Open Access
Add to BibSonomyAbstract
Black Si (b-Si) with gold or silver metal coating has been shown to be an extremely effective substrate for surface-enhanced Raman scattering (SERS). Here, we demonstrate that it is also a highly versatile SERS platform, as it supports a wide range of surface functionalizations. In particular, we report the use of a molecularly imprinted polymer (MIP) coating and a hydrophobic coating on b-Si to establish two different sensing modalities. First, using a MIP layer on Au-coated b-Si, we show selective sensing of two closely related varieties of tetracycline. Second, a hydrophobic coating was used to concentrate the analyte adsorbed on gold colloidal nanoparticles, thus increasing the sensitivity of the measurement by an order of magnitude. In this experiment, Au nanoparticles and analyte were mixed just before SERS measurements and were concentrated by drop-drying on the super-hydrophobic b-Si. These approaches are promising for SERS measurements that are sensitive to the aging of bare plasmonic metal-coated substrates.
© 2015 Optical Society of America
1. Introduction
Surface enhanced Raman scattering (SERS) is a highly sensitive technique for fingerprinting molecular [1–3], bacterial [4], and viral [5,6] materials. SERS uses a nano-rough surface which scatters light, resulting in localization of light intensity at particular locations on nano-tips and crevices [7]. Scattered and diffracted light can couple into a surface wave – known as the surface plasmon polariton (SPP) mode – at the interface between the substrate/metal/air (or liquid) where strong near fields are created. Parameters such as the surface morphology, granular texture of plasmonic metal coatings, their chemical and thermal stability, hydro-phobicity/philicity, and surface area [8] are all important for practical use of these substrates. Since the free propagation range of SPPs on a metal is up to several micrometers, they are scattered by nano-textured surfaces, thereby contributing to SERS. Hence, control of the nano-roughness of the substrate and plasmonic metal coating opens up intricate pathways to enhance SERS and even to make it a quantitative technique by using statistical analysis of the SERS intensity distribution on the surface [9]. Size, geometry, inter-particle separation and randomness of patterns are important to efficiently tune the Raman scattering. In particular, the raman scattering depends on the optical extinction and scales with size differently for the two contributions of surface scattering/reflection and volume losses [10, 11]. Arrays of plasmonic nanoparticles can be engineered to optimize sensing platforms for the particular excitation wavelength and Stokes component of Raman scattering. In this case localized plasmons create high intensity spots on the nanoscale, depending on the resonance between neighboring particles (shape and material dependent [11, 12]).
The active “hot-spot” areas in SERS sensors are on the scale of nanometers, hence it is difficult to directly detect analyte molecules diluted to femto- or atto-molar concentrations. Various solutions have been proposed to concentrate molecules in the desired locations of hot-spots. Detection of attomolar concentrations of rhodamine 6G was reported using a well controlled drop drying process on combined micro-nano structures [13] and, recently, on a nano-fabricated bull’s-eye structure [14]. However, these structures are very complicated to fabricate and require several processing steps. The concentration approach is very effective if the investigated solution is composed of only a few different analytes. Complex mixtures give overlapping Raman peaks and mathematical models should be applied to evaluate the measured spectra [15].
Progress in novel methods for bio-chemical surface functionalization is facilitating the high selectivity and sensitivity required for biomedical and forensic applications. These approaches can be successfully employed to localise analyte molecules in close proximity to the active SERS areas. Molecularly imprinted polymers (MIPs) have shown a high specificity to targeted molecules and various sensors for different proteins have been reported [16–19]. Moreover, the MIP surface is homogeneous and highly porous to concentrate target analytes within the limit of the surface plasmon field, which is responsible for the enhancement of Raman scattering [20]. In addition, the MIP matrix is selective and relatively stable to chemical interference from acids, bases, organic solvents and varying environmental conditions (temperature, pH etc.) [21, 22]. The combination of MIP and a lithographically-defined periodic SERS substrate has been tested [20]. However, the practical fabrication of SERS substrates would preferably avoid lithography steps, e.g. by using random patterns made by plasma or chemical etching. It has been demonstrated that the SERS intensity can be augmented by a factor of 2–3 by optimizing the randomness of SERS nanotextures or by introducing a size distribution of nanoparticles, as these attributes can be beneficial for light localization and the creation of higher intensity hot spots over a wider spectral range [11, 23, 24].
Here we have studied b-Si SERS substrates having tunable surface properties and a random distribution of the nano-features. Different aspect ratio nano- to micro-meter high Si needles were fabricated by dry reactive ion etching, which allows large area production in a single process, thus reducing the sensor cost. Hydro-phobicity/philicity of the substrate was tuned by surface chemistry and a thin MIP layer was used to augment selectivity and enhance the signal-to-noise ratio of the SERS signal from analytes.
2. Materials and methods
2.1. Black Silicon substrate
Black-Si was made out of a single side polished p-type 〈100〉 orientation silicon wafer using a reactive ion etching (RIE) process. First, substrates were rinsed in isopropanol to remove any remaining particles and dried under nitrogen flow. Then, the wafers were loaded directly into the RIE chamber for etching. It is to be noted that the surface morphology and spike geometry is highly dependent on etching parameters such as plasma power, etching gas flow rates and process pressure. Moreover, a self masking process randomly seeds the pillar positions at the start of the reactive etching process.
Two different etching systems were used to process the silicon wafers: Samco RIE-101iPH and Oxford PlasmaLab 100. The reason for using two tools was to deliver different spike geometries: the Samco tool was capable of producing pyramidal shaped spikes, while pillars with close-to-vertical sidewalls were made with the Oxford etcher. The latter was used to make different height pillars for the investigation of hydrophobicity. Etching parameters for b-Si fabrication are given in Table 1.
Table 1. Summary of b-Si etching parameters and spike heights.
2.2. Surface silanisation
Nanotextured b-Si shows super-hydrophilic behaviour due to both high roughness and hydrophilic surface chemistry. The b-Si was tuned superhydrophobic by deposition of a trichloro(1H,1H,2H,2H-perfluorooctyl)silane (PFTS) monolayer on the surface. First, the b-Si surfaces were treated for 1 minute under oxygen plasma to remove any organic contaminants and to generate surface silanol Si-OH groups. Then, the substrates were dipped into a fresh solution of PFTS (3 × 10−3 M) in hexane for 4 hours at room temperature in a desiccator. The resulting surfaces were rinsed twice in chloroform, twice in ethanol and finally dried under a gentle nitrogen flow. Figure 1 presents the static contact angle measurements over a range of nanotextured b-Si substrates with various spike heights from h = 70 nm to 3.8 μm. By changing the spike height, the static contact angle was tuned from θ = 117° (h = 70 nm) to 160° (3.8 μm); the flat Si surface had a contact angle of 90–100° after PFTS treatment. It is noteworthy that the separation between needles followed a diffusional dependence on the etching time (i.e. proportional to and was increasing together with the height of the needles. Hysteresis contact angle (HCA) measurements showed that only b-Si substrates with spike height greater than 500 nm presented a rolling ball effect (θH = 0°), confirming a Cassie-Baxter state (color-filled region in Fig. 1). A smaller spike height resulted in larger HCA (20°), confirming the Wenzel state and impalement of the droplets onto the nanotexturated surface.
Fig. 1 a) Dependence of water drop contact angle on b-Si spike height after PFTS adsorption onto the b-Si surface. The difference in shape of the needles: cylindrical (Oxford) and pyramidal (Samco); b) Images showing droplets on b-Si surfaces. See Media 1 and Media 2 for a droplet landing on b-Si.
Thiophenol was selected as an analyte for SERS measurements on the hydrophobic surfaces. A solution of 50 nm gold nanoparticles was used as an active agent for SERS. First, a 10 mM thiophenol solution was prepared in ethanol. Then, 10 μL of this solution was mixed with 10 μL of gold nanoparticles and 20 μL of Millipore deionized (DI) water, resulting in a 40 μL drop which was left on the b-Si surface to dry.
2.3. Molecular imprints
Black-Si substrates were magnetron sputtered with a 200-nm-thick layer of gold (AXXIS, Kurt J. Lesker Ltd.), making them SERS-active. A selective MIP layer was then formed on top to obtain a hybrid MIP-plasmonic sensor on a randomly nano-textured SERS substrate.
Two MIP surfaces were prepared for different species of molecules by imprinting the target molecules in an acrylamide/N,N-methyllene-bisacrylamide (AM/BIS) matrix. Widely used antibiotics such as tetracycline hydrochloride (TC) and oxytetracycline hydrochloride (OTC) were the target molecules. First, phosphate buffer of 0.1 M and pH 7 was prepared using Na2HPO4 · 2H2O and NaH2PO4 · 2H2O in DI water. Master solutions of the two tetracyclines were made separately by mixing AM/BIS (4 g AM with 0.2 g BIS) and 0.8 g TC (0.8 g OTC for OTC molecular imprints) molecules in DI water and stirring for 10 min in a nitrogen atmosphere. Three different polymerization mediums were then made by mixing DI water, master solution and buffer in the different ratios shown in Table 2 and by adding 7.5 mg ammonium persulfate and 20 μL of N-tetramethylenthylenediamide.
Table 2. Water, master solution, and buffer mixing ratios for fabrication of tetracycline MIP imprints.
Diced b-Si chips were drop coated with the prepared solution (Fig. 2(a)) and kept in an oven at 50°C for 3 hours for polymerization. After polymerization, the substrates were dipped into an aqueous solution of 10 wt% sodium lauryl sulphate and 1 mL acetic acid for 2 hours at room temperature for removal of the target molecules from the matrix (Fig. 2(b)), then washed in DI water. For SERS detection of tetracyclines, 10 mM TC and OTC aqueous solutions were prepared separately. Both types of substrate were immersed in each solution for 1 hour at room temperature. Finally, the samples were rinsed with DI water and dried under nitrogen flow.
Fig. 2 Selectivity of a hybrid sensor: a) target analyte molecules are mixed with polymer matrix and drop cast on 200 nm Au coated b-Si substrates; b) target molecules are removed from the matrix leaving steric-voids ready for molecular recognition; c) the sensor is immersed into a mixture of different molecules and only target molecules can enter into close proximity with the gold surface; d) during SERS measurement the light is concentrated on tips and crevices of the substrate; e) tilted-45° angle SEM image of b-Si coated with 200 nm Au.
3. Results and discussion
Gold-coated b-Si is a highly sensitive SERS substrate [25], however, it is strongly dependent on the focusing conditions of excitation and collection of the back-scattered SERS signal. When the axial extent of the focal region is comparable with the height of the b-Si needles, then the strongest SERS signals are obtained and they increase as would be expected from a linear dependence of SERS on the solid angle [26]. In this study, different treatments were applied to enhance the performance and to demonstrate the versatility of b-Si substrates. First, the problem of substrate aging was addressed by modifying the surface hydrophobicity/philicity to increase the analyte concentration in a SERS-active Au nanoparticle suspension. Then, the selectivity problem in SERS was tackled by means of a MIP coating.
3.1. Analyte concentration
The initial hydrophilic substrate of b-Si was changed into a super-hydrophobic one by surface silanization. Then, the potential to concentrate analyte into a single spot was tested by drying the drop-cast solution of interest onto a bare (i.e. not Au-coated) b-Si. The SERS signal was collected from the light scattered by gold nanoparticles which were combined with analyte just before measurement. Such an in situ mixing of Au nanoparticles and analyte favours practical applications as the Au-surface aging issue [3] is avoided.
A favorable analyte concentration is achieved as shown in Fig. 3. In the case of the hydrophilic surface (Fig. 3(a)), nanoparticles and analyte are distributed over a large area, thus reducing significantly the number of molecules which scatter light and are inside the small laser focal spot. On the other hand, the hydrophobic surface makes the droplet dry in a single small spot, which brings all of its contents into a tiny area defined by the Cassie-Baxter wetting. This serves to concentrate the solution.
Fig. 3 Concentration of analyte on a non-aging hydrophobic SERS substrate: a) gold nanoparticles with analyte are distributed over a large surface area with a “coffee stain” rim formed at the outside edge, leading to non-uniform analyte deposition on the sample; b) analyte is concentrated on the tips via a drop-drying process. The superhydrophobicity of the surface allows for a small contact area between the aqueous droplet and the nanotextured surface.
First, a b-Si substrate without any surface treatment was measured as a reference for SERS detection. As-fabricated b-Si with 500–600 nm spikes (Samco recipe) showed hydrophilic surface properties. Once the analyte with nanoparticles comes into contact with the surface, the drop spreads all over the chip and distributes the analyte over a large area. Raman spectra from 850 cm−1 to 1150 cm−1 were measured using 785 nm laser excitation, as presented in Fig. 4 (black line). Three characteristic thiophenol peaks at 998 cm−1, 1022 cm−1 and 1073 cm−1 are clearly distinguished in the spectra. However, the signal was one order of magnitude higher when the PFTS treated surface was used as a substrate (Fig. 4). This demonstrates that the analyte can be effectively concentrated into single spot using superhydrophobic b-Si.
Fig. 4 Solution of the SERS substrate aging problem: SERS spectra of thiophenol mixed with gold nanoparticles on as-fabricated and PFTS-treated b-Si substrates. Thiophenol signature peaks at 998 cm−1, 1022 cm−1 and 1073 cm−1 were an order of magnitude more intense on the hydrophobic surface (following the preparation shown in Fig. 3(b).
Since the Au-colloidal particles are mixed with analyte solution just before SERS measurement, the surface aging of Au on a SERS substrate is avoided, provided that a common solvent of the Au colloidal solution and the analyte is utilized. This solves a long standing issue of surface aging and sensitivity drift in Au-coated SERS substrates.
As discussed above, concentration of the analyte by drop-drying helps to increase the SERS signal and allows trace detection of the analyte molecules. This method increases sensitivity when there is only one kind of molecule in the solution, or when spectral features do not overlap in the case of multi-analyte mixtures. If different molecules in solution have similar Raman spectral signatures, then the drop-drying approach would fail in the task of SERS fingerprinting, since Au nanoparticles would amplify Raman signals from different species inside the mixture. It is critically important to have some level of selectivity when dealing with such mixtures.
3.2. Selectivity of SERS with MIP
To solve the problem of selectivity without complex and specific surface functionalisation of Au nanoparticles, it is demonstrated here that MIP films can be used to differentiate between closely matching derivatives of tetracycline. SERS spectra from 400 cm−1 to 2000 cm−1 were measured on b-Si using the same SERS excitation conditions as described above.
The “empty” steric voids of the target molecules were prepared inside the gel matrix on the surface of Au-coated b-Si. Then, the 3D matching voids were “filled” by the target tetracycline molecules: TC or OTC. The TC recognition by the TC imprinted MIP spectra is shown in Fig. 5(a). In this case the molecules and the imprinted voids match each other and allow some of the TC molecules to approach the nano-rough Au regions with the “hot spots”. The gel matrix is permeable to the analyte molecules. The measured “empty” TC MIP imprint spectrum (the reference) was subtracted from the filled TC imprint data to improve peak recognition. The resulting spectrum is plotted in Fig. 5(b). A reference spectrum of TC powder was also measured and is shown for comparison (Fig. 5(b) blue line). TC peaks between 1200 and 1400 cm−1 indicate an ability to detect and identify matching molecules.
Fig. 5 SERS spectra of tetracycline hydrochloride (TC) imprint with matching and mismatching molecules: (a,b) TC molecules in TC imprint, (c,d) oxytetracycline hydrochloride (OTC) molecules in TC imprint. The excitation laser was operating at 785 nm and the signal was collected using a NA = 0.5 microscope objective lens.
A cross check of MIP selectivity was carried out by immersing the TC MIP into the OTC solution with a non-matching analyte. In this case OTC molecules should not be detected as they can not enter the polymer gel and reach the vicinity of the gold layer. Figure 5(c,d) shows recorded and subtracted spectra, respectively. The spectrum does not show OTC peaks, due to the mismatch between imprint and molecules, as one would expect.
The same procedure of OTC recognition by the OTC MIP matrix was repeated and cross checked using non matching TC molecules. The results were similar: OTC molecules were detected with the OTC imprint and a very low TC signal was observed with the same imprint.
The influence of the MIP layer thickness on the SERS signal was investigated by preparing three different imprint matrices. Figure 6 shows the signal from the match-filled MIP for different matrix recipes (Table 2). The different recipes led to different viscosities of the matrices and this resulted in different imprint thicknesses on the b-Si surface after drop-casting. All three samples were immersed in the TC solution before SERS measurement. One hour at room temperature was found to provide enough time for the target molecules to enter the imprint layer (Fig. 2(c)) and reach the surface of the gold (Fig. 2(e)), where plasmonic light concentration occurs at hot-spots. Only steric voids within 1–10 nm from the Au surface contribute to the signal during SERS measurement (Fig. 2(d)). It was observed that the target molecules entered MIP voids of the matching analyte more readily in the thinner gel layers. The highest TC fingerprint signal was observed for the third recipe, which resulted in 20-μm-thick coating on the b-Si surface and was three times thinner than the other MIP layers.
Fig. 6 SERS spectra of match-filled TC imprint using various imprint preparation recipes 1–3. Different MIPs have variation in permeability and cross-linking. Spectra are offset-shifted for clarity.
4. Conclusion
Black-Si produced by reactive ion etching is a promising candidate for SERS substrates. Here, hybrid sensors combining b-Si substrates and surface treatments were investigated. Surface hydro-phobicity/philicity was tuned by chemical means, thereby enabling precise control of analyte distribution on the surface. The control of b-Si hydrophobicity and analyte preparation in Au-colloidal suspension demonstrated efficient concentration of analyte molecules and solved the issue of Au surface aging. A ten-fold increase in SERS signal was measured from the treated substrate as compared to the initial as-fabricated hydrophilic b-Si.
A molecularly imprinted polymer layer on top of gold coated b-Si demonstrated selective detection of two tetracycline derivatives. Recognition of the matching steric analyte molecules by the MIP on random b-Si shows the potential for SERS substrate selectivity. Further concentration of the solution using the volume-phase transition in gels [27] will be tested in future experiments using thermally- or pH-responsive MIPs. Mechanical rather than chemical control of selectivity using MIPs and a combination of SERS and MIPs on randomly nano-textured substrates have been demonstrated.
Acknowledgments
We are grateful to Pierrette Michaux for assisting in the development of b-Si recipes. This work was supported by Australian Research Council Discovery Projects DP130101205, DP120102980, DP110100262, and an ARC Linkage Project with Raith-Asia LP120100161. SJ acknowledges the Australian Academy of Science senior fellowship support for a visit of Prof. Banshi Gupta laboratory at the Indian Institute of Technology, Delhi. FC thanks the Victorian Government for funding to support the UROP Small Technology program. This work was performed in part at the Melbourne Centre for Nanofabrication (MCN) in the Victorian Node of the Australian National Fabrication Facility (ANFF).
References and links
1. B. Zhang, H. Wang, L. Lu, K. Ai, G. Zhang, and X. Cheng, “Large-area silver-coated silicon nanowire arrays for molecular sensing using surface-enhanced Raman spectroscopy,” Adv. Func. Mat. 18, 2348–2355 (2008). [CrossRef]
2. S. K. Srivastava, A. Shalabney, I. Khalaila, C. R. Grüner, B. Rauschenbach, and I. Abdulhalim, “SERS biosensor using metallic nano-sculptured thin films for the detection of endocrine disrupting compound biomarker vitellogenin,” Small 10, 3579–3587 (2014). [CrossRef] [PubMed]
3. A. Chou, E. Jaatinen, R. Buividas, G. Seniutinas, S. Juodkazis, E. L. Izake, and P. M. Fredericks, “SERS substrate for detection of explosives,” Nanoscale 4, 7419–7424 (2012). [CrossRef] [PubMed]
4. W. Premasiri, D. Moir, M. Klempner, N. Krieger, G. Jones, and L. Ziegler, “Characterization of the surface enhanced Raman scattering (SERS) of bacteria,” J. Phys. Chem. B 109, 312–320 (2005). [CrossRef]
5. J. D. Driskell, K. M. Kwarta, R. J. Lipert, M. D. Porter, J. D. Neill, and J. F. Ridpath, “Low-level detection of viral pathogens by a surface-enhanced Raman scattering based immunoassay,” Analyt. Chem. 77, 6147–6154 (2005). [CrossRef]
6. S. Shanmukh, L. Jones, J. Driskell, Y. Zhao, R. Dluhy, and R. A. Tripp, “Rapid and sensitive detection of respiratory virus molecular signatures using a silver nanorod array SERS substrate,” NanoLett. 6, 2630–2636 (2006). [CrossRef]
7. G. Duan, W. Cai, Y. Luo, Y. Li, and Y. Lei, “Hierarchical surface rough ordered au particle arrays and their surface enhanced Raman scattering,” Appl. Phys. Lett. 89, 181918 (2006). [CrossRef]
8. R. Buividas, N. Fahim, J. Juodkazytė, and S. Juodkazis, “Novel method to determine the actual surface area of a laser-nanotextured sensor,” Appl. Phys. A 14, 169–175 (2013).
9. R. Buividas, N. Dzingelevičius, R. Kubiliūtė, P. R. Stoddart, V. Khanh Truong, E. P. Ivanova, and S. Juodkazis, “Statistically quantified measurement of an Alzheimer’s marker by surface-enhanced Raman scattering,” J. Biophot. (published online) (2014). [CrossRef]
10. Y. Nishijima, Y. Hashimoto, L. Rosa, J. B. Khurgin, and S. Juodkazis, “Scaling rules of SERS intensity,” Adv. Opt. Mat. 2, 382–388 (2014). [CrossRef]
11. Y. Nishijima, J. B. Khurgin, L. Rosa, H. Fujiwara, and S. Juodkazis, “Randomization of gold nano-brick arrays: a tool for SERS enhancement,” Opt. Express 21, 13502–13514 (2013). [CrossRef] [PubMed]
12. Y. Nishijima, Y. Hashimoto, G. Seniutinas, L. Rosa, and S. Juodkazis, “Engineering gold alloys for plasmonics,” Appl. Phys. A 117, 641–645 (2014). [CrossRef]
13. F. De Angelis, F. Gentile, F. Mecarini, G. Das, M. Moretti, P. Candeloro, M. Coluccio, G. Cojoc, A. Accardo, C. Liberale, et al., “Breaking the diffusion limit with super-hydrophobic delivery of molecules to plasmonic nanofocusing SERS structures,” Nat. Photonics 5, 682–687 (2011). [CrossRef]
14. W. Song, D. Psaltis, and K. B. Crozier, “Superhydrophobic bull’s-eye for surface-enhanced Raman scattering,” Lab on Chip 14, 3907–3911 (2014). [CrossRef]
15. A. C. De Luca, P. Reader-Harris, M. Mazilu, S. Mariggiò, D. Corda, and A. Di Falco, “Reproducible surface-enhanced Raman quantification of biomarkers in multicomponent mixtures,” ACS Nano 8, 2575–2583 (2014). [CrossRef] [PubMed]
16. W. Song, Y. Chen, J. Xu, X.-R. Yang, and D.-B. Tian, “Dopamine sensor based on molecularly imprinted electrosynthesized polymers,” J. Sol. State Electrochem. 14, 1909–1914 (2010). [CrossRef]
17. D. Cai, L. Ren, H. Zhao, C. Xu, L. Zhang, Y. Yu, H. Wang, Y. Lan, M. F. Roberts, J. H. Chuang, et al., “A molecular-imprint nanosensor for ultrasensitive detection of proteins,” Nature Nanotechn. 5, 597–601 (2010). [CrossRef]
18. N. Karimian, A. P. Turner, and A. Tiwari, “Electrochemical evaluation of troponin T imprinted polymer receptor,” Biosens. Bioelectron. 59, 160–165 (2014). [CrossRef] [PubMed]
19. R. Verma and B. D. Gupta, “Optical fiber sensor for the detection of tetracycline using surface plasmon resonance and molecular imprinting,” Analyst 138, 7254–7263 (2013). [CrossRef] [PubMed]
20. E. L. Holthoff, D. N. Stratis-Cullum, and M. E. Hankus, “A nanosensor for TNT detection based on molecularly imprinted polymers and surface enhanced Raman scattering,” Sensors 11, 2700–2714 (2011). [CrossRef] [PubMed]
21. S. Chen, X. Li, Y. Zhao, L. Chang, and J. Qi, “High performance surface-enhanced Raman scattering via dummy molecular imprinting onto silver microspheres,” Chem. Commun. 50, 14331–14333 (2014). [CrossRef]
22. L. Chang, Y. Ding, and X. Li, “Surface molecular imprinting onto silver microspheres for surface enhanced Raman scattering applications,” Biosensors and Bioelectronics 50, 106–110 (2013). [CrossRef] [PubMed]
23. Y. Nishijima, L. Rosa, and S. Juodkazis, “Surface plasmon resonances in periodic and random patterns of gold nano-disks for broadband light harvesting,” Opt. Express 20, 11466–11477 (2012). [CrossRef] [PubMed]
24. Y. Nishijima, L. Rosa, and S. Juodkazis, “Long-range interaction of localized surface plasmons in periodic and random patterns of Au nanoparticles,” Appl. Phys. A 2, 409–414 (2014). [CrossRef]
25. G. Gervinskas, G. Seniutinas, J. S. Hartley, S. Kandasamy, P. R. Stoddart, N. F. Fahim, and S. Juodkazis, “Surface-enhanced Raman scattering sensing on black silicon,” Ann. Physik 525, 907–914 (2013). [CrossRef]
26. S. Jayawardhana, L. Rosa, R. Buividas, P. R. Stoddart, and S. Juodkazis, “Light enhancement in surface-enhanced Raman scattering at oblique incidence,” Photon. Sensors 2, 283–288 (2012). [CrossRef]
27. S. Juodkazis, N. Mukai, R. Wakaki, A. Yamaguchi, and H. Misawa, “Reversible phase transitions in polymer gels induced by radiation forces,” Nature 408, 178–181 (2000). [CrossRef] [PubMed]
