Home
Search
Collections
Journals
About
Contact us
My IOPscience
Ag films annealed in a nanoscale limited area for surface-enhanced Raman scattering detection
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 Nanotechnology 25 235301 (http://iopscience.iop.org/0957-4484/25/23/235301) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 134.151.40.2 This content was downloaded on 05/09/2014 at 07:09
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 25 (2014) 235301 (7pp)
doi:10.1088/0957-4484/25/23/235301
Ag films annealed in a nanoscale limited area for surface-enhanced Raman scattering detection Dan Jiang, Shuping Xu, Hailong Wang, Ming Cong, Yuyang Wang and Weiqing Xu State Key Laboratory of Supramolecular Structure and Materials and Institute of Theoretical Chemistry, Jilin University, Changchun, 130012, People’s Republic of China E-mail: [email protected] Received 24 January 2014, revised 24 March 2014 Accepted for publication 26 March 2014 Published 21 May 2014 Abstract
By constructing a limited two-dimensional (2D) area to regulate the coalescence of deposited Ag nanoparticles and to change the degree of aggregation of nanoparticles under annealing, a configuration of particles on pillar (POP) was fabricated successfully and high SERS enhancement similar to ‘hot spots’ was generated on the POP. With the assistance of nanosphere lithography (NSL), we constructed a silicon pillar array with a column diameter of 200 nm and a height of 155 nm. A thin layer of Ag film with a several nanometer thickness was deposited on the top of the silicon pillar array and then was annealed under 200 °C. Differing from the annealed Ag on a flat surface, the aggregation of Ag on the top of a pillar was predominated by the limited area, forming more Ag nanoaggregates. The SERS spectra demonstrated that the POP contributed to higher SERS enhancement than the silicon slide modified by the some thickness annealed Ag film. As a consequence, by means of POP configuration, the lowest SERS detection concentration for 4-mercaptopyridine was able to reach the level of 1.0 × 10−10 M. S Online supplementary data available from stacks.iop.org/NANO/25/235301/mmedia Keywords: SERS, Ag film, silicon pillar, hot spots (Some figures may appear in colour only in the online journal)
generally gives only 10–102 enhancement for probed molecules but EM field magnification contributes more than 106 to SERS enhancement [1, 2]. Therefore, currently, the EM enhancement mechanism is considered as the prevailing view [1–4]. Owing to the fact that the extreme enhancement derives from the interstitial regions of intense field amplification (socalled ‘hot spots’) between two close metal nanoparticles [5–7], approaches to fabricate novel SERS substrates are mainly by constructing more controllable narrow gap-type or tip-type configurations [8]. At present, nanopatterning [9] and electron and ion beam lithography [10, 11] are popularized to build hot spots, due to the merits that they can control and regulate the morphology
1. Introduction Surface-enhanced Raman scattering (SERS), considered as one of the most promising tools for molecular analysis, has attracted a great deal of attention in the last few decades, because of its distinctive vibrational modes for the detection and identification of chemical and biological analytes. In terms of SERS enhancement mechanism, there are two perspectives accepted widely. One is electromagnetic (EM) enhancement that originates from the coupling between noble nanoparticles and the other is chemical mechanism (CM), i.e. the interaction between metal and molecules adsorbed on its surface, e.g., the charge-transfer effect. In terms of CM, it 0957-4484/14/235301+07$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
Nanotechnology 25 (2014) 235301
D Jiang et al
Scheme 1. Schematic illustration of the fabrication of two kinds of SERS substrates. Steps (a) to (h) describe the preparation of the POP configuration, while steps (a) to (i) show the Ag nanoaggregates formed on a silicon flat slide.
and the particle position down to several nanometers. For example, K D Alexander et al [12] displayed a highly-ordered SERS substrate which yielded an enhancement factor as high as 109 by means of nanopatterning, assisted with the meniscus force deposition technique that merely located two Ag nanoparticles (diameter (D) = ∼60 nm) into the 120-nm diameter holes. In addition, by applying electron beam lithography, A Chen et al [13] built a grating template with the period of 550 nm to contain four Au nanoparticles (D = ∼80 nm) within the grating grooves. This novel SERS substrate produced an average SERS enhancement factor (EF) up to 108. These examples both used the meniscus force to assemble metal colloids into a limited space for SERS detection. Besides the self-assembly of colloidal metal nanoparticles, physical routes [14–17], mainly referring to metallic cluster deposition, were also proposed to fabricate the hot-spot SERS substrates. X Li et al [18] constructed a highly ordered bowl array to gather Ag nanoparticles. After annealing, many hot spots were created within the bottoms of the bowls, presenting high SERS behavior. The above methods propose strategies for the construction of hot spots, i.e., holding metal in a micro or submicro meter-scale limited space. Thus, the micro-fabrication and micro-machining technologies are essential to support these designs. As is known, nanosphere lithography (NSL) is very effective and widely used for constructing periodic arrays at micro or submicro meter scale. Via the self-assembly of the nanospheres of silica, polystyrene (PS) or other materials on a substrate, a closepacked hexagonal structure was achieved. With these nanospheres
as a shelter followed with reactive ion etching (RIE) treatment, we can copy its hexagonal structure on solid-supported substrates (e.g Al [19], silicon [20], and glass [8, 21] etc) and obtain positive/ negative hexagonal, cylindrical or discal arrays. In this study, with the assistance of NSL and RIE, we fabricated silicon pillars with a two-dimensional (2D) hexagonal pattern. Then, we deposited a thin layer of Ag films onto the silicon pillars by vacuum evaporation deposition. After annealing, the discontinuous Ag islands commenced to merge, as well as the size and gap of the Ag nanoparticles gradually changed with the thickness of the Ag film, which is vital to the localized surface plasmon (LSP) coupling and has an affinity for SERS enhancement. For the purpose of comparative research, an Ag film deposited on a flat silicon slide was prepared, followed by an annealing process. The SERS performance of two kinds of substrates was evaluated and the results showed that the designed particle on pillar (POP) substrate supported higher SERS signal.
2. Experimental section 2.1. Materials
PS nanospheres, 362 nm in diameter, were synthesized by the emulsion polymerization method [22, 23]. The silver coils of purity grade of 99.99% for evaporation was purchased from Sinopharm (China). 4-mercaptopyridine (4-MPY, 95%) was purchased from Sigma-Aldrich. 2
Nanotechnology 25 (2014) 235301
D Jiang et al
I 156.65nm
155.83nm
100
300.0 nm
159.32nm
0.0 nm
-100
0
150.0 nm
0
1.00
2.00
µm
Figure 1. SEM images of A to H present the substrates (b)–(i) in scheme 1. Scales all are 500 nm. I shows an AFM image of the silicon pillar array (substrate (b)) with the scan areas of 2.0 μm × 2.0 μm and a height profile along the black line.
5 mTorr, an RF power of 20 W, an etching time of 100 s, and a CHF3 and SF6 flow rate of 30 and 6 SCCM. Finally, the PS was removed by RIE (step 5). The etching conditions were same as the step 2, except the etching time of 9 min. Consequently, a highly orderly silicon pillar array was generated. We deposited a thin layer of Ag films onto the silicon pillar array using vacuum evaporate deposition (step 6). The thickness of the Ag film was controlled as 3–9 nm, respectively. Then, the Ag film on the silicon pillar array was annealed at 200 °C under a nitrogen atmosphere for 2.0 h. Finally, the POP configuration was obtained (Substrate h in scheme 1). For comparison, we prepared a control sample using a flat silicon slide with the same thickness of Ag deposition, followed with annealing with the conditions above. This control substrate (Substrate i in scheme 1) shows an example of another kind of Ag nanoaggregates induced by an unlimited area. The scanned electron microscopy (SEM, Hitachi H-8020 IV operated at 5.0 kV) was employed to characterize the morphologies of each step. Atomic force microscope (AFM, Digital Instruments Dimension 3100) was used to reveal the morphology and the height of the silicon pillar. The reflection spectra of the silicon flat and pillar substrates were measured by an Ocean Optics USB4000 spectrometer to evaluate the
2.2. Preparation and characterization of POP
Flat silicon wafers were cleaned using a boiling piranha solution consisting of a 3:7 mixture of concentrated H2O2 (30%) and H2SO4 for 30 min. After cooling, silicon wafers were rinsed repeatedly with purified water. Scheme 1 presents the procedures of constructing the POP SERS substrate. It mainly contains two processes: (1) the preparation of a silicon pillar array by NSL and RIE methods and (2) Ag deposition followed with annealing. A monolayer of close-packed PS nanospheres (D1 = 362 nm) was first assembled onto a cleaned silicon wafer (the details consult ref [22]) (step 1). Then, the PS spheres were treated with a Plasmalab Oxfold 80 plus (ICP65) system (Oxford Instrument Co., Bristol, UK) to decrease the diameter D2 into 200 nm (step 2). The etching conditions were a pressure of 10 mTorr, an RF power of 30 W, an etching time of 160 s and an O2 flow rate of 50 SCCM. After that, we heated the etched PS spheres on the silicon substrate to 120 °C for 10 min to make PS closely attach to the silicon surface (step 3). We continued to etch the bottom silicon slide with the PS nanosphere array as a mask to fabricate a silicon pillar array (step 4). The etching conditions were a pressure of 3
Nanotechnology 25 (2014) 235301
D Jiang et al
recorded on a Jobin Yvon/HORIBA LabRam ARAMIS Raman spectrometer. The integration time for both was 50 s and the laser powers were 8.3 and 33 mW, respectively.
3. Results and discussion 3.1. SEM characterization of the POP
Scheme 1 shows the procedures of constructing the POP structure. To confirm each step, we scanned the SEM images of the prepared silicon substrates from each step of the construction. Figure 1(A) presents the PS sphere (diameter 362 nm) monolayer assembled on a flat silicon slide surface with a close-packed hexagonal array, and then the RIE was used to decrease the diameter of PS spheres to 200 nm (figure 1(B)). Through heating, the PS spheres collapsed on the silicon slide surface (figure 1(C)) and by the treatment of RIE the silicon pillar array was achieved (figure 1(D)). It should be noted that only the use of collapsed PS spheres as a mask can lead to a cylindrical silicon array under RIE. If the PS spheres without heating are used as a mask, the conical silicon array will be obtained. Figure 2(E) presents the highly ordered silicon pillar array after PS was removed by RIE. The picture indicates that the pillar array is still of hexagonal arrangement. Figure 1(F) reveals deposited Ag films of a thickness of 7 nm on the silicon pillar surface. It is observed that there exists a uniform Ag island film via the vacuum evaporation deposition. Figures 1(G) and (H) present the aggregates of Ag nanoparticles on the top of the silicon pillar and the flat silicon slide after the annealing process. It can be found that all Ag nanoparticles coalesced on the substrates or only on the top of the silicon pillars. Note that there are no Ag nanoparticles forming onto the sidewalls of silicon pillars. However, on the top of the silicon pillar, they formed more compact aggregates and rather loose aggregates on the unlimited surface. 3.2. Optimization of the thickness of deposited Ag
The different thickness Ag films were deposited on the silicon pillar arrays to investigate the morphology of the POP and optimize the performance of the SERS substrate. Figures 2(A)–(C) present the SEM images of three kinds of silicon pillar arrays decorated with different Ag nanoaggregates, which were produced by annealing 3, 7, and 8 nm thicknesses of Ag films at 200 °C for 2.0 h in nitrogen ambience. From figures 1(F) and (G), it is obvious that discrete Ag islands commenced to coalesce and finally Ag films were transformed into regularly-shaped and -sized nanoparticles. However, there are many differences between them. It is observed in figure 2(A) that the Ag nanoparticles on the silicon pillar display a large interspace and the average size is approximately 20 nm. In contrast, the Ag nanoparticles in figure 2(B) exhibit a relatively small gap and their average size approaches 60 nm, which is supposed to have a large contribution to the SERS enhancement [24]. In terms of figure 2(C), the mean size of the Ag islands approximately reaches 100 nm, constructing a bigger gap between the nanoparticles than the 7 nm Ag film, thus forming a smaller
Figure 2. SEM images of the POPs with different thicknesses of Ag
film after annealing (A) to (C); 3, 7, and 8 nm). They are in the same magnification and the scale represents 500 nm.
plasmonic properties. The parameters for recording the reflection spectra were as follows, integration time of 8 ms, scans to average of 50 and a boxcar width of 10. 2.3. SERS characterization
4-MPY as the probe molecule was used to evaluate the SERS activities of the prepared substrates. First, 4-MPY molecules (10 μL, 1.0 × 10−6 M) were adsorbed onto the construction of the POP by dripping. The diameter of the natural drying drop of 4-MPY was 4.32 mm. The SERS spectra of 4-MPY were recorded by a portable BWTEK Raman spectrometer with 532 nm excitation wavelength. The laser power was 25.4 mW, and the integration time was 5 s. Raman spectra under the excitation of 633 and 785 nm wavelengths were 4
Nanotechnology 25 (2014) 235301
D Jiang et al
Figure 3. (A) The SERS spectra characterize the SERS enhancement of the 4-MPY (1.0 × 10−6 M) under 633 nm wavelength excitation, with
the Ag films of different thicknesses after annealing. From bottom to top, the thicknesses of the Ag films are 3, 5, 7, 8, and 10 nm, respectively. As SERS signals of the POP with 7, 8 nm were considerably stronger compared to ones of the others, for the sake of visual effect, the corresponding SERS signals of 7 and 8 nm were divided by 4 and 2, respectively. (B) The SERS band intensity at 1092 cm−1 with the Ag film thickness when detecting 4-MPY.
number of nanoparticles due to the huge volume compared to figures 2(A) and (B). It is reported that the distance between the nanoparticles and the size of nanoparticles both highly affect the LSPR and in turn, the SERS activity [1]. To display the SERS of different annealed Ag films, we used 4-MPY to evaluate their enhancement ability. Figure 3(A) shows the SERS spectra of 4-MPY on the POPs with different thickness of annealed Ag films. From figure 3(A), we can observe the existence of characteristic peaks of 4-MPY at 1003, 1060, 1092, 1214, 1575, and 1606 cm−1. Also, several characteristic peaks of 4-MPY (e.g. 1575 and 1606 cm−1) on the POP with 3, 5, 10 nm-thickness annealed Ag films (a, b, e line) can be obscurely identified. The profile of the Raman intensity with the Ag film thickness is shown in figure 3(B), which indicates that the maximal SERS enhancement appeared with the 7 nm annealed Ag film. So we used the POP with the 7 nm annealed Ag film in the following experiment.
Figure 4. The reflection spectra of silicon pillar array and flat silicon substrate with and without Ag nanoparticles.
reflectance spectra of a periodic moth-eye silicon nanostructure, and found that this structure can be thought of as an effective medium that approximates a single layer thin film when the shape of silicon pillars is with near-vertical side walls and flat tops [25]. Similar to the moth-eye silicon nanostructure they used, in the present work, we assigned the multiple resonance modes in 482 and 650 nm in the reflection spectra of the silicon pillar array to the interference effect. After the 7 nm thickness Ag film was deposited on the flat silicon slide followed with annealing, there appears a broad band in the range of 450–740 nm, which is ascribed to the dipolar resonance of Ag nanoparticles from annealed Ag film. Compared with the reflection spectrum of the silicon pillar, the interference peaks at 482 and 650 nm slightly red shifted to 497 and 678 nm in POP. We attributed the redshifts to the light scattering of the formed Ag nanoparticles,
3.3. Comparison of SERS enhancement on a limited and an unlimited area 3.3.1. Reflection spectra.
The SERS activity of a substrate is strongly affected by its plasmonic properties. So we measured the reflection spectrum of the prepared POP substrates (figure 4) to reflect their plasmonic properties. For comparison, the reflection spectrum of the annealed Ag on a flat silicon slide was recorded. We also recorded the reflection spectra of the silicon pillar array and the flat silicon slide as a reference. Figure 4 shows the reflection spectra of the silicon pillar array and the flat silicon substrate with and without Ag nanoparticles. It is obviously observed that the spectra of the silicon slide show a high reflectivity in 400–800 nm range with no obvious peak appearing, while the configuration of the silicon pillar presents multiple resonance modes in 482 and 650 nm. Boden et al studied that the 5
Nanotechnology 25 (2014) 235301
D Jiang et al
−6 Figure 5. SERS spectra of 4-MPY (10 μL, 1.0 × 10 M) on the POP (a) and the flat silicon slide (b). They were both decorated by an annealed 7 nm thickness silver film. A normal Raman spectrum of 4MPY (10 μL, 0.1 M) on a flat silicon slide (c).
which can be proved by the SEM images of POP (figure 1(G)) and Ag particles on a flat slide (figure 1(H)) after annealing, in which Ag films have been transformed into regularlyshaped and -sized nanoparticles. In addition, the simulated reflection spectra (see figure S1 available in the supplementary data at stacks.iop.org/NANO/ 25/235301/mmedia) show the same trend. The reflection spectra make a reference to the choice of the SERS excitation wavelength. To exactly match the resonance band of the POP for the strongest SERS, the excitation wavelength of 633 nm was adopted.
Figure 6. (A) SERS spectra of 4-MPY at different concentrations on
the POPs measured with a 633 nm laser. (B) The plot of SERS intensities at 1092 cm−1 in (A) as the 4-MPY concentration.
silicon substrate modified by the annealed 7 nm thickness Ag film, respectively. The SERS determination of 4-MPY was presented in figure 6(A), and the profile of the SERS intensities with the 4MPY concentration was given by figure 6(B). We can discern that the lowest detection concentration of 4-MPY reaches 1.0 × 10−10 M.
3.3.2. SERS behavior. Figure 5 shows the SERS spectra of
4-MPY on three different substrates under the excitation with 633 nm. The SERS spectra under 532 and 785 nm excitation were provided in the supplementary information figure S3. Three substrates, respectively, are an unvarnished silicon slide (curve c), a silicon slide whose surface is decorated by an annealed 7 nm thickness silver film (curve b), and a POP substrate with an annealed 7 nm thickness silver film (curve a). Obviously, curve a in figure 5 supports the strongest SERS signal. The SERS signals on the POP are considerably stronger than the flat silicon substrate because the limited 2D area of the silicon pillar top decreases the distance between Ag nanoparticles and regulates the size and morphology of the nanoparticles, which facilitates LSP coupling and EM field enhancement, and the silicon pillar array also displays some influence to SERS enhancement which is evidenced by the FDTD simulation in the supplementary information. The EFs were calculated by comparing curve a (or b) to c (in figure 5). The details of the EFs calculation are also provided in the supplementary information. EFs for the 1092 cm−1 band are 1.5 × 106 and 2.4 × 105 for the POP and the flat
4. Conclusions In this work, we explored an approach to fabricate a hot spot SERS substrate by depositing Ag film onto the limited 2D area under annealing. This method is beneficial to the aggregation between nanoparticles and the improvement of the SERS signal. By observing the effect of the deposited silver films of different thicknesses on SERS enhancement, it is founded that the thickness of 7 nm is best suited. In addition, the comparative research between POP and the flat silicon substrate demonstrates that POP is preferable in SERS enhancement. Moreover, our experiments support these 6
Nanotechnology 25 (2014) 235301
D Jiang et al
[4] Moskovits M 1985 Rev. Mod. Phys. 57 783 [5] Giannini V, Fernández-Domínguez A I, Heck S C and Maier S A 2011 Chem. Rev. 111 3888 [6] Halas N J, Lal S, Chang W, Link S and Nordlander P 2011 Chem. Rev. 111 3913 [7] Xu H, Aizpurua J, Kall M and Apell P 2000 Phys. Rev. E 62 4318 [8] Guo H, Xu W, Zhou J, Xu S and Lombardi J R 2013 Nanotechnology 24 045608 [9] Bhuvana T and Kulkarni G U 2008 Small 4 670 [10] Cialla D, Hübner U, Schneidewind H, Möller R and Popp J 2008 ChemPhysChem 9 758 [11] Jeon H C, Heo C, Lee S Y and Yang S 2012 Adv. Funct. Mater. 22 4268 [12] Alexander K D, Hampton M J, Zhang S, Dhawan A, Xu H and Lopeza R 2009 J. Raman Spectrosc. 40 2171 [13] Chen A, Miller R L, DePrince A E, Joshi-Imre A, Shevchenko E, Ocola L E, Gray S K, Welp U and Vlasko-Vlasov V K 2013 Small 9 1939 [14] Xu L, Tan L S and Hong M H 2011 Appl. Opt. 50 G74 [15] Upender G, Sathyavathi R, Raju B, Bansal C and Rao D N 2012 J. Mol. Struct. 1012 56 [16] Giorgis F, Descrovi E, Chiodoni A, Froner E, Scarpa M, Venturello A and Geobaldo F 2008 Appl. Surf. Sci. 254 7494 [17] Van Duyne R P, Hulteen J C and Treichel D A 1993 J. Chem. Phys. 99 2101 [18] Li X, Zhang Y, Shen Z X and Fan H J 2012 Small 8 2548 [19] Wang X, Xu S, Cong M, Li H, Gu Y and Xu W 2012 Small 8 972 [20] Wang C, Ruan W, Ji N, Ji W, Lv S, Zhao C and Zhao B 2010 J. Phys. Chem. C 114 2886 [21] Li H, Xu S, Gu Y, Wang H, Ma R, Lombardi J R and Xu W 2013 J. Phys. Chem. C 117 19154 [22] Li Y, Cai W and Duan G 2008 Chem. Mater. 20 615 [23] Li Y et al 2009 Adv. Mater. 21 4731 [24] Emory S R, Haskins W E and Nie S 1998 J. Am. Chem. Soc. 120 8009 [25] Boden S A and Bagnall D M 2008 Appl. Phys. Lett. 93 133108
conclusions and discover the lowest detection concentration of 4-MPY on the POP up to 1.0 × 10−10 M. Our research provides a simple approach to constructing a SERS substrate and the obtained POP configuration achieves a highly sensitive SERS measurement, which proves that the limited 2D area plays a dominating role in the formation of hot spots. Compared with other similar researches, the POP structure may present comparable SERS enhancement ability. Whereas, the silicon –based SERS substrate itself has an interior standard peak at 520 cm−1, which is very vital to SERS detection by decreasing detection error. More importantly, the POP substrate can be provided in wafer-scale, which is difficult for other micro fabrication technology, such as nanopatterning, electron and ion beam lithography etc. and it is promising for becoming a commercial SERS substrate.
Acknowledgments This work was supported by the National Natural Science Foundation of China NSFC Grant (Nos 21373096, 21073073, and 91027010) and National Instrumentation Program (NIP) of the Ministry of Science and Technology of China No. 2011YQ03012408 and Innovation Program of the State Laboratory of Supramolecular Structure and Meterials.
References [1] Guerrini L and Graham D 2012 Chem. Soc. Rev. 41 7085 [2] Sebastian S 2009 Chem. Phys. Chem. 10 1344 [3] Lombardi J R, Birke R L, Lu T and Xu J 1986 J. Chem. Phys. 84 4174
7
Search
Collections
Journals
About
Contact us
My IOPscience
Ag films annealed in a nanoscale limited area for surface-enhanced Raman scattering detection
This content has been downloaded from IOPscience. Please scroll down to see the full text. 2014 Nanotechnology 25 235301 (http://iopscience.iop.org/0957-4484/25/23/235301) View the table of contents for this issue, or go to the journal homepage for more
Download details: IP Address: 134.151.40.2 This content was downloaded on 05/09/2014 at 07:09
Please note that terms and conditions apply.
Nanotechnology Nanotechnology 25 (2014) 235301 (7pp)
doi:10.1088/0957-4484/25/23/235301
Ag films annealed in a nanoscale limited area for surface-enhanced Raman scattering detection Dan Jiang, Shuping Xu, Hailong Wang, Ming Cong, Yuyang Wang and Weiqing Xu State Key Laboratory of Supramolecular Structure and Materials and Institute of Theoretical Chemistry, Jilin University, Changchun, 130012, People’s Republic of China E-mail: [email protected] Received 24 January 2014, revised 24 March 2014 Accepted for publication 26 March 2014 Published 21 May 2014 Abstract
By constructing a limited two-dimensional (2D) area to regulate the coalescence of deposited Ag nanoparticles and to change the degree of aggregation of nanoparticles under annealing, a configuration of particles on pillar (POP) was fabricated successfully and high SERS enhancement similar to ‘hot spots’ was generated on the POP. With the assistance of nanosphere lithography (NSL), we constructed a silicon pillar array with a column diameter of 200 nm and a height of 155 nm. A thin layer of Ag film with a several nanometer thickness was deposited on the top of the silicon pillar array and then was annealed under 200 °C. Differing from the annealed Ag on a flat surface, the aggregation of Ag on the top of a pillar was predominated by the limited area, forming more Ag nanoaggregates. The SERS spectra demonstrated that the POP contributed to higher SERS enhancement than the silicon slide modified by the some thickness annealed Ag film. As a consequence, by means of POP configuration, the lowest SERS detection concentration for 4-mercaptopyridine was able to reach the level of 1.0 × 10−10 M. S Online supplementary data available from stacks.iop.org/NANO/25/235301/mmedia Keywords: SERS, Ag film, silicon pillar, hot spots (Some figures may appear in colour only in the online journal)
generally gives only 10–102 enhancement for probed molecules but EM field magnification contributes more than 106 to SERS enhancement [1, 2]. Therefore, currently, the EM enhancement mechanism is considered as the prevailing view [1–4]. Owing to the fact that the extreme enhancement derives from the interstitial regions of intense field amplification (socalled ‘hot spots’) between two close metal nanoparticles [5–7], approaches to fabricate novel SERS substrates are mainly by constructing more controllable narrow gap-type or tip-type configurations [8]. At present, nanopatterning [9] and electron and ion beam lithography [10, 11] are popularized to build hot spots, due to the merits that they can control and regulate the morphology
1. Introduction Surface-enhanced Raman scattering (SERS), considered as one of the most promising tools for molecular analysis, has attracted a great deal of attention in the last few decades, because of its distinctive vibrational modes for the detection and identification of chemical and biological analytes. In terms of SERS enhancement mechanism, there are two perspectives accepted widely. One is electromagnetic (EM) enhancement that originates from the coupling between noble nanoparticles and the other is chemical mechanism (CM), i.e. the interaction between metal and molecules adsorbed on its surface, e.g., the charge-transfer effect. In terms of CM, it 0957-4484/14/235301+07$33.00
1
© 2014 IOP Publishing Ltd Printed in the UK
Nanotechnology 25 (2014) 235301
D Jiang et al
Scheme 1. Schematic illustration of the fabrication of two kinds of SERS substrates. Steps (a) to (h) describe the preparation of the POP configuration, while steps (a) to (i) show the Ag nanoaggregates formed on a silicon flat slide.
and the particle position down to several nanometers. For example, K D Alexander et al [12] displayed a highly-ordered SERS substrate which yielded an enhancement factor as high as 109 by means of nanopatterning, assisted with the meniscus force deposition technique that merely located two Ag nanoparticles (diameter (D) = ∼60 nm) into the 120-nm diameter holes. In addition, by applying electron beam lithography, A Chen et al [13] built a grating template with the period of 550 nm to contain four Au nanoparticles (D = ∼80 nm) within the grating grooves. This novel SERS substrate produced an average SERS enhancement factor (EF) up to 108. These examples both used the meniscus force to assemble metal colloids into a limited space for SERS detection. Besides the self-assembly of colloidal metal nanoparticles, physical routes [14–17], mainly referring to metallic cluster deposition, were also proposed to fabricate the hot-spot SERS substrates. X Li et al [18] constructed a highly ordered bowl array to gather Ag nanoparticles. After annealing, many hot spots were created within the bottoms of the bowls, presenting high SERS behavior. The above methods propose strategies for the construction of hot spots, i.e., holding metal in a micro or submicro meter-scale limited space. Thus, the micro-fabrication and micro-machining technologies are essential to support these designs. As is known, nanosphere lithography (NSL) is very effective and widely used for constructing periodic arrays at micro or submicro meter scale. Via the self-assembly of the nanospheres of silica, polystyrene (PS) or other materials on a substrate, a closepacked hexagonal structure was achieved. With these nanospheres
as a shelter followed with reactive ion etching (RIE) treatment, we can copy its hexagonal structure on solid-supported substrates (e.g Al [19], silicon [20], and glass [8, 21] etc) and obtain positive/ negative hexagonal, cylindrical or discal arrays. In this study, with the assistance of NSL and RIE, we fabricated silicon pillars with a two-dimensional (2D) hexagonal pattern. Then, we deposited a thin layer of Ag films onto the silicon pillars by vacuum evaporation deposition. After annealing, the discontinuous Ag islands commenced to merge, as well as the size and gap of the Ag nanoparticles gradually changed with the thickness of the Ag film, which is vital to the localized surface plasmon (LSP) coupling and has an affinity for SERS enhancement. For the purpose of comparative research, an Ag film deposited on a flat silicon slide was prepared, followed by an annealing process. The SERS performance of two kinds of substrates was evaluated and the results showed that the designed particle on pillar (POP) substrate supported higher SERS signal.
2. Experimental section 2.1. Materials
PS nanospheres, 362 nm in diameter, were synthesized by the emulsion polymerization method [22, 23]. The silver coils of purity grade of 99.99% for evaporation was purchased from Sinopharm (China). 4-mercaptopyridine (4-MPY, 95%) was purchased from Sigma-Aldrich. 2
Nanotechnology 25 (2014) 235301
D Jiang et al
I 156.65nm
155.83nm
100
300.0 nm
159.32nm
0.0 nm
-100
0
150.0 nm
0
1.00
2.00
µm
Figure 1. SEM images of A to H present the substrates (b)–(i) in scheme 1. Scales all are 500 nm. I shows an AFM image of the silicon pillar array (substrate (b)) with the scan areas of 2.0 μm × 2.0 μm and a height profile along the black line.
5 mTorr, an RF power of 20 W, an etching time of 100 s, and a CHF3 and SF6 flow rate of 30 and 6 SCCM. Finally, the PS was removed by RIE (step 5). The etching conditions were same as the step 2, except the etching time of 9 min. Consequently, a highly orderly silicon pillar array was generated. We deposited a thin layer of Ag films onto the silicon pillar array using vacuum evaporate deposition (step 6). The thickness of the Ag film was controlled as 3–9 nm, respectively. Then, the Ag film on the silicon pillar array was annealed at 200 °C under a nitrogen atmosphere for 2.0 h. Finally, the POP configuration was obtained (Substrate h in scheme 1). For comparison, we prepared a control sample using a flat silicon slide with the same thickness of Ag deposition, followed with annealing with the conditions above. This control substrate (Substrate i in scheme 1) shows an example of another kind of Ag nanoaggregates induced by an unlimited area. The scanned electron microscopy (SEM, Hitachi H-8020 IV operated at 5.0 kV) was employed to characterize the morphologies of each step. Atomic force microscope (AFM, Digital Instruments Dimension 3100) was used to reveal the morphology and the height of the silicon pillar. The reflection spectra of the silicon flat and pillar substrates were measured by an Ocean Optics USB4000 spectrometer to evaluate the
2.2. Preparation and characterization of POP
Flat silicon wafers were cleaned using a boiling piranha solution consisting of a 3:7 mixture of concentrated H2O2 (30%) and H2SO4 for 30 min. After cooling, silicon wafers were rinsed repeatedly with purified water. Scheme 1 presents the procedures of constructing the POP SERS substrate. It mainly contains two processes: (1) the preparation of a silicon pillar array by NSL and RIE methods and (2) Ag deposition followed with annealing. A monolayer of close-packed PS nanospheres (D1 = 362 nm) was first assembled onto a cleaned silicon wafer (the details consult ref [22]) (step 1). Then, the PS spheres were treated with a Plasmalab Oxfold 80 plus (ICP65) system (Oxford Instrument Co., Bristol, UK) to decrease the diameter D2 into 200 nm (step 2). The etching conditions were a pressure of 10 mTorr, an RF power of 30 W, an etching time of 160 s and an O2 flow rate of 50 SCCM. After that, we heated the etched PS spheres on the silicon substrate to 120 °C for 10 min to make PS closely attach to the silicon surface (step 3). We continued to etch the bottom silicon slide with the PS nanosphere array as a mask to fabricate a silicon pillar array (step 4). The etching conditions were a pressure of 3
Nanotechnology 25 (2014) 235301
D Jiang et al
recorded on a Jobin Yvon/HORIBA LabRam ARAMIS Raman spectrometer. The integration time for both was 50 s and the laser powers were 8.3 and 33 mW, respectively.
3. Results and discussion 3.1. SEM characterization of the POP
Scheme 1 shows the procedures of constructing the POP structure. To confirm each step, we scanned the SEM images of the prepared silicon substrates from each step of the construction. Figure 1(A) presents the PS sphere (diameter 362 nm) monolayer assembled on a flat silicon slide surface with a close-packed hexagonal array, and then the RIE was used to decrease the diameter of PS spheres to 200 nm (figure 1(B)). Through heating, the PS spheres collapsed on the silicon slide surface (figure 1(C)) and by the treatment of RIE the silicon pillar array was achieved (figure 1(D)). It should be noted that only the use of collapsed PS spheres as a mask can lead to a cylindrical silicon array under RIE. If the PS spheres without heating are used as a mask, the conical silicon array will be obtained. Figure 2(E) presents the highly ordered silicon pillar array after PS was removed by RIE. The picture indicates that the pillar array is still of hexagonal arrangement. Figure 1(F) reveals deposited Ag films of a thickness of 7 nm on the silicon pillar surface. It is observed that there exists a uniform Ag island film via the vacuum evaporation deposition. Figures 1(G) and (H) present the aggregates of Ag nanoparticles on the top of the silicon pillar and the flat silicon slide after the annealing process. It can be found that all Ag nanoparticles coalesced on the substrates or only on the top of the silicon pillars. Note that there are no Ag nanoparticles forming onto the sidewalls of silicon pillars. However, on the top of the silicon pillar, they formed more compact aggregates and rather loose aggregates on the unlimited surface. 3.2. Optimization of the thickness of deposited Ag
The different thickness Ag films were deposited on the silicon pillar arrays to investigate the morphology of the POP and optimize the performance of the SERS substrate. Figures 2(A)–(C) present the SEM images of three kinds of silicon pillar arrays decorated with different Ag nanoaggregates, which were produced by annealing 3, 7, and 8 nm thicknesses of Ag films at 200 °C for 2.0 h in nitrogen ambience. From figures 1(F) and (G), it is obvious that discrete Ag islands commenced to coalesce and finally Ag films were transformed into regularly-shaped and -sized nanoparticles. However, there are many differences between them. It is observed in figure 2(A) that the Ag nanoparticles on the silicon pillar display a large interspace and the average size is approximately 20 nm. In contrast, the Ag nanoparticles in figure 2(B) exhibit a relatively small gap and their average size approaches 60 nm, which is supposed to have a large contribution to the SERS enhancement [24]. In terms of figure 2(C), the mean size of the Ag islands approximately reaches 100 nm, constructing a bigger gap between the nanoparticles than the 7 nm Ag film, thus forming a smaller
Figure 2. SEM images of the POPs with different thicknesses of Ag
film after annealing (A) to (C); 3, 7, and 8 nm). They are in the same magnification and the scale represents 500 nm.
plasmonic properties. The parameters for recording the reflection spectra were as follows, integration time of 8 ms, scans to average of 50 and a boxcar width of 10. 2.3. SERS characterization
4-MPY as the probe molecule was used to evaluate the SERS activities of the prepared substrates. First, 4-MPY molecules (10 μL, 1.0 × 10−6 M) were adsorbed onto the construction of the POP by dripping. The diameter of the natural drying drop of 4-MPY was 4.32 mm. The SERS spectra of 4-MPY were recorded by a portable BWTEK Raman spectrometer with 532 nm excitation wavelength. The laser power was 25.4 mW, and the integration time was 5 s. Raman spectra under the excitation of 633 and 785 nm wavelengths were 4
Nanotechnology 25 (2014) 235301
D Jiang et al
Figure 3. (A) The SERS spectra characterize the SERS enhancement of the 4-MPY (1.0 × 10−6 M) under 633 nm wavelength excitation, with
the Ag films of different thicknesses after annealing. From bottom to top, the thicknesses of the Ag films are 3, 5, 7, 8, and 10 nm, respectively. As SERS signals of the POP with 7, 8 nm were considerably stronger compared to ones of the others, for the sake of visual effect, the corresponding SERS signals of 7 and 8 nm were divided by 4 and 2, respectively. (B) The SERS band intensity at 1092 cm−1 with the Ag film thickness when detecting 4-MPY.
number of nanoparticles due to the huge volume compared to figures 2(A) and (B). It is reported that the distance between the nanoparticles and the size of nanoparticles both highly affect the LSPR and in turn, the SERS activity [1]. To display the SERS of different annealed Ag films, we used 4-MPY to evaluate their enhancement ability. Figure 3(A) shows the SERS spectra of 4-MPY on the POPs with different thickness of annealed Ag films. From figure 3(A), we can observe the existence of characteristic peaks of 4-MPY at 1003, 1060, 1092, 1214, 1575, and 1606 cm−1. Also, several characteristic peaks of 4-MPY (e.g. 1575 and 1606 cm−1) on the POP with 3, 5, 10 nm-thickness annealed Ag films (a, b, e line) can be obscurely identified. The profile of the Raman intensity with the Ag film thickness is shown in figure 3(B), which indicates that the maximal SERS enhancement appeared with the 7 nm annealed Ag film. So we used the POP with the 7 nm annealed Ag film in the following experiment.
Figure 4. The reflection spectra of silicon pillar array and flat silicon substrate with and without Ag nanoparticles.
reflectance spectra of a periodic moth-eye silicon nanostructure, and found that this structure can be thought of as an effective medium that approximates a single layer thin film when the shape of silicon pillars is with near-vertical side walls and flat tops [25]. Similar to the moth-eye silicon nanostructure they used, in the present work, we assigned the multiple resonance modes in 482 and 650 nm in the reflection spectra of the silicon pillar array to the interference effect. After the 7 nm thickness Ag film was deposited on the flat silicon slide followed with annealing, there appears a broad band in the range of 450–740 nm, which is ascribed to the dipolar resonance of Ag nanoparticles from annealed Ag film. Compared with the reflection spectrum of the silicon pillar, the interference peaks at 482 and 650 nm slightly red shifted to 497 and 678 nm in POP. We attributed the redshifts to the light scattering of the formed Ag nanoparticles,
3.3. Comparison of SERS enhancement on a limited and an unlimited area 3.3.1. Reflection spectra.
The SERS activity of a substrate is strongly affected by its plasmonic properties. So we measured the reflection spectrum of the prepared POP substrates (figure 4) to reflect their plasmonic properties. For comparison, the reflection spectrum of the annealed Ag on a flat silicon slide was recorded. We also recorded the reflection spectra of the silicon pillar array and the flat silicon slide as a reference. Figure 4 shows the reflection spectra of the silicon pillar array and the flat silicon substrate with and without Ag nanoparticles. It is obviously observed that the spectra of the silicon slide show a high reflectivity in 400–800 nm range with no obvious peak appearing, while the configuration of the silicon pillar presents multiple resonance modes in 482 and 650 nm. Boden et al studied that the 5
Nanotechnology 25 (2014) 235301
D Jiang et al
−6 Figure 5. SERS spectra of 4-MPY (10 μL, 1.0 × 10 M) on the POP (a) and the flat silicon slide (b). They were both decorated by an annealed 7 nm thickness silver film. A normal Raman spectrum of 4MPY (10 μL, 0.1 M) on a flat silicon slide (c).
which can be proved by the SEM images of POP (figure 1(G)) and Ag particles on a flat slide (figure 1(H)) after annealing, in which Ag films have been transformed into regularlyshaped and -sized nanoparticles. In addition, the simulated reflection spectra (see figure S1 available in the supplementary data at stacks.iop.org/NANO/ 25/235301/mmedia) show the same trend. The reflection spectra make a reference to the choice of the SERS excitation wavelength. To exactly match the resonance band of the POP for the strongest SERS, the excitation wavelength of 633 nm was adopted.
Figure 6. (A) SERS spectra of 4-MPY at different concentrations on
the POPs measured with a 633 nm laser. (B) The plot of SERS intensities at 1092 cm−1 in (A) as the 4-MPY concentration.
silicon substrate modified by the annealed 7 nm thickness Ag film, respectively. The SERS determination of 4-MPY was presented in figure 6(A), and the profile of the SERS intensities with the 4MPY concentration was given by figure 6(B). We can discern that the lowest detection concentration of 4-MPY reaches 1.0 × 10−10 M.
3.3.2. SERS behavior. Figure 5 shows the SERS spectra of
4-MPY on three different substrates under the excitation with 633 nm. The SERS spectra under 532 and 785 nm excitation were provided in the supplementary information figure S3. Three substrates, respectively, are an unvarnished silicon slide (curve c), a silicon slide whose surface is decorated by an annealed 7 nm thickness silver film (curve b), and a POP substrate with an annealed 7 nm thickness silver film (curve a). Obviously, curve a in figure 5 supports the strongest SERS signal. The SERS signals on the POP are considerably stronger than the flat silicon substrate because the limited 2D area of the silicon pillar top decreases the distance between Ag nanoparticles and regulates the size and morphology of the nanoparticles, which facilitates LSP coupling and EM field enhancement, and the silicon pillar array also displays some influence to SERS enhancement which is evidenced by the FDTD simulation in the supplementary information. The EFs were calculated by comparing curve a (or b) to c (in figure 5). The details of the EFs calculation are also provided in the supplementary information. EFs for the 1092 cm−1 band are 1.5 × 106 and 2.4 × 105 for the POP and the flat
4. Conclusions In this work, we explored an approach to fabricate a hot spot SERS substrate by depositing Ag film onto the limited 2D area under annealing. This method is beneficial to the aggregation between nanoparticles and the improvement of the SERS signal. By observing the effect of the deposited silver films of different thicknesses on SERS enhancement, it is founded that the thickness of 7 nm is best suited. In addition, the comparative research between POP and the flat silicon substrate demonstrates that POP is preferable in SERS enhancement. Moreover, our experiments support these 6
Nanotechnology 25 (2014) 235301
D Jiang et al
[4] Moskovits M 1985 Rev. Mod. Phys. 57 783 [5] Giannini V, Fernández-Domínguez A I, Heck S C and Maier S A 2011 Chem. Rev. 111 3888 [6] Halas N J, Lal S, Chang W, Link S and Nordlander P 2011 Chem. Rev. 111 3913 [7] Xu H, Aizpurua J, Kall M and Apell P 2000 Phys. Rev. E 62 4318 [8] Guo H, Xu W, Zhou J, Xu S and Lombardi J R 2013 Nanotechnology 24 045608 [9] Bhuvana T and Kulkarni G U 2008 Small 4 670 [10] Cialla D, Hübner U, Schneidewind H, Möller R and Popp J 2008 ChemPhysChem 9 758 [11] Jeon H C, Heo C, Lee S Y and Yang S 2012 Adv. Funct. Mater. 22 4268 [12] Alexander K D, Hampton M J, Zhang S, Dhawan A, Xu H and Lopeza R 2009 J. Raman Spectrosc. 40 2171 [13] Chen A, Miller R L, DePrince A E, Joshi-Imre A, Shevchenko E, Ocola L E, Gray S K, Welp U and Vlasko-Vlasov V K 2013 Small 9 1939 [14] Xu L, Tan L S and Hong M H 2011 Appl. Opt. 50 G74 [15] Upender G, Sathyavathi R, Raju B, Bansal C and Rao D N 2012 J. Mol. Struct. 1012 56 [16] Giorgis F, Descrovi E, Chiodoni A, Froner E, Scarpa M, Venturello A and Geobaldo F 2008 Appl. Surf. Sci. 254 7494 [17] Van Duyne R P, Hulteen J C and Treichel D A 1993 J. Chem. Phys. 99 2101 [18] Li X, Zhang Y, Shen Z X and Fan H J 2012 Small 8 2548 [19] Wang X, Xu S, Cong M, Li H, Gu Y and Xu W 2012 Small 8 972 [20] Wang C, Ruan W, Ji N, Ji W, Lv S, Zhao C and Zhao B 2010 J. Phys. Chem. C 114 2886 [21] Li H, Xu S, Gu Y, Wang H, Ma R, Lombardi J R and Xu W 2013 J. Phys. Chem. C 117 19154 [22] Li Y, Cai W and Duan G 2008 Chem. Mater. 20 615 [23] Li Y et al 2009 Adv. Mater. 21 4731 [24] Emory S R, Haskins W E and Nie S 1998 J. Am. Chem. Soc. 120 8009 [25] Boden S A and Bagnall D M 2008 Appl. Phys. Lett. 93 133108
conclusions and discover the lowest detection concentration of 4-MPY on the POP up to 1.0 × 10−10 M. Our research provides a simple approach to constructing a SERS substrate and the obtained POP configuration achieves a highly sensitive SERS measurement, which proves that the limited 2D area plays a dominating role in the formation of hot spots. Compared with other similar researches, the POP structure may present comparable SERS enhancement ability. Whereas, the silicon –based SERS substrate itself has an interior standard peak at 520 cm−1, which is very vital to SERS detection by decreasing detection error. More importantly, the POP substrate can be provided in wafer-scale, which is difficult for other micro fabrication technology, such as nanopatterning, electron and ion beam lithography etc. and it is promising for becoming a commercial SERS substrate.
Acknowledgments This work was supported by the National Natural Science Foundation of China NSFC Grant (Nos 21373096, 21073073, and 91027010) and National Instrumentation Program (NIP) of the Ministry of Science and Technology of China No. 2011YQ03012408 and Innovation Program of the State Laboratory of Supramolecular Structure and Meterials.
References [1] Guerrini L and Graham D 2012 Chem. Soc. Rev. 41 7085 [2] Sebastian S 2009 Chem. Phys. Chem. 10 1344 [3] Lombardi J R, Birke R L, Lu T and Xu J 1986 J. Chem. Phys. 84 4174
7
