Analyst View Article Online
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
PAPER
View Journal | View Issue
Cite this: Analyst, 2014, 139, 2207
DNA-embedded Au–Ag core–shell nanoparticles assembled on silicon slides as a reliable SERS substrate† Zhong Zhang, Sha Zhang and Mengshi Lin* This study aimed at developing a sensitive and reliable SERS substrate by assembling DNA-embedded Au–Ag core–shell nanoparticles (NPs) on silicon slides. First, a monolayer of well separated DNA-functionalized Au NPs (40 nm) was decorated on (3-aminopropyl)triethoxysilane modified silicon slides. The DNA-embedded Au–Ag core–shell NPs were assembled on the 40 nm Au–DNA NPs to form a core–satellite structure through DNA hybridization. Using 4-MBA as a Raman dye, the SERS performance of the substrates was evaluated after being cleaned by low oxygen and argon plasma. The Raman intensity of the assembly using DNA-embedded Au–Ag core–shell NPs was 8–10 times higher than the intensity of the assembly using Au NPs as satellites. In addition, the signal-to-noise ratio of the assembly was 2.6 times higher than that of a commercial substrate (Klarite™) when a 785 nm laser was used. The SERS enhancements of the assembled substrates were 2.2 to 2.8 times higher than the Klarite when an acquisition time of 5 s was used at an excitation wavelength of 633 nm. The assembled
Received 13th November 2013 Accepted 9th February 2014
substrates also show a good spot-to-spot and substrate-to-substrate reproducibility at the excitation wavelengths of 633 and 785 nm. These results demonstrate that the fabrication process is simple and
DOI: 10.1039/c3an02116e
cost-effective for assembling DNA-embedded Au–Ag core–shell NPs on silicon slides that can be used
www.rsc.org/analyst
as a reliable SERS substrate.
1. Introduction Surface enhanced Raman scattering (SERS), a technique which can dramatically enhance the Raman signals of analyte molecules adsorbed on the metallic nanostructures, has been applied in chemical analysis, detection of food contaminants, biological sensing, and environmental monitoring.1–5 For example, SERS has been used to detect melamine, pesticides, and pathogens in food using commercial gold substrates.6–9 SERS was employed to analyze the endocrine disruptor using gold- and silver-decorated microspheres as substrates.10 SERS was also developed for detection of tumors using labeled gold nanoparticles (NPs) or gold-patterned microarrays as substrates.11,12 A variety of substrates have been developed in recent years for SERS applications. The substrates fabricated by nanolithography show promising SERS enhancement and great reproducibility.13–15 Sub-10 nm metallic nanogap arrays with precise control of the gap morphology have been produced by the nanolithography technique.16 However, it is still a challenge to fabricate smaller nanogaps (1–2 nm) by nanolithography.
Division of Food Systems & Bioengineering, University of Missouri, Columbia, MO, 65211-5160, USA. E-mail: [email protected]; Fax: +1-573-884-7964; Tel: +1-573884-6718 † Electronic supplementary 10.1039/c3an02116e
information
(ESI)
This journal is © The Royal Society of Chemistry 2014
available.
See
DOI:
Nanolithography also requires specic equipment for the development of nano-patterns and the deposition of a Au or Ag layer. Another lithography method, called nanosphere lithography, has been adopted by many researchers because it is inexpensive and able to create a large area of ordered nanostructures.17–19 However, nanosphere lithography still needs to pattern the substrate surface by a layer of nanospheres in advance, which is difficult to be tailored for different SERS detection strategies. In addition, SERS substrates can also be fabricated by assembling Au or Ag NPs on the glass, silicon surface, or liquid/liquid interface.20–25 Among all the fabrication methods, the self-assembling approach is the most convenient and cost-effective method because it can be performed without using expensive equipment. Between the junctions of the self-assembled NPs, numerous hot spots for SERS can be created where an intense electromagnetic eld is generated at this area. However, a problem of self-assembled substrates is that the distance between NPs is difficult to control and may vary in different locations on the same substrate. This could inuence the reproducibility of the substrate and also cause big variations in Raman signals in different batches. To solve this problem, DNA has been used to direct and organize the assembly of Au NPs on the silicon or glass surface modied with (3-aminopropyl)triethoxysilane (APTES).26 The DNA-directed assembly provides an excellent control of the distance between Au NPs using rigid
Analyst, 2014, 139, 2207–2213 | 2207
View Article Online
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
Analyst
DNA molecules and offers a good reproducibility for different batches, making it an ideal method for preparing SERS substrates. The single molecule sensitivity of SERS has been achieved by placing the Raman dye into the junction of a DNA-tethered dimer of Au NPs and coating a thin layer of Ag shell on the dimer.27,28 The coating of the Ag shell signicantly enhances the Raman signals of the dye molecules because Ag induces a higher enhancement than Au and the gap size between two NPs was reduced to less than 1.5 nm. It was proved that the hybridized DNA was embedded in the Ag shell during its growth.29,30 Inspired by the DNA-directed assembly of Au NPs and the signicant enhancement by the Ag shell, we direct the assembly of ssDNA embedded Au–Ag core–shell NPs on the silicon surface to form a SERS-active core–satellite structure. As shown in Fig. 1, the silicon surface was functionalized by a layer of APTES, which was used to adsorb negatively charged ssDNAcoated Au NPs by the positive amino group of APTES. Thus, a monolayer of Au–DNA NPs was decorated on the silicon surface through this electrostatic interaction.31–35 Using Au–DNA NPs as the core, the ssDNA embedded Au–Ag core–shell NPs will be assembled as satellites around the core NPs by DNA hybridization.36–38 In this structure, the gap distance between the core NPs and the satellite NPs was smaller due to the presence of the Ag shell as compared to core–satellite structures constructed only by Au NPs. Meanwhile, the Au–Ag core–shell NPs provide a much better SERS enhancement than Au NPs.27,28 Furthermore,
Paper
the gap distances between the core NPs and satellite NPs were almost identical because they were separated by the rigid DNA chains with the same length, giving a high reproducibility for different core–satellite structures. This structure could be used as a reliable SERS substrate with high enhancement and high reproducibility.
2. 2.1
2208 | Analyst, 2014, 139, 2207–2213
Materials and chemicals
Hydrogen tetrachloroaurate solution (HAuCl4, 30 wt% in dilute HCl), silver nitrate, sodium citrate dihydrate (>99%), L-ascorbic acid, 4-mercaptobenzoic acid, DL-dithiothreitol (DTT), (3-aminopropyl)triethoxysilane (APTES), and sodium dodecyl sulfate (SDS) were purchased from Sigma-Aldrich (St Louis, MO, USA) and used without any purication. Mono-thiolated ssDNA was purchased from Integrated DNA Technologies (IDT, Iowa, USA). All the ultrapure water (18.2 MU cm1) was prepared from a Millipore water purication system. 2.2
Modication of silicon substrates by APTES
The silicon slides (3 mm 4 mm) were rst cleaned and oxidized by piranha solution (3 : 1 H2SO4 : H2O2) for 60 min with gentle shaking.35,39 The silicon slides were then thoroughly rinsed with pure water and dried at room temperature. The modication of the silicon surface with APTES was performed by immersing the slides into a mixture of APTES, water, and ethanol (3 : 3 : 94) for 3 h with shaking. Aer modication, the slides were rinsed with copious amounts of water and baked at 105 C for 30 min. 2.3
Fig. 1 Schematic illustration of the fabrication of core–satellite structures as SERS substrates: (a) silicon slide was modified by APTES to form a layer of the NH2 group; (b) ssDNA-modified 40 nm Au NPs were adsorbed on the silicon surface; (c) ssDNA-embedded Au–Ag core–shell NPs were conjugated to 40 nm Au NPs through the hybridization of complementary ssDNA.
Materials and methods
Modication of Au NPs by thiolated ssDNA
Gold NPs with average diameters of 17 and 40 nm were synthesized by the reduction of HAuCl4 using sodium citrate.40–42 The procedure of DNA loading was similar to a previous report.43 Briey, thiolated ssDNA (Link-1 or Link-2) was rst reduced by 0.1 M DTT for an hour and puried by the NAP-5 column using ultrapure water as an eluent. The puried ssDNA (0.33 mg mL1, 1 mL) was added into 10 mL of Au NPs with 0.01% SDS and 1 mL of 0.01 M phosphate buffer. The mixture of ssDNA and Au NPs was subjected to ultrasonication for 30 s and then incubated for 20 min with shaking. Aer incubation, 0.3 mL of 2.0 M NaCl in 0.01 M phosphate buffer (pH ¼ 8.0) was added to the mixture. The mixture was then subjected to 30 s sonication and 20 min incubation. This process was repeated aer each addition of 0.3, 0.6, 0.65, 0.7, 0.77, and 0.80 mL of 2.0 M NaCl. Finally, the mixture was incubated overnight for the full adsorption of ssDNA on Au NPs. In this study, the ssDNA (Link-1) with a sequence of TTA TAA CTA TTC CTA AAA AAA AAA A C(6)SH was loaded on 17 nm Au NPs. The complementary ssDNA (Link-2) with a sequence of TAG GAA TAG TTA TAA AAA AAA AAA A C(6)SH was loaded on 40 nm Au NPs. 2.4
Fabrication of DNA embedded Au–Ag core–shell NPs
The modied Au NPs (17 nm, 1 mL) were centrifuged twice to remove excess Link-2 ssDNA and re-dispersed in 0.90 mL of
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
Paper
ultrapure water in a 1.5 mL Eppendorf (EP) tube. Then 0.1 mL of phosphate buffer (0.18 M, pH ¼ 8.0) and 0.1 mL of NaCl (2.0 M) were added to adjust the pH and ionic strength of the Au NP solution. This step is very important for successful Ag coating because high ionic strength reduces the repulsion between the DNA chains and Ag atoms, facilitating the nucleation of Ag on the surface of Au NPs. Ascorbic acid (0.1 M, 150 mL) and AgNO3 (1 mM, 0–150 mL) were mixed with the NP solution. The tube was subsequently incubated in a 45 C water bath for 60 min to accelerate the growth of the Ag shell. In the meantime, the enhanced temperature also unfolded the DNA and reduced the adsorption of DNA base on the Au surface for the purpose of keeping the active part of DNA out of the Ag shell. Aer the growth of the Ag shell, the excess ascorbic acid was removed by double centrifugation. The Au–Ag core–shell NPs were nally dispersed in 300 mL phosphate buffer (10 mM, pH ¼ 7.45) with 0.5 M NaCl. 2.5 Assembling DNA embedded Au–Ag core–shell NPs into a core–satellite structure on silicon The DNA-modied 40 nm Au NPs (1 mL) were puried by three consecutive centrifugations and nally dispersed in 400 mL water. The APTES-modied silicon slide was immersed into the NP solution and kept in the solution overnight. Aer that, a monolayer of 40 nm Au NPs was formed on the silicon surface by the electrostatic interaction of the Au and NH2 groups. The silicon slide was rinsed with water three times to remove weakly bound NPs and then incubated in phosphate buffer solution (10 mM, pH ¼ 7.45) with 2 mM sulfosuccinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate (Sulfo-SMCC) at 45 C for 2 h to block the excess NH2 groups on the silicon surface.26 Aer cleaning by water, the silicon slides were immersed into an EP tube containing 100 mL of DNA-embedded Au–Ag core–shell NPs. The EP tube was then heated in a 65 C water bath for 5 min and le overnight at room temperature for the hybridization of DNA. Finally, the silicon slides were rinsed with pure water for three times and dried at room temperature. 2.6
Analyst
2.7
Characterization by TEM and SEM
TEM (Jeol 1400, JEOL, Tokyo, Japan) was used to characterize the 40 nm Au NPs and the shell thickness of the Au–Ag core– shell NPs under an accelerating voltage of 120 kV. Before the SEM characterization, the substrates were attached on the aluminum stub by double-sided adhesive carbon tape. The SEM images of the substrates were obtained using a FEI Quanta 600 FEG Scanning Electron Microscope (FEI, Oregon, USA).
3. 3.1
Results & discussion Decoration of 40 nm Au–DNA NPs on the silicon surface
Au NPs with an average diameter of 40 nm were fabricated by the citrate reduction method. As shown in Fig. 2a, Au NPs are monodispersed with uniform size distribution with spherical/ oval shapes. Aer modication with ssDNA, Au NPs were used to decorate the surface of APTES-functionalized silicon. A very uniform monolayer of Au–DNA NPs was attached and dispersed on the silicon surface (Fig. 2b). Only a few clusters of dimers, trimers and tetramers of NPs could be observed in the SEM image. These clusters have little inuence on the reproducibility of the SERS substrate because it is estimated that 97% of the Au–DNA NPs are well separated from each other due to the strong repulsion of the negatively charged DNA coating (Fig. 2c). According to the scheme, the 40 nm Au–DNA NPs were used to attract and direct the assembly of the Au–DNA–Ag NPs through the interaction of the complementary ssDNA. Therefore, the number of core–satellite structures formed on per mm2 silicon will be determined by the number of Au–DNA NPs in this area. The Raman spot size was estimated to be 1 mm2 when an excitation wavelength of 785 nm and a 50 objective were used. The number of Au NPs per mm2 was counted for ten randomly
SERS measurement
The as-prepared substrates were cleaned by plasma treatment under 4 min of oxygen/argon ow (5 : 95) and 2 min of argon ow at a power of 50 mW and a gas ow rate of 30 sccm. Aer cleaning, the silicon were immersed into 400 mL ethanol solution of 4-MBA (106 M) for 3 h. A commercial substrate, Klarite™, was used as control and also kept in the same concentration of 4-MBA solution for 3 h. The substrates were rinsed with 400 mL water and air dried at room temperature. A 785 nm diode laser and a 633 nm He–Ne laser were used to evaluate the SERS performance of the substrates. SERS spectra were recorded using a Renishaw 1000 micro-Raman system at a wavelength of 785 nm using an acquisition time of 10 s and a laser power of 30 mW using a 50 objective. Raman data were also collected using a Horiba LabRam-HR 800 micro-Raman spectrometer at an excitation wavelength of 633 nm using a laser power of 4.75 mW and acquisition time of 1 s and 5 s. All the SERS data were recorded from at least ve randomly selected spots on the substrate. This journal is © The Royal Society of Chemistry 2014
Characterization of ssDNA-modified 40 nm Au NPs: (a) TEM image of pure Au NPs with an average size of 40 nm; (b) and (c) SEM images of a monolayer of Au NPs adsorbed on the APTES modified silicon surface at different magnifications; (d) average number of Au NPs per mm2 at ten randomly selected locations.
Fig. 2
Analyst, 2014, 139, 2207–2213 | 2209
View Article Online
Analyst
Paper
selected areas in the SEM image (Fig. 2b). It can be observed that the number per mm2 was narrowly distributed between 47 and 55 (Fig. 2d). This is crucial for a good reproducibility in SERS because the narrow distribution will reduce the signal uctuation from spot to spot and contribute to a good reproducibility for SERS.
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
3.2 Assembly of DNA-embedded Au–Ag core–shell NPs on Au–DNA NP decorated silicon The 17 nm Au NPs were rst functionalized by ssDNA (Link-1) which is complementary to the ssDNA (Link-2) used for 40 nm Au NPs. The DNA-embedded Au–Ag core–shell NPs were synthesized by coating a thin layer of the Ag shell on the 17 nm Au–DNA NPs. Ag could be easily reduced by ascorbic acid using Au NPs as seeds.44 However, the reduction of Ag on Au–DNA NPs is different from the deposition of Ag on citrate stabilized Au NPs at room temperature because the growth of the Ag shell will be hindered by dense DNA chains on the Au surface while it will not be inuenced by citrate.29,45,46 Thus, an appropriate NaCl concentration and an enhanced temperature should be maintained during the growth of the Ag shell. The Ag shell was successfully formed on the 17 nm Au–DNA NPs which was proved by the color change of NP solution from pink to yellow. Otherwise, the color of the 17 nm Au–DNA NP solution would not change from pink to yellow aer the addition of AgNO3 and ascorbic acid if NaCl was not used to enhance the ionic strength of the mixture. The thickness of the Ag shell has to be carefully controlled to avoid burying the whole ssDNA chain in the shell. It is obvious that the Au–DNA–Ag NPs would not be able to link with the 40 nm Au–DNA NPs on the silicon surface if the ssDNA molecule was hidden under the Ag shell. The efficiency of assembling will also be affected if some of complementary bases on the DNA chain were embedded in the Ag shell. In this study, the thickness of the Ag shell was controlled by adding different amounts of 1 mM AgNO3 (50, 100, and 150 mL) while maintaining excess ascorbic acid in the solution. The Au–DNA–Ag NPs with different thicknesses were then incubated with the 40 nm Au–DNA NP decorated silicon slide in the 0.5 M NaCl solution. The excess NH2 groups on the silicon surface were blocked by Sulfo-SMCC before the assembling process. Aer assembling at room temperature for 12 h, the 40 nm Au–DNA NPs were linked with Au–DNA–Ag NPs and surrounded by these NPs via DNA interaction (Fig. 3b, d and f). The average numbers of satellite NPs assembled on each 40 nm Au–DNA NP were 6.8 1.4, 5.3 1.6, and 4.9 1.6 respectively when 50, 100, and 150 mL of AgNO3 were used to grow the Ag shell (Fig. 3a, c and e). The corresponding thickness of the Ag shell was around 1.1, 2.1, and 3.0 nm measured according to the TEM images (Fig. S1†). The Au– DNA–Ag NPs were still able to link to the 40 nm Au–DNA NPs because the length of the DNA spacer used in this study was 4 nm.44 There are two possible reasons for the decreased number of satellite NPs assembled on each 40 nm Au–DNA NP. First, more ssDNA molecules were buried in the Ag shell
2210 | Analyst, 2014, 139, 2207–2213
Characterization of the core–satellite structure assembled by ssDNA-embedded Au–Ag core–shell NPs with different Ag thicknesses by SEM and TEM (embedded): (a) and (b) 50 mL of AgNO3 (1 mM) coating; (c) and (d) 100 mL of AgNO3 (1 mM) coating; (e) and (f) 150 mL of AgNO3 (1 mM) coating. Fig. 3
when the Ag shell became thicker. In other words, the DNA molecules tended to lose their active bases as the thickness of the Ag shell increased. Furthermore, the assembly of Au–DNA NPs and Au–DNA–Ag into the core–satellite structure on silicon is different from the assembly of both NPs in solution. The steric hindrance is larger for the Au–DNA NPs immobilized on the silicon surface compared to that of Au–DNA in solution. The steric hindrance was also enhanced due to the increased size of satellite NPs, reducing the efficiency of DNA hybridization. Nevertheless, the average number of assembled satellite NPs was similar for the NPs with 100 and 150 mL of AgNO3 coating. It is observed in Fig. 3c and e that most of the 40 nm Au–DNA NPs were still uniformly surrounded by Au–DNA–Ag NPs. In this case, 150 mL of AgNO3 solution was chosen to build Au–DNA–Ag NPs due to the thicker Ag shell, which could lead to a smaller gap distance between the 40 nm core and the DNA-embedded Au–Ag NPs. As a comparison, 27 nm Au NPs have also been used as satellite NPs to assemble the core–satellite structure. It was found that 27 nm Au NPs impose a much higher steric hindrance than 17 nm Au NPs, which makes it unable to uniformly attach on the core NPs and form perfect core–satellite structures on silicon (Fig. S1f†). The Au NPs with a diameter of 40 nm were selected as core nanoparticles because 40 nm Au NPs could be conveniently synthesized by a classic citrate reduction method while it is necessary to use the seed-mediated method to grow nanoparticles larger than 40 nm.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
3.3 Comparison of the SERS enhancement between the assemblies using 17 nm Au–DNA NPs and Au–DNA–Ag NPs as satellites The DNA molecules on the surface of NPs should be removed before the SERS measurements because the DNA coating would hinder the adsorption of the Raman dye on the substrate. Unlike the assembly of pure Au NPs, the Ag shell of DNA-embedded core–shell NPs could be easily oxidized by the UV–ozone treatment. To solve this problem, low oxygen plasma treatment was used in this study. The plasma treatment would not cause any substantial change in the core–satellite structure as evidenced by the TEM images (Fig. S2†). However, the SERS enhancement could be improved by the plasma treatment under suitable conditions. Fig. 4 shows the Raman spectra of 4-MBA on the substrates which were subjected to different plasma treatments. The Raman peak at 1078 cm1, a typical band of 4-MBA, was found to be the highest for the substrate treated by 5% oxygen plasma. The Raman intensity is much lower for the substrate without any treatment. It was also found that there are no visible peaks of 4-MBA for the substrates treated by 25% oxygen plasma due to the oxidation of the Ag layer. Fig. 5 shows the Raman spectra of 4-MBA adsorbed on the core–satellite assemblies using 17 nm Au–DNA NPs or Au–DNA–Ag NPs as satellite NPs. It is clearly observed that the Raman peak at 1078 cm1 is much higher for the assembly using Au–DNA–Ag NPs as satellites compared to that using Au–DNA NPs as satellites. The Raman peaks of 4-MBA from 1100 to 1300 cm1 were barely recognizable for the assembly using Au–DNA NPs as satellite NPs. It is estimated that enhancements by 8–10 times were induced by coating the satellite NPs with a thin layer of Ag shell. This enhancement could be attributed to the better SERS response of Ag, the reduction of gap size between the core and satellites, and the larger size of satellite NPs aer the growth of the Ag shell.47 The assembly using DNA-embedded Au–Ag core–shell NPs as satellites shows a much more promising SERS enhancement than the assembly using Au–DNA NPs as satellites.
Fig. 5 Comparison of SERS intensity of 4-MBA (106 M) using different NPs as satellites: top spectrum: ssDNA embedded Au–Ag core–shell NPs as satellites; bottom spectrum: ssDNA-modified Au NPs as satellites.
3.4 Reproducibility of the assembly and comparison with a commercial substrate Fig. 6a shows the Raman spectra of 4-MBA adsorbed on the assembled substrates and a commercial substrate (Klarite) when excited by a 785 nm laser. It is found that the intensities of peaks at 1078 cm1 and 1585 cm1 were comparable for both substrates. But the signal-to-noise ratio of the assembly (47.3) was 2.6 times higher than that of Klarite (18.4), showing a better performance of the assembly. Fig. 6b and c show the Raman intensities of peaks at 1078 and 1585 cm1 from three different substrates and the Klarite. It can be found that there is no signicant difference for the Raman intensity of the three assembled substrates, indicating a good reproducibility of the substrates (Fig. S3a†). The variation of Raman intensity at
Comparison between the assembled substrates and a commercial substrate using a 785 nm laser: (a) Raman spectra of 4-MBA (106 M) on the assembly and the Klarite; (b) Raman intensity of 4-MBA at 1078 cm1 from three assembled substrates and the Klarite; (c) Raman intensity of 4-MBA at 1585 cm1 from three assembled substrates and the Klarite. Fig. 6
Fig. 4 Raman spectra of 4-MBA (106 M) from the assembled substrates treated under different plasma conditions.
This journal is © The Royal Society of Chemistry 2014
Analyst, 2014, 139, 2207–2213 | 2211
View Article Online
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
Analyst
different locations on the same substrate has also been evaluated and shown as error bars of the peak intensities at 1078 and 1585 cm1. The results show that the spot-to-spot variation of the three assembled substrates is close to that of the Klarite produced by the lithography method, suggesting a uniform distribution of the core–satellite structures on the silicon surface. A drawback of the substrate is that the Ag shell on the satellite NPs can be easily oxidized in air. The SERS performance of the substrates was largely reduced aer 4 month storage at room temperature (Fig. S3b†). Therefore, the substrate should be used right aer the preparation or stored in an oxygen-free atmosphere for long term stability. The SERS performance of the assembled substrates was evaluated and compared with the Klarite by measuring 4-MBA with a 633 nm laser. It was found that the 633 nm laser induced much higher enhancement than the 785 nm laser. For the assembled core–satellite structure by pure Au NPs, the surface plasmon resonance has been found to be at 650 nm.26 However, in this study, the surface plasmon resonance of the core–satellite structure blue-shied because of the Ag shell on the satellite nanoparticles, leading the surface plasmon resonance closer to 633 nm. Therefore, the 633 nm laser resulted in a higher enhancement than the 785 nm laser. As shown in Fig. 7a, the Raman intensity of the band at 1078 and 1585 cm1 from the assembled substrates was 2.2 and 2.8 times higher than the Raman intensity from the Klarite when 25% of laser power and 5 s of acquisition time were used. Meanwhile, the original Raman spectra from the Klarite have a much higher baseline dri than the assembled substrates (Fig. S4†). The high baseline dri causes problems for the acquisition of Raman data when a higher laser power and a longer acquisition time
Fig. 7 Comparison of the assembled substrates and a commercial substrate using a 633 He–Ne laser: (a) Raman spectra of 4-MBA (106 M) on the assembly and the Klarite using an acquisition time of 5 s; (b) Raman intensity of 4-MBA at 1078 cm1 acquired from three assembled substrates and the Klarite by a 1 s acquisition time; (c) Raman intensity of 4-MBA at 1585 cm1 acquired from three assembled substrates and the Klarite by a 1 s acquisition time.
2212 | Analyst, 2014, 139, 2207–2213
Paper
are needed to increase the Raman signals. The Raman spectra of 4-MBA from the Klarite exceeded the detection limit when the acquisition time was extended to 10 s due to the baseline dri (Fig. S5†). The reproducibility of the substrates was also evaluated using a shorter acquisition time (1 s) at an excitation wavelength of 633 nm. The Raman intensities of the band at 1078 and 1585 cm1 were very consistent for all of the tested substrates, indicating a good substrate-to-substrate reproducibility. The Raman intensities of the peak at 1078 and 1585 cm1 for the assembled substrates were 1.25 times and 2.1 times higher than the intensity of spectra acquired using the Klarite when 1 s of acquisition time was applied, demonstrating a better SERS performance of assembled substrates. More importantly, the cost to fabricate this substrate is much lower than the commercial substrate, which is estimated to be around 5.22 US dollars for each substrate. Meanwhile, the synthesis of the substrate does not require pattern development, metal deposition, ion milling, or other complex steps in the lithography method.
4. Conclusions In this study, DNA embedded Au–Ag core–shell NPs were assembled with 40 nm Au–DNA NPs on the surface of silicon slides through the hybridization of DNA attached on the NPs. The monolayer of core–satellite structures was uniformly distributed on the silicon surface using well separated 40 nm Au–DNA NPs as cores. Around 4.9 1.6 Au–DNA–Ag NPs were assembled on each 40 nm Au core when the thickness of the Ag shell was 3.1 nm. In addition, the SERS performance of the substrates was evaluated aer low oxygen and argon plasma cleaning. The Raman intensity of 4-MBA is 8–10 times higher for the assembly using DNA Au–DNA–Ag NPs as satellites compared to that of the assembly using Au–DNA NPs without Ag coating as satellites. It is found that the Raman enhancement of substrates was similar to a commercial substrate when a 785 nm laser was used. But the signal-noise ratio of the assembly was 2.6 times higher than that of a commercial substrate. The SERS enhancements of the assembled substrates were 2.2 to 2.8 times higher than a commercial substrate when 5 s of acquisition time was used at an excitation wavelength of 633 nm. The assembled substrates also show a good spot-tospot and substrate-to-substrate reproducibility at the excitation wavelengths of 633 nm and 785 nm. The assembled substrates show great potential for SERS applications due to their high enhancement, good reproducibility, low cost, and simple fabrication. Future study is needed to reduce the steric hindrance for assembling bigger Au–DNA–Ag NPs and to further improve the SERS enhancement.
Acknowledgements We thank Dr Richard K. Brow and Jaime George at Missouri University of Science and Technology for their assistance in Raman measurement and the Electron Microscopy Center at the University of Missouri for TEM and SEM characterization.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
This research was supported by the USDA NIFA Nanotechnology Program project #2011-67021-30391.
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
References 1 M. Fleischmann, P. Hendra and A. McQuillan, Chem. Phys. Lett., 1974, 26, 163–166. 2 R. P. Cooney, E. S. Reid, M. Fleischmann and P. J. Hendra, J. Chem. Soc., Faraday Trans. 1, 1977, 73, 1691–1698. 3 B. Liu, G. Han, Z. Zhang, R. Liu, C. Jiang, S. Wang and M. Y. Han, Anal. Chem., 2012, 84, 255–261. 4 X. Qian, X. H. Peng, D. O. Ansari, Q. Yin-Goen, G. Z. Chen, D. M. Shin, L. Yang, A. N. Young, M. D. Wang and S. Nie, Nat. Biotechnol., 2008, 26, 83–90. 5 W. Ren, C. Zhu and E. Wang, Nanoscale, 2012, 4, 5902–5909. 6 M. Lin, L. He, J. Awika, L. Yang, D. Ledoux, H. Li and A. Mustapha, J. Food Sci., 2008, 73, T129–T134. 7 B. Liu, P. Zhou, X. Liu, X. Sun, H. Li and M. Lin, Food Bioprocess Technol., 2013, 6, 710–718. 8 C. Fan, Z. Hu, L. K. Riley, G. A. Purdy, A. Mustapha and M. Lin, J. Food Sci., 2010, 75, M302–M307. 9 T. A. Alexander and D. M. Le, Appl. Opt., 2007, 46, 3878–3890. 10 C. Fang, N. M. Bandaru, A. V. Ellis and N. H. Voelcker, Biosens. Bioelectron., 2013, 42, 632–639. 11 M. Lee, S. Lee, J.-h. Lee, H.-w. Lim, G. H. Seong, E. K. Lee, S.-I. Chang, C. H. Oh and J. Choo, Biosens. Bioelectron., 2011, 26, 2135–2141. 12 S. Lee, S. Kim, J. Choo, S. Y. Shin, Y. H. Lee, H. Y. Choi, S. Ha, K. Kang and C. H. Oh, Anal. Chem., 2007, 79, 916–922. 13 R. Alvarez-Puebla, B. Cui, J.-P. Bravo-Vasquez, T. Veres and H. Fenniri, J. Phys. Chem. C, 2007, 111, 6720–6723. 14 N. A. Abu Hatab, J. M. Oran and M. J. Sepaniak, ACS Nano, 2008, 2, 377–385. 15 J. Theiss, P. Pavaskar, P. M. Echternach, R. E. Muller and S. B. Cronin, Nano Lett., 2010, 10, 2749–2754. 16 H. Im, K. C. Bantz, N. C. Lindquist, C. L. Haynes and S.-H. Oh, Nano Lett., 2010, 10, 2231–2236. 17 G. H. Chan, J. Zhao, E. M. Hicks, G. C. Schatz and R. P. Van Duyne, Nano Lett., 2007, 7, 1947–1952. 18 H. Im, K. C. Bantz, S. H. Lee, T. W. Johnson, C. L. Haynes and S. H. Oh, Adv. Mater., 2013, 25, 2678–2685. 19 X. Zhang, J. Zhao, A. V. Whitney, J. W. Elam and R. P. Van Duyne, J. Am. Chem. Soc., 2006, 128, 10304–10309. 20 Q. Su, X. Ma, J. Dong, C. Jiang and W. Qian, ACS Appl. Mater. Interfaces, 2011, 3, 1873–1879. 21 M. P. Cecchini, V. A. Turek, J. Paget, A. A. Kornyshev and J. B. Edel, Nat. Mater., 2012, 12, 165–171. 22 R. A. Alvarez-Puebla, A. Agarwal, P. Manna, B. P. Khanal, P. Aldeanueva-Potel, E. Carb´ o-Argibay, N. Pazos-P´ erez, L. Vigderman, E. R. Zubarev and N. A. Kotov, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 8157–8161. 23 B. Peng, G. Li, D. Li, S. Dodson, Q. Zhang, J. Zhang, Y. H. Lee, H. V. Demir, X. Yi Ling and Q. Xiong, ACS Nano, 2013, 7, 5993–6000.
This journal is © The Royal Society of Chemistry 2014
Analyst
24 L. Tong, T. Zhu and Z. Liu, Chem. Soc. Rev., 2011, 40, 1296– 1304. 25 A. Chen, A. E. DePrince, A. Demorti` ere, A. Joshi-Imre, E. V. Shevchenko, S. K. Gray, U. Welp and V. K. VlaskoVlasov, Small, 2011, 7, 2365–2371. 26 Y. Zheng, T. Thai, P. Reineck, L. Qiu, Y. Guo and U. Bach, Adv. Funct. Mater., 2013, 23, 1519–1526. 27 D.-K. Lim, K.-S. Jeon, H. M. Kim, J.-M. Nam and Y. D. Suh, Nat. Mater., 2010, 9, 60–67. 28 J.-H. Lee, J.-M. Nam, K.-S. Jeon, D.-K. Lim, H. Kim, S. Kwon, H. Lee and Y. D. Suh, ACS Nano, 2012, 6, 9574–9584. 29 D.-K. Lim, I.-J. Kim and J.-M. Nam, Chem. Commun., 2008, 5312–5314. 30 J.-H. Lee, G.-H. Kim and J.-M. Nam, J. Am. Chem. Soc., 2012, 134, 5456–5459. 31 S. M¨ uhlig, A. Cunningham, S. Scheeler, C. Pacholski, T. B¨ urgi, C. Rockstuhl and F. Lederer, ACS Nano, 2011, 5, 6586–6592. 32 D. Tsoutsi, J. M. Montenegro, F. Dommershausen, U. Koert, L. M. Liz-Marz´ an, W. J. Parak and R. A. Alvarez-Puebla, ACS Nano, 2011, 5, 7539–7546. 33 S. Liu, T. Zhu, R. Hu and Z. Liu, Phys. Chem. Chem. Phys., 2002, 4, 6059–6062. 34 M. Gellner, D. Steinigeweg, S. Ichilmann, M. Salehi, M. Sch¨ utz, K. K¨ ompe, M. Haase and S. Schl¨ ucker, Small, 2011, 7, 3445–3451. 35 K. M. Mayer, S. Lee, H. Liao, B. C. Rostro, A. Fuentes, P. T. Scully, C. L. Nehl and J. H. Hafner, ACS Nano, 2008, 2, 687–692. 36 L. Xu, H. Kuang, C. Xu, W. Ma, L. Wang and N. A. Kotov, J. Am. Chem. Soc., 2012, 134, 1699–1709. 37 R. C. Mucic, J. J. Storhoff, C. A. Mirkin and R. L. Letsinger, J. Am. Chem. Soc., 1998, 120, 12674–12675. 38 J. J. Storhoff, A. A. Lazarides, R. C. Mucic, C. A. Mirkin, R. L. Letsinger and G. C. Schatz, J. Am. Chem. Soc., 2000, 122, 4640–4650. 39 J. Zheng, Z. Zhu, H. Chen and Z. Liu, Langmuir, 2000, 16, 4409–4412. 40 J. Turkevich, P. C. Stevenson and J. Hillier, Discuss. Faraday Soc., 1951, 11, 55–75. 41 G. Frens, Nature, 1973, 241, 20–22. 42 X. Ji, X. Song, J. Li, Y. Bai, W. Yang and X. Peng, J. Am. Chem. Soc., 2007, 129, 13939–13948. 43 S. J. Hurst, A. K. Lytton-Jean and C. A. Mirkin, Anal. Chem., 2006, 78, 8313–8318. 44 E. Chung, R. Gao, J. Ko, N. Choi, D. W. Lim, E. K. Lee, S.-I. Chang and J. Choo, Lab Chip, 2013, 13, 260–266. 45 T. Y. Olson, A. M. Schwartzberg, C. A. Orme, C. E. Talley, B. O'Connell and J. Z. Zhang, J. Phys. Chem. C, 2008, 112, 6319–6329. 46 L. Lu, H. Wang, Y. Zhou, S. Xi, H. Zhang, J. Hu and B. Zhao, Chem. Commun., 2002, 144–145. 47 P. Xu, J. H. Lee, K. Ma, C. Choi, S. Jin, J. Wang and J. N. Cha, Chem. Commun., 2013, 49, 8994–8996.
Analyst, 2014, 139, 2207–2213 | 2213
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
PAPER
View Journal | View Issue
Cite this: Analyst, 2014, 139, 2207
DNA-embedded Au–Ag core–shell nanoparticles assembled on silicon slides as a reliable SERS substrate† Zhong Zhang, Sha Zhang and Mengshi Lin* This study aimed at developing a sensitive and reliable SERS substrate by assembling DNA-embedded Au–Ag core–shell nanoparticles (NPs) on silicon slides. First, a monolayer of well separated DNA-functionalized Au NPs (40 nm) was decorated on (3-aminopropyl)triethoxysilane modified silicon slides. The DNA-embedded Au–Ag core–shell NPs were assembled on the 40 nm Au–DNA NPs to form a core–satellite structure through DNA hybridization. Using 4-MBA as a Raman dye, the SERS performance of the substrates was evaluated after being cleaned by low oxygen and argon plasma. The Raman intensity of the assembly using DNA-embedded Au–Ag core–shell NPs was 8–10 times higher than the intensity of the assembly using Au NPs as satellites. In addition, the signal-to-noise ratio of the assembly was 2.6 times higher than that of a commercial substrate (Klarite™) when a 785 nm laser was used. The SERS enhancements of the assembled substrates were 2.2 to 2.8 times higher than the Klarite when an acquisition time of 5 s was used at an excitation wavelength of 633 nm. The assembled
Received 13th November 2013 Accepted 9th February 2014
substrates also show a good spot-to-spot and substrate-to-substrate reproducibility at the excitation wavelengths of 633 and 785 nm. These results demonstrate that the fabrication process is simple and
DOI: 10.1039/c3an02116e
cost-effective for assembling DNA-embedded Au–Ag core–shell NPs on silicon slides that can be used
www.rsc.org/analyst
as a reliable SERS substrate.
1. Introduction Surface enhanced Raman scattering (SERS), a technique which can dramatically enhance the Raman signals of analyte molecules adsorbed on the metallic nanostructures, has been applied in chemical analysis, detection of food contaminants, biological sensing, and environmental monitoring.1–5 For example, SERS has been used to detect melamine, pesticides, and pathogens in food using commercial gold substrates.6–9 SERS was employed to analyze the endocrine disruptor using gold- and silver-decorated microspheres as substrates.10 SERS was also developed for detection of tumors using labeled gold nanoparticles (NPs) or gold-patterned microarrays as substrates.11,12 A variety of substrates have been developed in recent years for SERS applications. The substrates fabricated by nanolithography show promising SERS enhancement and great reproducibility.13–15 Sub-10 nm metallic nanogap arrays with precise control of the gap morphology have been produced by the nanolithography technique.16 However, it is still a challenge to fabricate smaller nanogaps (1–2 nm) by nanolithography.
Division of Food Systems & Bioengineering, University of Missouri, Columbia, MO, 65211-5160, USA. E-mail: [email protected]; Fax: +1-573-884-7964; Tel: +1-573884-6718 † Electronic supplementary 10.1039/c3an02116e
information
(ESI)
This journal is © The Royal Society of Chemistry 2014
available.
See
DOI:
Nanolithography also requires specic equipment for the development of nano-patterns and the deposition of a Au or Ag layer. Another lithography method, called nanosphere lithography, has been adopted by many researchers because it is inexpensive and able to create a large area of ordered nanostructures.17–19 However, nanosphere lithography still needs to pattern the substrate surface by a layer of nanospheres in advance, which is difficult to be tailored for different SERS detection strategies. In addition, SERS substrates can also be fabricated by assembling Au or Ag NPs on the glass, silicon surface, or liquid/liquid interface.20–25 Among all the fabrication methods, the self-assembling approach is the most convenient and cost-effective method because it can be performed without using expensive equipment. Between the junctions of the self-assembled NPs, numerous hot spots for SERS can be created where an intense electromagnetic eld is generated at this area. However, a problem of self-assembled substrates is that the distance between NPs is difficult to control and may vary in different locations on the same substrate. This could inuence the reproducibility of the substrate and also cause big variations in Raman signals in different batches. To solve this problem, DNA has been used to direct and organize the assembly of Au NPs on the silicon or glass surface modied with (3-aminopropyl)triethoxysilane (APTES).26 The DNA-directed assembly provides an excellent control of the distance between Au NPs using rigid
Analyst, 2014, 139, 2207–2213 | 2207
View Article Online
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
Analyst
DNA molecules and offers a good reproducibility for different batches, making it an ideal method for preparing SERS substrates. The single molecule sensitivity of SERS has been achieved by placing the Raman dye into the junction of a DNA-tethered dimer of Au NPs and coating a thin layer of Ag shell on the dimer.27,28 The coating of the Ag shell signicantly enhances the Raman signals of the dye molecules because Ag induces a higher enhancement than Au and the gap size between two NPs was reduced to less than 1.5 nm. It was proved that the hybridized DNA was embedded in the Ag shell during its growth.29,30 Inspired by the DNA-directed assembly of Au NPs and the signicant enhancement by the Ag shell, we direct the assembly of ssDNA embedded Au–Ag core–shell NPs on the silicon surface to form a SERS-active core–satellite structure. As shown in Fig. 1, the silicon surface was functionalized by a layer of APTES, which was used to adsorb negatively charged ssDNAcoated Au NPs by the positive amino group of APTES. Thus, a monolayer of Au–DNA NPs was decorated on the silicon surface through this electrostatic interaction.31–35 Using Au–DNA NPs as the core, the ssDNA embedded Au–Ag core–shell NPs will be assembled as satellites around the core NPs by DNA hybridization.36–38 In this structure, the gap distance between the core NPs and the satellite NPs was smaller due to the presence of the Ag shell as compared to core–satellite structures constructed only by Au NPs. Meanwhile, the Au–Ag core–shell NPs provide a much better SERS enhancement than Au NPs.27,28 Furthermore,
Paper
the gap distances between the core NPs and satellite NPs were almost identical because they were separated by the rigid DNA chains with the same length, giving a high reproducibility for different core–satellite structures. This structure could be used as a reliable SERS substrate with high enhancement and high reproducibility.
2. 2.1
2208 | Analyst, 2014, 139, 2207–2213
Materials and chemicals
Hydrogen tetrachloroaurate solution (HAuCl4, 30 wt% in dilute HCl), silver nitrate, sodium citrate dihydrate (>99%), L-ascorbic acid, 4-mercaptobenzoic acid, DL-dithiothreitol (DTT), (3-aminopropyl)triethoxysilane (APTES), and sodium dodecyl sulfate (SDS) were purchased from Sigma-Aldrich (St Louis, MO, USA) and used without any purication. Mono-thiolated ssDNA was purchased from Integrated DNA Technologies (IDT, Iowa, USA). All the ultrapure water (18.2 MU cm1) was prepared from a Millipore water purication system. 2.2
Modication of silicon substrates by APTES
The silicon slides (3 mm 4 mm) were rst cleaned and oxidized by piranha solution (3 : 1 H2SO4 : H2O2) for 60 min with gentle shaking.35,39 The silicon slides were then thoroughly rinsed with pure water and dried at room temperature. The modication of the silicon surface with APTES was performed by immersing the slides into a mixture of APTES, water, and ethanol (3 : 3 : 94) for 3 h with shaking. Aer modication, the slides were rinsed with copious amounts of water and baked at 105 C for 30 min. 2.3
Fig. 1 Schematic illustration of the fabrication of core–satellite structures as SERS substrates: (a) silicon slide was modified by APTES to form a layer of the NH2 group; (b) ssDNA-modified 40 nm Au NPs were adsorbed on the silicon surface; (c) ssDNA-embedded Au–Ag core–shell NPs were conjugated to 40 nm Au NPs through the hybridization of complementary ssDNA.
Materials and methods
Modication of Au NPs by thiolated ssDNA
Gold NPs with average diameters of 17 and 40 nm were synthesized by the reduction of HAuCl4 using sodium citrate.40–42 The procedure of DNA loading was similar to a previous report.43 Briey, thiolated ssDNA (Link-1 or Link-2) was rst reduced by 0.1 M DTT for an hour and puried by the NAP-5 column using ultrapure water as an eluent. The puried ssDNA (0.33 mg mL1, 1 mL) was added into 10 mL of Au NPs with 0.01% SDS and 1 mL of 0.01 M phosphate buffer. The mixture of ssDNA and Au NPs was subjected to ultrasonication for 30 s and then incubated for 20 min with shaking. Aer incubation, 0.3 mL of 2.0 M NaCl in 0.01 M phosphate buffer (pH ¼ 8.0) was added to the mixture. The mixture was then subjected to 30 s sonication and 20 min incubation. This process was repeated aer each addition of 0.3, 0.6, 0.65, 0.7, 0.77, and 0.80 mL of 2.0 M NaCl. Finally, the mixture was incubated overnight for the full adsorption of ssDNA on Au NPs. In this study, the ssDNA (Link-1) with a sequence of TTA TAA CTA TTC CTA AAA AAA AAA A C(6)SH was loaded on 17 nm Au NPs. The complementary ssDNA (Link-2) with a sequence of TAG GAA TAG TTA TAA AAA AAA AAA A C(6)SH was loaded on 40 nm Au NPs. 2.4
Fabrication of DNA embedded Au–Ag core–shell NPs
The modied Au NPs (17 nm, 1 mL) were centrifuged twice to remove excess Link-2 ssDNA and re-dispersed in 0.90 mL of
This journal is © The Royal Society of Chemistry 2014
View Article Online
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
Paper
ultrapure water in a 1.5 mL Eppendorf (EP) tube. Then 0.1 mL of phosphate buffer (0.18 M, pH ¼ 8.0) and 0.1 mL of NaCl (2.0 M) were added to adjust the pH and ionic strength of the Au NP solution. This step is very important for successful Ag coating because high ionic strength reduces the repulsion between the DNA chains and Ag atoms, facilitating the nucleation of Ag on the surface of Au NPs. Ascorbic acid (0.1 M, 150 mL) and AgNO3 (1 mM, 0–150 mL) were mixed with the NP solution. The tube was subsequently incubated in a 45 C water bath for 60 min to accelerate the growth of the Ag shell. In the meantime, the enhanced temperature also unfolded the DNA and reduced the adsorption of DNA base on the Au surface for the purpose of keeping the active part of DNA out of the Ag shell. Aer the growth of the Ag shell, the excess ascorbic acid was removed by double centrifugation. The Au–Ag core–shell NPs were nally dispersed in 300 mL phosphate buffer (10 mM, pH ¼ 7.45) with 0.5 M NaCl. 2.5 Assembling DNA embedded Au–Ag core–shell NPs into a core–satellite structure on silicon The DNA-modied 40 nm Au NPs (1 mL) were puried by three consecutive centrifugations and nally dispersed in 400 mL water. The APTES-modied silicon slide was immersed into the NP solution and kept in the solution overnight. Aer that, a monolayer of 40 nm Au NPs was formed on the silicon surface by the electrostatic interaction of the Au and NH2 groups. The silicon slide was rinsed with water three times to remove weakly bound NPs and then incubated in phosphate buffer solution (10 mM, pH ¼ 7.45) with 2 mM sulfosuccinimidyl 4-(N-maleimidomethyl)cyclohexane-1-carboxylate (Sulfo-SMCC) at 45 C for 2 h to block the excess NH2 groups on the silicon surface.26 Aer cleaning by water, the silicon slides were immersed into an EP tube containing 100 mL of DNA-embedded Au–Ag core–shell NPs. The EP tube was then heated in a 65 C water bath for 5 min and le overnight at room temperature for the hybridization of DNA. Finally, the silicon slides were rinsed with pure water for three times and dried at room temperature. 2.6
Analyst
2.7
Characterization by TEM and SEM
TEM (Jeol 1400, JEOL, Tokyo, Japan) was used to characterize the 40 nm Au NPs and the shell thickness of the Au–Ag core– shell NPs under an accelerating voltage of 120 kV. Before the SEM characterization, the substrates were attached on the aluminum stub by double-sided adhesive carbon tape. The SEM images of the substrates were obtained using a FEI Quanta 600 FEG Scanning Electron Microscope (FEI, Oregon, USA).
3. 3.1
Results & discussion Decoration of 40 nm Au–DNA NPs on the silicon surface
Au NPs with an average diameter of 40 nm were fabricated by the citrate reduction method. As shown in Fig. 2a, Au NPs are monodispersed with uniform size distribution with spherical/ oval shapes. Aer modication with ssDNA, Au NPs were used to decorate the surface of APTES-functionalized silicon. A very uniform monolayer of Au–DNA NPs was attached and dispersed on the silicon surface (Fig. 2b). Only a few clusters of dimers, trimers and tetramers of NPs could be observed in the SEM image. These clusters have little inuence on the reproducibility of the SERS substrate because it is estimated that 97% of the Au–DNA NPs are well separated from each other due to the strong repulsion of the negatively charged DNA coating (Fig. 2c). According to the scheme, the 40 nm Au–DNA NPs were used to attract and direct the assembly of the Au–DNA–Ag NPs through the interaction of the complementary ssDNA. Therefore, the number of core–satellite structures formed on per mm2 silicon will be determined by the number of Au–DNA NPs in this area. The Raman spot size was estimated to be 1 mm2 when an excitation wavelength of 785 nm and a 50 objective were used. The number of Au NPs per mm2 was counted for ten randomly
SERS measurement
The as-prepared substrates were cleaned by plasma treatment under 4 min of oxygen/argon ow (5 : 95) and 2 min of argon ow at a power of 50 mW and a gas ow rate of 30 sccm. Aer cleaning, the silicon were immersed into 400 mL ethanol solution of 4-MBA (106 M) for 3 h. A commercial substrate, Klarite™, was used as control and also kept in the same concentration of 4-MBA solution for 3 h. The substrates were rinsed with 400 mL water and air dried at room temperature. A 785 nm diode laser and a 633 nm He–Ne laser were used to evaluate the SERS performance of the substrates. SERS spectra were recorded using a Renishaw 1000 micro-Raman system at a wavelength of 785 nm using an acquisition time of 10 s and a laser power of 30 mW using a 50 objective. Raman data were also collected using a Horiba LabRam-HR 800 micro-Raman spectrometer at an excitation wavelength of 633 nm using a laser power of 4.75 mW and acquisition time of 1 s and 5 s. All the SERS data were recorded from at least ve randomly selected spots on the substrate. This journal is © The Royal Society of Chemistry 2014
Characterization of ssDNA-modified 40 nm Au NPs: (a) TEM image of pure Au NPs with an average size of 40 nm; (b) and (c) SEM images of a monolayer of Au NPs adsorbed on the APTES modified silicon surface at different magnifications; (d) average number of Au NPs per mm2 at ten randomly selected locations.
Fig. 2
Analyst, 2014, 139, 2207–2213 | 2209
View Article Online
Analyst
Paper
selected areas in the SEM image (Fig. 2b). It can be observed that the number per mm2 was narrowly distributed between 47 and 55 (Fig. 2d). This is crucial for a good reproducibility in SERS because the narrow distribution will reduce the signal uctuation from spot to spot and contribute to a good reproducibility for SERS.
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
3.2 Assembly of DNA-embedded Au–Ag core–shell NPs on Au–DNA NP decorated silicon The 17 nm Au NPs were rst functionalized by ssDNA (Link-1) which is complementary to the ssDNA (Link-2) used for 40 nm Au NPs. The DNA-embedded Au–Ag core–shell NPs were synthesized by coating a thin layer of the Ag shell on the 17 nm Au–DNA NPs. Ag could be easily reduced by ascorbic acid using Au NPs as seeds.44 However, the reduction of Ag on Au–DNA NPs is different from the deposition of Ag on citrate stabilized Au NPs at room temperature because the growth of the Ag shell will be hindered by dense DNA chains on the Au surface while it will not be inuenced by citrate.29,45,46 Thus, an appropriate NaCl concentration and an enhanced temperature should be maintained during the growth of the Ag shell. The Ag shell was successfully formed on the 17 nm Au–DNA NPs which was proved by the color change of NP solution from pink to yellow. Otherwise, the color of the 17 nm Au–DNA NP solution would not change from pink to yellow aer the addition of AgNO3 and ascorbic acid if NaCl was not used to enhance the ionic strength of the mixture. The thickness of the Ag shell has to be carefully controlled to avoid burying the whole ssDNA chain in the shell. It is obvious that the Au–DNA–Ag NPs would not be able to link with the 40 nm Au–DNA NPs on the silicon surface if the ssDNA molecule was hidden under the Ag shell. The efficiency of assembling will also be affected if some of complementary bases on the DNA chain were embedded in the Ag shell. In this study, the thickness of the Ag shell was controlled by adding different amounts of 1 mM AgNO3 (50, 100, and 150 mL) while maintaining excess ascorbic acid in the solution. The Au–DNA–Ag NPs with different thicknesses were then incubated with the 40 nm Au–DNA NP decorated silicon slide in the 0.5 M NaCl solution. The excess NH2 groups on the silicon surface were blocked by Sulfo-SMCC before the assembling process. Aer assembling at room temperature for 12 h, the 40 nm Au–DNA NPs were linked with Au–DNA–Ag NPs and surrounded by these NPs via DNA interaction (Fig. 3b, d and f). The average numbers of satellite NPs assembled on each 40 nm Au–DNA NP were 6.8 1.4, 5.3 1.6, and 4.9 1.6 respectively when 50, 100, and 150 mL of AgNO3 were used to grow the Ag shell (Fig. 3a, c and e). The corresponding thickness of the Ag shell was around 1.1, 2.1, and 3.0 nm measured according to the TEM images (Fig. S1†). The Au– DNA–Ag NPs were still able to link to the 40 nm Au–DNA NPs because the length of the DNA spacer used in this study was 4 nm.44 There are two possible reasons for the decreased number of satellite NPs assembled on each 40 nm Au–DNA NP. First, more ssDNA molecules were buried in the Ag shell
2210 | Analyst, 2014, 139, 2207–2213
Characterization of the core–satellite structure assembled by ssDNA-embedded Au–Ag core–shell NPs with different Ag thicknesses by SEM and TEM (embedded): (a) and (b) 50 mL of AgNO3 (1 mM) coating; (c) and (d) 100 mL of AgNO3 (1 mM) coating; (e) and (f) 150 mL of AgNO3 (1 mM) coating. Fig. 3
when the Ag shell became thicker. In other words, the DNA molecules tended to lose their active bases as the thickness of the Ag shell increased. Furthermore, the assembly of Au–DNA NPs and Au–DNA–Ag into the core–satellite structure on silicon is different from the assembly of both NPs in solution. The steric hindrance is larger for the Au–DNA NPs immobilized on the silicon surface compared to that of Au–DNA in solution. The steric hindrance was also enhanced due to the increased size of satellite NPs, reducing the efficiency of DNA hybridization. Nevertheless, the average number of assembled satellite NPs was similar for the NPs with 100 and 150 mL of AgNO3 coating. It is observed in Fig. 3c and e that most of the 40 nm Au–DNA NPs were still uniformly surrounded by Au–DNA–Ag NPs. In this case, 150 mL of AgNO3 solution was chosen to build Au–DNA–Ag NPs due to the thicker Ag shell, which could lead to a smaller gap distance between the 40 nm core and the DNA-embedded Au–Ag NPs. As a comparison, 27 nm Au NPs have also been used as satellite NPs to assemble the core–satellite structure. It was found that 27 nm Au NPs impose a much higher steric hindrance than 17 nm Au NPs, which makes it unable to uniformly attach on the core NPs and form perfect core–satellite structures on silicon (Fig. S1f†). The Au NPs with a diameter of 40 nm were selected as core nanoparticles because 40 nm Au NPs could be conveniently synthesized by a classic citrate reduction method while it is necessary to use the seed-mediated method to grow nanoparticles larger than 40 nm.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
Analyst
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
3.3 Comparison of the SERS enhancement between the assemblies using 17 nm Au–DNA NPs and Au–DNA–Ag NPs as satellites The DNA molecules on the surface of NPs should be removed before the SERS measurements because the DNA coating would hinder the adsorption of the Raman dye on the substrate. Unlike the assembly of pure Au NPs, the Ag shell of DNA-embedded core–shell NPs could be easily oxidized by the UV–ozone treatment. To solve this problem, low oxygen plasma treatment was used in this study. The plasma treatment would not cause any substantial change in the core–satellite structure as evidenced by the TEM images (Fig. S2†). However, the SERS enhancement could be improved by the plasma treatment under suitable conditions. Fig. 4 shows the Raman spectra of 4-MBA on the substrates which were subjected to different plasma treatments. The Raman peak at 1078 cm1, a typical band of 4-MBA, was found to be the highest for the substrate treated by 5% oxygen plasma. The Raman intensity is much lower for the substrate without any treatment. It was also found that there are no visible peaks of 4-MBA for the substrates treated by 25% oxygen plasma due to the oxidation of the Ag layer. Fig. 5 shows the Raman spectra of 4-MBA adsorbed on the core–satellite assemblies using 17 nm Au–DNA NPs or Au–DNA–Ag NPs as satellite NPs. It is clearly observed that the Raman peak at 1078 cm1 is much higher for the assembly using Au–DNA–Ag NPs as satellites compared to that using Au–DNA NPs as satellites. The Raman peaks of 4-MBA from 1100 to 1300 cm1 were barely recognizable for the assembly using Au–DNA NPs as satellite NPs. It is estimated that enhancements by 8–10 times were induced by coating the satellite NPs with a thin layer of Ag shell. This enhancement could be attributed to the better SERS response of Ag, the reduction of gap size between the core and satellites, and the larger size of satellite NPs aer the growth of the Ag shell.47 The assembly using DNA-embedded Au–Ag core–shell NPs as satellites shows a much more promising SERS enhancement than the assembly using Au–DNA NPs as satellites.
Fig. 5 Comparison of SERS intensity of 4-MBA (106 M) using different NPs as satellites: top spectrum: ssDNA embedded Au–Ag core–shell NPs as satellites; bottom spectrum: ssDNA-modified Au NPs as satellites.
3.4 Reproducibility of the assembly and comparison with a commercial substrate Fig. 6a shows the Raman spectra of 4-MBA adsorbed on the assembled substrates and a commercial substrate (Klarite) when excited by a 785 nm laser. It is found that the intensities of peaks at 1078 cm1 and 1585 cm1 were comparable for both substrates. But the signal-to-noise ratio of the assembly (47.3) was 2.6 times higher than that of Klarite (18.4), showing a better performance of the assembly. Fig. 6b and c show the Raman intensities of peaks at 1078 and 1585 cm1 from three different substrates and the Klarite. It can be found that there is no signicant difference for the Raman intensity of the three assembled substrates, indicating a good reproducibility of the substrates (Fig. S3a†). The variation of Raman intensity at
Comparison between the assembled substrates and a commercial substrate using a 785 nm laser: (a) Raman spectra of 4-MBA (106 M) on the assembly and the Klarite; (b) Raman intensity of 4-MBA at 1078 cm1 from three assembled substrates and the Klarite; (c) Raman intensity of 4-MBA at 1585 cm1 from three assembled substrates and the Klarite. Fig. 6
Fig. 4 Raman spectra of 4-MBA (106 M) from the assembled substrates treated under different plasma conditions.
This journal is © The Royal Society of Chemistry 2014
Analyst, 2014, 139, 2207–2213 | 2211
View Article Online
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
Analyst
different locations on the same substrate has also been evaluated and shown as error bars of the peak intensities at 1078 and 1585 cm1. The results show that the spot-to-spot variation of the three assembled substrates is close to that of the Klarite produced by the lithography method, suggesting a uniform distribution of the core–satellite structures on the silicon surface. A drawback of the substrate is that the Ag shell on the satellite NPs can be easily oxidized in air. The SERS performance of the substrates was largely reduced aer 4 month storage at room temperature (Fig. S3b†). Therefore, the substrate should be used right aer the preparation or stored in an oxygen-free atmosphere for long term stability. The SERS performance of the assembled substrates was evaluated and compared with the Klarite by measuring 4-MBA with a 633 nm laser. It was found that the 633 nm laser induced much higher enhancement than the 785 nm laser. For the assembled core–satellite structure by pure Au NPs, the surface plasmon resonance has been found to be at 650 nm.26 However, in this study, the surface plasmon resonance of the core–satellite structure blue-shied because of the Ag shell on the satellite nanoparticles, leading the surface plasmon resonance closer to 633 nm. Therefore, the 633 nm laser resulted in a higher enhancement than the 785 nm laser. As shown in Fig. 7a, the Raman intensity of the band at 1078 and 1585 cm1 from the assembled substrates was 2.2 and 2.8 times higher than the Raman intensity from the Klarite when 25% of laser power and 5 s of acquisition time were used. Meanwhile, the original Raman spectra from the Klarite have a much higher baseline dri than the assembled substrates (Fig. S4†). The high baseline dri causes problems for the acquisition of Raman data when a higher laser power and a longer acquisition time
Fig. 7 Comparison of the assembled substrates and a commercial substrate using a 633 He–Ne laser: (a) Raman spectra of 4-MBA (106 M) on the assembly and the Klarite using an acquisition time of 5 s; (b) Raman intensity of 4-MBA at 1078 cm1 acquired from three assembled substrates and the Klarite by a 1 s acquisition time; (c) Raman intensity of 4-MBA at 1585 cm1 acquired from three assembled substrates and the Klarite by a 1 s acquisition time.
2212 | Analyst, 2014, 139, 2207–2213
Paper
are needed to increase the Raman signals. The Raman spectra of 4-MBA from the Klarite exceeded the detection limit when the acquisition time was extended to 10 s due to the baseline dri (Fig. S5†). The reproducibility of the substrates was also evaluated using a shorter acquisition time (1 s) at an excitation wavelength of 633 nm. The Raman intensities of the band at 1078 and 1585 cm1 were very consistent for all of the tested substrates, indicating a good substrate-to-substrate reproducibility. The Raman intensities of the peak at 1078 and 1585 cm1 for the assembled substrates were 1.25 times and 2.1 times higher than the intensity of spectra acquired using the Klarite when 1 s of acquisition time was applied, demonstrating a better SERS performance of assembled substrates. More importantly, the cost to fabricate this substrate is much lower than the commercial substrate, which is estimated to be around 5.22 US dollars for each substrate. Meanwhile, the synthesis of the substrate does not require pattern development, metal deposition, ion milling, or other complex steps in the lithography method.
4. Conclusions In this study, DNA embedded Au–Ag core–shell NPs were assembled with 40 nm Au–DNA NPs on the surface of silicon slides through the hybridization of DNA attached on the NPs. The monolayer of core–satellite structures was uniformly distributed on the silicon surface using well separated 40 nm Au–DNA NPs as cores. Around 4.9 1.6 Au–DNA–Ag NPs were assembled on each 40 nm Au core when the thickness of the Ag shell was 3.1 nm. In addition, the SERS performance of the substrates was evaluated aer low oxygen and argon plasma cleaning. The Raman intensity of 4-MBA is 8–10 times higher for the assembly using DNA Au–DNA–Ag NPs as satellites compared to that of the assembly using Au–DNA NPs without Ag coating as satellites. It is found that the Raman enhancement of substrates was similar to a commercial substrate when a 785 nm laser was used. But the signal-noise ratio of the assembly was 2.6 times higher than that of a commercial substrate. The SERS enhancements of the assembled substrates were 2.2 to 2.8 times higher than a commercial substrate when 5 s of acquisition time was used at an excitation wavelength of 633 nm. The assembled substrates also show a good spot-tospot and substrate-to-substrate reproducibility at the excitation wavelengths of 633 nm and 785 nm. The assembled substrates show great potential for SERS applications due to their high enhancement, good reproducibility, low cost, and simple fabrication. Future study is needed to reduce the steric hindrance for assembling bigger Au–DNA–Ag NPs and to further improve the SERS enhancement.
Acknowledgements We thank Dr Richard K. Brow and Jaime George at Missouri University of Science and Technology for their assistance in Raman measurement and the Electron Microscopy Center at the University of Missouri for TEM and SEM characterization.
This journal is © The Royal Society of Chemistry 2014
View Article Online
Paper
This research was supported by the USDA NIFA Nanotechnology Program project #2011-67021-30391.
Published on 13 March 2014. Downloaded by University of Chicago on 06/10/2014 22:16:23.
References 1 M. Fleischmann, P. Hendra and A. McQuillan, Chem. Phys. Lett., 1974, 26, 163–166. 2 R. P. Cooney, E. S. Reid, M. Fleischmann and P. J. Hendra, J. Chem. Soc., Faraday Trans. 1, 1977, 73, 1691–1698. 3 B. Liu, G. Han, Z. Zhang, R. Liu, C. Jiang, S. Wang and M. Y. Han, Anal. Chem., 2012, 84, 255–261. 4 X. Qian, X. H. Peng, D. O. Ansari, Q. Yin-Goen, G. Z. Chen, D. M. Shin, L. Yang, A. N. Young, M. D. Wang and S. Nie, Nat. Biotechnol., 2008, 26, 83–90. 5 W. Ren, C. Zhu and E. Wang, Nanoscale, 2012, 4, 5902–5909. 6 M. Lin, L. He, J. Awika, L. Yang, D. Ledoux, H. Li and A. Mustapha, J. Food Sci., 2008, 73, T129–T134. 7 B. Liu, P. Zhou, X. Liu, X. Sun, H. Li and M. Lin, Food Bioprocess Technol., 2013, 6, 710–718. 8 C. Fan, Z. Hu, L. K. Riley, G. A. Purdy, A. Mustapha and M. Lin, J. Food Sci., 2010, 75, M302–M307. 9 T. A. Alexander and D. M. Le, Appl. Opt., 2007, 46, 3878–3890. 10 C. Fang, N. M. Bandaru, A. V. Ellis and N. H. Voelcker, Biosens. Bioelectron., 2013, 42, 632–639. 11 M. Lee, S. Lee, J.-h. Lee, H.-w. Lim, G. H. Seong, E. K. Lee, S.-I. Chang, C. H. Oh and J. Choo, Biosens. Bioelectron., 2011, 26, 2135–2141. 12 S. Lee, S. Kim, J. Choo, S. Y. Shin, Y. H. Lee, H. Y. Choi, S. Ha, K. Kang and C. H. Oh, Anal. Chem., 2007, 79, 916–922. 13 R. Alvarez-Puebla, B. Cui, J.-P. Bravo-Vasquez, T. Veres and H. Fenniri, J. Phys. Chem. C, 2007, 111, 6720–6723. 14 N. A. Abu Hatab, J. M. Oran and M. J. Sepaniak, ACS Nano, 2008, 2, 377–385. 15 J. Theiss, P. Pavaskar, P. M. Echternach, R. E. Muller and S. B. Cronin, Nano Lett., 2010, 10, 2749–2754. 16 H. Im, K. C. Bantz, N. C. Lindquist, C. L. Haynes and S.-H. Oh, Nano Lett., 2010, 10, 2231–2236. 17 G. H. Chan, J. Zhao, E. M. Hicks, G. C. Schatz and R. P. Van Duyne, Nano Lett., 2007, 7, 1947–1952. 18 H. Im, K. C. Bantz, S. H. Lee, T. W. Johnson, C. L. Haynes and S. H. Oh, Adv. Mater., 2013, 25, 2678–2685. 19 X. Zhang, J. Zhao, A. V. Whitney, J. W. Elam and R. P. Van Duyne, J. Am. Chem. Soc., 2006, 128, 10304–10309. 20 Q. Su, X. Ma, J. Dong, C. Jiang and W. Qian, ACS Appl. Mater. Interfaces, 2011, 3, 1873–1879. 21 M. P. Cecchini, V. A. Turek, J. Paget, A. A. Kornyshev and J. B. Edel, Nat. Mater., 2012, 12, 165–171. 22 R. A. Alvarez-Puebla, A. Agarwal, P. Manna, B. P. Khanal, P. Aldeanueva-Potel, E. Carb´ o-Argibay, N. Pazos-P´ erez, L. Vigderman, E. R. Zubarev and N. A. Kotov, Proc. Natl. Acad. Sci. U. S. A., 2011, 108, 8157–8161. 23 B. Peng, G. Li, D. Li, S. Dodson, Q. Zhang, J. Zhang, Y. H. Lee, H. V. Demir, X. Yi Ling and Q. Xiong, ACS Nano, 2013, 7, 5993–6000.
This journal is © The Royal Society of Chemistry 2014
Analyst
24 L. Tong, T. Zhu and Z. Liu, Chem. Soc. Rev., 2011, 40, 1296– 1304. 25 A. Chen, A. E. DePrince, A. Demorti` ere, A. Joshi-Imre, E. V. Shevchenko, S. K. Gray, U. Welp and V. K. VlaskoVlasov, Small, 2011, 7, 2365–2371. 26 Y. Zheng, T. Thai, P. Reineck, L. Qiu, Y. Guo and U. Bach, Adv. Funct. Mater., 2013, 23, 1519–1526. 27 D.-K. Lim, K.-S. Jeon, H. M. Kim, J.-M. Nam and Y. D. Suh, Nat. Mater., 2010, 9, 60–67. 28 J.-H. Lee, J.-M. Nam, K.-S. Jeon, D.-K. Lim, H. Kim, S. Kwon, H. Lee and Y. D. Suh, ACS Nano, 2012, 6, 9574–9584. 29 D.-K. Lim, I.-J. Kim and J.-M. Nam, Chem. Commun., 2008, 5312–5314. 30 J.-H. Lee, G.-H. Kim and J.-M. Nam, J. Am. Chem. Soc., 2012, 134, 5456–5459. 31 S. M¨ uhlig, A. Cunningham, S. Scheeler, C. Pacholski, T. B¨ urgi, C. Rockstuhl and F. Lederer, ACS Nano, 2011, 5, 6586–6592. 32 D. Tsoutsi, J. M. Montenegro, F. Dommershausen, U. Koert, L. M. Liz-Marz´ an, W. J. Parak and R. A. Alvarez-Puebla, ACS Nano, 2011, 5, 7539–7546. 33 S. Liu, T. Zhu, R. Hu and Z. Liu, Phys. Chem. Chem. Phys., 2002, 4, 6059–6062. 34 M. Gellner, D. Steinigeweg, S. Ichilmann, M. Salehi, M. Sch¨ utz, K. K¨ ompe, M. Haase and S. Schl¨ ucker, Small, 2011, 7, 3445–3451. 35 K. M. Mayer, S. Lee, H. Liao, B. C. Rostro, A. Fuentes, P. T. Scully, C. L. Nehl and J. H. Hafner, ACS Nano, 2008, 2, 687–692. 36 L. Xu, H. Kuang, C. Xu, W. Ma, L. Wang and N. A. Kotov, J. Am. Chem. Soc., 2012, 134, 1699–1709. 37 R. C. Mucic, J. J. Storhoff, C. A. Mirkin and R. L. Letsinger, J. Am. Chem. Soc., 1998, 120, 12674–12675. 38 J. J. Storhoff, A. A. Lazarides, R. C. Mucic, C. A. Mirkin, R. L. Letsinger and G. C. Schatz, J. Am. Chem. Soc., 2000, 122, 4640–4650. 39 J. Zheng, Z. Zhu, H. Chen and Z. Liu, Langmuir, 2000, 16, 4409–4412. 40 J. Turkevich, P. C. Stevenson and J. Hillier, Discuss. Faraday Soc., 1951, 11, 55–75. 41 G. Frens, Nature, 1973, 241, 20–22. 42 X. Ji, X. Song, J. Li, Y. Bai, W. Yang and X. Peng, J. Am. Chem. Soc., 2007, 129, 13939–13948. 43 S. J. Hurst, A. K. Lytton-Jean and C. A. Mirkin, Anal. Chem., 2006, 78, 8313–8318. 44 E. Chung, R. Gao, J. Ko, N. Choi, D. W. Lim, E. K. Lee, S.-I. Chang and J. Choo, Lab Chip, 2013, 13, 260–266. 45 T. Y. Olson, A. M. Schwartzberg, C. A. Orme, C. E. Talley, B. O'Connell and J. Z. Zhang, J. Phys. Chem. C, 2008, 112, 6319–6329. 46 L. Lu, H. Wang, Y. Zhou, S. Xi, H. Zhang, J. Hu and B. Zhao, Chem. Commun., 2002, 144–145. 47 P. Xu, J. H. Lee, K. Ma, C. Choi, S. Jin, J. Wang and J. N. Cha, Chem. Commun., 2013, 49, 8994–8996.
Analyst, 2014, 139, 2207–2213 | 2213
