InSitu SurfaceEnhanced Raman Spectroscopy ... - Wiley Online Library

DOI: 10.1002/cctc.201403032

Full Papers

In Situ Surface-Enhanced Raman Spectroscopy Study of Plasmon-Driven Catalytic Reactions of 4-Nitrothiophenol under a Controlled Atmosphere Leilei Kang,[a] Xijiang Han,*[a] Jiayu Chu,[a] Jie Xiong,[b] Xiong He,[a] Hsing-Lin Wang,*[c] and Ping Xu*[a, b] We demonstrate the plasmon-driven catalytic reactions of 4-nitrothiophenol (4NTP) on a single Ag microsphere by an in situ surface-enhanced Raman spectroscopy (SERS) technique. The highly SERS-active hierarchical Ag structures served as an ideal platform to study plasmon-driven catalytic reactions. This single-particle surface-enhanced Raman spectroscopy (SPSERS) technique coupled with inbuilt apparatus allow us to study the impact of reaction atmospheres and laser power on the rate of dimerization and reduction of 4NTP. Contrary to

that found in previous studies, 4NTP could be transformed into 4-aminothiophenol under H2O or H2 atmosphere. The broadening and splitting of the n(CC) band during the reaction results from the frequency shift of the n(CC) band that arises from different products. Our results suggest that the SPSERS technique is ideally suited to study plasmon-driven catalytic reactions because of the possibility to monitor the reaction under controlled atmospheres in real time.

Introduction In recent years, the rapid development of nanoscience and nanotechnology has attracted increasing attention because of the unique optical, electrical, and electromagnetic properties of nanomaterials.[1] The design and fabrication of hierarchical structures assembled from nanoscale building blocks are of great importance as such structures are endowed with unique properties that are distinctly different from those of the individual components.[2] Unfortunately, the controlled synthesis of nanostructured materials with desirable properties remains a great challenge. Various approaches have been attempted to promote practical applications of nanostructured materials. Transparent and opaque Ag hydrogels and aerogels have been fabricated by the oxidation-induced self-assembly of Ag nanoshells with tunable plasmon bands in the visible spectrum.[3] Layer-by-layer self-assembled TiO2 hierarchical nanosheets with exposed {0 0 1} facets have been reported as an effective bi-

[a] L. Kang, Prof. X. Han, J. Chu, X. He, Prof. P. Xu Department of Chemistry Harbin Institute of Technology Harbin 150001 (China) E-mail: [email protected] [email protected] [b] Prof. J. Xiong, Prof. P. Xu State Key Laboratory of Electronic Thin Films and Integrated Devices University of Electronic Science and Technology of China Chengdu 610054 (China) [c] Dr. H.-L. Wang Chemistry Division Los Alamos National Laboratory Los Alamos, NM 87545 (USA) E-mail: [email protected] Supporting Information for this article is available on the WWW under http://dx.doi.org/10.1002/cctc.201403032.

ChemCatChem 2015, 7, 1004 – 1010

functional layer for dye-sensitized solar cells.[4] Hollow microspheres assembled by VO2 nanowires exhibit a high capacity and excellent cyclability because of their high surface area and efficient self-expansion and self-shrinkage buffering.[5] The successful synthesis of monodispersed Au nanotriangles leads to extended self-assembly at the air–liquid interface, which allows the development of a promising surface-enhanced Raman scattering (SERS) substrate.[6] Uniform Ag microspheres with a nanoscale surface roughness have been synthesized as SERS substrates by the addition of poly(vinylpyrrolidone),[7] small molecular acids,[8] or amino acids[9] as capping agents. The development of nanomaterials for practical applications requires synthetic methods to shift gradually from what it is to what we design. Thus, an understanding of how the chemical structures (types of functional groups, number of carbon atoms, etc.) of capping agents impact the assemblies with well-defined morphologies is of great interest. The renaissance of Raman spectroscopy depends heavily on the discovery of the SERS phenomenon related to substrates with roughened surfaces. For a long time, SERS was deemed as a noninvasive technique for chemical detection,[10] which means the analyte will remain intact under irradiation from the light source (i.e., laser) in Raman spectroscopy. Thus chemical enhancement (charge transfer) mechanisms were interpreted from the modified Raman spectrum of 4-aminothiophenol (4ATP) in 1994.[11] These new peaks were considered as a robust experimental evidence of charge transfer mechanisms.[12] p,p’-Dimercaptoazobenzene (DMAB) transformed from 4ATP by a surface catalytic coupling reaction on Ag nanoparticles was predicted theoretically[13] and confirmed by experimental study.[14] Since then, a number of publications have validated this hypothesis through combined theoretical

1004

2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Papers models and experimental methods.[9, 15] More recently, it was discovered that DMAB can also be produced from 4-nitrothiophenol (4NTP) with the assistance of so-called surface-plasmon resonance.[16] The conversion of 4ATP into DMAB can be realized under high vacuum, and 4-nitrothioanisole (4NTA) can be dimerized into DMAB in an electrochemical aqueous environment.[17] After the discovery of the dimerization of amino or nitro groups into an azo group, various catalytic reactions on SERS substrates have been realized. For instance, ethylene epoxidation, CO oxidation, and NH3 oxidation have been reported on plasmonic Ag nanostructures;[18] a prototypical stilbene photoreaction can be monitored in situ through SERS on selfassembled Au nanoparticles linked by the sub-nm macrocycle cucurbit[n]uril (CB[n]);[19] molecular synthesis through the removal or decoration of functional groups can be realized by using so-called “plasmonic scissors”;[20] the conversion of Fe3+ to Fe2+ by hot electrons generated plasmonically from Ag nanoparticles has been evidenced by SERS;[21] the plasmon-induced dissociation of H2 and the doping of graphene by hot electrons generated from Au nanoparticles have also been witnessed.[22] However, the influence of gas components in air, which can impact the plasmon-assisted catalytic reaction on metal nanoparticles, has yet to be studied carefully. Recently, we investigated the underlying mechanism of the surface-plasmon-assisted catalytic reaction of 4ATP and DMAB systematically under different atmospheres using our self-designed gas flow cell.[23] Herein, we report the in situ SERS monitoring of the surfaceplasmon-driven catalytic reactions of 4-NTP under controlled atmospheres. We have taken advantage of our single-particle surface-enhanced Raman spectroscopy (SP-SERS) technique reported previously[16c, 23] to investigate the influence of various gas components (N2, O2, H2, H2O) on the plasmon-driven reaction of 4NTP using Ag microspheres as a single reactor. We use a series of dicarboxylic acids as directing agents for the synthesis of hierarchical Ag microspheres to investigate how the carbon-chain length impacts the nanostructural features on these Ag microsphere surfaces. Contrary to earlier reports,[16a, b] 4NTP is not just dimerized into DMAB, but it can be further converted into 4ATP under H2 or H2O atmospheres. We believe that plasmon-driven reactions performed under controlled atmospheres may provide insights into the emerging field of plasmon-assisted photocatalytic reactions.

Results and Discussion Our previous work has verified that the acids introduced into the chemical reaction system can effectively direct the growth of Ag microspheres with well-defined nanostructures.[8] The subtle roughness at the surface of these Ag microspheres can provide “hot spots” for SERS detection. Of particular importance is that as-prepared Ag microspheres are large enough for the optical microscope on the confocal Raman spectrometer to focus on such that a single Ag particle can readily serve as an independent SERS substrate, and we have named this technique SP-SERS.[23] Recently, we have found that some dicarboxylic acids are ideal directing agents for the direct formation ChemCatChem 2015, 7, 1004 – 1010

www.chemcatchem.org

of hierarchical structures that consist of Ag nanosheets, which are not accessible by using monocarboxylic acids. Monocarboxylic acids lead to Ag microspheres with a smooth surface, which resemble the morphology of Ag particles prepared without any capping agents (Figure S1). Comparative studies between mono- and dicarboxylic acids suggest that monocarboxylic acids do not possess the functionality to direct Ag nanoparticles to assemble into specific structures. The morphology evolution of Ag microspheres in the presence of a set of dicarboxylic acids is shown in Figure 1. All pro-

Figure 1. SEM images of Ag microspheres fabricated with the assistance of a) malonic acid, b) succinic acid, c) glutaric acid, and d) adipic acid.

duced Ag particles are actually assembled by numerous nanosheets of ~ 80 nm thick. If malonic acid was introduced into the reaction solution, a Ag architecture with a high yield and uniform size can be prepared (Figure 1 a). Nevertheless, a large amount of half-finished Ag particles appeared if adipic acid was used as the directing agent (Figure S2). We observe a trend that the yield and uniformity of Ag microspheres became worse with the increase in the number of carbon atoms between the two carboxylic groups, which can be observed from the Gaussian nonlinear fitting of the size histograms of the corresponding Ag particles. This phenomenon can be rationalized by the decreased solubility in aqueous solution with increasing carbon-chain length, which suppress the dispersion of capping agents and further affects the yield and size distribution of the Ag microparticles. Pimelic acid, with five carbon atoms between the two carboxylic groups, cannot be dissolved in water. The good solubility of oxalic acid should have the right structure and functional groups to direct nanosheet self-assembly to produce uniform Ag microspheres. To

1005


Full Papers our surprise, the experimental results show that oxalic acid leads to scattered Ag nanosheets and loosely packed Ag assemblies that consist of numerous nanosheets, without the formation of a well-assembled spherical morphology (Figure S3). The magnified SEM images of these Ag particles reveal that the stacking pattern of Ag microspheres can be affected by the carbon-chain length (Figure 1 a–d). Typically, more closely packed Ag particles can be obtained through the use of a dicarboxylic acid with more carbon atoms. We found that hydroxyl groups on the carbon chain of the dicarboxylic acid can also affect the morphology of the Ag microspheres. With malic acid and tartaric acid as the capping agent, the surface of the Ag microspheres presents grainlike nanostructures instead of nanosheets (Figure S4). XRD patterns show typical face-centered cubic (fcc) Ag crystal phases (Figure S5), in which the diffraction peaks can be indexed to the (111), (2 0 0), (2 2 0), (3 11), and (2 2 2) crystal planes of Ag (PDF #65-2871). The almost identical intensity ratio of the diffraction peaks implies that the change of carbon-chain length would not induce anisotropic growth along a specific crystal plane but simply vary the assembly of Ag nanosheets.[8] A statistical study shows that the range of the size distribution of Ag microspheres becomes wider with the increase of the number of carbon atoms between the two carboxylic groups of the directing agents, and oxalic acid is the exception (with which no Ag microspheres are produced). According to the statistical data, we see that the yield of Ag microspheres is a function of the number of carbon atoms (Figure 2). This shows that as we increase the number of carbon atoms between the two carboxylic groups, the yield of well-defined spherical particles was reduced gradually. The plasmon-driven catalytic dimerization of 4NTP to DMAB on Au, Ag, and Cu films shows very different behaviors, which indicates that this reaction is substrate dependent.[24] Here, Ag microspheres directed by dicarboxylic acids, which consist of the same material but a different morphology of nanosheet ensembles, are first tested for their SERS response to find the one with the strongest enhancement for this research. Well-re-

Figure 2. Yield of Ag microspheres as a function of the number of carbon atoms between the two carboxylic groups.

ChemCatChem 2015, 7, 1004 – 1010

www.chemcatchem.org

solved Raman spectra can be obtained for Rhodamine B at a concentration of 105 m on all Ag samples (Figure S6), but Ag particles directed by longer dicarboxylic acids have relatively lower Raman intensities. Ag microspheres directed by malonic acid were selected because of their highly sensitive SERS response. We used the SP-SERS technique to investigate the influence of different atmospheres on the plasmon-driven catalytic reaction of 4NTP by using a gas-control flow cell (Figure 3). The main advantage of SP-SERS is that the efficient

Figure 3. Schematic illustration of the gas flow cell for the plasmon-induced catalytic reactions on a single particle with the in situ SERS technique.

Raman signal comes from a single Ag particle, which eliminates interference from the surrounding environment outside the laser spot. Raman images recorded on a single particle allow us to track the intensity variations of the characteristic bands during the reaction processes (Figure S7). The dimerization of 4NTP into DMAB is a reduction reaction, but the presence of O2 in air is not enough to hinder this reaction. However, if pure O2 is introduced into the reaction system, the intensity of the n(NO2) band at n˜ = 1335 cm1 decreases slightly, but the n(N=N) bands at n˜ = 1380 and 1440 cm1 can hardly be observed after irradiation by a continuous 633 nm laser (1.5 mW) for 20 min (Figure 4 a). The reduction of 4NTP is halted in a saturated O2 environment. In contrast to O2, we find that such a dimerization reaction can be completed quickly within 5 min under a N2 atmosphere, as evidenced by the disappearance of the n(NO2) band (Figure 4 b). Importantly, the new b(CH) band at n˜ = 1140 cm1 arises from the connection of two benzene rings in the formation of trans-DMAB.[15d] The sequence of the reaction rate under various atmospheres from fast to slow is N2, air, and O2. This result indicates that the dimerization of DMAB from 4NTP would be suppressed in concentrated O2. Hot electrons generated from surface plasmons are necessary to trigger this reduction reaction (4NTP!DMAB). However, these hot electrons concentrated at the metal surface may be quenched by O2 molecules,[18a] which leads to a decreased reaction rate. The laser power used to induce surface plasmons is another important parameter that affects the reaction rate.[16c] Here, at an elevated laser power of 3 mW, the n(N=N) bands and b(CH) band become distinct in 20 min under a pure O2 atmosphere (Figure 5 a), but the n(NO2) band can still be distinguished clearly. This result again suggests that O2 is the hot-electron quencher and thus impedes the conversion of 4NTP to DMAB. However, the quenching of hot electrons can be overcome by

1006


Full Papers

Figure 4. Time-dependent SERS spectra of 4NTP under continuous 633 nm laser excitation with a laser power of 1.5 mW under a) O2 and b) N2.

Figure 5. Time-dependent SERS spectra of 4NTP under continuous exposure to the 633 nm laser with a laser power of 3 mW under a) O2 and b) N2.

a strong enough laser power. In N2, this plasmon-driven reaction can be completed within 100 s (Figure 5 b), which is too fast for us to capture any specific changes of Raman fingerprints at the early stage. These results show clearly that a high laser power can accelerate the rate of this plasmon-driven reaction regardless of the gas atmosphere, mainly because of the generation of more hot electrons to induce this reduction reaction. After we ascertained the influence of the excited energy, the time-dependent SERS spectra were recorded in N2-saturated H2O and H2 atmospheres with a high power (3 mW) to monitor the plasmon-driven reaction of 4NTP. Interestingly, if H2O vapor was introduced into the reaction station by N2, very different Raman features were recorded compared to that in pure N2 (Figure 6 a). The n(NO2) band is diminished significantly, and we can see very weak n(N=N) bands only at the beginning that then disappear quickly with prolonged laser excitation. This means that DMAB will be consumed under these conditions. Another interesting and important change in the Raman spectra is that the band of 4NTP at n˜ = 1570 cm1, which corresponds to the stretching vibrational mode of CC bond, becomes slowly reduced and broadened and eventually splits into two peaks. The new peak is located at n˜ = 1590 cm1 and becomes dominant toward the end of the reaction. Under H2,

a similar phenomenon was observed (Figure 6 b), in which the disappearance of the n(NO2) band is accompanied by the broadening and splitting of the n(CC) band at n˜ = 1570 cm1. Here, we are particularly interested in the underlying physical and chemical mechanisms of this broadened and split n(CC) band. On the basis of SERS spectra and DFT calculations, we discovered that the n(CC) Raman bands of benzenethiol (BT), DMAB to 4ATP are redshifted compared to that of 4NTP (Figure S8). We usually focus on the growth or disappearance of Raman fingerprints and overlook this band broadening and splitting. If we consider the variation of the environment of the N atom, the distribution of electron density on the benzene ring will definitely affect the Raman vibration of the CC bond. Here, the distribution of the molecular electron density located in the HOMO was calculated. The benzene ring of 4ATP has the majority of the electron density from the N atom and that of 4NTP has the least. The n(CC) band would shift to a higher wavenumber with an increase in the electron density as manifested by the calculated Raman spectra of these molecules.[25] Here, SERS spectra show that the n(CC) band for 4NTP, BT, DMAB, and 4ATP is located at n˜ = 1570, 1574, 1577, and 1590 cm1, respectively. The broadening and subsequent splitting into two Raman bands at n˜ = 1570 and 1590 cm1 suggest that 4NTP can be converted into 4ATP under H2O or

ChemCatChem 2015, 7, 1004 – 1010

www.chemcatchem.org

1007


Full Papers

Figure 7. Laser-power-dependent plasmon-driven reaction rate of 4NTP under controlled atmospheres. The reaction time is defined as the point when the Raman peak intensity ratio of the NO2 band and the formed N=N or NH2 band reaches a stationary state. Inset is shown a magnified vision of the high-power region.

Figure 6. Time-dependent SERS spectra of 4NTP under continuous exposure to the 633 nm laser with a laser power of 3 mW under a) N2-saturated H2O and b) H2.

H2 atmospheres, consistent with the failure to show a b(CH) band at n˜ = 1140 cm1 (benzene rings are not connected together). Previously, we have reported that H2O and H2 can act as proton source in the plasmon-driven conversion of DMAB into 4ATP.[23] Here, we hypothesize that under a N2-saturated H2O or H2 atmosphere, 4ATP can be formed by the dimerization of 4NTP into DMAB and then the reduction of DMAB into 4ATP. Previous studies have reported the monitoring of the conversion of 4NTP to 4ATP by SERS on metal nanostructures, but a strong reducing agent, KBH4, and Pt catalyst were introduced into the reaction system.[25, 26] In these cases, 4NTP was reduced by KBH4 with the assistance of Pt catalysis. Notably, a n(CC) band shift to higher frequency could also be seen in their studies but was not discussed in detail. In our case, all reactions were performed on the Ag surface, and 4NTP was converted into 4ATP via an intermediate, DMAB. These results led us to study laser-power-dependent plasmon-driven reactions of 4NTP under controlled atmospheres (Figure 7). Notably, the NO2 band does not always disappear completely if the applied laser power is low, even after a very long excitation period. Therefore, the reaction time is defined as the point when the Raman peak intensity ratio of the NO2 band and formed N=N or NH2 band becomes constant. For instance, 8 min is defined as the reaction time if the relative inChemCatChem 2015, 7, 1004 – 1010

www.chemcatchem.org

tensity of the NO2 band and the new N=N band is stable upon excitation with a laser power of 0.3 mW in the presence of O2. Under a N2, H2, or H2O atmosphere, a higher laser power leads typically to an accelerated reaction rate. Under these conditions, 4NTP was converted into either DMAB or 4ATP, and the atmosphere does not quench the hot electrons generated from the surface plasmons. Therefore, a higher laser power generates more hot electrons at the Ag surface, which leads to faster reaction rates. However, under an O2 atmosphere, Ag surface plasmons can transfer excited electrons to O2 either through chemical interface damping or plasmon decay.[27] The reduction of 4NTP into DMAB may occur only after surface-adsorbed O2 on Ag was dissociated by an electron-driven O2 dissociation process.[18] Here, although the plot shows a reaction rate that is inversely proportional to the laser power, there is actually no discernible reaction detected with a laser power less than 1.5 mW (the time was recorded when the intensity of the NO2 band was not changing). We believe it can be rationalized by the fact that with a low laser power, the limited number of generated hot electrons were mostly transferred to O2, and the transient negative ions (O2) that remain on the metal surface will inhibit the plasmon-driven reaction of 4NTP into DMAB. These data, taken together, give a greater understanding of the evolution of SERS spectra during the in situ monitoring of the plasmon-driven reactions of 4NTP under controlled atmospheres (Scheme 1). Under an O2 or N2 atmosphere, the broadened n(CC) band results from an overlap of CC bond vibrations from 4NTP (n˜ = 1570 cm1) and DMAB (n˜ = 1577 cm1; Figures 4 and 5). However, the new Raman peak at n˜ = 1590 cm1 (from 4ATP) under N2-saturated H2O or H2 atmosphere suggests a reduction of the azo group into two amine functional groups (Figure 6). Therefore, we suggest that the plasmon-driven reaction of 4NTP occurs as follows: First, the conversion of 4NTP to DMAB will occur naturally as long as the decay of the surface plasmons generates enough hot elec-

1008


Full Papers water (10 mL) in a 25 mL beaker with a magnetic stirrer in an icewater bath. After 10 min, an aqueous solution of ascorbic acid (1 m) was injected quickly into the mixture, which was stirred vigorously. The color of the solution became gray or black, and a large number of Ag particles were produced in a few minutes. The reaction was terminated after 15 min by centrifugation, and the Ag particles were collected. The resulting particles were rinsed repeatedly with deionized water and ethanol to remove the surface-adsorbed molecules. The samples were dried in a vacuum drier to prevent the oxidation of the Ag surface.

Characterization

Scheme 1. Schematic illustration of the surface-plasmon-induced catalytic reactions between 4NTP, DMAB, and 4ATP.

trons. Therefore, 4NTP can be converted into DMAB under any atmosphere with high laser powers, even though the presence of O2 will hinder this reduction reaction. Under a H2 or H2O atmosphere, 4NTP is converted easily into DMAB, which can be further transformed into 4ATP, during which H2 and H2O serve as proton sources. Our study differs from previous reports on the plasmon-driven reactions of 4NTP in that we have successfully controlled the reaction atmosphere and the in situ SPSERS technique to monitor a sequence of reduction reaction that transform 4NTP to DMAB and to 4ATP. We raise the importance of the Raman band broadening and splitting at n˜ = 1590 cm1, a Raman signature represents the presence of 4ATP that has been overlooked in the past.

SEM images were obtained by using a Helios NanoLab 600i (FEI) to analyze the morphology and size of the Ag microspheres. Powder XRD data were recorded by using an XRD-6000 X-ray diffractometer (Shimadzu) with a CuKa radiation source (40.0 kV, 30.0 mA). The Ag particles were soaked in Rhodamine B or 4NTP ethanol solution for 30 min and then rinsed several times with water to remove the surface residuals. The resulting Ag particles were dispersed on glass substrates before the SERS responses or Raman images were measured. The SERS spectra and Raman images were recorded by using a Renishaw inVia micro-Raman spectroscopy system under single scanning measurement and chemical image mode, respectively, by using a TE air-cooled 576 400 CCD array in a confocal Raman system with a laser wavelength of 633 nm. The number of accumulations was one, and a total exposure time of 10 s was employed. Laser powers of 3, 1.5, 0.3, 0.15, and 0.025 mW were applied. The beam size was adjusted to ~ 2 mm diameter.

Conclusions

Acknowledgements

We have established a correlation between the chemical structure of capping agents and the morphology of hierarchical Ag nanostructures in the acid-directed solution chemistry route by using a series of dicarboxylic acids. The optimized microsphere is used to monitor the plasmon-driven catalytic reactions of 4nitrothiophenol (4NTP) by an in situ single-particle surface-enhanced Raman spectroscopy technique under controlled atmospheres and laser power. Our results conclude that O2 can effectively quench hot electrons; however, at a high enough laser power, 4NTP can be dimerized into dimercaptoazobenzene even under an O2-saturated atmosphere, and dimercaptoazobenzene can be further reduced to form 4-aminothiophenol under a H2O or H2 atmosphere. The broadening and splitting of the n(CC) band of 4NTP during the plasmon-driven reactions have been interpreted. We believe this single-particle surface-enhanced Raman spectroscopy technique coupled with atmosphere control is imperative to unravel the reaction mechanisms of surface-plasmon-assisted/induced catalytic reactions.

P.X. is grateful for support from the Natural Science Foundation of China (21471039, 21203045, 21101041), China Postdoctoral Science Foundation (2014M560253), the Fundamental Research Funds for the Central Universities (HIT.BRETIII.201223), and the Open Foundation of State Key Laboratory of Electronic Thin Films and Integrated Devices (KFJJ201401). H.L.W. is grateful for support from the Laboratory Directed Research Development (LDRD) program, Los Alamos National Laboratory. Keywords: photochemistry · Raman spectroscopy · reaction mechanisms · silver · surface plasmon resonance

Experimental Section Synthesis of Ag microspheres Ag microspheres were synthesized according to a literature procedure.[8] Typically, an aqueous solution of AgNO3 (1 mL, 1 m) and dior monocarboxylic acids (50 mL, 0.1 m) were added to deionized

ChemCatChem 2015, 7, 1004 – 1010

www.chemcatchem.org

1009

[1] a) A. C. Blair, E. R. Fisher, D. Rickey, Science 2012, 337, 1056 – 1057; b) T. Grancha, J. Ferrando-Soria, M. Castellano, M. Julve, J. Pasan, D. Armentano, E. Pardo, Chem. Commun. 2014, 50, 7569 – 7585; c) S. Dolai, P. R. Nimmala, M. Mandal, B. B. Muhoberac, K. Dria, A. Dass, R. Sardar, Chem. Mater. 2014, 26, 1278 – 1285; d) S. Tang, M. S. Dresselhaus, J. Mater. Chem. C 2014, 2, 4710 – 4726; e) A. C. Manikas, F. Causa, R. Della Moglie, P. A. Netti, ACS Appl. Mater. Interfaces 2013, 5, 7915 – 7922. [2] a) R. Klajn, K. J. M. Bishop, M. Fialkowski, M. Paszewski, C. J. Campbell, T. P. Gray, B. A. Grzybowski, Science 2007, 316, 261 – 264; b) H. Wang, J. K. Keum, A. Hiltner, E. Baer, B. Freeman, A. Rozanski, A. Galeski, Science 2009, 323, 757 – 760. [3] X. Gao, R. J. Esteves, L. Thi Thu Hien, R. Jaini, I. U. Arachchige, J. Am. Chem. Soc. 2014, 136, 7993 – 8002. [4] W. Sun, T. Peng, Y. Liu, W. Yu, K. Zhang, H. F. Mehnane, C. Bu, S. Guo, X.Z. Zhao, ACS Appl. Mater. Interfaces 2014, 6, 9144 – 9149. [5] C. Niu, J. Meng, C. Han, K. Zhao, M. Yan, L. Mai, Nano Lett. 2014, 14, 2873 – 2878.


Full Papers [6] L. Scarabelli, M. Coronado-Puchau, J. J. Giner-Casares, J. Langer, L. M. Liz-Marzan, ACS Nano 2014, 8, 5833 – 5842. [7] H. Liang, Z. Li, W. Wang, Y. Wu, H. Xu, Adv. Mater. 2009, 21, 4614 – 4618. [8] B. Zhang, P. Xu, X. Xie, H. Wei, Z. Li, N. H. Mack, X. Han, H. Xu, H.-L. Wang, J. Mater. Chem. 2011, 21, 2495 – 2501. [9] L. Kang, P. Xu, D. Chen, B. Zhang, Y. Du, X. Han, Q. Li, H.-L. Wang, J. Phys. Chem. C 2013, 117, 10007 – 10012. [10] M. Sun, H. Xu, Small 2012, 8, 2777 – 2786. [11] M. Osawa, N. Matsuda, K. Yoshii, I. Uchida, J. Phys. Chem. 1994, 98, 12702 – 12707. [12] a) L. L. Zhao, L. Jensen, G. C. Schatz, J. Am. Chem. Soc. 2006, 128, 2911 – 2919; b) M. Sun, S. Liu, M. Chen, H. Xu, J. Raman Spectrosc. 2009, 40, 137 – 143; c) W.-H. Park, Z. H. Kim, Nano Lett. 2010, 10, 4040 – 4048. [13] a) D.-Y. Wu, X.-M. Liu, Y.-F. Huang, B. Ren, X. Xu, Z.-Q. Tian, J. Phys. Chem. C 2009, 113, 18212 – 18222; b) Y. R. Fang, Y. Z. Li, H. X. Xu, M. T. Sun, Langmuir 2010, 26, 7737 – 7746. [14] a) Y.-F. Huang, H.-P. Zhu, G.-K. Liu, D.-Y. Wu, B. Ren, Z.-Q. Tian, J. Am. Chem. Soc. 2010, 132, 9244 – 9246; b) K. Kim, K. L. Kim, K. S. Shin, Langmuir 2013, 29, 183 – 190. [15] a) V. Canpean, M. Iosin, S. Astilean, Chem. Phys. Lett. 2010, 500, 277 – 282; b) D.-Y. Wu, L.-B. Zhao, X.-M. Liu, R. Huang, Y.-F. Huang, B. Ren, Z.Q. Tian, Chem. Commun. 2011, 47, 2520 – 2522; c) Y.-F. Huang, D.-Y. Wu, H.-P. Zhu, L.-B. Zhao, G.-K. Liu, B. Ren, Z.-Q. Tian, Phys. Chem. Chem. Phys. 2012, 14, 8485 – 8497; d) M. T. Sun, Y. Z. Huang, L. X. Xia, X. W. Chen, H. X. Xu, J. Phys. Chem. C 2011, 115, 9629 – 9636. [16] a) E. M. van Schrojenstein Lantman, T. Deckert-Gaudig, A. J. G. Mank, V. Deckert, B. M. Weckhuysen, Nat. Nanotechnol. 2012, 7, 583 – 586; b) M. T. Sun, Z. L. Zhang, H. R. Zheng, H. X. Xu, Sci. Rep. 2012, 2, 647; c) L. Kang, P. Xu, B. Zhang, H. Tsai, X. Han, H.-L. Wang, Chem. Commun. 2013, 49, 3389 – 3391.

ChemCatChem 2015, 7, 1004 – 1010

www.chemcatchem.org

[17] a) M. Sun, Y. Fang, Z. Zhang, H. Xu, Phys. Rev. E 2013, 87, 020401; b) X. Zhang, P. Wang, Z. Zhang, Y. Fang, M. Sun, Sci. Rep. 2014, 4, 5407. [18] a) P. Christopher, H. L. Xin, S. Linic, Nat. Chem. 2011, 3, 467 – 472; b) P. Christopher, H. L. Xin, A. Marimuthu, S. Linic, Nat. Mater. 2012, 11, 1044 – 1050. [19] R. W. Taylor, R. J. Coulston, F. Biedermann, S. Mahajan, J. J. Baumberg, O. A. Scherman, Nano Lett. 2013, 13, 5985 – 5990. [20] M. Sun, Z. Zhang, Z. H. Kim, H. Zheng, H. Xu, Chem. Eur. J. 2013, 19, 14958 – 14962. [21] K. Kim, S. H. Lee, J.-Y. Choi, K. S. Shin, J. Phys. Chem. C 2014, 118, 3359 – 3365. [22] a) Z. Y. Fang, Y. M. Wang, Z. Liu, A. Schlather, P. M. Ajayan, F. H. L. Koppens, P. Nordlander, N. J. Halas, ACS Nano 2012, 6, 10222 – 10228; b) S. Mukherjee, F. Libisch, N. Large, O. Neumann, L. V. Brown, J. Cheng, J. B. Lassiter, E. A. Carter, P. Nordlander, N. J. Halas, Nano Lett. 2013, 13, 240 – 247. [23] P. Xu, L. L. Kang, N. H. Mack, K. S. Schanze, X. J. Han, H. L. Wang, Sci. Rep. 2013, 3, 2997. [24] B. Dong, Y. Fang, X. Chen, H. Xu, M. Sun, Langmuir 2011, 27, 10677 – 10682. [25] W. Xie, C. Herrmann, K. Kompe, M. Haase, S. Schlucker, J. Am. Chem. Soc. 2011, 133, 19302 – 19305. [26] R. Liu, J. F. Liu, Z. M. Zhang, L. Q. Zhang, J. F. Sun, M. T. Sun, G. B. Jiang, J. Phys. Chem. Lett. 2014, 5, 969 – 975. [27] L. Brus, Acc. Chem. Res. 2008, 41, 1742 – 1749.

Received: December 17, 2014 Published online on February 6, 2015

1010
