We investigated enhancement of sensitivity of sensors based on metallic photonic crystals through tuning the thickness of the waveguide layer by pulsed laser deposition. Thicker waveguides made of InGaZnO allow double resonance of Fano coupling modes due to plasmonic-photonic interactions. Tuning the angle of incidence enables overlap between these doubly resonant modes, which induces much enlarged and spectrally narrowed sensor signals, leading to significantly enhanced sensitivity of the sensor device. The thickness of the waveguide layer is found to be a crucial structural parameter to improve sensitivity of the MPC sensors.
© 2014 Optical Society of America
The sensitive spectroscopic response of the metallic nanostructures based on localized surface plasmon resonance has been applied widely in sensors for detecting label-free molecular binding or refractive-index changes. A variety of plasmonic sensors have been demonstrated [1–6]. The sensitivity of such sensors is dependent strongly on the shape, the size, and the arrangement of the metallic nanostructure arrays [7,8]. However, pure plasmonic response to the environmental refractive index change is limited [9–13], which is due to the small spectral shift and broad-band resonance of localized surface plasmons. It is crucial to incorporate new mechanisms so that the effect of plasmon resonance may be amplified in its amplitude and spectral shift. Arranging the metallic nanostructures into periodical arrays, which produces the so-called metallic photonic crystals, is an effective approach to induce plasmonic-photonic coupling and consequently enhance the response sensitivity of the corresponding device. Waveguide metallic photonic crystals (MPCs) have been demonstrated as a robust sensor to detect bioreactions with high sensitivity , which is based on the coupling between plasmon resonance of gold nanostructures and resonance of the waveguide grating structures [15,16]. The sensor signal may be amplified and the sensitivity may be enhanced largely through this kind of Fano resonance.
Recently, we demonstrate that the sensitivity of this kind of sensor may be enhanced by optimizing the angle of incidence, so that the waveguide resonance mode overlaps the most sensitive spectral position on the extinction spectrum of plasmon resonance . However, optimization of the structural parameters is another aspect for improving the sensitivity of such sensors. This involves adjustment of the configurations of both the plasmonic gratings and the waveguides. In this work, we achieve significant enhancement of the sensitivity through adjusting the thickness of the waveguide layer, which enables simultaneous resonance of two Fano coupling modes and consequent overlapping between them through tuning the angle of incidence.
2. Pulsed laser deposition of waveguide layers with controllable thickness for MPCs
Pulse laser deposition (PLD) has been employed to fabricate waveguide layers on substrates of fused silica with an area 10 × 10 mm2 and a thickness of 1 mm. The mixture of In2O3, Ga2O3, and ZnO powders has been used to prepare the target with a composition ratio of 0.75:0.1:0.15. A frequency-tripled Nd:YAG laser (Spectra-Physics GCR-170) with a pulse duration of 10 ns and a pulse energy of 40 mJ was used to perform the deposition of the waveguide layers. The UV laser beam at 355 nm has a diameter of about 10 mm at the direct output and was focused to an area of about 500 μm onto the surface of the target. A deposition speed of about 9 nm/minute has been measured for such a PLD system. InGaZnO thin films deposited homogeneously onto the fused-silica substrate supplied the waveguide layers with excellent transmission in the visible spectrum. Figure 1(a) shows the atomic force microscopic (AFM) image of a 100-nm InGaZnO film, which shows a surface modulation of about 5 nm, implying excellent surface smoothness and homogeneity in the applications as waveguides. Figure 1(b) shows the transmission spectrum of the deposited films with a thickness of about 100 nm, where different composition ratios have been used. More than 90% transmission has been measured almost for the whole visible band from 450 to 750 nm for a composition ratio of In2O3:Ga2O3:ZnO = 0.75:0.1:0.15 (red curve). Thus, InGaZnO films by PLD are suitable to be used as waveguides of MPCs and PLD supplies a convenient approach to adjust their thickness. Furthermore, the deposited InGaZnO films are thermally stable and they are not damaged by the annealing process in the fabrication of the MPCs. In this work, three different InGaZnO waveguide layers with a thickness of 110, 270, and 350 nm have been employed as the waveguides, where a thin-film analyzer F20 from Filmetrics, Inc. was used to measure the thickness of the deposited InGaZnO films.
3. Fabrication of MPCs consisting of periodically gold nanolines on InGaZnO waveguides
Solution processed method was used to fabricate the MPC structures, as described in our previous publications [15,17]. Here, in the fabrication of MPCs on the InGaZnO waveguides, similar procedures have been employed and were found to work well. First, a photoresist (PR) template grating was produced through interference lithography using a He-Cd laser at 325 nm, where S1805 photoresist was used as the recording medium and a grating period of 400 nm was employed for all of the fabrications in this work. Chemically synthesized gold nanoparticles in xylene with a concentration of 100 mg/ml were then spin-coated onto the PR grating. A subsequent annealing process at 350 °C in a Muffel furnace finishes the fabrication procedures, which enables melting and confinement of the gold into PR grating grooves to form gold nanolines. Figure 2 shows schematically the structural composition of the waveguide MPCs and the light-incident geometry for optical extinction spectroscopic measurements, where Λ is the grating period, w is the width of the gold nanolines, T is the thickness of the InGaZnO waveguide. For the light incident at an angle of θi, TM and TE polarizations are perpendicular and parallel to the gold nanolines, respectively.
Figures 3(a) and 3(b) show the scanning electron microscopic (SEM) and AFM images of the resultant MPC structures, respectively. As has been discussed before , the gold nanolines are easily broken into segments during the heating/cooling process, although this does not influence the spectroscopic response of Fano coupling. In the MPCs shown in Fig. 3, an InGaZnO layer with a thickness of about 110 nm has been deposited as the waveguide. According to Fig. 3, the gold nanolines have a period of about 400 nm and a modulation depth of more than 20 nm. It should be noted that photoresist was not removed in the annealing process at 350 °C. Thus, the modulation depth of the Au/PR-interlaced grating actually measured the height of the gold nanolines above the PR surface. The width of each gold nanoline is about 180 nm, corresponding to a duty cycle (w/Λ) of about 37.5%. The duty cycle, the width, and the height of the gold nanolines were not influenced much by the thickness of the InGaZnO waveguide. Therefore, the structural parameters evaluated by Fig. 3 are typical for all of the MPCs investigated in this work.
4. Spectroscopic characterization of MPCs with different waveguide thicknesses in air
4.1 Fano resonance modes Simultaneously excited for thick waveguides
Multiple resonance modes may be allowed for thick waveguide in waveguide MPCs. Figure 4 shows the optical extinction spectroscopic measurements on the waveguide MPCs with different InGaZnO waveguide thickness, where the optical extinction spectrum has been calculated by -log10[IS(λ)/I0(λ)], where I0(λ) denotes the transmission spectrum through the substrate coated with the waveguide layer and IS(λ) through the MPC structures. As shown in the measured extinction spectra at normal incidence (θi = 0) in Figs. 4(a) and 4(b) for TM and TE polarizations, respectively, single-mode resonance is observed for T = 110 nm and double-mode resonance is observed for T = 270 nm and T = 350 nm. For T = 110 nm, the waveguide resonance mode is observed at 660 nm for TE polarization and the Fano resonance through plasmon-photonic coupling at 658 nm for TM polarization, which are identified by a peak and a dip in the extinction spectrum, respectively. As T is increased to 270 nm, double resonance modes are observed, which can be observed at 604 and 735 nm for TE polarization and at 613 and 724 nm for TM polarization. As T is increased further to 350 nm, the double resonance modes both shifts further to the red, at 629 and 740 nm for TM polarization and at 623 and 749 nm for TE polarization. According to Fig. 4, the amplitude of the Fano coupling is as large as 0.66, as shown in Fig. 4(a), and that of the waveguide resonance mode is larger than 0.4, as shown in Fig. 4(b). The strong waveguide resonance mode and its strong coupling with plasmon resonance imply excellent performance of the MPCs. Both the single- and double-resonance modes are actually degenerate at normal incidence, where + 1 and −1 orders of diffraction take place at the same wavelength.
4.2 Overlapping of Fano resonance modes through tuning the angle of incidence
With increasing the incident angle of light, the spectral features of both the Fano- and the waveguide-resonance modes will be split into two branches, corresponding to the + 1 and −1 orders of diffraction. For the double-mode resonance at a thick waveguide, there is a chance for longer-wavelength branch of the resonance at shorter wavelength and the shorter-wavelength branch of that at longer wavelength to cross each other, so that there exists an overlap between the double resonance modes at a specific angle of incidence.
Figures 5(a) and 5(b) show the angle-resolved tuning properties of the extinction spectrum for TE and TM polarization in air, respectively, for a waveguide thickness of T = 270 nm, where the angle of incidence (θi) is increased from 0 to 26 degrees. Figures 5(c) and 5(d) show the case for T = 350 nm. For T = 270 nm, the double waveguide resonance modes get overlapped at an incident angle of 12° and at a wavelength of 666 nm for TE polarization, as shown in Fig. 5(a). Consequently, the double Fano resonance modes for TM polarization are also overlapped at the same angle of incidence while at a different wavelength of 670 nm, as identified by the strong extinction dip in Fig. 5(b) and marked out by the downward arrow. For T = 350 nm, the doubly resonant waveguide modes are also overlapped around θi = 12° and at a wavelength of 680 nm, as shown in Fig. 5(c), although a slight deviation is observed from perfect overlap. This deviation becomes more obvious for TM polarization, as shown in Fig. 5(d), and means that the perfect overlap needs to be found by much finer tuning of the incident angle around 12°. However, Figs. 5(c) and 5(d) already indicate possible overlap of the doubly resonant Fano modes at about 684 nm.
Thus, through tuning the angle of incidence, we can easily obtain the overlapped Fano resonance modes for doubly resonant MPCs based on thick waveguides. These overlapped resonance modes exhibit more sensitive response to the environmental change in refractive index, as will be demonstrated in section 5.
5. Sensitivity enhancement of MPC sensor through overlapping doubly resonant Fano modes
5.1 Sensor measurements on glucose solutions at different concentrations using MPCs with different waveguide thicknesses
Sensor measurements have been performed on glucose solutions in water using MPCs with different waveguide thickness for light incident at the normal of the substrates (θi = 0), as shown in Fig. 6. The design of the sensor device has been described in detail in . Four different samples with concentrations of 1%, 5%, 7%, and 9% have been measured, where the sensor signal has been calculated by –log[IS(λ)/I0(λ)]. Here IS(λ) denotes the transmission spectrum through the glucose solutions with different concentrations and I0(λ) denotes that through pure water. Figures 6(a), 6(b), and 6(c) show measurement results using MPCs with waveguide thickness of T = 110, 270, and 350 nm, respectively. A single feature in the extinction spectrum extending from about 550 to 700 nm is observed for a waveguide thickness of T = 110 nm, which consists of a dip at 560 nm and peak at 638 nm. The amplitude (A) of the sensor signal is defined by the peak-to-dip difference, as shown in Fig. 6(a). However, for T = 270 and 350 nm two features in the sensor signal spectrum are observed, corresponding to the doubly resonant Fano coupling, where the feature at shorter wavelength is much stronger than that at longer wavelength, implying strong plasmon resonance and large spectral shift at shorter wavelength. The dip and peak features are located at 570 and 710 nm for T = 270 nm and at 557 and 615 nm for T = 350 nm. Much larger amplitude of the sensor signal is observed for T = 350 nm than those for T = 110 and 270 nm. As summarized in Fig. 6(d), with increasing the thickness of the waveguide layer from 110, 270, to 350 nm, both the amplitude of the sensor signal and the slope of the variation dynamics with solution concentration become larger. This implies better visibility or contrast of the sensor signal and sensor sensitivity with increasing the thickness of the waveguide, according to the three MPC devices investigated here. Therefore, in section 5.2, we employed the MPCs with T = 350 nm to investigate the improvement of the sensitivity by overlapping the double Fano resonance.
5.2 Improvement of the sensitivity using the overlapped Fano resonance modes of MPCs with a waveguide thickness of T = 350 nm
As has been discussed in section 5.1, the sensitivity of the MPC sensor device is dependent strongly on the thickness of the waveguide layer. If looking back on Fig. 4, we can find that the spectroscopic response becomes stronger with increasing the thickness of the waveguide layer and understand that stronger optical extinction due to Fano coupling leads to higher sensitivity. Thus, further enhanced sensitivity may be achieved at the overlap of the doubly resonant Fano modes, where the strongest spectroscopic response or largest modulation on the extinction spectrum is induced.
Figure 7(a) shows the angle-resolve tuning properties of TM-polarized resonance modes of the MPCs with a waveguide thickness of T = 350 nm, where the MPC device was mounted in the sensor system using the geometry as depicted in  and pure water is being circulated in the chamber and channels of the sensor system. Clearly, the double Fano resonance modes get overlapped at an incident angle of about 10°, where a narrowest extinction dip with large amplitude can be observed. Thus, sensor measurements on glucose solutions are performed at θi = 10°, where the spectroscopic response is featured with a single mode degenerate of double Fano resonance. Figure 7(b) shows the sensor signals with the concentration of the glucose solution increased from 1% to 9%. A pronounced signal is observed at about 630 nm, which consists of a very narrow and deep dip followed a relatively broad peak, leading to a sensor signal increased from 0.006, 0.037, 0.0512, to 0.0704 as the concentration is increased from 1%, 5%, 7%, to 9%. The transmission spectrum through pure water has been used as the blank to calculate the extinction spectra in Fig. 7(b). Clearly, much stronger and narrower spectra of the sensor signal have been measured, as compared with the results in Fig. 6, implying much enhanced sensitivity with significantly improved visibility of the sensor signal.
Systematic sensor measurements were then performed at different angles of incidence, as shown in Fig. 8. According to Fig. 7(a), the modulation depth of the optical extinction spectra in Fig. 7(a) becomes smaller as the incident angle is larger than 10°, implying even lower response sensitivity than smaller angles. Figure 8 makes a comparison between the evolution dynamics of the sensor signal with concentration of the glucose solution for different angles (θi = 0°, 4°, 6°, 10°, 12°, 14°) of incidence. Obviously, the highest sensitivity (slope of the evolution dynamics) and largest sensor signals are obtained at θi = 10°. A simple quantitative evaluation found an enhancement factor as large as 2.25 if comparing the sensor signal at θi = 10° with that at θi = 6°.
Thus, the largest sensor signal or the highest sensitivity of the MPC sensor device has been observed for a waveguide thickness of 350 nm and at an incident angle of 10°. Inducing double Fano resonance by thick waveguides and overlapping them through optimizing the angle of incidence have been the main mechanisms.
Additionally, as the incident angle is tuned to overlap the two resonance modes, another effect has also contributed to the modulation of the response sensitivity of the sensor. We have investigated both theoretically and experimentally how the sensitivity may be optimized through changing the incident angle such that the waveguide resonance mode is coupled with the most sensitive spectral position on plasmon resonance . It can be understood that this is not a monotonic relationship, since it is dependent on the shape of the practically measured spectrum of plasmon resonance, as we can also conclude from Fig. 8.
5.3 Mechanisms for sensitivity enhancement with theoretical simulations
For better understanding the mechanisms how the overlapped resonance modes enhanced the response sensitivity of the MPC sensor, we performed numerical simulations using the GSOLVER software, as shown in Fig. 9. Basically, we used the same parameters as employed in the experiments. However, it is difficult to obtain the precise dielectric constant and its dispersion of the deposited InGaZnO waveguide, therefore, the shapes and spectral positions of the simulated spectra are not well consistent with those in the measurements. Figure 9(a) shows the transmission spectra through MPC device with 350-nm InGaZnO waveguide as the incident angle is increased from 0 to 14 degrees, where the two resonance modes get overlapped at an incident angle of about 10°. The dotted blue and pink lines in Fig. 9(a) are used to guide the observation of the tuning of the two resonance modes as enhanced transmission. Based on different diffraction and waveguiding processes, these two resonance modes have different response sensitivity to the environmental change in refractive index, so that they have different tuning rate with increasing environmental refractive index, as shown in Fig. 9(b). The inset of Fig. 9(b) shows a locally enlarged spectrum that is as enclosed by the blue circle on the red spectrum in Fig. 9(a), corresponding to an incident angle of 10° with the two resonance modes basically overlapped. For clearer observation, the incident angle is set such that the two resonance modes are slightly deviated from perfect overlapping. Then, the two resonance modes are slightly separated, leaving two peaks (A and B) and a dip within the enhanced transmission spectrum. The red and black circles in Fig. 9(b) shows the tuning curves of resonance modes A and B, respectively, with increasing the environmental refractive index from 1.33 to 1.36. Clearly, the shorter-wavelength branch (A) of the resonance at a longer wavelength shifts to the red much more quickly than the longer-wavelength branch (B) of the resonance at a shorter wavelength. As a result, the separation gap between these two resonance modes becomes wider and larger in amplitude with increasing the environmental refractive index. Thus, this overlapping effect introduces an additional mechanism to the Fano resonance mode and leads to enhanced sensor signals, which explains our experimental observations in Fig. 7(b).
We achieved flexible tuning of the waveguide thickness of the MPCs using pulsed laser deposition of InGaZnO films, which enables control of the number of resonant modes of MPCs for sensor applications. Two Fano resonance modes were excited simultaneously for the InGaZnO waveguide layers with thicknesses of 270 and 350 nm. Overlapping these two resonance modes through tuning the angle of incidence enables significantly enhanced sensor signal and response sensitivity, which was verified by sensing measurements on the glucose/water solutions with different concentrations. Thus, tuning such structural parameters as waveguide thickness, in combination with tuning of the angle of the incidence, is the practical approach to optimize the performance of both the MPCs and the corresponding sensor devices.
We acknowledge the 973 program (2013CB922404) and the National Natural Science Foundation of China (11274031) for the support.
References and links
1. A. J. Haes, L. Chang, W. L. Klein, and R. P. Van Duyne, “Detection of a biomarker for Alzheimer’s disease from synthetic and clinical samples using a nanoscale optical biosensor,” J. Am. Chem. Soc. 127(7), 2264–2271 (2005). [CrossRef]
2. J. Zhao, A. Das, X. Zhang, G. C. Schatz, S. G. Sligar, and R. P. Van Duyne, “Resonance surface plasmon spectroscopy: low molecular weight substrate binding to cytochrome P450,” J. Am. Chem. Soc. 128(34), 11004–11005 (2006). [CrossRef]
3. P. Jia, H. Jiang, J. Sabarinathan, and J. Yang, “Plasmonic nanohole array sensors fabricated by template transfer with improved optical performance,” Nanotechnology 24(19), 195501 (2013). [CrossRef]
4. N. Zhang, X. Su, P. Free, X. Zhou, K. G. Neoh, J. Teng, and W. Knoll, “Plasmonic metal nanostructure array by glancing angle deposition for biosensing application,” Sens. Actuators B Chem. 183, 310–318 (2013). [CrossRef]
5. A. Abbas, L. Tian, J. J. Morrissey, E. D. Kharasch, and S. Singamaneni, “Hot spot-localized artificial antibodies for label-free plasmonic biosensing,” Adv. Funct. Mater. 23(14), 1789–1797 (2013). [CrossRef]
8. S. K. Ghosh and T. Pal, “Interparticle coupling effect on the surface plasmon resonance of gold nanoparticles: from theory to applications,” Chem. Rev. 107(11), 4797–4862 (2007). [CrossRef]
9. M. E. Stewart, N. H. Mack, V. Malyarchuk, J. A. N. T. Soares, T.-W. Lee, S. K. Gray, R. G. Nuzzo, and J. A. Rogers, “Quantitative multispectral biosensing and 1D imaging using quasi-3D plasmonic crystals,” Proc. Natl. Acad. Sci. U.S.A. 103(46), 17143–17148 (2006). [CrossRef]
10. J. McPhillips, A. Murphy, M. P. Jonsson, W. R. Hendren, R. Atkinson, F. Höök, A. V. Zayats, and R. J. Pollard, “High-performance biosensing using arrays of plasmonic nanotubes,” ACS Nano 4(4), 2210–2216 (2010). [CrossRef]
11. J. C. Yang, J. Ji, J. M. Hogle, and D. N. Larson, “Metallic nanohole arrays on fluoropolymer substrates as small label-free real-time bioprobes,” Nano Lett. 8(9), 2718–2724 (2008). [CrossRef]
13. J. Ji, J. G. O’Connell, D. J. D. Carter, and D. N. Larson, “High-throughput nanohole array based system to monitor multiple binding events in real time,” Anal. Chem. 80(7), 2491–2498 (2008). [CrossRef]
14. X. P. Zhang, X. M. Ma, F. Dou, P. X. Zhao, and H. M. Liu, “A biosensor based on metallic photonic crystals for the detection of specific bioreactions,” Adv. Funct. Mater. 21(22), 4219–4227 (2011). [CrossRef]
15. X. P. Zhang, B. Q. Sun, R. H. Friend, H. C. Guo, D. Nau, and H. Giessen, “Metallic photonic crystals based on solution-processible gold nanoparticles,” Nano Lett. 6(4), 651–655 (2006). [CrossRef]
16. X. P. Zhang, S. F. Feng, J. Zhang, T. R. Zhai, H. M. Liu, and Z. G. Pang, “Sensors based on plasmonic-photonic coupling in metallic photonic crystals,” Sensors (Basel) 12(12), 12082–12097 (2012). [CrossRef]
17. X. P. Zhang, H. M. Liu, and S. F. Feng, “Solution-processible fabrication of large-area patterned and unpatterned gold nanostructures,” Nanotechnology . 20, 425303 (2009).