Tunable Silver Nanocap Superlattice Arrays for Surface-Enhanced

Nov 7, 2011 - Surface-enhanced Raman scattering (SERS) has drawn wide- ... fabricate PAA-templated SERS-active substrates using small pro- trusions ...
0 downloads 0 Views 4MB Size
ARTICLE pubs.acs.org/JPCC

Tunable Silver Nanocap Superlattice Arrays for Surface-Enhanced Raman Scattering Xianzhong Lang,†,‡ Teng Qiu,*,†,‡ Wenjun Zhang,*,‡,§ Yin Yin,† and Paul K. Chu*,‡ †

Department of Physics, Southeast University, Nanjing 211189, P. R. China Department of Physics and Materials Science, City University of Hong Kong, Kowloon, Hong Kong, P. R. China § Department of Microelectronics, Fudan University, 220 Handan Road, Shanghai 200433, P. R. China ‡

bS Supporting Information ABSTRACT: We report on a convenient nanotechnique to fabricate large-area silver nanocap superlattice arrays templated by the base of porous anodic alumina membranes as robust and cost-efficient surface-enhanced Raman scattering substrate. The topography can be tuned to optimize the enhancement factor by adjusting anode voltages or the time of silver magnetron sputtering. Our technique is especially promising considering their easy fabrication and evenly distributed plasmonic fields to cm-dimensions featuring high average enhancement factor, thereby boding well for application in the sensing device.

’ INTRODUCTION Surface-enhanced Raman scattering (SERS) has drawn widespread attention as a powerful spectroscopic technique capable of nondestructive and ultrasensitive characterization down to single molecular level.1,2 The SERS effect has an electromagnetic (EM) origin that arises from an increase in the local optical field exciting the molecule and multiplicative amplification of the reradiated Raman scattered light. The fact that particle plasmon allows direct coupling of light to resonant electron plasmon oscillation has spurred tremendous efforts in the design and fabrication of highly SERS-active nanostructured films and metallic nanoparticles.3 5 However, “hot spots” on traditional SERSactive substrates are found at unpredictable positions, and the SERS intensity changes dramatically from one “hot spot” to another even on the same sample due to the irregular geometry. Therefore, the success of an SERS-active substrate in practice depends on the production of subwavelength nanostructures with evenly distributed plasmonic fields featuring high average enhancement factor (EF).6 9 One well-known fabrication method is lithography such as electron beam lithography,10 nanosphere lithography,11 nanoimprint lithography,12 or lithographically defined templates to produce tightly spaced plasmonic geometries.8 However, the fabrication process tends to be labor intensive and costly, and it is sometimes impossible to extend to large dimensions. Hence, the use of porous anodic alumina (PAA) membranes as the template to fabricate SERS-active substrates is promising considering the easy fabrication, excellent reproducibility, modest cost, and large area production.7,13 17 Moreover, the technique has the following advantages: (1) the ability to optimize periodically plasmonic geometries and tune the “hot junction” in the sub-10 nm regime, (2) creation of long-range r 2011 American Chemical Society

uniform plasmonic structures to the cm dimensions, and (3) production of 3D SERS-active substrates using PAA membranes with inner walls. Our group recently reported a convenient nanotechnique to fabricate PAA-templated SERS-active substrates using small protrusions along the surface of the PAA pore wall.16 In this paper, we describe a new approach to fabricate tunable silver nanocap superlattice arrays (SNSAs) using PAA membranes as robust and cost-efficient SERS substrates exhibiting significant SERS enhancement. The topography can further be tuned to optimize the EF by adjusting the anode voltages (VPAA) and silver deposition time (tAg).

’ EXPERIMENTAL SECTION Synthesis of the Base of PAA Substrates. High purity aluminum (Al) foils (99.99%, 20 mm  20 mm  0.3 mm) were degreased by acetone and electro-polished using a mixture of ethanol and perchloric acid with a volume ratio of 5:1 under a constant DC voltage of 15 V for 3 min to further remove surface impurities (Figure 1a). After rinsing in distilled water and drying, the Al foils were anodized separately in a 0.5 M oxalic acid solution at a constant DC voltage of 20 V (30, 40, 50, and 60 V) at 10 °C. In order to obtain an ordered nanopore array, a two-step anodizing process was adopted. The Al foils were first anodized for 2 h followed by immersion into a mixture of chromic acid (1.8 wt %) and Received: September 27, 2011 Revised: November 6, 2011 Published: November 07, 2011 24328

dx.doi.org/10.1021/jp2093302 | J. Phys. Chem. C 2011, 115, 24328–24333

The Journal of Physical Chemistry C

ARTICLE

Figure 1. Schematic of the fabrication method of SNSAs: (a) Pretreated Al sheets, (b) PAA templates, (c) reinforced and inversed PAA templates, (d) the base of PAA, (e) SNSAs, and (f) alumina cell.

phosphoric acid (6 wt %) at 75 °C (1:1 in volume). After 2 h, the alumina layer, which grew during the first step, was removed and the surface of the foil became bright. The anodizing time in the second step was 2 h and the PAA templates were obtained (Figure 1b). The PAA templates were reinforced by copper tape as a protective layer (Figure 1c). Finally, the aluminum base was dissolved in a saturated solution of copper sulfate by wetchemical etching to obtain the PAA substrates with alumina protrusion arrays (Figure 1d). Synthesis of SNSAs. SNSAs were prepared on the PAA templates at room temperature under argon (20 sccm) in a conventional direct-current magnetron sputtering system. The power of the magnetron was about 40 W (voltage of 410 V and current of 100 mA). During deposition, the pressure in the magnetron chamber was 3.8  10 1 Pa and the distance between the target and substrate was 20 cm. After deposition, the samples were protected from contamination and oxidation in a vacuum desiccator until they were subjected to further experiments (Figure 1e). Instrumentation and Data Acquisition. SEM (JEOL JSM6335F) and AFM (Vecco Nano Scope V) were used to investigate the nanostructures. AFM probes are composed by pyramidal tip with 14 16 μm height and 6 nm radius of curvature connecting with rectangular silicon cantilever with 30 ( 5 nm aluminum reflex side coating. It should be noted that the height measurements of the nanocap configurations remain fairly accurate because the tip may not touch bottom between two adjacent nanocaps. The Raman measurements were performed on a Jobin Yvon LabRAM HR800 micro-Raman spectrometer with the 514 nm laser line at room temperature. The area ∼3 μm in diameter was probed by a 50 objective lens (nominal aperture 0.45) and the incident power at the sample was 0.04 mW. The signal collection time was 50 s. In order to evaluate the substrate Raman-enhancing capability, a Rhodamine 6G (R6G) 10 5 M water solution was used. In order to allow molecule adsorption, the substrate was maintained for 30 min in the R6G solution and then taken out and rinsed thoroughly. The acquisition time and laser power were the same for all Raman spectra. The SERS spectra were recorded from multiple sites on the substrate surface to confirm reproducibility. Similar SERS spectral characteristics such as enhancement, position, and relative intensity of the bands were determined from various locations to confirm large area production of uniform geometries.

Figure 2. (a) A 3D AFM image, (b) magnified 2D AFM image acquired from the base of a typical PAA substrate (VPAA = 50 V) with the scale bar being 200 nm, (c) analysis along the blue line in (b) and corresponding structure diagram (inset).

’ RESULTS AND DISCUSSION In this technique, silver is directly deposited on the PAA substrate using direct-current magnetron sputtering (see Figure 1). The template consists of periodic arrays of close-packed protuberances as shown in Figure 2a. The protuberances are the results of a self-ordering process under a high local hemisphere field and mechanical stress.18 20 The high-resolution 2D atomic force microscopy (AFM) image of the same template is displayed in Figure 2b. The size of each protuberance and interparticle gap are controlled to around 100 nm and less than 10 nm respectively, which are necessary for big SERS enhancement.21,22 To describe this feature more clearly, Figure 2b presents the corresponding result in Figure 2c. The top-to-bottom fluctuations of the cell boundary between two protrusions forming a natural V-shape dent are obvious. The cross-sectional geometry 24329

dx.doi.org/10.1021/jp2093302 |J. Phys. Chem. C 2011, 115, 24328–24333

The Journal of Physical Chemistry C of the PAA substrate is shown schematically in Figure 2c. These protrusions have similar shape and size. These protuberances can be regarded as part of the overlapping spheres where the duple curvature radius (2R) is longer than the center distance (d) between two adjacent spheres. Using this procedure, different protrusion diameters (D) which can be controlled by adjusting VPAA can be produced. Almost all the protrusions on the templates have periodical hexagonal arrangements and the templates are easily scaled up to contiguous areas cm2 in size (see Figures S1 and S2). Thus, this technique opens new possibilities for future design and engineering of SERS substrates. In particular, this method allows one to study the origin of the enhancement, optimize the amplification factors, and replicate structures that provide consistent and optimized response. We first obtain a series of SNSAs with different sizes by adjusting VPAA and subsequently the deposition processing (tAg = 15 min) as illustrated in Figure 3a. The silver film thickness is calculated to be 60 nm based on the deposition parameters. The shape resembles a cap based on the ratio of the height to radius of the alumina protrusions. The size of each silver cap is consistent with the template consisting of the deposited alumina protusions. That is, the value of D is almost unchanged after silver deposition (Figure S3). Figure 3b depicts the typical scanning electron microscopy (SEM) images of SNSAs and the functional relationship between the values of D and VPAA. Each silver nanocap is surrounded by six equivalent adjacent ones forming the nanostructure unit and the average value of D is a function of VPAA with a slope of 2 nm/V in the voltage region of 30 60 V. These silver nanocaps allow direct coupling between the electric field arising from the incoming radiation and resonant electron plasmon oscillations, and two coupled adjacent silver nanocaps with close voids are verified to support extremely intense local EM fields, known as “hot spots” upon optical excitation.23 Our technique can tailor the density of these “hot spots” from ∼1  1010 to ∼6  1010 cm 2 and produce a very large number of such “hot spots” over an area of a few cm2 (inset in Figure 3b). This would be extremely difficult to obtain using existing nanofabrication methods. To evaluate the Ramanenhancing capability of the SNSAs, an aqueous solution (10 5 M) of R6G is used on the substrate. Figure 3c shows the typical Raman spectrum of the R6G solution obtained on the SNSAs (VPAA = 40 V) which is compared to that acquired from the flat silver sheet. Many salient peaks can be observed, and the more pronounced ones at 1510, 1537, 1574, and 1648 can be assigned to the totally symmetric modes of in-plane C C stretching vibration. However, the normal Raman signature of 10 2 M of R6G on the silver coated aluminum sheet is only barely recognizable. The empirical EF is estimated to be larger than 105 by comparing ratios of the average SERS peak intensity (at 1510 cm 1) of R6G to the corresponding average unenhanced signals. We also evaluate the performance of other SNSAs formed under different VPAA and the intensity of the Raman signal reaches a maximum at VPAA = 40 V. Figure 3d shows the normalized average Raman peak intensity (at 1510 cm 1) versus VPAA. We found in the experiments that all the SNSAs have large UV vis absorption from 330 to 500 nm representing the localized surface plasmon energy due to various gap sizes between neighboring silver nanocaps on the samples. Thus, the variation in the SERS signal intensity can at first glance be associated with the effect of resonance coupling between neighboring nanocaps. In order to assess the relative contributions

ARTICLE

Figure 3. (a) Schematic of SNSAs tailored by VPAA. (b) Average D size as a function of VPAA (tAg = 15 min) and corresponding SEM images of SNSAs. The scale bar is 200 nm. The inset in (b) shows large-area SNSA with size of about 1 cm2. (c) SERS spectral comparison of R6G adsorbed on the SNSA and the flat silver sheet. The inset in (c) shows a typical simulated EM-field distribution map of the SNSA (VPAA = 40 V, tAg = 15 min). (d) Normalized average SERS signal at 1510 cm 1 as a function of VPAA (tAg = 15 min). (e) SERS map (40.0  40.0 μm2) obtained from the SNSA with VPAA = 40 V and tAg = 15 min. The map is obtained from the integrated intensity of the 1510 ( 5 cm 1 band of 10 5 M R6G adsorbed on the sample.

Table 1. Calculated Data of the Maximum Local Electric Field for SNSAs under Different VPAA Figure 4a

Figure 4b

Figure 4c

Figure 4d

Figure 4e

voltage (V)

20

30

40

50

60

|Emax|(V/m)

104.0

137.0

432.3

223.1

167.8

of different geometries to the experimentally observed SERS intensity, the local EM fields are calculated using the finite-difference 24330

dx.doi.org/10.1021/jp2093302 |J. Phys. Chem. C 2011, 115, 24328–24333

The Journal of Physical Chemistry C

ARTICLE

Figure 5. (a) Schematic of SNSAs tailored by tAg. (b) Average R size as a function of tAg (VPAA = 40 V) and corresponding SEM images of SNSAs. The scale bar is 200 nm. (c) Normalized average SERS signal at 1510 cm 1 as a function of tAg (VPAA = 40 V). (d) Average SERS signal at 1510 cm 1 as a function of the R6G molecular concentration on a logarithmic scale (VPAA = 40 V, tAg = 15 min).

Figure 4. Simulated EM-field distribution maps of the SNSAs (tAg = 15 min): VPAA = 20 (a), 30 (b), 40 (c), 50 (d), and 60 V (e). (f) Calculated densities of Ag nanocaps and “hot spots” as a function of VPAA (tAg = 15 min). (g) Normalized average SERS signal at 1510 cm 1 divided by the density of “hot spots” and the maximum local electric field as a function of VPAA (tAg = 15 min), demonstrating good correlation between experiment and theory. The calculated data of the maximum local electric field for SNSAs under different VPAA are shown in Table 1.

time-domain (FDTD) method. A typical planar and crosssectional view of the calculated radial EM field components of

the SNSA (VPAA = 40 V) is displayed in the inset of Figure 3c. The SNSA is approximated by the fundamental nanostructure unit composed of seven hexagonally arranged hemispheres using dimensional parameters (D, d, and R illustrated in Figure 3c) equal to the mean values of the samples produced experimentally. The radiation at 514 nm is assumed to be normal to the sample surface. Clearly, the incident radiation excites a plasmon trapped between the nanocaps to create a huge EM field locally. Our calculation shows that these localized resonant plasmon modes are able to produce maximum local enhancement as large as 1010 in the Raman signal of the molecules adsorbed at these V-shaped gaps (see Table 1). This exceeds the average over the entire SNSA surface area by several orders of magnitude because most of the optical excitations are localized in these “hot spots”.24 The maximum localized EF at the junction of the silver nanocaps appears in the SNSA at 40 V and the local EM enhancement is as high as 3.5  1010, demonstrating good correlation between experiments and theory (see Figure 4). The homogeneity of the SNSAs is evaluated by 2D point-bypoint SERS mapping of R6G molecules. Figure 3e shows a typical SERS map of the SNSA (VPAA = 40 V). Excellent uniformity over a large area of 40.0  40.0 μm2 is achieved by measuring 1600 points with a regular scanning step of 1 μm. The relative SERS peak intensity of the collection spots is centered in a narrow range, and the spot-to-spot relative standard deviation is calculated to be 15%. This indicates that the substrate homogeneity is 24331

dx.doi.org/10.1021/jp2093302 |J. Phys. Chem. C 2011, 115, 24328–24333

The Journal of Physical Chemistry C

ARTICLE

Table 2. Calculated Data of the Maximum Local Electric Field for SNSAs under Different tAg Figure 6a time (min) |Emax|(V/m)

Figure 6b

Figure 6c

Figure 6d

Figure 6e

5

10

15

20

25

88.0

196.6

432.3

339.6

132.6

The maximum local electric field of the SNSAs calculated also appears from the sample for tAg equal to 15 min (see Table 2). The result shows good correlation with the experimental data (see Figure 6). During routine, online trace analysis, the SERS substrates should have a large SERS dynamic range. Since the SNSAs are demonstrated to have all the characteristics that render them excellent SERS substrates, we also study the SERS dynamic range of the sample with optimized EF (VPAA = 40 V and tAg = 15 min) using R6G solutions with different concentrations. Figure 5d shows the intensity of the SERS signal at 1510 cm 1 as a function of the molar concentration on a logarithmic scale. The linear correlation from 10 9 to 10 6 M with a proportionality constant of unity suggests that the number of adsorption sites with high Raman enhancement is large enough to accommodate a considerable range of sample concentrations. A nonlinear dependence emerges for concentrations above 10 6 M, indicating that adsorption of R6G onto sites with high enhancement becomes saturated beyond this level. In conjunction with the large EF obtained across the entire substrate, the large dynamic range facilitates the use of SERS in chemical/biological sensing applications with high sensitivity.

’ CONCLUSION In summary, a simple technique to fabricate tunable SNSAs using the base of PAA membranes as robust and cost-efficient SERS substrate with high signal area-averaged EF is described. This technique allows scaling up to cm2 areas enabling preparation of large-area SERS sensors. Further increase in EFs can be achieved by tuning the surface structure according to FDTD calculation. ’ ASSOCIATED CONTENT

bS

Supporting Information. A series of SEM images acquired from the base of PAA membranes under different VPAA and schematic of SNSAs tailored by tAg. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author Figure 6. Simulated EM-field distribution maps of the SNSAs (VPAA = 40 V): (a) tAg = 5 (a), 10 (b), 15 (c), 20 (d), and 25 (e). (f) Normalized average SERS signal at 1510 cm 1 and the maximum local electric field as a function of tAg (VPAA = 40 V), demonstrating good correlation between experiment and theory. Calculated data of the maximum local electric field for SNSAs under different tAg are shown in Table 2.

quite good and strict control of the preparation conditions can ensure good reproducibility among different batches. Apart from the observation that the SNSAs can be tuned by varying VPAA to optimize the SERS, we also increase tAg from 5 to 25 min to tailor the geometries of SNSAs as illustrated in Figure 5a. The changes in these SNSAs have two main characteristics. First of all, both the value of D and density of “hot spots” are almost fixed. Second, the value of R goes up by increasing tAg (Figure 5b). The Raman-enhancing capability of these SNSAs is evaluated. The variation in the SERS signal intensity at 1510 cm 1 is presented in Figure 5c. The intensity of the R6G Raman signal reaches a maximum when tAg is 15 min.

*E-mail: [email protected] (T.Q.); [email protected] (W.Z.); [email protected] (P.K.C.).

’ ACKNOWLEDGMENT This work was jointly supported by the National Natural Science Foundation of China under Grant Nos. 50801013 and 51071045, Excellent Young Teachers Program of Southeast University, and Hong Kong Research Grants Council (RGC) General Research Funds (GRF) CityU 112510. The FDTD simulations are done at Southeast University. ’ REFERENCES (1) Nie, S. M.; Emery, S. R. Probing single molecules and single nanoparticles by surface-enhanced Raman scattering. Science 1997, 275, 1102–1106. (2) Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R.; Feld, M. S. Single molecule detection using surface-enhanced Raman scattering (SERS). Phys. Rev. Lett. 1997, 78, 1667–1670. 24332

dx.doi.org/10.1021/jp2093302 |J. Phys. Chem. C 2011, 115, 24328–24333

The Journal of Physical Chemistry C (3) Banholzer, M. J.; Millstone, J. E.; Qin, L. D.; Mirkin, C. A. Rationally designed nanostructures for surface-enhanced Raman spectroscopy. Chem. Soc. Rev. 2008, 37, 885–897. (4) Li, J. F.; Huang, Y. F.; Ding, Y.; Yang, Z. L.; Li, S. B.; Zhou, X. S.; Fan, F. R.; Zhang, W.; Zhou, Z. Y.; Wu, D. Y.; Ren, B.; Wang, Z. L.; Tian, Z. Q. Shell-isolated nanoparticle-enhanced Raman spectroscopy. Nature 2010, 464, 392–395. (5) Rycenga, M.; Cobley, C. M.; Zeng, J.; Li, W. Y.; Moran, C. H.; Zhang, Q.; Qin, D.; Xia, Y. N. Controlling the Synthesis and Assembly of Silver Nanostructures for Plasmonic Applications. Chem. Rev. 2011, 111, 3669–3712. (6) Qiu, T.; Wu, X. L.; Shen, J. C.; Chu, P. K. Silver nanocrystal superlattice coating for molecular sensing by surface-enhanced Raman spectroscopy. Appl. Phys. Lett. 2006, 89, 131914. (7) Choi, D.; Choi, Y.; Hong, S.; Kang, T.; Lee, L. P. Self-Organized Hexagonal-Nanopore SERS Array. Small 2010, 6, 1741–1744. (8) Liberman, V.; Yilmaz, C.; Bloomstein, T. M.; Somu, S.; Echegoyen, Y.; Busnaina, A.; Cann, S. G.; Krohn, K. E.; Marchant, M. F.; Rothschild, M.; Nanoparticle, A Convective Directed Assembly Process for the Fabrication of Periodic Surface Enhanced Raman Spectroscopy Substrates. Adv. Mater. 2010, 22, 4298–4302. (9) Caldwell, J. D.; Glembocki, O.; Bezares, F. J.; Bassim, N. D.; Rendell, R. W.; Feygelson, M.; Ukaegbu, M.; Kasica, R.; Shirey, L.; Hosten, C. Plasmonic Nanopillar Arrays for Large-Area, High-Enhancement Surface-Enhanced Raman Scattering Sensors. ACS Nano 2011, 5, 4046–4055. (10) Yu, Q. M.; Guan, P.; Qin, D.; Golden, G.; Wallace, P. M. Inverted size-dependence of surface-enhanced Raman scattering on gold nanohole and nanodisk arrays. Nano Lett. 2008, 8, 1923–1928. (11) Hicks, E. M.; Zhang, X.; Zou, S.; Lyandres, O.; Spears, K. G.; Schatz, G. C.; Van Duyne, R. P. Plasmonic properties of film over nanowell surfaces fabricated by nanosphere lithography. J. Phys. Chem. B 2005, 109, 22351–22358. (12) Alvarez-Puebla, R.; Cui, B.; Bravo-Vasquez, J. P.; Veres, T.; Fenniri, H. Nanoimprinted SERS-active substrates with tunable surface plasmon resonances. J. Phys. Chem. C 2007, 111, 6720–6723. (13) Lee, S. J.; Morrill, A. R.; Moskovits, M. Hot spots in silver nanowire bundles for surface-enhanced Raman spectroscopy. J. Am. Chem. Soc. 2006, 128, 2200–2201. (14) Wang, H. H.; Liu, C. Y.; Wu, S. B.; Liu, N. W.; Peng, C. Y.; Chan, T. H.; Hsu, C. F.; Wang, J. K.; Wang, Y. L. Highly Ramanenhancing substrates based on silver nanoparticle arrays with tunable sub-10 nm gaps. Adv. Mater. 2006, 18, 491–495. (15) Ko, H.; Tsukruk, V. V. Nanoparticle-decorated nanocanals for surface-enhanced Raman scattering. Small 2008, 4, 1980–1984. (16) Qiu, T.; Zhang, W.; Lang, X.; Zhou, Y.; Cui, T.; Chu, P. K. Controlled assembly of highly Raman-enhancing silver nanocap arrays templated by porous anodic alumina membranes. Small 2009, 5, 2333–2337. (17) Huang, Z. L.; Meng, G. W.; Huang, Q.; Yang, Y. J.; Zhu, C. H.; Tang, C. L. Improved SERS Performance from Au Nanopillar Arrays by Abridging the Pillar Tip Spacing by Ag Sputtering. Adv. Mater. 2010, 22, 4136–4139. (18) Thompson, G. E. Porous anodic alumina: Fabrication, characterization and applications. Thin Solid Films 1997, 297, 192–201. (19) Li, A. P.; Muller, F.; Birner, A.; Nielsch, K.; Gosele, U. Hexagonal pore arrays with a 50 420 nm interpore distance formed by selforganization in anodic alumina. J. Appl. Phys. 1998, 84, 6023–6026. (20) Ono, S.; Saito, M.; Asoh, H. Self-ordering of anodic porous alumina formed in organic acid electrolytes. Electrochim. Acta 2005, 51, 827–833. (21) GarciaVidal, F. J.; Pendry, J. B. Collective theory for surface enhanced Raman scattering. Phys. Rev. Lett. 1996, 77, 1163–1166. (22) Jiang, J.; Bosnick, K.; Maillard, M.; Brus, L. Single molecule Raman spectroscopy at the junctions of large Ag nanocrystals. J. Phys. Chem. B 2003, 107, 9964–9972. (23) Hildebrandt, P.; Stockburger, M. Surface-Enhanced Resonance Raman-Spectroscopy of Rhodamine-6g Adsorbed on Colloidal Silver. J. Phys. Chem. 1984, 88, 5935–5944.

ARTICLE

(24) Kim, W.; Safonov, V. P.; Shalaev, V. M.; Armstrong, R. L. Fractals in microcavities: Giant coupled, multiplicative enhancement of optical responses. Phys. Rev. Lett. 1999, 82, 4811–4814.

24333

dx.doi.org/10.1021/jp2093302 |J. Phys. Chem. C 2011, 115, 24328–24333