The production of inexpensive, large-scale, uniform substrates for surface-enhanced Raman scattering (SERS) is a key to popularize its usage in chemical and biological detection. We demonstrate a flexible nano-imprinted hexagonally patterned SERS-active substrate. Its electromagnetic enhancement factor was optimized by the thickness adjustment of its silver over-coated film. The experimental data show a good correspondence with the theoretical prediction. Such substrate was shown to exhibit high uniformity and reproducibility with a variation of less than 2%, offering a potential of greatly exploiting such substrate in infield biocide monitoring.
© 2011 OSA
Surface enhanced Raman scattering (SERS) has attracted a lot of attention lately, owing to its ability to provide finger-print vibrational signature of molecular species with high sensitivity and thus its potential usage in chemical and biological sensors [1–3]. Recently, researchers have developed many nanostructures, such as inverted pyramidal pits , nanorods [5, 6], and other specially engineered geometries (nano-egg, nano-rice, nano-crescent, nano-bowl, etc [7–9].), to obtain prominent SERS signals. One of the most important issues in the development of the SERS technique is to realize uniform, reproducible SERS-active substrates with high electromagnetic enhancement by simple and inexpensive means. One well-known fabrication method realized by Van Duyne’s group is to use a close-packed monolayer of nanospheres as a mask in the metal deposition process to fabricate different shapes of nanoparticles arrays with tunable localized surface plasmon resonance spectra . Another related approach is to use the array as a substrate to fabricate silver film over nanosphere structures . Successful in vivo glucose detection was demonstrated with such substrates . The other notable method is to fabricate two-dimensional silver nanoparticle arrays embedded in nanochannels of anodic aluminum oxide . Adjusting the gap between adjacent nanoparticles tunes its optical properties. Such substrate has been applied to bacterial detection and to the monitoring of their antibiotic-induced changes . In addition to the above two methods, imprinting with precisely made two-dimensional nano-scale molds also confers a simple approach for mass production of SERS-active substrates [14–16].
The basic principle of making SERS-active substrates lies on producing subwavelength structures that enable the resonant interaction with applied electromagnetic radiation (plasmon resonance) in both absorption and emission , thus enhancing local electric field for Raman scattering. Based on this principle, we show here another large-scale, cheap fabrication method of SERS-active substrates based on imprinted two-dimensional hexagonally patterned structures with a silver nanometer-scaled over-coating to induce plasmon resonance. The intention is to use this structure to enhance the absorption of the excitation laser beam while performing Raman measurements. The silver film thickness was adjusted to optimize its SERS performance. The absorption spectra of the substrates with different silver film thicknesses presented a good correspondence with electrodynamic calculation based on finite difference time-domain method. The relationship between the SERS signal and the absorption spectrum is discussed. With such substrate, the trace detection of malachite green (MG) in water is demonstrated down to 10−8 M, corresponding to <5 μg/l. MG is highly effective against protozoal and fungal infections and had been extensively used as biocide in aquaculture industry, scientific evidences have however indicated that residual MG might cause carcinogenesis and teratogenesis [18, 19]. For that reason, in Ireland, the concentration of MG in fish farm water effluent should not exceed 100 μg/l. This study thus shows the potential usage of SERS with the developed inexpensive, large-scale substrates in on-site monitoring of biocide concentration in water systems. The method developed in this study is to be compared with other relatively low-speed, bulky techniques, such as liquid chromatography , amperometer , capillary electrophoresis , and enzyme-linked immunosorbent assay .
2. Experiments and calculation
The two-dimensional hexagonally patterned mold was fabricated by two-beam laser interference plus nickel electroless plating, which was reported previously . Briefly, spin-coated photo-resist film on silicon wafers was triply exposed with two 351 nm Ar-ion laser beams, crossing at about 80°. The substrates were rotated laterally by 60° between exposures, resulting in hexagonal periodic patterns. After a development process, the pattern was then formed on the surface of the silicon wafers with reactive ion etching. The subsequent nickel electroless plating process then created nickel-plated mold with the hexagonal pattern. Before nano-imprinting, a self-assembled monolayer of fluoroalkyl silane was deposited on the mold surface with atmospheric-plasma chemical surface treatment to decrease its surface energy, thus improving its anti-adhesive property. The patterned mold was then imprinted onto the poly-ethylenetelephthalate (PET) substrates by hot embossing process. This was followed by raising the temperature of the polymer just above its glass transition temperature. The PET substrates were then separated from the mold, creating the replicated hexagonally patterned structure on their surfaces. Silver films of different thicknesses were subsequently deposited on the substrates. Surface profilometric calibration on separate silver films that were co-deposited on flat silicon substrates gave the resultant silver film thicknesses of d = 13.1, 21.8, 30.6, 43.7, 65.5 and 87.3 nm. Figure 1 shows the schematic cross-section, top-view scanning electron microscopic image, and topographic atomic force microscopic (AFM) image of a typical nano-imprinted SERS-active substrate with a 25 nm silver over-coating. The period (p) and modulation depth (h) of the resultant structures determined with the AFM image are about 280 nm and 140 nm, respectively.
The SERS-active substrates were placed inside an integrating sphere of a UV-visible spectrophotometer for light absorption characterization. Their absorptance spectra in air were determined by the subtraction of the measured reflection and transmission spectra from unity. Dropping a water droplet on the surface and then placing a cover glass slip on top forms a uniform water layer above the substrate. The corresponding absorptance spectra of the substrates with a top water layer were also determined similarly. Distinct quantities of MG molecules were dissolves in water to prepare solutions of different concentrations. Rhodamine 6G (R6G) molecules in water was also prepared to compare its SERS performance with that of MG. 1 μL of the sample solution was introduced to the substrate for Raman measurements. The following application of a cover glass on the substrate prevented water evaporation during the measurements. All the Raman measurements were performed with a commercial micro-Raman system. A 515 nm solid-state laser (632.8 nm HeNe laser) served as the excitation source for MG (R6G) to avoid direct photoexcitation, making the Raman measurements being nearly non-resonant and thus minimizing the disturbance of laser-induced fluorescence and resonance Raman scattering. With a quarter-wave plate, the excitation polarization was set to be circular, thus eliminating the possible polarization dependence caused by the directional character of the hexagonally ordered pattern of the substrates. The 1 mW laser beam was focused with a 20× objective lens, resulting in a focal spot diameter of about 6.5 μm and therefore an illuminating intensity of 3 kW/cm2. The scattered radiation was collected in a backward direction with the same objective lens and was sent through a spectrograph to a liquid nitrogen-cooled charge-coupled device for spectral recording. The spectral resolution was calibrated to be less than 0.68 cm−1. The signal integration time was varied from 1 to 10 sec, depending on the sample concentration.
Finite-difference time-domain calculations were performed to simulate the electrodynamic behaviors of the substrates whereas the two-dimensional hexagonally patterned structure is approximated with periodical hemispheres, as depicted in Fig. 1(a). The refractive index of PET in the wavelength of interest was set to 1.58. The incident direction of the excitation light source is downward towards the silver film. The simulation was performed in a unit cell with anti-symmetry or symmetry boundary conditions on the substrate surface — depending on the polarization of the excitation light source — and with perfect matching layers on both the upper and lower boundaries of the calculation domain. The calculations were performed for the polarization direction of the incident optical wave being parallel with and perpendicular to the axis along the two adjacent hemispheres, in which the linear polarization of the excitation source was assumed to elucidate the underlying optical characteristics. The mesh sizes are 1 nm in the lateral direction but 1.5 nm in the vertical direction. The extracted transmission and reflection spectra for the two polarization configurations were averaged to mimic the absorptance spectra with a random polarization of the lamp source in the absorption measurement, as no chiral asymmetry is expected in the hexagonal ordered structure.
3. Results and discussion
Figure 2 shows the absorption spectra of the SERS substrates with different silver film thicknesses (d) in air and with a top water layer. All the spectra exhibit a prominent peak in the wavelength range from 400 nm to 900 nm, reflecting plasmon resonant excitation. The absorption peak wavelength is decreased with the increment of d and reaches a plateau as d > 65 nm. The absorption peak wavelength obtained with a top water layer is red-shifted with respect to the one obtained in air, which is caused by the extra dielectric screening effect from the top water layer. The calculation also shows an absorption peak at a shorter wavelength (not shown here). The consistency between the experimental and calculation results is demonstrated in Fig. 3 . The resonant wavelengths extracted from the absorption spectra of the substrates at different silver film thicknesses agree with the ones obtained from calculations. The two resonant peaks correspond to two localized plasmon modes: symmetric and asymmetric . Their identification has been similarly pointed out in the investigation of light transmission through corrugated silver films at normal incidence by Dvoynenko, Samoylenko and Wang . The symmetric mode corresponds to the low-frequency resonant peak, while the asymmetric mode corresponds to the high-frequency one. They have different dispersion relations. Similar to a flat metallic slab, as the slab thickness is increased, the asymmetric mode undergoes red-shifting, while the symmetric mode experiences blue-shifting [27, 28]. This trend, as demonstrated in Fig. 3, can be interpreted with plasmon hybridization theory  in which the coupling between these two modes decreases with the increment of d. As a final note, the experimentally observed absorption spectra were broader than that obtained from the calculation. This may be due to the following two facts. Firstly, the surface roughness of the deposited silver films, as shown in the inset of Fig. 1(b), provides extra electron scattering mechanism that facilitates the dephasing of plasmon oscillation and therefore increases the spectral width of plasmon resonance . Secondly, the slight non-uniformity in the fabricated hexagonally patterned structure renders an additional inhomogeneous contribution to the overall spectral characteristics. Owing to the spherical symmetry of the hemispheric protrusion in the structure, the electrical field component expressed in spherical coordinate — in which the origin is at the spherical center of the hemisphere (Fig. 1) — is anticipated to reveal its spatial symmetric properties. Figure 4 shows a cross-sectional view of calculated radial electric field components of the SERS-active substrate with a 15 nm silver film at the two resonant wavelengths of 338 and 643 nm, corresponding to asymmetric () and symmetric () modes, respectively. The chosen thickness of the silver over-coated film is thin enough to portray the two distinct plasmon modes that are sustained by such hemispheric structure. Note that changes phase across the PET-silver and silver-air interfaces along the radial direction, while changes phase across the silver film. Such distinct characteristics are illustrated in the rescaled distributions in Fig. 4. As a final note, these two plasmon modes are not propagating surface plasmon modes. The excitation of such delocalized electromagnetic modes requires the satisfaction of both the dispersion relation of surface plasmon at the silver-air (or silver-water) interface and the Bragg diffraction condition. The calculation yields a period of >349 nm for both excitation wavelengths (515nm and 633nm), which is larger than the period of our substrates (280 nm).
Figure 5 shows the Raman spectrum of R6G solution obtained on the SERS-active substrate with a 25-nm silver over-coating film which is compared with that acquired on the flat glass substrate with the same silver coating. Note that the Raman signature of 10−3 M of R6G in water on the silver-coated flat glass substrate is barely recognizable. The Raman enhancement factor is estimated to be larger than 105. The performance of the SERS-active substrate with a 25-nm silver coating in the detection of MG in water is illustrated in Fig. 6 . The standard error obtained by analyzing ten Raman spectra of 10−6 M of MG in water measured at ten different locations of a typical SERS-active substrate shows less than 2% of percentage error at the most intense Raman peak of 1620 cm−1. MG is a triarylmethane dye with dimethylamino substituents at the para positions of two of the three phenyl rings. In the range from 1100 to 1700 cm−1, nine prominent peaks emerge and match with the peaks observed with normal Raman scattering excited at 514.5 nm . In the concentration curve shown in Fig. 6(b), the recorded Raman intensity at 1620 cm−1 increases linearly with the MG concentration ranging from 10−8 to 10−6 M. The intensity then saturates from 10−6 to 10−5 M, suggesting that the average MG molecule does not enjoy the same amount of Raman enhancement as multilayers of MG molecules accumulate on the substrate surface. This concentration curve represents the detection of MG with this SERS substrate reaching less than 5 μg/L and a dynamical range of 100.
In surface-enhanced Raman scattering, the electromagnetic enhancement factor is related to the localized electric field of the enhancer where the molecules reside [32, 33]. Specifically, in this study, localized enhanced electric field at the plasmon resonant wavelength interacts with the molecules adsorbed on the surface of the silver over-coating, producing oscillating space dipoles at a Raman-shifted frequency. The propagating field produced by these dipoles is then greatly amplified by the plasmonic interaction with the nearby nanostructure. In other words, the electromagnetic enhancement factor for Raman scattering in a backward-collected configuration  is given by35], MEM is proportional to A(ωL)A(ωS) according to Eq. (1). The calculated values of the normalized electromagnetic enhancement factors were obtained by product of A(ωL) and A(ωS) of the measured absorptance in Fig. 2(b), where A(ωL) and A(ωS) are the absorptances at the excitation wavelength and the wavelength of the Raman-shifted Stokes radiation, respectively. Figure 7 shows normalized Raman intensities of MG at 1620 cm−1 excited at 514.5 nm and R6G at 1365 cm−1 excited at 632.8 nm on the SERS-active substrates as a function of the silver film thickness. The Raman intensities reach maxima for d = 43.7 and 21.8 nm, respectively. For comparison, the calculated MEM are also plotted in Fig. 7 (open squares). Notice that the relation of silver thicknesses with the experimental Raman intensities and the calculated MEM are in good agreement. This correspondence suggests that the absorptance measurement is a simple approach to estimate the optical performance of the hexagonally patterned, silver-coated, hemispherically protruded substrates.
As shown in Fig. 7(b), the SERS-active substrate with a silver film thickness of 25 nm gives the optimal performance in the case of using the 632.8 nm laser as the excitation source in Raman measurements performed in water. In this case, the resonant wavelength of the low-frequency symmetric surface plasmon mode coincides with the excitation wavelength. The calculated electric field intensity, shown in Fig. 8 , exhibits high values at the gaps between adjacent hemispherical protrusion structures. It is anticipated that the molecules residing at these so-called “hot spots” enjoy the maximal Raman enhancement factor. This scenario can be compared with the case of the SERS substrates based on two-dimensional silver nanoparticle arrays grown on nanochannels in anodic alumina [5, 36], except that, in such case, the maximal plasmon-mediated localized field is created by adjacent silver nanoparticles. Similarly, the electrical field intensity in this case, shown in Figs. 8(b) and (c), decays drastically away from the top surface of the silver over-coated film, exhibiting an evanescent character. The intensity enhancement factor at the hot spots of our SERS substrates reaches its maximal value of >1600 at the silver-water interface, as presented in Fig. 8(c), resulting in an expected Raman enhancement factor of more than 2×106. Such theoretical prediction can be compared with the experimental Raman data presented in Fig. 5, in which the experimental enhancement factor confers the average electromagnetic enhancement contribution over the whole silver surface.
Surface-enhanced Raman scattering has been successfully demonstrated on silver-coated, two-dimensional hexagonally patterned, hemispherically protruded structures fabricated with nano-imprint on flexible PET substrates. Their optical properties have been analyzed experimentally and investigated with electrodynamic calculation. The surface plasmon polariton modes have been identified with the calculations, and their resonant wavelengths agree well with the experimental results. Based on the acquired absorptance data, we show that the dependence of the SERS signal on the silver film thickness is in good agreement with the corresponding dependence of the calculated electromagnetic enhancement factor. We demonstrate that the SERS detection of malachite green with such substrates can reach 10−8 M, indicating that this technique can potentially become a quick and simple detection method of residual antiseptics in aquaculture industry.
References and links
1. R. J. C. Brown and M. J. T. Milton, “Nanostructures and nanostructured substrates for surface-enhanced Raman scattering (SERS),” J. Raman Spectrosc. 39(10), 1313–1326 (2008). [CrossRef]
2. S. Lal, N. K. Grady, J. Kundu, C. S. Levin, J. B. Lassiter, and N. J. Halas, “Tailoring plasmonic substrates for surface enhanced spectroscopies,” Chem. Soc. Rev. 37(5), 898–911 (2008). [CrossRef] [PubMed]
4. N. M. B. Perney, J. J. Baumberg, M. E. Zoorob, M. D. B. Charlton, S. Mahnkopf, and C. M. Netti, “Tuning localized plasmons in nanostructured substrates for surface-enhanced Raman scattering,” Opt. Express 14(2), 847–857 (2006). [CrossRef] [PubMed]
5. H. H. Wang, C. Y. Liu, S. B. Wu, N. W. Liu, C. Y. Peng, T. H. Chan, C. F. Hsu, J. K. Wang, and Y. L. Wang, “Highly Raman-enhancing substrates based on silver nanoparticle arrays with tunable sub-10 nm gaps,” Adv. Mater. 18(4), 491–495 (2006). [CrossRef]
7. M. W. Knight and N. J. Halas, “Nanoshells to nanoeggs to nanocups: optical properties of reduced symmetry core-shell nanoparticles beyond the quasistatic limit,” N. J. Phys. 10(10), 105006 (2008). [CrossRef]
9. J. Ye, L. Lagae, G. Maes, G. Borghs, and P. Van Dorpe, “Symmetry breaking induced optical properties of gold open shell nanostructures,” Opt. Express 17(26), 23765–23771 (2009). [CrossRef]
11. J. A. Dieringer, A. D. McFarland, N. C. Shah, D. A. Stuart, A. V. Whitney, C. R. Yonzon, M. A. Young, X. Y. Zhang, and R. P. Van Duyne, “Surface enhanced Raman spectroscopy: new materials, concepts, characterization tools, and applications,” Faraday Discuss. 132, 9–26 (2006). [CrossRef] [PubMed]
12. D. A. Stuart, J. M. Yuen, N. Shah, O. Lyandres, C. R. Yonzon, M. R. Glucksberg, J. T. Walsh, and R. P. Van Duyne, “In vivo glucose measurement by surface-enhanced Raman spectroscopy,” Anal. Chem. 78(20), 7211–7215 (2006). [CrossRef] [PubMed]
13. T. T. Liu, Y. H. Lin, C. S. Hung, T. J. Liu, Y. Chen, Y. C. Huang, T. H. Tsai, H. H. Wang, D. W. Wang, J. K. Wang, Y. L. Wang, and C. H. Lin, “A High Speed Detection Platform Based on Surface-Enhanced Raman scattering for monitoring Antibiotic-Induced Chemical Changes in Bacteria Cell Wall,” Plos One 4(5), e5470 (2009). [CrossRef] [PubMed]
14. R. Alvarez-Puebla, B. Cui, J. P. Bravo-Vasquez, T. Veres, and H. Fenniri, “Nanoimprinted SERS-active substrates with tunable surface plasmon resonances,” J. Phys. Chem. C 111(18), 6720–6723 (2007). [CrossRef]
15. B. Cui and T. Veres, “Fabrication of metal nanoring array by nanoimprint lithography (NIL) and reactive ion etching,” Microelectron. Eng. 84(5-8), 1544–1547 (2007). [CrossRef]
16. B. D. Lucas, J. S. Kim, C. Chin, and L. J. Guo, “Nanoimprint lithography based approach for the fabrication of large-area, uniformly oriented plasmonic arrays,” Adv. Mater. 20(6), 1129–1134 (2008). [CrossRef]
17. A. Campion and P. Kambhampati, “Surface-enhanced Raman scattering,” Chem. Soc. Rev. 27(4), 241–250 (1998). [CrossRef]
19. E. Sudova, J. Machova, Z. Svobodova, and T. Vesely, “Negative effects of malachite green and possibilities of its replacement in the treatment of fish eggs and fish: a review,” Vet. Med. 52, 527–539 (2007).
20. J. L. Allen, J. R. Meinertz, and J. E. Gofus, “Determination of malachite green and its leuco form in water,” J. AOAC Int. 77, 646 (1992).
21. K. Sagar, M. Smyth, J. Wilson, and K. McLaughin, “High-performance liquid chromatographic determination of the triphenylmethane dye, malachite green, using amperometric detection at a carbon fibre microelectrode,” J. Chromatogr. A 659(2), 329–336 (1994). [CrossRef]
22. C. H. Tsai, J. D. Lin, and C. H. Lin, “Optimization of the separation of malachite green in water by capillary electrophoresis Raman spectroscopy (CE-RS) based on the stacking and sweeping modes,” Talanta 72(2), 368–372 (2007). [CrossRef] [PubMed]
23. M.-C. Yang, J.-M. Fang, T.-F. Kuo, D.-M. Wang, Y.-L. Huang, L.-Y. Liu, P.-H. Chen, and T.-H. Chang, “Production of antibodies for selective detection of malachite green and the related triphenylmethane dyes in fish and fishpond water,” J. Agric. Food Chem. 55(22), 8851–8856 (2007). [CrossRef] [PubMed]
24. T.-L. Chang, K.-Y. Cheng, T.-H. Chou, C.-C. Su, H.-P. Yang, and S.-W. Luo, “Hybrid-polymer nanostructures forming an anti-reflection film using two-beam interference and ultraviolet nanoimprint lithography,” Microelectron. Eng. 86(4-6), 874–877 (2009). [CrossRef]
25. F. Pigeon, I. F. Salakhutdinov, and A. V. Tishchenko, “Identity of long-range surface plasmons along asymmetric structures and their potential for refractometric sensors,” J. Appl. Phys. 90(2), 852–859 (2001). [CrossRef]
26. M. M. Dvoynenko, I. I. Samoylenko, and J. K. Wang, “Suppressed light transmission through corrugated metal films at normal incidence,” J. Opt. Soc. Am. A 23(9), 2315–2319 (2006). [CrossRef]
28. P. Berini, “Long-range surface plasmon polaritons,” Adv. Opt. Photon. 1(3), 484–588 (2009). [CrossRef]
30. H. Hövel, S. Fritz, A. Hilger, U. Kreibig, and M. Vollmer, “Width of cluster plasmon resonances: Bulk dielectric functions and chemical interface damping,” Phys. Rev. B Condens. Matter 48(24), 18178–18188 (1993). [CrossRef] [PubMed]
31. H. B. Lueck, D. C. Daniel, and J. L. McHale, “Resonance Raman Study of Solvent Effects on a series of Triarylmethane Dyes,” J. Raman Spectrosc. 24(6), 363–370 (1993). [CrossRef]
32. J. G. Bergman, D. S. Chemla, P. F. Liao, A. M. Glass, A. Pinczuk, R. M. Hart, and D. H. Olson, “Relationship between surface-enhanced Raman scattering and the dielectric properties of aggregated silver films,” Opt. Lett. 6(1), 33–35 (1981). [CrossRef] [PubMed]
33. N. Félidj, J. Aubard, G. Levi, J. R. Krenn, M. Salerno, G. Schider, B. Lamprecht, A. Leitner, and F. R. Aussenegg, “Controlling the optical response of regular arrays of gold particles for surface-enhanced Raman scattering,” Phys. Rev. B 65(7), 075419 (2002). [CrossRef]
36. B. Y. Lin, H. C. Hsu, C. H. Teng, H. C. Chang, J. K. Wang, and Y. L. Wang, “Unraveling near-field origin of electromagnetic waves scattered from silver nanorod arrays using pseudo-spectral time-domain calculation,” Opt. Express 17(16), 14211–14228 (2009). [CrossRef] [PubMed]