A Simple Low-Temperature Glass Bonding Process with Surface Activation by Oxygen Plasma for Micro/Nanoﬂuidic Devices

micromachines

Article A Simple Low-Temperature Glass Bonding Process with Surface Activation by Oxygen Plasma for Micro/Nanoﬂuidic Devices

Koki Shoda †, Minori Tanaka †, Kensuke Mino † and Yutaka Kazoe * Department of System Design Engineering, Faculty of Science and Technology, Keio University, 3-14-1 Hiyoshi, Kohoku, Kanagawa 223-8522, Japan; [email protected] (K.S.); [email protected] (M.T.); [email protected] (K.M.) * Correspondence: [email protected]; Tel.: +81-45-566-1843 These authors contributed equally to this work. † Received: 9 June 2020; Accepted: 23 August 2020; Published: 25 August 2020

Abstract: The bonding of glass substrates is necessary when constructing micro/nanofluidic devices for sealing micro- and nanochannels. Recently, a low-temperature glass bonding method utilizing surface activation with plasma was developed to realize micro/nanofluidic devices for various applications, but it still has issues for general use. Here, we propose a simple process of low-temperature glass bonding utilizing typical facilities available in clean rooms and applied it to the fabrication of micro/nanofluidic devices made of different glasses. In the process, the substrate surface was activated with oxygen plasma, and the glass substrates were placed in contact in a class ISO 5 clean room. The pre-bonded substrates were heated for annealing. We found an optimal concentration of oxygen plasma and achieved a bonding energy of 0.33–0.48 J/m2 in fused-silica/fused-silica glass bonding. The process was applied to the bonding of fused-silica glass and borosilicate glass, which is generally used in optical microscopy, and revealed higher bonding energy than fused-silica/fused-silica glass bonding. An annealing temperature lower than 200 ◦C was necessary to avoid crack generation by thermal stress due to the different thermal properties of the glasses. A fabricated micro/nanofluidic device exhibited a pressure resistance higher than 600 kPa. This work will contribute to the advancement of micro/nanofluidics.

Keywords: microﬂuidics; nanoﬂuidics; fabrication; bonding; glass

1. Introduction The field of microfluidics has rapidly developed to realize various applications, including chemical analysis and synthesis, medical diagnosis, and tissue engineering [1–3]. Recently, the field has been further downscaled to nanofluidics, exploiting volumes of attoliters to femtoliters where surface effects are dominant [4]. Novel applications have been reported such as single-molecule sorting [5] and analysis [6], high-efficiency separation utilizing solid/liquid [7] and liquid/liquid phases [8], single-cell proteomics [9], and autonomous solar-light-driven fuel cells [10]. To realize these devices, micro/nano fabrication technologies are important. Recently, the direct writing technologies, such as 3D printing using polymer materials, has rapidly developed to realize easy and high-throughput production [11–13]. However, the fabrication of nanostructures like nanopillars and nanochannels is still difficult. Currently, polydimethylsiloxane (PDMS)-based fabrication technologies are most widely used. Microchannels can be easily fabricated by replica molding technique (soft lithography) [14]. PDMS-made devices have optical transparency suitable for detections, and surface chemical modification utilizing surface silanol groups is available. However, maintaining the shape of nanostructure is difficult due to its softness, and the material’s solubility to organic solvents and gas permeability restrict its use. On the other hand,

Micromachines 2020, 11, 804; doi:10.3390/mi11090804 www.mdpi.com/journal/micromachines Micromachines 2020, 11, 804 2 of 12 glass-based fabrication technologies are challenging compared with others, but can achieve glass-made devices with mechanical robustness suitable for realizing nanostructures, chemical stability to organic solvents, optical transparency for detections and availability of surface chemical modification. Fabrication methods for micro/nanofluidic devices made of glass have been reported. Micro- and nanochannels are usually fabricated on a glass substrate by wet etching and plasma etching, respectively, after the channel patterns are formed by lithography [15,16]. Surface modification of micro- and nanochannels is performed by treating with reagents [7,10] or by patterning [6,8]. The methods for bonding glass substrates are essential to construct devices by sealing microchannels and nanochannels on a substrate with another substrate. The strongest form of bonding is by thermal fusion of glass substrates at high temperatures (1080 ◦C in the case of fused silica, 550 ◦C in the case of borosilicate) [17–19]. Glass bonding at room temperature by surface pre-treatment with a solution of hydrofluoric acid has been reported [16,20]. Anodic bonding method at temperatures of 300 to 500 ◦C has been used for bonding of borosilicate glass substrates with a conductive interlayer [19,21,22]. These bonding methods are problematic for the fabrication of micro/nanofluidic devices integrating various functional materials. Most of the functional materials utilized, such as self-assembled monolayers, catalysts and modified electrodes in micro- and nanochannels, do not tolerate the extremely high temperatures needed for thermal fusion. Using hydrofluoric acid, which is a typical etchant for glass, can damage micro/nanostructures fabricated on the glass substrate. The conductive interlayer required for anodic bonding often reduces the chemical tolerance of the devices. In order to solve these problems in bonding, recent studies have developed a method for bonding glass substrates by surface treatment with plasma and heating at low temperatures [23,24]. In this method, the glass surface is treated with oxygen plasma containing fluorine atoms for surface activation with moderate hydrophobization. After bringing the glass surfaces into contact, the substrates are heated at relatively low temperatures for annealing while pressing at 5000 N. The bonding strength was sufficient for endurance during driving fluids by pressures in micro- and nanochannels on the order of 1000 kPa. The low-temperature glass bonding method has contributed to the realization of micro/nanofluidic devices integrating various functional materials, as reported in recent studies [6,8–10]. In addition, detachable glass bonding with controlled bonding strength was realized by optimizing the bonding conditions [25]. Besides the methods utilizing oxygen plasma containing fluorine atoms, a sequential plasma activation process consisting of oxygen reactive ion etching plasma and nitrogen radical plasma was developed for room-temperature bonding of Si/glass and glass/glass wafers [26]. To decrease damages to the surface, a low-temperature bonding process utilizing surface activation by ultraviolet/ozone was also developed to achieve bonding of Si/Si and quartz/quartz wafers [27]. However, the low-temperature glass bonding method still has issues for general use. Adding fluorine atoms into oxygen plasma and heating the substrates while pressing with a controlled pressure requires the use of customized facilities in some cases. Furthermore, the bonding of different types of glass has not been achieved. Previous studies have verified the low-temperature bonding of fused-silica glass substrates [23–25,28], which is identical to the glass material used for semiconductor fabrication. However, since microscope objective lenses are optimally designed for borosilicate cover glass (refractive index: 1.52), the use of micro/nanofluidic devices made of fused-silica glass (refractive index: 1.46) results in reduced spatial resolution during microscopic observation, as reported in a previous study [29]. Thus, the bonding of fused-silica and borosilicate glass substrates is important for combining micro/nanofluidic devices and microscopic optical detection. In the present study, we developed a low-temperature glass bonding process utilizing typical facilities available in clean rooms, and applied the process to fabrication of micro/nanofluidic devices. A bonding process with surface activation by oxygen plasma and without pressing substrates during the annealing was proposed. The oxygen plasma and heating temperature conditions were investigated, and the bonding of fused-silica glass substrates was verified by measuring the bonding energy. The verified bonding process was applied to the bonding of fused-silica and borosilicate glass Micromachines 2020, 11, 804 3 of 12 substrates. The bonding energy of fused-silica/borosilicate glass substrates was evaluated. In addition, the pressure resistance of a micro/nanofluidic device made of fused-silica and borosilicate glass was evaluated by investigating liquid leakage from a nanochannel. This work provides a simple and easy low-temperature glass bonding process for the fabrication of micro/nanofluidic devices.

2. Materials and Methods

2.1. Materials Fused-silica glass substrates with thicknesses of 0.25 and 0.70 mm (70 mm 30 mm, VIO-SILSX, × Shin-Etsu Quartz Co., Ltd., Tokyo, Japan) and borosilicate glass substrates with thicknesses of 0.17 and 0.25 mm (70 mm 30 mm, Matsunami Glass Ind., Ltd., Osaka, Japan) were prepared. The 0.17-mm × borosilicate glass is used as a standard cover glass for microscopy. The Young’s modulus of fused-silica glass and borosilicate glass was 7.38 1010 Pa and 7.29 1010 Pa, respectively, according to the × × manufactures. The surface roughness of glass substrates was less than 0.3 nm. For the confirmation of liquid leakage from nanochannels, a fluorescent aqueous solution, including 0.1-mmol/L fluorescein and phosphate-buffered saline was prepared.

2.2. Fabrication of Micro/Nanoﬂuidic Devices Micro/nanoﬂuidic devices were fabricated by the top-down fabrication of micro- and nanochannels on glass substrates, based on reported methods [8,30,31]. Nanochannels with sizes of 100 to 1000 nm were fabricated by electron beam lithography and dry etching. In the electron beam lithography process (F5112+VD01, Advantest Corp., Tokyo, Japan), ZEP-520A (Zeon Corp., Tokyo, Japan) was used as the resist. After forming the channel pattern, dry etching (NLD-570, ULVAC Co., Ltd., Kanagawa, Japan) was performed using a mixture of gaseous CHF3 and SF6. To fabricate nanochannels of 10 to 100-µm width and 100 to 1000-nm depth, the channel pattern was formed by photolithography using THB-111N (JSR Corp., Tokyo, Japan) as a photoresist. On the other hand, microchannels were fabricated by photolithography and dry etching. KMPR®1035 (Kayaku Co., Ltd., Tokyo, Japan) was used as a photoresist for micrometer-scale dry etching using a mixture of gaseous C3F8, CHF3 and Ar. Inlet holes with a diameter of 0.7 mm were produced on the substrate using a diamond-coated drill. After fabricating the channels and inlet holes, the device was constructed by the low-temperature bonding of the glass substrates as described in the following section.

2.3. Low-Temperature-Bonding of Glass Substrates In the present study, we propose a process of low-temperature glass bonding utilizing typical facilities available in clean rooms. Figure1 illustrates schematics of the bonding process. In a class ISO 7 clean room, washing and activation of the surfaces of the glass substrates were performed. In the washing step, the substrates were immersed in a three-to-one mixture of sulfuric acid and hydrogen peroxide for 8 min. After cleaning the substrates by ultrasonication for 8 min and air-drying, the substrate surface was activated by irradiation with an oxygen plasma for 40 s at a power of 200 W, utilizing a gas plasma reactor (PR510, Yamato Scientiﬁc Co., Ltd., Tokyo, Japan). Then, the substrates were immersed in pure water to prevent contamination by particles in the atmosphere and transported to a class ISO 5 clean room. After washing the substrates with running pure water and air-drying, the substrates were brought into contact for pre-bonding. In this step, air often remained in the bonding interface and generate voids. In such case, air in the bonding interface was ejected by manual pressure. Finally, the pre-bonded substrates were heated for annealing in an electric furnace (FO200, Yamato Scientiﬁc Co., Ltd., Tokyo, Japan) at a constant temperature for 5 h with a heating rate of 100 ◦C/h. Micromachines 2020, 11, 804 4 of 12 Micromachines 2020, 11, x FOR PEER REVIEW 4 of 12

Micromachines 2020, 11, x FOR PEER REVIEW 4 of 12

Figure 1. SchematicSchematic diagram diagram of of low low-temperature-temperature bonding process for glass substrates utilizing typical facilities in clean rooms.

2.4. ETovaluation optimize of Bonding the proposed Strength bonding process, the eﬀects of the pressure in the plasma reaction chamber and the heating temperature used for annealing on the bonding strength were investigated. The bonding strength of glass substrates was measured by a crack-opening test [32], which is a The pressure in the reaction chamber was set to 40, 51 and 60 Pa by providing oxygen at a ﬂow rate of standard method for evaluating bonding. A stainless-steel blade with a thickness of 0.1 mm (Hi- 45, 70 and 100 mL/min, respectively, with evacuation by a vacuum pump. The heating temperature Stainless, Feather Safety Razor Co., Ltd., Osaka, Japan) was used. As shown in Figure 2a, the stainless was set at 20 (room temperature), 50, 110, 200, 300 and 400 ◦C. bladeFigure was inserted1. Schematic at the diagram bonding of low interface,-temperature and bondingthe crack process propagation for glass length,substrates L, utilizingwas measured. typical The 2.4.bonding Evaluationfacilities energy, in clean of Bondingγ, betweenrooms. Strength substrate 1 and substrate 2 is given by:

The bonding strength of glass substrates= was measured by, a crack-opening test [32], which(1) is a 2.4. Evaluation of Bonding Strength 2 3 3 3𝑡𝑡𝑏𝑏𝐸𝐸𝑠𝑠1𝑡𝑡𝑠𝑠1𝐸𝐸𝑠𝑠2𝑡𝑡𝑠𝑠2 standard method for evaluating bonding. A stainless-steel4 3 blade3 with a thickness of 0.1 mm (Hi-Stainless, 16𝐿𝐿 �𝐸𝐸𝑠𝑠1𝑡𝑡𝑠𝑠1+𝐸𝐸𝑠𝑠2𝑡𝑡𝑠𝑠2� FeatherwhereThe t Safetyb bondingis the Razor thickness strength Co., Ltd.,of of theglass Osaka, blade, substrates Japan) ts𝛾𝛾 is wasthe was thickness used. measured As shownof bythe a inglasscrack Figure -substrate,opening2a, the teststainless and [32 E]s, iswhich blade Young’s wasis a standardinsertedmodulus at method[2 the8]. As bonding shownfor evaluating interface,in Figure bonding. and2b, the the blade crackA stainless was propagation inserted-steel intoblade length, the with bondingL, wasa thickness measured. interface of between 0.1 The mm bonding glass(Hi- Stainless,energy,substratesγ ,Feather, between and the Safety substratecrack Razor propagation 1 Co., and Ltd., substrate length Osaka, 2was is Japan) given measu was by:red. used . As shown in Figure 2a, the stainless blade was inserted at the bonding interface, and the crack propagation length, L, was measured. The 2 3 3 bonding energy, γ, between substrate 1 and substrate3tbEs1ts1 E2s 2ists given2 by: γ = , (1) 16L4 E t3 + E t3 = s1 s1 s2 s2 2 3 3 , (1) 3𝑡𝑡𝑏𝑏𝐸𝐸𝑠𝑠1𝑡𝑡𝑠𝑠1𝐸𝐸𝑠𝑠2𝑡𝑡𝑠𝑠2 4 3 3 where t is the thickness of the blade, ts is the16𝐿𝐿 � thickness𝐸𝐸𝑠𝑠1𝑡𝑡𝑠𝑠1+𝐸𝐸𝑠𝑠2𝑡𝑡𝑠𝑠2 of� the glass substrate, and Es is Young’s where tbb is the thickness of the blade, ts𝛾𝛾 is the thickness of the glass substrate, and Es is Young’s Figure 2. (a) Principle of crack-opening test and (b) photograph of bonded glass substrates with a modulusmodulus [2 [288].]. As As shown shown in in Figure Figure 2b,b, the blade was inserted into the bonding interface between glass blade inserted into the bonding interface. substratessubstrates,, and and the the crack crack propagation propagation length was measu measured.red. The pressure resistance of the micro/nanofluidic device was examined by observing the leakage of a fluorescein solution from a nanochannel. The fluorescence was observed by an inverted fluorescence microscope (IX71, Olympus Corp., Tokyo, Japan) combined with an objective lens (20×, NA = 0.75), a 1.6× magnification lens, and a complementary metal-oxide-semiconductor (CMOS)

camera (C11440-36U, Hamamatsu Photonics K. K., Hamamatsu, Japan). The pixel size of the CMOS cameraFigureFigure was 2. 2. 5.86 ((aa)) Principle Principleμm, and of ofthe crack crack-opening exposure-opening time test test wasand and (set (bb)) photograph photographat 70 ms. The of of bonded bondedfluorescein glass glass solutionsubstrates substrates was with with injected a a into bladebladethe deviceinserted inserted by into into air the the pressure bonding bonding interface. interface.generated by a pressure controller (Institute of Microchemical Technology Co., Ltd., Kanagawa, Japan). TheThe pressure pressure resistance resistance of of the the micro/ micro/nanofluidicnanofluidic device was was examined examined by by observing observing the the leakage leakage of a fluorescein solution from a nanochannel. The fluorescence was observed by an inverted of3. Results a fluorescein and Discussion solution from a nanochannel. The fluorescence was observed by an inverted fluorescencefluorescence microscope microscope (IX71, (IX71, Olympus Olympus Corp., Corp., Tokyo, Tokyo, Japan) Japan) combined combined with with an objective an objective lens (20 lens×, NA(20 =, NA0.75)=, 0.75a 1.6),× a 1.6magnificationmagnification lens, lens, and and a complementary a complementary metal metal-oxide-semiconductor-oxide-semiconductor (CMOS) (CMOS) 3.1.× Bonding of Fused-Silica/Fused× -Silica Glass Substrates camera (C11440-36U, Hamamatsu Photonics K. K., Hamamatsu, Japan). The pixel size of the CMOS cameraFigure was 5.863 shows μm, and photographs the exposure of timefused was-silica set atglass 70 ms.substrates The fluorescein bonded solutionby low -wastemperature injected intobonding the deviceaccording by toair the pressure process generatedshown in Figureby a pressure 1. The pressure controller in the(Institute plasma of reaction Microchemical chamber Technology Co., Ltd., Kanagawa, Japan).

3. Results and Discussion

3.1. Bonding of Fused-Silica/Fused-Silica Glass Substrates Figure 3 shows photographs of fused-silica glass substrates bonded by low-temperature bonding according to the process shown in Figure 1. The pressure in the plasma reaction chamber

Micromachines 2020, 11, 804 5 of 12 camera (C11440-36U, Hamamatsu Photonics K. K., Hamamatsu, Japan). The pixel size of the CMOS camera was 5.86 µm, and the exposure time was set at 70 ms. The ﬂuorescein solution was injected into the device by air pressure generated by a pressure controller (Institute of Microchemical Technology Co., Ltd., Kanagawa, Japan).

3. Results and Discussion

3.1. Bonding of Fused-Silica/Fused-Silica Glass Substrates Micromachines 2020, 11, x FOR PEER REVIEW 5 of 12 Figure3 shows photographs of fused-silica glass substrates bonded by low-temperature bonding accordingand the annealing to the process temperature shown were in Figure set at 151. ThePa and pressure 400 °C, in respectively. the plasma reactionAs shown chamber in Figure and 3a, the the annealingglass substrates temperature were bonded were set without at 51 Pa residual and 400 v◦oidsC, respectively. at the interface Asshown between in Figurethe substrates.3a, the glass The substratesbonding process were bonded was applied without to residualthe fabrication voids at of the a micro/ interfacenanofluidic between thedevice substrates. made of The fused bonding-silica processglass. As was shown applied in toFigure the fabrication 3b, the device of a micro include/nanoﬂuidicd nanochannels device madeof 400 of μ fused-silicam wide and glass. 900 Asnm shown deep, inwhich Figure were3b, theinterfaced device includedwith microchannels nanochannels of of1.7 400 mmµm wid widee and and 30 900 μm nm de deep,ep. As which well wereas the interfaced bonding withof glass microchannels substrates without of 1.7 mm any wide fabrication, and 30 µ tmhe deep.bonding As wellof glass as the substrates bonding with of glass fabricated substrates micro without- and anynano fabrication,channels was the realized bonding. We of glass note substratesthat, among with the fabricatedbonding process micro- and(Figure nanochannels 1), when we was conducted realized. Weprocesses note that, of “washing among the with bonding running process water” (Figure and1), “pre when-bonding” we conducted in a class processes ISO 7 of clean “washing room, with the runningsubstrate water” surface and was “pre-bonding” contaminated in aby class a large ISO 7number clean room, of particles the substrate from surfacethe atmosphere, was contaminated and the bybonding a large failed. number This of result particles suggests from the the atmosphere, importance andof using the bonding a class ISO failed. 5 clean This room result during suggests these the importanceprocesses to ofachieve using asuccessful class ISO bonding. 5 clean room during these processes to achieve successful bonding.

Figure 3. 3. LowLow-temperature-temperature bonding bonding of fused of fused-silica-silica glass glass substrates. substrates. (a) Photograph (a) Photograph of bonded of bonded fused- fused-silicasilica glass subst glassrates. substrates. (b) Photograph (b) Photograph of a micro/na of a micronofluidic/nanoﬂuidic device devicefabricated fabricated by bonding by bonding of a fused of a- fused-silicasilica glass subst glassrates substrates with microchannels with microchannels and nanochannels. and nanochannels.

In order toto evaluate the bondingbonding strength of thethe fused-silicafused-silica glassglass substrates,substrates, we performed a crack-openingcrack-opening test.test. Figure4 4aa shows shows the the bonding bonding energy energy as as a a function function of of the the heating heating temperature temperature for for a dia ffdifferenterent plasma plasma reaction reaction chamber chamber pressure pressure conditions. conditions. For each For condition, each wecondition, repeated we experiments repeated atexperiments least four timesat least and four obtained times and standard obtained deviation. standard The deviation bonding. The energy bonding increased energy with increased increasing with heatingincreasing temperature. heating temperature. We found thatWe afound moderate that a concentration moderate concentration of oxygen plasma of oxygen with plasma the pressure with the of 51pressure Pa could of 51 achieve Pa could high achieve bonding high energies bonding ranging energies from ranging 0.33 to 0.48from J/ m0.332 with to 0.48 heating J/m2 temperatureswith heating rangingtemperatures from ranging 200 to 400 from◦C, 200 which to 400 are °C comparable, which are comparable to that corresponding to that corresponding to a pressure to a resistance pressure onresistance the order on the of 100 order kPa, of 100 as reported kPa, as reported previously previously [23–25]. [23 Therefore,–25]. Therefore, in case in of case bonding of bonding fused-silica fused- glasssilica substrates,glass substrates, we verified we verified the process the process of low-temperature of low-temperature glass bonding glass bonding utilizing utilizing typical facilities typical availablefacilities available in clean rooms.in clean rooms. WeWe alsoalso investigatedinvestigated the eeffectsffects of surface activation by oxygen plasma plasma on on the the bonding bonding energy. energy. Figure 44bb showsshows a comparison comparison of of the the bonding bonding energ energyy at atdifferent di fferent plasma plasma powers powers (0 and (0 and200 W 200). The W). Thepressure pressure in the in thereaction reaction chambe chamberr was was set set at at51 51Pa. Pa. At At a aheating heating temperature temperature of of 200 200 °◦CC,, the the bonding energy with surface activation for a power of 200 W increased 1.7 times compared to that without surface activationactivation (power:(power: 00 W).W).At At a a heating heating temperature temperature of of 400 400◦C, °C the, th bondinge bonding energies energies at at powers powers of 0of and 0 and 200 200 W became W beca comparable.me comparable The. resultsThe result indicates indicate that, for that lower, for heating lower heating temperatures, temperatures, the bonding the strengthbonding isstrength increased is increased by activating by activating the surface. the surface.

Micromachines 2020, 11, 804 6 of 12 Micromachines 2020, 11, x FOR PEER REVIEW 6 of 12

FigureFigure 4.4. Bonding energy measured by aa crack-openingcrack-opening test.test. ( a)) Bonding energy as a functionfunction ofof heatingheating temperaturetemperature at at di ﬀdifferenterent pressures pressures in ain reaction a reaction chamber, chamber, indicating indicating the concentration the concentration of oxygen of plasma.oxygen plasma. (b) Comparison (b) Comparison of bonding of bonding energy for energy plasma for power plasma of power 0 and 200of 0 W. and (c) 200 Bonding W. (c) energy Bonding as functionenergy as of function heating of temperature heating temperature at a pressure at a in press reactionure in chamber reaction of chamber 51 Pa. of 51 Pa.

The results were evaluated to consider the mechanism involved in low-temperature glass bonding, The results were evaluated to consider the mechanism involved in low-temperature glass which is not yet fully understood. Recently, Wang et al. [28] experimentally investigated surface bonding, which is not yet fully understood. Recently, Wang et al. [28] experimentally investigated properties during bonding by contact angle measurement, atomic force microscopy, Raman scattering surface properties during bonding by contact angle measurement, atomic force microscopy, Raman spectrometry and X-ray photoelectron spectroscopy, and proposed the following bonding mechanism. scattering spectrometry and X-ray photoelectron spectroscopy, and proposed the following bonding Upon activation by oxygen plasma, the oxygen free radicals break the covalent bonds of siloxane mechanism. Upon activation by oxygen plasma, the oxygen free radicals break the covalent bonds of (Si-O-Si) on the glass surface and produce dangling Si-O- groups by the reaction, Si-O-Si + O* Si- + siloxane (Si-O-Si) on the glass surface and produce dangling Si-O- groups by the reaction, Si→-O-Si + Si-O-. After activation, water molecules are adsorbed on the surface to form silanol groups (Si-OH) O* → Si- + Si-O-. After activation, water molecules are adsorbed on the surface to form silanol groups by the reaction, Si- + Si-O- + H2O Si-OH + Si-OH, and the hydrophilicity of the surface increases. (Si-OH) by the reaction, Si- + Si-O→- + H2O → Si-OH + Si-OH, and the hydrophilicity of the surface During the annealing after the pre-bonding, a dehydration reaction of silanol groups between the pair increases. During the annealing after the pre-bonding, a dehydration reaction of silanol groups of substrates occurs to form covalent bonds of siloxane, according to Si-OH + Si-OH Si-O-Si + H O. between the pair of substrates occurs to form covalent bonds of siloxane, according →to Si-OH + Si-OH2 As a result, the bonding of the substrates is achieved. Xu et al. [23,24] and Ohta et al. [25] indicated → Si-O-Si + H2O. As a result, the bonding of the substrates is achieved. Xu et al. [23,24] and Ohta et that an increase in bonding energy could be achieved by adding fluorine plasma in the activation al. [25] indicated that an increase in bonding energy could be achieved by adding fluorine plasma in process. They suggested that moderate hydrophobicity of the surface is necessary for the ejection of the activation process. They suggested that moderate hydrophobicity of the surface is necessary for residual water molecules from the bonding interface to enhance the bonding. The results obtained in the ejection of residual water molecules from the bonding interface to enhance the bonding. The the present study basically support the models considered in these previous studies, and provide new results obtained in the present study basically support the models considered in these previous insights to understand the bonding mechanism. Namely, finding the optimal concentration of oxygen studies, and provide new insights to understand the bonding mechanism. Namely, finding the plasmaoptimal asconcentration shown in Figure of oxygen4a suggests plasma that as strongshown bondingin Figure can 4a suggests be achieved thatby strong an optimal bonding balance can be betweenachieved the by amountan optimal of hydrophilic balance between silanol the groups amount (Si-OH) of hydrophilic to produce silanol bonding groups andthat (Si-OH) of hydrophobic to produce siloxanebonding (Si-O-Si)and that to of eject hydrophobic water molecules siloxane from (Si the-O bonding-Si) to eject interface. water In molecules addition, thefrom result the shownbonding in Figureinterface.4bsuggests In addition, that the eresultffect ofshown increased in Figure Si-OH 4b due suggest to surfaces that activationthe effect onof increased the bonding Si- strengthOH due isto significantsurface activation at 200 ◦onC, the while bonding it becomes strength insignificant is significant at 400at 200◦C °C because, while it the becomes dehydration insignificant reaction at of400 Si-OH °C because for bonding the isdehydration sufficiently reaction accelerated of atSi- temperaturesOH for bonding higher is thansufficiently 200 ◦C, asaccelerated reported inat previoustemperatures studies higher on the than surface 200 ° chemistryC, as reported of silica in previous [33]. studies on the surface chemistry of silica [33]. At the optimized pressure in reaction chamber of 51 Pa, we further investigated the heating temperatureAt the optimized for annealing. pressure Figure in4c reaction shows the chamber results ofof a51 crack-opening Pa, we further test investigated to measure the the bonding heating energytemperature as a function for annealing of heating. Figure temperature 4c shows ranging the results from of 20 a crack (room-opening temperature) test to to measure 400 ◦C. Thethe bonding energyenergy significantlyas a function decreased of heating when temperature the heating ranging temperature from 20 decreased (room temperature) to 110 ◦C, probably to 400 because°C. The ofbonding the insu energyfficient significantly dehydration decreased reaction when of silanol the heating groups. temperature Therefore, in decreased case of bonding to 110 ° fused-silicaC, probably because of the insufficient dehydration reaction of silanol groups. Therefore, in case of bonding

Micromachines 2020, 11, 804 7 of 12 Micromachines 2020, 11, x FOR PEER REVIEW 7 of 12 fusedglass substrates,-silica glass the substrates, heating temperature the heating higher temperature than 200 higher◦C is necessary than 200 to°C obtain is necessary a suﬃcient to obtain bonding a sufficientenergy for bonding nanoﬂuidic energy devices, for nanofluidic based on previous devices, studiesbased on [23 previous]. studies [23].

3.2. Bonding of F Fused-Silicaused-Silica/Borosilicate/Borosilicate Glass Subsrates Based on the results shown in Figure4 4,, thethe low-temperaturelow-temperature glassglass bondingbonding methodmethod withwith anan optimized pressure of 51 Pa was used to bond fused-silicafused-silica and borosilicate glass substrates. Figure 5 5 shows thethe results results of of bonding bonding at di atff erentdifferent heating heating temperatures. temperatures. The 0.17-mm The 0.17 borosilicate-mm borosilicate glass substrate, glass whichsubstrate, is a which standard is a standard cover glass cover for glass microscopy, for microscopy, was used. was Inused. the In case the ofcase substrates of substrates without without any anyfabrication fabrication (Figure (Figure5a), the 5a bonding), the bonding was successful was successful at heating at temperaturesheating temperature of 200 ands of 250 200◦ C.and However, 250 °C. However,at a heating at temperature a heating temperature of 300 ◦C, cracks of 300 were °C, generatedcracks were owing generated to the thermalowing to stress the thermal caused bystress the 6 1 causedorder of by magnitude the order diofff magnitudeerence in thermal difference expansion in thermal coeffi expansioncients between coefficient fused-silicas between (0.5 fused10−-silicaK− ) 6 1 × and borosilicate−6 −1 (7.2 10 K ). On−6 the−1 other hand, in the case of substrates with micro- and (0.5 × 10 K ) and borosilicate× − (7.2− × 10 K ). On the other hand, in the case of substrates with micro- andnanochannels nanochannels (Figure (Figure5b), cracks 5b), cracks were generatedwere generated even at even a temperature at a temperature of 250 ◦ C,of probably250 °C, probably because becauseof stress of concentration stress concentration at the places at the where places the where channels the channels were fabricated. were fabricated At a temperature. At a temper ofature 200 ◦ ofC, we200 succeeded°C, we succeeded in fabricating in fabricating a micro a /micronanofluidic/nanofluidic device device containing containing microchannels microchannels (width: (width: 500 µ500m, depth:μm, depth: 10 µ 10m) μ andm) and nanochannels, nanochannels as shown, as shown in Figure in Figure5c. The5c. The results results suggest suggest that that the the fabrication fabrication of micro-of micro and- and nanochannels nanochannels aff ectsaffects an appropriatean appropriate heating heating temperature temperature for bondingfor bonding different different glasses glasses due todue the to concentration the concentration of thermal of thermal stress. Therefore,stress. Therefore, to fabricate to fabricate micro/nanofluidic micro/nanofluidic devices bydevices bonding by fused-silicabonding fused and-silica borosilicate and borosilicate glass substrates, glass substrates, using a heating using temperature a heating temperature lower than 200 lower◦C isthan important 200 °C tois important avoid crack to generation avoid crack due generation to the thermal due to stress. the thermal stress.

Figure 5. LowLow-temperature-temperature bonding of fused fused-silica-silica glass substrate (thickness: 0.7 mm) and and borosilicate borosilicate glass substrate (thickness: 0.17 mm). ( (aa)) Photograph Photograph of of bonded fused fused-silica-silica and borosilicate glass substrates. (b(b)) Photograph Photograph of microof micro/nanofluidic/nanoﬂuidic device device fabricated fabricated by the by bonding the bonding of fused-silica of fused substrate-silica withsubstrate micro- with and micro nanochannels- and nanochannels and a borosilicate and a borosilicate glass substrate. glass (substrate.c) X-shaped (c) nanochannel X-shaped nanochannel of 4000 nm wideof 4000 and nm 2000 wid nme and deep 2000 fabricated nm de inep a boxedfabrica areated in indicated a boxed on area the photographindicated on of the micro photograph/nanoﬂuidic of micdevicero/nanofluidic (Figure4b). device (Figure 4b).

SSinceince the the borosilicate borosilicate glass glass substrate substrate with with a 0.17 a 0.17-mm-mm thickness thickness was wastoo fragile too fragile to perform to perform a crack a- openingcrack-opening test, we test, instead we instead used the used borosilicate the borosilicate glass substrate glass substrate with a 0.25 with-mm a 0.25-mmthickness thickness to examine to theexamine bonding the energy bonding at energythe heating at the temperature heating temperature lower than lower 200 ° thanC. The 200 bonding◦C. The energies bonding of energies fused- silica/fusedof fused-silica silica,/fused fused silica,-silica/borosilicate fused-silica/borosilicate and borosilicate/borosilicate and borosilicate/borosilicate glass substrates glass substrates at a heating at a temperatureheating temperature of 110 °C of are 110 shown◦C are in shown Figure in 6. Figure To confirm6. To conﬁrmthe effect the of etheﬀect substrate of the substrate thickness thickness,, a fused- silicaa fused-silica glass substrate glass substrate with a thickness with a thicknessof 0.25 mm of was 0.25 also mm used was. The also results used. suggest The results that suggestthe bonding that energythe bonding is independent energy is independenton the substrate on the thickness substrate but thickness dependent but on dependent the glass on material. the glass When material. the borosilicate glass was used, the bonding energy was more than three times higher than that in case of fused-silica/fused-silica glass bonding. The bonding energies of fused-silica/borosilicate and

Micromachines 2020, 11, 804 8 of 12

WhenMicromachines the borosilicate 2020, 11, x FOR glass PEER was REVIEW used, the bonding energy was more than three times higher than8 thatof 12 in case of fused-silica/fused-silica glass bonding. The bonding energies of fused-silica/borosilicate 2 andborosilicate/borosilicate borosilicate/borosilicate glass glass substrates substrates were were around around 0.5 0.5 J/m J/m2 at ata aheating heating temperature temperature of of 110 110 °◦C, which is susufficientfficient forfor nanofluidicnanofluidic devices based on previous studies [[2233]. We note that, at a heatingheating temperature of 200 ◦°C, we could not insert the blade without breaking the glass because of further increased bonding bonding energy. energy. Therefore, Therefore, in in case case of ofbonding bonding fused fused-silica-silica/borosilicate/borosilicate glass glass substrates, substrates, the theheating heating temperature temperature lower lower than than200 °C 200 is an◦C appropriate is an appropriate condition condition for nanofluidic for nanofluidic devices. devices.Several Severalreasons reasonsfor the forincreased the increased bonding bonding energy energy by using by using borosilicate borosilicate glass glass can canbe beconsidered considered,, such such asas a adifferent different density density of of silanol silanol group group on on the the surface surface and and B 2OO33 oror other other cations included in borosilicate glass. TheThe previousprevious study study revealed revealed that that washing washing the substratethe substrate with calciumwith calcium acetate acetate solution solution can enhance can theenhance low-temperature the low-temperature bonding of borosilicate bonding/ borosilicateof borosilicate/borosilicate glasses [34]. Strong borosilicateglasses [34/borosilicate]. Strong glassborosilicate/borosilicate bonding without annealingglass bonding was alsowithout achieved annealing by washing was also substrates achieved withby washing sulfuric substrates acid and high-flow-ratewith sulfuric acid tap and water high [35-flow]. Although-rate tap water the mechanism [35]. Although is still the not mechanism understood, is still the not higher understood, bonding energythe higher of borosilicate bonding energy glass of compared borosilicate with glass that compared of fused-silica with glass that mayof fused support-silica the glass validity may support of these previousthe validity studies of these on theprevious low-temperature studies on the bonding low-temperature of borosilicate bonding glass. of borosilicate glass.

Figure 6. 6.Bonding Bonding energies energie of fused-silicas of fused/fused-silica,-silica/fused fused-silica-silica, fused/borosilicate-silica/borosilicate and borosilicate and/ borosilicate/ glassborosilicate substrates glass at asubstrates heating temperature at a heating of temperature 110 ◦C, measured of 110 by ° crack-openingC, measured by test. crack- opening test. Finally, we examined the pressure resistance of a micro/nanofluidic device by fluorescence microscopy,Finally, which we examined previous studiesthe pressure have employed resistance to of evaluate a micro/nanofluidic the bonding strength device [ 23by–25 fluor]. Figureescence7a showsmicroscopy, the experimental which previous setup. studies The devicehave employed was fabricated to evaluate by bonding the bonding a fused-silica strength glass[23–25 substrate]. Figure with7a shows microchannels the experimental (width: setup. 500 µ m,The depth: device 10 wasµm) fabricated and nanochannels by bonding (width: a fused 50-silicaµm, depth: glass substrate 390 nm), aswith shown microchannels in Figure7 (width:b, and a500 0.17-mm μm, depth: borosilicate 10 μm) and glass nanochannels substrate without (width: any 50 μ patterning.m, depth: 390 In nm), the bondingas shown process, in Figure the 7 heatingb, and a temperature 0.17-mm borosilicate was set at 200glass◦C. substrate A fluorescein without solution any patterning. was injected In intothe thebonding nanochannels process, the via heating the microchannel. temperature Since was theset pressureat 200 °C. loss A fluorescein in microchannels solution is was negligibly injected small into comparedthe nanochannels to that in via nanochannels, the microchannel. an external Since pressure the pressure applied loss to in the microchannels device generated is negligibly by the pressure small controllercompared wasto that regarded in nanochannels, as that applied an to external the nanochannels. pressure applied Fluorescence to the images device of generated the nanochannel by the atpressure applied controller pressures was of 150 regarded and 600 as kPa that were applied obtained, to the as nanochannels. shown in Figure Fluorescence7c. images of the nanochannelFigure7d at shows applied profiles pressures of theof 150 fluorescence and 600 kPa intensity were obtained, of the nanochannel.as shown in Figure The area7c. of the nanochannelFigure 7d shows shows bright profiles fluorescence, of the fluorescen while thatce outsideintensity the of nanochannel the nanochannel. is dark, The with area an imageof the intensitynanochannel similar shows tothe bright background. fluorescence, In addition,while that the outside image the intensity nanochannel outside is the dark nanochannel, with an image was constantintensity evensimilar when to the the background. applied pressure In addition, was increased the image from intensit 150 toy 600 outside kPa. Thesethe nanochannel results indicate was thatconstant no detectable even when leakage the applied of the pressure fluorescein was solution increased from from the 150 nanochannel to 600 kPa. occurred These results and bondingindicate ofthat the no glass detectable substrates leakage was of achieved. the fluorescein Therefore, solution it is from concluded the nanochannel that the proposed occurred bonding and bonding process of wasthe glass successfully substrates applied was achieved to the micro. Therefore,/nanofluidic it is concluded device made that of the fused-silica proposed and bonding borosilicate process glass, was andsuccessfully the device applied had a to pressure the micro/nanofluidic resistance higher device than made 600 kPa. of fused Micro-silica/nanofluidic and borosilicate devices fabricatedglass, and the device had a pressure resistance higher than 600 kPa. Micro/nanofluidic devices fabricated by the proposed bonding process can be used for various microfluidic applications [1,2], nanofluidic applications with pressure-driven flows generated by high external pressures on the order of 100 kPa, such as single-molecule immunoassays [6], femtoliter solvent extraction [8], and single-cell target

Micromachines 2020, 11, 804 9 of 12 by the proposed bonding process can be used for various microfluidic applications [1,2], nanofluidic applicationsMicromachines 20 with20, 11 pressure-driven, x FOR PEER REVIEW flows generated by high external pressures on the order of 1009 kPa,of 12 such as single-molecule immunoassays [6], femtoliter solvent extraction [8], and single-cell target proteomicsproteomics [[99],], andand otherother nanofluidicnanofluidic applicationsapplications suchsuch asas electrophoreticelectrophoretic single-moleculesingle-molecule sortingsorting [[55]] andand miniaturizedminiaturized fuelfuel cellscells [[1010].]. In addition,addition, the construction of devicesdevices mademade ofof fused-silicafused-silica andand borosilicateborosilicate glassglass enablesenables aa combination of micromicro/nanofluidics/nanofluidics andand advancedadvanced opticaloptical microscopy suchsuch asas super-resolutionsuper-resolution microscopymicroscopy [[36,3736,37]] withoutwithout aa reductionreduction inin performance.performance.

FigureFigure 7.7. ExaminationExamination of of the pressure resistance of a micromicro/nanofluidic/nanofluidic devicedevice byby fluorescencefluorescence microscopy.microscopy. ((aa)) ExperimentalExperimental setup.setup. ((b)) Cross-sectionalCross-sectional profileprofile ofof aa nanochannelnanochannel measuredmeasured byby stylusstylus surfacesurface profiler.profiler. ((cc)) FluorescenceFluorescence imagesimages ofof nanochannelnanochannel wherewhere fluoresceinfluorescein solutionsolution waswas injectedinjected byby externalexternal pressure.pressure. ((dd)) ProfilesProfiles ofof fluorescentfluorescent intensityintensity inin transversetransverse directiondirection ofof nanochannelnanochannel (A-A (A-A’).0).

4.4. Conclusions WeWe developeddeveloped aa simplesimple processprocess forfor low-temperaturelow-temperature glassglass bondingbonding utilizingutilizing typicaltypical facilitiesfacilities availableavailable inin cleanclean roomsrooms andand appliedapplied itit to the fabrication of micromicro/nanofluidic/nanofluidic devicesdevices byby thethe bonding ofof fused-silicafused-silica/fused/fused-silica-silica glass glass substrates substrates and fused-silicaand fused/borosilicate-silica/borosilicate glass substrates. glass substrates The proposed. The bondingproposed process bonding was process verified, was and verified the bonding, and energy the bonding for fused-silica energy/ fused-silicafor fused-silica/fused substrates at-silica the 2 annealingsubstrates temperatureat the annealing of 200–400temperature◦C was of 200 0.33–0.48–400 °C Jwas/m ,0.33 which–0.48 is J/m comparable2, which is tocomparable that reported to that in previousreported studies.in previous We found studies. that We a moderate found that concentration a moderate of oxygenconcentration plasma withof oxygen a pressure plasma of 51 with Pa for a activatingpressure of the 51 surface Pa for ofactivating glass substrates the surface could of achieve glass substrate strong bonding.s could Theachieve enhancement strong bonding. of bonding The strengthenhancement by activating of bonding the surfacestrength with by activating oxygen plasma the surface was significant with oxygen for annealing plasma was at 200 sign◦C,ificant while for it wasannealing insignificant at 200 ° atC, 400while◦C. it These was insignificant results provide at 400 new °C. insights These results to help provide understanding new insights the bonding to help mechanism.understanding The the optimized bonding bonding mechanism process. wasTheapplied optimized to the bonding fabrication process of micro was/nanofluidic applied devicesto the madefabrication of a fused-silicaof micro/nanofluidic glass substrate devices and made a borosilicate of a fused-silica glass glass substrate substrate (0.17 and mm a borosilicate thick), which glass is generallysubstrate used(0.17 formm microscopic thick), which optical is generally measurements. used for In microscopic the bonding optical process, measurements using an annealing. In the temperaturebonding process, lower using than 200an annealing◦C was important temperature to avoid lower crack than generation 200 °C was by important thermal stress to avoid due tocrack the digenerationfference in by thermal thermal expansion stress due coe toffi thecients difference between in fusedthermal silica expansion and borosilicate. coefficient Ats between an annealing fused 2 temperaturesilica and borosilicate of 110 ◦C,. theAt an bonding annealing energy temperature was around of 0.5110 J /°cmC, theand bonding three times energy higher was than around that 0.5 of fused-silicaJ/cm2 and three/fused times silica higher glass bonding. than that The of fused results-silica/fused suggest that silica the bonding glass bonding. energy depends The results on the suggest glass materialthat the andbonding increases energy in casedepends of using on borosilicatethe glass material glass. Weand succeeded increases inin devicecase of fabrication, using borosilicate and the fabricatedglass. We succeeded micro/nanofluidic in device device fabrication had a, pressureand the fabricated resistance micro/nanofluidic higher than 600 kPa, device which had is a su pressurefficient forresistance various higher micro/nanofluidic than 600 kPa, applications. which is Insufficient previous for studies various [23 –micro/nanofluidic25], conditions of low-temperatureapplications. In previous studies [23–25], conditions of low-temperature bonding for fused-silica/fused-silica glass substrates utilizing oxygen plasma with fluorine atoms were investigated to achieve bonding energy of 0.45–1.00 J/m2, which corresponds to pressure resistance of nanochannels on the order of 100–1000 kPa. In the present study, we developed conditions of low-temperature glass bonding utilizing oxygen plasma without fluorine atoms and verified bonding of fused-silica/fused-silica and fused-

Micromachines 2020, 11, 804 10 of 12 bonding for fused-silica/fused-silica glass substrates utilizing oxygen plasma with fluorine atoms were investigated to achieve bonding energy of 0.45–1.00 J/m2, which corresponds to pressure resistance of nanochannels on the order of 100–1000 kPa. In the present study, we developed conditions of low-temperature glass bonding utilizing oxygen plasma without fluorine atoms and verified bonding of fused-silica/fused-silica and fused-silica/borosilicate glass substrates. Bonding energy in similar order to that by previous studies was achieved. The results provide a simple and easy process of low-temperature glass bonding for the fabrication of micro/nanofluidic devices, enable the use of high-resolution optical microscopy designed for borosilicate cover glass, and will greatly contribute to advancements in micro/nanofluidics.

Author Contributions: Conceptualization, supervision, funding acquisition and writing—original draft preparation, Y.K.; experiments, K.S., M.T. and K.M. All authors have read and agreed to the published version of the manuscript. Funding: This research was funded by JSPS KAKENHI Grant Numbers JP19H02061, JP19K21930 and TEPCO Memorial Foundation. Acknowledgments: The authors would like to thank Kyojiro Morikawa at The University of Tokyo for his assistance in conducting experiments and fruitful discussion on the relationship between the bonding energy and the conditions of oxygen plasma and annealing. Fabrication facilities were provided in part by the Academic Consortium for Nano and Micro Fabrication of four universities (The University of Tokyo, Tokyo Institute of Technology, Keio University and Waseda University, JAPAN). This work was also conducted at Takeda Sentanchi Supercleanroom, The University of Tokyo, supported by “Nanotechnology Platform Program” of the Ministry of Education, Culture, Sports, Science and Technology (MEXT), Japan, Grant Number JPMXP09F-19-UT-0109. Conﬂicts of Interest: The authors declare no conﬂict of interest.

References

1. Janasek, D.; Franzke, J.; Manz, A. Scaling and the design of miniaturized chemical-analysis systems. Nature 2006, 442, 374–380. [CrossRef][PubMed] 2. Mawatari, K.; Kazoe, Y.; Aota, A.; Tsukahara, T.; Sato, K.; Kitamori, T. Microflow systems for chemical synthesis and analysis: Approaches to full integration of chemical process. J. Flow Chem. 2011, 1, 3–12. [CrossRef] 3. Choi, N.W.; Cabodi, M.; Held, B.; Gleghorn, J.P.; Bonassar, L.J.; Stroock, A.D. Microfluidic scaffolds for tissue engineering. Nat. Mater. 2007, 6, 908–915. [CrossRef][PubMed] 4. Mawatari, K.; Kazoe, Y.; Shimizu, H.; Pihosh, Y.; Kitamori, T. Extended-nanofluidics: Fundamental technologies, unique liquid properties, and application in chemical and bio analysis methods and devices. Anal. Chem. 2014, 86, 4068–4077. [CrossRef] 5. Cipriany, B.R.; Murphy, P.J.; Hagarman, J.A.; Cerf, A.; Latulippe, D.; Levy, S.L.; Benítez, J.J.; Tan, C.P.; Topolancik, J.; Soloway, P.D.; et al. Real-time analysis and selection of methylated DNA by fluorescence- activated single molecule sorting in a nanofluidic channel. Proc. Natl. Acad. Sci. USA 2012, 109, 8477–8482. [CrossRef][PubMed] 6. Shirai, K.; Mawatari, K.; Ohta, R.; Shimizu, H.; Kitamori, T. A single-molecule ELISA device utilizing nanofluidics. Analyst 2018, 143, 943–948. [CrossRef] 7. Ishibashi, R.; Mawatari, K.; Kitamori, T. Highly efficient and ultra-small volume separation by pressure-driven liquid chromatography in extended nanochannels. Small 2012, 8, 1237–1242. [CrossRef] 8. Kazoe, Y.; Ugajin, T.; Ohta, R.; Mawatari, K.; Kitamori, T. Parallel multiphase nanofluidics utilizing nanochannels with partial hydrophobic surface modification and application to femtoliter solvent extraction. Lab Chip 2019, 19, 3844–3852. [CrossRef] 9. Nakao, T.; Kazoe, Y.; Mori, E.; Morikawa, K.; Fukasawa, T.; Yoshizaki, A.; Kitamori, T. Cytokine analysis on a countable number of molecules from living single cells on nanofluidic devices. Analyst 2019, 144, 7200–7208. [CrossRef] 10. Pihosh, Y.; Uemura, J.; Turkevych, I.; Mawatari, K.; Kazoe, Y.; Smirnova, A.; Kitamori, T. From extended nanofludics to an autonomous solar-light-driven micro fuel-cell device. Angew. Chem. 2017, 56, 8130–8133. [CrossRef] Micromachines 2020, 11, 804 11 of 12

11. Bhattacharjee, N.; Urrios, A.; Kang, S.; Folch, A. The upcoming 3D-printing revolution in microfluidics. Lab Chip 2016, 16, 1720–1742. [CrossRef][PubMed] 12. Grilli, S.; Coppola, S.; Nasti, G.; Vespini, V.; Gentile, G.; Ambrogi, V.; Carfagna, C.; Ferraro, P. Hybrid ferroelectric-polymer microfluidic device for dielectrophoretic self-assembling of nanoparticles. RSC Adv. 2014, 4, 2851–2857. [CrossRef] 13. Coppola, S.; Nasti, G.; Todino, M.; Olivieri, F.; Vespini, V.; Ferraro, P. Direct writing of microfluidic footpaths by pyro-EHD printing. ACS Appl. Mater. Interfaces 2017, 9, 16488–16494. [CrossRef][PubMed] 14. Whitesides, G.M.; Ostuni, E.; Takayama, S.; Jiang, X.; Ingber, D.E. Soft lithography in biology and biochemistry. Annu. Rev. Biomed. Eng. 2001, 3, 335–373. [CrossRef][PubMed] 15. Tokeshi, M.; Minagawa, T.; Uchiyama, K.; Hibara, A.; Sato, K.; Hisamoto, H.; Kitamori, T. Continuous-flow chemical processing on a microchip by combining microunit operations and a multiphase flow network. Anal. Chem. 2002, 74, 1565–1571. [CrossRef] 16. Hibara, A.; Saito, T.; Kim, H.-B.; Tokeshi, M.; Ooi, T.; Nakao, M.; Kitamori, T. Nanochannels on a fused-silica microchip and liquid properties investigation by time-resolved fluorescence measurements. Anal. Chem. 2002, 74, 6170–6176. [CrossRef] 17. Renberg, B.; Sato, K.; Tsukahara, T.; Mawatari, K.; Kitamori, T. Hands on: Thermal bonding of nano- and microfluidic chips. Microchim. Acta 2009, 166, 177–181. [CrossRef] 18. Mellors, J.S.; Gorbounov, V.; Ramsey, R.S.; Ramsey, J.M. Fully integrated glass microfluidic device for performing high-efficiency capillary electrophoresis and electrospray ionization mass spectrometry. Anal. Chem. 2008, 80, 6881–6887. [CrossRef] 19. 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] 20. Chen, L.; Luo, G.; Liu, K.; Ma, J.; Yao, B.; Yan, Y.; Wang, Y. Bonding of glass-based microfluidic chips at low- or room-temperature in routine laboratory. Sens. Actuators B Chem. 2006, 119, 335–344. [CrossRef] 21. Fonslow, B.R.; Bowser, M.T. Free-flow electrophoresis on an anodic bonded glass microchip. Anal. Chem. 2005, 77, 5706–5710. [CrossRef][PubMed] 22. Queste, S.; Salut, R.; Clatot, S.; Rauch, J.-Y.; Khan Malek, C.G. Manufacture of microfluidic glass chips by deep plasma etching, femtosecond laser ablation, and anodic bonding. Microsyst. Technol. 2010, 16, 1485–1493. [CrossRef] 23. Xu, Y.; Wang, C.; Dong, Y.; Li, L.; Jang, K.; Mawatari, K.; Suga, T.; Kitamori, T. Low-temperature direct bonding of glass nanofluidic chips using a two-step plasma surface activation process. Anal. Bioanal. Chem. 2012, 402, 1011–1018. [CrossRef][PubMed] 24. Xu, Y.; Wang, C.; Matsumoto, N.; Jang, K.; Dong, Y.; Mawatari, K.; Suga, T.; Kitamori, T. Bonding of glass

nanofluidic chips at room temperature by a one-step surface activation using an O2/CF4 plasma treatment. Lab Chip 2013, 13, 1048–1052. [CrossRef][PubMed] 25. Ohta, R.; Mawatari, K.; Takeuchi, T.; Morikawa, K.; Kitamori, T. Detachable glass micro/nanofluidic device. Biomicrofluidics 2019, 13, 024104. [CrossRef][PubMed] 26. Howlader, M.M.R.; Suehara, S.; Takagi, H.; Kim, T.H.; Maeda, R.; Suga, T. Room-temperature microfluidics packaging using sequential plasma activation process. IEEE. Trans. Adv. Packag. 2006, 29, 448–456. [CrossRef] 27. Xu, J.; Wang, C.; Wang, T.; Wang, Y.; Kang, Q.; Liu, Y.; Tian, Y. Mechanisms for low-temperature direct

bonding of Si/Si and quartz/quartz via VUV/O3 activation. RSC Adv. 2018, 8, 11528–11535. [CrossRef] 28. Wang, C.; Fang, H.; Zhou, S.; Qi, X.; Niu, F.; Zhang, W.; Tian, Y.; Suga, T. Recycled low-temperature direct bonding of Si/glass and glass/glass chips for detachable micro/nanofluidic devices. J. Mater. Sci. Technol. 2020, 46, 156–157. [CrossRef] 29. Kazoe, Y.; Mawatari, K.; Sugii, Y.; Kitamori, T. Development of a measurement technique for ion distribution in an extended nanochannel by super-resolution-laser-induced fluorescence. Anal. Chem. 2011, 83, 8152–8157. [CrossRef] 30. Kazoe, Y.; Pihosh, Y.; Takahashi, H.; Ohyama, T.; Sano, H.; Morikawa, K.; Mawatari, K.; Kitamori, T. Femtoliter nanofluidic valve utilizing glass deformation. Lab Chip 2019, 19, 1686–1694. [CrossRef] 31. Morikawa, K.; Matsushita, K.; Tsukahara, T. Rapid plasma etching for fabricating fused silica microchannels. Anal. Sci. 2017, 33, 1453–1456. [CrossRef][PubMed] 32. Maszara, W.P.; Goetz, G.; Caviglia, A.; McKitterick, J.B. Bonding of silicon wafers for silicon-on-insulator. J. Appl. Phys. 1988, 64, 4943–4950. [CrossRef] Micromachines 2020, 11, 804 12 of 12

33. Zhuravlev, L.T. The surface chemistry of amorphous silica. Zhuravlev model. Colloids Surf. A Physicochem. Eng. Asp. 2000, 173, 1–38. [CrossRef] 34. Allen, P.B.; Chiu, D.T. Calcium-assisted glass-to-glass bonding for fabrication of glass microfluidic devices. Anal. Chem. 2008, 80, 7153–7157. [CrossRef] 35. Jia, Z.-J.; Fang, Q.; Fang, Z.-L. Bonding of glass microfluidic chips at room temperatures. Anal. Chem. 2004, 76, 5597–5602. [CrossRef] 36. Hell, S.W. Toward fluorescence nanoscopy. Nat. Biotechnol. 2003, 21, 1347–1355. [CrossRef] 37. Rust, M.J.; Bates, M.; Zhuang, X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nat. Methods 2006, 3, 793–795. [CrossRef]

© 2020 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/).