Article pubs.acs.org/Langmuir
The Dynamic Effects of Surfactants on Droplet Formation in Coaxial Microfluidic Devices J. H. Xu,* P. F. Dong, H. Zhao, C. P. Tostado, and G. S. Luo The State Key Lab of Chemical Engineering, Department of Chemical Engineering, Tsinghua University, Beijing 100084, China ABSTRACT: Droplet emulsification in microfluidic devices involves the constant formation of fresh interfaces between two immiscible fluids. When the multiphase system contains surfactant, dynamic mass transfer of the surfactant onto the interface results in a dynamic interfacial tension different from the static interfacial tension measured in an equilibrium state. In this work, we have systematically investigated the effects of surfactant concentration and type on the dynamic interfacial tension of two different liquid−liquid two phase systems [Nhexane/water−sodium dodecyl sulfate (SDS) and N-hexane/ water−cetyltrimethylammonium bromide (CTAB)] rapidly producing relatively small droplets in coaxial microfluidic devices. Dynamic interfacial tension experiments using the pendent drop method and a tensiometer were conducted, and a semiempirical equation was developed to put into context the effects of surfactants and the experimental conditions on droplet formation and dynamic interfacial tension in dynamic microchannel flows. The results presented in this work provide a more in-depth understanding of the dynamic effects of surfactants on droplet formation and the precise controllable preparation of monodispersed droplets in microfluidic devices.
1. INTRODUCTION In the past decade, the application of microfluidic technology has grown rapidly in various fields, particularly because of the new methods it has provided for the controllable preparation of monodipersed droplets. Generally, monodispersed droplets can be generated in flow-focusing microchannels,1−6 T-junction microchannels,7−10 and coaxial microchannels.11−15 In the drop emulsification process, surfactants are usually used to adjust the interfacial tension and stabilize the emulsion droplets against coalescence.16−18 Unlike other conventional emulsification methods, emulsification in microfluidic devices involves the continuous formation of fresh interface between the immiscible fluids (e.g., oil and water) at a rate proportional to the surface area and frequency of formation for the droplets.19 If surfactants are dissolved in the liquids at sufficient concentrations, they will quickly transport from the bulk and adsorb onto the freshly created interface. This process requires a finite amount of time to complete, since it is often limited by the transport of surfactants to the interface. Furthermore, the surfactants adsorbed onto the interface may realign along the interface when subjected to sufficient shear. Together these phenomena create a surfactant concentration gradient, and thus an interfacial tension gradient at the interface. This gradient then creates tangential stresses or “Marangoni stresses”, which may affect the dynamics of flow around the interface.20−25 Thus, droplet formation in microfluidic devices is a typical © 2012 American Chemical Society
nonequilibrium process strongly affected by the dynamic adsorption of surfactants.18,19,26 In previous studies, to mitigate any effects surfactant gradients on the interface might have on dynamic interfacial tension, excessive amounts of surfactant, roughly 10 times higher than the critical micelle concentration (CMC), are usually used.8,17,18 At such high concentrations, surfactant transport and adsorption was assumed to be much faster than droplet formation, allowing for an equilibrium interfacial tension to be reached and avoiding any associated Marangoni stresses during the droplet formation process. In our previous work,26 we studied the kinetics of surfactant adsorption at the droplet interface in a T-junction microchannel, and concluded that the method used to determine dynamic interfacial tension depended on the type of surfactant used. In dynamic microchannel flow experiments, small molecule surfactants such as sodium dodecyl sulfate (SDS) required a bulk phase concentration 1.5 times higher than its CMC in order to saturate the interface and achieve an equilibrium interfacial tension equivalent to that observed in the static case. However, Tween20, a larger molecule with a slower adsorption rate, required a bulk phase concentration more than 400 times Received: January 5, 2012 Revised: May 14, 2012 Published: May 31, 2012 9250
dx.doi.org/10.1021/la301363d | Langmuir 2012, 28, 9250−9258
Langmuir
Article
higher than its CMC to reach saturation. Steegmans et al.19 used a Y-junction microfluidic device to form micrometer-sized droplets and concluded that the dynamic interfacial tension was much higher than the equilibrium interfacial tension, even though the concentration of SDS was 10 times higher than the CMC. Furthermore, the authors showed that surfactant transport was dominated by convection using the bursting membrane method. These prior studies have shown that the static equilibrium interfacial tension is very different from the dynamic interfacial tension observed during droplet formation in microfluidic channels with formation times in the submillisecond to millisecond range. In this work, we used coaxial microfluidic devices to systematically study the effects of surfactant type and concentration on microdroplet formation. Furthermore, using the dynamic interfacial tension measured using the pendent drop method, we developed a semiempirical equation to discuss and examine the effects of surfactants and experimental conditions on droplet formation and dynamic interfacial tension in microfluidic devices.
coaxiality was achieved by matching the outer diameter of the capillaries with the inner dimension of the square channel. The second capillary was tapered to approximately 500 μm. The microfluidic devices were sealed using another thin PMMA plate (1 mm in thickness) and cured at 75 °C for 1.5 min using a high-pressure thermal sealing technique. 2.2. Materials. Several systems were used to study the dynamic effects of surfactants on droplet formation in the microfluidic devices. N-hexane was used as the dispersed phase, while deionized water with either SDS (288.38 g/mol) or cetyltrimethylammonium bromide (CTAB, 364.45 g/mol) as a surfactant was used as the continuous phase. SDS solutions with concentrations of 0.24, 1.0, 2.0, 4.0, 6.0, 8.0, and 10.0 wt % were prepared by dissolving SDS in deionized water. The SDS concentrations were chosen relative to its CMC of 0.24 wt %28. CTAB solutions with concentrations of 0.03, 0.3, 0.5, 1.0, 2.0, and 4.0 wt % were prepared by dissolving CTAB in deionized water, relative to its CMC of 0.03 wt.%.28 The viscosities of the continuous phases were measured by an Ubbelohde viscometer, and the static interfacial tensions of N-hexane/surfactant solution systems were measured with a tensiometer using the pendent drop technique (OCAH200, DataPhysics Instruments GmbH, Germany). All the experiments were carried out at room temperature. Figure 2 shows the relationship between the interfacial tension and the viscosity with surfactant concentrations. 2.3. Method. The flow rates of the two phases were controlled with two syringe pumps and three gastight syringes (LSP01-1B, Baoding Longer Precision Pump Co., Ltd.), respectively. Droplet formation experiments were carried out with an optical microscope (BX61, Olympus, Japan) equipped with a high-speed camera with a frequency of up to 1000 images per second (DK-2740, Dantech, Danmark). The diameter and the formation time of droplets were measured directly from the video with a phantom video player, an accessory software to the high-speed camera. The average droplet size (dav) and the coefficient of variation (CV) were determined by measuring the sizes of at least 100 drops from recorded pictures using image analysis software (ImageTool 3.00). The CV is defined by the following equation CV = δ/dav × 100%, where δ is the standard deviation, and dav is the average droplet diameter. The average droplet formation time, t, was determined by measuring the formation time of up to 10 droplets. The actual dispersed phase flow rates, Qd, were calculated from the droplet volume V and the droplet formation frequency f = 1/t by using the following equation:
2. MICROFLUIDIC DEVICES AND MATERIALS 2.1. Microfluidic Device. Monodispersed droplets can be generated in flow-focusing, cross-junction, T-junction, and coaxial microfluidic device geometries. In the first three microfluidic device geometries, prior studies have shown that two-phase flow is strongly affected by the wetting properties of the microchannel wall.1,2,8,16 In coaxial microfluidic devices, however, the effects of wetting can be avoided altogether.11,27 Thus we used coaxial microfluidic devices to study the dynamic interfacial tension of different surfactant systems both for simplification and to focus exclusively on the influence of liquid−liquid interfacial tension. The flowchart of the experiments and the microfluidic device used are shown in Figure 1. The device was fabricated on a
Q d = V × f (μL /min)
(1)
These calculated dispersed-phase flow rates varied from the flow rates set at the pump because of slight variations in the pump itself and any associated pressure drop in the microfluidic device. In this study, the actual calculated dispersed-phase flow rates were used in order to avoid any inaccuracies. Droplet data was collected 2 min after any changes to the continuous phase flow rate were made. Data collection lasted