-only solutions Vd,SP and for MC-only solutions Vd,MC are measured by generating a quasi-steady-state bubble using the syringe pump with a constant and low flow rate of 1 μL/min. It is found that Vd,SP = 4.05 μL and Vd,MC = 3.96 μL. The difference between Vd,SP and Vd,MC agrees well with the surface tension change of 0.1 mM SP-DA-PEG–toluene solution before and after the photoisomerization. The bubble departure size can further be reduced to Vd,UV,G = 3.25 μL (Figure e) by applying UV and green light (5 mW) to the top and bottom of the bubble, respectively, as illustrated in Figure a. The photo-Marangoni force FM, as a percentage of the buoyancy force, herein is evaluated to be FM/FB ≅ (Vd,SP – Vd,UV,G)/Vd,SP = 19.8%, even with a modest change (2.1%) of surface tension of SP-DA-PEG solution. The above equation assumes that the bubbles are in quasi-static states and that the pinning force is relatively the same for all the cases. By designing new surfactants and characterizing the surface tension change in different solvents, the photo-Marangoni effect acting on gaseous bubbles can potentially be further enhanced.

Finite Element Simulations

We carried out several finite element simulations to investigate the lift force due to the photo-Marangoni effect (details in Methods), focusing on toluene droplets in DI water. We first simulate the departure of a 10.65 μL droplet (0.1 mM SP-DA-PEG) and assume a linear interfacial tension gradient with γtop = 8.59 mN/m at the top of the droplet and γb = 12.33 mN/m at the bottom. These values are based on the assumptions that the photoconversion equilibrium is established (Figure c) and the proportions of SP and MC molecules are determined by the local light intensity, which we will further discuss. Figure a presents the velocity field in both phases when the photo-Marangoni effect is activated. Within the droplet, the liquid travels downward near the interface due to the Marangoni stress and then circulates back to the top near the center axis. Outside the droplet, there is also a downward flow which moves faster at the interface and decays to almost zero away from the droplet. The pressure distribution is shown in Figure b. A high-pressure region appears near the base of the droplet in the outer phase, due to the stagnant flow near the bottom wall. This high-pressure region exerts an additional lift force on the droplet. Figure c plots the profiles of the z-component of total stress σ, viscous stress τ, and pressure p from the droplet bottom to its top. Interestingly, the result shows that pressure is the dominating factor near the bottom of the droplet, but then dips at a height below the maximum width of the droplet, slightly hindering departure. In contrast, the profile of τ is relatively flat, indicating that the outer phase is constantly pushing the droplet upward through surface shear.

Figure 5

Computational results of light-activated droplet departure enabled by the photo-Marangoni effect. (a) Streamline plots and w-velocity contour and (b) pressure contour around the droplet when the photo-Marangoni effect is activated. (c) Distribution of z-stresses acting on the droplet along the droplet interface. P is the vertical component of the pressure, τ is the viscous stress, and σ is the total stress. All quantities are positive if they push the drop upward. (d) Force contributed by pressure and viscous stress normalized by the buoyancy FB. Δγ is the difference between γb and γtop, and γ0 = 12.33 mN/m is the interfacial tension before the photoconversion. (e) Marangoni force FM and net force FN versus different interfacial tension gradients. FN = 0 is achieved at Δγ/γ0 = 21.5%.

However, the above simulation may be overpredicting FM, since it assumes that the maximum possible interfacial tension difference is realized across the drop, whereas in practice the finite-time kinetics of the photoconversion may not allow the MC-isomer to fully switch back to the SP-isomer before the fluid reaches the bottom of the droplet. Therefore, we reduced the interfacial tension difference and calculated the corresponding photo-Marangoni force FM, looking for the minimum interfacial tension difference that would enable departure. Since the forward reaction is fast (Figure c), we fixed γtop at 8.59 mN/m by assuming a complete photoconversion under UV irradiation, and gradually increased γb, corresponding to a decreasing proportion of MC molecules in the surfactant mixture. Figure d summarizes the upward force contributed by pressure Fp and viscous stress Fv under different Δγ/γ0, where Δγ = γb – γtop and γ0 is the interfacial tension before the photoconversion. Fp and Fv are of the same order of magnitude for Δγ/γ0 spanning from 0% to 30%, indicating that both pressure buildup at the bottom and the relatively constant viscous shear contribute substantially to the photo-Marangoni force FM. In Figure b, we have shown that the interfacial response of SP-containing solutions can be tuned by adjusting the UV light intensity. Therefore, the UV intensity can be used not only in depinning an originally pinned droplet, but also to tune the net force acting on the droplet and the subsequent departure dynamics. Note that as Fp and Fv increase with respect to Δγ/γ0, the net photo-Marangoni force FM = Fp + Fv eventually exceeds buoyancy FB at Δγ/γ0 = 29.4%. This suggests that the photo-Marangoni effect could even serve to completely replace buoyancy forces if using two density-matched fluids or if working in microgravity environments. The net force FN on the droplet can be estimated by considering the pinning force Fpin = πds γb sin θ, buoyancy FB = ΔρgV, and the photo-Marangoni force FM, such that FN = FM + FB – Fpin. In Figure e, it is shown that FM is of the same order of magnitude as FB. For small Δγ/γ0, FN is negative, indicative of a pinned droplet on the substrate; droplet departure (FN > 0) occurs when Δγ/γ0 exceeds 21.5%. Therefore, Δγ/γ0 required for departure is below the maximum Δγmax/γ0 = 30% that we previously measured in our tensiometer experiments, which further supports the feasibility of photo-Marangoni-assisted departure.

Discussion and Conclusions

In summary, we developed a dynamic, noninvasive, and sustainable manipulation scheme to enable bubble and droplet departure from solid surfaces using a light-responsive surfactant, namely, SP-DA-PEG. The excellent reversibility, stability, sensitivity to illumination, as well as fast photoisomerization kinetics make SP-DA-PEG ideal for applications involving bubble and droplet manipulation. In our experiments, an interfacial tension change as high as 30% and a droplet departure time of 0.38 s (for an otherwise pinned droplet) are achieved. Both experiments and simulations demonstrate that the photo-Marangoni force arises not only from the viscous shear along the interface but also from the pressure build-up due to flow stagnation at the bottom. As a matter of fact, the photo-Marangoni flow induces a Reynolds number (based on interface velocity, drop diameter, and ambient fluid properties) of approximately 600, indicative of a laminar flow where inertial effects may not be negligible. Therefore, even a modest change in surface tension (2.1%) can lead to a significant change in bubble behaviors (19.8% change in bubble departure volume). Note that the insights gained from the simulations can be extended to describe other photoresponsive surfactants and interfaces which may have a different interfacial response upon illumination. While the interfacial tension may increase or decrease upon illumination, as long as a downward Marangoni flow is generated, a lift force can be created to assist bubble/droplet depinning. Therefore, the depinning dynamics will remain unchanged. The photo-Marangoni effect can serve as an effective yet simple platform technology for noninvasive dynamic fluid manipulation that does not require electrical voltage, magnetic field, patterning complex electrodes, or engineering of well-designed surface textures. The surfactant concentration used in this work is around 0.1 mM, which is even smaller than the concentration of natural or man-made surfactants in environmental water.[40] The required light intensity for complete photoisomerization is only around 0.25 mW/mm2, comparable to the average solar irradiance received by the earth’s surface.[41] Therefore, the photo-Marangoni effect can contribute to developing new research capabilities not only in applications requiring fluid removal, such as heat transfer, electrolysis, and organic synthesis, but also in processes that entail biological compatibility and optical resistance, such as biochemical assays, dynamic patterning, and manufacturing of novel materials. Furthermore, the physical insights gained in the theoretical analysis, experimental observations, and numerical simulations will guide the design and synthesis of photosurfactants for future photodriven multiphase systems.

More work can be done to improve the photo-Marangoni effect to achieve faster response and more durable performance. The mechanism of interfacial tension change during photoconversions depends on the ring-opening process of the photoswitch unit, resulting in a polarity change and charge redistribution of SP-DA-PEG. Typically, the effect of spiropyran-based surfactants on nonpolar interfaces, such as the toluene–air interface, becomes less significant compared to that of an interfacial system involving polar liquids like water. Therefore, the meticulous design and optimization of molecular structures of the light-responsive surfactants is needed to achieve optimal interfacial tension change and reaction kinetics for different combinations of interfacial systems. In addition, more detailed numerical and theoretical studies on the coupling between multiphase flow, interface deformation, species transport, surfactant adsorption, and photoreactions can provide more insights to explore the surfactant transport and fluid dynamics associated with the photo-Marangoni effect for optimal performance.

Methods

Synthesis of SP-DA-PEG

SP-alcohol[42] and NBD-COOH[33] were synthesized following literature procedures, and mal-PEG was prepared from commercially available poly(ethylene glycol) monomethyl ether (PEG, Mn ≈ 550 g/mol) functionalized with 3-maleimidopropionic acid as previously described.[33] Norbornadiene functionalized spiropyran (SP-NBD) was prepared by first dissolving SP-alcohol (1.0 equiv) in anhydrous dichloromethane (DCM, 0.1 g/mL). To this solution was added and NBD-COOH (1.2 equiv), 1-ethyl-3-(3-(dimethylamino)propyl)carbodiimide (1.5 equiv), and 4-(dimethylamino)pyridine (0.10 equiv). The resulting mixture was allowed to stir for 24 h, then washed with water, dried over magnesium sulfate, and concentrated to afford a residue that was purified on a plug of silica. The residue was first rinsed with hexanes, then eluted from the silica plug with a mixture of ethyl acetate and hexanes (5:95). The organic solvent was removed under reduced pressure to afford SP-NBD as a viscous pink/green oil (61%) that was used in a one-pot deprotection to SP-cyclopentadiene and DA click to mal-PEG. To accomplish the one-pot DA click, SP-NBD (1.4 equiv), mal-PEG (1.0 equiv), and 3,6-di-2-pyridyl-1,2,4,5-tetrazine (2.0 equiv) were dissolved in chloroform (0.1 g/mL) and allowed to stir for 15–24 h. Reaction completion was determined by 1H NMR of the crude reaction mixture in deuterated chloroform. Upon disappearance of the maleimide protons (6.7 ppm)—indicating complete conversion of mal-PEG to SP-DA-PEG—norbornylene (2.0 equiv) was added to the solution to quench remaining tetrazine and facilitate purification. After stirring an additional 30 min, the solution was passed directly through a silica plug. The material was first rinsed with DCM to remove the yellow byproduct and then the pink product was eluted with a mixture of methanol in DCM (1:9). The solvent was removed under reduced pressure and the residue taken up in ethanol (25 mL/g polymer). The solution was transferred into centrifuge tubes and cooled in a dry ice/acetone bath (−78 °C) before the functionalized polymer was isolated by cold centrifugation. The supernatant was decanted from the pink pellet, and the cold ethanol precipitation and centrifugation was repeated once more. The resulting pink polymer pellet was redissolved in DCM and then dried under reduced pressure to afford the desired SP-DA-PEG (48%) as green/pink viscous liquid. Mn(NMR) = 1300 g/mol. Further synthetic detail and characterization of SP-DA-PEG can be found in the Supporting Information.

Sample Preparation

The photoactive surfactant selected for this work is a spiropyran-based photoisomerizable molecule (SP-DA-PEG). SP-DA-PEG can easily dissolve into toluene liquids by vortexing and sonication.

Light Sources

In the bubble experiment, a custom-built multiline argon ion UV laser (351.1 + 363.8 nm, beam diameter of 1.3 mm) is used with a beam diameter of 1.3 mm and light intensity of 44 mW/mm2. The fiber-coupled UV LED and white LED are directly bought from Thorlabs, Inc. (M365FP1-365 nm, 9.8 mW Fiber-Coupled LED, and MCWHF2-6200 K, 21.5 mW Fiber-Coupled LED) A high-intensity UV lamp (100 W over a lamp area of 216 mm × 140 mm, UVP High Intensity Lamp, Analytik Jena) is used to fully transform the SP isomers in 0.1 mM SP-DA-PEG solutions into the MC isomers so as to measure the bubble departure volume in pure MC solutions. A 532 nm green laser (5 mW, GMY-532-5F3-PP, Lasermate Group, Inc.) is used to expedite the reverse photoisomerization of SP-DA-PEG molecules in the bubble experiment.

Droplet and Bubble Experiments

A stainless steel tube with four glass windows is fabricated using the CNC milling machine and a round hole with a diameter of 0.6 mm is fabricated at the tube bottom using a mechanical drilling rig. The round hole is connected to a syringe pump and sealed by Dow Corning High Vacuum Grease. All experimental glassware are cleaned by sonication in water and isopropyl alcohol to prevent contamination by residual surfactant molecules and dust. Droplet is equilibrated for about 1 min before recording. For droplet manipulation with UV LED, the LED is placed about 2 cm away from the side of the droplet. For the bubble experiment, the UV LED is mounted around 2 cm above the air bubble. Videos are recorded from the side using Canon EOS 80D.

Pendant Drop Measurements

All measurements related to the surface tension or interfacial tension were carried out on the Theta Flex Tensiometer from Biolin Scientific. This instrument uses a high-speed camera to capture the outline of drops and bubbles. The pendant drop method is then used to calculate the surface tension or interfacial tension dynamically.

Particle Tracing

To demonstrate the flow pattern inside the toluene droplet, we added conductive silver-coated hollow glass microspheres (900 kg/m3, 25–65 μm, Cospheric LLC) into 0.1 mM toluene solutions before filling into the syringe. The particle-containing solution is then injected into DI water, and the droplet volume is tuned to make sure the droplet stays pinned after UV irradiation. The particle density is close to that of the toluene solution such that the effect of gravity or buoyancy on the motion of those microspheres is negligible. The argon ion UV laser was used to trigger the photo-Marangoni flow within the droplet, and all video frames were recorded using the Canon EOS 80D.

Finite Element Simulations

We capture the outline of the droplet right before it departs and assume that the droplet is in a quasi-steady state. The computational domain is axisymmetric (6 mm × 12 mm) and the flow is laminar due to low Reynolds number. The governing equations are continuity and Navier–Stokes equations:For simplicity, we used a fixed interface where Marangoni stresses are balanced by viscous stresses, consistently with previous work on interfaces with surfactants.[43] The curvature-dependent Laplace pressure is also considered and applied to the droplet[13]where Π is the stress tensor, is the unit outer normal vector, is the unit vector tangential to the interface and ▽ is the gradient along the interface. The operator δ represents the difference across the interface. No slip boundary condition is applied to the walls in the computational domain except the droplet bottom where an inlet boundary condition is applied and the top wall where an outlet boundary condition is enforced.