1924

OPTICS LETTERS / Vol. 40, No. 9 / May 1, 2015

Differential phase-detecting localized surface plasmon resonance sensor with self-assembly gold nano-islands Guangyu Qiu, Siu Pang Ng, and Chi Man Lawrence Wu* Department of Physics and Materials Science, City University of Hong Kong, Hong Kong SAR, Hong Kong *Corresponding author: [email protected] Received March 9, 2015; revised March 27, 2015; accepted March 31, 2015; posted April 1, 2015 (Doc. ID 235744); published April 20, 2015 Self-assembly (SAM) gold nano-islands are fabricated by two-step thin-film deposition-annealing method. Despite random distribution of the SAM, the p-polarized light after total internal reflection shows significant phase transition at the extinction wavelengths upon refractive index variation due to localized surface plasmon resonance (LSPR) effect. It resembles the sharp phase transition observed in conventional surface plasmon resonance (SPR) biosensors, so that the bulk sensitivity of the SAM-LSPR sensor is improved via the phase interrogation method. In this Letter, we present both computational and experimental investigations to the SAM-LSPR sensor and the results show excellent agreement with each other. With bulk refractive index resolution to 9.75 × 10−8 RIU, we believe the phase-detecting SAM-LSPR sensor would be an essential step toward low-cost labelfree sensing applications. © 2015 Optical Society of America OCIS codes: (240.6680) Surface plasmons; (120.3180) Interferometry; (280.1415) Biological sensing and sensors; (220.4241) Nanostructure fabrication. http://dx.doi.org/10.1364/OL.40.001924

Localized surface plasmon resonance (LSPR) is a strong electromagnetic near-field effect associated with noble metal nanostructures. Owing to the locally enhanced electromagnetic (EM) field confined to the vicinity of the nanostructures at resonance [1], regular array of gold nanostructures had demonstrated high sensitivity to local refractive index (RI) variation. Thus, LSPR is an ideal candidate for real-time label-free sensing applications of small biomolecules [2,3]. It is known that the regular array of gold nanostructures, i.e., dots [1], holes [4], and pillars [5], exhibit sharp spectral extinction and abrupt phase transition in resonance. Therefore, the amplitude component and phase difference of the incident light interacting with these LSPR arrays had been explored extensively via spectral ellipsometry (SE) [6]. Regular nanoarrays can be fabricated by lithography methods, i.e., E-beam and multi-step nanomasking techniques [7]. However, the lithography approaches impose considerable challenge on the effectiveness of mass production of the LSPR sensor in terms of cost and yield. On the other hand, random array of self-assembly (SAM) gold nanoislands were fabricated by two-step deposition-annealing methods [8], and biosensor based on these SAM was successfully reported [9]. An extensive list of applications employing nano-islands was recently reported by Yang et al. [10] To our awareness, most of these works done on SAM were performed with SE, and the LSPR phase was calculated from the two orthogonal polarizations separately [11]. On the other hand, we have introduced spectral interferometry (SI), which directly measures the differential phase between the orthogonal polarizations of SPR biosensor [12]. While SE employs divisional methodology that captures the p-and s-polarizations as numerator and denominator alternatively, SI is a convolution process that employs windowed Fourier transform (WFT) for real-time phase retrieval. In view of the evergrowing processing power of personal computers, SI prevails in terms of simplicity, speed, and robustness in operation. 0146-9592/15/091924-04$15.00/0

The SAM-LSPR sample was fabricated by sputtering gold of 5.0-nm nominal thicknesses onto a clean BK7 glass slide, followed by thermal annealing at 550°C for 3 h in air. We performed topographical scan of the SAM using tapping mode atomic force microscopy (TM-AFM), as shown in Fig. 1(a). The SAM nano-islands were found almost spherical with average diameter of 40 nm. Then the nano-islands were deliberately removed by chemical etching to confirm the embedment into the glass substrate by repeating the TM-AFM scan. The SAMs were found partially embedded into the substrate for about 8 nm (data not shown). To establish the theoretical

Fig. 1. (a) Tapping-mode atomic force microscopy (TM-AFM) scan of the SAM gold nano-islands. (b) Plane view of the numerical model with TFSF source for spectral absorption computation by FDTD. Side views of the numerical model with (c) p-polarized and (d) s-polarized incident light. Grid size of (b)–(d) is 50 nm. © 2015 Optical Society of America

May 1, 2015 / Vol. 40, No. 9 / OPTICS LETTERS

support of the SAM-LSPR sensor, it is necessary to construct the numerical model based on the TM-AFM data and determine its response to the orthogonal p-and s-polarizations. Before constructing the model, extensive statistical analysis of nano-islands based on TM-AFM was performed. Even though the locations of individual nano-islands were random, their density and dimension were repeatable in each sample prepared under the same fabrication condition. The resulting simulations of multiple samples were found similar to each other. Due to the random pattern as shown in Fig. 1(a), it is inappropriate to employ the rigorous coupled-wave analysis (RCWA) [13], which assumes two-dimensional (2D) periodicity in the model. Therefore, we adapted the finite difference time domain (FDTD) method instead. FDTD resolves Maxwell’s equations in time domain by discretizing the model into grids of sufficiently small volumes, and then the electric and magnetic components of each spatial coordinate are resolved in leaping manner [14]. Our threedimensional (3D) model consisted of a 1000-nm cube, and it was refined at the glass-SAM-water interface with spatial resolution of 1 nm. The top view of our model is shown in Fig. 1(b) with the nano-islands being simplified as nanospheres of 40 nm in diameter. The nanospheres are partially embedded into the glass substrate by 8 nm and probed into the analyte as shown in Fig. 1(c) and 1(d) for the p-and s-polarizations, respectively. We have also prescribed the inclined nominal incident angle to 72° so that total internal reflection occurs at the interface between glass and water. To evaluate the SAM spectral absorption, total-field scattered-field (TFSF) [15] source was used to excite the SAM-LSPR sensor. Bloch boundary condition (BC) was adopted for the x and y directions due to non-zero wave vector in the propagating direction of incident field. Perfectly matched layer BC of 512 layers was used in the z directions to absorb the transmitted and reflected lights. The TFSF source and the BC were separated far enough to avoid any interaction in between which might have introduced errors in calculation. The simulation was performed for the duration of 50 fs with time steps of 0.0038 fs. The results to be presented below are considered to have reached the condition close to that of steady state. This is because a longer duration of 500 fs was used in another simulation and the results between the simulations with 50 fs and 500 fs were nearly the same. To examine the phase response of the same structure, we replaced the TFSF source with a 2D plane-wave source with extended coverage over the simulated volume so as to avoid the edge effect. Both TFSF and plane-wave source covered spectral range from 400 nm to 1000 nm with 2-nm spectral resolution. The spectral absorption was represented by the Mie efficiency, which is defined as the ratio between the absorption cross-section and the individual geometrical area πr 2 . Thus, the greater the absorption cross section, the larger the Mie efficiency. It is known that the incident polarization is decisive to the SPR extinction and phase response. However, the LSPR extinction and the cooresponding phase transition of these random SAM gold nano-islands excited via total interal reflection (TIR) at a dielctric interface had not been reported to our awareness. The TIR scenario differs further from the reported case of direct transmission

1925

Fig. 2. (a) Mie absorption efficiency of p- (red) and s- (blue) polarized components calculated with TFSF source. (b) Corresponding phase transition with plane-wave source for each polarization and TIR incident condition.

[16] such that the SAM gold nano-islands interact solely with the evanescent field instead of those free-space propagating photons. We present the numerical results of spectral absorption and phase response under p-and s-polarization respectively in Fig. 2. As shown in Fig. 2(a), the absorption spectra of the two polarizations show remarkable differences. A broad absorption band with Mie efficiency of about 5.0 is observable in the spolarized spectrum, and it is centered at about 550 nm with full width at half-maximum (FWHM) of 50 nm. This is believed to be corresponding to the intrinsic extinction band of isolated gold nanospheres [17]. On the other hand, the p-polarized spectrum demonstrates two obvious absorption peaks. The first one shows Mie efficiency of about 13.0 and centered at approximately 600 nm with FWHM of 16 nm. The second peak is located at 680 nm with Mie efficiency up to about 45.0 and the FWHM is reduced to 21 nm. The amplification of Mie absorption efficiency and red-shift of the absorption peaks indicate that the SAM-LSPR effect is indeed enhanced by the p-polarization upon TIR incident. This is very similar to conventional SPR devices in which the s-polarized component may serve as the reference at the red-shifted wavelengths. The phase response was evaluated with frequency monitor being placed underneath the SAM gold nano-islands inside the glass substrate to record the spectral phase transition. Figure 2(b) shows the phase transitions of both p-and s-polarized components. The s-polarization undergoes absolute phase transition of about π radians from approximately 550 nm to 600 nm. The p-polarization demonstrates absolute phase transition of approximately 6.0 radians over the same spectral range. By mutual subtraction, the absolute differential phase between p-and s-polarized components is calculated to be 3.1 radians at 596 nm. Thus, the maximum differential phase transition coincides with the first p-polarized absorption peak at about 600 nm as observed in Fig. 2(a). However, there is no phase transition that can be observed for both p-and s-polarization beyond 600 nm in Fig. 2(b) even in the spectral range of the second absorption peak at around 680 nm in Fig. 2(a). In order to verify the above findings from the FDTD simulation and benchmark the SAM-LSPR differential phase performance for refractive index sensing, we incorporate our common-path spectral interferometer (CPSI) [18] with the SAM to test against sodium chloride

1926

OPTICS LETTERS / Vol. 40, No. 9 / May 1, 2015

Fig. 3. Common-path spectral interferometric sensing system used to measure the differential phase induced by LSPR effect of the SAM gold nano-islands (SAM AuNIs) at various refractive indices.

solutions of various concentrations. Figure 3 shows the common-path spectral interferometer we used in benchmarking the refractive index sensing performance of the SAM. This common-path sensing system consists of a temperature-stabilized white-light LED source (LedEngin, LZ1-00W00) with emission spectrum from 480 nm to 730 nm with FWHM of 150 nm. Two pieces of linear broadband polarizers (Edmund Optics, #89602) were inserted into the optical path. A customized undoped yttrium vanadate (YVO4 ) birefringent crystal (United Crystal) of 315-μm thickness was added to introduce sufficient optical path difference between the two polarizations. A BK7 right-angle glass prism (Edmund Optics, #32-336) was used to fulfill TIR and excite LSPR with the SAM gold nano-islands at the glass/analyte interface. A flow chamber was secured on top of the glass prism to bring sodium chloride solution into contact with the SAM. A spectrometer (Avantes, AvaSpec-ULS2048) covering 300 nm to 970 nm with resolution of 0.8 nm was used to capture the CPSI signal. The overall sensing spot was 10 mm2 . The convolution operation [19] of our CPSI system can be simplified by the following equation, E p λE s λ ⊗ E p λE s λ E 2p λE 2s λ2E p λEs λ cosφOPD φLSPR  Aλ  Bλ cosφOPD φLSPR ;

(1)

in which E p λ and E s λ are the electric field amplitudes of the p-and s-polarizations at a certain wavelength λ, Aλ is the mean spectral interference intensity, Bλ is the visibility term of the interferogram, φOPD is the optical path difference introduced by the birefringent crystal, and φLSPR is the differential phase induced by the SAM-LSPR effect of particular interest. To study the differential phase response of the SAM gold nano-islands to refractive index variation ΔRI, we introduced distilled water RI 1.3330 and aqueous sodium chloride (NaCl) solutions of 2% RI 1.3365 and 4% RI 1.3400 by weight into the flow chamber sequentially. The zero-mean spectral interferograms of the three cases are shown in Fig. 4(a). It is obvious that the interference amplitude is diminished dramatically from 595 nm to 600 nm, which is due to the transformation of photon to LSPR of the SAM gold nano-islands. Besides, differential phase transition emerges as the interference peak shifts at about 598 nm as the RI increases, whereas other portions of the spectral interferogram show no obvious change. The extinction wavelength and the

Fig. 4. (a) Spectral interferograms with water and NaCl solutions being injected to the flow cell. (b) Temporal phase shift recorded with alternative NaCl solutions flowing for 1500 s.

absorption bandwidth are in good agreement with the FDTD computation. By WFT calculation, the corresponding phase change was determined as 6.28 radians (water to 2% NaCl), i.e., the interference changed from a smooth green line (water) to a full cycle oscillation in red (2% NaCl) as shown in Fig. 4(a). The phase change was also determined as 2.01 radians from 2% to 4% NaCl. Since most label-free biosensing experiment was measured against time [20], we also evaluated the temporal response of the SAM to bulk refractive index alternation by changing the NaCl concentration. NaCl solution of 4% by weight was flowed continuously at a rate of 100 μl/min for the first 500 s, and then 8% NaCl by weight RI 1.3470 was injected into the flow chamber at the same rate as indicated by the first arrow in Fig. 4(b). After flowing continuously for another 500 s, 4% NaCl was injected back into the flow chamber and flowed nonstop as shown by the second arrow. By subtracting the initial reference phase from every data point, we obtained the temporal phase response as shown in Fig. 4(b). With injection of 8% NaCl solution, the RI increased by 7.0 × 10−3 , i.e., from 1.3400 to 1.3470, and the phase changed dramatically by about 1.23 radians at approximately 597 nm. As the 8% NaCl solution was displaced by the 4%, the phase recovered to its original value. The recovery of the phase value in the final state suggests that the SAM gold nano-islands had survived the flushing rate of 100 μl/min and confirms the structural integrity of the partially embedded SAM gold nano-islands formed on the BK7 glass substrate. Assuming the standard deviation of phase fluctuation is 1.75 × 10−4 radians and using the well-known formula, RIRbulk

δnbulk  stdδφLSPR ; δφLSPR

(2)

the bulk refractive index resolution RIRbulk  of the SAMLSPR device is estimated to be 9.75 × 10−8 RIU at approximately 600 nm. The spectral position of the maximum differential phase shift is in good correlation with the FDTD prediction. The calculated RIR is about the same order of magnitude as our previous SPR configuration using gold film [18], and it has improved substantially over those reported by Svedendahl et al. [21], who stated that the bulk sensitivity of LSPR devices were several orders of magnitude behind their SPR counterpart via wavelength interrogation. It is noted that by changing the gold film thickness as well as the optimal annealing condition, the diameter of the nano-islands changed accordingly. From this, it was found that the combination of these processing

May 1, 2015 / Vol. 40, No. 9 / OPTICS LETTERS

parameters provided a tunable sensing condition, such that the resulting sensitivity level from 560 nm to 620 nm is as good as that of 600 nm mentioned above. To conclude, we have successfully predicted the polarization-dependent spectral absorption and differential phase response of the randomly distributed SAM gold nano-island structure with FDTD computation. Besides, we have also verified the numerical model with our common-path spectral interference sensing system. Both numerical calculation and experimental results show good agreement with each other. With the simple fabrication method, the SAM-LSPR chip consumes only 1/10 of the gold material comparing to thinfilm SPR sensor. Yet, it is able to achieve bulk refractive index resolution of 9.75 × 10−8 RIU, which is comparable to its SPR counterpart. To confirm on the biosensing capability of the SAM gold nano-islands, biosensing test with surface functionalization and biomolecules interaction is being executed. The initial results are promising. The full results will be reported in a separate publication. With advantages on material cost, simple fabrication, and competitive performance, we believe the differential phase detecting SAM-LSPR sensor would be attractive to the global biosensing community. This work was supported by a grant from City University of Hong Kong (Project No. 7004013). The authors would like to thank the Center of Super-Diamond and Advanced Films (COSDAF) of City University of Hong Kong for their generous support of the FDTD Solutions software. References 1. V. G. Kravets, F. Schedin, and A. N. Grigorenko, Phys. Rev. Lett. 101, 087403 (2008). 2. J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao, and R. P. Van Duyne, Nat. Mater. 7, 442 (2008).

1927

3. K. M. Mayer and J. H. Hafner, Chem. Rev. 111, 3828 (2011). 4. L. Niu, K. Cheng, Y. Wu, T. Wang, Q. Shi, D. Liu, and Z. Du, Biosens. Bioelectron. 50, 137 (2013). 5. W. Kubo and S. Fujikawa, Nano. Lett. 11, 8 (2011). 6. V. G. Kravets, F. Schedin, R. Jalil, L. Britnell, R. V. Gorbachev, D. Ansell, B. Thackray, K. S. Novoselov, A. K. Geim, A. V. Kabashin, and A. N. Grigorenko, Nat. Mater. 12, 304 (2013). 7. M. Farrokh Baroughi, D. Dachhepati, U. Gautam, K. Bayat, and S. May, Opt. Lett. 38, 2153 (2013). 8. T. Karakouz, A. B. Tesler, T. A. Bendikov, A. Vaskevich, and I. Rubinstein, Adv. Mater. 20, 3893 (2008). 9. R. S. Moirangthem, Y. C. Chang, and P. K. Wei, Biomed. Opt. Express 2, 2569 (2011). 10. H. Yang, W. Lee, T. Hwang, and D. Kim, Opt. Express 22, 28412 (2014). 11. R. S. Moirangthem, M. T. Yaseen, P.-K. Wei, J.-Y. Cheng, and Y.-C. Chang, Biomed. Opt. Express 3, 899 (2012). 12. S. P. Ng, C. M. L. Wu, S. Y. Wu, and H. P. Ho, Opt. Express 19, 4521 (2011). 13. S. Pechprasarn and M. G. Somekh, Biomed. Opt. Express 5, 1744 (2014). 14. A. Taflove and S. C. Hagness, Computational Electrodynamics: The Finite-Difference Time-Domain Method, 3rd ed. (Artech House, 2005). 15. J. B. Schneider, IEEE T. Antenn. Propag. 52, 3280 (2004). 16. T. Yang and H. P. Ho, Opt. Express 17, 11205 (2009). 17. X. Liu, M. Atwater, J. Wang, and Q. Huo, Colloid. Surface B 58, 3 (2007). 18. S. P. Ng, F. C. Loo, S. Y. Wu, S. K. Kong, C. M. Wu, and H. P. Ho, Opt. Express 21, 20268 (2013). 19. A. Lipson, S. G. Lipson, and H. Lipson, Optical Physics, 4th ed. (Cambridge University, 2011). 20. F.-C. Loo, S.-P. Ng, C.-M. L. Wu, and S. K. Kong, Sensor. Actuat. B 198, 416 (2014). 21. M. Svedendahl, S. Chen, A. Dmitriev, and M. Kall, Nano. Lett. 9, 4428 (2009).

Differential phase-detecting localized surface plasmon resonance sensor with self-assembly gold nano-islands.

Self-assembly (SAM) gold nano-islands are fabricated by two-step thin-film deposition-annealing method. Despite random distribution of the SAM, the p-...
409KB Sizes 0 Downloads 10 Views