Ordered Ag/Si Nanowires Array: Wide-Range Surface-Enhanced Raman Spectroscopy for Reproducible Biomolecule Detection Jian-An Huang, 1 Ying-Qi Zhao, 1 Xue-Jin Zhang, 3 Li-Fang He, 1 Tai-Lun Wong, 1 Ying-San Chui, 1 Wen-Jun Zhang, 1 * Shuit-Tong Lee 2 * 1 Center of Super-Diamond and Advanced Films (COSDAF) and Department of Physics and Materials Science, City University of Hong Kong, Hong Kong SAR, China. 2 Institute of Functional Nano & Soft Materials (FUNSOM) and Collaborative Innovation Center of Suzhou Nano Science and Technology, Soochow University, Suzhou 215123, China. 3 National Laboratory of Solid State Microstructures, Nanjing University, Nanjing 210093, China Page 1 of 19
Supporting Information S1. Materials Polystyrene nanospheres of 300 nm and 470 nm in diameter were brought from Bangs Laboratories, Inc. and Microparticle GmbH Berlin, respectively. The silicon wafer was Boron-doped p-type <111> with a resistivity of 5 20 Ω cm -2. Rhodamine 6G (R6G, Sigma-Aldrich 252433), 4-Aminothiophenol (4-ABT, Sigma-Aldrich 422967), and 1,2-Di(4-pyridyl)ethylene (BPE, Sigma-Aldrich B52808) were brought from Sigma-Aldrich Co., LLC. Deoxyribonucleic acid (DNA) sodium salt with molecular weight of 50000 100000 Dalton was brought from ACROS Organics (CAS: 68938-01-2). S2. Fabrication of Ag/SiNWs array, AgNPs/SiNWs array and AgFON SiNWs arrays were prepared via Nano-Sphere Lithography and Metal-assisted chemical etching. 1 Briefly, 300-nm-diameter polystyrene nanospheres were diluted 1:1 (volume) in ethanol and injected slowly onto water surface to form a monolayer. The clean silicon wafer placed under the water in advance was lifted above the wafer surface slowly to land the polystyrene monolayer on it; afterwards the wafer was allowed to air dry. The as-made substrate was put into a Reactive Ion Etcher (Plasma-Therm 790 RIE) to etch the polystyrene nanospheres into 110-nm diameter by oxygen plasma of 40-sccm flow-rate, 37 mtorr pressure and 30 W power for 5.5 minutes. Then the substrate was coated by 1-nm-thick of titanium and 20-nm-thick layer of gold by an Electron Beam Evaporator and put into a Rapid Thermal Processer at 250 C for 1 hour to melt the polystyrene spheres. The SiNWs array was fabricated by immersing the substrate into an etchant Page 2 of 19
composed of 20 ml water, 2 ml 48% HF, and 0.2 ml 35% H 2 O 2 for 3 minutes. For Ag/SiNWs array, a ~20-nm-thick silver layer was coated on the as-prepared SiNWs array by Magnetron Sputtering Deposition in Argon plasma (20 sccm, 80 W and 1 minute). To construct AgNPs/SiNWs array, the SiNWs array was annealed at 900 C and 1 Bar in air for 1 hour by rapid thermal annealing before Ag sputtering deposition. Due to the slight oxidation of silicon surface during annealing, the sputtered Ag forms separated particles instead of a continuous layer on Si surface. 2, 3 Silver film on nanospheres (AgFON) was made by Nano-Sphere Lithography and Ag sputtering deposition according to Van Duyne group. 4 The 470-nm-diameter polystyrene nanospheres were diluted 1:1 (volume) in ethanol and injected slowly onto water surface to form a monolayer. The clean silicon wafer placed under the water in advance was lifted above the wafer surface slowly to land the polystyrene monolayer on it; afterwards the wafer was allowed to air dry. Then, ~200-nm-thick silver layer was coated on the as-prepared nanospheres array by Magnetron Sputtering Deposition in Argon plasma (20 sccm, 80 W) to form AgFON. Page 3 of 19
Figure S1. (a) SEM image and (b) TEM image of Ag/SiNWs array. Scale bars are 200 nm. Page 4 of 19
Page 5 of 19
Figure S2. (a) TEM image of Ag/SiNW and the HRTEM images of (b) SiNW/Ag interface, and (c) Ag/Air interface as indicated in (a). Figure S3. SEM image of AgFON made by 200 nm Ag film deposited on nanospheres of 470 nm in diameter. Figure S4. SEM image of AgNPs/SiNWs Page 6 of 19
S3. Finite-Difference-Time-Domain (FDTD) simulation We used Lumerical FDTD Solution (version 7.5.6) to perform FDTD simulation. As shown in Figure S2, the models were smooth or rough 20 nm-think Ag layer coated on a SiNW of 110-nm diameter and 700-nm length inside vacuum. The optical data of Ag and silicon are from Palik. 5 The plane wave source of 450-800 nm with polarization direction along x axis was incident from the head. Meshing enclosing the entire wire was set to 2 nm per grid. The z-axis boundaries of the simulation region were set to Perfect Matching Layer, while those of x- and y-axis set to Periodic. A z-normal 2D monitor was placed ~400 nm above the wire head for reflectivity spectra. Near-field cross-sectional electrical intensity profiles was obtained by a y-normal 2D monitor passing the wire axis. Page 7 of 19
Figure S5. FDTD model of smooth (left) and rough (right) 20 nm-thick Ag layer coated on a SiNW of 110 nm in diameter and 700 nm in length. Page 8 of 19
Figure S6. Comparison of reflectivity spectrum from FDTD simulation on smooth 20 nm Ag film coated on SiNW of 110 nm in diameter and 700 nm in length with the experimental reflectivity of the Ag/SiNW with the same dimensions. Curves are shifted vertically for clarity. In the simulated curve, the minima at 504 and 541 nm correspond to diameter-dependent leaky mode of SiNWs, and the one at 647nm was SPR valley. Page 9 of 19
Figure S7. FDTD-simulated images of near-field electrical intensity distribution of a smooth model 20nm-Ag/SiNW of 110-nm diameter and 700-nm length. The images at 504 nm and 514 nm show there are diameter-dependent leaky modes Page 10 of 19
inside the SiNW. 6-8 The wavelengths of incident light inside the SiNW are smaller than the wire diameter so that much resonant energy is mainly confined inside the SiNW and thus not sensitive to the Ag layer. This explains why the wire-dependent guiding modes at 487 nm and 560 nm in the experimental reflectivity curves changed little with Ag thicknesses. However, as the wavelength of incident light increases to 541 nm, the energy starts to leak out to the surface and affect the SERS performance. At 633 and 647 nm, most resonant energy is confined on the surface and resonantly coupled to the propagating surface plasmons of the Ag layer to induce SPR with enhanced electrical fields. Figure S8. (a) FDTD simulated electric field intensity of Ag/SiNW arrays at 633 nm; (b) Electric field intensity at position a in (a) δ is the decay length of the propagating surface plasmon which is defined as the length that the field intensity falls to the 1/e times of its maximum. 9 Page 11 of 19
Figure S9. (a) FDTD-simulated reflectance of Ag/SiNWs with different lengths. The minima are their SPRs. FDTD-simulated electric field intensity images at corresponding SPRs of length at (b) 500 nm (c) 700 nm, and (d) 900 nm, respectively. Page 12 of 19
S4. Wavelength-dependent SERS of Ag/SiNWs array We tuned the 514 laser and 633 one to the same power of 4.15 mw under a 20x objective and collected the Raman spectra of 4-ABT monolayer on Ag/SiNWs, 4-ABT powders on glass slides and silicon wafer under their excitations below. Page 13 of 19
Figure S10. Raman spectra under excitation by (a) 633 nm and (b)514 nm laser of 4.15 mw power. Both figures show that the 4-ABT is not resonant with either 514 or 633 lasers. Figure S11. 4-ABT SERS spectra normalized to the Raman band of Si wafer (520 cm -1 ) under excitation of 633 nm and 514 nm lasers of 4.15 mw power, respectively. As shown in Figure S7, the energy of Ag/SiNWs at 514 nm was mostly confined in the SiNWs and eventually absorbed by the SiNWs to emit stronger Raman peak at 518 cm -1. Besides, energy of Ag/SiNWs at 633 nm was confined on the Ag/SiNWs surface in the form of propagating surface plasmons that excited the absorbed 4-ABT monolayer to give much stronger SERS signals. Page 14 of 19
S5. Deposition of analyte molecules and Raman measurements Raman measurement and mapping were conducted on a RENISHAW InVia Raman spectrometer with an Argon laser of 514.5 nm with tunable power up to 50 mw and a 17 mw HeNe laser of 632.8 nm. The laser power could also be adjusted by the spectrometer software, WiRE 3.2, to obtain the optimum SERS spectra. A 20x objective delivering a laser spot of 3.0 µm diameter, 80 µm effective height of sampling volume and full power of 4.4 mw was used to map a surface area of 200 200 μm 2 at 3 μm step and 1 second collection time to obtain a total of 4624 Raman spectra. For mono-analyte detection of 4-ABT, the substrate was immersed into 10-4 M ethanol 10, 11 solution of 4-ABT for 8 hours to allow the molecules to form a monolayer on the Ag surface. Then the substrate was air dried, rinsed with ethanol to remove excessive molecules and blown to dry by nitrogen for Raman measurements. For bi-analyte detection, 4-ABT monolayer was firstly formed on the substrate by the above method. Then, 1 µm droplet of 10-4 M ethanol solution of R6G was dropped onto the substrate, spread out as a circle of ~5 mm in diameter and finally vaporized for Raman measurements. For tri-analyte detection, 4-ABT monolayer was firstly formed on the substrate by the above method. Then, 1 µm droplet of the mixed ethanol solution of BPE (10-4 M) and R6G (10-5 M) was dropped onto the substrate, spread out as a circle of ~5 mm in diameter and finally vaporized for Raman measurements. For DNA detection, salmon double-strand DNA (dsdna) was used as received and diluted in sterilized water to 10-5 M solution. The salmon dsdna had a molecular weight of 50000 Page 15 of 19
100000 Daltons. Since the average molecular weight of a nucleotide pair in dsdna is 660 Daltons and each nucleotide pair is 0.33 nm long on average, the salmon dsdna had the length of 25 50 nm. For the mapping of a surface area of 200 200 μm 2, 1 µm droplet of water-diluted 10-5 M DNA solution was dropped onto the substrate, spread out as a circle of ~1.5 mm in diameter and finally vaporized before Raman mapping. Figure S12. (a) SERS spectra of 4-ABT, R6G and bi-analytes containing these three moleculesre, respectively from Ag/SiNWs array. (b) 4624 SERS spectra of bi-analyte of 4-ABT and R6G adsorbed on Ag/SiNWs array. Page 16 of 19
Figure S13. (a) 4624 SERS spectra of 4-ABT monolayer adsorbed on AgNPs/SiNWs array. (b) Five SERS spectra extracted from (a). The 4624 SERS spectra from AgFON after deposition of (c) tri-analytes, and (d) salmon dsdna. S6. Calculation of Enhancement Factor To estimate the Enhancement Factor (EF) of the SERS substrate, 4-ABT solid films of known volume and weight was placed on a clean glass slide for the measurement of normal Raman spectrum. The SERS peaks of 1076 cm -1 and 1141 cm -1 correspond to the 4-ABT Raman peaks of 1093 cm -1 and 1172 cm -1, respectively. 12 EF of 4-ABT on Ag/SiNWs array was Page 17 of 19
calculated for the SERS peaks at 1076 cm -1 and 1141 cm -1 using the following formula specific for the solid SERS substrate of periodic nanostructure: 13 EF = ISERS /( µ M µ S AM ) I /( C H ) RS RS eff where I RS is the Raman intensity of the analyte molecules under non-sers conditions and I SERS the intensity under SERS conditions. C RS [µm -3 ] is the volume density of the molecules used for the non-sers measurement, H eff [µm] is the effective height of the scattering volume of the objective, µ M [µm -2 ] is the surface density of the individual Ag/SiNWs producing the enhancement, and µ S [µm -2 ] is the surface density of molecules on the metal. A M [µm 2 ] is the surface area of one Ag/SiNW wire. In our Raman measurement setup, µ M was 7 µm -2, µ S was 3.3 10 6 µm -2, 10, 11 and H eff was measured to be 80 µm for the 20x objective. 14 The C RS of 97% 4-ABT film was calculated to be 4.94 10 9 µm -3 with the density of neat 4-ABT being 1.06 10-12 g/µm -3. 15 I RS (1093 cm -1 ) = 5250 counts and I RS (1172 cm -1 ) = 3071 counts. For 4-ABT monolayer deposited on Ag/SiNWs of 150 nm in diameter and 700 nm in length, I SERS (1076 cm -1 ) = 61856 counts, I SERS (1141 cm -1 ) = 81780 counts and A M = 0.35 µm 2. So the EFs are: EF (1076 cm -1 ) = 5.76 10 5 EF (1141 cm -1 ) = 1.10 10 6 Reference of Supporting Information 1. Huang, J.A. et al. Enhanced Raman scattering from vertical silicon nanowires array. Applied Physics Letters 98 (2011). 2. Pillai, S., Catchpole, K.R., Trupke, T. & Green, M.A. Surface plasmon enhanced silicon solar cells. Journal of Applied Physics 101, 8 (2007). 3. Krishna, H. et al. Thickness-dependent spontaneous dewetting morphology of Page 18 of 19
ultrathin Ag films. Nanotechnology 21, 7 (2010). 4. Kleinman, S.L., Frontiera, R.R., Henry, A.I., Dieringer, J.A. & Van Duyne, R.P. Creating, characterizing, and controlling chemistry with SERS hot spots. Physical Chemistry Chemical Physics 15, 21-36 (2013). 5. Palik, E.D. HANDBOOK OF OPTICAL-CONSTANTS. Journal of the Optical Society of America a-optics Image Science and Vision 1 (1984). 6. Lopez, F.J. et al. Diameter and Polarization-Dependent Raman Scattering Intensities of Semiconductor Nanowires. Nano Letters 12, 2266-2271 (2012). 7. Cao, L.Y. et al. Engineering light absorption in semiconductor nanowire devices. Nature Materials 8, 643-647 (2009). 8. Wang, B.M. & Leu, P.W. Tunable and selective resonant absorption in vertical nanowires. Optics Letters 37, 3756-3758 (2012). 9. Maier, S.A. Plasmonics : fundamentals and applications (Springer, New York :, 2007). 10. Wang, H., Levin, C.S. & Halas, N.J. Nanosphere arrays with controlled sub-10-nm gaps as surface-enhanced Raman spectroscopy substrates. Journal of the American Chemical Society 127, 14992-14993 (2005). 11. Park, W.-H. & Kim, Z.H. Charge Transfer Enhancement in the SERS of a Single Molecule. Nano Letters 10, 4040-4048 (2010). 12. Wang, H., Kundu, J. & Halas, N.J. Plasmonic nanoshell arrays combine surface-enhanced vibrational spectroscopies on a single substrate. Angewandte Chemie-International Edition 46, 9040-9044 (2007). 13. Le Ru, E.C., Blackie, E., Meyer, M. & Etchegoin, P.G. Surface enhanced Raman scattering enhancement factors: a comprehensive study. Journal of Physical Chemistry C 111, 13794-13803 (2007). 14. Gartia, M.R. et al. Rigorous surface enhanced Raman spectral characterization of large-area high-uniformity silver-coated tapered silica nanopillar arrays. Nanotechnology 21 (2010). 15. Jackson, J.B. & Halas, N.J. Surface-enhanced Raman scattering on tunable plasmonic nanoparticle substrates. Proceedings of the National Academy of Sciences of the United States of America 101, 17930-17935 (2004). Page 19 of 19