Highly sensitive deep-silver-nanowell arrays - Springer Link

0 downloads 0 Views 4MB Size Report
Oct 23, 2016 - approximately 400–800 nm). ... from Beijing Chemical Works and used as received. ... performed using a Plasmalab Oxford 80 Plus system.
Nano Research 2017, 10(3): 908–921 DOI 10.1007/s12274-016-1348-7

Highly sensitive deep-silver-nanowell arrays (d-AgNWAs) for refractometric sensing Xueyao Liu, Wendong Liu, Liping Fang, Shunsheng Ye, Huaizhong Shen, and Bai Yang () State Key Laboratory of Supramolecular Structure and Materials, College of Chemistry, Jilin University, Changchun 130012, China

Received: 4 August 2016

ABSTRACT

Revised: 23 October 2016

Large-area deep-silver-nanowell arrays (d-AgNWAs) for plasmonic sensing were manufactured by combining colloidal lithography with metal deposition. In contrast to most previous studies, we shed light on the outstanding sensitivity afforded by deep metallic nanowells (up to 400 nm in depth). Using gold nanohole arrays as a mask, a silicon substrate was etched into deep silicon nanowells, which acted as a template for subsequent Ag deposition, resulting in the formation of d-AgNWAs. Various geometric parameters were separately tailored to study the changes in the optical performance and further optimize the sensing ability of the structure. After several rounds of selection, the best sensing d-AgNWA, which had a Ag thickness of 400 nm, template depth of 400 nm, hole diameter of 504 nm, and period of 1 μm, was selected. It had a sensitivity of 933 nm·RIU–1, which is substantially higher than those of most common thin metallic nanohole arrays. As a proof of concept, the as-prepared structure was employed as a substrate for an antigen-antibody recognition immunoassay, which indicates its great potential for label-free real-time biosensing.

Accepted: 28 October 2016 © Tsinghua University Press and Springer-Verlag Berlin Heidelberg 2016

KEYWORDS deep-silver-nanowell arrays, colloidal lithography, nanohole, plasmonic, nanostructure, refractometric sensing

1

Introduction

Surface-plasmon resonance (SPR) is the oscillation of free electrons at the metal-dielectric interface. By concentrating and channeling light via surface plasmons (SPs) using the subwavelength metallic structure, the light can be confined at the interface, which boosts the electric-field magnification [1]. By exploiting the enhanced electric field, which can be easily electrically modulated, the Raman signal can be dramatically altered, resulting in interesting electric–plasmonic Address correspondence to [email protected]

switching [2]. Moreover, the enhanced electric field is sensitive to its adjacent dielectric environment’s refractive index, resulting in the refractometric sensitivity of the metallic structure [3]. The heart of plasmonics involves tailoring the geometry or the assembly form of structures, thereby allowing the manipulation of the properties of SPs [4–7]. Shape- and size-distinguished structures show varied polar sensitivities. Thus far, plasmonic sensors with various structures, including nanoparticles [8–10], nanoholes [11, 12], nanopillars [13], and nanodisks [14], and structures derived from

909

Nano Res. 2017, 10(3): 908–921

these [15, 16], have been investigated for plasmonic sensing. Among them, periodic nanohole arrays exhibit intense resonance at a specific wavelength because of their extraordinary optical transmission [17], which has received great attention since being discovered. Owing to their tunable and scalable spectral properties, which offer simplicity [18], metallic nanohole arrays are widely exploited as plasmonic sensors [19, 20–31]. Most reported metallic nanohole arrays were fabricated using electron-beam lithography and focused ion-beam lithography [32]. Although these fabrication methods provide structures with high accuracy and a precise processing control ability, they are highly costly and are unsuitable for large-area nanostructure fabrication, which impedes their industrial application. Soft interference lithography can realize the fabrication of large-area, high-quality nanostructures [33]. However, the fabrication process is complex, limiting its practical operability. Soft nanoimprint lithography, with the aid of a polydimethylsiloxane (PDMS) soft mold [34], enables large-area nanostructure construction [11, 35]; however, the PDMS mold is so soft that the reproducibility and accuracy of smaller nanostructures are affected by the operation difficulty. Colloidal lithography cleverly employs self-assembled colloidal crystals with various colloidal-crystal-assisted nanofabrication techniques and has proven to be facile, low-cost, and efficient for fabricating large-area ordered nanohole arrays [36]. The effects of the fabrication material [37] and structural parameters (i.e., shape [19], size [38], depth [39], and periodicity [40]) on the optical performance of metallic nanohole arrays have been studied theoretically and experimentally. To our knowledge, most current plasmonic sensors based on metallic nanohole arrays have low metal thickness. Thus, their sensing performance is greatly affected by the substrate, resulting in lower sensitivity than that of optically thicker metallic nanohole arrays. For extremely thick perforated metal film, the SPs of the two interfaces are uncoupled, and the transmission decreases rapidly with increasing depth (from a thickness of approximately 400–800 nm). The peak then broadens and shifts for thinner metallic nanohole arrays [39]. Therefore, thickness of 400 nm is proposed to be ideal when the transmission reaches the maximum and the

linewidth remains narrow. The group of Nuzzo and Rogers reported a biosensor by employing a highly sensitive quasi-three-dimensional plasmonic crystal [11]. They fabricated a 350-nm-deep polymer nanowell, followed by 50-nm gold deposition, resulting in a physically isolated gold disk at the bottom of the gold nanohole. The achieved sensitivity ranged from 700 to 800 nm·RIU–1. Notably, the metal was thin. Subsequently, they improved the sensing performance by changing the gold coverage form, in which the gold layer was continuous [31]. In such a case, the metallic hole of the structure became deeper than that reported earlier, leading to improved sensing performance. Compared with normal metallic hole arrays with a low depth, its sensing performance is remarkably high. However, the metallic section is surrounded by the polymer, which affects the sensing performance of the metal film. Therefore, we suppose that a thickness of ~400 nm is the optimal value for metallic hole arrays to exhibit excellent sensing behavior owing to their high transmission and narrow linewidth. This inspired us to focus on pure metallic nanohole arrays with a relatively high thickness. Herein, we describe a facile approach by combining colloidal lithography with metal deposition to fabricate large-area deep-silver-nanowell arrays (d-AgNWAs). Using large-area ordered gold nanohole arrays as etching masks, a silicon substrate was etched into deep silicon-nanowell arrays, which served as a template for the subsequent Ag deposition. Owing to the depth of the template, the obtained silver nanowells were extremely deep. The structure was comprehensively modulated by regulating the silicon-nanowell template and the deposition conditions. Various structural parameters, including hole diameter, template depth, metallic-film thickness, and structure period, were thoroughly examined to optimize the sensing performance. A d-AgNWA with a Ag thickness of 400 nm, template depth of 400 nm, hole diameter of 504 nm, and period of 1 μm afforded the desired sensing performance, with sensitivity of up to 933 nm·RIU–1. This is substantially higher than those of most common thin metallic nanohole arrays. The fabrication approach is facile, low-cost, reproducible, and efficient for fabricating large-area ordered metallic nanowell arrays. Finally, the optimized d-AgNWAs were employed

www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano

Research

910

Nano Res. 2017, 10(3): 908–921

as a platform for application in immunoassay to demonstrate their potential in label-free biosensing.

2 2.1

Experimental Materials

Silicon-chip (2 cm × 2 cm) substrates were immersed in a mixture solution of H2SO4 and H2O2 with a volume ratio of 7:3 and boiled for over 2 h to make the substrate surface hydrophilic. Then, the chips were repeatedly rinsed with deionized (DI) water (ultrapure, 18.2 MΩ) from a Millipore water purification system and absolute ethanol alternately three times and finally stored in DI water. We purchased 1-μm polystyrene (PS) spheres from Sigma–Aldrich, Inc. and 620-nm PS spheres from Wuhan Tech Co., Ltd. We synthesized 470-nm PS spheres by emulsion polymerization as described in a reference [41]. The silver (99.9%) powder and gold (99.99%) powder for vapor deposition were purchased from Sinopharm Chemical Reagent Co. Ltd. All solvents were purchased from Beijing Chemical Works and used as received. Proteins were purchased from DingGuo Biotech. Co. Ltd. (Beijing, China) and dissolved in phosphatebuffered saline (PBS, pH 7.4) solutions when the immunoassay was performed. Mercaptohexadecanoic acid (MHA) and PBS were purchased from Sigma– Aldrich. 2.2

Fabrication of d-AgNWAs

Briefly, the silver-nanowell arrays were fabricated on a silicon substrate by the combination of colloidal lithography and thermal vapor deposition. First, an orderly packed monolayer of PS spheres was achieved by the self-assembly of PS at the interface between water and air. A PS dispersion with a concentration of 2.5% in a mixture of DI water and absolute ethanol (v/v = 1:1) was dropped onto the surface of water within a glass culture dish. One drop of a 3% sodium dodecyl sulfate solution was added when the PS dispersion had almost filled the whole surface. Then, the assembled PS spheres were transferred to the silicon substrate by lifting the entire surface on a silicon chip and gradually dried in the atmosphere. Second, a gold nanohole array mask was created by

colloidal lithography combined with metal deposition. The orderly packed PS sphere monolayer was etched by oxygen reactive-ion etching (RIE) at a pressure of 10 mTorr, a flow rate of 50 sccm, an ratio frequency (RF) power of 40 W, and an inductively coupled plasma (ICP) power of 120 W. The etching procedure was performed using a Plasmalab Oxford 80 Plus system (ICP 65) (Oxford Instrument Co., UK). Afterward, a 3-nm chromium adhesion layer and a 40-nm gold layer were deposited into the void space between the spheres of the samples. The gold-coated PS spheres were removed by toluene, leaving the gold nanohole array. Subsequently, the samples were rinsed by ethanol and dried by a nitrogen flow. Third, silicon nanowells were prepared by continuously etching the silicon substrate, using the gold nanohole arrays as a mask. The procedures for the mixture of CHF3 and SF6 gases with flow rates of 30 and 4 sccm, respectively, an RF power of 20 W, and an ICP power of 100 W were applied to the samples for different periods. The gold mask was removed by immersing the sample in aqua regia, and the sample was then rinsed with DI water and dried by nitrogen. Eventually, silver with a specific thickness was deposited on the silicon nanowell to construct the silver nanowell on the Si substrate. 2.3 Finite-difference time-domain (FDTD) simulations The electromagnetic fields were simulated using a commercial software package (FDTD Solutions v8.6.3, Lumerical Solutions Inc.). The simulation employed identical structural parameters to the experimental samples. The unit cell was set as rectangular, with one hole in the center and four quartering holes at the four corners. Periodic boundary conditions were set along both the x and y directions, and perfectly matched layers were employed in the ±z direction under normal incidence. A unit-magnitude plane wave propagated in the z direction with electric-field polarization along the x-axis. To increase the numerical accuracy, we set the mesh type as “auto non-uniform”. The mesh refinement was “conformal variant 2”. The frequency-domain field profile monitor and frequencydomain field & power monitor were used to simulate the distribution of the SP energy and the reflectance spectra. The electromagnetic-field enhancement was

| www.editorialmanager.com/nare/default.asp

911

Nano Res. 2017, 10(3): 908–921

thus evaluated. The material parameters of silicon and silver were taken from the Palik Handbook. 2.4 Characterization 2.4.1

Reflectance spectroscopy

All the reflectance spectra were measured with a Shimadzu 3600 ultraviolet–visible–near-infrared spectrophotometer. For the solvent test, the spectra of the d-AgNWAs in the atmosphere were recorded in advance. Then, a liquid cell was filled with a specific solvent, and the corresponding spectrum was logged once the cell was stable. Afterward, both the liquid cell and the nanowell chip were rinsed with ethanol and water alternately and purged by nitrogen to completely remove the solvent. Before the testing of the other solvents, the measurement of the initial spectrum of the nanowell was repeated to ensure its accuracy. The solvents used in the experiment were methanol; ethanol; isopropyl alcohol; N,N-dimethylformamide; and toluene. For the immunoassay, the original d-AgNWA chip was immersed in 4 mM 16-MHA overnight. Subsequently, protein immobilization was performed by the physical adsorption of 50 μg/mL human immunoglobulin G (IgG) for 20 min, followed by blocking with bovine serum albumin (BSA) of 200 μg/mL for 30 min. The control experiment of the recognition step was performed with 25 μg/mL goat anti-rabbit IgG for 30 min. Sequentially, the recognition of anti-human IgG was performed with concentrations of 1, 2, 5, 8, and 10 μg/mL over 20 min. 2.4.2

in Scheme 1. The morphology characterization after each step was performed by SEM, as shown in Fig. 1. The PS colloidal microspheres (e.g., 620 nm in diameter) in the monolayer were hexagonally close-packed on the silicon substrate (Fig. 1(a)). After etching with O2 plasma for 280 s, the diameter of the microsphere was reduced from 620 to 425 nm, showing the orderly rendered structure and the period (Fig. 1(b)). To obtain a hole mask, 3 nm of chromium and 40 nm of gold were vertically deposited onto the substrate successively, and the gold-coated PS microspheres were then removed (Fig. 1(c)). Using the gold hole as an etching mask, the flat silicon substrate was etched into siliconnanowell arrays with a relatively high aspect ratio (Fig. 1(d)). As shown in the cross-sectional image, the silicon nanowell template was straight and deep, maintaining a highly ordered lattice structure. After the removal of the gold mask, 400 nm of Ag was vertically deposited onto the silicon-nanowell template to obtain d-AgNWAs (Fig. 1(e)). Centimeter-sized highly ordered d-AgNWAs were fabricated with few defects (Fig. S1 in the Electronic Supplementary Material (ESM)). According to the SEM images of the final d-AgNWAs, the nanowell diameter and height were approximately 350 and 400 nm, respectively. The geometric parameters of the d-AgNWAs—the individual hole diameter (D); the nanowell height (the thickness of the Ag layer above the Si–Ag interface) (H); the silicon-nanowell template depth (d); and the lattice

Scanning electron microscopy (SEM)

SEM images were obtained using a JEOL JSM 6700F field-emission scanning electron microscope, and a ~2-nm-thick layer of platinum was sputtered onto the sample to improve the conductivity.

3 3.1

Results and discussion d-AgNWA fabrication

Large-area highly ordered d-AgNWAs were fabricated by well-developed colloidal lithography [36], as shown

Scheme 1 Schematic of the fabrication process for d-AgNWAs. The center displays the cross-section of one d-AgNWA. The main geometric parameters are included.

www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano

Research

912

Nano Res. 2017, 10(3): 908–921

Figure 1 SEM images of (a) close-packed PS nanospheres with a diameter of 620 nm, (b) PS spheres etched by O2 plasma after 280 s, (c) the gold hole mask, (d) the silicon-nanowell template, and (e) the d-AgNWAs. The scale bars all represent 1 μm. The insets in (c)–(e) are the corresponding cross-section SEM images. The scale bars in the insets all represent 100 nm.

constant, i.e., the period of the structure (P) (as shown in the center of Scheme 1)—can be easily modulated by regulating the duration of the PS microsphere etching, the deposition thickness of the silver in the last step, and the PS size. 3.2 Optical and sensing performance optimization of d-AgNWAs The optical performance of the patterned metallic film structure was precisely manipulated by altering the geometric parameters of the structure. In this regard, the effects of D, d, H, and P on the final optical performance of the structure, the dip feature of the reflectance spectrum (its position, intensity, and linewidth), and the corresponding sensing performance of the structure were thoroughly studied. 3.2.1 Optimization of d-AgNWAs by changing hole diameter D d-AgNWAs with different hole diameters were fabricated by varying the etching period of the colloidal mask and keeping other parameters fixed. The other structural parameters P, d, and H were fixed at 620, 400, and 400 nm, respectively. As shown in Figs. 2(a)–2(c), the extended etching of PS resulted

in a final hole structure with a reduced size. The hole diameters of the three d-AgNWAs shown in Figs. 2(a)–2(c) were 165, 237, and 350 nm, respectively. FDTD modeling facilitated the visualization of the electric-field distribution in the plasmonic structures. Figures 2(d)–2(f) show the electric-field distribution at the dip positions of the reflectance spectra in Figs. 2(a)–2(c), respectively. The electric field was mainly distributed at the surface outside the silvernanowell cavity in the case of the d-AgNWA with a hole diameter of 165 nm. When the hole diameter is increased, the electric field penetrated more deeply into the nanowell cavity. The maximum electric-field intensity was simulated to occur with a hole diameter of 237 nm. Increasing the hole diameter to 350 nm dramatically decreases the electric-field intensity. The optical performance of d-AgNWAs with various hole diameters was investigated using their experimental reflectance spectra (Fig. 3(a)) and simulated with FDTD calculations (Fig. 3(b)). According to the experimental reflectance spectra of the d-AgNWAs, the dip at 654 nm for a d-AgNWA hole diameter of 237 nm exhibits the maximum intensity and the minimum linewidth. As shown in Fig. 3(a), when the hole diameter is less than 237 nm, the resonance dip is sharper but

| www.editorialmanager.com/nare/default.asp

913

Nano Res. 2017, 10(3): 908–921

Figure 2 SEM images of d-AgNWAs with diameters of (a) 165, (b) 237, and (c) 350 nm. All the scale bars represent 500 nm. (d)–(f) show simulated distributions of the normalized electric-field intensity at the peak wavelengths indicated by the corresponding dip positions in the calculated reflectance spectra of (a)–(c), respectively.

Figure 3 (a) Experimental reflectance spectra of d-AgNWAs with different hole diameters. (b) Simulated reflectance spectra of d-AgNWAs with different hole diameters. In both the experiment and the simulation, the conditions were fixed as P = 620 nm and d = H = 400 nm.

weaker. When the hole size increases to 350 nm, the intensity of the reflectance dip gradually decreases, and the linewidth broadens simultaneously. On the whole, the dip position undergoes a blue-shift as the hole diameter decreases. The FDTD simulation coincides perfectly with the experimental results. The simulation results indicate that when the hole diameter is 200 nm, the reflectance-dip intensity is maximized. When the hole diameter is less than 200 nm, the dip becomes sharper yet weaker. When

the hole diameter is larger than 200 nm, the dip becomes broader, and its intensity decreases significantly. The aforementioned general evolution trends of the experiment and calculations agree well. Additionally, there are spectral features that cannot be observed in the simulation results. The small peak on the shoulder of the dip corresponds to a diffraction phenomenon known as Wood’s anomaly [4]. The observed redshift of the SPR in the experiment compared with the simulation is explained by the Fano

www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano

Research

914

Nano Res. 2017, 10(3): 908–921

analysis, in which the interfering resonant and nonresonant contributions are considered. Notably, except for the small peak aroused by Wood’s anomaly, the shoulder of the main dip in the experimental results differs slightly from the simulation line shape. This is caused by the inevitable defects in the process of building the structure. The real surface is rougher than the simulated surface, and the edge cannot be as sharp as that of the model structure. These imperfections give rise to small mismatches between the experimental and simulated results. The sensing performance of the d-AgNWAs was examined according to its refractive-index sensitivity (RIS) and relative RIS (RISrelat). The RIS is defined as the change in the dip position per refractive-index unit (RIU), which is also the slope value in the linear fitting of the dip shift intensity with respect to the refractive index. The RISrelat was proposed by ShumakerParry et al. [42] to describe the discrepancy between the RIS of samples with different morphologies and feature peaks. The RISrelat is defined by the following equation RIS relat 

1



 ( eV )  100% n

Table 1 RIS, determination coefficient, and RISrelat of d-AgNWAs with different hole diameters (D) D (nm)

RIS (nm·RIU–1)

Adjust R2

RISrelat (%·RIU–1)

165±23

609.9

0.9872

51

193±10

617.1

0.9973

49

237±15

633.2

0.9985

50

268±19

635.4

0.9992

52

285±12

645.8

0.9974

48

295±23

601.5

0.9991

42

333±17

711.4

0.9950

52

350±18

721.3

0.9991

50

(721 nm·RIU–1) and the second-highest RISrelat (50%·RIU–1), which is close to that of the d-AgNWA with a hole diameter of 333 nm (52%·RIU–1). Moreover, the linear fit determinant coefficients were as high as 0.9991. Therefore, for the samples with P = 620 nm, d = 400 nm, and H = 400 nm, a hole diameter of 350 nm yielded the best sensing performance. According to the holediameter study, a larger hole provides higher sensing performance because of the penetration of the electric field into the nanowell cavity.

(1)

3.2.2 Optimization of d-AgNWAs by changing siliconnanowell template depth d

where ωr is the resonance energy in units of eV, and Δω/Δn represents the energy-shift RIS in units of eV·RIU –1. Linear fits of the dip shift with increasing refractive index for d-AgNWAs with different values of D are shown in Fig. S2 in the ESM. Table 1 provides an overview of the sensing performance of the as-prepared d-AgNWAs. Overall, the RIS tended to increase with the hole diameter’s increase. Interestingly, although the resonance energy was highest when the diameter was 237 nm, the relative sensing performance of the d-AgNWAs was not optimized at this diameter. We attribute this phenomenon to the penetration of the electric field into the cavity of the d-AgNWAs with an increasing hole size, as revealed by the FDTD simulation. However, when the hole became extremely large, beyond its reasonable range, not only did the sensing performance decrease, but the dip split into two simultaneous dips, which also appeared for other periods (Fig. S3 in the ESM). The d-AgNWA with a hole diameter of 350 nm exhibited the highest RIS

The effect of d on the optical performance of the d-AgNWAs was studied by fixing the D, H, and P of the d-AgNWAs at 350, 350, and 620 nm, respectively. Considering that a previously reported polymer nanowell having a depth of 350 nm with a 50-nm continuous gold coverage structure exhibited a satisfying sensing performance [31], we propose 350 nm to be the d-AgNWA height that yields the best sensing performance and thus preliminarily set H to be 350 nm for a trial. d-AgNWAs samples with various values of d were prepared by changing the etching duration in the second etching process, i.e., silicon etching using the gold nanohole arrays as a mask. The extended etching duration resulted in a deeper silicon-nanowell template. By changing the etching duration to 2, 4, and 6 min, d-AgNWAs with d of 200, 400, and 600 nm were achieved (Figs. 4(a)–4(c)). These three states represent the situations where the bulk silver within the siliconnanowell template was beyond, close to, and far below the top edge of the silicon-nanowell template, respectively. The experimental reflectance spectra shown in

r

| www.editorialmanager.com/nare/default.asp

915

Nano Res. 2017, 10(3): 908–921

Table 2 RIS, determination coefficient, and RISrelat of d-AgNWAs with different silicon-nanowell template depths (d) d (nm)

RIS (nm·RIU–1)

Adjust R2

RISrelat (%·RIU–1)

200

616.5

0.9989

47

400

681.9

0.9990

47

600(D1)

577.6

0.8226

46

600(D2)

697.7

0.9943

46

3.2.3 Optimization of d-AgNWAs by changing nanowell height H Figure 4 SEM cross-section images of d-AgNWAs with different silicon-nanowell template depths of (a) 200, (b) 400, and (c) 600 nm. The scale bars all represent 200 nm. (d) Corresponding experimental reflectance spectra for the three depths. In all cases, P = 620 nm, H = 350 nm, and D = 350 nm.

Fig. 4(d) indicate that the dip broadened but became stronger when the hole depth increased within a certain range. However, the sample with a 600-nm hole depth exhibited two dips and therefore a substantially broader dip. We suppose that a deeper hole caused more of the silicon-nanowell template surface to be exposed to air, affecting the SPR performance of the silvernanowell film. The observed broadening of the plasmon bands could have been caused by the averaging of the through-air coupling and through-silicon coupling, which agrees with results for a previously reported Ag film over a nanowell [43]. A relatively sharp and intense dip, which is considered to be the most promising candidate for sensing, was achieved at a depth of 400 nm. The sensing performance of the as-prepared samples with different values of d was further tested. Figure S4 in the ESM shows linear fits of the dip shift with an increasing refractive index of d-AgNWAs for different values of d. For a d-AgNWA with d of 600 nm, the RISs of both dips were tested, which are labeled as D1 (located at 729 nm) and D2 (located at 806 nm) in Table 2. Although the RIS of the d-AgNWA with d of 400 nm was lower than that of D2 of the d-AgNWA with d of 600 nm, the reflectance-dip shape and intensity for the d-AgNWA with d of 400 nm was far more acceptable. Therefore, the silicon-nanowell depth of 400 nm is considered to be the most adaptable depth in this system.

The role of H in the optical performance of dAgNWAs was explored. The structural parameters of D, d, and P were fixed at 350, 400, and 620 nm, respectively. During the last step of the vertical silver deposition, silver with thicknesses of 150, 250, 400, and 500 nm was deposited onto the silicon-nanowell template respectively. The morphology was precisely controlled when the thickness of the deposited silver was less than 400 nm, as shown in Figs. 5(a)–5(c). In those cases, the thickness of the bulk silver within the silicon-nanowell template remained identical to the silver-nanowell height above the dielectric interface. The continued deposition of silver to a thickness of 500 nm resulted in a gradually reduced silver-nanowell hole size, leading to an uncontrolled deposition process of silver into the silicon-nanowell hole (Fig. S5 in the ESM). The bulk silver within the silicon-hole template can be cone-like or pillar-shaped, which affects the

Figure 5 SEM cross-section images of d-AgNWAs with different nanowell heights of (a) 150, (b) 250, and (c) 400 nm. The scale bar represents 200 nm. (d) Corresponding experimental reflectance spectra for the aforementioned three heights and the height of 500 nm. In all cases, P = 620 nm, d = 400 nm, and D = 350 nm.

www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano

Research

916

Nano Res. 2017, 10(3): 908–921

SPR of the silver film. As shown in Fig. 5(d), within a certain range of thickness, the height increase of the nanowell can make the line shape of the reflectance spectra significantly sharper. In the case of the d-AgNWAs with H of 150 and 250 nm, a substantially broader dip and even two dips appear. When the silver-nanowell height was still low, the bulk silver within the silicon hole had not reached the top edge of the hole. The increasing height led to the incorporation of two dips into one relatively broad dip. The more exposed area of the silicon-nanowell template influenced the SPR of the silver layer more severely, resulting in the splitting of the resonance dip. The spectrum of the d-AgNWA with H of 400 nm exhibits one dip of great intensity. The dip position coincides with the dip of the d-AgNWA with H of 250 nm, which indicates the complete incorporation of the two dips into one stronger dip. The dip of the d-AgNWA with H of 500 nm was blue-shifted and became sharper compared with that of the d-AgNWA with H of 400 nm. This is attributed to the uncertain morphology resulting from the uncontrolled metal-deposition process. According to the spectral features, the d-AgNWAs with H of 400 and 500 nm are expected to be good candidates for further experiments. To confirm the best H condition for sensing applications, the sensing performance of d-AgNWAs with different H values was tested, as shown in Fig. S6 in the ESM. The sensitivity results are summarized in Table 3. Because the first dip of the 150-nm-H d-AgNWA gradually disappeared in the solvent test, only the sensing performance of the second dip (located at 821 nm) is listed herein. Apparently, the d-AgNWA with H of 400 nm had the highest RIS and RISrelat. Although the 500-nm-H d-AgNWA appeared to have a better spectrum shape, its sensing performance was far less satisfying than that of the 400-nm-H d-AgNWAs. The results indicate that in the ideal situation, the height of the bulk silver within the Si nanowell template equals the depth of the template for H of 400 nm. To determine whether d = H when the sensing performance is optimized for d-AgNWAs with other H values, the sensing performance of d-AgNWAs with other H values was tested (except for H over 400 nm with the uncontrolled structure

Table 3 RIS, determination coefficient, and RISrelat of d-AgNWAs with different silver film thicknesses (H) H (nm)

RIS (nm·RIU–1)

Adjust R2

RISrelat (%·RIU–1)

150

539.7

0.9953

48

250

693.7

0.9763

34

400

721.3

0.9991

50

500

664.0

0.9999

45

formation, as shown in Fig. S5 in the ESM). As for the d-AgNWAs with H of 230 and 330 nm, the RISs of the other d-AgNWAs continuously increased with the template depth. The highest RISrelat appeared when d = H (Figs. S7 and S8, and Tables S1 and S2 in the ESM), which proves that d = H is the condition in which the sensing performance of the d-AgNWAs is optimized for all values of H. The sensing performance achieved with H < 400 nm is far less satisfying than that achieved with d = H = 400 nm. Hence, for d-AgNWAs, H of 400 nm is confirmed to be the ideal condition for sensing applications, which further demonstrates that the best situation is when the height of the bulk silver within the Si nanowell template equals the template depth. 3.2.4 Optimization of d-AgNWAs by changing structure period P Regarding hexagonal arrays of nanoholes at normal incidence, the following equation gives the position of the main dip

min 

 Ag  d

P

(2)

 Ag   d 4 2 (i  ij  j 2 ) 3

where P is the lattice constant; εAg and εd represent the dielectric constants of silver and the surrounding dielectric medium, respectively; and i and j are the orders of diffraction. According to the above equation, the theoretical sensitivity RISt can be derived and expressed as follows [18] d RIS t  min  dnd

| www.editorialmanager.com/nare/default.asp

  Ag     nd2 4 2 2 (i  ij  j )  Ag 3 P

   

3

(3)

917

Nano Res. 2017, 10(3): 908–921

where nd (nd2 = εd) is the refractive index of the surrounding environment. Although this equation does not consider the appearance of the nanowell—which allows only an approximate prediction—it at least indicates that the lattice constant, i.e., the period of the metallic structure, can affect its sensitivity. This speculation was verified by our experiment: the increased structure period enhanced the sensitivity of the structure significantly. The lattice constant P was regulated by altering the diameter of the PS sphere that was primarily used. Three groups of d-AgNWAs with various values of P—470 nm, 620 nm, and 1 μm—were fabricated. The optimal fabrication condition was determined by changing the hole diameter D while maintaining d and H at 400 nm. The sensing performance and spectral features were both considered in selecting the best condition for d-AgNWAs fabrication. As a result, the optimized hole diameters D for P of 470 nm, 620 nm, and 1 μm were 125, 350, and 504 nm, respectively, as shown in Figs. 6(a)–6(c). These three SEM images show prepared d-AgNWAs with a promising structure having smooth edges, which endows them with a

favorable RIS for sensing applications. The experimental results show that regulating the period can not only change the dip position notably but also simultaneously enhance the sensitivity remarkably. As shown in Fig. 6(d), the highest sensitivity achieved for each period was 439.5, 721.3, and 933.3 nm·RIU–1. A larger period led to a steeper linear fitting curve and a higher sensitivity. Notably, the highest sensitivity was 1,107 nm·RIU–1 for the period of 1 μm (Fig. S9 in the ESM). The coefficient of determination of 0.99738 for the linear fitting curve in the solvent test is considerable. Nevertheless, the reflectance spectra of the d-AgNWA in the solvent test expressed an unfavorable stability. Additionally, its dip intensity was rather low compared with the final selected condition of the best sensing performance. Therefore, it is considered to be a potentially outstanding structure with a high sensitivity, as long as the line-shape stability is improved slightly. Even so, the sensitivity of 933 nm·RIU–1 is 2–6 times higher than those of most metallic nanohole arrays [44] (300–600 nm·RIU–1) and other LSPR plasmonic sensors, including nanocubes [45], nanostars [46], nanoprisms

Figure 6 SEM images of d-AgNWAs with different periods—(a) 470 nm, (b) 620 nm, and (c) 1 μm—and optimized sensitivity. The scale bars all represent 1 μm. (d) Dip shift with respect to the refractive index for the different periods of (a)–(c). www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano

Research

918

Nano Res. 2017, 10(3): 908–921

[47], and nanotriangles [48], which range from 150– 350 nm·RIU–1. 3.3

Refractometric sensing of d-AgNWAs

To examine the sensing potential of the d-AgNWAs, a sample with a relatively high RIS and stable performance with a period of 1 μm was applied in an immunoassay. Prior to the device evaluation, a solvent test was performed for the illustration of the RIS determination. 3.3.1

Solvent sensing of d-AgNWAs

The d-AgNWAs (P = 1 μm, D = 504 nm, and d = H = 400 nm) were immersed in a series of solvents with different refractive indices. Before the solvent replacement, the samples were softly rinsed with distilled water and ethanol alternately and purged by N2 to ensure reproducibility. To clarify the dip-position movement in the reflectance spectra, only the normalized main dip part were kept. The original complete spectra of the d-AgNWAs in the solvent test are presented in Fig. S10 in the ESM. As shown in Fig. 7(a), the refractive-index discrepancy of the surrounding medium caused the spectrum dip to redshift distinctly. According to the linear relationship between the dip shift (the dip position in the solvent referenced to that in air) and the refractive index in Fig. 7(b), the coefficient of determination was relatively high, indicating an excellent linear relationship. The slope represents the sensitivity, i.e., the RIS, of the d-AgNWAs structure, which was 933.3 nm·RIU–1 in this case. 3.3.2 Immunoassay of d-AgNWAs The d-AgNWAs (P = 1 μm, D = 504 nm, and d = H = 400 nm) were employed as a sensing platform for an immunoassay to evaluate their potential for practical application. The immunoassay was performed using the method described by Lisboa et al. [49]. The sample was first modified with 16-MHA by immersion in an alcoholic solution of 4 mM overnight. Afterward, human IgG was immobilized onto the substrate by electrostatic adsorption, followed by the immobilization of BSA to block the surface from nonspecific adsorption. A specificity test, i.e., a control experiment using goat anti-rabbit IgG, was performed before the

Figure 7 (a) Typical reflectance spectra of d-AgNWAs (P = 1 μm, d = H = 400 nm, and D = 504 nm) soaked in solvents with different refractive indices. (Only the normalized main dip parts of the spectra were kept.) (b) Dip shift with respect to the refractive index.

final recognition step. Finally, anti-human IgG with a series of concentrations was added to the system for over 20 min in the recognition step. The dip movement in the reflectance spectrum of the sample that occurred upon the whole immobilization step is shown in Fig. 8(a). The successful modification of MHA on the d-AgNWAs resulted in a 4.8-nm redshift of the dip position. The sequential immobilization of human IgG induced a further dip redshift of 2.4 nm. According to the specificity test, the incubation of the d-AgNWAs with anti-rabbit IgG caused almost no dip shift, which suggests a favorable structure specificity. For instance, the concentration of anti-human IgG in Fig. 8(a) was selected to be 10 μg·mL–1. The final recognition of anti-human IgG led to a dip redshift of 5 nm. To assess the biosensing capability of the d-AgNWAs, a calibration curve for the immunoassay was established by a dip-shift test

| www.editorialmanager.com/nare/default.asp

919

Nano Res. 2017, 10(3): 908–921

Figure 8 (a) Reflectance spectra of d-AgNWAs used for the immunoassay. The concentration of anti-human IgG was 10 μg·mL–1 in this case. (b) Calibration curve for the immunoassay between human IgG and anti-human IgG.

using a series of standard anti-human IgG analytes with gradually increasing concentration, as shown in Fig. 8(b). Within the tested concentration scope, the d-AgNWAs exhibited a linear response to the concentration of the analyte. The slope of the calibration curve represents the biosensing sensitivity, which was calculated to be 0.47 nm·mL·μg–1, which indicates the potential of the d-AgNWAs for biosensing applications. The d-AgNWAs can be applied as a substrate for biosensing because of their facile fabrication process, large-area ordered operation, and label-free sensing.

4

Conclusions

Highly sensitive d-AgNWAs for refractometric sensing were fabricated over a large area via colloidal lithography and metal deposition. The silicon-nanowell

template depth and silver-nanowell height influenced the reflectance-spectrum line shape and sensitivity of the structure. When the bulk-silver height within the Si nanowell template equaled the depth of the template, the sensing performance was least influenced by the silicon substrate, and the spectrum line shape was the best (most sharp and intense). Deepened metallic hole arrays afforded improved sensing performance. However, nanowell arrays greater than 500 nm in depth exhibit weakened sensitivity. The d-AgNWAs with a nanowell template depth equivalent to the silver-nanowell height of 400 nm provided the best optical and sensing performance. After several rounds of optimization, structural parameters of D = 504 nm, d = 400 nm, H = 400 nm, and P = 1 μm were selected for the best sensing performance of the d-AgNWAs: sensitivity of 933.3 nm·RIU–1. In contrast to thin nanohole arrays employed as plasmonic sensors, the d-AgNWAs show superior sensing performance, and the line shape of their reflectance spectrum is sharp and intense, making them promising candidates for refractometric sensing. The great potential of the d-AgNWAs for biosensing was proven by employing them in immunoassay. Notably, profiting from their large-area ordered structure fabrication feasibility, the d-AgNWAs can be integrated with microfluidic devices, offering multiplex sensing superiority. Moreover, the fabrication procedure of the d-AgNWAs is straightforward and allows mass production, making them cost-efficient and thus promising for practical sensing application.

Acknowledgements This work was financially supported by the National Basic Research Program of China (973 program, No. 2012CB933800) and the National Natural Science Foundation of China (NSFC, No. 91123031). Electronic Supplementary Material: Supplementary material (obtained large area highly ordered dAgNWAs’ images; the dip shift of d-AgNWAs with different D as a function of refractive index; the dip splitting for d-AgNWAs with extremely large D; the dip shift of d-AgNWAs with different d as a function of refractive index; SEM image of d-AgNWAs with H

www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano

Research

920

Nano Res. 2017, 10(3): 908–921

of 500 nm; the dip shift of d-AgNWAs with different H as a function of refractive index; the dip shift of d-AgNWAs with different d as a function of refractive index (H = 230 nm and H = 330 nm); the sensing performance contrast table for different depth under H of 230 nm and H of 330 nm; reflectance spectra of d-AgNWAs with sensitivity of 1,107.8 nm·RIU–1; complete reflectance spectra of the d-AgNWAs’ solvent test) is available in the online version of this article at http://dx.doi.org/10.1007/s12274-016-1348-7.

[11] Stewart, M. E.; Mack, N. H.; Malyarchuk, V.; Soares, J. A. N. T.; Lee, T. W.; Gray, S. K.; Nuzzo, R. G.; Rogers, J. A. Quantitative multispectral biosensing and 1D imaging using quasi-3D plasmonic crystals. Proc. Natl. Acad. Sci. USA 2006, 103, 17143–17148. [12] Ye, S. S.; Zhang, X. M.; Chang, L. X.; Wang, T. Q.; Li, Z. B.;

[13]

References [1] Barnes, W. L.; Dereux, A.; Ebbesen, T. W. Surface plasmon subwavelength optics. Nature 2003, 424, 824–830. [2] Zhong, L. B.; Jiang, Y. Y.; Liow, C.; Meng, F. B.; Sun, Y. H.; Chandran, B. K.; Liang, Z. Q.; Jiang, L.; Li, S. Z.; Chen, X. D. Highly sensitive electro-plasmonic switches based on fivefold stellate polyhedral gold nanoparticles. Small 2015, 11, 5395–5401. [3] Krishnan, A.; Thio, T.; Kim, T. J.; Lezec, H. J.; Ebbesen, T. W.; Wolff, P. A.; Pendry, J.; Martin-Moreno, L.; Garcia-Vidal, F. J. Evanescently coupled resonance in surface plasmon enhanced transmission. Opt. Commun. 2001, 200, 1–7. [4] Gordon, R.; Sinton, D.; Kavanagh, K. L.; Brolo, A. G. A new generation of sensors based on extraordinary optical transmission. Acc. Chem. Res. 2008, 41, 1049–1057. [5] Jiang, L.; Chen, X. D.; Lu, N.; Chi, L. F. Spatially confined assembly of nanoparticles. Acc. Chem. Res. 2014, 47, 3009–3017. [6] Jiang, L.; Sun, Y. H.; Nowak, C.; Kibrom, A.; Zou, C. J.; Ma, J.; Fuchs, H.; Li, S. Z.; Chi, L. F.; Chen, X. D. Patterning of plasmonic nanoparticles into multiplexed one-dimensional arrays based on spatially modulated electrostatic potential. ACS Nano 2011, 5, 8288–8294. [7] Jiang, L.; Zou, C. J.; Zhang, Z. H.; Sun, Y. H.; Jiang, Y. Y.; Leow, W.; Liedberg, B.; Li, S. Z.; Chen, X. D. Synergistic modulation of surface interaction to assemble metal nanoparticles into two-dimensional arrays with tunable plasmonic properties. Small 2014, 10, 609–616. [8] Du, J. J.; Zhu, B. W.; Chen, X. D. Urine for plasmonic nanoparticle-based colorimetric detection of mercury ion. Small 2013, 9, 4104–4111. [9] Walsh, G. F.; Negro, L. D. Engineering plasmon-enhanced Au light emission with planar arrays of nanoparticles. Nano Lett. 2013, 13, 786–792. [10] Mulvaney, P. Surface plasmon spectroscopy of nanosized metal particles. Langmuir 1996, 12, 788–800.

[14]

[15]

[16]

[17]

[18] [19]

[20]

[21]

[22]

Zhang, J. H.; Yang, B. High-performance plasmonic sensors based on two-dimensional Ag nanowell crystals. Adv. Opt. Mater. 2014, 2, 779–787. Yang, S. C.; Hou, J. L.; Finn, A.; Kumar, A.; Ge, Y.; Fischer, W. J. Synthesis of multifunctional plasmonic nanopillar array using soft thermal nanoimprint lithography for highly sensitive refractive index sensing. Nanoscale 2015, 7, 5760–5766. Zheng, Y. B.; Kiraly, B.; Cheunkar, S.; Huang, T. J.; Weiss, P. S. Incident-angle-modulated molecular plasmonic switches: A case of weak exciton-plasmon coupling. Nano Lett. 2011, 11, 2061–2065. Sun, Y. H.; Jiang, L.; Zhong, L. B.; Jiang, Y. Y.; Chen, X. D. Towards active plasmonic response devices. Nano Res. 2015, 8, 406–417. Weiler, M.; Quint, S. B.; Klenka, S.; Pacholski C. Bottom-up fabrication of nanohole arrays loaded with gold nanoparticles: Extraordinary plasmonic sensors. Chem. Commun. 2014, 50, 15419–15422. Ebbesen, T. W.; Lezec, H. J.; Ghaemi, H. F.; Thio, T.; Wolff, P. A. Extraordinary optical transmission through subwavelength hole arrays. Nature 1998, 391, 667–669. Genet, C.; Ebbesen, T. W. Light in tiny holes. Nature 2007, 445, 39–46. Yue, W. S.; Wang, Z. H.; Yang, Y.; Li, J. Q.; Wu, Y.; Chen, L. Q.; Ooi, B.; Wang, X. B.; Zhang, X. X. Enhanced extraordinary optical transmission (EOT) through arrays of bridged nanohole pairs and their sensing applications. Nanoscale 2014, 6, 7917–7923. Brolo, A. G.; Gordon, R.; Leathem, B.; Kavanagh, K. L. Surface plasmon sensor based on the enhanced light transmission through arrays of nanoholes in gold films. Langmuir 2004, 20, 4813–4815. De Leebeeck, A.; Kumar, L. K. S.; de Lange, V.; Sinton, D.; Gordon, R.; Brolo, A. G. On-chip surface-based detection with nanohole arrays. Anal. Chem. 2007, 79, 4094–4100. Lesuffleur, A.; Im, H.; Lindquist, N. C.; Lim, K. S.; Oh, S. H. Laser-illuminated nanohole arrays for multiplex plasmonic microarray sensing. Opt. Express 2008, 16, 219–224.

[23] Feuz, L.; Jönsson, P.; Jonsson, M. P.; Höök, F. Improving the limit of detection of nanoscale sensors by directed binding to high-sensitivity areas. ACS Nano 2010, 4, 2167–2177.

| www.editorialmanager.com/nare/default.asp

921

Nano Res. 2017, 10(3): 908–921

[24] Zhang, X. M.; Li, Z. B.; Ye, S. S.; Wu, S.; Zhang, J. H.; Cui, L. Y.; Li, A. R.; Wang, T. Q.; Li, S. Z.; Yang, B. Elevated Ag nanohole arrays for high performance plasmonic sensors based on extraordinary optical transmission. J. Mater. Chem. 2012, 22, 8903–8910. [25] Yanik, A. A.; Huang, M.; Kamohara, O.; Artar, A.; Geisbert, T. W.; Connor, J. H.; Altug, H. An optofluidic nanoplasmonic biosensor for direct detection of live viruses from biological media. Nano Lett. 2010, 10, 4962–4969. [26] Sannomiya, T.; Scholder, O.; Jefimovs, K.; Hafner, C.; Dahlin, A. B. Investigation of plasmon resonances in metal films with nanohole arrays for biosensing applications. Small 2011, 7, 1653–1663. [27] Nakamoto, K.; Kurita, R.; Niwa, O.; Fujii, T.; Nishida, M. Development of a mass-producible on-chip plasmonic nanohole array biosensor. Nanoscale 2011, 3, 5067–5075. [28] Caballero, B.; García-Martín, A.; Cuevas, J. C. Hybrid magnetoplasmonic crystals boost the performance of nanohole arrays as plasmonic sensors. ACS Photonics 2016, 3, 203–208. [29] Sharma, N.; Keshmiri, H.; Zhou, X. D.; Wong, T. I.; Petri, C.; Jonas, U.; Liedberg, B.; Dostalek, J. Tunable plasmonic nanohole arrays actuated by a thermoresponsive hydrogel cushion. J. Phys. Chem. C 2016, 120, 561–568. [30] Xiong, K. L.; Emilsson, G.; Dahlin, A. B. Biosensing using plasmonic nanohole arrays with small, homogenous and tunable aperture diameters. Analyst 2016, 141, 3803–3810. [31] Yao, J. M.; Stewart, M. E.; Maria, J.; Lee, T. W.; Gray, S. K.; Rogers, J. A.; Nuzzo, R. G. Seeing molecules by eye: Surface plasmon resonance imaging at visible wavelengths with high spatial resolution and submonolayer sensitivity. Angew. Chem., Int. Ed. 2008, 47, 5013–5017. [32] Stewart, M. E.; Anderton, C. R.; Thompson, L. B.; Maria, J.; Gray, S. K.; Rogers, J. A.; Nuzzo, R. G. Nanostructured plasmonic sensors. Chem. Rev. 2008, 108, 494–521. [33] Gao, H. W.; Henzie, J.; Odom, T. W. Direct evidence for surface plasmon-mediated enhanced light transmission through metallic nanohole arrays. Nano Lett. 2006, 6, 2104–2108. [34] Delamarche, E.; Schmid, H.; Michel, B.; Biebuyck, H. Stability of molded polydimethylsiloxane microstructures. Adv. Mater. 1997, 9, 741–746. [35] Malyarchuk, V.; Hua, F.; Mack, N. H.; Velasquez, V. T.; White, J. O.; Nuzzo, R. G.; Rogers, J. A. High performance plasmonic crystal sensor formed by soft nanoimprint lithography. Opt. Express 2005, 13, 5669–5675. [36] Zhang, J. H.; Li, Y. F.; Zhang, X. M.; Yang, B. Colloidal

[37]

[38]

[39]

[40]

[41]

[42]

[43]

[44]

[45]

[46] [47]

[48]

[49]

self-assembly meets nanofabrication: From two-dimensional colloidal crystals to nanostructure arrays. Adv. Mater. 2010, 22, 4249–4269. Murray-Méthot, M. P.; Ratel, M.; Masson, J. F. Optical properties of Au, Ag, and bimetallic Au on Ag nanohole arrays. J. Phys. Chem. C 2010, 114, 8268–8275. Park, T. H.; Mirin, N.; Lassiter, J. B.; Nehl, C. L.; Halas, N. J.; Nordlander, P. Optical properties of a nanosized hole in a thin metallic film. ACS Nano 2008, 2, 25–32. Degiron, A.; Lezec, H. J.; Barnes, W. L.; Ebbesen, T. W. Effects of hole depth on enhanced light transmission through subwavelength hole arrays. Appl. Phys. Lett. 2002, 81, 4327–4329. Thio, T.; Ghaemi, H. F.; Lezec, H. J.; Wolff, P. A.; Ebbesen, T. W. Surface-plasmon-enhanced transmission through hole arrays in Cr films. J. Opt. Soc. Am. B 1999, 16, 1743–1748. Zhang, J. H.; Chen, Z.; Wang, Z. L.; Zhang, W. Y.; Ming, N. B. Preparation of monodisperse polystyrene spheres in aqueous alcohol system. Mater. Lett. 2003, 57, 4466–4470. Bukasov, R.; Shumaker-Parry, J. S. Highly tunable infrared extinction properties of gold nanocrescents. Nano Lett. 2007, 7, 1113–1118. Hicks, E. M.; Zhang, X. Y.; Zou, S. L.; 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. Valsecchi, C.; Brolo, A. G. Periodic metallic nanostructures as plasmonic chemical sensors. Langmuir 2013, 29, 5638–5649. Sherry, L. J.; Chang, S.-H.; Schatz, G. C.; van Duyne. R. P.; Wiley, B. J.; Xia, Y. N. Localized surface plasmon resonance spectroscopy of single silver nanocubes. Nano Lett. 2005, 5, 2034–2038. Nehl, C. L.; Liao, H. W.; Hafner, J. H. Optical properties of star-shaped gold nanoparticles. Nano Lett. 2006, 6, 683–688. McFarland, A. D.; van Duyne, R. P. Single silver nanoparticles as real-time optical sensors with zeptomole sensitivity. Nano Lett. 2003, 3, 1057–1062. Mock, J. J.; Smith, D. R.; Schultz, S. Local refractive index dependence of plasmon resonance spectra from individual nanoparticles. Nano Lett. 2003, 3, 485–491. Lisboa, P.; Valsesia, A.; Mannelli, I.; Mornet, S.; Colpo, P.; Rossi, F. Sensitivity enhancement of surface-plasmon resonance imaging by nanoarrayed organothiols. Adv. Mater. 2008, 20, 2352–2358.

www.theNanoResearch.com∣www.Springer.com/journal/12274 | Nano

Research