ARTICLE pubs.acs.org/JPCC
Reporter-Embedded TiO2 Core-Mixed Metal Shell Nanoparticles with Enormous Average Surface-Enhanced Raman Scattering Enhancement Factors Wenbing Li,† Xiaoyu Miao,‡ Ting Shan Luk,‡ and Peng Zhang*,§ †
Department of Chemistry, New Mexico Tech, Socorro, New Mexico 87801, United States Center for Integrated Nanotechnologies, Sandia National Laboratories, Albuquerque, New Mexico 87123, United States § Department of Chemistry, University of Cincinnati, Cincinnati, Ohio 45221, United States ‡
bS Supporting Information ABSTRACT: We report the development of Raman reporter-embedded TiO2 nanoparticles coated with mixed Ag and Au shells, showing enormous ensemble surface-enhanced Raman scattering (SERS) enhancement factors (up to 1010). Effects of shell composition on the enhancement are investigated both experimentally and theoretically. Colloidal TiO2 nanoparticles are first tagged with meso-tetra(4-carboxyphenyl)porphine or tris(2,20 -bipyridyl)ruthenium(II) chloride, used as reporter molecules. They are subsequently coated with either a Ag shell or a mixed Au-Ag shell with different compositions. The resulting nanostructures are characterized by UV-visible spectroscopy, transmission electron microscopy, energy-dispersive X-ray spectroscopy, particle size analyzer, and Raman spectroscopy. These Raman reporter-embedded TiO2 core-metal shell nanoparticles exhibit reproducible SERS signals of the reporter molecules with very high average enhancement factors. Interestingly, depending on the excitation wavelength, bimetallic Au-Ag shell nanostructures with proper Au/Ag ratios display higher enhancement factors than Ag-only shell. Simulation results based on equivalent dielectric functions show a very good match with the experimental observations.
’ INTRODUCTION Surface-enhanced Raman scattering (SERS) has been intensely studied since being observed over three decades ago.1-3 A large number of various substrates have been discovered and developed to give rise to SERS, such as electrochemically roughened silver and gold surfaces, colloidal aggregates of silver, monodispersed silver sols, cold-deposited silver films, metal nanoparticles of various shapes (spherical, prismatic, and triangular), and metal nanoshells, among others.4-6 The strongest SERS effects are usually observed in metal substrates, typically Ag and Au, rough on a nanoscopic scale.7-11 The effort to exploit SERS for applications can be classified into two categories. Typically, nanostructures of various designs are being developed and fabricated/synthesized to act as SERS substrates. These substrates can be used for detections of numerous targets, including biological molecules, pollutants, explosives, and bacteria, among others.12-15 Alternatively, SERS tags are being developed and synthesized as labeling agents for bioimaging and bioassays.16-18 These SERS tags usually consist of Raman reporter molecules being placed in the proximity of an Au or Ag nanoparticle, together being surrounded by a silica or polymer shell. The advantage of using SERS tags as opposed to fluorescence-tags is the inherently narrow and highly specific features of the spectra, offering a greater degree of multiplexing capability. Unfortunately, the average enhancement factors (EFs) r 2011 American Chemical Society
of currently available SERS tags are not sufficiently high to allow them to compete with fluorescence-based techniques. Currently the consensus view is that SERS enhancement comes from two components: an electromagnetic (EM) enhancement as the predominant contribution and a chemical enhancement as the minor contribution.19 It is believed that the EM enhancement is solely due to the optical properties of the substrates independent of the nature of the Raman probe molecules, while the chemical enhancement is more pertinent to the nature of probe molecules. The excitation of surface plasmons in the SERS-active substrates leads to the increase in the magnitude of both the incident and the scattered EM fields at some regions, combining to generate an enormous increase in SERS intensity of the reporter molecules locating at these socalled “hot spots”, which makes single-molecule Raman spectroscopy possible.20,21 In practice, unfortunately, the actual positions of such hot spots are difficult, if at all, to predict or locate individually, where single molecule EFs are expected to be on the order of ∼1010.22 Most studies concentrating on measuring the ensemble SERS EFs report the average values on the order of 104-106, and sometimes as high as 108.23-26 Received: December 8, 2010 Revised: January 23, 2011 Published: February 09, 2011 3318
dx.doi.org/10.1021/jp1116849 | J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C The significant difference in SERS EFs between the single molecular measurements and the ensemble measurements largely lies in the fact that placing the Raman reporter molecules onto these hot spots remains a technical challenge, which has been the objective of a lot of research efforts. We previously reported the observation of SERS signals of reporter molecules residing inside a metal nanoshell experimentally for the first time.3 The Raman reporter molecules, usually dye molecules, are encapsulated inside the SiO2 nanoparticles through doping during the nanoparticle formation. The dye-doped SiO2 nanoparticles are then coated with a thin layer of either Au or Ag nanoshell. The resulting core-shell nanostructures display highly enhanced Raman signals of the embedded molecules. Preliminary estimation leads to the EFs of the dye-embedded SiO2 core-metal shell nanostructures to be on the order of ∼1011. More recently, we replaced the SiO2 nanoparticles with TiO2 nanoparticles and developed a general approach to incorporate a wide variety of Raman reporter molecules with different surface charge properties into the design.27 The adoption of TiO2 nanoparticles as the core material retained the highly enhanced SERS signal, while reducing the overall size of the nanostructures by approximately 1 order of magnitude (from 70-100 nm to ∼5 nm). In these studies, we noticed that Ag-shell nanostructures tended to have higher EFs than the Au-shell nanostructures. Yet the latter displayed much better stability than the former. We thus decided to replace the single metal shells with mixed Au/Ag shells, in hope of combining the high EFs of the Ag shell with the good stability of the Au shell. Here we report the synthesis of Raman reporter-embedded TiO2 core metal shell nanostructures with different Ag/Au ratios, all displaying enormous average SERS EFs. The adoption of mixed-metal shells leads to some unusual effects on the SERS EFs of the nanostructures, i.e., mixed Au/Ag shells in some cases lead to higher SERS EFs than that using only a Ag shell, a counterintuitive behavior since Ag is always considered to be a better enhancer. This behavior is supported by the results from the theoretical modeling.
’ EXPERIMENTAL SECTION Materials. All chemicals were obtained commercially and used without further purification. Titanium chloride (TiCl4) was obtained from Fisher. meso-Tetra(4-carboxyphenyl)porphine (TCPP) was purchased from Frontier Scientific. Tris(2,20 -bipyridyl)ruthenium(II) chloride (RuBpy), sodium citrate, sodium borohydride (NaBH4), and hydrogen tetrachloroaurate (HAuCl4, 99.99%) were obtained from Sigma-Aldrich. All chemicals, unless specified, were of reagent grade. Deionized water, with a resistivity greater than 18.0 MΩ 3 cm (Millipore Milli-Q system), was used in preparing the aqueous solutions. All glassware used in the experiments was washed with freshly prepared aqua regia (3:1 of HCl:HNO3) and rinsed thoroughly in tap water first and then deionized (DI) water prior to use. Preparation of Colloidal TiO2 Nanoparticles. Colloidal stock solutions of TiO2 were prepared by the hydrolysis of TiCl4 as previously described.27,28 In brief, 2.50 mL of 99.8% TiCl4 at 0 °C was introduced under a stream of argon into 35.0 mL of icechilled 0.10 M HCl solution under vigorous stirring. After 30 min of stirring at 0 °C, the solution was dialyzed against aqueous HCl of pH 2.3 using Spectra/Por membrane tubing (Spectrum Medical Industries) with MW 6000-8000 cutoff pores. A transparent solution containing ca. 20 g/L TiO2 at pH 2.3 was
ARTICLE
thus obtained. The solution was kept in the refrigerator at 4 °C and used within 3 months. Preparation of Colloidal TCPP-TiO2 Nanoparticles. TCPPTiO2 nanoparticles were prepared by mixing 2.60 mg of TCPP with 10.00 mL of 0.20 g/L colloidal TiO2 solution (pH = 2.3) for 7-8 days under vigorous stirring. The mixture was kept in the dark during the preparation in order to avoid photocatalytic decomposition of TCPP. The excessive TCPP molecules were then separated from the aqueous phase as precipitate after centrifugation. Preparation of Colloidal RuBpy-TiO2 Nanoparticles. Ten milliliters of 0.20 g/L colloidal TiO2 solution (pH = 2.3) was mixed with 2.00 mL of 1.0 mM RuBpy solution under vigorous stirring. Then 0.10 M of NaOH was added into the mixture to adjust the pH to 8. Excessive RuBpy molecules were separated from aqueous phase in supernatant after centrifugation. Coating of Ag nanoshell onto the Surface of Raman Reporter-Embedded TiO2 Nanostructures. Citrate stabilized Ag nanoshells on the surface of reporter-embedded TiO2 nanoparticles were prepared from the NaBH4 reduction of AgNO3. In brief, 1.0 mL of 25 mM aqueous sodium citrate solution was added to 10 mL of a 0.20 g/L reporter-embedded TiO2. Then the pH of the mixture was adjusted to be above 7 with 0.10 M NaOH, so that the TiO2 surface became negatively charged. Next, 1.0 mL of 10 mM aqueous AgNO3 solution was mixed thoroughly with the above solution for an hour and the Agþ ions would adsorb to the reporter-embedded TiO2 surface. Lastly, a total of 0.1 mL of 0.10 M freshly prepared aqueous NaBH4 solution was added dropwise under vigorous stirring, leading to a yellowish brown solution. The resulting reporter-embedded TiO2 core-Ag shell nanoparticles were aged for 24 h to decompose the residual NaBH4 before they were used in subsequent steps. Coating of Au Nanoshell onto the Surface of Raman Reporter-Embedded TiO2 Core-Ag Shell Nanoparticles. To prepare reporter-embedded TiO2 core-Au/Ag shell nanoparticles, a controlled amount of 1.0 mM aqueous HAuCl4 solution was added dropwise to the reporter-embedded TiO2 core-Ag shell nanoparticles solution under vigorous stirring. For example, in the preparation of TiO2 core-Au/Ag shell nanoparticles with Au/Ag molar ratio 1/9, 0.10 mL of 1.0 mM aqueous HAuCl4 solution was added dropwise to 12 mL of the reporter-embedded TiO2 core-Ag shell nanoparticles solution (the Ag amount being equivalent to 0.1 mM AgNO3 solution) under vigorous stirring. The yellow color of the reporter-embedded TiO2 coreAg shell nanoparticles gradually diminished, while the pink color of the Au shell evolved, indicating the reduction of chloroaurate ions by the Ag nanoshell. In all cases, the color of the initial Ag shell nanoparticles solution gradually changed from yellowish brown to different shades of pink. Here, no extra NaBH4 solution was added to induce the formation Au shell on the moleculeembedded TiO2 core-Ag shell nanoparticles. Preparation of Ag and Au nanoparticles. Silver and gold nanoparticles used as reference systems in this study were prepared following the procedures reported in the literature with some variations.29,30 Briefly, for Ag nanoparticles, 0.1 mL of 10 mM AgNO3 was added dropwise to 9.9 mL of 25 mM sodium citrate solution. The solution was vigorously stirred for 5 min. Then 0.5 mL of an aqueous 20 mM NaBH4 in 25 mM sodium citrate solution was added dropwise to the above solution. The resulting solution was stirred for 1 h and aged for at least 24 h at room temperature before use. For Au nanoparticles, 1.0 mL of 10 mM HAuCl4 was added to 9.0 mL of 25 mM sodium citrate solution under vigorous stirring for 10 min at room temperature. 3319
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C Then 0.5 mL of freshly prepared aqueous 0.05 M of NaBH4 in 25 mM sodium citrate solution was added dropwise to the above solution. The color of the solution changed to pink immediately. The final solution was stirred for 1 h and aged for 24 h to decompose the residual NaBH4 at room temperature. The extinction spectra of these Ag and Au nanoparticles are shown in the Supporting Information (Figure S2), with distinct surface plasmon absorption bands around 408 and 530 nm, characteristic of Ag and Au nanoparticles, respectively. Extinction Spectra and Particle Size Analyzer Measurements. The extinction spectra were recorded on a Shimadzu 2550 spectrophotometer (Shimadzu Corp., Japan) using a 1 cm path length quartz cuvette at room temperature. Particle size analyzing was carried out by dynamic light scattering with a Nanotrac particle size analyzer (PSA) from Microtrac, Inc. The average values of the particle size and polydispersity, defined as a relative width of the size distribution, were determined from the PSA measurements. TEM and EDX Measurements. A drop of well-sonicated solution containing the nanoparticles was deposited on a Formvarcovered carbon-coated copper grid (Electron Microscopy Sciences, PA). The samples were allowed to dry at room temperature overnight. A JEOL 2010 high-resolution transmission electron microscope (HRTEM) was used to obtain the TEM images and EDX spectra at 200 kV. Raman Measurements. The Raman reporter-embedded TiO2 core-Au/Ag shell nanoparticles were dispersed thoroughly in DI water by sonication. For the liquid samples, an argon ion laser (Melles Griot) with multiple laser wavelengths was used. The laser beam was focused by a 10 objective onto the sample contained in a 1 cm path length quartz cell. Raman signals were collected by the same objective, passed through an appropriate notch filter, and coupled into a spectrometer (SpectraPro-2300i, Acton Research, MA) equipped with an air-cooled CCD detector (Spec-10, Roper Scientific, NJ). Laser intensities at the samples were approximately 9.2 mW for the 488 nm line and 9.5 mW for the 514 nm line. For the solid samples, a drop of the diluted solution was cast onto a tilted, clear coverslip and allowed to dry at room temperature. Raman measurements were carried out on a LabRAM Raman microscope (HORIBA Jobin Yvon Inc., NJ). Laser intensities at the samples were set at approximately 0.13 mW for a 632.8 nm HeNe laser as excitation source. Between different Raman sessions, the 520.7 cm-1 peak of a silicon wafer was used to calibrate the spectrograph and to normalize the measured intensity to compensate for possible fluctuation of the Raman system. Exposure time for all measurements was 1 s. Each spectrum was the average of 10 scans. The nanoparticle concentrations in the Raman measurements are estimated to be on the order of 0.1 μM.
’ RESULTS AND DISCUSSION General Strategy of Nanoparticle Synthesis. The scheme for preparing the Raman reporter-embedded TiO2 core-Au/Ag shell nanoparticle is illustrated in Figure 1. We recently reported a general strategy to synthesize such TiO2 core-metal shell nanoparticles based on the different surface charges of TiO2 nanoparticles under different pH environments.27 Here we follow this strategy to first synthesize the reporter-embedded TiO2 coreAg shell nanoparticles, using RuBpy and TCPP as the reporter molecules. Then the replacement reaction between the chloroaurate ions and the Ag shell on the nanoparticle surface lead to
ARTICLE
Figure 1. Preparation of Au/Ag-coated TiO2-based core-shell nanoparticles: (i) hydrolysis of TiCl4 in 0.1 M HCl to form colloidal TiO2 solution; (ii) addition of SERS dyes; (iii) controlled aggregation using NaOH; (iv) growth of Ag shell using AgNO3; (v) growth of Au shell using HAuCl4.
the deposition of Au onto the Ag shell, forming a mixed Au/Ag shell. By controlling the amounts of chloroaurate ions, different compositions of Au/Ag shell can be achieved. This scheme has proved to be versatile toward many Raman reporter molecules with different charge properties. Colloidal TiO2 nanoparticles prepared by our method usually result in an average diameter of less than 3 nm. Using the molecular TiO2 molar extinction coefficient (6050 M-1 cm-1) at 215 nm, we can calculate the total TiO2 concentration in the solution. During the synthesis of the Raman reporter-nanoparticles, we can estimate the TiO2 particle concentration following the method described in ref 28, using the TiO2 density of 4 g/cm3 and an average diameter of 3 nm. Extinction Spectra of Raman Reporter-Embedded TiO2 Core-Au/Ag Shell Nanoparticles. Figure 2 shows the extinction spectra of TiO2, TCPP, RuBpy, and their various combinations. Similar to our previous report, the spectra indicate the adsorption of dyes to the colloidal TiO2 nanoparticles. The absorption peak of free TCPP at 405 nm is red-shifted to 418 nm for TCPP-TiO2, while for RuBpy-TiO2 the absorption peak of free RuBpy at 452 nm is also slightly red-shifted to 456 nm. The different shifts of the absorption peak suggest the different extents of charge transfer between the respective dye and TiO2. Since TiO2 nanoparticles are porous, the dye molecules probably penetrate throughout the pores of the TiO2 nanoparticles. There is no indication of dye molecules being desorbed from the TiO2 nanoparticles during the metal coating process, as the supernatant after the metal coating does not contain free dye molecules (by checking the dye fluorescence). The coating of Au shell on top of Ag shell is further confirmed by the appearance of the Au shell plasmon band at ∼530 nm and the decrease of the Ag shell plasmon band at ∼408 nm. In the case of RuBpy-embedded nanoparticles, the RuBpy absorption band seems to be overwhelmed by the metal plasmon band after the metal coating. Still, there are RuBpy molecules entrapped underneath the metal shell, as indicated in the SERS measurement results shown later. The less intense RuBpy absorption may be due to less RuBpy molecules adsorbed to the TiO2 core, as the TiO2-RuBpy interaction is not as strong as that of TiO2TCPP. Again, the Au/Ag-coated RuBpy-embedded TiO2 nanoparticles distinctively display the combined surface plasmon features of both Ag and Au shells. TEM, EDX, and PSA Characterizations of Raman ReporterEmbedded TiO2 Core-Au/Ag Shell Nanoparticles. TEM 3320
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C
ARTICLE
Figure 2. Extinction spectra of TiO2, TCPP, RuBpy, and their combinations: (a) TCPP as Raman reporter; (b) RuBpy as Raman reporter.
Figure 3. TEM images of (a) TiO2, (b) TiO2-TCPP, (c) Ag-TiO2-TCPP, (d) Au-Ag-TiO2-TCPP, (e) TiO2-RuBpy, (f) Ag-TiO2-RuBpy, and (g) Au-Ag-TiO2-RuBpy. All scale bars represent 20 nm.
and EDX results provide more evidence to support the formation of Raman reporter-embedded TiO2 core-Au/Ag shell nanostructures. Figure 3 shows the typical TEM images of the (a) TiO2, (b) TiO2-TCPP, (c) Ag-TiO2-TCPP, (d) Au-AgTiO2-TCPP, (e) TiO2-RuBpy, (f) Ag-TiO2-RuBpy and (g) Au-Ag-TiO2-RuBpy. The colloidal TiO2 nanoparticles are rather uniform. After the adsorption of dyes and the coating of Ag shells, the particle sizes increase slightly. EDX data collected on
the nanoparticles in respective samples (see Figure S3 in the Supporting material) clearly indicate the presence of the major elements expected in the final reporter-embedded TiO2 coreAu/Ag shell nanoparticles. PSA results show the particle size distribution of TiO2, TiO2-TCPP, Ag-TiO2-TCPP, and AuAg-TiO2-TCPP. Figure 4 shows the typical PSA histograms of (a) TiO2, average 1.4 nm; (b) TiO2-TCPP, average 2.5 nm; (c) Ag-TiO2-TCPP, average 3.7 nm; (d) Au-Ag-TiO2-TCPP, 3321
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C
ARTICLE
Figure 4. Typical PSA histograms of (a) TiO2 (average 1.4 nm), (b) TiO2-TCPP (average 2.5 nm), (c) Ag-TiO2-TCPP (average 3.7 nm), and (d) Au-Ag-TiO2-TCPP (average 5 nm).
average 5 nm. All are consistent with the TEM results. On the basis of the TEM and PSA data of the nanoparticles before and after the metal coating, the metal shell thickness can be estimated to be ∼2-3 nm. Determination of the Average SERS EFs. Determination of the average SERS EFs can be carried out by a few different ways. In one method, we compare the SERS signals of these reporterembedded TiO2 core-Au/Ag shell nanoparticles with the fluorescence signals of reporter-embedded TiO2-core nanoparticles. Take the RuBpy-embedded TiO2 core-Au/Ag shell nanoparticles as an example. After the SERS measurement under 488 nm excitation, we add a small amount of NaCN, which is in excess compared to the Au/Ag in the solution, to the cuvette. Immediately the RuBpy SERS peaks disappear and the fluorescence band corresponding to RuBpy (∼585 nm, see Figure S4 in the Supporting material) appears. This is due to the reaction between Au/Ag and CN- ions, resulting in the removal of the metal shells. While maintaining the same laser intensity and spectra integration time, the intensities of the fluorescence signals after metal shell removal and the SERS signals with the metal shell intact are observed to be at the same order of magnitude. The fluorescence cross section of common dyes, RuBpy included, is typically ∼10-16 cm2. The Raman cross section of a molecule very similar to RuBpy, under non-surface enhanced but resonance Raman process, was reported to be on the order of 10-26 cm2.31 Note that our RuBpy measurements are also under resonant conditions. Thus the average SERS EFs for our RuBpy-embedded TiO2 core-Au/Ag shell nanoparticles can be calculated to be on the order of 1010. This result is in line
with the conclusions from the recent in-depth investigations in the literature.22 A similar treatment on the TCPP-embedded TiO2 core-Au/Ag shell nanoparticles yields the average SERS EFs on the order of 108. The difference in EFs between the two systems can be accounted for by considering that the measurements for the RuBpy system are under resonant conditions while those for the TCPP systems are not. In a second method, we carry out a different experiment to help determine the average SERS EFs of these RuBpy-embedded TiO2 core-Au/Ag shell nanoparticles. Here, different concentrations of pure RuBpy solution are mixed with Ag nanoparticleonly solution as described in the Experimental Section, and the SERS spectra are taken under the same conditions (laser intensity, integration time, etc.). As shown in Figure 5a, the SERS intensity of the RuBpy-embedded TiO2 core-Au/Ag shell nanoparticles is slightly higher than that of the RuBpy-Ag nanoparticle mixture (containing approximately the same amount of Ag) when the RuBpy concentration in the mixture is 5 μM. Independently, based on the fluorescence signals obtained in the RuBpyembedded TiO2 core-Au/Ag shell nanoparticles treated with CN- ions, we can estimate that the concentration of RuBpy molecules in the mixture to be ∼10-11 M (see Figure S4 in the Supporting Information). Thus, the average SERS EFs of these RuBpy-embedded TiO2 core/Au/Ag shell nanoparticles are ∼106 higher than that of the Ag-nanoparticle-only solution. Next, we determine the average SERS EF of the Ag-nanoparticle-only system, following the widely accepted method in the literature.32 The results are shown in Figure 5b, where SERS signals from 10-5 M free RuBpy solution and 10-10 M free 3322
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C
ARTICLE
Figure 5. SERS spectra of (a) RuBpy-embedded TiO2 core-Au/Ag shell nanoparticle solution and 5 μM of free RuBpy solution mixed with Ag-nanoparticle-only solution and (b) 10-5 M free RuBpy solution and 10-10 M free RuBpy solution mixed with Ag-nanoparticle-only solution, under 514 nm excitation.
Figure 6. SERS spectra of (a) RuBpy-embedded TiO2 core-Ag shell nanoparticles solution containing RuBpy molecules equivalent to 10-11 M and (b) TiO2 core-Ag shell nanoparticles solution mixed with 10-7 M free RuBpy, under the same experimental conditions.
RuBpy solution mixed with Ag nanoparticle only are similar. This leads to a SERS EF of ∼105 for the Ag-nanoparticle-only system, indicating the SERS EFs for the RuBpy-embedded TiO2 coreAu/Ag shell nanoparticles to be on the order of ∼1011. In a third method, we directly compare the ensemble SERS EF of the TiO2 core-Ag shell nanoparticles with the Raman reporter molecules either inside or outside the metal shell. Here, we have prepared TiO2 core-Ag shell nanoparticles in a similar manner as before but without the Raman reporter molecules (RuBpy) embedded inside the TiO2 core. Instead, we add the RuBpy solution to the TiO2 core-Ag shell nanoparticles after they are synthesized, washed, and redispersed. The results collected under the same experimental conditions, as shown in Figure 6, indicate the ensemble EF to be >104 higher when the RuBpy molecules are inside the metal shell. We should emphasize that such enormous ensemble EFs are achieved as average EFs in aqueous solution, which are orders of magnitude higher than those usually reported in the literature.22 This is directly in contrast to many other studies where similarly or even higher EFs were reported. Those studies tend to focus on
Figure 7. SERS spectra of TCPP-embedded TiO2 core-Au/Ag shell nanoparticles solution under 514 nm laser excitation. The numbers in parentheses indicate the molar ratio of Au/Ag. Insert: Raman intensity at 1474 cm-1 vs Au/Ag molar ratio.
the SERS EFs of some “hot spots”, which are sporadically distributed within the substrate. When taking into consideration the broad distribution of EFs present throughout the substrate, the average EFs of those systems, such as the Ag and Au nanoparticles used as the reference systems in this study, are easily brought down by a few orders of magnitude. In contrast, the core region of the TiO2 core-metal shell nanostructures reported here has uniform field enhancement, as supported by results from theoretical modeling shown hereafter and in the literature.33 With the interior of such nanostructures being a “hot zone”, all embedded Raman reporter molecules will experience Raman enhancement and contribute to the high average SERS EFs. Last, but not least, we note that such TiO2 core-metal shell nanostructures can be constructed and prepared easily and reproducibly. Combining the features of very high and reproducible ensemble EFs, versatility in embedding various reporter molecules, and good stability, these nanostructures can be superior SERS tags with multiplexing capability and can provide 3323
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C an alternative to the fluorescence-based tags for bioimaging and labeling applications. Effect of Shell Composition on the SERS of Raman Reporter-Embedded TiO2 Core-Au/Ag Shell Nanoparticles. In an effort to combine the high EFs of Ag shell and the good stability of Au shell, we introduce the mixed Au/Ag shells in place of the single metal (Ag or Au) shells into the nanostructures. Results of Raman measurements of the reporter-embedded TiO2 core-Au/Ag shell nanoparticles are shown in Figures 7 and 8, under laser excitations of different wavelengths. It can be seen that all reporter-embedded TiO2 core-Au/Ag shell nanoparticles display very strong SERS, similar to those SiO2 core and TiO2 core-Au shell nanoparticles reported previously.3,27 It is usually believed that silver is a better material than gold as SERS substrates, an observation also confirmed in our own experiences. In that regard, it is not surprising in observing that, for the TCPP-embedded TiO2 core-Au/Ag shell nanoparticles under 514 nm excitation, the SERS intensity decreases with the increasing thickness of Au shell (Figure 7 inset). However, in the case of RuBpy-embedded TiO2 coreAu/Ag shell nanoparticles under 488 nm excitation, the SERS intensity first increases with the increasing thickness of Au shell
ARTICLE
and then decreases with the thicker Au shell (Figure 8 inset), contrary to the conventional behavior that Au, compared to Ag, would reduce the EF. Simulation of SERS Enhancement of the Raman ReporterEmbedded TiO2 Core-Au/Ag Shell Nanoparticles. In order to further understand the behavior of the SERS enhancement factor as the metal composition changes, as shown in Figures 7 and 8, we model the electromagnetic enhancement to include the effect of Au/Ag molar ratio in the dielectric constant. Such changes in SERS EFs should be attributed to the modification of the electromagnetic enhancement, as the material composition of the metal shell only influences the optical properties of the surrounding environment of the Raman molecules but not the intrinsic properties of the molecules. The electromagnetic enhancement of SERS is partly due to the enhancement of local field intensity at the location of the Raman molecule and partly due to the enhancement of the radiation emission rate of the Raman molecule.2,34 Extensive work in computational electrodynamics would be required to exactly determine the SERS EFs,35 which is out of the scope of this paper. Here we use the electric field enhancement in the core region as an indicator to study how the SERS intensity changes with the Au/Ag molar ratio. As the particles in our experiment are very small in size, we adopt the quasi-static approximation and numerically calculate the electric field enhancement in the core region of the nanoparticle by solving Laplace’s equation for the electric potential. The dielectric function of the metal shell is determined according to a semiempirical model based on the Drude theory and the related experimental data36 εðωÞ ¼ εðωÞbulk þ
Figure 8. SERS spectra of RuBpy-embedded TiO2 core-Au/Ag shell nanoparticles solution under 488 nm laser excitation. The numbers in parentheses indicate the molar ratio of Au/Ag. Insert: Raman intensity at 1481 cm-1 vs Au/Ag mole ratio.
ωp 2 ωp 2 ω2 þ iωγ0 ω2 þ iωðγ0 þ ΔγÞ
where ε(ω)bulk is the experimental dielectric function value of the bulk material,37 ωp is the plasma frequency, γ0 is the bulk collision frequency, and Δγ is the electron-surface collision frequency. The Au/Ag shell is modeled as a homogeneous material with physical properties obtained by averaging those of Au and Ag based on the molar ratio used in the preparation.38 Figure 9 shows the real and imaginary parts of the equivalent dielectric function of the Au/Ag shell when the molar ratio is set as 1/25 and 1/3, respectively. The dielectric function of the core material is treated in the similar manner by averaging those of TiO2 (rutile phase) and TCPP/RuBpy. Our calculation shows that the electric field enhancement occurs both outside and inside the shell. For the outer electric field,
Figure 9. (a) Real part and (b) imaginary part of the equivalent dielectric function of metal shell when the Au/Ag molar ratio is set as 1/9 (blue curves). The red curve and black curve show the experimental dielectric function values of the bulk material of Au and Ag, respectively.37. 3324
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C
ARTICLE
’ ASSOCIATED CONTENT
bS
Supporting Information. Molecular structures of TCPP and RuBpy, extinction spectra of Ag and Au nanoparticles, EDX spectra of TiO2, TiO2-TCPP, Ag-TiO2-TCPP, and Au-AgTiO2-TCPP, and fluorescence spectra of pure RuBpy and of RuBpy-embedded TiO2 core Au/Ag shell nanoparticles after NaCN treatment. This information is available free of charge via the Internet at http://pubs.acs.org. Figure 10. The fourth power of electric field enhancement (normalized) as a function of Au/Ag molar ratio at 514 and 488 nm, respectively.
large field enhancement only occurs along the polarization direction of the incident light, and it decays rapidly with the distance away from the metal surface. These are consistent with the results calculated using a discrete dipole approximation (DDA) approach.33 To further study the influence of Au/Ag molar ratio on SERS intensity, we examine the electric field enhancement at 514 and 488 nm, which are the two laser wavelengths used in the SERS experiments with the TCPP-embedded and RuBpy-embedded nanoparticles, respectively. The normalized electric field enhancement as a function of Au/Ag molar ratio is shown in Figure 10. At 514 nm, the field enhancement monotonically decreases when the Au/Ag molar ratio increases from 0 to 1. However, at 488 nm, the field enhancement first increases when the Au/Ag molar ratio increases from 0 to approximately 0.09, reaching a maximum, and then starts to decrease when the Au/Ag molar ratio continues to increase toward 1. These trends are in a very good agreement with the experimental results shown in Figure 7 and Figure 8. It is worth mentioning that the significant influence of the Au/Ag molar ratio on the electric field enhancement only occurs in the wavelength range between 450 and 500 nm, where the surface plasmon resonance locates. In the longer wavelength range, change of Au/Ag molar ratio does not influence the electric field enhancement much and thereby the SERS EFs. To sum up, the simulation results confirm that the SERS intensity of the TiO2 core-mixed metal shell nanoparticles can be tuned by adjusting the Au/Ag molar ratio of the shell. It also suggests that theoretical modeling could provide insight for designing an optimized core-shell nanoparticle with maximum SERS intensity.
’ CONCLUSIONS In conclusion, we report the synthesis of Raman reporterembedded TiO2 core-Au/Ag shell nanoparticles, which display enormous ensemble SERS EFs of the encapsulated reporter molecules. Au/Ag shells are formed by the replacement reaction between the preformed Ag shell and AuCl4- ions. Average SERS EFs on the order of 108-1010 can be achieved easily and reproducibly. Under some excitation wavelengths, the mixed Au/Ag shells are shown to display higher ensemble SERS EFs than the Ag shells, an observation supported by our simulation results. Theoretical modeling proves to be able to provide insight for the design of an optimized core-shell nanoparticle with maximum SERS intensity. The Raman reporter-embedded TiO2 core-Au/Ag shell nanoparticles should have great potential as efficient SERS tags in labeling and imaging applications.
’ AUTHOR INFORMATION Corresponding Author
*E-mail:
[email protected].
’ ACKNOWLEDGMENT Support from the Natural Science Foundation (CHE0632071) and the National Center for Research Resources (NCRR) of the National Institutes of Health (RR-016480) is gratefully acknowledged. This work was performed, in part, at the Center for Integrated Nanotechnologies, a U.S. Department of Energy, Office of Basic Energy Sciences user facility at Los Alamos National Laboratory (Contract DE-AC52-06NA25396) and Sandia National Laboratories (Contract DE-AC04-94AL85000). ’ REFERENCES (1) Fleischmann, M.; Hendra, P. J.; McQuillan, A. J. Chem. Phys. Lett. 1974, 26, 163. (2) Moskovits, M. Rev. Mod. Phys. 1985, 57, 783. (3) Zhang, P.; Guo, Y. J. Am. Chem. Soc. 2009, 131, 3808. (4) Jadzinsky, P. D.; Calero, G.; Ackerson, C. J.; Bushnell, D. A.; Kornberg, R. D. Science 2007, 318, 430. (5) Kneipp, K.; Kneipp, H.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Chem. Rev. 1999, 99, 2957. (6) Campion, A.; Kambhampati, P. Chem. Rev. 1998, 27, 241. (7) Constantino, C. J. L.; Lemma, T.; Antunes, P. A.; Aroca, R. Anal. Chem. 2001, 73, 3674. (8) Ren, B.; Lin, X. F.; Yang, Z. L.; Liu, G. K.; Aroca, R. F.; Mao, B. W.; Tian, Z. Q. J. Am. Chem. Soc. 2003, 125, 9598. (9) Dou, X.-M.; Ozaki, Y. Rev. Anal. Chem. 1999, 18 (4), 285. (10) Zhang, P.; Smith, S.; Rumble, G.; Himmel, M. Langmuir 2005, 21, 520. (11) Moskovits, M. J. Raman Spectrosc. 2005, 36, 485. (12) Pavel, I.; McCarney, E.; Elkhaled, A.; Morrill, A.; Plaxco, K.; Moskovits, M. J. Phys. Chem. C 2008, 112 (13), 4880. (13) Alvarez-Puebla, R. A.; dos Santos, D. S.; Aroca, R. F. Analyst 2007, 132, 1210. (14) Braun, G.; Lee, S. J.; Dante, M.; Nguyen, T.; Moskovits, M.; Reich, N. J. Am. Chem. Soc. 2007, 129, 637. (15) Premasiri, W. R.; Moir, D. T.; Klempner, M. S.; Krieger, N.; Jones, G; Ziegler, L. D. J. Phys. Chem. B 2005, 109 (1), 312. (16) Penn, S. G.; He, L.; Natan, M. J. Curr. Opin. Chem. Biol. 2003, 7, 609. (17) Cao, Y. C.; Jin, R.; Mirkin, C. A. Science 2002, 297, 1536. (18) Qian, X.; Peng, X.-H.; Ansari, D. O.; Yin-Goen, Q.; Chen, G. Z.; Shin, D. M.; Yang, L.; Young, A. N.; Wang, M. D.; Nie, S. Nat. Biotechnol. 2008, 26 (1), 83. (19) Schatz, G. C.; Van Duyne, R. P. In Handbook of Vibrational Spectroscopy; Chalmers, J. M., Griffiths, P. R., Eds.; Wiley & Sons: New York, 2002. (20) Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R.; Feld, M. S. Phys. Rev. Lett. 1997, 78, 1667. (21) Nie, S. M.; Emory, S. R. Science 1997, 275, 1102. 3325
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326
The Journal of Physical Chemistry C
ARTICLE
(22) Le Ru, E. C.; Blackie, E.; Meyer, M.; Etchegoin, P. G. J. Phys. Chem. C 2007, 111, 13794. (23) Cai, W. B.; Ren, B.; Li, X. Q.; She, C. X.; Liu, F. M.; Cai, X. W.; Tian, Z. Q. Surf. Sci. 1998, 406, 9. (24) Felidj, N.; Aubard, J.; Levi, G.; Krenn, J. R.; Salerno, M.; Schider, G.; Lamprecht, B.; Leitner, A.; Aussenegg, F. R. Phys. Rev. B 2002, 65, No. 075419. (25) Lal, S.; Grady, N. K.; Goodrich, G. P.; Halas, N. J. Nano Lett. 2006, 6, 2338. (26) Grabar, K. C.; Freeman, G. R.; Hommer, M. B.; Natan, M. J. Anal. Chem. 1995, 67 (4), 735and references therein. (27) Li, W.; Guo, Y.; Zhang, P. J. Phys. Chem. C 2010, 114, 7263. (28) Li, W.; Gandra, N.; Ellis, E; Courtney, S.; Li, S; Bulter, El; Gao, R. ACS Appl. Mater. Interfaces 2009, 1, 1778. (29) Zou, X.; Dong, S. J. Phys. Chem. B 2006, 110, 21545. (30) Jana, N. R.; Gearheart, L.; Murphy, C. J. Langmuir 2001, 17 (22), 6782. (31) Shoute, L. C. T.; Loppnow, G. R. J. Am. Chem. Soc. 2003, 125, 15636. (32) McFarland, A. D.; Young, M. A.; Dieringer, J. A.; Duyne, R. P. V. J. Phys. Chem. B 2005, 109, 11279. (33) Hao, E.; Li, S.; Bailey, R. C.; Zou, S.; Schatz, G. C.; Hupp, J. T. J. Phys. Chem. B 2004, 108, 1224. (34) Schatz, G. C.; Young, M. A.; Van Duyne, R. P. Top. Appl. Phys. 2006, 103, 19. (35) Goude, Z. E.; Leung, P. T. Solid State Commun. 2007, 143, 416. (36) Averitt, R. D.; Saker, D.; Halas, N. J. Phys. Rev. Lett. 1997, 78, 4217. (37) Johnson, P. B.; Christy, C. Phys. Rev. B 1972, 6, 4370. (38) Cottancin, E.; Lerme, J.; Gaudry, M.; Vialle, J.-L.; Broyer, M.; Prevel, B.; Treilleux, M.; Melinon, P. Phys. Rev. B 2000, 62, 51.
3326
dx.doi.org/10.1021/jp1116849 |J. Phys. Chem. C 2011, 115, 3318–3326