Fast Surface Restructuring within the Gap of Au Nanocube Dimer for the Enhancement of Catalytic Efficiency

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 Mar 2022Fast Surface Restructuring within the Gap of Au Nanocube Dimer for the Enhancement of Catalytic Efficiency Xiao Wang, Zhongju Ye, Jianhao Hua, Lin Wei, Shen Lin and Lehui Xiao Xiao Wang State Key Laboratory of Medicinal Chemical Biology, Tianjin Key Laboratory of Biosensing and Molecular Recognition, College of Chemistry, Nankai University, Tianjin 300071 , Zhongju Ye State Key Laboratory of Medicinal Chemical Biology, Tianjin Key Laboratory of Biosensing and Molecular Recognition, College of Chemistry, Nankai University, Tianjin 300071 , Jianhao Hua State Key Laboratory of Medicinal Chemical Biology, Tianjin Key Laboratory of Biosensing and Molecular Recognition, College of Chemistry, Nankai University, Tianjin 300071 , Lin Wei College of Chemistry and Chemical Engineering, Hunan Normal University, Changsha 410081 , Shen Lin State Key Laboratory of Medicinal Chemical Biology, Tianjin Key Laboratory of Biosensing and Molecular Recognition, College of Chemistry, Nankai University, Tianjin 300071 and Lehui Xiao *Corresponding author: E-mail Address: [email protected] State Key Laboratory of Medicinal Chemical Biology, Tianjin Key Laboratory of Biosensing and Molecular Recognition, College of Chemistry, Nankai University, Tianjin 300071 https://doi.org/10.31635/ccschem.021.202100770 SectionsAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Identification of the catalytic dynamics and plasmonic effects plays a critical role in the design of heterogeneous catalysts. However, the knowledge of plasmonic effect on catalytic dynamics remains limited at the single-particle level. Using the non-fluorescent amplex red to fluorescent resorufin as a model reaction, significant enhancement in catalytic efficiency from the coupled Au nanocube dimer (AuCD) was clearly revealed with the single-molecule fluorescence microscopy. AuCD exhibits noticeably higher catalytic efficiency than the monomer, which is attributed to the spontaneous dynamic surface restructuring. Spatiotemporally resolved dynamics suggest that the active catalytic sites essentially originate from the plasmonic nanogap where an electromagnetic (EM) hot spot exists. The enhanced EM field accelerates the generation of hot carriers and promotes the spontaneous surface restructuring by enhancing the lattice vibrations, which ultimately improves the catalytic activity. These microscopic views provide new insights into the effect of EM fields on surface restructuring dynamics of nanocatalysts. Download figure Download PowerPoint Introduction Nanoparticles with excellent catalytic properties have attracted considerable attention in many important chemical transformations, including oxidation of hydrocarbons, cross coupling, and hydrogenation–dehydrogenation.1–6 Great efforts have been made to correlate the structure and catalytic property and modulate the catalytic activities.7–9 With advanced transmission electron microscopy (TEM),10 scanning probe microscopy (SPM),11 and atomic force microscopy (AFM),12 the structure of nanoparticles can be characterized at the single-particle level with nanometer resolution. However, the catalytic property of nanoparticles is commonly studied ex situ in bulk solution, which only provides ensemble-averaged properties not spatially resolved real-time information. Furthermore, because of structural dispersions, heterogeneous distribution of surface sites, and surface restructuring dynamics, the catalytic activity of individual nanoparticle varies as a function of time.9,13–17 To elucidate deep insight into the catalytic performance of the catalyst, a direct approach is to explore the catalytic reaction in a spatially resolved manner in real-time (i.e., at the single-particle level).18–23 Single-molecule fluorescence microscopy has recently proved to be a sensitive and informative tool for real-time in situ study of catalytic events.13,24 By removing ensemble averaging, one can follow the reactions at single-turnover resolution. For instance, real-time single-turnover kinetics reveal size-, catalysis-, and metal-dependent temporal activity from single pseudospherical Au nanoparticles.17,21 The time-dependent catalytic efficiency variation is essentially attributable to both spontaneous and catalysis-induced dynamic surface restructuring. In addition, the surface dynamic restructuring process also depends on the constantly changing adsorbate–surface interaction. More interestingly, spatiotemporally resolved single-molecule imaging further enables the direct quantification of catalytic activities at different surface sites of individual particles, clearly uncovering the radial activity gradients within the same surface facet.20,25,26 These spatially resolved activity patterns within Au nanostructures can be ascribed to the fluctuation of low-coordinated surface sites, including corner, edge, and defect sites, among which the distributions of these sites are associated with the nanocrystals' morphology and growth mechanism. Notably, the single-molecule technique has uniquely unveiled the microscopic views of spatiotemporal variations in catalytic reactivity which are otherwise difficult to disclose with classical ensemble experiments. To realize desirable catalytic performance, it is necessary to construct nanocatalysts with ideal geometric structure and low-coordinated surface sites. Among many plasmonic metal nanostructures, Au nanocube (AuNC) is a promising nanocatalyst because of its low-coordinated corners and edges.27,28 These sites on AuNC normally give rise to a noticeably enhanced localized electromagnetic (EM) field on individual nanoparticles or between coupled nanoparticles. Moreover, because of the flat surface, it is often used as a building block for functional nanostructure assembly such as nanocube dimer, where a strongly enhanced EM field exists within the nanogap.10,29 So far, localized surface plasmon resonance (LSPR) has been determined as a robust property to promote the plasmonic catalysis,1,8 which can generate a strong EM field localized around the nanoparticle and confine or focus EM energy in the near-field.4,16 This surface plasmon (SP)-enhanced phenomenon opens a new opportunity to promote catalytic conversions by generating highly energetic electrons, that is, hot electrons. Hot electrons can not only greatly facilitate chemical conversions but also offer new pathways to manipulate product selectivity.5,6,30–32 Even though these measurements have been extensively studied in bulk solution, it remains a challenge to evaluate the hot electron effect on microscopic dynamics due to the lack of spatiotemporal resolution. Therefore, it is urgent to explore the relationship between hot electrons and surface dynamic restructuring using the single-molecule technique. In this work, the catalytic efficiency from Au nanocube dimer (AuCD) promoted by EM field-enhanced surface restructuring was revealed at the single-particle level for the first time (Scheme 1). In detail, a substrate-supported assembly strategy was adopted to engineer this dimeric structure that exhibited strong plasmonic coupling. Using the non-fluorescent amplex red (AR) to fluorescent resorufin as a model reaction, different catalytic properties between Au nanocube monomers (AuCMs) and AuCDs were detected with single-turnover resolution. Interestingly, AuCDs possessed higher catalytic activity that can be ascribed to the enhanced spontaneous surface restructuring induced by the strong EM field within the nanogap, which accords well with the weaker substrate binding affinity and the enhanced fluorescence intensity. Furthermore, the boosted hot carriers from AuCDs were directly verified by surface-enhanced Raman spectroscopy (SERS) characterizations and theoretical simulations. These new views demonstrated herein as well as the spatiotemporally resolved catalysis mapping may facilitate the development of highly efficient nanocatalysts. Scheme 1 | The single-molecule catalysis from Au nanostructures. Download figure Download PowerPoint Experimental Methods Chemicals and materials Hexadecyltrimethylammonium bromide (CTAB), hydrogen peroxide (H2O2), and ethanol (EtOH) were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Gold(III) chloroauric acid (HAuCl4·3H2O) was obtained from Aladdin Reagent Co., Ltd (Shanghai, China). Sodium borohydride (NaBH4), sodium bromide (NaBr), 11-mercaptoundecanoic acid (MUA), p-nitrothiophenol (pNTP), and 1,6-hexanedithiol (C6DT) were purchased from Sigma-Aldrich Trading Co., Ltd. (Shanghai, China). AR was obtained from Invitrogen Ltd. (Paisley, UK). All other chemicals not mentioned here were purchased from Sinopharm Chemical Reagent Co., Ltd.. The experiments without special instructions were performed at room temperature. All aqueous solutions were prepared with ultrapure water. Synthesis of AuCMs AuCMs with edge length of 45.7 ± 1.1 nm were synthesized according to a seed-growth method based on previous work.10,27 First, seed particles were prepared in a mixed solution containing 125 μL of HAuCl4 (10 mM) and 3.75 mL of CTAB (100 mM). Then, 300 μL of freshly prepared, ice-cold NaBH4 aqueous solution (10 mM) was rapidly added. The mixture turned brownish immediately. The seed solution was stirred for 2 min and then kept undisturbed for 1 h before use. The resulting solution was diluted 10 times by water. Second, the growth solution was prepared by sequentially adding 100 μL of HAuCl4 (10 mM), 800 μL of CTAB (100 mM) and 4 mL of H2O into the vial. Then, 600 μL of ascorbic acid (100 mM) was added and the growth solution became colorless. Next, 2.5 μL of diluted seed solution was added into the growth solution. The resultant solution was mixed by gentle inversion for 10 s and stored overnight. Synthesis of AuCDs The synthesis of ideal dimers comprises five steps: (1) Adsorption of the first AuNCs. The glass slide cleaned with piranha solution was immersed in 150 μL of solution containing CTAB-capped AuNCs (10 pM) for 3 h at 30 °C. All subsequent steps were performed at 30 °C. The glass slide with adsorbed AuNCs was washed with water and EtOH. (2) Thiol functionalization of the first AuNCs. The AuNC-adsorbed glass slide from step 1 was immersed in a mixture containing 150 μL of ethanolic 1,6-hexanedithiol solution (1 mM) and 6 μL of aqueous NaBr solution (25 mM). After 1 h, the glass substrate was washed with EtOH. (3) Adsorption of the second AuNCs. The glass slide was dipped into a mixture containing 150 μL of AuNCs (150 pM) and 6 μL of NaBr (25 mM) for 5 h. (4) Stablizing the dimers. The glass substrate was washed with EtOH and immersed in 150 μL of ethanolic MUA solution (1 mM) for 1 h. (5) Collection of the dimers. After washing with EtOH, the glass substrate was sonicated in 100 μL of MUA (10 μM) for 30 s. Characterization of AuCMs and AuCDs The size and morphology of nanoparticles were characterized by TEM (JEM 2100, JEOL, Tokyo, Japan). UV–vis absorption spectra were recorded by a UV-2450 spectrophotometer (Shimadzu, Kyoto, Japan). The scattering spectra and polarization characterizations were performed on an upright microscope (Ni-U, Nikon, Tokyo, Japan) that was equipped with a 40× objective and a true-color complementary metal oxide semiconductor camera (sCMOS; Digiretina 16, Tucsen photonics Co., Ltd., Fujian, China). The corresponding scattering spectra were recorded by a monochrome charge-coupled device (CCD; Hamamatsu, Japan). SERS measurement In a typical experiment, the synthesized Au nanostructures were first concentrated fivefold. Then, 10 μL of the concentrated AuCMs or AuCDs was mixed with 10 μL of pNTP (5 μM). Next, 10 μL of NaBH4 aqueous solution (0.6 mM) was added to the mixture. Afterward, the mixtures were placed in the capillary and tested by a LabRAM HR Evolution microscope (Horiba Jobin Yvon, France) installed with a 10× objective (NA = 0.25). A He−Ne laser of 633 nm with a power of 86 μW μm−2 was used as the excitation source. SERS experiments were taken on an extended range (900–1700 cm−1). Single-nanoparticle catalysis experiments Single-molecule fluorescence measurements were performed on a home-built imaging system based on an Eclipse Ti-U inverted epi-fluorescence microscope (Nikon), equipped with the DF imaging systems. In the experiment, nanoparticles were first injected into a flow channel. Reactants with various concentrations (0.5–5.0 μM AR and 50 mM H2O2) were injected into the channel at a flow rate of 5 μL min−1. A tightly focused Gaussian laser beam (532 nm) was utilized to directly excite the fluorescence of resorufin. The laser beam was reflected by a 565 nm dichroic mirror. The excited fluorescence from the molecules was filtered by a 605/55 nm band-pass filter (Semrock, Rochester, NY). The fluorescent signals of the resorufin molecules were collected by a 100× oil objective (NA 0.5–1.3) and then captured by an EMCCD (iXon ultra 897, Andor, Belfast, United Kingdom). More than 4000 frames were captured to track the catalytic trajectory and the frame rate was operated at 20 fps. All movies were processed and analyzed with MATLAB (MathWorks, Inc., Natick, MA) and Image J (National Institutes of Health, Bethesda, MD). Finite-difference time-domain simulation The simulations of the near-field distributions of isolated and coupled nanoparticles were performed by a software package, finite-difference time-domain (FDTD) Solution 8.0. The total-field scattered-field (TF/SF) source from 500 to 900 nm was used in simulation. The geometric parameters of the monomer and dimer were consistent with TEM images. For AuCM, the edge length was set as 50 nm and the radius curvature of corner was set as 5 nm. In addition, the gap size of AuCD was set as 1 nm. The dielectric constants of gold were obtained by fitting the measured data of Johnson and Christy.4,10 In the calculations, the mesh around the Au nanostructure was 1 nm × 1 nm × 1 nm and in the gap the mesh was 0.1 nm × 2 nm × 2 nm where 0.1 nm is along the gap. The refractive index of the surrounding medium was set to 1.33. COMSOL simulation The rate and density of hot electron generation for the AuCM and AuCD were calculated using COMSOL Multiphysics software (COMSOL Inc., Burlington, MA). In our model, the wavelength of incident light and electric field amplitude were set as 532 nm and 1 V/m, respectively. The electric field propagated down along the z-axis, and the polarization direction was along the x-axis. The model of two nanostructures had the same settings as described in the FDTD simulation. The outermost layer was a perfectly matched layer (PML), and its thickness was set to half wavelength. The local generation rate of hot electrons was calculated according to the published formula.33 Results and Discussion Spectroscopic and microscopic characterizations of AuCMs and AuCDs The synthesis of AuCDs is based on a substrate-supported assembly strategy using alkanedithiol as the linker.10,29 A monolayer of dithiol was first assembled on AuCMs' surfaces which were fixed on the cover glass surface. After that, the sample was washed with EtOH and immersed in AuCMs solution again. AuCDs were then formed through Au-thiol chemistry. To improve the colloidal stability of the dimers, MUA was added to the solution to prevent the aggregation process. The size and morphology of AuCMs and AuCDs were characterized by TEM. As shown in Figure 1a, cube-like AuCMs are well dispersed with edge length of 45.7 ± 1.1 nm. From the TEM image (Figure 1b), the dimer yield is approximately 75%, which is in good accordance with the result reported previously.29 The inset is the statistical analysis of different types of dimer, consisting of corner to corner, face to face, and edge to face configurations. From the UV–vis absorption spectra, before the dimerization process (Figure 1c, green line), AuCMs display a characteristic LSPR peak at 542 nm. After the assembly process, a strong peak in the range of 700–750 nm emerges (Figure 1c, red line). This new absorption peak is ascribed to the longitudinal plasmonic coupling between two adjacent nanocubes. Figure 1 | (a) Typical TEM image of AuCMs and the statistically analyzed size distribution of AuCMs from the TEM image (inset plot). (b) Typical TEM image of AuCDs and the statistically analyzed type distribution of AuCDs (I, II, and III columns correspond to the coupled mode of corner to corner, face to face, and edge to face, respectively). (c) UV–vis absorption spectra of AuCMs and AuCDs. (d and g) Dark-field images of AuCMs and AuCDs, respectively. (e and h) Scattering intensity spectra of a single AuCM and AuCD, respectively. The black lines are Lorentz fitting. The red dotted line in (h) represents the location of the transverse plasmon coupling band. (f and i) The measured polarization-dependent scattering responses from single AuCM and AuCD, respectively. The plots in (c), (e), (f), (h), and (i) are normalized. Download figure Download PowerPoint Then, dark-field optical microscopy (DFM) was used to characterize the scattering feature of these particles at the single-particle level.9,12,34–39 As shown in Figure 1d, all the AuCMs disperse separately with uniform and characteristic green color. However, for the AuCDs, noticeable red-shift in the dark-field image can be observed after the dimerization process (Figure 1g). Representative scattering spectra of single AuCM and AuCD are shown in Figures 1e and 1h. An approximate 100 nm red-shift is observed in the scattering spectrum after the plasmonic coupling. Besides DFM characterization, polarization measurement was further used to characterize the anisotropic morphology of these two nanostructures beyond the optical diffraction limit.9,36 Figures 1f and 1i illustrate polarization-dependent scattering intensity measurements of AuCM and AuCD, respectively. The orientation-dependent intensity distribution of AuCM is noticeably different from that of AuCD. In the far-field image plane, the scattered signal from individual AuCM displays semi-isotropic effect in each direction (Figure 1f). However, for a quasi rod-shaped AuCD with anisotropic property, the scattered signal displays transverse and longitudinal SP modes with oscillation directions parallel to the short and long axes, respectively. Given the target particle has dimeric morphology, the period of polarization-dependent signal response of AuCD is π due to the symmetric effect (Figure 1i). In this case, AuCD exhibits a well-defined dipole feature under scattering mode in contrast to AuCM. In the following experiments, to discriminate the morphology of the observed nanoparticle on the glass slide surface, microscopic polarization measurement, an effective and sensitive strategy, was operated in situ. Enhanced catalytic activities from AuCDs To explore the catalytic activity of AuCMs and AuCDs, fluorogenic reaction from AR to resorufin in the presence of H2O2 was adopted as a model reaction.25,40 For the single-molecule experiments, the particles were first adsorbed on a quartz slide surface within a flow channel. H2O2-saturated (50 mM) reactant solution with different AR concentrations (ranging from 0.5 to 5.0 μM) was then added to the flow channel. The fluorescence from the oxidative product (i.e., resorufin) after the catalytic reaction was excited by a 532 nm laser and then recorded by an EMCCD. A representative time-dependent trajectory of fluorescence intensity from individual nanoparticle is shown in Figure 2, which essentially consists of stochastic "off"/"on" signals. Each fluorescent burst represents the formation of a fluorescent molecule (i.e., resorufin). Each "off"/"on" cycle corresponds to a single turnover of the product formation and its subsequent dissociation process from the catalytic site. The narrow burst on the trajectory indicates the fast desorption of the product molecules. The fluorescent signal can only be detected before it dissociates from the catalytic site. To correlate the catalytic process with the desired nanoparticle, the dark-field imaging system was combined with the single-molecule fluorescence imaging modality (Figures 2a, 2b, 2d, and 2e).9,41,42 The catalytic reactions from the target nanoparticle can then be recorded accurately. Figure 2 | (a and d) The dark-field images of AuCMs and AuCDs, respectively. (b and e) The corresponding fluorescence microscopic characterizations of the catalytic process from individual AuCMs and AuCDs, respectively (taken at 50 ms frame rate). (c and f) Fluorescence intensity vs time trajectories of single AuCM and AuCD under respectively. and h) The distribution of fluorescence intensity in the trajectory from AuCMs and AuCDs, respectively. (i) AR of from AuCMs and AuCDs and The distributions of from turnover trajectory of AuCM and AuCD, respectively. lines are with a of = and for AuCMs and AuCDs, respectively. of from 50 trajectories of AuCMs and AuCDs and the lines are Gaussian fitting. Download figure Download PowerPoint Figures and the time-dependent catalytic from AuCM and AuCD, respectively. many fluorescent (i.e., higher turnover in the track from AuCD within the same with the intensity of the signal from AuCMs (Figure the catalytic from AuCDs display a fluorescence response (Figure which be ascribed to the fluorescence enhancement effect from the plasmonic To disclose the catalytic kinetics from these two nanostructures, the reactant turnover rate from these nanoparticles was first the of AR was H2O2 was kept in a (50 mM). the turnover of these two nanostructures with AR In contrast to AuCM, the statistically analyzed turnover rate from individual AuCDs was higher catalytic activity from AuCDs (Figure were under the of and μM for AuCMs and AuCDs, respectively. the turnover rate from AuCDs is around times than the kinetics of this reaction can be described by a for heterogeneous The of two and H2O2) a model, in which different sites on a Au the turnover rate kinetics the of AR is H2O2 is kept Notably, the microscopic time of is than which can be in one In this case, the type of kinetics is by the following 1 = 1 = AR AR 1 AR AR 1 is turnover is is the density function of is the single-particle catalytic rate the reactivity of an AR and for AR and respectively. and are the corresponding constants of AR and respectively. the of H2O2 is 1 can be 1 = AR AR 1 AR AR = AR AR 1 AR AR is the rate for one catalytic is the of surface catalytic sites, to the on one In addition, the of by the surface of single nanoparticles indicates the catalytic reactivity surface for an AuCD assembled by two be two times than that of a single AuCM because active sites on the surface are the data in Figure with 2, the determined for AuCMs and AuCDs is ± and ± respectively. the reaction activity of AuCDs is approximately times higher than the determined for AuCMs and AuCDs is × and × a higher catalytic reactivity for AuCDs. this the higher catalytic of AuCDs be ascribed to the of active sites of two isolated This can be further from the of constants from these two ± and ± for AuCMs and AuCDs, respectively). have shown that is by the of the A from AuCDs indicates a surface is To further the catalytic enhancement effect from AuCDs, analyzed the time-dependent turnover rate In this case, the distributions of microscopic reaction time from each single-particle turnover trajectory were on the reaction the density function of (i.e., can be = AR AR 1 AR AR 1 AR AR 1 AR AR 1 = represents the rate for the generation of single resorufin. Figures and the distributions of AuCM and AuCD at a AR The distributions all follow an function with a of The are the based on By the distribution of from 50 the distribution of and the can be determined (Figure the distribution of from AuCDs is than reaction from the dimeric The distribution in the activity be ascribed to the structural where face to face or edge to face coupling can all as shown in the TEM However, it is that the statistically analyzed of from AuCDs ± is higher than that from AuCMs ± Surface restructuring dynamics from AuCMs and AuCDs the catalytic reaction corresponds to the restructuring of surface or generation of low-coordinated time-dependent catalytic from individual particles give rise to on the catalytic enhancement Interestingly, the fluorescence turnover trajectory not only the direct of the reaction activity between different but also enables the of time-dependent reaction fluctuation from the same The temporal variation in catalytic efficiency the dynamic activity oscillation of a single nanoparticle, which can be determined by the rate of turnover as shown in Figures and the of "off"/"on" time varies dynamic activity fluctuation from an individual The time of dynamic activity fluctuation can be by the function of the microscopic reaction time from the of = represents is the turnover index from the and =

Topics & Concepts