Optical immunosensor using carbon nanotubes coated with a photovoltaic polymer

Biosensors and Bioelectronics 34 (2012) 208–214

Contents lists available at SciVerse ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

Optical immunosensor using carbon nanotubes coated with a photovoltaic polymer Joon S. Shim ∗ , Chong H. Ahn ∗ Department of Electrical and Computer, University of Cincinnati, Cincinnati, OH, 45221, USA

a r t i c l e

i n f o

Article history: Received 25 October 2011 Received in revised form 29 January 2012 Accepted 6 February 2012 Available online 14 February 2012 Keywords: Optical immunosensor Carbon nanotube Nano-biosensor Point-of-care testing

a b s t r a c t In this work, an on-chip optical immunosensor using an individually assembled carbon nanotube (CNT) coated with a photovoltaic polymer has been proposed, developed, characterized, and applied for the detection of cardiac biomarkers. An individual CNT was self-assembled on a nickel (Ni)-patterned electrode by magnetically attracting the residual iron catalyst at one end of the CNT. After the CNT self-assembled electrode was prepared, it was coated with a photovoltaic polymer to implement a CNT photodetector. Under an incident light, the photovoltaic polymer generated electrons that changed the conductivity of the CNT. The CNT photodetector was finally insulated with parylene to prevent interruptions of charged molecules in a sample solution, such as non-specifically bound proteins and various ions. Chemiluminescent immunoassay was directly performed on the CNT photodetector for an on-chip detection of cardiac troponin T (cTnT) with a detection limit of 12 pg/mL. High sensitivity and reliable selectivity have been achieved through the use of on-chip measurement of chemiluminescent light by the CNT photodetector. As a result, the developed device is envisaged as a new platform for optical immunosensing using the individually self-assembled CNT for point-of-care (POC) clinical diagnostics. Published by Elsevier B.V.

1. Introduction Over the recent decades, carbon nanotubes (CNTs) have been explored for highly sensitive sensors that detect various types of biomolecules such as cells, proteins, and DNA (Panchapakesan et al., 2005; Chen et al., 2003; Star et al., 2006). With the basic structure of the field-effect transistor (FET) assembled with an ideal 1D structure of CNT, the charged molecules around the CNT modulate an extremely narrow current path through the CNT (Wanekaya et al., 2006). As a result, the CNT-based biosensors have achieved a highly sensitive detection of target proteins with a simple device structure. Additionally, the electrical sensing mechanism can be flexibly realized with electronic components, allowing a miniaturized device and simple steps for biological analysis. In spite of these advances in the use of CNT-based biosensors, a practical application of the CNT biosensor for clinical diagnosis is still retarded. Because the FET-type CNT biosensors respond very sensitively to most surrounding charges, it is difficult to selectively detect target molecules without interruptions from ions or non-specifically bound proteins in a sample solution (Collins et al., 2000; Cui et al., 2001). To improve the selectivity of the CNT biosensor for detecting target molecules, various efforts have been made through

∗ Corresponding author. E-mail addresses: [email protected] (J.S. Shim), [email protected] (C.H. Ahn). 0956-5663/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.bios.2012.02.004

functionalizing the surface of CNTs with surfactants (Chen et al., 2003), proteins (Kim et al., 2009), and polymers (Star et al., 2003; Shim et al., 2002). Particularly, polymers such as polyethylene glycol (PEG) (Star et al., 2003) and polyethylenimine (PEI) (Shim et al., 2002) were used to completely coat the surface of the CNT, leading to the selective detection of target proteins. Even though these functionalizing techniques on the surface of the CNT significantly reduced the non-specific binding of proteins, the detection limit of the functionalized CNT has not yet been proved to fully explore the extremely high sensitivity of the CNT (Zheng et al., 2005). It should be noted that the disturbance from various ions in a sample solution is a more serious problem in realizing a reliable biosensor than the disturbance from non-specifically bound proteins. Many reports have shown that the electrical properties of CNTs are extremely sensitive to the interaction of ionic molecules around the CNT (Cui et al., 2001; Collins et al., 2000). Owing to the smaller size of the ionic molecules, the surface of the CNT should be thoroughly coated with functionalizing materials for blocking the intervention of the ions. Moreover, human blood contains various ionic molecules, making the isolation of the CNT from these molecules essential for the application of the CNT-assembled device to a clinical diagnosis. Thus, reliable selectivity without the loss of sensitivity for target molecules needs to be achieved in order to implement a practical biosensor with CNTs. In this work, an innovative mechanism for the CNT-based immunosensor has been proposed and developed to realize the selective detection of target proteins while maintaining high sensitivity. To achieve an excellent selectivity as well as

J.S. Shim, C.H. Ahn / Biosensors and Bioelectronics 34 (2012) 208–214

209

sensitivity, a new optical immunosensor utilized a photovoltaic polymer-coated CNT which served as a CNT-based photodetector. Chemiluminescent-based immunoassay was performed directly on the CNT photodetector. When light was produced from the chemiluminescent assay, the generated charges in the photovoltaic polymer accumulated around the surface of the CNT, resulting in a change of the CNT conductivity. Because the generated charges increased with an increase in the concentration of target proteins by the chemiluminescent immunoassay, the CNT photodetector served as an on-chip optical immunosensor. Moreover, since the CNT photodetector was insulated with parylene, the signal variation by ionic molecules were completely blocked, thereby finally achieving highly selective biosensing. With this configuration of the biosensor, a cardiac biomarker (cTnT) could be successfully detected with an extremely low concentration of 12 pg/mL. Thus, the FET-type CNT biosensor with high selectivity and sensitivity has been successfully realized for the detection of a cardiac biomarker, and it would also be widely applicable for detecting various types of proteins and peptides for point-of-care (POC) clinical diagnostics. 2. Principle 2.1. Chemiluminescent immunoassay In order to achieve highly selective and sensitive biosensing of a target molecule, the principle of the proposed biosensor consists of the generation of chemiluminescent light and the detection of the chemiluminescent light by a CNT photodetector, which are fully performed on a lab-on-a-chip (LOC) platform. For the procedure of chemiluminescent immunoassay, the capturing antibody is functionalized inside a microfluidic chamber, which is built on the CNT photodetector. While the sample solution flows to the functionalized region, the target antigen in a sample is tethered by the capturing antibody. Because the chemiluminescent enzyme of horse redox peroxidase (HRP) is conjugated with the secondary antibody, the capturing antibody and the secondary antibody are linked by the target antigen, finally cross-linking the HRP. When this linked HRP reacts with the enzyme substrate, chemiluminescent light is generated. Since the intensity of this optical signal is proportional to the number of linked HRPs, which increases with the concentration of the target antigen, the intensity of the chemiluminescent light also increases according to the concentration of the target antigen. As a result, the target antigen can be detected by measuring the intensity of chemiluminescent light by the CNT photodetector. 2.2. On-chip optical sensing through photovoltaic-polymer-coated CNTs To achieve on-chip signal sensing, the self-assembled carbon nanotube was coated with a photovoltaic polymer. Fig. 1 shows a schematic illustration of light detection by the photovoltaic-polymer-coated CNT. The photovoltaic polymer of poly(m-phenylenevinylene-co-2,5-dioctoxy-p-phenylenevinylene) (PmPV) produces excitons under the incident light. Because a valence band of the PmPV is aligned with the CNT, the holes in the produced excitons are injected into the carbon nanotube, leaving the electrons in the vicinity of the CNT (Star et al., 2004). Then, the electrons are trapped at the photovoltaic polymer–SiO2 interface, and provide a gate potential to the CNT (Borghetti et al., 2006; Shi et al., 2008). As a result, the incident light modulates a current flowing through the self-assembled CNT between two electrodes, enabling a detection of the chemiluminescent light by measuring the conductivity of the CNT. Additionally, the UV–vis spectra of PmPV show the peak absorption at a wavelength of 450 nm (Star

Fig. 1. Device concept. (a) Conceptual diagram of the CNT photodetector insulated with parylene and (b) schematic illustration of charge generation in PmPV changing the conductivity of CNT.

et al., 2004). This selective absorption of a specific wavelength is well matched with a wavelength of 430 nm produced by HRP. Thus, the optical response of the PmPV-coated CNT leads to a highly selective and sensitive detection of biological molecules by chemiluminescent immunoassay. 2.3. Parylene insulation In this work, the developed CNT photodetector was insulated with parylene to implement a selective biosensor with a long-term stability. Parylene has a great optical transparency for a wide range of light wavelengths (Jeong et al., 2002), which makes it very desirable as a coating layer for optical detection by CNT photodetectors. In addition, its biocompatible and non-toxic material properties provide an excellent platform for on-chip immunosensors. Owing to these properties, the capturing proteins were immobilized on parylene, which served as an intermediate moderator to provide a protein-binding surface on the PmPV-coated CNT (Chang et al., 2007). Finally, as an insulation layer, parylene improves the longterm stability of the fabricated device and completely isolates the device from disturbances from ionic molecules, which significantly enhances the selectivity of CNT biosensors for highly accurate clinical diagnostics (Schmidt, 1983; Loeb et al., 1977) 3. Material and methods 3.1. Fabrication of individual CNT-assembled electrode A gap between the electrodes and the self-aligned Ni pattern was fabricated by the previously reported technique of controlled undercut and metallization (Shim et al., 2010). The precise alignment of the Ni pattern on the electrode was achieved without delicate aligning procedures, and a sub-micron-sized gap between electrodes was successfully fabricated using standard i-line optical lithography with the 365 nm wavelength, which is usually available in most clean-room facilities. Thus, this fabrication technique provided an inexpensive fabrication of the CNT-assembled device,

210

J.S. Shim, C.H. Ahn / Biosensors and Bioelectronics 34 (2012) 208–214

Fig. 2. Image of the fabricated electrode. Sub-micron gap of electrode with self-aligned Ni pattern was fabricated by optical lithography. The Ni self-aligned electrode with nanogap was applied to a self-assembled SWNT by magnetic capturing and fluidic alignment.

suitable for a disposable-type biosensor. As shown in Fig. 2, the fabricated device had 20 pairs of electrodes with nano-scale gaps. Each electrode has a self-aligned Ni pattern to magnetically assemble the single-walled nanotubes (SWNT) at the edge of the Ni pattern. The vertical-array-type SWNT was purchased from FirstNano (FirstNano Inc., USA). The SWNT had a length of 60 ␮m and was synthesized by a water-assisted CVD process (Hata et al., 2004). The SWNT array was sonicated in a 1 wt% mixture of DI water and Tween 20 for 30 min. After the SWNT solution was placed on the magnet for 24 h to magnetically separate the metal impurities (Shim et al., 2009), the top half of the SWNT solution was carefully pipetted and used for device fabrication. To enable the flow of the CNT solution onto the Ni self-aligned electrodes, soft lithography with PDMS was performed using an AZ4620 photoresist to fabricate a microchannel with a thickness of 20 ␮m. The regular bonding procedure of the oxygen plasma treatment was not carried out to avoid the permanent bonding of PDMS to the SiO2 substrate. Without permanent bonding, the SWNT solution flowed through the microchannel without any leakage. After the self-assembly of the SWNT, the PDMS microchannel was detached from the substrate for subsequent procedures. The SWNT was self-assembled at the Ni pattern by the previously developed method (Shim et al., 2009). An external magnetic field (0.7 T) was applied to the device, while the dispersed SWNT flowed through the devices. Using the induced magnetic force, the Fe catalyst at the end of the SWNT was captured at the edge of the Ni pattern and aligned parallel to the flow direction. After flowing 2 ␮L

of the SWNT dispersed solution with a flow velocity of 4.2 mm/s, 80% of the electrode was assembled with the SWNTs.

3.2. Fabrication of CNT photodetector The SWNT-assembled electrode was coated with a photovoltaic polymer by a drop-casting method (Star et al., 2004). The photovoltaic polymer of PmPV was dissolved in chloroform with a 1 wt% concentration. After fully dissolving the PmPV, a 1 ␮L volume of PmPV in chloroform was pipetted and then carefully dropped on the assembled SWNT. By evaporating the chloroform at room temperature, the SWNT-assembled electrode was uniformly coated with PmPV. The PmPV-coated device was insulated with parylene by a vaporized deposition process (Parylene Coater, Model PDS 2010 Labcoter 2, USA). A raw parylene pellet (1.5 g) was deposited to attain a 2 ␮m-thick parylene layer on the PmPV-coated SWNT. Parylene was coated on the whole device with a uniform thickness. After parylene was deposited on the device, oxygen plasma etching (March CS1701 RIE, USA) was performed to etch out parylene at the contact electrodes that were connected to the measuring equipment (Meng et al., 2008). Because the photoresist was also etched during the oxygen plasma, a thick photoresist of AZ4620 was patterned on the device with a thickness of 20 ␮m to protect the sensing area. After etching out parylene at the contact electrode, the photoresist was rinsed by acetone, methanol, and DI water.

J.S. Shim, C.H. Ahn / Biosensors and Bioelectronics 34 (2012) 208–214

211

3.3. Equipment setup

3.4. Chemiluminescent immunoassay

To electrically characterize our developed device, the contact pad of the CNT assembled electrode was connected with an external detection circuit utilizing a spring-loaded electrode connector. Then, the device was enclosed in a shielding box which was covered with an aluminum foil to reduce electromagnetic interference (EMI). After the electrical cables and the fluidic tubes were connected through small holes at the box, the shielding box was tightly sealed to prevent the external light during the measurement. By connecting the device with the outside Picoammeter (Keith 780, Keithly Inc., USA), the current–voltage (I–V) characteristics were measured for the developed devices. To reduce the loss of chemiluminescent light, on-chip detection was achieved by performing immunoassay directly above the parylene-insulated CNT photodetector. For this purpose, a microchannel with a 20 ␮m height was attached on the device, as shown in Fig. 3. For the procedures of ELISA, the inlet and outlet of the microchannel was connected with the syringe pump (Harvard Apparatus Inc., USA) outside the shielding box by flexible silicone tubes. Thus, the equipment for the electrical measurement and the flow control could be executed outside the shielded device, which also blocked external light during the optical measurement.

Standard steps of chemiluminescent immunoassay were performed inside the microchannel for the detection of the cTnT protein, as illustrated in Fig. 3. To enable the flow of the reagents through the attached PDMS microchannel and to prevent the microchannel from leaking, a suction pressure was applied to the outlet of the microchannel, and the reagents were pulled into the microchannel. As a first step of the ELISA procedure, 5 ␮L of the capturing antibody (cardiac Troponin T antibody, AbCam Inc., USA) was injected into the microchannel and immobilized for 15 min. The parylene coated surface provides a suitable platform for protein immobilization because of protein-fouling on the hydrophobic polymer surface (Chang et al., 2007). After the capturing antibody was successfully immobilized on the detection microchannel, non-immobilized antibodies were washed out by applying a washing buffer through the microchannel for 30 s with the flow velocity of 10 ␮L/min. To prevent any further immobilization of proteins to the surface of microchannel, 5 ␮L of the protein blocking solution (PBS, Thermo Fisher Scientific Inc.) was injected to the detection chamber and stayed for 15 min, which allows the blocking proteins to cover the empty spot of microchannel where

Fig. 3. On-chip chemiluminescent immunoassay inside the microchannel. (a) Schematic cut-view of on-chip chemiluminescent reaction, (b) picture of attached PDMS microchannel on the CNT photodetector for a microfluidic control. (c and f) Procedures of chemiluminescent immunoassay. (c) Immobilizing 1st antibody on the surface of the parylene-coated CNT photodetector, (d) capturing target protein (cTnT) from the sample, (e) immobilizing the HRP-conjugated 2nd antibody on cTnT, and (f) chemiluminescent reaction with the substrate solution by the enzyme of HRP.

J.S. Shim, C.H. Ahn / Biosensors and Bioelectronics 34 (2012) 208–214

the capturing antibody was not immobilized. Following these steps to functionalize the microchannel, 3 ␮L of the target sample containing the cTnT antigen (cardiac Troponin T protein, AbCam Inc., USA) was applied and immobilized for 3 min. After washing out the microchannel, 3 ␮L of the HRP-conjugated secondary antibody (cardiac Troponin T antibody (HRP), AbCam, USA) was flowed into the microchannel and retained for 3 min. Then, the HRP substrate was applied to the microchannel, and the chemiluminescent light was detected by measuring the conductivity change in the PmPV-coated CNT. The HRP substrate (SuperSignal ELISA Femto Maximum Sensitivity Substrate, Thermo Fisher Scientific Inc, USA) was prepared by mixing Luminol/Enhancer solution and Peroxide solution with 1:1 ratio. In order to sustain the uniform chemiluminescent intensity during the signal measurement, 20 ␮L of the prepared substrate was flowed through the microchannel with the flow velocity of 5 ␮L/min. The enzyme reaction was performed inside a microchannel at room temperature and the external light was shielded by placing the developed device inside a black EMI shielding box.

(a)

0.8 0.6 0.4

Current (uA)

212

0.2 0 -0.2 -0.4

CNT -0.6

PmPV

-0.8 -4

-3

-2

-1

0

1

2

3

4

Voltage (V) 5

(b)

4

4. Results and discussion

3

4.1. Electrical characterization

Current (uA)

1 0 -1 -2 -3

Light_On Light_Off

-4 -5 -4

-3

-2

-1

0

1

2

3

4

Voltage (V)

1000

(c) Current Change (ΔI/I)

The CNT-assembled electrode and the PmPV-coated electrode without CNTs were electrically characterized, as plotted in Fig. 4(a). The PmPV-coated electrode without CNTs did not show any response to the incident light. In addition, the conductivity change in bare CNTs without PmPV coating was not detected for the incident light. These results showed that the electrical response for the incident light was mainly attributed to the interaction between PmPV and CNTs. Fig. 4(b) shows the electrical response under the incident light, where the light has a 365 nm wavelength and 8.4 mW/cm2 intensity. The conductivity of the PmPV-coated CNT changed in response to the incident light as the charges from the PmPV surrounded the assembled CNT. To achieve a highly sensitive detection of chemiluminescent light, the electrical conditions need to be optimized by maximizing the output signal of the CNT photodetector for the incident light. For this purpose, the drain-source voltage (Vds ) of the device was characterized to increase the current change ((I/I = ((Ion − Ioff )/Ioff )) from the device, as plotted in Fig. 4(c). Previous reports have shown that the signal from the CNT-FET sensors can be maximized at the sub-threshold regime (Heller et al., 2009). Our characterization result also agreed that the optimized range of the drain-source voltage should be chosen at the sub-threshold regime. Additionally, when the output signal was too low in spite of the largest change in conductivity, it was too difficult to differentiate the signal from the base noise-level. Considering these signal conditioning factors, 0.4 V of the drain-source voltage (Vds ) was selected to improve the detection limit of the developed CNT immunosensor.

2

800

600

400

200

0

-4

-3

-2

-1

0

1

2

3

4

-200

4.2. Chemiluminescent light detection Fig. 5(a) shows the conductance change of our optical immunosensor while the HRP enzyme substrate flowed through the microchannel. For different concentrations of the target antigen of cTnT, the target biomarker of cTnT was caught by the capturing antibody, which was immobilized above the CNT photodetector. Because cTnT provides the cross-linking site to the secondary antibody-conjugated HRP in the procedures of ELISA, the number of the tethered HRP is proportional to the concentration of cTnT, resulting in the increased intensity of chemiluminescent light. As expected, Fig. 5(b) clearly shows the increasing signal change according to the increase in the cTnT concentration. In

Voltage (V) Fig. 4. Electrical characterizations of the CNT photodetector. (a) I–V curve from the bare CNT and the PmPV without CNT, (b) conductivity change for the CNT coated with PmPV under the incident light, and (c) characterization of Vds for maximum output.

our experiments, cTnT with a concentration of 12 pg/mL has been detected by the developed CNT immunosensor. In the standard setup for measuring the chemiluminescent light, the photodetector was placed outside the reaction chamber which was fabricated with transparent substrates such as glass and plastics. Compared with the result from this type of off-chip

Conductance Change (ΔS/S)

J.S. Shim, C.H. Ahn / Biosensors and Bioelectronics 34 (2012) 208–214

213

2.5 2 1.2 ug/ml

1.5

120 ng/ml 12 ng/ml 1.2 ng/ml

1

120 pg/ml 12 pg/ml

0.5

Control

0 0

5

10

15

20

Time (sec)

(a)

Conductance Change (ΔS/S)

10

1

0.1 0.001

0.1

10

1000

Concentration (ng/ml) (b) Fig. 5. Conductivity change of photovoltaic polymer-coated SWNT according to the concentration of cTnT. (a) Real-time signal change during enzyme substrate flowing and (b) conductance change according to the concentration of cTnT.

detection, the limit of detection (LOD) by the direct on-chip measurement was much lower than the one by the off-chip detection (Silva et al., 2010). This higher performance of the developed device was attributed to the prevention of light loss by scattering through the substrate. If the chemiluminescent light was measured outside the reaction chamber, the transparent substrate of the reaction chamber acted likes an optical waveguide to disperse the generated optical signal. Therefore, the optical signal transferring to the outside photodetector was significantly reduced from the originally generated chemiluminescent light. In addition, the high cost of sensitive photodetectors such as photomultiplier tubes (PMT) and avalanche photodiodes makes it difficult to integrate the photodetectors inside the reaction chamber as a disposable device for on-chip detection. However, the developed immunosensor used

a thin film (∼2 ␮m) of transparent parylene to enable the chemiluminescent immunoassay directly above the CNT photodetector. As a result, the light loss through the substrate was considerably decreased, allowing the sensitive detection of chemiluminescent light, even for extremely low concentrations of the target antigen. Furthermore, the implementation of the device involved low-cost procedures for the application to a disposable-type device. In addition, the parylene was utilized in this work to completely isolate the CNT photodetector from the surrounding environment. Specifically in this work, the ELISA procedure required consecutive injections of multiple solutions containing various ions and proteins. Thus, uncontrollable noises could be generated by ionic conduction at non-passivated electrodes, non-specific absorption of proteins at the CNT, and ionic gating effect on the CNT FET

214

J.S. Shim, C.H. Ahn / Biosensors and Bioelectronics 34 (2012) 208–214

5. Conclusions

1

Conductance Change (ΔS/S)

0.9 0.8

In this work, a highly sensitive and selective CNT biosensor was developed using a photovoltaic-polymer-coated CNT, which successfully detected 12 pg/mL of cTnT. The developed optical immunosensor was developed for clinical diagnosis at a low cost. In spite of this inexpensive method of fabrication, superior functionality was achieved utilizing CNTs. In addition, the fabricated device was electrically insulated by parylene, which significantly enhanced the selectivity of the CNT-based biosensor and provided an excellent platform for protein immunoassay. Furthermore, since the CNT photodetector was directly contacted to the site of chemiluminescent assay, the loss of optical signal was extremely low by minimizing the scattering light through the substrate. As a result, the developed CNT biosensor can be applied to low-cost POCT devices, achieving highly specific and sensitive optical detection.

120 pg/ml

0.7

pH 4

0.6

pH 10

0.5 0.4 0.3 0.2 0.1 0 0

5

10

15

20

Time (sec) Fig. 6. Effect of ionic solution by flowing the solutions of pH 4 and pH 10. There is no signal change when the ionic solutions were applied to the device. When the enzyme substrate was provided after flowing 120 pg/ml of cTnT in ELISA procedure, there was a sharp increase of conductivity change due to the chemiluminescent light.

Acknowledgments The authors gratefully acknowledge the partial support of this work by National Science Foundation (NSF) under the program of Electronics, Photonics and Device Technologies (EPDT, #0622036). Appendix A. Supplementary data

device. For the verification of parylene insulation effect, the solutions with various pH levels were flowed through the microchannel on the CNT photodetector, causing both the ionic current and the gating effect on the CNT FET device. Fig. 6 shows the response from the device while pH 4 and pH 10 solutions flowed through the device. Because the CNT photodetector was insulated with the parylene, the solutions did not make any signal change from the CNT photodetector. This result clearly showed the insulation effect by parylene, which isolated the CNT biosensor from the surrounding media and only transferred the light signal to the CNT photodetector. The high sensitivity of our optical immunosensor also obtained from the structural superiority of the CNT coated with PmPV. Since all the atoms of the CNT made a direct contact with the surrounding photovoltaic polymer, a highly sensitive response to the optical signal was accomplished by the generation of electrons in PmPV. Furthermore, the matched optical spectra between PmPV and chemiluminescent light reduced the interruption of external light and maximized the signal change for the chemiluminescent light by HRP. Thus, even if there was a small leakage of outside light during the measurement, the PmPV-coated CNT could be more robust than common photodetectors in detecting the targeted chemiluminescent light owing to the wavelength match between the chemiluminescent light and the absorption of PmPV. The developed CNT immunosensor has been implemented with inexpensive fabrication procedures to be clinically utilized as a disposable type of device. In addition, parylene insulation excluded the interruption of ions and non-specifically bound proteins, attaining highly selective detection of the biomolecules. Furthermore, although our CNT biosensor was constructed with an FET-type structure for easy measurement in a miniaturized system, high selectivity has been achieved by an optical measurement scheme. Thus, the developed optical immunosensor is highly desirable for point-of-care clinical diagnosis as a CNT-based nanobiosensor.

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bios.2012.02.004. References Borghetti, J., Derycke, V., Lenfant, S., Chenevier, P., Filoramo, A., Goffman, M., Vuillaume, D., Bourgoin, J.-P., 2006. Adv. Mater. 18, 2535–2540. Chang, T.Y., Yadav, V.G., Leo, S.D., Mohedas, A., Rajalingam, B., Chen, C., Selvarasah, S., Dokmeci, M.R., Khademhosseini, A., 2007. Langmuir 23, 11718–11725. Chen, R.J., Bangsaruntip, S., Drouvalakis, K.A., Kam, N.W.S., Shim, M., Li, Y., Kim, W., Utz, P.J., Dai, H., 2003. Proc. Natl. Acad. Sci. U.S.A. 100, 4984–4989. Collins, P.G., Bradley, K., Ishigami, M., Zettl, A., 2000. Science 287, 1801–1804. Cui, Y., Wei, Q., Park, H., Lieber, C.M., 2001. Science 293, 1289–1292. Hata, K., Futaba, D.N., Mizuno, K., Namai, T., Yumura, M., Iijima, S., 2004. Science 306, 1362–1364. Heller, I., Mannik, J., Lemay, S.G., Dekker, C., 2009. Nano Lett. 9, 377–382. Jeong, Y.S., Ratier, B., Moliton, A., Guyard, L., 2002. Synth. Met. 127, 189–193. Kim, J.P., Lee, B.Y., Lee, J., Hong, S., Sim, S.J., 2009. Biosens. Bioelectron. 24, 3372–3378. Loeb, G.E., Bak, M.J., Salcman, M., Schmidt, E.M., 1977. IEEE Trans. Biomed. Eng. 24, 121–128. Meng, E., Li, P., Tai, Y., 2008. J. Micromech. Microeng. 18, 045004. Panchapakesan, B., Cesarone, G., Liu, S., Teker, K., Wickstrom, E., 2005. Nanobiotechnology 1, 353–360. Schmidt, E.M., 1983. J. Electrophys. Technol. 10, 19–29. Shi, Y., Dong, X., Tantang, H., Weng, C.-H., Chen, F., Lee, C., Zhang, K., Chen, Y., Wang, J., Li, L.-J., 2008. J. Phys. Chem. C 112, 18201–18206. Shim, J.S., Yun, Y.H., Cho, W., Shanov, v., Schulz, M.J., Ahn, C.H., 2010. Langmuir 26, 11642–11647. Shim, J.S., Yun, Y.H., Rust, M.J., Do, J., Shanov, V., Schulz, M.J., Ahn, C.H., 2009. Nanotechnology 20, 325607. Shim, M., Kam, N.W.S., Chen, R.J., Li, Y., Dai, H., 2002. Nano Lett. 2, 285–288. Silva, B.V., Cavalcanti, I.T., Mattos, A.B., Moura, P., Sotomayor, M.P., Dutra, R.F., 2010. Biosens. Bioelectron. 26, 1062–1067. Star, A., Gabriel, J.P., Bradley, K., Gruner, G., 2003. Nano Lett. 3, 459–463. Star, A., Lu, Y., Bradley, K., Gruner, G., 2004. Nano Lett. 4, 1587–1591. Star, A., Tu, E., Niemann, J., Gabriel, J.P., Joiner, C.S., Valcke, C., 2006. Proc. Natl. Acad. Sci. U.S.A. 103, 921–926. Wanekaya, A.K., Chen, W., Myung, N.V., Mulchandani, A., 2006. Electroanalysis 18, 533–550. Zheng, G., Patolsky, F., Cui, Y., Wang, W.U., Lieber, C.M., 2005. Nat. Biotechnol. 23, 1294–1301.