sensors Article
Fabrication of Glass Microchannel via Glass Imprinting using a Vitreous Carbon Stamp for Flow Focusing Droplet Generator Hyungjun Jang 1 ID , Muhammad Refatul Haq 1 , Youngkyu Kim 1 , Jun Kim 1 , Pyoung-hwa Oh 1 , Jonghyun Ju 1 , Seok-Min Kim 1, * and Jiseok Lim 2, * 1
2
*
School of Mechanical Engineering, College of Engineering, Chung-Ang University, 84 Heukseok-ro, Dongjak-gu, Seoul 06974, Korea;
[email protected] (H.J.);
[email protected] (M.R.H.);
[email protected] (Y.K.);
[email protected] (J.K.);
[email protected] (P.-h.O.);
[email protected] (J.J.) School of Mechanical Engineering, Yeungnam University, 280 Daehak-Ro, Gyeongsan, Gyeongbuk 38541, Korea Correspondence:
[email protected] (S.-M.K.);
[email protected] (J.L.); Tel.: +82-2-820-5877 (S.-M.K.); +82-53-810-2577 (J.L.)
Received: 28 October 2017; Accepted: 27 December 2017; Published: 29 December 2017
Abstract: This study reports a cost-effective method of replicating glass microfluidic chips using a vitreous carbon (VC) stamp. A glass replica with the required microfluidic microstructures was synthesized without etching. The replication method uses a VC stamp fabricated by combining thermal replication using a furan-based, thermally-curable polymer with carbonization. To test the feasibility of this method, a flow focusing droplet generator with flow-focusing and channel widths of 50 µm and 100 µm, respectively, was successfully fabricated in a soda-lime glass substrate. Deviation between the geometries of the initial shape and the vitreous carbon mold occurred because of shrinkage during the carbonization process, however this effect could be predicted and compensated for. Finally, the monodispersity of the droplets generated by the fabricated microfluidic device was evaluated. Keywords: autofluorescence; glass imprinting process; vitreous carbon mold; glass microchannel; droplet based microfluidic
1. Introduction Microfluidic technology is a rapidly developing field of research that focuses on the manipulation of small quantities of fluids. It is a powerful tool for applications that require quantitative, high-throughput analysis in biochemistry, material science, and fluid physics [1–5]. The fabrication technologies used to make microfluidic devices are key to practical applications in this field. One type of polymer microfluidic device is a poly-dimethylsiloxane (PDMS) part fabricated via soft lithography [6,7]. This device enables a relatively simple procedure for the replication and sealing of chips. In addition to the advantages of a convenient fabrication process, PDMS has high gas permeability and optical transparency, which are beneficial in a variety of biological applications. However, such elastomer-based polymer microfluidic devices have limitations [8–11] in many applications that require mechanical, thermal, and chemical stability. Another drawback is the limited applicability of PDMS in high-sensitivity fluorescence-based read-out applications, due to its high autofluorescence level [12]. Glass microfluidic devices may be an alternative for applications that require physical or chemical characteristics which are not offered by PDMS devices. Glass microfluidic devices provide exceptional chemical resistance, biocompatibility and optical properties, as well as mechanical stability, which prevents swelling and deformation [13–16]. In the realization of
Sensors 2018, 18, 83; doi:10.3390/s18010083
www.mdpi.com/journal/sensors
Sensors 2018, 18, 83
2 of 9
glass microfluidic chips, the pattering of the microfluidic channel is the most important issue. Selective material removal processes based on conventional semiconductor manufacturing techniques including photolithography and wet- or dry-etching are typically used to fabricate the glass microfluidic channel. Even with the advantages of glass, their high cost of fabrication limits their use, especially for disposable devices [17–19]. Therefore, considerable attention is focused in developing cost-effective fabrication solutions for glass microfluidic channels. A glass imprinting process for micro/nano replication was recently studied using commercially available glasses such as quartz, Pyrex, and soda-lime glass [20,21]. This may be a promising approach to low-cost glass microchannel fabrication. The key issue in the mass production of glass microfluidic channels is the cost of the mold fabrication method, which must support small structures dispersed over a large area. This is because multiple molds are required in progressive glass imprinting systems which can provide a high production rate [22]. Since the mold used in the glass imprinting process must have superior thermal resistance, sufficient mechanical strength and good release properties, the selection of mold materials is limited. Some materials such as tungsten carbide, silicon carbide, nickel alloy, and glassy carbon [23–26] have been proposed for use in molds. Molds must be microfabricated at high resolution. However, the micro-texturing methods available for these materials are not suitable for large area applications because of their high processing costs and low throughput [27]. In this study, we propose a vitreous carbon (VC) mold, made via a combination of replication and carbonization, for the fabrication of glass microfluidic channels, using a glass imprinting process. In our previous research, we demonstrated large area glass imprinting at micro and nano scale resolutions using VC molds [28,29]. Since texturing of the VC mold is conducted via thermal replication, a large area mold compatible with glass imprinting can be fabricated at a low cost. Finally, the initial autofluorescnece of the fabricated micro channel was evaluated. To verify the feasibility of the proposed solution, a flow-focusing droplet generator was fabricated using a soda-lime glass microchannel plate and PDMS top plate. 2. Materials and Methods 2.1. Fabrication of A Vitreous Carbon Stamp The VC mold was fabricated via the carbonization of a Furan-based replica, which was itself fabricated via a series of replication steps using PDMS and a furan-based thermal curable polymer. Since the demolding properties of the furan-based resin are extremely poor, it is not feasible to replicate a silicon wafer master pattern directly. The first replication step generated an elastomeric intermediate mold made from PDMS. PDMS has superior mechanical and surface properties, and can transfer a pattern to a Furan-based resin with high fidelity. The intermediate PDMS mold was fabricated via a conventional soft-lithography process using 10 parts of Sylgard 184 (Dow Corning Korea Ltd., Seoul, Korea) and 1 part of a curing agent. The master pattern of the microfluidic structure was fabricated on a 4 inch silicon wafer using a photolithography process that employed SU-8 3050 (MicroChem Co., Westborough, MA, USA) as a photoresist. The height of the master pattern was 40 µm. The second replication step was performed using a mixture that included a furan resin (Kangnam Chemical Co. Ltd., Gwangju, Gyeonggi, Korea), p-TSA (p-Toluenesulfonic acid), and ethanol. The furan mixture was poured onto the PDMS mold. Before solidification, the mixture was degassed to remove air bubbles created during mixing. In the first curing process, the mixture was allowed to polymerize naturally over 5 d under atmospheric conditions. Next, thermal curing was performed in a conduction oven at up to 100 ◦ C for 2 d. The furan precursor synthesized using the aforementioned method was carbonized in a furnace at 1000 ◦ C under N2 for 5 d. During this process, pyrolysis phenomena caused shrinkage of the furan precursor to occur. The pyrolysis is when all the molecules in the furan precursor, except carbon, are thermally decomposed and only carbon remains. Molecules which are thermally decomposed in this way include hydrogen and oxygen. The pyrolysis process results in a mass reduction
Sensors 2018, 18, 83 Sensors 2018, 18, 83
3 of 9 3 of 9
volumetric shrinkage. Figure 1a–d shows a schematic diagram of the proposed VC stamp fabrication and a volumetric shrinkage. Figure 1a–da shows a schematic diagram of the VC proposed VC stamp volumetric shrinkage. Figure 1a–d shows schematic diagram of the proposed stamp fabrication method. fabrication method. method.
Figure 1.1.Proposed fabrication method for thefor glass device. PDMS is polyProposed fabrication method themicrofluidic glass microfluidic device. PDMS is polyFigure 1. Proposed fabrication method for the glass microfluidic device. PDMS is polydimethylsiloxane; VC is vitreous carbon. dimethylsiloxane; VC is vitreous carbon.
2.2. The Glass Imprinting Process 2.2. The Glass Imprinting Process evaluate the theprocess processofofimprinting imprinting glass using a VC stamp, a high temperature thermal To evaluate glass using a VC stamp, a high temperature thermal press To evaluate the imprinting using a VC stamp, a high temperature thermal press press system consisting anof infrared (IR) glass heater, a motor-driven pressure module, a controller system consisting of process an of infrared (IR) heater, a motor-driven pressure module, and aand controller was ◦ system consisting of an infrared (IR) heater, a motor-driven pressure module, and a controller was developed. system allows precise temperature and pressure controlupuptoto1050 1050°CCand and900 900was N, developed. The The system allows precise temperature and pressure control N, developed. The system allows precise temperature and pressure control up to 1050 °C and 900 N, respectively. Figure respectively. Figure 2a 2a shows shows the the imprinting imprinting system. system. respectively. 2a shows the was imprinting system. glassFigure imprinting process wasperformed performed using system. A soda-lime substrate The glass imprinting process using thisthis system. A soda-lime glassglass substrate with The glass imprinting process was performed using this system. A soda-lime glass substrate with with a thickness of 3 was mm used was used asimprinting the imprinting material. 1e–f show the schematic a thickness of 3 mm as the material. FigureFigure 1e–f show the schematic flow a thickness of 3 mm was used as the imprinting material. Figure 1e–f show the schematic flow diagram of glass the glass imprinting process. First, mold wasinstalled installedininthe themiddle middle of offlow the diagram of the imprinting process. First, thethe VCVC mold was diagram of the glass imprinting process. First, the VC mold was installed in the middle of the imprinting system and the glass substrate was placed so as to cover the VC mold. To prevent the imprinting system and the glass substrate was placed so as to cover the VC mold. To prevent the oxidization of the VC mold during the process, the chamber was maintained in an inert environment oxidization the VC mold during process,were the chamber was in antemperature inert environment The mold andthe substrate heated up tomaintained with a via nitrogenof purging. the imprinting ◦ via nitrogen purging. The mold and substrate were heated up to the imprinting temperature with a heating rate of 70 °C/min C/minand andthe thetemperature temperature was was maintained maintained for for 10 10 min min in in order order to obtain uniform heating rate of 70 °C/min and the temperature was maintained for 10 min in order to obtain uniform temperature distribution. distribution. After After the the holding holding time, time, aa compression compression pressure pressure was was applied applied for for 20 20 min. min. temperature After thethe holding time, a compression pressure was applied for 20 min. After cooling distribution. to room temperature, temperature, the glass replica was removed from the VC mold. glass replica was removed from the VC mold. After cooling to room temperature, the glass replica was removed from the VC mold.
Figure 2. (a) Photograph of the glass imprinting system and (b) the temperature and pressure Figure Photograph of the glass imprinting systemsystem and (b)and the temperature and pressure Figure2.2.(a)(a) Photograph the glass imprinting (b) the temperature andconditions pressure conditions during the glass of imprinting process. during the glass imprinting process. conditions during the glass imprinting process.
Sensors 2018, 18, 83 Sensors 2018, 18, 83
4 of 9 4 of 9
To optimum imprinting imprintingcondition, condition,the theeffects effectsofof imprinting temperature To obtain obtain the optimum imprinting (a)(a) temperature andand (b) (b) pressure height of microchannel at the orifice and channel regions were analyzed shown pressure on on thethe height of microchannel at the orifice and channel regions were analyzed as as shown in in Figure clearlyshows showsthat thatthe theheights heightsof ofthe the imprinted imprinted glass glass microchannels microchannels increased as the Figure 3. 3. It It clearly the temperature temperatureand andpressure pressureincreased, increased,and andthat that the the measured measured heights heights reached reached appropriate appropriate values values when when ◦ Cand the higher than than680 680°C andthe thepressure pressurewas was higher than 163.2 kPa. Therefore, the temperature was higher higher than 163.2 kPa. Therefore, we ◦ we selected imprinting temperature 680°C Cand anda amaximum maximumimprinting imprintingpressure pressure of of 163.2 kPa as selected an an imprinting temperature ofof 680 as the the optimal optimal conditions. conditions.
Figure Theheight heightofofthe theorifice orifice and channel parts of imprinted microfluidic in Figure 3. 3. The and channel parts of imprinted glassglass microfluidic chipschips in with with relation (a) temperature at the same pressure (163.2 kPa) and (b)atpressure the same relation to (a) to temperature at the same pressure (163.2 kPa) and (b) pressure the same at temperature temperature (680 ◦ C). (680 °C).
2.3. Sealing Sealing The The Glass-Based, Glass-Based, Droplet-Generating Droplet-Generating Microfluidic MicrofluidicChip Chip 2.3. Toexamine examinethe thefeasibility feasibilityof of using using the the glass glass replica replica for for microfluidic microfluidic applications, applications, the the glass-based glass-based To microfluidic pattern with a flat blockblock via an via O2 plasma microfluidic patternwas wassealed sealed with a PDMS flat PDMS an O2 bonding plasma process. bondingAlthough process. sealing was performed with a PDMS block insteadblock of theinstead glass substrate, the purpose ofthe thispurpose test was Although sealing was performed with a PDMS of the glass substrate, to this verify the fidelity of the glass imprinting process. Conventional glass-to-glass bonding solutions of test was to verify the fidelity of the glass imprinting process. Conventional glass-to-glass such as thermal fusion [30] fusion and anodic bonding available. Fabrication all-glass bonding solutions suchbonding as thermal bonding [30] [31] and are anodic bonding [31] areofavailable. microfluidics may be achieved by combining the glass the aforementioned Fabrication of all-glass microfluidics may be achieved by imprinting combining process the glasswith imprinting process with glass-to-glass bonding methods. bonding We made holes for outlets on the the aforementioned glass-to-glass methods. Weinlets madeand holes for inlets andPDMS outlets block on theusing PDMSa biopsyusing punching toolpunching with a diameter ofa1.5 mm. Subsequently, a glass replica and the PDMS block block a biopsy tool with diameter of 1.5 mm. Subsequently, a glass replica and the were bonded via O 2 plasma treatment at treatment 18 W for 30 PDMS block were bonded viasurface O2 plasma surface ats.18 W for 30 s. 3. 3. Results Results To To verify verify the the fidelity fidelity of of the the glass glass imprinting imprinting process, process, the the samples samples were measured at each step. That Thatis, is, (a) (a) the the silicon silicon master, master, (b) (b) the the intermediate intermediate PDMS PDMS mold, mold, (c) (c) the the furan furan replica, replica, (d) the VC stamp, and and (e) (e) the the replicated replicated glass glass microfluidic microfluidic structure structure were were each eachmeasured. measured. Figure Figure 44 shows shows photographs photographs and scanning electron microscopy (SEM) images of samples from each step. For quantitative evaluation, and scanning electron microscopy (SEM) images of samples from each step. For quantitative measurements were taken were usingtaken a confocal (OLS-4000, OlympusOlympus Co., Tokyo, evaluation, measurements using amicroscope confocal microscope (OLS-4000, Co., Japan). Tokyo, Figure show 3D surface profilesprofiles of (a) the master,master, (b) the (b) PDMS mold, (c) the (c) furan Japan).5a–e Figure 5a–ethe show the 3D surface of silicon (a) the silicon the PDMS mold, the precursor, (d) the VC stamp and (e) the replicated glass microfluidic structure obtained by the confocal furan precursor, (d) the VC stamp and (e) the replicated glass microfluidic structure obtained by the microscope measurement results. Figure 5f–g show5f–g the comparison of the surface profiles measured confocal microscope measurement results. Figure show the comparison of the surface profiles at (f) the channel and (g) the orifice. As shown in Figure 5f–g, the difference between the surface measured at (f) the channel and (g) the orifice. As shown in Figure 5f–g, the difference between the morphologies of the silicon master Furanand replica is negligible. However, considerable shrinkage surface morphologies of the siliconand master Furan replica is negligible. However, considerable occurred the carbonization process. To analyze characteristic and repeatability shrinkageduring occurred during the carbonization process.the To shrinkage analyze the shrinkage characteristic and of the proposed glass imprinting process with a VC mold, the geometrical properties of three PMDS repeatability of the proposed glass imprinting process with a VC mold, the geometrical properties of molds, Furanmolds, precursors, VC molds a single silicon master, as well as 9asglass three PMDS Furanand precursors, andfabricated VC moldsfrom fabricated from a single silicon master, well replicas the three fabricated VC molds, analyzed as summarized in Table The1.total as 9 glassfrom replicas from the three fabricated VCwere molds, were analyzed as summarized in 1. Table The shrinkage ratios of the VC were 30.43% for orifice width, 30.17% for channel width, and 30.33% total shrinkage ratios of the VC were 30.43% for orifice width, 30.17% for channel width, and 30.33%
for height. It is clear that the almost isotropic shrinkage occurred in the VC mold fabrication process. The percent coefficient of variance of measured values in the VC mold were less than 1%. This means
Sensors 2018, 18, 83
5 of 9
for height. It is clear that the almost isotropic shrinkage occurred in the VC mold fabrication process. The percent coefficient of variance of measured values in the VC mold were less than 1%. This means that the repeatability error of the proposed VC mold fabrication is under 1%, which is acceptable Sensors 2018, 18, 83 5 of 9 accuracy in microchannel applications. The dimensional difference between the VC molds and glass replicas was negligible because of the similar thermal expansion ratio of VC and glass materials. that the repeatability error of the proposed VC mold fabrication is under 1%, which is acceptable In addition, surface roughness of theThe VCdimensional mold mustdifference be assessed to determine that it would accuracythe in microchannel applications. between the VC molds and glass be acceptable in microfluidic applications. Table 2 shows the channel wall surface roughness as measured replicas was negligible because of the similar thermal expansion ratio of VC and glass materials. In in each step. Five different positions measured on be each sample, the results averaged. addition, the surface roughness of were the VC mold must assessed to and determine that itwere would be acceptable in that microfluidic applications. Table 2 shows the channel roughness and as the The results show the deterioration of the surface roughness duringwall VC surface mold fabrication in each step.was Fivenegligible. different positions were measured on each sample, and the results were glassmeasured imprinting process averaged. were The results show using that thethe deterioration of thetosurface roughness VC mold Droplets generated chip in order investigate the during applicability of the fabrication and the glass imprinting process was negligible. fabrication method. We prepared an apparatus for the observation of droplet generation. The apparatus Droplets were generated using the chip in order to investigate the applicability of the fabrication included an inverted microscope (CKX-41, Olympus Co., Tokyo, Japan), two precision syringe pumps method. We prepared an apparatus for the observation of droplet generation. The apparatus included (Legato 200 & KDS 100, KD Scientific.), andCo., a high-speed camera (CR600x2, Optronis an inverted microscope (CKX-41, Olympus Tokyo, Japan), two precision syringe pumpsGmbH, (LegatoKehl, Germany). Droplet generation was conducted using deionized water as a dispersed phase and 200 & KDS 100, KD Scientific.), and a high-speed camera (CR600x2, Optronis GmbH, Kehl, Germany). fluorocarbon oil (HFE-7500, 3M Co.,using St. Paul, MN, water USA)as asaadispersed continuous phase. The microchannel Droplet generation was conducted deionized phase and fluorocarbon oil surfaces were made via treatment withphase. a commercial reagentsurfaces (Repel-Silane ES, GE (HFE-7500, 3M Co.,hydrophobic St. Paul, MN, USA) as a continuous The microchannel were made hydrophobic via treatment with aTable commercial reagentthe (Repel-Silane ES, GE Healthcare Co., fixed Healthcare Co., Waukesha, WI, USA). 3 summarizes test results with an aqueous phase WI, USA). Table 3 summarizes the conditions. test results with an aqueous phase fixed flow rate of flow Waukesha, rate of 30 µl/min and 4 different oil phase 30 µl/min and 4 different oil phase conditions.
Figure 4. (a) Design of the droplet-generating microfluidic chip, and photograph and SEM images of Figure 4. (a) Design of the droplet-generating microfluidic chip, and photograph and SEM images of the replicated glass microfluidic structure: (b) the silicon master, (c) the PDMS mold, (d) the furan the replicated glass microfluidic structure: (b) the silicon master, (c) the PDMS mold, (d) the furan precursor, (e) the vitreous carbon stamp, and (f) the replicated glass microfluidic structure. precursor, (e) the vitreous carbon stamp, and (f) the replicated glass microfluidic structure.
Sensors 2018, 2018, 18, 18, 83 83 Sensors
6 of 9
Figure results of of (a) Figure 5. 5. 3D 3D microscope microscope measurement measurement results (a) the the silicon silicon master, master, (b) (b) the the PDMS PDMS mold, mold, (c) (c) the the furan precursor, (d) the vitreous carbon stamp, and (e) the replicated glass microfluidic structure. furan precursor, (d) the vitreous carbon stamp, and (e) the replicated glass microfluidic structure. Comparison Comparison of of the the surface surface profiles profiles measured measuredin in(f) (f)the thechannel channeland and(g) (g)the theorifice. orifice. Table 1. Table 1. Comparison Comparison of of the the pattern pattern widths widths and and heights heights measured measured after after each each fabrication fabrication process. process. Si Master Si 62.53 Master
Mean (µm) Standard deviation Mean (µm) (µm) (coefficient of variance) (µm) Standard deviation
Orifice width
Orifice width
Total Shrinkage (coefficient of variance) ratio (from master)
62.53
- (-)
- (-) -
PDMS Mold
Furan Precursor Furan
PDMS Mold 61.83
61.83
0.1559 (0.252%)
0.1559 (0.252%) 0.98%
61.33 Precursor
61.33
0.1179 (0.192%)
0.1179 (0.192%) 1.91%
VC Mold
VC Mold
43.50
43.50
0.2700 (0.621%)
0.2700 (0.621%) 30.43%
Glass Replica Glass 43.54 Replica
43.54
0.2569 (0.590%)
0.2569 (0.590%) 30.30%
Total Shrinkage ratio 0.98% 1.91% 30.43% 30.30% Mean (µm) 127.35 126.22 125.06 88.93 88.87 (from master) Standard deviation Mean (µm) 127.35 0.0236 (0.019%) 126.22 125.06 0.1671 (0.188%) 88.93 0.2718 (0.306%) 88.87 (µm) (coefficient of - (-) 0.2585 (0.207%) Channel width Standard deviation 0.0236 0.2585 0.1671 0.2718 variance) (µm) Channel - (-) (coefficient of variance) (0.019%) (0.207%) (0.188%) (0.306%) Shrinkage ratio 0.89% 1.80% 30.17% 30.22% width (fromratio master) Shrinkage (from 0.89% 1.80% 30.17% 30.22% Mean (µm) 40.29 40.13 39.84 28.07 27.91 master) Standard deviation Mean (µm) 40.29 40.13 39.84 28.07 27.91 (µm) (coefficient of 0.0125 (0.031%) 0.0535 (0.134%) 0.1337 (0.476%) 0.1808 (0.648%) Height (µm) Standard deviation 0.0125 0.0535 0.1337 0.1808 variance) (µm) Height (coefficient of variance) (0.031%) (0.134%) (0.476%) (0.648%) Shrinkage ratio (µm) 0.41% 1.12% 30.33% 30.74% (from master) Shrinkage ratio 0.41% 1.12% 30.33% 30.74% (from master) Table 2. Comparison of the roughness generated by each of the fabrication processes. Table 2. Comparison of the roughness generated by each of the fabrication processes. Arithmetic Average of the Roughness (Ra, nm) Si master PDMS mold Furan Precursor Si master VC mold Glass replica PDMS mold
Furan Precursor VC mold Glass replica
Arithmetic Average of the 1.67 ± 0.52 3.83 ± 1.47 (Ra, nm) Roughness 1.67 ± 0.52 1.67 ± 0.52 2.83 ± 0.41 5.50 ± 1.76± 1.47 3.83 1.67 ± 0.52 2.83 ± 0.41 5.50 ± 1.76
Root Mean Squared Roughness (Rq, nm)
Root Mean Squared 2.17 ± 0.41 5.17 ± 1.47 nm) Roughness (Rq, 2.17 ± 0.41 2.173.67 ± 0.41 ± 0.52 ± 3.02 5.177.50 ± 1.47 2.17 ± 0.41 3.67 ± 0.52 7.50 ± 3.02
Table 3. Comparison of droplet generation frequencies achieved under various flow rate conditions.
Sensors 2018, 18, 83
7 of 9
Sensors 2018, 18, 83 Sensors 2018, 18, 83 Sensors Sensors 2018, 2018, 18, 18, 83 83
7 of 9 7 of 9 77 of of 99
Table 3. Comparison of droplet generation frequencies achieved under various flow rate conditions.
Flow Flow Rate Rate of of Flow of Oil Flow Rate of Flow Rate Rate of Oilof Oil FlowFlow RateRate of of Aqueous Aqueous Phase Flowof Rate Flow Rate Aqueous Phase Flow Rate of Oil Oil Phase Aqueous Phase Phase (μL/min) (μL/min) Phase (µL/min) Phase Phase (µL/min) Aqueous Phase (μL/min) (μL/min) (μL/min) Phase (μL/min) (μL/min) (μL/min)
30
Sensors 2018, 18, 83
30 30 30 30
2.5 2.5 2.5 2.5
55 55
Flow Rate of Aqueous Phase (μL/min)
10 10 10 10
30
2.5
15 15 15 15
5 Flow Rate of Oil Phase (μL/min)
2.5
Generation Generation Generation Generation Generation Frequency (Hz) Frequency (Hz) Frequency (Hz) Frequency Frequency (Hz) (Hz)
469 469 469 469
872 872 872 872
1588 1588 1588 469 1588
15
2521 2521 2521 2521
5
469
7 of 9
872
Generation Frequency (Hz)
10
Captured Captured Image Image Captured Image Captured Image from the Movie Captured Image from from the the Movie Movie from from theCaptured Movie the Movie by Captured by High Captured by by High High Captured High Speed Camera Captured by High Speed Camera Speed Camera Camera Speed Speed Camera
Captured Image from the Movie Captured by High Speed Camera
1588
872
10
2521
1588
Finally, Finally, the the initial initial autofluorescence autofluorescence of of the the fabricated fabricated micro micro channel channel structure structure was was measured measured for for Finally, initial of the fabricated micro channel structure measured for Finally, the the initial autofluorescence autofluorescence ofnm theand fabricated micro channel structure was was measured for two different excitation wavelengths: 532 635 nm. Those were compared with results from two different excitation wavelengths: 532 nm and 635 nm. Those were compared with results from two excitation wavelengths: 532 nm and nm. Those were compared results from Finally, the initial autofluorescence of the micro channel structure waswith measured two two different different excitation wavelengths: 532 nmfabricated and 635 635including nm. Those were compared with resultsfor from other commonly used materials for microfluidics, PDMS, Poly(methyl methacrylate) 2521including PDMS, Poly(methyl methacrylate) other commonly commonly used used materials materials 15for for microfluidics, microfluidics, other PDMS, Poly(methyl methacrylate) different excitation wavelengths: 532 and 635fluorescence nm.including Those were compared with results from other other commonly used materials for nm microfluidics, including PDMS, Poly(methyl methacrylate) (PMMA) and polycarbonate (PC). Laser-induced detection (PMMA) and polycarbonate (PC). Laser-induced fluorescence detection was was applied applied to to measure measure the the (PMMA) and polycarbonate (PC). Laser-induced fluorescence detection was applied to measure the commonly used materials for microfluidics, including PDMS, Poly(methyl methacrylate) (PMMA) (PMMA) and polycarbonate (PC). Laser-induced fluorescence detection was6, applied to measure the initial autofluorescence of the microfluidic chips. As shown in Figure the fabricated glass initial autofluorescence of the microfluidic chips. As shown in Figure 6, the fabricated glass Finally, the initial autofluorescence of the fabricated micro channelin structure was measured for initial autofluorescence of the microfluidic chips. As shown Figure 6, the fabricated glass and polycarbonate (PC). Laser-induced fluorescence detection was applied to measure the initial initial autofluorescence of the microfluidic chips. As shown in Figure 6, the fabricated glass microfluidic has initial autofluorescence We believe that two different excitation wavelengths: 532 nm and 635 level. nm. Those were compared with this resultsglass from microfluidics microfluidic device device has low low initial autofluorescence level. We believe that this glass microfluidics microfluidic device has initial autofluorescence level. We believe that this glass autofluorescence ofcommonly the low microfluidic chips. As shown in Figure 6,Poly(methyl the fabricated glassmicrofluidics microfluidic other used materials for microfluidics, including PDMS, methacrylate) microfluidic device has low initial autofluorescence level. We believe that this glass microfluidics fabrication method could be a powerful solution for the devices used in applications requiring fabrication method could be a powerful solution for the devices used in applications requiring (PMMA) and polycarbonate (PC). Laser-induced fluorescence detection was applied to measure the fabrication method could be a powerful solution for the devices used in applications requiring device has low initialcould autofluorescence level. We believe glass microfluidics fabrication method fabrication method be a at powerful solution forthat the this devices used in applications requiring sensitive fluorescence detection low cost. initial autofluorescence of the microfluidic chips. As shown in Figure 6, the fabricated glass sensitive fluorescence fluorescence detection detection at at low low cost. cost. sensitive could befluorescence a powerful solution for the devices used in applications requiring sensitive fluorescence sensitive detection at low cost. microfluidic device has low initial autofluorescence level. We believe that this glass microfluidics detection at lowfabrication cost. method could be a powerful solution for the devices used in applications requiring sensitive fluorescence detection at low cost.
Figure 6. Measurement results for the autofluorescence in different materials commonly used for microfluidic devices; The unit of autofluorescece is absolute unit (AU).
Sensors 2018, 18, 83
8 of 9
4. Conclusions In this study, we developed a glass imprinting process for the manufacturing of glass microfluidic chips. This method uses a VC mold to provide cost-effective microstructure texturing on a glass substrate. The VC mold was fabricated by combining a furan-based thermal replication process with carbonization. The overall geometrical deviation between the silicon master and the VC mold, due to the shrinkage of the furan resin during carbonization, was approximately 30%. Degradation of the surface integrity during carbonization was negligible. Since the shrinkage was isotropic and repeatable (30.31% with less than ~1% error), it can be predicted and could be compensated for by enlarging the initial master pattern. To verify the applicability of this method, a microfluidic chip designed for droplet generation via a flow-focusing structure was fabricated. Droplet generation was demonstrated at various flow rates. We demonstrated that our approach can provide cost-effective glass microfluidic chip fabrication. The fabrication of the all-glass chip via thermal fusion bonding and the practical application is the subject of ongoing research. In addition, the proposed VC mold fabrication and glass imprinting technique will be extended to nanoscale structured devices. Acknowledgments: This research was supported by a National Research Foundation of Korea (NRF) grant funded by the Korean Government (MSIP) (No. 2015R1A5A1037668). Author Contributions: Hyungjun Jang, Muhammad Refatul Haq, and Pyoung-hwa Oh, conducted the VC mold fabrication. Youngkyu Kim, Jun Kim and Jonghyun Ju performed the glass imprinting process. Seok-Min Kim and Jiseok Lim worked on data analysis and supervise all researchers. All authors were involved in writing the paper. Conflicts of Interest: The authors declare no conflict of interest.
References 1. 2. 3. 4. 5.
6. 7. 8. 9. 10. 11. 12. 13. 14.
Whitesides, G.M. The origins and the future of microfluidics. Nature 2006, 442, 368–373. [CrossRef] [PubMed] Dittrich, P.S.; Manz, A. Lab-on-a-chip: Microfluidics in drug discovery. Nat. Rev. Drug Discov. 2006, 5, 210–218. [CrossRef] [PubMed] Ma, Z.; Teo, A.J.T.; Tan, S.H.; Ai, Y.; Nguyen, N.T. Self-aligned interdigitated transducers for acoustofluidics. Micromachines 2016, 7, 216. [CrossRef] Xi, H.D.; Guo, W.; Leniart, M.; Chong, Z.Z.; Tan, S.H. AC electric field induced droplet deformation in a microfluidic T-junction. Lab Chip 2016, 16, 2982–2986. [CrossRef] [PubMed] Castro-Hernandez, E.; Garcia-Sanchez, P.; Tan, S.H.; Ganan-Calvo, A.M.; Baret, J.-C.; Ramos, A. Breakup length of AC electrified jets in a microfluidic flow-focusing junction. Microfluid. Nanofluid. 2016, 19, 787–794. [CrossRef] Link, D.; Anna, S.L.; Weitz, D.; Stone, H. Geometrically mediated breakup of drops in microfluidic devices. Phys. Rev. Lett. 2004, 92, 054503. [CrossRef] [PubMed] McDonald, J.C.; Whitesides, G.M. Poly(dimethylsiloxane) as a material for fabricating microfluidic devices. Acc. Chem. Res. 2002, 35, 491–499. [CrossRef] [PubMed] Lee, J.N.; Park, C.; Whitesides, G.M. Solvent compatibility of poly(dimethylsiloxane)-based microfluidic devices. Anal. Chem. 2003, 75, 6544–6554. [CrossRef] [PubMed] Mukhopadhyay, R. When PDMS isn’t the best. Anal. Chem. 2007, 79, 3248–3253. [CrossRef] [PubMed] Lee, J.; Kim, M.J.; Lee, H.H. Surface modification of poly(dimethylsiloxane) for retarding swelling in organic solvents. Langmuir 2006, 22, 2090–2095. [CrossRef] [PubMed] Abate, A.R.; Lee, D.; Do, T.; Holtze, C.; Weitz, D.A. Glass coating for PDMS microfluidic channels by sol-gel methods. Lab Chip. 2008, 8, 516–518. [CrossRef] [PubMed] Lim, J.; Gruner, P.; Konrad, M.; Baret, J.-C. Micro-optical lens array for fluorescence detection in droplet-based microfluidics. Lab Chip 2013, 13, 1472–1475. [CrossRef] [PubMed] Song, C.; Tan, S.H. A Perspective on the Rise of Optofluidics and the Future. Micromachines 2017, 8, 152. [CrossRef] Teo, A.J.; Li, K.H.H.; Nguyen, N.T.; Guo, W.; Heere, N.; Xi, H.D.; Tsao, C.W.; Li, W.; Tan, S.H. Negative Pressure Induced Droplet Generation in a Microfluidic Flow-Focusing Device. Anal. Chem. 2017, 89, 4387–4391. [CrossRef] [PubMed]
Sensors 2018, 18, 83
15.
16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.
27.
28. 29.
30. 31.
9 of 9
Grover, W.H.; Skelley, A.M.; Liu, C.N.; Lagally, E.T.; Mathies, R.A. Monolithic membrane valves and diaphragm pumps for practical large-scale integration into glass microfluidic devices. Sens. Actuators B Chem. 2003, 89, 315–323. [CrossRef] Chen, Q.; Li, G.; Jin, Q.H.; Zhao, J.L.; Ren, Q.S.; Xu, Y.S. A Rapid and Low-Cost Procedure for Fabrication of Glass Microfluidic Devices. J. Microelectromech. Syst. 2007, 16, 1193–1200. [CrossRef] Ren, K.; Zhou, J.; Wu, H. Materials for microfluidic chip fabrication. Acc. Chem. Res. 2013, 46, 2396–2406. [CrossRef] [PubMed] Nge, P.N.; Rogers, C.I.; Woolley, A.T. Advances in microfluidic materials, functions, integration, and applications. Chem. Rev. 2013, 113, 2550–2583. [CrossRef] [PubMed] Fiorini, G.S.; Chiu, D.T. Disposable microfluidic devices: Fabrication, function, and application. Biol. Tech. 2005, 38, 429–446. [CrossRef] Tseng, S.-F.; Chen, M.-F.; Hsiao, W.-T.; Huang, C.-Y.; Yang, C.-H.; Chen, Y.-S. Laser micromilling of convex microfluidic channels onto glassy carbon for glass molding dies. Opt. Laser Eng. 2014, 57, 58–63. [CrossRef] Chen, Q.; Chen, Q.; Maccioni, G. Fabrication of microfluidics structures on different glasses by simplified imprinting technique. Curr. Appl. Phys. 2013, 13, 256–261. [CrossRef] Chang, S.-H.; Lee, Y.-M.; Shin, K.-H.; Heo, Y.-M. A study on the aspheric glass tens forming analysis in the progressive gmp process. J. Opt. Soc. Korea 2007, 11, 85–92. [CrossRef] Choi, W.; Lee, J.; Kim, W.-B.; Min, B.-K.; Kang, S.; Lee, S.-J. Design and fabrication of tungsten carbide mould with micro patterns imprinted by micro lithography. J. Micromech. Microeng. 2004, 14, 1519. [CrossRef] Huang, C.-Y.; Kuo, C.-H.; Hsiao, W.-T.; Huang, K.-C.; Tseng, S.-F.; Chou, C.-P. Glass biochip fabrication by laser micromachining and glass-molding process. J. Mater. Process. Technol. 2012, 212, 633–639. [CrossRef] Zhou, T.; Yan, J.; Liang, Z.; Wang, X.; Kobayashi, R.; Kuriyagawa, T. Development of polycrystalline Ni–P mold by heat treatment for glass microgroove forming. Precis. Eng. 2015, 39, 25–30. [CrossRef] Youn, S.W.; Takahashi, M.; Goto, H.; Maeda, R. A study on focused ion beam milling of glassy carbon molds for the thermal imprinting of quartz and borosilicate glasses. J. Micromech. Microeng. 2006, 16, 2576. [CrossRef] Komori, M.; Uchiyama, H.; Takebe, H.; Kusuura, T.; Kobayashi, K.; Kuwahara, H.; Tsuchiya, T. Micro/nanoimprinting of glass under high temperature using a CVD diamond mold. J. Micromech. Microeng. 2008, 18, 065013. [CrossRef] Ju, J.; Lim, S.; Seok, J.; Kim, S.-M. A method to fabricate low-cost and large area vitreous carbon mold for glass molded microstructures. Intl. J. Precis. Eng. Man. 2015, 16, 287–291. [CrossRef] Ju, J.; Han, Y.; Seok, J.; Kim, S.-M. Development of Low-cost and Large-area Nanopatterned Vitreous Carbon Stamp for Glass Nanoreplication. In Proceedings of the 2012 12th IEEE Conference on Nanotechnology (IEEE-NANO), Birmingham, UK, 20–23 August 2012; pp. 1–4. Mao, P.; Han, J. Fabrication and characterization of 20 nm planar nanofluidic channels by glass-glass and glass-silicon bonding. Lab Chip. 2005, 5, 837–844. [CrossRef] [PubMed] Rogers, T.; Kowal, J. Selection of glass, anodic bonding conditions and material compatibility for silicon-glass capacitive sensors. Sens. Actuators. A. Phys. 1995, 46, 113–120. [CrossRef] © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).
