Biosensors and Bioelectronics 63 (2015) 444–449

Contents lists available at ScienceDirect

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

Ultra-sensitive plasmonic nanometal scattering immunosensor based on optical control in the evanescent field layer Seungah Lee a, Guenyoung Park b, Suresh Kumar Chakkarapani b, Seong Ho Kang a,b,n a b

Department of Applied Chemistry, College of Applied Science, Kyung Hee University, Yongin-si, Gyeonggi-do 446-701, Republic of Korea Department of Chemistry, Graduate School, Kyung Hee University, Yongin-si, Gyeonggi-do 446-701, Republic of Korea

art ic l e i nf o

a b s t r a c t

Article history: Received 30 June 2014 Received in revised form 21 July 2014 Accepted 25 July 2014 Available online 2 August 2014

Novel, fluorescence-free detection of biomolecules on nanobiochips was investigated based on plasmonic nanometal scattering in the evanescent field layer (EFL) using total internal reflection scattering (TIRS) microscopy. The plasmonic scattering of nanometals bonded to biomolecules was observed at different wavelengths by an electromagnetic field in the EFL. The changes in the scattering of nanometals on the gold-nanopatterned chip in response to the immunoreaction between silver nanoparticles and antibodies allowed fluorescence-free detection of biomolecules on the nanobiochips. Under optimized conditions, the TIRS immunoassay chip detected different amounts of immobilized antigen, i.e., human cardiac troponin I. The sandwich immuno-reaction was quantitatively analyzed in the dynamic range of 720 zM–167 fM. The limit of detection (S/N¼ 4) was 600 zM, which was  140 times lower than limits obtained by previous total internal reflection fluorescence and dark field methods. These results demonstrate the possibility for a fluorescence-free biochip nanoimmunoassay based on the scattering of nanometals in the EFL. & 2014 Elsevier B.V. All rights reserved.

Keywords: Nanobiochip Total internal reflection scattering microscopy Fluorescence-free Plasmonic nanometal scattering Nanoparticle detection

1. Introduction Detection of single fluorescent-tagged molecules has been demonstrated as a powerful tool for a wide range of applications ranging from biophysics to quantum optics (Michalet et al., 2005). Although fluorescence detection allows for signal amplification, thereby providing sufficient sensitivity for single-molecule detection, such methods have limitations regarding photostability, photobleaching, and time-dependent fluctuations in fluorescence signals (Moerner and Orrit, 1999). Therefore, several works have focused on the development of fluorescence-free detection technologies to overcome these limitations (Espina et al., 2004). Over the past few years, researchers have developed extinction detection and spectroscopy as an alternative to fluorescence in order to investigate single molecules and nanoscale objects such as metallic nanoparticles and viruses (Kukura et al., 2009). Recently, in array biochip studies, fluorescence-free detection methods have provided information about biological analytes; these include surface plasmon resonance (Esseghaier et al., 2014), surfaceenhanced Raman scattering (Hu et al., 2011), interferometric reflectance (He et al., 2014), colorimetric (Liang et al., 2005), n Corresponding author at: Department of Applied Chemistry, College of Applied Science, Kyung Hee University, Yongin-si, Gyeonggi-do 446-701, Republic of Korea. Tel.: þ 82 31 201 3349; fax: þ 82 31 201 2340. E-mail address: [email protected] (S.H. Kang).

http://dx.doi.org/10.1016/j.bios.2014.07.071 0956-5663/& 2014 Elsevier B.V. All rights reserved.

chemiluminesence (Chen et al., 2013), and electrochemical (Chai and Takhistov, 2012) methods. In addition, we have previously reported fluorescent-free detection of biomolecules on nanobiochips based on enhanced dark field microscopy, one of the scattering-based detection techniques (Lee et al., 2013), and based on differential interference contrast (DIC) (Lee and Kang, 2014). However, these approaches limit the specimens that can be studied and are susceptible to scattering by dust and other particles (Anker et al., 2008). In the case of DIC, typical signal-tonoise ratios (SNRs) are low, and, even with adequate SNRs, it is difficult to discriminate between the probes and other particles (Xiao et al., 2011). Total internal reflection (TIR) occurs when a propagating wave strikes a medium boundary at an angle larger than a particular critical angle with respect to the normal to the surface. If the refractive index is lower on the other side of the boundary and the incident angle is greater than the critical angle, the wave cannot pass through the surface and is reflected entirely. When a particle is located near the evanescent wave formed in this manner, it will scatter the wave with a measurable intensity that reflects its distance from the interface. The intensity of the scattered light depends on the distance from the interface to the particle, following the relationship I(h)¼I0exp(  βh), where h is the particle–surface separation, I0 is the scattered light intensity at h¼ 0, and β is the decay index (Prieve et al., 1987). Recently, TIR microscopy based on plasmon scattering has been used to collect

S. Lee et al. / Biosensors and Bioelectronics 63 (2015) 444–449

information to allow for the tracking and localization of particles (Ha et al., 2012). Metallic nanoparticles and nanostructures are useful in many applications including light guiding and manipulation on the nanoscale, biodetection on the single-molecule level, enhanced optical transmission through sub-wavelength apertures, and highresolution optical imaging below the diffraction limit (Deng and Goldys, 2012). The merits of metallic nanoparticles are chiefly related to their optical properties (Sonnichsen and Alivisatos, 2005), large scattering and absorption cross-sections (Sperling et al., 2008), high photostability (Sperling et al., 2008; Wu and Yeow, 2008), and excellent biocompatibility (Murphy et al., 2008). In particular, the plasmonic properties of noble metals have become one of the most researched topics in optical sensors (Barnes et al., 2003). While these aforementioned properties of nanoparticles are known, there has been relatively little exploration in regard to their potential application in fluorescence-free immunoassays with high SNRs based on plasmon scattering of noble nanometals on biochips in the evanescent field layer (EFL) using TIR. This detection system can overcome problems related to low SNRs through the use of DIC and of scattering through the use of a dark field. Herein, we describe a novel setup in which two noble metals, configured as a gold-patterned biochip containing silver nanoparticles, were incorporated to form a total internal reflection scattering (TIRS) device. This device allows for highly sensitive detection without the commonly used TIR fluorescence in the EFL (Lee and Kang, 2013a). The importance of this approach is that it allows fluorescence-free immunoassay detection with a highresolution CCD camera. It is also not susceptible to photobleaching. This method includes various means of optical control (i.e., light source, bandpass filter, incident angle, numerical aperture of the objective, and rotatable analyzer) and uses the scattering of laser light by noble metals in the EFL.

2. Materials and methods 2.1. Reagents 11-Mercaptoundecanoic acid (MUA, 95%), 6-mercapto-1-hexanol (MCH, 97%), 1-ethyl-3-(3-dimethylamino-propyl)carbodiimide hydrochloride (EDC), dimethyl sulfoxide (DMSO, 99.5%), 2-(morpholino)ethanesulfonic acid (MES), glycine, and phosphate-buffered saline (PBS) were supplied by Sigma-Aldrich (St. Louis, MO,

USA). Dithiobis(succinimidyl propionate) (DSP) and Protein A/G were purchased from Pierce (Rockford, IL, USA). Tris(base) was purchased from Mallinckrodt Baker, Inc. (Phillipsburg, NJ, USA). StabilGuard was purchased from Surmodics (Eden Prairie, MN, USA). Silver nanoparticles (SNPs, 20 nm in diameter, 7.0  1010 particles/mL) were obtained from BBI Life Sciences (Cardiff, UK). Monoclonal mouse anti-cardiac troponin I antibody (clone 19C7 and 16A11) and standard human cardiac troponin I (cTnI, clone 8T53) were purchased from HyTest (Turku, Finland). Thyroid stimulating hormone protein (TSH-antigen, 30-AT09) purchased from Fitzgerald (North Acton, MA, USA). 2.2. Gold-nanopatterned chip The gold-nanopatterned chip (GNC) was fabricated as previously reported (Lee and Kang, 2013a). The cleaned glass wafer was coated with a 150-nm-thick polymethylmethacrylate (PMMA) layer that served as an electron-sensitive photoresist. An electron beam (Elionix E-beam system, 100 keV/100 pA) was then used to burn off the polymer in a desired pattern. Au/Cr was deposited by thermal evaporation, and PMMA was removed by dichloromethane to form 4  5 array patterns of gold spots (500 nm in diameter and with a pitch of 10 μm) on a 10 mm2 glass wafer. 2.3. Total internal reflection scattering (TIRS) detection system The schematic representation and physical layout of the apparatus were modified from previously published configurations (Fig. 1 and Supplementary Fig. S1) (Lee and Kang, 2013b). The use of label-free TIRS imaging techniques based on an upright Olympus BX51 microscope (Olympus Optical Co., Ltd., Tokyo, Japan) was applied for the first time to a sandwich immunoassay. For TIR illumination, two lasers with vertical polarization (p-pol) were used: a 20-mW, 405-nm laser (SOL-405-LM-020 T, Shanghai Laser & Optics Century Co., Ltd., China) for illumination of the 20nm SNPs and a 30-mW, solid-state, 671-nm laser (SDL-671-040T, Shanghai Laser & Optics Century Co., Ltd.) for illumination of the gold spots. Various optical components were used in the TIRS detection system to observe the scattering shape and intensity of the raw gold spots and the SNP conjugated with antibody and reacted on the gold-nanopatterned chip. The scattered signal was collected with an NA 0.6–1.3 objective lens (UPLANFLN,  100). Wavelength selection was accomplished with a set of filters purchased from Semrock (Rochester, NY, USA) with central wavelengths of 406/15 nm, 473/10 nm, 520/15 nm, 575/15 nm, 605/

EM CCD

Iss

penetration depth (d)

z

BF 0.6-1.3 NA

SNP

Iz Isg

Y

RA

L1

445

Y

EFL L2

Unit: nm

~20 14.5

14.5 ~5 1.2

GNC

Prism Before

After

Fig. 1. (A) Schematic of a TIRS detection system and (B) scattering of light from GNC (left) and SNP (right) in the evanescent field layer. The following abbreviations are used: L, laser; GNC, gold-nanopatterned chip; NA, numerical aperture; RA, rotatable analyzer; BF, bandpass filter; EM CCD, electron-multiplying charge-coupled device; SNP, silver nanoparticle; EFL, evanescent field layer; Isg, scattering intensity of gold spots; and Iss, scattering intensity of silver nanoparticles.

446

S. Lee et al. / Biosensors and Bioelectronics 63 (2015) 444–449

10 nm, and 670/30 nm. Scattering was measured at various wavelengths with fixed prisms and an analyzer that could be rotated 360° (U-AN360P-2, Olympus). Moreover, maximization of light scattering from the metals was carried out by adjusting the incident angle of light illumination using the mirror manipulator in the TIRS detection system. TIRS images were acquired with an electron-multiplying cooled charge-coupled device (EM-CCD) camera (512  512 pixel imaging array, QuantEM 512SC, Photometrics, AZ, USA) after the sandwich immunoprocess of cTnI on the gold substrate was prepared following a previously reported procedure (Lee and Kang, 2013b). Using a shutter control, the exposure time was set to 100 ms. Images were acquired and the scattered intensity was calculated as the difference between the intensities of the selected signal regions and the intensities of the background regions with the same CCD image area using MetaMorph 7.1 software.

3.2. Sandwich immunoreaction with SNPs on a gold-nanopatterned chip A change in the surface roughness of a GNC with a 30.5-nm average deposition thickness (Lee et al., 2013) was confirmed with atomic force microscopy after the sandwich immunoreaction with 20-nm SNP antibodies. The height of the 500-nm-diameter gold spot increased by 51 nm after immunoreaction with 85 aM of the cTnI protein (Supplementary Fig. S3). Although the gold spots were not exactly the same in terms of their aspect ratio (height difference), due to either the size heterogeneity of the 20-nm SNPs (their size distribution was 26.6 74.9 nm; see Supplementary Fig. S4) or to their strong dependence on spatial orientation or form (such as dimer forms) of the SNPs (Stender et al., 2010), the target cTnI was successfully detected using the SNP detection probe based on the sandwich immunoreaction on GNC. 3.3. Optical control of plasmonic nanometals scattered in the evanescent field layer

3. Results and discussion 3.1. Characterization of gold and silver metals Yguerabide et al. have reported that, in normalized terms, silver and gold show the most efficient plasmon scattering: Ag 4Au 4Cu EAl (Yguerabide and Yguerabide, 1998). A 60-nm gold nanoparticle, for example, scatters with an intensity equivalent to the fluorescence of 3  105 fluorescein molecules (Yguerabide and Yguerabide, 1998; Aslan et al., 2005). Metal nanoparticles of two materials (i.e., gold and silver) were used for highly sensitive, selective non-fluorescence detection based on their scattering. The first material, gold, was patterned in 500-nm spots on a goldnanopatterned chip (GNC) and had an absorption maximum at the wavelength of 675 nm (Lee et al., 2013). Detection of molecular binding to the surface of the nanoparticles was confirmed by a shift in this absorbance peak. Silver nanoparticles (SNPs, 20 nm diameter) were used as the secondary noble metal and showed absorption maxima at 399 and 408 nm before and after conjugation with the cTnI detection antibody, respectively (Supplementary Fig. S2). In addition, differences in the absorbance wavelength of these nanometals were detected selectively under specific illuminations (i.e., 405 nm and 671 nm lasers).

θ = 74.2°

A

B

θ = 71.3°

Scattering intensity

500 nm 112 1×10 108 104

188

θ = 74.2°

148 S/N = 24

θ = 71.3°

109

500 nm 900

1×10

1×10 700

107 105 S/N = 2.8

500 S/N = 22 300

103

108

D

500 nm 111

1×10

128 100

C

500 nm

168 S/N = 11

Scattered metals are difficult to incorporate into objective-type TIR setups because the totally reflected beam remains in the detection path and dominates the scattered signal due to the fact that both light beams share the same wavelength (Schneider et al., 2013; Ueno et al., 2010). To solve this problem, we selected a prism-type TIRS setup because it naturally prevents the totally reflected illumination beam from entering the detection path (Fig. 1 and Supplementary Fig. S1). Two kinds of lasers and various bandpass filters were used to selectively detect the scattered signals of the two nanometals. To maximize the resolution of the total internal reflection scattering imaging technique, the intensity of the scattering from the metals was optimized by controlling the penetration depth of the EFL. The principle of the height measurement method was based on the fact that the intensity of the evanescent field (Iz) decreases exponentially with perpendicular distance (z) from the TIR surface (Fig. 1B) (Helden et al., 2006). This means that the intensity of the scattered light from a particle in this region strongly depends on its distance from the TIR surface. Light sources with wavelengths of 405 nm and 671 nm are illuminated on the interface of two mediums, combining to create the incident angle chosen such that it is larger than the critical angle (62.9°). We constructed a trapezoidal prism where the incident angle of

100

101

0 10 20 30 40 50 60

0 10 20 30 40 50 60

0 10 20 30 40 50 60

0 10 20 30 40 50 60

Pixel distance

Pixel distance

Pixel distance

Pixel distance

Fig. 2. Typical image showing the effect on scattered intensity with changes to the incident angle θ before conjugation with the cTnI detection antibody: (A, B) 405 nm illumination and (C, D) 671 nm illumination.

S. Lee et al. / Biosensors and Bioelectronics 63 (2015) 444–449

3.4. Fluorescence-free detection of cTnI molecules by scattering intensity

light illumination could be adjusted between 71.3° and 74.2° by means of the total reflection mirror manipulator. When the incident angle was changed from 74.2° to 71.3°, the relative scattering intensity was increased by 8.3 times for the 405 nm illumination and by 290 times for the 671 nm illumination (Fig. 2); the respective SNRs for the 74.2° and 71.3° incident angle increased 2.1 and 7.9 times, respectively. The angle of 71.3° was selected as the optimum incident angle of illumination. The numerical aperture (NA) of the objective lens affected both the scattering intensity and the shape of the resulting image. The NA of the objective is defined as nsinθ, where n is the refractive index of the medium, and θ is the half-angle of the maximum cone of light that can enter the objective. The higher-NA objective lens collected more light and generally produced brighter images with more diffraction patterns. When a 1.3-NA objective lens was used in TIRS microscopy (including a 406-nm bandpass filter), the resulting scattering signal of the SNP-antibody reacted on a 500nm GNC and illuminated by a 405-nm laser was about 1.8 times higher than that collected by the 0.6-NA objective lens (Fig. 3A and B). In TIRS microscopy with a 670-nm bandpass filter, the scattering signal of the raw 500-nm gold spots illuminated by the 671nm laser as collected by the 1.3-NA objective lens was 18.5 times higher than that collected by the 0.6-NA objective lens (Fig. 3C and D). This observed difference in SNR agreed with a previous report that an objective with a larger NA visualizes finer details than an objective with a smaller NA (Ha et al., 2012). The scattering image collected using the higher-NA objective was dumbbell-shaped due to a shadow in the propagation direction of the surface plasmonic wave. The cross-sectional profile across the gold spot was also measured (Fig. 3). Sample contrast arose from the rotation of polarized light through the sample. We confirmed the scattering shapes for the gold spots by rotating the analyzer at intervals of 30° between 0° and 360° for both 405-nm and 671-nm illuminations; high scattering intensity and a dumbbell shape for the GNC were observed at angles near 90° or near 270° (Fig. 4). Analyzer orientations of 0°, 180°, and 360° corresponded to low intensities of the scattering signal.

B

NA = 1.3

0

2

4

6

8

10 12

Pixel distance

4 16π 4a6nmed I0 m2 − 1 cos2θ , 2 4 m2 + 2 r λ

where I0 is the incident intensity of monochromatic light, nmed is the refractive index surrounding the particle, m is the refractive index of the bulk particle material, and r is the distance between the particle and the location of scattered light detection. Our experimental conditions were adjusted to obtain the desired Isc, which depends on the variables described in the Rayleigh equation above. The intensities of light scattered by the metals before and after the protein immunoreaction on the GNC were measured. The RSI was calculated as the difference between the signal intensity and the intensity from the background within a selected area. The scattering image of the TIRS nanoimmunoassay chip could be observed using either the 671-nm or 405-nm light source, a 1.3NA objective lens, a 90° analyzer angle, and either the 670-nm or 406-nm bandpass filter (corresponding to the light source used) (Fig. 5). Although the reflection depends on the distance of the particles from the surface, the scattering intensities of the GNC before and after the reaction were different. This difference is caused by the binding of the antigen or the 20-nm SNP-detection antibody complex, which had a 408-nm absorption wavelength on GNC and created a sheltering effect on the GNC surface. The scattering intensity of the raw GNC (black circle) was 3.3 times that of the GNC after the sandwich reaction (white circle) under 671-nm laser illumination and using the 670-nm bandpass filter (Fig. 5A). Under 405 nm laser illumination and with a 406 nm bandpass filter, the scattering intensity of the GNC after sandwich immunoreaction was increased about 6.5 times that of the raw

0

2

4

6

8

10 12

Pixel distance

NA = 0.6

500 nm

500 nm

Intensity (a.u.)

50

D

NA = 1.3

500 nm

Intensity (a.u.)

Intensity (a.u.)

50

2

Isc =

C

NA = 0.6

500 nm

The scattering of light by sub-wavelength-sized particles is well described by the Rayleigh theory (Yguerabide and Yguerabide, 1998). For incident light polarized and observed in the same plane, the intensity of light scattered, Isc, in the direction θ by a homogeneous spherical particle with radius a, given that this particle is much smaller than the wavelength of the incident beam, λ, is described by the Rayleigh expression (Yguerabide and Yguerabide, 1998)

Intensity (a.u.)

A

447

50

0

2

4

6

8

10 12

Pixel distance

0

50

2

4

6

8

10 12

Pixel distance

Fig. 3. TIRS images showing the effect of the NA of the objective lens on the resulting images of the 500-nm GNC before conjugation with the cTnI detection antibody for laser illumination at (A, B) 405 nm and (C, D) 671 nm. Line-sectional profiles are included across the center of the scattering patterns of the GNC with the different NAs of 0.6 and 1.3. The following abbreviations is used: NA, numerical aperture.

448

S. Lee et al. / Biosensors and Bioelectronics 63 (2015) 444–449

0°

30°

60°

90°

120°

150°

180°

210°

240°

270°

300°

330°

360°

0°

30°

60°

90°

120°

150°

180°

210°

240°

270°

300°

330°

360°

405 nm illumination

671 nm illumination

20

RSI (a.u.)

2

30

90

150

210

270

330

Angle of rotatable analyzer (°)

30

90

150

210

270

330

Angle of rotatable analyzer (°)

Fig. 4. Changes in (A, B) shape and (C) relative scattering intensity (RSI) for 500-nm GNC before conjugation with the cTnI detection antibody with respect to changes in analyzer angle for (A) 405- and (B) 671-nm illuminations.

GNC

GNC+SNP

GNC

decreased 3.3 times

GNC+SNP

increased 6.5 times 500 nm GNC 500 nm GNC + 20 nm SNP

2.5×103

5×10

RSI (a.u.)

500 nm GNC 500 nm GNC + 20 nm SNP

406

473

520

575

605

670

Bandpass filter (nm)

406

473

520

575

605

670

Bandpass filter (nm)

Fig. 5. Scattering intensities before and after the sandwich reaction, including the use of a bandpass filter of 670 or 406 nm to match the (A) 671 nm illumination and (B) 405 nm illumination source. RSI: relative scattering intensity. Error bar is equivalent to 500 nm.

GNC (black circle, Fig. 5B). These results indicate that a sandwich immunoreaction occurs between the SNPs and the detection antibody on raw GNC. 3.5. Quantitative analysis of cTnI molecules on chip by total internal scattering microscopy Recently, we quantitated cTnI protein on a gold nanoarray chip based on total internal reflection fluorescence (TIRF) and enhanced dark field (EDF) methods (Lee and Kang, 2013b; Lee et al., 2013). These results showed a concentration detection limit of approximately 85 aM cTnI. The standard curve for serially diluted

standard cTnI protein using TIRS is shown in the linear range between 720 zM and 167 fM (correlation coefficient, R¼0.9876) under 405-nm laser illumination and with a 406-nm bandpass filter, as in Fig. 6. The TIRS method achieved a limit of detection (LOD) of 600 zM; this value is 142 times lower than the LOD obtained by previous methods. In addition, to demonstrate the selectivity about cTnI protein on our immunochip, a mixture of cTnI protein and another protein (i.e., TSH) were reacted on GNC and then detected under 405 nm laser illumination and with a 406 nm bandpass filter. The experimental concentration for cTnI based on quantitative calibration curves in Fig. 6 showed in Fig. S5 (white bar) after reaction with

S. Lee et al. / Biosensors and Bioelectronics 63 (2015) 444–449

Acknowledgments

10

This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) and was funded by the Ministry of Education, Science and Technology (2012R1A2A2A01013466).

RSI (×102)

8 6 S/N = 4

y = 143.98x + 137 R = 0.9876

4 LOD = 600 zM

Appendix A. Supplementary information

2 0 -1

449

0

1

2

3

4

5

6

Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2014.07.071.

Log[cTnI (aM)] Fig. 6. Quantitative calibration curves for concentrations of target protein molecules (standard cTnI) in the linear range of 720 zM to 167 fM with 600 zM LOD at a wavelength of 405 nm. RSI: relative scattering intensity.

mixture sample (cTnI and TSH). In the experiments, the mixture samples of cTnI and TSH were prepared in the ratios of 3:1, 1:1, and 1:5, in which cTnI concentration were 300 aM, 1 fM, and 10 fM respectively. The experiment results showed that the scattering intensity was only correlated with the concentration of cTnI, not depending on the concentration of TSH or the ratios of the mixture.

4. Conclusions Herein, we developed a fluorescence-free immunoassay based on noble metal scattering in a trapezoid prism-type TIR setup with high SNR. The use of a trapezoidal prism allowed the penetration depth of the evanescent field to be adjusted by controlling the incident angle; the use of a manipulator-equipped mirror allowed this adjustment to be carried out while maintaining the position of the illuminated area. In addition, observation of the scattering shape and intensity of the raw gold spot and the SNP-antibody complex that reacted on the gold spot was made possible by controlling various optical components in the TIRS microscopy setup. Because the TIRS system is based on the scattering intensity observed in the prism-type system before and after the sandwich immunoreaction on GNC, quantitative analysis is possible. The scattering intensity of the gold spot at 671 nm before the immunoreaction was about 3.3 times that after the immunoreaction, whereas the scattering intensity of GNC at 405 nm after the reaction was 6.5 times higher than the intensity before the reaction. The limit of detection (LOD) was observed at 600 zM over a wide linear range (from 720 zM to 167 fM, R¼0.9876). These results demonstrate the possibility of detection with high SNRs using a non-fluorescence nanoimmunoassay chip and a TIRS technique with controlled optics.

References Anker, J.N., Hall, W.P., Lyandres, O., Shah, N.C., Zhao, J., Van Duyne, R.P., 2008. Nat. Mater. 7, 442–453. Aslan, K., Lakowicz, J.R., Geddes, C.D., 2005. Curr. Opin. Chem. Biol. 9, 538–544. Barnes, W.L., Dereux, A., Ebbesen, T.W., 2003. Nature 424, 824–830. Chen, F., Mao, S., Zeng, H., Xue, S., Yang, J., Nakajima, H., Lin, J.-M., Uchiyama, K., 2013. Anal. Chem. 85, 7413–7418. Chai, C., Takhistov, P., 2012. Appl. Surf. Sci. 263, 104–110. Deng, W., Goldys, E.M., 2012. Langmuir 28, 10152–10163. Espina, V., Woodhouse, E.C., Wulfkuhle, J., Asmussen, H.D., Petricoin 3rd, E.F., Liotta, L.A., 2004. J. Immunol. Methods 290, 121–133. Esseghaier, C., Suaifan, G.A.R.Y., Ng, A., Zourob, M., 2014. J. Biomed. Nanotechnol. 10, 1123–1129. Ha, J.W., Marchuk, K., Fang, N., 2012. Nano Lett. 12, 4282–4288. He, Y., Li, X., Que, L., 2014. J. Biomed. Nanotechnol. 10, 767–774. Helden, L., Eremina, E., Riefler, N., Hertlein, C., Bechinger, C., Eremin, Y., Wriedt, T., 2006. Appl. Opt. 45, 7299–7308. Hu, H., Wang, Z., Wang, S., Zhang, F., Zhao, S., Zhu, S., 2011. J. Alloys Compd. 509, 2016–2020. Kukura, P., Ewers, H., Muller, C., Renn, A., Helenius, A., Sandoghdar, V., 2009. Nat. Methods 6, 923–927. Lee, S., Kang, S.H., 2013a. Analyst 138, 3478–3482. Lee, S., Kang, S.H., 2013b. Talanta 104, 32–38. Lee, S., Kang, S.H., 2014. Biosens. Bioelectron. 60, 45–51. Lee, S., Yu, H., Kang, S.H., 2013. Chem. Commun. 49, 8335–8337. Liang, R.Q., Li, W., Li, Y., Tan, C.Y., Li, J.X., Jin, Y.X., Ruan, K.C., 2005. Nucleic Acids Res. 33, e17. Michalet, X., Pinaud, F.F., Bentolila, L.A., Tsay, J.M., Doose, S., Li, J.J., Sundaresan, G., Wu, A.M., Gambhir, S.S., Weiss, S., 2005. Science 307, 538–544. Moerner, W.E., Orrit, M., 1999. Science 283, 1670–1676. Murphy, C.J., Gole, A.M., Stone, J.W., Sisco, P.N., Alkilany, A.M., Goldsmith, E.C., Baxter, S.C., 2008. Acc. Chem. Res. 41, 1721–1730. Prieve, D.C., Luo, F., Lanni, F., 1987. Faraday Discuss. 83, 297–307. Schneider, R., Glaser, T., Berndt, M., Diez, S., 2013. Opt. Express 21, 3523–3539. Sonnichsen, C., Alivisatos, A.P., 2005. Nano Lett. 5, 301–304. Sperling, R.A., Gil, P.R., Zhang, F., Zanella, M., Parak, W., 2008. Chem. Soc. Rev. 37, 1896–1908. Stender, A.S., Wang, G., Sun, W., Fang, N., 2010. Anal. Chem. 4, 7667–7675. Ueno, H., Nishikawa, S., Iino, R., Tabata, K.V., Sakakihara, S., Yanagida, T., Noji, H., 2010. Biophys. J. 98, 2014–2023. Wu, X., Yeow, E.K., 2008. Nanotechnology 19, 035706. Xiao, L., Qiao, Y., He, Y., Yeun, E.S., 2011. J. Am. Chem. Soc. 133, 10638–10645. Yguerabide, J., Yguerabide, E.E., 1998. Anal. Biochem. 262, 137–156.