Micro Scale
Micro Scale
a r t i c l e
i n f o
a b s t r a c t
Commonly used assembly methods for microuidic devices rely heavily on direct, manual bonding of different components with limited visual aids. The manual operation is likely to cause damage to existing structures, especially to nanoscale sensors on the substrate. Here we report a novel approach to integrate nanosensors into a lab-on-a-chip device with total elimination of operational errors from the manual bonding process. The microuidic components are composed of an ultraviolet (UV) light-dened, cross-linked SU-8 microchannel and a polydimethylsiloxane (PDMS) top cover. The hybrid microuidic structure provides a fully sealed microchannel and precisely positioned features. Well-organized singlewalled carbon nanotube (SWNT) thin lms are deposited and aligned across the electrodes on a silicon substrate with dielectrophoresis. The assembly of SWNTs is carried out in a sealed microchannel which eliminates the potential damage of the nanosensors during the bonding process. The SWNT devices are congured as two-terminal resistor-type sensors with the metal electrodes as the probing pads and the dielectrophoretically captured SWNTs as the sensing elements. To demonstrate the feasibility of this integration approach, an assembled SWNT device is measured as an integrated ow sensor to monitor the ow rate in the microchannel. This lab-on-a-chip device is designed as a platform that can be expanded for more applications. For example, with minimal modications, the device can be used in chemical sensing, biosensing, and medical research. 2012 Elsevier B.V. All rights reserved.
Article history: Received 3 September 2011 Received in revised form 8 February 2012 Accepted 19 February 2012 Available online 24 February 2012 Keywords: Lab-on-a-chip device Nanosensor Microchannel Bonding Single-walled carbon nanotube (SWNT) Dielectrophoresis
1. Introduction The innovations in microuidics and microuidic devices have given sensor technology new meaning and new opportunities. For example, a microuidic exchange system superimposed on a sensor can actively transport target analytes to the sensor surface, enabling the detection of analytes at ultra-low concentrations [1,2]. The successful integration of sensors and microuidics has led to the creation of complex lab-on-a-chip microsystems, providing a low-cost, low-volume alternative approach to traditional methods used in analytical chemistry. Research activities and advances of lab-on-a-chip devices have revolutionized the biomedical research eld. Many innovative devices with a wide range of functionalities have been successfully developed and used in various applications in the past decade. For example, such lab-on-a-chip devices have been developed for drug delivery [3], neuron manipulation [4], bioand immune-assays [5], and fuel cell-based energy conversion [6]. The latest advances in the fabrication and applications of lab-ona-chip devices can be found in a number of recent review articles [79].
Corresponding author. Tel.: +1 360 546 9250; fax: +1 360 546 9438. E-mail address: [email protected] (W. Xue). 0925-4005/$ see front matter 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.snb.2012.02.047
A common method to fabricate lab-on-a-chip devices is to bond plasma-treated polydimethylsiloxane (PDMS) to solid substrates. This method has proven effective in forming fully sealed microuidic channels on silicon and glass substrates. However, the existing assembly technology relies heavily on direct, manual bonding of different parts with limited visual aids. When sensors and electronic components are included in the system, the bonding process becomes unreliable due to the lack of precise positioning control mechanisms. With the critical dimensions at the micro- and nano-scales, the bonding process often causes problems such as part-to-part misalignment, damage to the sensors, and leakage from the channels. Consequently, these processes can only be used to create individual devices; reliable, high-yield integration methods of nanosensors and microuidics need to be further explored. Alternative approaches have been investigated by researchers to obtain strong bonding between various substrates including thermoplastics and epoxy polymers [10,11]. In particular, the use of photo-denable polymers such as SU-8 photoresist has provided positioning capabilities in the bonding process [1215]. Single-walled carbon nanotubes (SWNTs), one of the most studied 1D cylindrical nanomaterials, are well known for their remarkable mechanical, electrical, and chemical properties. The inherent properties such as large surface areas and high conductivity make SWNTs promising candidates for biosensing applications
871
[16,17]. The SWNTs exhibit excellent sensing performance due to the large number of sites on their tubular surfaces where gaseous and aqueous molecules can react. The adsorption of molecules onto the surface of an SWNT alters its electrical properties, resulting in a detectable resistance change [18,19]. Compared with the traditional uorescent labeling method, the detection of resistance change of an SWNT is label free and requires less sample preparation. In addition, the recorded signals are easier to interpret and the required testing system is simpler with lower cost. The effective combination of the SWNTs with microuidics represents a signicant step towards real-time and label-free detection in biomedical research. The integration of SWNT-based nanosensors into lab-on-achip microsystems has been investigated by various groups. Many innovative devices have been developed. For example, Bourlon et al. studied electrokinetic phenomena at the nanoscale using a nanotube transistor-based probe [20]. Basuray et al. reported an open-ow impedance-sensing platform for DNA hybridization on a nanotube surface [21]. Li et al. developed a nanotube-based microuidic platform. With proper modications, the platform has been used as a chemical sensor [2], an air micro-ow sensor [22], and an aqueous shear stress sensor [23]. These reported devices have demonstrated interesting functionalities and satisfactory performance. However, most of these devices are based on the direct bonding of PDMS on glass substrates. The yield and success rate of the bonding processes remain unknown. In this work, we investigate the effective integration of nanomaterials, electronics and nanosensors, and microuidic structures into a lab-on-a-chip device. In particular, we develop a novel approach to integrate nanosensors into the microuidic device after the microchannel is sealed. The nal device has a hybrid structure with a sealed microchannel, well-aligned features, and precisely positioned nanosensors. The microchannel is constructed using ultraviolet (UV) light-patterned, cross-linked SU-8 photoresist on a silicon wafer. Using SU-8 in the process ensures the perfect positioning of the microchannel relative to the positions for nanosensors. A silane solution, 3-aminopropyltriethoxysilane (APTES), is used to introduce amine groups (Si NH2 ) to the SU8 surface. The microchannel is sealed by bonding a PDMS cover (with surface silanol groups: Si OH) to the SU-8 structure based on strong, irreversible Si O Si covalent bonds between the two materials. A solution containing dispersed SWNTs is injected into the microchannel and the SWNTs are deposited between pre-fabricated electrode pairs using dielectrophoresis. For demonstration purposes, the aligned SWNTs are used to monitor the ow rate in the microchannel. The integrated lab-on-a-chip device is designed as a sensing platform which has great potential for future expansion in various applications.
Fig. 1. Schematic of a lab-on-a-chip device with integrated carbon nanotube-based sensors. (a) Isometric view and (b) side view.
2. Materials and device design The substrates used in this work were 4-in. silicon wafers with surface coating of 200 nm thermal oxide; they were purchased from Silicon Inc. (Boise, ID). The elastomer PDMS, Sylgard 184 from Dow Corning (Midland, MI), was provided in a two-part silicone elastomer kit with a base polymer and a curing agent. The two components were mixed together with a weight ratio of 10:1 to form PDMS. The air bubbles generated during the mixing step were removed using a centrifuge rotating at 2500 rpm for 5 min. The mixed PDMS was poured over a silicon mold and cured at 60 C for 8 h in an oven. After curing, the PDMS was peeled off from the mold and cut into small pieces. The high-viscosity negative photoresist SU-8 3050 and SU-8 developer were purchased from MicroChem Corp. (Newton, MA) and used as is. The surface treating silane solution APTES (H2 N(CH2 )3 Si(OC2 H5 )3 ; purity 98%), purchased from
SigmaAldrich Co. (St. Louis, MO), was diluted to 5% in volume with deionized (DI) water. The pristine SWNTs (outer diameter < 2 nm; length 515 m; purity > 90%), in the form of powder, were purchased from SES Research (Houston, TX). Because dielectrophoresis is a solutionbased method, it requires the deposited material to be able to move and rotate freely in a medium. Therefore, the pristine SWNTs were treated with a mixture of strong acids H2 SO4 :HNO3 (volume ratio 3:1) at 110 C for 45 min to increase their solubility in DI water [24]. Repeated ltration steps were carried out to remove the acid residual from the SWNTs. The acid-treated SWNTs were diluted with DI water to a concentration of 0.2 mg/ml. Fig. 1 shows the schematic isometric and side views of a lab-ona-chip device, which is composed of microuidic components and a sensor array. The device is fabricated on a Si/SiO2 wafer. A layer of SU-8 photoresist is patterned on the surface with a microchannel and two reservoirs for uid ow. The dimensions of the microchannel are 1.8 cm long, 500 m wide, and 40 m deep. The diameter of each reservoir is 4 mm. A PDMS piece with a ow inlet and an outlet is used as a top cover to seal the microchannel. A sensor array composed of 1119 individual sensors is integrated in the microuidic channel. These sensors are congured as two-terminal resistor-type sensors with each sensor consisting of two metal electrodes and an aligned SWNT thin lm. The electrodes are used as probing pads to connect the device to external electronics for testing. The SU-8 structure and the PDMS cover are patterned to expose the electrodes for the electronic connection. The SWNTs are rst assembled across the electrodes using dielectrophoresis and later used as the sensing elements for ow rate detection. The entire labon-a-chip device is approximately 3 cm long, 2 cm wide, and 3 mm thick. 3. Experimental methods The lab-on-a-chip device design contains components with varying dimensions ranging from macro- to nano-scales. A fully functional device would require the precise positioning of all the components and the seamless interface among structures with different dimensions. Therefore, a reliable and repeatable fabrication process is the key for the creation of such a device. In our design,
872
the inner tips of the electrodes for the dielectrophoretic assembly of SWNTs in a later step. 3.2. SU-8 to PDMS bonding The heterogeneous bonding between thermoplastics and PDMS has been investigated previously [10,11]. However, the method presented in these early reports was only used to evaluate the bonding strength of different substrates without considering the integration of nanosensors. Nevertheless, it is a suitable approach that can be used to bond SU-8 and PDMS with the capability to integrate nanosensors. It is also a critical step towards the damage-free integration of nanosensors and microuidics in the fabrication of lab-on-a-chip devices. Because the two materials have different properties in terms of surface chemistry, additional treatment is needed to achieve strong bonding between them. Fig. 4 illustrates the surface treatment process of the two materials. The cross-linked SU-8 layer is treated with a two-step process before bonding. First, it is etched by oxygen plasma at 40 W for 20 s (chamber pressure 120 mTorr). This etching process can activate the SU-8 layer, resulting in a hydrophilic SU-8 surface. Second, the plasma-treated device is soaked in a 5% APTES solution, heated at 80 C on a hot plate, for 20 min. The APTES solution introduces a silylated layer on the substrate, forming amine groups (Si NH2 ) on the SU-8 surface. The plasma- and APTEStreated SU-8 sample is rinsed under DI water for 10 s and dried with compressed air.
a patterned SU-8 photoresist layer is used as the microchannel structure. As a result, the fabrication can rely on well-developed lithography systems to minimize errors caused by the manual operation steps during bonding. 3.1. Lab-on-a-chip device fabrication The fabrication process of the lab-on-a-chip device is illustrated in Fig. 2. First, the process starts with a 4-in. Si/SiO2 wafer (Fig. 2(a)). Second, metal layers of Cr (100 nm, adhesion material) and Au (200 nm, electrode material) are coated on the wafer surface with sputtering. The metals are patterned as electrodes with UV lithography and wet etching (Fig. 2(b)). Third, the SU-8 3050 photoresist is spin coated on the surface at a rotation speed of 3000 rpm for 30 s, resulting in a 40 m thick layer. The substrate is soft baked at 95 C for 13 min on a hot plate followed by 20 s of UV light exposure in a hard-contact aligner (OAI 200 Mask Aligner). A post exposure bake step, 95 C for 4 min on a hot plate, is carried out to harden the SU8 layer. The device is soaked in the SU-8 developer for 8 min with steady hand shaking of the solution. The development step dissolves the protected SU-8 layer under the photomask, leaving the exposed, cross-linked SU-8 structures on the surface of the substrate (Fig. 2(c)). Fourth, a piece of PDMS is bonded to the SU-8 structure to seal the microchannel (Fig. 2(d)). The detailed bonding process will be described and discussed in the next section. Fifth, a small volume of SWNT dispersion is introduced into the microchannel for the assembly of SWNTs using dielectrophoresis (Fig. 2(e)). Sixth, multiple sensors can be fabricated using the same dielectrophoresis process (Fig. 2(f)). Because all the sensors are fabricated separately and each one of them is individually controlled, this lab-on-a-chip device has the potential to be congured as a multiplexed sensor array or addressable electronics. A device with fabricated SU-8 microuidic components is shown in Fig. 3. Most of the area is covered by the cross-linked SU8. The light-patterned SU-8 structure consists of a microchannel, two reservoirs, and two open windows for sensor pads. Because the device is fabricated using the standard lithography techniques, alignment marks are included in the design for precise wafer-to-photomask registration during the fabrication process. The microchannel is placed on top of the sensor array, exposing
Fig. 4. Plasma and chemical treatment of SU-8 and PDMS to obtain a hybrid, bonded structure.
873
The PDMS piece only needs to be treated once. It is activated by oxygen plasma at 40 W for 20 s with a chamber pressure of 120 mTorr, resulting in a surface with silanol groups. Immediately after the treatment of both SU-8 and PDMS, the two pieces are pressed together manually and placed on a hot plate for bonding. The optimized bonding temperature and time for our experiments are 70 C and 10 min, respectively. During the bonding process, a stainless steel cylinder weighing 42 N is placed on top of the sample to ensure a conformal contact between the SU-8 and the PDMS layers. This cylinder has a large surface area (approximately 10 cm in diameter) that can cover the entire surface of the chip. After the treatment steps, both SU-8 and PDMS contain surface coating with similar surface chemistry properties. During the bonding process, the activated amine groups and silanol groups condense to reduce their surface free energy. Consequently, Si O Si covalent bonds are formed between the SU-8 and PDMS pieces, leading to strong, irreversible bonding of the two materials. 3.3. Formation of inlet and outlet in PDMS The inlet and outlet of the lab-on-a-chip device can be obtained by simply punching small vertical through holes (shown in Fig. 1(b)) with a needle in the PDMS cover. However, the inlet and outlet created this way sometimes have slight leaking problems in longterm tests, especially at high ow rates. In addition, we notice that the vertically attached tubing is difcult to maneuver and can be torn apart from the chip occasionally. Alternatively, the inlet/outlet can be included in the photomask layout design and later patterned into horizontal SU-8 structures. However, the embedding of small tubes into the chip can be problematic. Here we use an approach that involves an embedded tube inside PDMS, a vertical through hole, and a sealing PDMS piece to form the inlet/outlet. The device has a sandwiched structure, as shown in Fig. 5(a). The tube (Dow Corning Silastic Laboratory Tubing, Midland, MI) is cut into small pieces and inserted into the liquid PDMS before the curing process. One side of the tube is tied into a knot to prevent the lling of PDMS into the tube; the other side is cut open as the ow inlet. A sharp needle is used to punch a vertical through hole that goes through both the PDMS piece and the embedded tube. After bonding the PDMS to the bottom SU-8 device, the tube is connected to the microchannel through the punched hole. A second piece of PDMS is bonded to the rst one to seal the hole. The fabrication of such a sandwiched structure is relatively simple and it enables the precise interconnecting interface for structures at different scales. Fig. 5(c) shows a fully assembled device using this method. The inlet/outlet needles can be easily inserted into the embedded tubes for uid transport. 3.4. Nanosensor assembly in a sealed microchannel with dielectrophoresis Sensor fabrication and microchannel bonding are two key steps in the development of lab-on-a-chip devices. In many cases, the sensors, especially nanosensors, need to be created before the bonding process. However, the manual operation during bonding can cause damage to the sensors. Furthermore, the required plasma/chemical treatment on the sample surface may change the properties of the sensors and affect their performance. Therefore, ideally the sensor fabrication should take place after the microchannel is sealed. We believe that this sequence can minimize, or even eliminate, the damage to the sensors during the bonding process. Dielectrophoresis, a solution-based nanomaterial assembly method, provides a suitable approach to fabricate nanosensors in sealed microchannels. It is a simple electrical technique using alternating current (AC) signals. It has been successfully used in the controlled assembly of nanowires [25], carbon nanotubes [26],
Fig. 5. (a) Isometric and (b) side views of the ow inlet and outlet design. (c) An assembled lab-on-a-chip device next to a dime.
and graphene [27]. The mechanisms of dielectrophoresis have been extensively studied experimentally and numerically [28,29]. A major advantage of dielectrophoresis is that it can be easily integrated into high-yield device fabrication [30,31]. During the dielectrophoresis process, the AC source applied across the electrodes generates an electric eld in the solution. The charges including electrons and protons in the dispersed SWNTs are moved away from their initial balanced positions. This charge redistribution creates electric dipole moments which force the SWNTs to rotate along the electric eld lines. Strong and concentrated electric elds are generally preferred in the dielectrophoresis process because the net force, called dielectrophoretic force, applied to an SWNT is proportional to the eld strength. As a result, the electrodes are designed with narrow tips at the end to induce highly concentrated electric elds. Two types of electrode patterns are included in our device [32,33]. One is singletip (10 m width) electrodes with a 4 m gap between an opposite pair. The other one contains multiple (321) tips on each electrode with the beam width of 3 m and the electrode gap of 3 m. The multiple-tip design is used in this study to assemble SWNTs because our initial investigation shows that it generally has higher yield and better reproducibility than the single-tip design. Fig. 6 shows the experimental system conguration for the assembly of SWNTs using dielectrophoresis. The device is placed on a probe station and is connected to external electronics through
874
micropositioner-controlled metal probes. A function generator (Agilent Technologies 81150A) is used as the AC signal source to generate an electric eld between the electrodes. The frequency f and the peak-to-peak voltage VPP of the signal are 5 MHz and 10 V, respectively. An oscilloscope (Agilent Technologies MSO 7054A) is used to monitor the real-time potential change across the electrodes. A small volume of SWNT dispersion is injected into the microchannel for the assembly of nanotubes. When the SWNTs are assembled across the electrodes, a small potential drop of 0.52 V can be observed from the oscilloscope. The entire dielectrophoresis process takes approximately 3060 s. The SWNT dispersion is removed from the microchannel using a syringe pump. 4. Results and discussion Optical microscopy (Nikon NM-40) and scanning electron microscopy (SEM, FEI Quanta 3D 200i) are used to inspect the fabricated electrodes and the assembled SWNTs on the device. Fig. 7(a) shows an optical image of an electrode array with a multiple-tip design in the center. The design contains 21 pairs of tips which are evenly distributed along the edges of two 400 m wide electrode pads, as shown in Fig. 7(b). Fig. 7(c) shows a close-up view of the assembled SWNTs between an individual pair of tips. The SWNTs are in the form of small bundles and they can only be observed in between the electrode tips. This demonstrates that the dielectrophoresis is an active approach to assemble SWNTs and only place them on locations where the electric elds have high magnitude. Furthermore, we should point out that even though the exact number of aligned SWNTs varies from one pair to another, the alignment results and the amount of aligned SWNTs are relatively consistent. The real-time potential change shown on the oscilloscope is rst recorded to a video and later extracted as data points for analysis. Fig. 8 shows the extracted voltage as a function of time during the dielectrophoresis process for a multiple-tip device. The function generator is turned on at the beginning. The probes are lowered to make contact with the electrode pads at the tenth second, applying AC signals to the device. A small potential drop of 0.5 V is observed within a 20 s period, indicating the assembly of SWNTs across the electrodes. The SWNT dispersion is then removed by slowly pumping air through the microchannel. A second potential drop of 0.5 V can be observed and the voltage stabilizes at 9.5 V after another 20 s. The air-pumping step is critical for the SWNTs to precipitate from the dispersion and form strong bonds to the substrate. Similar measurement results are obtained from other devices, including single-tip devices. For example, the inset of Fig. 8 illustrates the voltagetime plot of a single-tip device. We believe that both the dielectrophoresis and liquidsolid interface during the dispersion removal play important roles in the assembly process of the SWNTs. The dielectrophoretic forces rotate the SWNTs to align with the electric eld lines and move them to the vicinity of the electrode gap. During this period, the electrodes
Fig. 7. (a) An optical image of the electrode array with one multiple-tip pattern and six single-tip patterns. (b) An SEM image of the multiple-tip electrode design. (c) The assembled SWNTs between a pair of tips.
can capture some SWNTs which are loosely bonded to the surface while others are still dispersed in the solution. In the dispersion removal step, the liquid is pushed away from the electrode area by the air but the dielectrophoretic forces are able to hold the positions of the SWNTs. As the liquid residue dries in the air, the capillary force (F 1/r, where r is the distance between the SWNTs and the substrate) gets stronger due to the decreasing distance r. Finally it pulls the SWNTs tightly onto the substrate surface.
Fig. 8. Real-time potential change during the dielectrophoresis process for a multiple-tip device. The inset shows the measurement result of a single-tip device.
875
Fig. 9. Measurement of resistance of the assembled SWNTs. (a) Currentvoltage plots of a device before (black squares) and after (red circles) the thermal treatment. (b) Measured resistance of six randomly selected devices. The rst set of values is obtained right after the assembly of SWNTs on the substrate (solid bars). The second set is obtained after the heating (patterned bars). (For interpretation of the references to color in this gure legend, the reader is referred to the web version of the article.)
Fig. 10. Flow rate sensing using an integrated SWNT sensor. (a) Recorded resistance of a sensor over time at various ow rates. (b) Resistance change versus ow rate of this sensor.
A thermal treatment step is conducted to further dry the bonding surface of the SWNTs. The device is heated at 100 C for 10 min on a hot plate. The medium-range temperature is chosen due to the presence of SU-8 and PDMS on the device. This step drives off water vapor from the microchannel and further promotes adhesion of the assembled SWNTs to the substrate. After the dielectrophoresis process, the gap between the electrodes is covered with the assembled SWNTs, which allow the transmission of charge carriers across the electrodes. Therefore, the devices can be measured as regular resistors with the electrodes as the probing pads and dielectrophoretically assembled SWNTs as the conductive material. In our experiments, the currentvoltage (IV) characteristics of each device are obtained by a semiconductor device analyzer (Agilent Technologies B1500A). Fig. 9(a) shows the IV plots of a multiple-tip device before (solid black squares) and after (red circles) the heat treatment. The plots are highly linear in the range of 2 to 2 V, suggesting that the assembled SWNTs are mostly metallic instead of semiconducting. This is because the metallic SWNTs have a larger ClausiusMossotti factor fcm , and the dielectrophoretic forces exerted on them are much stronger than those experienced by the semiconducting SWNTs [28]. Therefore, the metallic SWNTs dominate the movement of SWNT bundles in
the dielectrophoresis process. The device shows a slight increase in resistance after the thermal treatment. This is due to the change of environment from humid to dry for the assembled SWNTs. Similar changes can be observed in other devices. Fig. 9(b) compares the measurement results of six randomly selected devices. The solid bars show the resistances of the assembled SWNTs right after the dielectrophoresis process; the patterned bars illustrate the corresponding values after the heating step. Changes in resistance have been observed in all six devices with ve of them showing a resistance increase with one showing a resistance decrease. Although the changes are not uniform from one device to another, the nal resistances of all six devices are in a relatively dened range of 4001600 . We believe that with optimized parameters and a more controlled procedure during the dielectrophoresis process, the variation of the resistance can be minimized. The hybrid SU-8/PDMS lab-on-a-chip device is a complex microsystem which can be used as a platform to study uidic dynamics at the micro/nano-scale or as sensors for sensing applications. As an initial demonstration, we use the device to monitor the ow rate in the microchannel. The electrodes with the assembled SWNTs can be used as integrated ow sensors. Fig. 10(a) shows the recorded real-time resistance reading of one sensor in DI water with changed ow rates. The ow rate starts at 10 l/min and then is increased sequentially to 100, 200, 300, and 400 l/min. The rate is then reduced sequentially through the same values. The measurement period at each ow rate is approximately 60 s. The recorded
876
data demonstrate that the ow rate can signicantly affect the electrical properties of the SWNTs. A higher ow rate can increase the resistance of the SWNTs. The relationship between the resistance change and ow rate appears nonlinear, as shown in Fig. 10(b). An increase of 4.34% in resistance is obtained when the ow rate is increased from 10 to 400 l/min. Similar effects have been reported by other groups, indicating that nanotubes and nanowires can be used to monitor the ow rate [20,34]. However, due to the fact that the solutions used in their experiments are different from ours, the fundamental sensing principles are expected to be entirely different. In these previous reports, the devices were congured as eld-effect transistors and the sensing of ow rate is achieved through the detection of streaming potential in the solution. Ionic solutions were used to induce the electrical double layer at the surface of the substrate, where the sensors are located. By contrast, the solution under test in our experiments is DI water and the devices are congured as twoterminal resistor-type sensors. Due to the signicantly reduced ion concentration and the gate-free device conguration, the inuence from the electrical double layer is therefore negligible. We believe that the resistance change in our experiments is caused by the structural deformation of the SWNTs under uid ow. Previous research has demonstrated, both theoretically and experimentally, that the distortion of an individual SWNT could change its electromechanical characteristics, leading to higher resistance at higher bending and twisting angles [35,36]. Our measurement results of higher resistance at higher ow rates agree with these early ndings. However, the results are still preliminary. A thorough and in-depth investigation will be conducted in our laboratory to further explore the underlying principles of these SWNT-based ow sensors.
References
[1] G. Zheng, F. Patolsky, Y. Cui, W.U. Wang, C.M. Lieber, Multiplexed electrical detection of cancer markers with nanowire sensor arrays, Nat. Biotechnol. 23 (2005) 12941301. [2] L. Yang, M. Li, Y. Qu, Z. Dong, W.J. Li, Carbon nanotube-sensor-integrated microuidic platform for real-time chemical concentration detection, Electrophoresis 30 (2009) 31983205. [3] N.-C. Tsai, C.-Y. Sue, Review of MEMS-based drug delivery and dosing systems, Sens. Actuators A: Phys. 134 (2007) 555564. [4] J. Wang, L. Ren, L. Li, W. Liu, J. Zhou, W. Yu, D. Tong, S. Chen, Microuidics: a new cosset for neurobiology, Lab. Chip 9 (2009) 644652. [5] E.D. Goluch, J.-M. Nam, D.G. Georganopoulou, T.N. Chiesl, K.A. Shaikh, K.S. Ryu, A.E. Barron, C.A. Mirkin, C. Liu, A bio-barcode assay for on-chip attomolarsensitivity protein detection, Lab. Chip 6 (2006) 12931299. [6] E. Kjeang, N. Djilali, D. Sinton, Microuidic fuel cells: a review, J. Power Sources 186 (2009) 353369. [7] P. Abgrall, A.M. Gu, Lab-on-chip technologies: making a microuidic network and coupling it into a complete microsystema review, J. Micromech. Microeng. 17 (2007) R15R49. [8] M.B. Fox, D.C. Esveld, A. Valero, R. Luttge, H.C. Mastwijk, P.V. Bartels, A. Berg, R.M. Boom, Electroporation of cells in microuidic devices: a review, Anal. Bioanal. Chem. 385 (2006) 474485. [9] H. Becker, L.E. Locascio, Polymer microuidic devices, Talanta 56 (2002) 267287. [10] V. Sunkara, D.-K. Park, H. Hwang, R. Chantiwas, S.A. Soper, Y.-K. Cho, Simple room temperature bonding of thermoplastics and poly(dimethylsiloxane), Lab. Chip 11 (2011) 962965. [11] M.E. Vlachopoulou, A. Tserepi, P. Pavli, P. Argitis, M. Sanopoulou, K. Misiakos, A low temperature surface modication assisted method for bonding plastic substrates, J. Micromech. Microeng. 19 (2009) 015007. [12] S. Talaei, O. Frey, P.D. van der Wal, N.F. de Rooij, M. Koudelka-Hep, Hybrid microuidic cartridge formed by irreversible bonding of SU-8 and PDMS for multi-layer ow applications, Proc. Chem. 1 (2009) 381384. [13] S. Tuomikoski, S. Franssila, Free-standing SU-8 microuidic chips by adhesive bonding and release etching, Sens. Actuators A: Phys. 120 (2005) 408415. [14] R.J. Jackman, T.M. Floyd, R. Ghodssi, M.A. Schmidt, K.F. Jensen, Microuidic systems with on-line UV detection fabricated in photodenable epoxy, J. Micromech. Microeng. 11 (2001) 263269. [15] S. Li, C.B. Freidhoff, R.M. Young, R. Ghodssi, Fabrication of micronozzles using low-temperature wafer-level bonding with SU-8, J. Micromech. Microeng. 13 (2003) 732738. [16] E. Katz, I. Willner, Biomolecule-functionalized carbon nanotubes: applications in nanobioelectronics, ChemPhysChem 5 (2004) 10841104. [17] W. Yang, P. Thordarson, J.J. Gooding, S.P. Ringer, F. Braet, Carbon nanotubes for biological and biomedical applications, Nanotechnology 18 (2007) 412001. [18] T. Zhang, S. Mubeen, N.V. Myung, M.A. Deshusses, Recent progress in carbon nanotube-based gas sensors, Nanotechnology 19 (2008) 332001. [19] C.B. Jacobs, M.J. Peairs, B.J. Venton, Review: carbon nanotube based electrochemical sensors for biomolecules, Anal. Chim. Acta 662 (2010) 105127. [20] B. Bourlon, J. Wong, C. Mik, L. Forr, M. Bockrath, A nanoscale probe for uidic and ionic transport, Nat. Nanotechnol. 2 (2007) 104107. [21] S. Basuray, S. Senapati, A. Aijian, A.R. Mahon, H.-C. Chang, Shear and AC eld enhanced carbon nanotube impedance assay for rapid, sensitive, and mismatch-discriminating DNA hybridization, ACS Nano 3 (2009) 18231830. [22] W.W.Y. Chow, Y. Qu, W.J. Li, S.C.H. Tung, Integrated SWCNT sensors in microwind tunnel for air-ow shear-stress measurement, Microuid Nanouid 8 (2009) 631640. [23] Y. Qu, W.W.Y. Chow, M. Ouyang, S.C.H. Tung, W.J. Li, X. Han, Ultra-low-powered aqueous shear stress sensors based on bulk EG-CNTs integrated in microuidic systems, IEEE Trans. Nanotechnol. 7 (2008) 565572. [24] W. Xue, T. Cui, Characterization of layer-by-layer self-assembled carbon nanotube multilayer thin lms, Nanotechnology 18 (2007) 145709. [25] S. Raychaudhuri, S.A. Dayeh, D. Wang, E.T. Yu, Precise semiconductor nanowire placement through dielectrophoresis, Nano Lett. 9 (2009) 22602266. [26] M. Dimaki, P. Bggild, Dielectrophoresis of carbon nanotubes using microelectrodes: a numerical study, Nanotechnology 15 (2004) 10951102. [27] A. Vijayaraghavan, C. Sciascia, S. Dehm, A. Lombardo, A. Bonetti, A.C. Ferrari, R. Krupke, Dielectrophoretic assembly of high-density arrays of individual graphene devices for rapid screening, ACS Nano 3 (2009) 17291734. [28] D. Padmaraj, W. Zagozdzon-Wosik, L.M. Xie, V.G. Hadjiev, P. Cherukuri, J. Wosik, Parallel and orthogonal E-eld alignment of single-walled carbon nanotubes by ac dielectrophoresis, Nanotechnology 20 (2009) 035201. [29] N. Peng, Q. Zhang, J. Li, N. Liu, Inuences of ac electric eld on the spatial distribution of carbon nanotubes formed between electrodes, J. Appl. Phys. 100 (2006) 024309. [30] B.R. Burg, F. Lutolf, J. Schneider, N.C. Schirmer, T. Schwamb, D. Poulikakos, High-yield dielectrophoretic assembly of two-dimensional graphene nanostructures, Appl. Phys. Lett. 94 (2009) 053110. [31] A.H. Monica, S.J. Papadakis, R. Osiander, M. Paranjape, Wafer-level assembly of carbon nanotube networks using dielectrophoresis, Nanotechnology 19 (2008) 085303. [32] P. Li, W. Xue, Selective deposition and alignment of single-walled carbon nanotubes assisted by dielectrophoresis: from thin lms to individual nanotubes, Nanoscale Res. Lett. 5 (2010) 10721078.
5. Conclusions In conclusion, we have developed a novel approach to integrate nanosensors into microuidic devices without damaging the devices. Using this approach, we have fabricated a hybrid lab-ona-chip device using well-organized microuidic components and precisely positioned nanosensors. The microchannel is fabricated using a patterned SU-8 photoresist structure and a PDMS cover. Plasma treatment and silane solution soaking are needed to form such a hybrid structure. The SWNTs are deposited on the substrate using dielectrophoresis in the sealed microchannel. The electric elds generated by an external AC source can attract the SWNTs to the sensors and eventually assemble them across the electrodes. The results from the microscopy inspection and electrical characterization demonstrate that the dielectrophoresis process is an active approach to assemble SWNTs and only place them on locations where the electric elds have high magnitude. The SWNTbased sensors can be used as integrated ow sensors to monitor the ow rate in the microchannel. The lab-on-a-chip device presented in this study can be used as a platform which has high potential for future expansion in a wide range of applications, especially in the studies of uidic dynamics at the micro/nano-scale and the research in the biomedical eld.
Acknowledgments This research is supported by a Washington State University (WSU) Vancouver Faculty Research Mini-Grant and partially supported by a DARPA Young Faculty Award.
P. Li et al. / Sensors and Actuators B 166167 (2012) 870877 [33] P. Li, C.M. Martin, K.K. Yeung, W. Xue, Dielectrophoresis aligned single-walled carbon nanotubes as pH sensors, Biosensors 1 (2011) 2335. [34] D.R. Kim, C.H. Lee, X. Zheng, Probing ow velocity with silicon nanowire sensors, Nano Lett. 9 (2009) 19841988. [35] T.W. Tombler, C. Zhou, L. Alexseyev, J. Kong, H. Dai, L. Liu, C.S. Jayanthi, M. Tang, S.-Y. Wu, Reversible electromechanical characteristics of carbon nanotubes under local-probe manipulation, Nature 405 (2000) 769772. [36] A. Rochefort, P. Avouris, F. Lesage, D.R. Salahub, Electrical and mechanical properties of distorted carbon nanotubes, Phys. Rev. B 60 (1999) 1382413830.
877
Debra A. Sheadel received the BS in pulp and paper science and the BS in chemical engineering from the University of Washington, Seattle, WA, in 1995. Currently she is a graduate student of medical engineering at the University of Washington, Seattle, WA. Her research interest is the integration of microuidics and sensors for biological and chemical applications. Jie Xu received the BS in thermal engineering from Tsinghua University, Beijing, China, in 2005, and the PhD degree in mechanical engineering from Columbia University, New York, NY, in 2010. He is currently an assistant professor of mechanical engineering at Washington State University, Vancouver, WA. His research interest includes micro interfacial sciences and micro/nanouidic systems for health, energy, and environmental applications. His research results have been published in more than 30 technical papers in international journals and conference proceedings. He is a member of ASME, AAAS, APS, IEEE, and Sigma Xi. Wei Xue received the BS and the MS degrees in electrical engineering from Shandong University, Jinan, China, in 1997 and 2000, respectively, and the PhD degree in mechanical engineering from the University of Minnesota, Minneapolis, MN, in 2007. He is currently an assistant professor of mechanical engineering at Washington State University, Vancouver, WA. His main research interest includes microfabrication, nanotechnology, polymer/silicon microelectromechanical systems (MEMS), micro/nano electronics, and chemical/biological sensors. He is a member of ACS, ASME, IEEE, and Sigma Xi.
Biographies
Pengfei Li received the BE in energy and power engineering from Shandong University, Jinan, Shandong, China, in 2007 and the MS degree in mechanical engineering from Washington State University, Vancouver, WA, in 2011. Currently he is pursuing his PhD degree at the Department of Mechanical Engineering, Carnegie Mellon University, Pittsburgh, PA. His main research interest includes carbon nanotube- and graphene-based sensors and electronics. Nan Lei received the BS in aerospace engineering from Beihang University, Beijing, China, in 2010. Currently he is a graduate student of mechanical engineering at Washington State University, Vancouver, WA. His research interest is developing carbon-based microelectromechanical systems (MEMS) for biological and chemical applications.