Ordered Packing of Emulsion Droplets toward the Preparation of

May 1, 2014 - Monodisperse emulsion droplets with a high volume fraction form crystalline phases that can potentially serve as adjustable photomasks i...
1 downloads 0 Views 2MB Size
Download PDF
Article pubs.acs.org/Langmuir

Ordered Packing of Emulsion Droplets toward the Preparation of Adjustable Photomasks Ju Hyeon Kim,† Jae-Hoon Choi,‡ Jae Young Sim,† Woong Chan Jeong,‡ Seung-Man Yang,† and Shin-Hyun Kim*,† †

Department of Chemical and Biomolecular Engineering, KAIST, Daejeon, South Korea LG Chem. Ltd. CRD Center, Daejeon, South Korea

‡

S Supporting Information *

ABSTRACT: Monodisperse emulsion droplets with a high volume fraction form crystalline phases that can potentially serve as adjustable photomasks in photolithography. Such photomasks were prepared using a microfluidic device in which a flow-focusing junction, side channels, and a reservoir were connected in series. Transparent oil droplets were generated in a dye-containing continuous water phase at the flow-focusing junction. The droplets were then concentrated through the selective removal of the continuous phase using the side channels. This process led to the formation of a regular array of droplets in the reservoir with a configuration that depended on the relative height of the reservoir to the droplet diameter. The configurations could be selected among a single-layered hexagonal array, a bilayered square array, and a bilayered hexagonal array. The droplet arrays were used as a photomask to create hexagonal or square arrays of microdots. The transmittance profile of the ultraviolet (UV) light from each droplet was parabolic, which enabled the dot size to be tuned by controlling the UV irradiation time. This mask effect is otherwise difficult to achieve using conventional photomasks. The dot size and array periodicity could be adjusted by the in-situ control of the droplet size at the flow-focusing droplet maker. The combination of droplet size adjustments and the UV irradiation time provided independent control over the dot size and array periodicity to enable the preparation of a series of hexagonal microarrays with a wide spectrum of array parameters using a single microfluidic device.

1. INTRODUCTION Photolithography has been widely used to prepare micropatterns.1−3 The selective exposure of ultraviolet (UV) light onto a photoresist results in local cross-linking or degradation of the resist molecules, thereby providing patterns after the removal of the relatively small molecules. Selective UV exposure is performed through a photomask, which allows the transmission of light only through the transparent regions. Photolithography faces several technical limitations, in that only two-dimensional micropatterns may be obtained, with a minimum feature size that is limited by the wavelength of light. To overcome these limitations, self-assembled nanobuilding blocks, such as colloids4,5 and block copolymers,6 have been used as nanofabrication templates. Such novel approaches partially address the current limitations on photolithography. The low reconfigurability of a photomask pattern presents another limitation: one photomask can produce a single micropattern, and therefore, the number of photomasks required is equal to the number of micropatterns available. Conventional photomasks are static and allow only either the transmission or blocking of light. Grayscale photomasks have been developed using microfluidic devices, which enables the in-situ control over light transmittance;7 however, such devices have prefixed pattern shapes, and only line patterns can be © 2014 American Chemical Society

prepared. In addition, the intensity profile of the transmitted light is uniform throughout the microchannel, thereby preventing control over the line width. The development of a novel photomask that can provide a microarray with an adjustable pattern size and periodicity presents an important challenge in this field. This paper reports a microfluidic approach to preparing ordered arrays of emulsion droplets that can serve as flexible photomasks for the creation of microdot arrays with an adjustable dot size and periodicity. Monodisperse emulsion droplets produce regular structures or crystallites when concentrated. For example, small emulsion droplets confined in another drop form a droplet cluster with a unique configuration.8 Droplets confined in planar containers form hexagonal arrays at a high volume fraction. Multilayered structures may be prepared in face-centered cubic or hexagonal close-packed structures.9 The formation of these regular structures is driven by the minimization of the local interfacial energy.9 Microfluidic emulsification techniques enable the production of highly monodisperse droplets with precisely Received: February 25, 2014 Revised: April 28, 2014 Published: May 1, 2014 5404

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411

Langmuir

Article

controlled sizes.10 The self-assembly of monodisperse droplets is dynamic, and drop interfaces tend to be curved to minimize the interfacial energy. Such regular arrays of emulsion droplets can be implemented in microfluidic devices to provide adjustable photomasks. Droplet photomasks may be implemented by designing microfluidic devices composed of a series of three functional parts: a flow-focusing geometry for droplet generation, two side channels for droplet concentration, and a 2D reservoir for array formation. To prevent the movement of the photomasking droplets during UV exposure, we integrated a pneumatic valve on the device. This device was used to generate transparent oil droplets in the dye-containing continuous water phase. These droplets then formed densely packed structures through the autonomous removal of the continuous phase. The resulting arrangement of the droplets in the reservoir was determined by the height of the reservoir relative to the droplet size. A singlelayered hexagonal array, a bilayered square array, and a bilayered hexagonal array could be selected. The regular arrays of droplets acted as a photomask to provide a distinctive microdot array of negative photoresist. The dot size and array periodicity could be adjusted by controlling the flow rate, which determined the droplet size. The light transmitted through each droplet formed a parabolic intensity profile that provided an additional degree of freedom to control the dot size through the UV dose. The droplet-based microfluidic photomasks provided independent control over the dot size and array periodicity in a single device, which is potentially useful for the creation of a variety of distinct micropatterns in a wide spectrum of array parameters.

2. EXPERIMENTAL SECTION

Figure 1. (a) Schematic diagram of a microfluidic device composed of a drop maker, side channels, and a reservoir in series. The pneumatic valve was positioned between the side channels and the reservoir to regulate the droplet flow. The black part had a round top surface with a height of 15 μm, and the yellow part had a rectangular cross section with a height of 40 μm, as shown in the right inset. The design of the drop maker is shown in the left inset. Two side channels were designed to flow 31.4% of the total volume through them (15.7% through each channel), whereas 68.6% flowed through the main channel. The flow direction is denoted by the arrows. (b, c) Optical microscopy (OM) images collected in the region of the droplet maker and side channels (b) without or (c) with pneumatic valve actuation. Without the valve actuation, monodisperse oil droplets were generated in the droplet maker and flowed through the widening channel, where a portion of the continuous-phase aqueous dye solution flowed through the side channels. With valve actuation, both the dispersed and continuous phases flowed through the side channels. (d) OM image showing a hexagonal array of droplets in the reservoir.

Materials. Fluorocarbon oil (FC40, 3M) was used as a dispersed phase, and an aqueous solution of 5 wt % poly(vinyl alcohol) (PVA, Sigma-Aldrich) and 22.4 wt % Allura Red AC (Sigma-Aldrich) was used as a continuous phase. The PVA stabilized the fluorocarbon oil droplets and the Allura Red AC absorbed UV light during irradiation. The concentration of the dye was adjusted to its maximum solubility to efficiently block UV light. An adhesion layer (Omnicoat, Microchem) was spin-coated onto an oxygen plasma-treated cover glass (Fisherfinest), and a negative photoresist (SU-8 2 and SU-8 10, Microchem) was spin-coated onto the adhesion layer. SU-8 2 and SU8 10 have different viscosities and, therefore, provide different film thicknesses. The photoresist film was subjected to prebaking to evaporate the solvent prior to conducting the photolithography step. Preparation of the Microfluidic Device. The microfluidic device was composed of a flow channel and a valve channel, as illustrated in Figure 1a. The master molds for both channels were prepared using conventional photolithography. The flow channel was composed of a round channel part (denoted in black) and a rectangular channel part (denoted in yellow), the molds of which were prepared using a positive photoresist (AZ9260, Clariant) and a negative photoresist (SU-8), respectively. The negative photoresist was micropatterned to form a rectangular channel using photolithography. The positive photoresist was spin-coated onto the substrate and was micropatterned using an aligned photomask to form the round channel. The whole pattern was then subjected to thermal annealing at 125 °C for 20 min, leading to the selective reflow of the positive photoresist pattern and the creation of a round mold.11 After treating the pattern with hexamethyldisilazane (HMDS, Sigma-Aldrich) using vapor deposition, a mixture of the polydimethylsiloxane (PDMS, Sylgard 184, DowCorning) prepolymer and the curing agent in a 10:1 weight ratio was spin-coated over the pattern to form a thin PDMS film on top of the mold. The film was then thermally cured. The mold for the valve channel was created by micropatterning the negative photoresist and treating it with HMDS in the same manner. The valve channel was

molded using the PDMS, which was then mounted on top of the PDMS-coated flow mold and bonded at 75 °C for 30 min. Both channels were treated with oxygen plasma prior to assembly and were carefully aligned using an optical microscope. The whole PDMS structure was peeled away from the mold to create the flow channel and was mounted on a 20 μm thick PDMS film that had been precoated onto a plastic substrate after oxygen plasma treatment. 2[Methoxy(polyethylenoxy)propyl]trimethoxysilane (Gelest, Inc.) was then injected through the flow channel to render the inner surface hydrophilic, and the device was subjected to thermal treatment at 75 °C for 30 min. The whole PDMS microfluidic device was detached from the plastic substrate, and then part of the bottom surface was mounted onto a slide glass; the remaining part (reservoir) was mounted onto the negative photoresist-coated cover glass. Preparation of Emulsions. Monodisperse oil-in-water (O/W) emulsion droplets were generated by injecting a fluorocarbon oil through one inlet of a microfluidic device (denoted FC40 in Figure 1a) and an aqueous solution containing 5 wt % PVA and 22.4 wt % Allura Red AC through the other inlet (denoted by the dye solution). The fluorocarbon oil was emulsified into the aqueous solution at the cross-junction of the drop maker. Drop generation was observed using an inverted optical microscope (Eclipse TE 2000-U, Nikon) equipped with a high-speed camera (Motionscope M1, Redlake). The flow rates of the oil and aqueous solution were independently controlled using two syringe pumps (KD Scientific) and were typically set to be 50 and 70 μL h−1, respectively. The droplet size was controlled by adjusting the flow rate of the aqueous solution from 50 to 100 μL h−1. Photolithography and Characterization. After making droplet arrays, UV light was directed through the arrays using a 5× lens from a Ag lamp installed in an optical microscope. The micropatterns were developed using propylene glycol monomethyl ether acetate (PGMEA, Sigma-Aldrich) after postexposure baking (PEB). The 5405

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411

Langmuir

Article

height of 40 μm, comparable to the diameter of the spherical droplets. This shape transformation decreased the diameter of the spherical droplets relative to the size of the deformed droplets in the drop maker. In the reservoir, the droplets were aligned along two sidewalls, which guided the crystal orientations. A crystal formed in the middle of the reservoir, which produced polycrystalline and grain boundaries. The domain size was approximately 20 times the droplet diameter. When the width of the reservoir channel was smaller than the domain size, a single crystal was obtained within the channel. Fabrication of Microdot Arrays Using the Droplet Photomasks. A hexagonal array of droplets in the reservoir may be used as a photomask, as illustrated in Figure 2a. To prevent droplet movement during photolithography, the

fabricated micropatterns were analyzed using optical microscopy in the reflection mode (Eclipse LV 150, Nikon) and scanning electron microscopy (SEM, S-4800, Hitachi). The absorption coefficient of the dye was estimated by preparing a continuous phase dilution over factors of 250−25 000, and the absorbance at 365 nm was measured using UV−vis spectrometry (Optizen 3220UV, Mecasys).

3. RESULTS AND DISCUSSION Formation of the Droplet Arrays in the Microfluidic Device. The microfluidic device was designed to have a series of three functional parts in the flow channel, as shown in Figure 1a: a flow-focusing junction for droplet generation, two side channels for droplet concentration, and a reservoir for array formation. The drop maker and side channels had rounded top surfaces with a maximum height of 15 μm, whereas the reservoir had a rectangular cross section with a height of 40 μm. The pneumatic valve was located above the round channel, between the side channels and the reservoir, to prevent droplet movement during the photolithography UV irradiation step. The round shape of the channel ensured the complete closure of the flow channel without permitting leakage under the pressurization of the pneumatic valve.12,13 We injected a dispersed phase of FC40 and a continuous phase of the aqueous dye solution into the device to generate monodisperse oil-in-water (O/W) emulsion droplets at the flow-focusing droplet maker, as shown in Figure 1b. The arched microchannel with a maximum height of 15 μm deformed the droplets, which made the droplets appear polydisperse. The volumetric flow rates were typically set to 50 μL h−1 for the dispersed phase, Qd, and 70 μL h−1 for the continuous phase, Qc. The two side channels were connected to the main channel to selectively remove the continuous phase, leading to a higher volume fraction of droplets and facilitating the formation of a regular array. The width and length of the side channels were carefully determined to remove 31.4% of the total flow volume (15.7% in each side channel), which corresponded to a 53.8% flow volume in the continuous phase under typical flow rates. Therefore, 68.6% of the total volume flowed through the main channel (see Figure S1 of the Supporting Information for details14,15). This resulted in a droplet volume fraction of 0.607, which was appropriate for preparing a regular array of droplets while maintaining the curved droplet interfaces.16,17 The droplets deformed into a disk shape and flowed through the center of the widening channel with a round cross section after side channels. The continuous phase of aqueous dye solution dominantly flowed through two corners of the channel and gaps between the deformed droplets and the channel surfaces. This made the droplets appear more concentrated in the center of channel than 60.7%. The side channels provided an exit for both the dispersed and continuous phases when the pneumatic valve was actuated, as shown in Figure 1c. When the valve was closed, the pressure distribution in the drop maker changed, leading to the formation of polydisperse droplets. This prevented the accumulation of pressure in the flow channel, even under the continuous feeding of the fluids. The motions of the fluids at the droplet maker and side channels, either without or with valve actuation, are shown in the Movie S1 of the Supporting Information. The droplets finally arrived at the reservoir with a rectangular cross section and formed 2D hexagonal arrays, as shown in Figure 1d. Volume fraction of droplets was conserved as 60.7% in the reservoir from that in the widening channel immediately after two side channels. The droplets recovered their spherical shapes in the reservoir with a

Figure 2. (a) Schematic diagram showing the microfluidic droplet photomask, in which the transparent droplets allowed the transmission of UV light, and the dye-containing continuous phase adsorbed UV light. (b) Illustration showing the light path through the droplet (denoted by the violet-colored solid line), where two refraction events at the droplet interfaces and one refraction event at the water−PDMS interface were considered to obtain the path. (c) Transmittance profile of the light from three neighboring droplets, where the Beer−Lambert law was used to consider the adsorption at the continuous phase. (d, e) OM images of (d) a hexagonal array of droplets in the reservoir and (e) a dot array prepared by photolithography. (f, g) Scanning electron microscopy (SEM) images of the dot array: (f) Top view and (g) tilted view. 5406

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411

Langmuir

Article

Figure 3. (a, b) OM and SEM images of the micropatterns prepared using the three different dot diameters and a constant periodicity (prepared by controlling the UV irradiation time). The UV irradiation time was set to 5, 7, and 10 s (left to right). (c) The transmittance profile from the droplet photomask with a diameter of 47 μm and an interdroplet gap of 1 μm, where the red horizontal line was drawn to obtain the value of the critical UV dose with a dot diameter of 35 μm for an irradiation time of 5 s. The two blue horizontal lines were drawn to estimate the dot diameters for irradiation times of 7 and 10 s. (d) Irradiation time dependence of the dot diameter, where the squares denote the diameters measured from the SEM images, and the circles denote the diameters obtained from the transmittance profile shown in (c). The error bars denote the standard deviations, and the dotted line denotes the interdroplet distance.

pneumatic valve was completely closed by pressurizing the valve channel with approximately 400 kPa. All liquids injected flowed only through the two side channels, and no flow was permitted in the reservoir, thereby preventing the formation of a pressure gradient. Transparent droplets in the form of an array caused the selective transmission of UV light and created a hexagonal intensity profile. This enables the formation of a hexagonal dot array in a negative photoresist or a hole array in a positive photoresist. Here, we used the negative photoresist, SU-8. High definition was achieved by separating the droplets and photoresist by only a 20 μm thick PDMS film, where the PDMS film and the photoresist provided conformal contact that minimized scattering at the interface. The thinner PDMS membrane was too fragile to handle. Prior to UV irradiation, the pneumatic valve was actuated to prevent the movement of the droplet photomask. Under UV irradiation through the droplet arrays, the negative photoresist was selectively exposed to the light. The profile of the UV dose was estimated by calculating the transmittance of light through each droplet using simple ray tracing techniques, where we assumed that the absorption of light occurred only at the continuous phase volume and followed the Beer−Lambert law:

T=

I = 10−εLC I0

(1)

where I0 is the initial intensity of light, I is the intensity of the transmitted light, ε is the extinction coefficient, L is the length of the light path, and c is the concentration of the dye solution (0.451 M). The value of ε at a concentration of 0.451 M was estimated as 1265.8 M−1 cm−1 based on the absorbance of the diluted solutions, as shown in Figure S2 of the Supporting Information. The light path was obtained by considering refraction at the droplet interfaces and at the water−PDMS interface at which the refractive indices of the dispersed and continuous phases were 1.290 and 1.341, respectively, and that of the PDMS was 1.400. An incident light ray with an angle β was refracted at the first interface of the droplet through an angle α, according to Snell’s law. The light ray then entered the second interface with an incident angle α and was refracted to an angle β, as illustrated in Figure 2b. The light ray was refracted again at the water−PDMS interface. The violetcolored solid line indicates the light path. These estimates were used to express the length of the light path at the continuous phase as L(b) = (R − a(b)) + x(b), and the distance from the center of the drop to the ray arrival point at the surface of the photoresist could be expressed as P(b) = b + l(b) + k(b). All other parameters are denoted in Figure 2b. The calculations are 5407

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411

Langmuir

Article

Figure 4. (a, b) Three illustrations and OM images, showing droplets confined in the reservoirs. The first two droplets were deformed into disk shapes because the droplet size exceeded the height of the reservoir. The last droplets remained spherical due to their smaller size. These droplets having three different diameters were prepared by controlling the flow rate of the continuous phase, Qc, between 60, 70, and 100 μL h−1, from left to right, while maintaining the flow rate of the dispersed phase, Qd, at a constant value of 50 μL h−1. The coefficients of variation of droplet sizes were 1.70%, 2.36%, and 1.15%, respectively. (c) OM images of the microdot arrays prepared from the photomasks in (b). (d) Transmittance profiles from the three different droplet sizes. The horizontal line indicated the transmittance required to cross-link the photoresist under the experimental irradiation conditions. The intersections between the line and the curves provided three different dot diameters of 34.1, 31.3, and 26.2 μm, respectively. (e) Flow rate (continuous phase, Qc) dependence of the droplet diameter (denoted by the red triangles) and the interdot microarray distance (denoted by the cyan inverted triangles), where Qd was maintained at 50 μL h−1. The dot diameters measured from (c) and obtained from (d) are also included in the plot with squares and circles, respectively, in good agreement.

explained in detail in the Supporting Information. A combination of the light path and the Beer−Lambert law enabled the calculation of the transmittance profile at the surface of the photoresist film. For example, the transmittance profile, for droplets that were slightly flattened by a channel height, H, of 40 μm to a disk diameter of 44 μm, is shown in in Figure 2c, where we assume a 0.5 μm gap between the flattened part of the drop and the PDMS surface and a 1 μm gap between droplets. Here, three neighboring droplets were included in the calculation. A small rectangular profile appeared over the main parabolic profile due to the flattened regions on the top and bottom of the droplet. Although the refractive index of the droplets was smaller than that of the continuous phase, which defocused the light in a manner similar to that of a concave lens, the transmittance followed a parabolic profile because the light path at the continuous phase increased as the

position deviated from the center of the droplet. We produced hexagonal dot arrays of the negative photoresist using droplet arrays as photomasks. Hexagonal array of droplets was flattened similarly to those shown in the illustration in Figure 2b, as shown in Figure 2d. This droplet array was used as a photomask, through which we irradiated the 5 μm negative photoresist (SU-8 2) using a 5× lens and 4 s of UV light from a Ag lamp mounted on a microscope. The intensity of the UV light, I0, was found to be 42.44 mW/cm2. After development, we obtained a hexagonal dot array, as shown in Figures 2e−g. The top surfaces of the dots were slightly curved, as shown in Figure 2g. This was attributed to the parabolic profile of the UV dose. The average diameter of the dot was 31 μm, and its coefficient of variation was 3.71%. The size of the dot and the transmittance profile were used to estimate the critical UV dose, Dc, expected to lead to the polymerization of the SU-8 2, 5408

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411

Langmuir

Article

42.78 mW. Dc = I0 × Tc × t, where Tc is the transmittance a distance from the center of the droplet equal to the dot radius (Tc = 0.252 at a distance of 15.65 μm) and t is the irradiation time (t = 4 s). Control over the Dot Size and Array Periodicity. The parabolic profile of the UV transmittance from each droplet enabled the dot diameter to be controlled via the UV irradiation time. This was difficult to achieve using conventional photomasks with rectangular profiles. Longer UV irradiation times cross-linked the negative photoresist in the region with a low transmittance. Therefore, as the irradiation time increased, so did the diameter of the microdots, while preserving the periodicity of the dot array. This effect was clearly shown in the OM and SEM images shown in Figures 3a and 3b, in which a photomask composed of a droplet with a diameter of 47 μm and an interdroplet gap of 1 μm was employed. The SU-8 10 negative photoresist film 10 μm thick was used. As the irradiation time increased from 5 to 7 s and then to 10 s, the diameters of the fabricated dots increased from 35 to 40 μm and then to 48 μm. The longest exposure time (10 s) produced partial connections between the microdots. The changes in the dot diameter could be quantitatively explained according to the transmittance profile, as shown in Figure 3c. As the irradiation time, t, increased, the transmittance required for cross-linking decreased according to the equation of Tc = Dc/(I0 × t). The value of Dc could be estimated based on data collected from the 35 μm dots for t = 5 s, as denoted by the red horizontal line in the transmittance plot shown in Figure 3c. This provided a value of Dc = 52.63 mW. This value of Dc was used to determine two additional blue horizontal lines, corresponding to t = 7 and 10 s, as shown in the plot. This provided a 40 μm diameter for t = 7 s, as indicated by the intersection between the horizontal lines and the transmittance curve. For t = 10 s, the horizontal line was placed immediately below the minima of the transmittance curve, indicating partial connections between the dots, as confirmed in the SEM image shown in the last panel of Figure 3b. We treated the dot diameter under these overexposure conditions as 48 μm. The diameters measured from the SEM images and obtained from the transmittance profile agreed well, as shown in Figure 3d. The irradiation time influenced the dot diameter, although the periodicity of the dot arrays did not change significantly. The periodicity could be adjusted by controlling the droplet diameter. The flow-focusing droplet maker was used to generate an emulsion droplet in a dripping mode, in which the continuous phase exerted a drag force on the droplet hanging over the short jet of the dispersed phase. An increase in the flow rate of the continuous phase would, therefore, decrease the droplet size. To demonstrate the on-demand control over the droplet size, the volumetric flow rate of the continuous phase, Qc, was varied from 60 to 70 μL h−1 and then to 100 μL h−1, while the flow rate of the dispersed phase, Qd, remained fixed at 50 μL h−1. The variations in Qc produced three different droplet diameters, d, of 47, 44, and 37 μm, as shown in Figures 4a and 4b. Two of the droplets were larger than the height of the reservoir, H = 40 μm, and were deformed into a disk shape (H < d), whereas one of the droplets was smaller and maintained its spherical shape (H > d), as illustrated in Figure 4a. The diameters of the two large droplets were measured in the deformed state. Hexagonal arrays of three different-sized droplets were used as photomasks to create microdot arrays, as shown in Figure 4c. We used a 5 μm thick SU-8 2 negative photoresist film and irradiated the photoresist

for 4 s. The periodicity of the dot array was varied from 46.7 to 43.6 μm and then to 37.4 μm, nearly identical to the corresponding droplet diameter. At the same time, the dot diameter was varied from 34.0 to 31.3 μm and then to 26.6 μm. These changes may also be understood in terms of the transmittance profiles of the three different droplet sizes, as shown in Figure 4d. The profiles obtained from the two large droplets had a small rectangular profile due to the flattened regions on the top and bottom, whereas the profile obtained from the small droplets was not rectangular. The value of Dc of SU-8 2, obtained in Figure 2, was used to draw a single horizontal line into the transmittance plots and provide three different dot diameters of 34.1, 31.3, and 26.2 μm based on the intersections. The dot diameters measured from the OM images in Figure 4c were consistent with the diameters obtained from the transmittance profiles, as shown in Figure 4e. The diameters of the droplets and the interdot distance (or periodicity) were in good agreement. Because the dot diameter may be further controlled using the UV irradiation time, as shown in Figure 3, the dot diameter and periodicity could be independently controlled through a combination of on-demand control over the droplet diameter and irradiation time. Bilayered Droplet Photomasks. The micropattern type was diversified by employing bilayered droplet arrays as photomasks. Monodisperse droplets can form bilayered arrays with structures that are determined by the height of the reservoir relative to the droplet diameter. For a height-todroplet diameter ratio, H/d, of 1 + (1/2)1/2 = 1.707, the droplets are arranged into a square array, in which the droplets in the top layer sit on the interstices between the droplets in the bottom layer, as shown in Figure 5a.16 For a ratio of 1 + (2/3)1/2 = 1.816, the droplets are arranged in a hexagonal array in a manner similar to that of the bilayered square array, as shown in Figure 5b.16 The bilayered arrays can provide unique micropatterns that are useful as photomasks. The bilayered arrays allow the penetration of UV light only through the vertically overlaid regions of the transparent droplets, as denoted by the violet color in Figures 5a and 5b. The bilayered droplet arrays may be implemented by designing a reservoir with a height of 80 μm, and droplets were generated with two different diameters of 47 and 44 μm. The droplets with diameters of 47 μm, yielding an H/d ratio of 1.702, formed a square array, as shown in Figure 5c. This droplet photomask could prepare a special array of squares, in which four dots formed each square, as shown in Figure 5d. The droplets with diameters of 44 μm, with an H/d ratio of 1.818, formed a hexagonal array, as shown in Figure 5e. We prepared a hexagonal array of triangles in which three dots formed each triangle, as shown in Figure 5f. Patterning at Two Different Scales Using Parallelized Microfluidic Channels. The reservoir with a wide 2D space was replaced with parallelized microfluidic channels to produce droplet arrays. Droplets formed a hexagonal array without grain boundaries in each channel as the width of the channel was much smaller than crystal domain size.18,19 This enabled patterning on two distinct scales and shapes: the line pattern was a few hundreds of micrometers in scale, whereas the hexagonal dot pattern was a few tens of micrometers in scale. This structure was demonstrated by generating droplets with diameters of 44 μm in the drop makers. These droplets then flowed through the parallelized channels of height 40 μm, width 200 μm, and with an interchannel gap of 90 μm. The channel design is shown in Figure 6a. The droplets formed a hexagonal 5409

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411

Langmuir

Article

by applying photolithography, as shown in Figure 6c. Hexagonal arrays of droplets were transferred into a negative photoresist as an array of dots. Because PDMS channels are transparent, the photoresist film between the lines remained after development. The microchannels could be further designed to have a complex shape that can potentially provide useful micropatterns for advanced applications.

4. CONCLUSIONS We developed novel microfluidic photomask systems using ordered arrays of emulsion droplets. A microfluidic device composed of a droplet maker, side channels, and a reservoir in series yielded crystalline-phase transparent emulsion droplets in an opaque continuous phase. The droplet arrays were used as photomasks to create micropatterns comprising hexagonal arrays of dots. The dot size and periodicity of the array could be independently controlled through a combination of in-situ control over the droplet size and UV irradiation time in a single microfluidic device. The micropatterns could be diversified by employing either bilayered arrays of droplets or parallelized microchannels. The enhanced reconfigurability of this microfluidic photomask system provides new opportunities for creating a series of microdot patterns with several distinct dot sizes and array periodicities. The system may be further developed to achieve advanced photomask performances. For example, the volume fraction of droplets may be increased to prepare an array of hexagonal disks, and the refractive indices of the droplet and continuous fluids may be inverted to provide light-focusing effects similar to those obtained using a convex lens array.

Figure 5. (a, b) Schematic diagram showing the bilayered droplet arrays: (a) The square array and (b) hexagonal arrays, in which the first panels showed the top views and the second panels showed the side view. The light only penetrated the vertically overlaid regions of the transparent droplets, as indicated by the violet color. (c, d) OM images of the square array of droplets 47 μm in diameter, confined within a reservoir of height 80 μm, and a micropattern prepared using the droplet array as a photomask. (e, f) OM images of a hexagonal array of droplets prepared with a diameter of 44 μm confined in the same reservoir and micropattern.

■

ASSOCIATED CONTENT

S Supporting Information *

array in each channel such that the direction of the shortest interdroplet distance was aligned along the channel wall, as shown in Figure 6b. The array of droplets prepared in the parallelized channels was then used to prepare a micropattern

Detailed calculations of the channel resistance and transmission profile and estimations of the absorption coefficient; a movie clip showing the fluid motion without or with valve actuation.

Figure 6. (a) Schematic diagram showing a microfluidic device containing parallelized channels, instead of wide reservoir. (b) OM image showing a hexagonal array of droplets in each parallelized channel. (c) OM image of micropattern prepared using the droplet photomask shown in (b). 5410

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411

Langmuir

Article

(15) Beebe, D. J.; Mensing, G. A.; Walker, G. M. Physics and applications of microfluidics in biology. Annu. Rev. Biomed. Eng. 2002, 4, 261−286. (16) Hatch, A. C.; Fisher, J. S.; Pentoney, S. L.; Yang, D. L.; Lee, A. P. Tunable 3D droplet self-assembly for ultra-high-density digital microreactor arrays. Lab Chip 2011, 11 (15), 2509−2517. (17) Priest, C.; Herminghaus, S.; Seemann, R. Generation of monodisperse gel emulsions in a microfluidic device. Appl. Phys. Lett. 2006, 88 (2), 024106. (18) Claussen, O.; Herminghaus, S.; Brinkmann, M. Packings of monodisperse emulsions in flat microfluidic channels. Phys. Rev. E 2012, 85 (6), 061403. (19) Surenjav, E.; Priest, C.; Herminghaus, S.; Seemann, R. Manipulation of gel emulsions by variable microchannel geometry. Lab Chip 2009, 9 (2), 325−330.

This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected] (S.-H.K.). Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by KAIST VRPGP (Project No. N01130054) and a grant from the Creative Research Initiative Program of the Ministry of Science and Technology for “Complementary Hybridization of Optical and Fluidic Devices for Integrated Optofluidic Systems”.

■

DEDICATION Professor Yang passed away unexpectedly on September 26, 2013. We dedicate this work as a memorial to him.

■

REFERENCES

(1) Vilkner, T.; Janasek, D.; Manz, A. Micro total analysis systems. Recent developments. Anal. Chem. 2004, 76 (12), 3373−3386. (2) Gates, B. D.; Xu, Q. B.; Stewart, M.; Ryan, D.; Willson, C. G.; Whitesides, G. M. New approaches to nanofabrication: Molding, printing, and other techniques. Chem. Rev. 2005, 105 (4), 1171−1196. (3) Levenson, M. D.; Viswanathan, N. S.; Simpson, R. A. Improving resolution in photolithography with a phase-shifting mask. IEEE Trans. Electron Devices 1982, 29 (12), 1828−1836. (4) Yang, S.-M.; Jang, S. G.; Choi, D.-G.; Kim, S.; Yu, H. K. Nanomachining by colloidal lithography. Small 2006, 2 (4), 458−475. (5) Haynes, C. L.; McFarland, A. D.; Smith, M. T.; Hulteen, J. C.; Van Duyne, R. P. Angle-resolved nanosphere lithography: Manipulation of nanoparticle size, shape, and interparticle spacing. J. Phys. Chem. B 2002, 106 (8), 1898−1902. (6) Ruiz, R.; Kang, H.; Detcheverry, F. A.; Dobisz, E.; Kercher, D. S.; Albrecht, T. R.; de Pablo, J. J.; Nealey, P. F. Density multiplication and improved lithography by directed block copolymer assembly. Science 2008, 321 (5891), 936−939. (7) Chen, C.; Hirdes, D.; Folch, A. Gray-scale photolithography using microfluidic photomasks. Proc. Natl. Acad. Sci. U. S. A. 2003, 100 (4), 1499−1504. (8) Kim, S.-H.; Hwang, H.; Lim, C. H.; Shim, J. W.; Yang, S.-M. Packing of emulsion droplets: Structural and functional motifs for multi-cored microcapsules. Adv. Funct. Mater. 2011, 21 (9), 1608− 1615. (9) Shui, L.; Kooij, E. S.; Wijnperle, D.; van den Berg, A.; Eijkel, J. C. T. Liquid crystallography: 3D microdroplet arrangements using microfluidics. Soft Matter 2009, 5 (14), 2708−2712. (10) Yobas, L.; Martens, S.; Ong, W.-L.; Ranganathan, N. Highperformance flow-focusing geometry for spontaneous generation of monodispersed droplets. Lab Chip 2006, 6 (8), 1073−1079. (11) He, M.; Bu, J.; Ong, B. H.; Yuan, X. Two-microlens coupling scheme with revolved hyperboloid sol-gel microlens arrays for highpower-efficiency optical coupling. J. Lightwave Technol. 2006, 24 (7), 2940−2945. (12) Unger, M. A.; Chou, H.-P.; Thorsen, T.; Scherer, A.; Quake, S. R. Monolithic microfabricated valves and pumps by multilayer soft lithography. Science 2000, 288 (5463), 113−116. (13) Studer, V.; Hang, G.; Pandolfi, A.; Ortiz, M.; Anderson, W. F.; Quake, S. R. Scaling properties of a low-actuation pressure microfluidic valve. J. Appl. Phys. 2004, 95 (1), 393−398. (14) Choi, S.; Lee, M. G.; Park, J.-K. Microfluidic parallel circuit for measurement of hydraulic resistance. Biomicrofluidics 2010, 4 (3), 034110. 5411

dx.doi.org/10.1021/la5007692 | Langmuir 2014, 30, 5404−5411