Scanning probe microscopy for real-space observations of local chemical reactions induced by a localized surface plasmon

Emiko Kazuma and Yousoo Kim*
Surface and Interface Science Laboratory, RIKEN, 2-1 Hirosawa, Wako-shi, Saitama 351-0198, Japan. E-mail:

Received 15th April 2019 , Accepted 8th July 2019

First published on 8th July 2019

Localised surface plasmon (LSP) resonance has attracted considerable attention in recent years as an efficient driving force for chemical reactions. The chemical reactions induced by LSP are classified into two types, namely, redox reactions based on plasmon-induced charge separation (PICS) and chemical reactions induced by the direct interaction between LSP and molecules (plasmon-induced chemical reactions). Although both types of reactions have been extensively studied, the mechanisms of PICS and plasmon-induced chemical reactions remain unexplained and controversial because conventional macroscopic methods can hardly grasp the local chemical reactions induced by LSP. In order to obtain mechanistic insights, nanoscale observations and investigations are necessary. Scanning probe microscopy (SPM) is a powerful experimental tool to investigate not only the surface morphology but also the physical and chemical properties of samples at a high spatial resolution. In this perspective review, we first explain SPM combined with optical excitation, and then, review the recent studies using SPM techniques for real-space observations of the chemical reactions induced by LSP.

1 Introduction

Metal nanostructures (NSs) such as Au, Ag, and Cu nanoparticles absorb and scatter light due to localised surface plasmon (LSP) resonance, which is the collective oscillation of the free electrons of the metals coupled with the electric field of the incident light. The resonant energy and intensity of LSP strongly depend on the size and shape of the metal NSs, gap distance between the NSs, and the permittivity of the surrounding medium.1 Furthermore, a strong localised oscillating electric field (optical near field) of LSP is focused in the close vicinity of the metal surfaces beyond the diffraction limit.2 These unique properties of the plasmonic metal NSs have attracted great interest since 2000, and have been exploited for various applications, including fabrication of optical materials,3 plasmon sensors,4,5 local spectroscopies such as surface-enhanced Raman spectroscopy (SERS),6 photovoltaic cells,7,8 fluorescence enhancement,9 and photocatalysts.10,11 The nanoscale morphology of the metal NSs directly influences the function and performance of the plasmonic devices. The structural anisotropy is also an important factor to determine the spatial distribution of the optical near field, which is known as a hot spot for the enhancement of Raman signals and light energy conversion efficiency of photovoltaics and photocatalysts.12,13 Thus, the formation of the plasmonic metal NSs having precisely controlled structures has been studied in earnest,14–16 and the analysis of the morphology is crucial to evaluate the relation between the structures of the NSs and performances of the plasmonic devices.

Scanning probe microscopy (SPM) techniques such as scanning tunnelling microscopy (STM) and atomic force microscopy (AFM), which have revolutionised surface science through the observation of the surface with atomic spatial resolution, are powerful and useful tools to characterise the three-dimensional morphology of the plasmonic metal NSs.17–22 In addition to morphological information, SPM can also provide local mapping of the chemical and physical properties on the surfaces.23–25 With the progress in SPM techniques, SPM combined with optical excitation by light has led to the evolution of local optical spectroscopies such as scanning near field optical microscopy (SNOM) and tip-enhanced Raman spectroscopy (TERS), and has also been used to understand the local phenomena induced by LSP and those induced by light in optoelectronic devices, photovoltaics, and photocatalysts.

Chemical reactions induced by LSP, which are one of the most promising applications of the plasmon, are novel photocatalytic reactions because LSP can convert solar energy to chemical energy with high efficiency. These chemical reactions can be classified into two types, namely, redox reactions based on plasmon-induced charge separation (PICS)26–31 (Fig. 1a) and chemical reactions induced by the direct interaction between LSP and molecules (plasmon-induced chemical reactions)32–46 (Fig. 1b). PICS proceeds at the interface between a plasmon excited metal NS and a semiconductor such as TiO2, ZnO, CeO2, or GaN, and the separated electrons and holes result in reduction and oxidation reactions, respectively (Fig. 1a). In contrast, the plasmon-induced chemical reactions are initiated from the molecules excited by the excitation sources and generated through the decay of LSP such as an electric field, hot carriers, and heat (Fig. 1b).46 Both bond formation and dissociation reactions on the basis of the direct interaction between LSP and molecules have attracted increasing attention and have been used in direct plasmonic photocatalysis. Although the chemical reactions induced by LSP have been enthusiastically studied in the past decade, the details of both the mechanisms of PICS and the interaction between LSP and molecules remain unexplained, and thus still controversial. In order to clarify these, nanoscale observations and investigations are necessary. Recently, SPM combined with optical excitation by light has been successfully utilised for nanoscale investigation of PICS and the local chemical reactions induced by LSP. This combination provided mechanistic insights that could not be obtained with conventional macroscopic measurements because of the strong localisation of the plasmon.38,43,47–60

image file: c9cp02100k-f1.tif
Fig. 1 (a) Schematic illustrations of a redox reaction based on PICS at the interface between a metal NS and an n-type semiconductor such as TiO2, ZnO, CeO2, and GaN. The Schottky junction enables PICS and suppresses the recombination of the separated electrons and holes. (b) A chemical reaction induced by the direct interaction between LSP and a molecule. The molecule is excited by (i) electric field, (ii) hot electrons, or (iii) heat generated through the excitation and decay of LSP.

In this perspective, we describe SPM techniques for analysing local chemical reactions induced by LSP, and explain how important these techniques are to reveal reaction behaviours in real-space and obtain insights into reaction mechanisms. Section 2 summarises the SPM techniques combined with optical excitation by light and the features of each method applicable to plasmon chemistry. Section 3 introduces several examples of how the SPM techniques have been applied to the investigation of local chemical reactions caused by the PICS. Section 4 describes the SPM observations and analyses of plasmon-induced chemical reactions of molecules adsorbed on metal surfaces. Finally, a perspective for the advanced use of the SPM techniques is presented.

2 Scanning probe microscopy combined with optical excitation

SPM provides three-dimensional information of the sample surfaces by detecting the interaction between the SPM tip and the sample during scanning with the tip along the surface. SPM is a versatile observation tool and there are many types of SPMs that are available for conducting measurements under ultra-high vacuum (UHV) conditions, in air, and in a solution. Here, we have focused on describing the combination of SPM with optical excitation by light, including STM, AFM, TERS, and Kelvin probe force microscopy (KPFM), which have been applied to plasmon chemistry, in terms of analytical methods. However, the detailed mechanics of the instruments have not been explained.

2.1 Scanning tunnelling microscopy (STM)

STM, the first SPM, was invented by G. Binnig and H. Rohre at IBM Zurich in 1981.61 After its emergence, the understanding of local phenomena on surfaces has dramatically advanced and this has led to great strides in nanotechnology, owing mainly to the high spatial resolution of STM. A sharp metal tip such as the W, Pt, or PtIr tip is used for STM and the bias voltage is applied between the tip and a conductive sample (Fig. 2a). When the gap distance between the tip and the sample becomes close to ∼1 nm, the tunnelling current starts to flow. Tunnelling current reflects the density of states of the sample surface and is highly sensitive to changes in the gap distance, which enables observations of surfaces at atomic spatial resolution. STM has been used not only for imaging but also for local spectroscopies, including scanning tunnelling spectroscopy (STS),62 inelastic electron tunnelling spectroscopy (IETS),63 and STM action spectroscopy (STM-AS).64 STS and IETS are the first and second derivatives of an IV curve, and are applied to reveal the electronic structures and vibrational information of the sample, respectively. STM-AS has been developed to clarify the elementary processes of IET-induced vibrational excitation leading to molecular motions and reactions. In addition to these conventional functions gained by analysing the tunnelling current, combining STM with optical excitation and detection has led to local optical spectroscopies such as STM-TERS65 and STM-induced luminescence.66–68 STM combined with optical detection has been used to visualize and map the spatial distribution of plasmon modes in metal nanoparticles.23–25
image file: c9cp02100k-f2.tif
Fig. 2 Schematic illustrations of (a) STM and (b) AFM.

Regarding the analysis of chemical reactions, UHV-LT-STM can be combined with optical excitation and utilised for real-space observation of surface photochemical reactions such as photon-induced molecular motions and dynamics,69–71 photoisomerisation,72,73 and photodissociation.74 In the experiments of the surface photochemical reactions, the STM tip needs to be retracted from the surface during light irradiation to eliminate the influence of the tip. In contrast, recently, UHV-LT-STM combined with optical excitation was successfully applied to observe a single-molecule reaction induced by LSP excited at the nanogap between the STM tip and a metal substrate.43 To excite LSP in the STM junction, a sharp STM tip made of a plasmonic metal such as Au and Ag is necessary. The resonant energy and intensity are tuneable by various factors such as the tip shape, gap distance, and materials of the tip and substrate.75,76 In particular, the fabrication of Au and Ag tips having well-defined shape is one of the key technologies for regulating LSP generated in the STM junction.77

Although STM is a powerful technique to examine surface structures, dynamics and reactions, only conductive samples are available and samples with high roughness are not appropriate. Accordingly, STM has not been used very much for industrial applications but mainly applied to fundamental research.

2.2 Atomic force microscopy (AFM)

AFM was invented to overcome the limitation on conductive samples required in the STM measurements, and is applicable to any sample regardless of the conductivity. Thus we can evaluate sample surfaces as it is in various environments such as in air, gas, and a solution, without the pretreatment of samples. This enables us to study biomaterials including biological macromolecules and living cells. AFM can not only scan surfaces but also measure various forces and apply force locally. A cantilever with a sharp tip is used in AFM to detect the attractive or repulsive force between the tip and the sample (Fig. 2b). AFM has several measurement modes, which are mainly divided into the contact mode and the dynamic force mode (tapping mode). In the contact mode, the tip is in contact with the sample surface and experiences repulsive force, and the feedback signal is applied to keep the cantilever at a constant position during scanning. On the other hand, in the dynamic force mode, the cantilever oscillates at or near the resonance frequency without contact with the sample surface and detects attractive force, and the feedback signal is applied to make vibrational amplitude constant during scanning. In general, the spatial resolution and images of AFM strongly depend on both the size and shape of the tip. The influence of the tip morphology on the AFM image is unavoidable, which is emphasized on samples of small size. Although AFM is a versatile and convenient technique to investigate sample surfaces, it does not have the ability to distinguish materials and chemical species.

In the field of plasmonics, the dynamic force mode is the most commonly used method to evaluate the structural and morphological properties of plasmonic functional materials for photoelectrochemical solar cells, photocatalysts, optical films, and plasmonic sensors.17–22,47,48 Furthermore, AFM can manipulate atoms and molecules.78 By using this technique, plasmonic nanoparticles were manipulated to form a hot spot at the gap between two nanoparticles for SERS measurements.79 AFM combined with optical excitation can visualise the morphological changes of samples induced by light, and has allowed the observation of chemical reactions induced by LSP of plasmonic metal nanoparticles loaded on the sample surfaces.47–49 In the AFM system, LSP can be excited by using a tip coated with plasmonic metals such as Au and Ag instead of a conventional tip made of Si or SiN.80

2.3 Tip-enhanced Raman spectroscopy (TERS)

TERS is an SPM technique that is combined with SERS to achieve the detection of Raman signals, which provides chemical information on the sample surface at a high spatial resolution. Early TERS was based on the configuration of AFM and the Raman signals were strongly enhanced by LSP excited at the AFM tip coated with Au or Ag. Owing to earnest research effort, AFM-TERS systems under ambient conditions are now commercially available. The spatial resolution of a typical AFM-TERS is ∼10 nm, which is mainly determined by the tip size. Furthermore, AFM-TERS has been widely used for chemical characterisation of sample surfaces, large molecules, and biomolecules. STM incorporated with TERS instead of AFM has been developed to improve the spatial resolution, and, recently, 1.7 nm spatial resolution has been obtained even under ambient conditions.81 The biggest problem in the TERS technique is to assure the reproducibility. Under the ambient conditions at room temperature, it is difficult to control the fluctuation of Raman signals. In addition, the quality of the tip including shape and cleanness directly influences the sensitivity of TERS. Ultimately, STM-TERS having spatial resolution at a single molecule level has been attained by combining with UHV-LT-STM,65 suppressing the instabilities caused by heat and gas atmospheres.

In TERS, some molecules have been found to be transformed during measurement, which was indicated by time-dependent changes in TERS spectra.38 In short, chemical reactions of the molecules proceeded in the presence of LSP of the tip. Although both bond formation and dissociation reactions have been investigated with AFM- or STM-TERS,38,56–58 the underlying mechanism is still unrevealed and engages the researchers’ interest.

2.4 Kelvin probe force microscopy (KPFM)

KPFM, a derivative of the AFM technique, can map the local potential at a sample surface with a lateral resolution of a sub-10 nm scale and potential resolution of a few mV. Using this technique, topographic and potential images can be simultaneously obtained. KPFM measures the contact potential difference (CPD, VCPD) between the tip and the sample, which corresponds to the relative difference of the work function between the tip (ϕTip) and the sample (ϕSample); VCPD = (ϕSampleϕTip)/e. In KPFM measurements, the distance between the tip and sample is kept constant to avoid the influence of capacitance gradients which contribute to the electrical force between the tip and sample. An offset bias voltage (Voff) is applied to cancel out the electrostatic force between the tip and the sample, and the obtained surface potential equals −Voff. KPFM has been utilised to evaluate the local electric properties of the various surfaces such as electronic devices, organic thin films, and nanomaterials.82 Although KPFM can visualize the surface potential at a high spatial resolution, the calibration of the tip against a reference sample with a well-defined work function is necessary to obtain the absolute values of the surface potential or work function. This indicates that contamination or oxidation of the tip should be avoided. Measurement time to acquire both topographic and potential images is relatively long. However, severe cross-talk between topographic and surface potential signals occurs at a high speed scan rate.

KPFM combined with optical excitation by light is an advanced approach to visualise the distribution of photogenerated carriers in the local structures of photoresponsive materials and opto-electronic devices.83–86 For example, the charge transport dynamics at the donor–acceptor interfaces of solar cells has been examined by in situ measurements with KPFM to gain requisite knowledge for the improvement of device performance.83 Furthermore, by taking advantage of the strengths of high spatial resolution, KPFM combined with optical excitation has been recently adopted for the real-space investigations of local phenomena induced by the LSP of metal NSs.51–54,87–89

3 Applications to chemical reactions based on plasmon-induced charge separation (PICS)

PICS occurs at the interface between a plasmonic metal NS and a semiconductor (Fig. 1a) under light irradiation with energy that corresponds to the resonance energy of LSP and is lower than the band gap of the semiconductor. The widely accepted understanding of PICS in previous studies has been that electron transfer takes place from the metal NS to the semiconductor. Furthermore, a Schottky junction formed by the direct contact between the metal and semiconductor is necessary to allow PICS and suppresses the recombination of the separated electrons and holes at the interface. Recently, an interfacial charge-transfer transition at the interface with strong mixing of the metal NS and semiconductor, where the electrons are directly excited from the metal to semiconductor, has been proposed.30 PICS has been applied to photovoltaic and photoelectric devices, photocatalysts, chemical sensors and biosensors, and storage of information.10,31 Although PICS has been studied eagerly with respect to the aforementioned applications and its mechanism ever since it was reported for the first time in 2005,26 a detailed description of the mechanism has been unestablished so far. In this section, we review the studies using SPM techniques to provide mechanistic insights into PICS based on nanoscale observations.

3.1 Visualisation of PICS sites with AFM

Ag nanoparticles exhibit brilliant colour in the entire range of visible light by tuning the resonant wavelength of LSP. In the case of the Ag nanoparticles loaded on TiO2, PICS occurs at the interface and results in oxidative dissolution from Ag to Ag+ ions. The oxidative dissolution reaction of Ag nanoparticles is wavelength-dependent due to LSP, which has been applied to multicolour photochromism.90,91 The morphological changes of Ag nanoparticles caused by PICS were ascertained by AFM observations of Ag nanoparticles deposited on a TiO2 single crystal surface before and after light irradiation.47–49 The real-space observation revealed that the reaction was completely suppressed when the relative humidity was ∼0% and there was negligible adsorbed water on TiO2.92 In other words, the adsorbed water layer on the TiO2 surface was necessary to initiate the reaction. The generated Ag+ ions diffused in the adsorbed water layer and recombined with the electrons transferred to TiO2, leading to redeposition as small satellite nanoparticles surrounding the original nanoparticle (Fig. 3a). It was possible to control the distance between the satellite nanoparticles and the original nanoparticle by humidity because the ionic mobility was regulated by the thickness of the adsorbed water layer. Thus, at appropriate humidity, the satellite nanoparticles can serve as markers to visualise the PICS sites instead of observations of a slight morphological change in the original nanoparticle.
image file: c9cp02100k-f3.tif
Fig. 3 (a) Schematic illustration of oxidative dissolution of a plasmon resonant Ag nanoparticle (NP) and the redeposition process in PICS. (b) Morphological changes in an Ag nanorod on TiO2(100) caused by sequential irradiation with the light at 1000, 900, and 800 nm polarised along the long axis. White arrows show the direction of the incident electric field. (c) Ag nanorod before and after irradiation with 560 nm light polarised along the short axis. (d) Maps of redeposited satellite Ag NPs around the original Ag nanorod with aspect ratios (>2.7) when (left) the longitudinal mode and (right) the transverse mode were excited. (e) Spatial distributions of simulated electric fields of (left) longitudinal and (right) transverse modes surrounding the Ag nanorod (length = 57 nm and width = 19 nm) with the FDTD method. Reproduced with permission from ref. 48. Copyright 2011 The Royal Society of Chemistry.

The oxidative dissolution reaction caused by PICS of pyramidal Ag nanorods photocatalytically deposited on a rutile TiO2(100) single crystal was investigated using AFM combined with optical excitation in N2 gas at 50% relative humidity.48 Nanorods have two resonant modes that are excited with light polarised along the long axis (the longitudinal mode) and along the short axis (the transverse mode). The resonant wavelength of the longitudinal mode is longer than that of the transverse mode. After the longitudinal mode was excited by polarised light in the near-infrared (NIR) region, the nanorods resonant with light became shorter and the small satellite nanoparticles were deposited in the vicinity of the short sides of the nanorods (Fig. 3b), which indicated that oxidative dissolution occurred from the short sides. The plasmon resonant wavelength of the longitudinal mode of the nanorods strongly depends on the aspect ratio. Thus, the aspect ratio-selective reaction was possible by selecting the irradiation wavelength, and the sequential shortening of a pyramidal Ag nanorods on TiO2 was achieved by sequential irradiation with 1000, 900, and 800 nm light (Fig. 3b). It was demonstrated that the excitation of the multipole resonance mode of LSP also resulted in PICS, followed by the oxidative dissolution of Ag nanorods with a high aspect ratio.93 On the other hand, when the transverse mode was excited with visible light, the oxidative dissolution occurred from the long sides of the nanorods, and the small satellite nanoparticles appeared near the long sides (Fig. 3c).

Since the satellite nanoparticles serve as markers reflecting the PICS sites, the statistically obtained anisotropic spatial distribution of the satellite nanoparticles indicates that the PICS sites are distributed anisotropically according to the excitation modes (Fig. 3d). Anisotropic plasmonic metal nanoparticles such as nanorods and nanoprisms have an anisotropic distribution of the optical near field depending on the direction of the incident electric field.1 In particular, the longitudinal mode of the nanorods generates an electric field localised at the short side of the nanorods. In contrast, the transverse mode localises the electric field at the long sides. To compare the PICS sites with the spatial distribution of the electric fields surrounding the pyramidal Ag nanorods, simulations with a finite-difference time domain (FDTD) method were performed.

The PICS sites revealed by real-space observations with AFM (Fig. 3d) showed good agreement with the spatial distribution of the electric fields simulated with the FDTD method (Fig. 3e). This result suggests that the optical near field is responsible for the electron transfer from the metal nanoparticle to the conduction band (CB) of the oxide semiconductor.

3.2 In situ observation of charge distribution caused by PICS with KPFM

KPFM combined with optical excitation has been applied to in situ observations of the spatial distribution of the separated charge carriers in PICS. The photoinduced surface potential changes on truncated triangular Au nanoplates and polydisperse spheroids, which were photocatalytically deposited on a 0.05 wt% Nb-doped rutile TiO2(100) single crystal, were evaluated with KPFM in a closed chamber filled with dry N2 gas at atmospheric pressure (Fig. 4a).51 Technically, the laser spot was focused on the midpoint of the cantilever to remove the influence of the laser beam (830 nm) used for oscillating the cantilever and so as to not excite the nanoplates under the tip. The surface potential difference between the Au nanoplate and TiO2 (EAuETiO2) was +31 ± 18 mV in the dark, but −25 ± 12 mV under UV light irradiation. The polarity change of the surface potential difference indicates that the electrons in the CB of TiO2 excited from the valence band (VB) by UV light irradiation were transferred to Au. This result is consistent with the reports that metal nanoparticles on TiO2 trap excited electrons in the CB and promote charge separation.94,95 The electron transfer from the TiO2 CB to the metal is interrupted by the band bending due to the Schottky barrier at the interface between the TiO2 and metal. However, when the excited electrons accumulate in the CB, the potential of the CB shifts negatively, and finally, the electrons overflow from the CB to the metal (Fig. 4b).
image file: c9cp02100k-f4.tif
Fig. 4 (a) Schematic illustration of the KPFM combined with optical excitation. (b and c) (left) Energy level diagrams and (right) cross-sectional illustrations of charge distributions at the Au nanoplate–TiO2 interface under (b) UV and (c) visible-NIR light irradiation. (b) Under UV light irradiation, the holes migrate to the TiO2 surface while excited electrons go into TiO2. After the accumulation of the electrons in the CB, resulting in the decrease in the band bending, the electrons spill over into the Au nanoplate. (c) Electrons in the plasmon resonant Au nanoplate are transferred to the CB of TiO2. Reproduced with permission from ref. 51. Copyright 2014 John Wiley and Sons.

The Au nanoplates deposited on TiO2 (Fig. 5a) exhibit absorption in the NIR region based on LSP (Fig. 5b). When the sample was irradiated with NIR light, the surface potential on the nanoplates (EAu) shifted to the positive side, although the change of ETiO2 was negligible. This suggested that the electrons excited by LSP were injected from the resonant Au nanoplate to TiO2, leading to the accumulation of holes in the nanoplate (Fig. 4c), which is the opposite process to the case of TiO2 excitation by UV light irradiation. Note that the Au nanoparticles were not oxidatively dissolved by PICS even in air because of their chemical stability. The plasmon resonant wavelength and intensity of the truncated triangular nanoplates depend on their size, thickness, and the degree of truncation.1 Fig. 5c shows the action spectra of the photoinduced surface potential changes (Δ(EAuETiO2)) acquired on nanoplates with different morphologies (Fig. 5a). The peak wavelengths of the action spectra for plates A, B, and C were ∼60 nm shorter than those of the corresponding simulated absorption spectra, which resulted from a slightly bigger permittivity value of undoped rutile TiO2 used for the simulation than Nb-doped rutile TiO2. In contrast, the peak position difference of plate D was too large to explain only by the difference of the permittivity. Furthermore, although the absorption intensity of plate D was larger than that of the other plates, the peak intensity in the action spectrum was smaller. This suggested that the threshold wavelength of the potential response was 1300–1400 nm (0.95–0.89 eV), which is consistent with the study reporting that PICS-based photocurrents for Au nanorods on rutile TiO2 were observed only at wavelengths ≤1300 nm.96 The threshold energy of PICS is determined mainly by the Schottky barrier. The theoretically estimated value of the Schottky barrier at the Au–TiO2 interface was 1.23 eV, which was slightly higher than the experimentally obtained values, and could be explained in terms of the difference between the surface under vacuum and the epitaxial interface with a certain mismatch.97 In summary, the real-space and in situ observations of charge distributions caused by PICS at the Au nanoparticle/TiO2 interfaces with KPFM revealed a positive potential shift on the Au surfaces, indicating that the electrons in the resonant Au nanoparticles overcame the Schottky barrier and were transferred to the conduction band of TiO2.

image file: c9cp02100k-f5.tif
Fig. 5 (a) Morphological images of Au nanoplates on Nb-doped rutile TiO2(100). The scale bar is 100 nm. (b) Absorption spectra simulated by the FDTD method. The inset shows the simulation models ([h, l, t] (in nm) = [15, 140, 25], [11.5, 140, 25], [11.5, 150, 27], and [12.5, 220, 40] for A, B, C, and D, respectively). (c) Action spectra of Δ(EAuETiO2). Dashed curves were obtained by Lorentzian fitting around the peaks. Reproduced with permission from ref. 51. Copyright 2014 John Wiley and Sons.

The separated charge carriers in PICS at different interfaces were also visualised using the KPFM technique. Lee et al. revealed that PICS induced a back-electron transfer at the interface between Ag nanoparticles and p-type poly(pyrrole) (PPy) nanowires (NW) (Fig. 6a).53 The contact potential difference (VCPD) between the tip and the sample surface was measured by KPFM (Fig. 6b). At the interface, electron transfer from Ag to p-type PPy NW occurred in the dark (Fig. 6c) because of energy level alignment between the work function of Ag (∼4.62 eV) and the Fermi level of PPy (∼4.93 eV), and thus, the direction of the band bending was opposite to the case of the Au–TiO2 interface (Fig. 6d). Both the plasmon of Ag nanoparticles and the band gap of PPy NW were excited with blue light and the surface potential on the Ag nanoparticles and the PPy NW shifted negatively and positively, respectively (Fig. 6c). This observation with KPFM indicated that the back-electron transfer from the PPy NW to the Ag nanoparticles occurred in PICS at the metal/p-type semiconductor nanostructure interface. Fig. 6d shows two PICS pathways available under blue light irradiation. Hot electrons generated by the plasmon were injected from Ag to the PPy NW (the curved blue arrow in Fig. 6d), with resonance energy transfer (RET) also occurring simultaneously (purple arrow). The electrons excited in the CB of the PPy NW were transferred to Ag along the band bending (the red arrow in Fig. 6d) and the holes in the VB move to the inner region of the PPy NW (black arrow).

image file: c9cp02100k-f6.tif
Fig. 6 (a) Topographic AFM image and (b) VCPD mapping of Ag nanoparticles attached on PPy NW. (c) ΔVCPD = VCPD(PPyAgNW) − VCPD (PPy NW or Ag) for Ag nanoparticles and PPy NW without illumination and under blue light irradiation, respectively. (d) Two available pathways of PICS at the interface between Ag nanoparticles and p-type PPy NW under blue light irradiation. Reproduced with permission from ref. 53. Copyright 2018 American Chemical Society.

On the other hand, at the interface between the Au nanoparticles and n-type ZnO NW, a positive shift on Au and a negative shift on ZnO NW were visualised with KPFM, which resulted from electron transfer from Au to ZnO. The back-electron transfer was never observed in the Au/ZnO system, which corresponds to the case of Au (Ag) nanoparticles on TiO2. In conclusion, the band structures at the interface between the metal and semiconductor determine the pathway of electron transfer in PICS.

3.3 AFM-based technique for the measurement of photocurrent caused by PICS

Giugni et al. developed a novel nanoscopy technique based on an AFM setup by exploiting PICS at a Schottky junction (Fig. 7a).50 They evaluated photocurrent generated by the PICS at the junction between the Au-coated AFM tip and n-type GaAs substrate. The surface plasmon was adiabatically focused at the tip apex by fabricating grating structures along the tip surface for efficient coupling with light. Owing to adiabatic focusing of the plasmon, electrons were efficiently transferred from the Au tip to the n-type GaAs substrate (Fig. 7b), resulting in the generation of a photocurrent under light irradiation. By using this nanoscopy technique, both topographic and photocurrent maps of the GaAs surfaces with local patterns of thin oxide and ion-implanted lines were successfully obtained.
image file: c9cp02100k-f7.tif
Fig. 7 (a) Schematic of a nanoscopy technique based on an AFM setup. (b) Energy diagram at the metal/semiconductor interface indicating the transfer of the electrons generated through plasmon decay to the conduction band of GaAs. Reproduced with permission from ref. 50. Copyright 2013 Nature Publishing Group.

Lee et al. successfully visualized the active sites of PICS with photoconductive AFM (pc-AFM).60 In the pc-AFM measurements, the sample of triangular Au nanoprisms on n-type TiO2 was irradiated with a laser light, which passed through a fused silica prism and was focused on the backside of the sample (Fig. 8a). Topographic and current images were simultaneously obtained (Fig. 8b and c). Since a Schottky junction was formed at the interface between a Au nanoprism and n-type TiO2, PICS caused electron transfer from Au to TiO2, resulting in the generation of photocurrent (Fig. 8a). The local photocurrent mapping revealed that the boundary of the Au nanoprism exhibited the highest activity under the light at the plasmon resonant wavelength, and the spatial distribution of photocurrent (Fig. 8c) well matched with the distribution of plasmonic electric field simulated with the FDTD method (Fig. 8d). This result indicates that the enhancement of electron transfer preferentially occurred at the boundary where the electric field was confined.

image file: c9cp02100k-f8.tif
Fig. 8 (a) Schematic illustration of the pc-AFM measurement on a triangular Au nanoprism/n-type TiO2 interface, and a Schottky junction at the interface. (b) Topographic AFM image of the Au nanoprism/n-type TiO2 interface. (c) Current image obtained under irradiation with light at 640 nm. (d) Electric field distribution simulated with the FDTD method. Reproduced with permission from ref. 60. Copyright 2019 American Chemical Society.

4 Applications to plasmon-induced chemical reactions

The number of reports on plasmon-induced chemical reactions (Fig. 1b) including both bond formation and dissociation reactions has steadily increased since 2010. Various reactions, such as N[double bond, length as m-dash]N bond formation, carbon–carbon coupling, epoxidation, oxidation, reduction, and dissociation reactions, have been accomplished.46 Although most studies have been conducted using conventional macroscopic methods, several others have successfully applied SPM techniques to the real-space observations of the reactions. In this section, we summarise the studies where plasmon-induced chemical reactions were followed using SPM techniques.

4.1 Plasmon-induced chemical reactions investigated with TERS

The plasmon-induced N[double bond, length as m-dash]N bond formation in the chemical transformation of p-aminothiophenol (pATP) (or p-nitrothiophenol (pNTP)) to 4,4′-dimercaptoazobenzene (DMAB) was proposed to explain the changes in the Raman signals in SERS measurements.33,34,98–100 This chemical reaction has attracted considerable attention of researchers and been earnestly studied to clarify the reaction mechanism. Deckert and Weckhuysen et al. reported nanoscale monitoring of the N[double bond, length as m-dash]N bond formation by using time-resolved TERS for the first time.38 A Ag-coated AFM tip was used not only to enhance the Raman signal but also to induce the reaction by the gap-plasmon between the tip and Au nanoplate (Fig. 9). The tip was positioned on a self-assembled monolayer of pNTP adsorbed on the Au nanoplates during the measurements. The TERS signals were detected under 633 nm light irradiation and the reaction was activated by irradiation with 532 nm laser light. Measuring the TERS spectrum and activating the reaction were performed in turn, which enabled time-dependent TERS measurements. In the subsequent studies, TERS techniques were also applied to monitor the dissociation reaction56,57 and dehydrogenation.58
image file: c9cp02100k-f9.tif
Fig. 9 Schematic overview of the AFM-TERS setup. The Ag-coated AFM tip is used not only as a TERS tip but also as a photocatalyst for the conversion of pNTP to DMAB. Reproduced with permission from ref. 38. Copyright 2012 Nature Publishing Group.

4.2 Visualisation of plasmon-assisted deformation of a photosensitive polymer with AFM

Azobenzene polymers have been used to visualise the spatial distribution of the electric field intensity of the light because the surface topography of such polymers changes when the wavelength of the incident light matches the absorption of azobenzene. Galarreta et al. studied the distribution of the plasmonic electric field generated around metallic nanotriangles by observing the morphological changes of an azobenzene polymer consisting of 4′-(((2-(methacryloyloxy)ethyl)ethyl)amino)-4-nitroazobenzene-co-methyl methacrylate (p(DR1M-co-MMA)) using the AFM technique.55 The periodic hexagonal arrays of Au and Ag nanotriangles were fabricated by electron beam lithography. The length of the triangle base and thickness were 320 ± 20 nm and 40 ± 5 nm, respectively. The azobenzene polymer was deposited on the metallic nanotriangles by spin-coating and a uniform thin film was formed (Fig. 10a). The AFM images scanned before and after irradiation with 532 nm light revealed that the thin film of the azobenzene polymer was deformed by light irradiation and the deformation occurred preferentially at the location where the electric field of the plasmon was strong (Fig. 10b). In addition, the morphological change caused by the deformation of the polymer near the Ag nanotriangles was more drastic than that near the Au nanotriangles. This is consistent with the results of FDTD simulations, indicating that the electric field intensity near the Ag nanotriangles was higher than Au. The direction of linearly polarised light also influenced the deformation behaviour. In summary, the AFM observations led to the conclusion that surface deformation of the azobenzene polymer could be induced when the wavelength of the light matched the absorption of both LSP and azobenzene moieties.
image file: c9cp02100k-f10.tif
Fig. 10 AFM images of periodic hexagonal arrays of Ag nanotriangles coated with a thin film of p(DR1M-co-MMA) before (a) and after (b) irradiation with 532 nm light. Reproduced with permission from ref. 55. Copyright 2011 American Chemical Society.

4.3 Single-molecule study of plasmon-induced bond dissociation with STM

The optical near field of the LSP is highly localised near the metal nanostructure and its intensity dramatically decreases with an increase in distance from the metal surface. In order to obtain mechanistic insights into plasmon-induced chemical reactions, a real-space observation of the reactions induced in the optical near field is necessary. We applied UHV-LT-STM combined with optical excitation to analyse the plasmon-induced chemical reaction of dimethyl disulfide, (CH3S)2, adsorbed on Ag(111) and Cu(111) surfaces (Fig. 11a).43 The gap plasmon between the Ag tip and metal substrate was excited by monochromatic visible light. The STM images before and after the excitation of the plasmon clarified that the dissociation of the S–S bond in (CH3S)2 was induced preferentially under the tip where the plasmonic field intensity was the strongest (Fig. 11b). UHV-LT-STM not only visualises the reactive sites but is also utilised for the investigation of the reaction mechanism and pathway. The action spectrum, which is the light energy dependence of the reaction yield, is crucial for understanding the fundamental aspects of the reaction. Therefore, we obtained the action spectra for the plasmon-induced dissociation as well as photodissociation. By comparing the energy dependence between plasmon-induced dissociation and photodissociation, it was concluded that the S–S bond dissociation induced by the plasmon occurred through the direct electronic excitation from the HOMO-derived molecular orbital (nS) to the LUMO-derived molecular orbital (σSS*). This excitation pathway was revealed by the combination study of STM and DFT calculations. Furthermore, the dissociation pathway was clarified by analysing the changes in tunnelling current caused by the dissociation reaction at the single molecule level. Finally, it was concluded that the plasmon-induced S–S bond dissociation occurred along the dissociative potential energy surface without vibrational excitation and the pathway initiated from a transient negative ion state could be excluded. Additional details of the mechanism and methodologies have been described in previous reports.43,101
image file: c9cp02100k-f11.tif
Fig. 11 (a) Experimental setup for the real-space investigation of the plasmon-induced chemical reaction in the nanogap between the Ag tip and a metal substrate. STM images of (CH3S)2 molecules adsorbed on Ag(111) (b) before and (c) after irradiation with p-polarised light at 532 nm. The tip was positioned at the centre of 10 nm wide concentric rings during light irradiation. Reprinted with permission from Science (, ref. 43. Copyright 2018 American Association for the Advancement of Science.

5. Conclusions and outlook

The combination of SPM with optical excitation by light has provided mechanistic insights into chemical reactions induced by LSP on the basis of real-space observations of the reactions in the optical near field of LSP. The real-space observations with tapping mode AFM have revealed that charge separation at the interface between a plasmonic metal NS and a semiconductor occurs preferentially in the optical near field, leading to chemical reactions. In addition, the AFM-based technique was developed and used to measure the photocurrent generated in PICS. Based on the in situ KPFM measurements of charge distribution, it was proven that the pathway of the electron transfer in PICS is determined by the energy band structure at the interface between a plasmonic metal NS and a semiconductor. In plasmon-induced chemical reactions, the deformation of a photosensitive polymer induced in the optical near field of LSP was visualised with AFM. The reactions were induced by LSP generated at the gap between the tip and the sample, and the chemical information was analysed in real-time with TERS. Furthermore, UHV-LT-STM visualised the plasmon-induced chemical reaction at a single molecule level, and helped clarify the reaction mechanism and pathway.

At present, the number of studies using SPM techniques to analyse the chemical reactions induced by LSP is small, and SPM combined with optical excitation by light has the potential to give deeper mechanistic insights into both PICS and the plasmon-induced chemical reactions. For further sophisticated studies, advanced SPM techniques are required.

Characterization of chemical species at a single molecule level would be pivotal especially for the investigation of associative bond formation reactions induced by LSP. Although STM-IETS has been used to identify chemical species, the application to unknown molecules is difficult because its selection rule is still undiscovered. Recently, visualizing vibrational normal modes of a single molecule has been attained with UHV-STM-TERS.102 This technique will contribute to the understanding of the surface reaction processes, and bring breakthrough in the field of both surface science and plasmon chemistry. Non-contact AFM,103,104 which can visualise chemical bonds within a single molecule, will also be useful in gaining direct evidence of chemical transformation. In electrochemical environments, electrochemical TERS (EC-TERS)105 will become a prominent tool to investigate the redox reactions based on PICS. Although technical improvement and stabilization of EC-TERS have been achieved recently,106,107 further research effort is needed to make EC-TERS a versatile analytical method. For clarification of the excitation process in PICS and the excitation pathway of molecules by LSP, the development of SPM techniques combined with ultrafast laser spectroscopies is eagerly awaited. All of these advanced SPM techniques are difficult but merit further studies and development.

Conflicts of interest

There are no conflicts to declare.


This work was supported in part by KAKENHI (Grant No. 21225001, 18H05257, and 18H01947) of the Ministry of Education, Culture, Sports, Science, and Technology (MEXT) of Japan.


  1. K. L. Kelly, E. Coronado, L. L. Zhao and G. C. Schatz, J. Phys. Chem. B, 2003, 107, 668 CrossRef CAS.
  2. J. A. Schuller, E. S. Barnard, W. Cai, Y. C. Jun, J. S. White and M. L. Brongersma, Nat. Mater., 2010, 9, 193 CrossRef CAS PubMed.
  3. Z. W. Liu, J. M. Steele, W. Srituravanich, Y. Pikus, C. Sun and X. Zhang, Nano Lett., 2005, 5, 1726 CrossRef CAS PubMed.
  4. J. Homola, S. S. Yee and G. Gauglitz, Sens. Actuators, B, 1999, 54, 3 CrossRef CAS.
  5. J. N. Anker, W. P. Hall, O. Lyandres, N. C. Shah, J. Zhao and R. P. Van Duyne, Nat. Mater., 2008, 7, 442 CrossRef CAS PubMed.
  6. P. L. Stiles, J. A. Dieringer, N. C. Shah and R. R. Van Duyne, Annu. Rev. Anal. Chem., 2008, 1, 601 CrossRef CAS PubMed.
  7. V. E. Ferry, J. N. Munday and H. A. Atwater, Adv. Mater., 2010, 22, 4794 CrossRef CAS PubMed.
  8. H. Nakanishi, K. J. M. Bishop, B. Kowalczyk, A. Nitzan, E. A. Weiss, K. V. Tretiakov, M. M. Apodaca, R. Klajn, J. F. Stoddart and B. A. Grzybowski, Nature, 2009, 460, 371 CrossRef CAS PubMed.
  9. S. Kuhn, U. Hakanson, L. Rogobete and V. Sandoghdar, Phys. Rev. Lett., 2006, 97, 017402 CrossRef PubMed.
  10. C. Clavero, Nat. Photonics, 2014, 8, 95 CrossRef CAS.
  11. S. Linic, P. Christopher and D. B. Ingram, Nat. Mater., 2011, 10, 911 CrossRef CAS PubMed.
  12. E. Hao and G. C. Schatz, J. Chem. Phys., 2004, 120, 357 CrossRef CAS PubMed.
  13. K. L. Wustholz, A. I. Henry, J. M. McMahon, R. G. Freeman, N. Valley, M. E. Piotti, M. J. Natan, G. C. Schatz and R. P. Van Duyne, J. Am. Chem. Soc., 2010, 132, 10903 CrossRef CAS PubMed.
  14. R. C. Jin, Y. C. Cao, E. C. Hao, G. S. Metraux, G. C. Schatz and C. A. Mirkin, Nature, 2003, 425, 487–490 CrossRef CAS PubMed.
  15. C. J. Murphy, T. K. San, A. M. Gole, C. J. Orendorff, J. X. Gao, L. Gou, S. E. Hunyadi and T. Li, J. Phys. Chem. B, 2005, 109, 13857 CrossRef CAS PubMed.
  16. Y. N. Xia, Y. J. Xiong, B. Lim and S. E. Skrabalak, Angew. Chem., Int. Ed., 2009, 48, 60 CrossRef CAS PubMed.
  17. N. Chandrasekharan and P. V. Kamat, J. Phys. Chem. B, 2000, 104, 10851 CrossRef CAS.
  18. T. Ung, L. M. Liz-Marzan and P. Mulvaney, J. Phys. Chem. B, 2001, 105, 3441 CrossRef CAS.
  19. O. Crespo-Biel, B. Dordi, D. N. Reinhoudt and J. Huskens, J. Am. Chem. Soc., 2005, 127, 7594 CrossRef CAS PubMed.
  20. S. Gupta, P. Uhlmann, M. Agrawal, S. Chapuis, U. Oertel and M. Stamm, Macromolecules, 2008, 41, 2874 CrossRef CAS.
  21. P. Sangpour, F. Hashemi and A. Z. Moshfegh, J. Phys. Chem. C, 2010, 114, 13955 CrossRef CAS.
  22. B. Lahiri, G. Holland, V. Aksyuk and A. Centrone, Nano Lett., 2013, 13, 3218 CrossRef CAS PubMed.
  23. P. Myrach, N. Nilius and H. J. Freund, Phys. Rev. B: Condens. Matter Mater. Phys., 2011, 83, 035416 CrossRef.
  24. E. Le Moal, S. Marguet, D. Canneson, B. Rogez, E. Boer-Duchemin, G. Dujardin, T. V. Teperik, D. C. Marinica and A. G. Borisov, Phys. Rev. B, 2016, 93, 035418 CrossRef.
  25. A. Yu, S. W. Li, G. Czap and W. Ho, Nano Lett., 2016, 16, 5433 CrossRef CAS PubMed.
  26. Y. Tian and T. Tatsuma, J. Am. Chem. Soc., 2005, 127, 7632 CrossRef CAS PubMed.
  27. W. Hou and S. B. Cronin, Adv. Funct. Mater., 2013, 23, 1612 CrossRef CAS.
  28. C. Wang and D. Astruc, Chem. Soc. Rev., 2014, 43, 7188 RSC.
  29. M. Valenti, M. P. Jonsson, G. Biskos, A. Schmidt-Otta and W. A. Smith, J. Mater. Chem. A, 2016, 4, 17891 RSC.
  30. K. Wu, J. Chen, J. R. McBride and T. Lian, Science, 2015, 349, 632 CrossRef CAS PubMed.
  31. T. Tatsuma, H. Nishi and T. Ishida, Chem. Sci., 2017, 8, 3325 RSC.
  32. K. Watanabe, D. Menzel, N. Nilius and H.-J. Freund, Chem. Rev., 2006, 106, 4301 CrossRef CAS PubMed.
  33. Y.-F. Huang, H.-P. Zhu, G.-K. Liu, D.-Y. Wu, B. Ren and Z.-Q. Tian, J. Am. Chem. Soc., 2010, 132, 9244 CrossRef CAS PubMed.
  34. Y. Fang, Y. Li, H. Xu and M. Sun, Langmuir, 2010, 26, 7737 CrossRef CAS PubMed.
  35. H. Zhu, X. Ke, X. Yang, S. Sarina and H. Liu, Angew. Chem., Int. Ed., 2010, 49, 9657 CrossRef CAS PubMed.
  36. P. Christopher, H. Xin and S. Linic, Nat. Chem., 2011, 3, 467 CrossRef CAS PubMed.
  37. P. Christopher, H. Xin, A. Marimuthu and S. Linic, Nat. Mater., 2012, 11, 1044 CrossRef CAS PubMed.
  38. E. M. van Schrojenstein Lantman, T. Deckert-Gaudig, A. J. G. Mank, V. Deckert and B. M. Weckhuysen, Nat. Nanotechnol., 2012, 7, 583 CrossRef CAS PubMed.
  39. M. J. Kale, T. Avanesian and P. Christopher, ACS Catal., 2014, 4, 116 CrossRef CAS.
  40. M. L. Brongersma, N. J. Halas and P. Nordlander, Nat. Nanotechnol., 2015, 10, 25 CrossRef CAS PubMed.
  41. S. Linic, U. Aslam, C. Boerigter and M. Morabito, Nat. Mater., 2015, 14, 567 CrossRef CAS PubMed.
  42. Y. Zhang, S. He, W. Guo, Y. Hu, J. Huang, J. R. Mulcahy and W. D. Wei, Chem. Rev., 2018, 118, 2927 CrossRef CAS PubMed.
  43. E. Kazuma, J. Jung, H. Ueba, M. Trenary and Y. Kim, Science, 2018, 360, 521 CrossRef CAS PubMed.
  44. U. Aslam, V. Govind Rao, S. Chavez and S. Linic, Nat. Catal., 2018, 1, 656 CrossRef.
  45. L. Zhou, D. F. Swearer, C. Zhang, H. Robatjazi, H. Zhao, L. Henderson, L. Dong, P. Christopher, E. A. Carter, P. Nordlander and N. J. Halas, Science, 2018, 362, 69 CrossRef CAS PubMed.
  46. E. Kazuma and Y. Kim, Angew. Chem., Int. Ed., 2019, 58, 4800 CrossRef CAS PubMed.
  47. K. Matsubara and T. Tatsuma, Adv. Mater., 2007, 19, 2802 CrossRef CAS.
  48. E. Kazuma, N. Sakai and T. Tatsuma, Chem. Commun., 2011, 47, 5777 RSC.
  49. I. Tanabe and T. Tatsuma, Nano Lett., 2012, 12, 5418 CrossRef CAS PubMed.
  50. A. Giugni, B. Torre, A. Toma, M. Francardi, M. Malerba, A. Alabastri, R. P. Zaccaria, M. I. Stockman and E. Di Fabrizio, Nat. Nanotechnol., 2013, 8, 845 CrossRef CAS PubMed.
  51. E. Kazuma and T. Tatsuma, Adv. Mater. Interfaces, 2014, 1400066 CrossRef.
  52. M. Gwon, A. Sohn, Y. Cho, S. H. Phark, J. Ko, Y. S. Kim and D. W. Kim, Sci. Rep., 2015, 5, 16727 CrossRef CAS PubMed.
  53. S. H. Lee, S. W. Lee, T. Oh, S. H. Petrosko, C. A. Mirkin and J. W. Jang, Nano Lett., 2018, 18, 109 CrossRef CAS PubMed.
  54. M. M. Giangregorio, M. Losurdo, G. V. Bianco, A. Operamolla, E. Dilonardo, A. Sacchetti, P. Capezzuto, F. Babudri and G. Bruno, J. Phys. Chem. C, 2011, 115, 19520 CrossRef CAS.
  55. B. C. Galarreta, I. Rupar, A. Young and F. Lagugne-Labarthet, J. Phys. Chem. C, 2011, 115, 15318 CrossRef CAS.
  56. M. Sun, Z. Zhang, Z. H. Kim, H. Zheng and H. Xu, Chem. – Eur. J., 2013, 19, 14958 CrossRef CAS PubMed.
  57. Z. Zhang, S. Sheng, H. Zheng, H. Xu and M. Sun, Nanoscale, 2014, 6, 4903 RSC.
  58. S. Chaunchaiyakul, A. Setiadi, P. Krukowski, F. C. Catalan, M. Akai-Kasaya, A. Saito, N. Hayazawa, Y. Kim, H. Osuga and Y. Kuwahara, J. Phys. Chem. C, 2017, 121, 18162 CrossRef CAS.
  59. H. Böckmann, S. Gawinkowski, J. Waluk, M. B. Raschke, M. Wolf and T. Kumagai, Nano Lett., 2018, 18, 152 CrossRef PubMed.
  60. H. Lee, H. Lee and J. Y. Park, Nano Lett., 2019, 19, 891 CrossRef PubMed.
  61. G. Binning and H. Rohrer, Helv. Phys. Acta, 1982, 55, 726 Search PubMed.
  62. R. M. Feenstra, Surf. Sci., 1994, 965, 299 Search PubMed.
  63. B. C. Stipe, M. A. Rezaei and W. Ho, Science, 1998, 280, 1732 CrossRef CAS PubMed.
  64. Y. Kim, K. Motobayashi, T. Frederiksen, H. Ueba and M. Kawai, Prog. Surf. Sci., 2015, 90, 85 CrossRef CAS.
  65. R. Zhang, Y. Zhang, Z. C. Dong, S. Jiang, C. Zhang, L. G. Chen, L. Zhang, Y. Liao, J. Aizpurua, Y. Luo, J. L. Yang and J. G. Hou, Nature, 2013, 498, 82 CrossRef CAS PubMed.
  66. X. H. Qiu, G. V. Nazin and W. Ho, Science, 2003, 299, 542 CrossRef CAS PubMed.
  67. Y. Zhang, Y. Luo, Y. Zhang, Y.-J. Yu, Y.-M. Kuang, L. Zhang, Q.-S. Meng, Y. Luo, J.-L. Yang, Z.-C. Dong and J. G. Hou, Nature, 2016, 531, 623 CrossRef CAS PubMed.
  68. H. Imada, K. Miwa, M. Imai-Imada, S. Kawahara, K. Kimura and Y. Kim, Nature, 2016, 538, 364 CrossRef CAS PubMed.
  69. L. Bartels, F. Wang, D. Möller, E. Knoesel and T. F. Heinz, Science, 2004, 305, 648 CrossRef CAS PubMed.
  70. M. Mehlhorn, H. Gawronski and K. Morgenstern, Phys. Rev. Lett., 2010, 104, 076101 CrossRef PubMed.
  71. H. Böckmann, S. Liu, J. Mielke, S. Gawinkowski, J. Waluk, L. Grill, M. Wolf and T. Kumagai, Nano Lett., 2016, 16, 1034 CrossRef.
  72. M. J. Comstock, N. Levy, A. Kirakosian, J. Cho, F. Lauterwasser, J. H. Harvey, D. A. Strubbe, J. M. J. Fréchet, D. Trauner, S. G. Louie and M. F. Crommie, Phys. Rev. Lett., 2007, 99, 038301 CrossRef PubMed.
  73. J. Henzl, P. Puschnig, C. Ambrosch-Draxl, A. Schaate, B. Ufer, P. Behrens and K. Morgenstern, Phys. Rev. B: Condens. Matter Mater. Phys., 2012, 85, 035410 CrossRef.
  74. E. Kazuma, J. Jung, H. Ueba, M. Trenary and Y. Kim, J. Am. Chem. Soc., 2017, 139, 3115 CrossRef CAS PubMed.
  75. R. J. Hermann and M. J. Gordon, Opt. Express, 2018, 26, 27668 CrossRef CAS PubMed.
  76. T.-X. Huang, S.-C. Huang, M.-H. Li, Z.-C. Zeng, X. Wang and B. Ren, Anal. Bioanal. Chem., 2015, 407, 8177 CrossRef CAS PubMed.
  77. B. Yang, E. Kazuma, Y. Yokota and Y. Kim, J. Phys. Chem. C, 2018, 122, 16950 CrossRef CAS.
  78. O. Custance, R. Perez and S. Morita, Nat. Nanotechnol., 2009, 4, 804 CrossRef PubMed.
  79. T. Fukui, H. Naiki and S. Masuo, J. Phys. Chem. C, 2017, 121, 19329 CrossRef CAS.
  80. A. Sanders, R. W. Bowman, L. Zhang, V. Turek, D. O. Sigle, A. Lombardi, L. Weller and J. J. Baumberg, Appl. Phys. Lett., 2016, 109, 153110 CrossRef.
  81. C. Chen, N. Hayazawa and S. Kawata, Nat. Commun., 2014, 5, 3312 CrossRef PubMed.
  82. W. Melitz, J. Shena, A. C. Kummela and S. Lee, Surf. Sci. Rep., 2011, 66, 1 CrossRef CAS.
  83. E. J. Spadafora, R. Demadrille, B. Ratier and B. Grevin, Nano Lett., 2010, 10, 3337 CrossRef CAS PubMed.
  84. Q. Chen, H. P. Zhou, T. B. Song, S. Luo, Z. R. Hong, H. S. Duan, L. T. Dou, Y. S. Liu and Y. Yang, Nano Lett., 2014, 14, 4158 CrossRef CAS PubMed.
  85. S. Panigrahi, S. Jana, T. Calmeiro, D. Nunes, R. Martins and E. Fortunato, ACS Nano, 2017, 11, 10214 CrossRef CAS PubMed.
  86. J. L. Garrett, E. M. Tennyson, M. Hu, J. S. Huang, J. N. Munday and M. S. Leite, Nano Lett., 2017, 17, 2554 CrossRef CAS PubMed.
  87. S. Diefenbach, E. Parzinger, J. Kiemle, J. Wierzbowski, S. Funke, B. Miller, R. Csiki, P. Thiesen, A. Cattani-Scholz, U. Wurstbauer and A. W. Holleitner, J. Phys. Chem. C, 2018, 122, 9663 CrossRef CAS.
  88. J.-Z. Wang, Z.-Q. Guo, J.-P. Zhou and Y.-X. Lei, Nanotechnology, 2018, 29, 305709 CrossRef PubMed.
  89. F. Zhao, W. M. Yang, T. M. Shih, S. L. Feng, Y. J. Zhang, J. F. Li, J. W. Yan and Z. L. Yang, ACS Photonics, 2018, 5, 3519 CrossRef.
  90. Y. Ohko, T. Tatsuma, T. Fujii, K. Naoi, C. Niwa, Y. Kubota and A. Fujishima, Nat. Mater., 2003, 2, 29 CrossRef CAS PubMed.
  91. E. Kazuma and T. Tatsuma, Chem. Commun., 2012, 48, 1733 RSC.
  92. K. Matsubara, K. L. Kelly, N. Sakai and T. Tatsuma, Phys. Chem. Chem. Phys., 2008, 10, 2263 RSC.
  93. E. Kazuma and T. Tatsuma, J. Phys. Chem. C, 2013, 117, 2435 CrossRef CAS.
  94. Z. Zhang and J. T. Yates Jr., Chem. Rev., 2012, 112, 5520 CrossRef CAS PubMed.
  95. V. Subramanian, E. E. Wolf and P. V. Kamat, J. Am. Chem. Soc., 2004, 126, 4943 CrossRef CAS PubMed.
  96. Y. Nishijima, K. Ueno, Y. Yokota, K. Murakoshi and H. Misawa, J. Phys. Chem. Lett., 2010, 1, 2031 CrossRef CAS.
  97. A. H. M. Abdul Wasey, R. Batabyal, J. C. Mahoto, B. N. Dev, Y. Kawazoe and G. P. Das, Phys. Status Solidi B, 2013, 7, 1313 CrossRef.
  98. M. Sun, Y. Huang, L. Xia, X. Chen and H. Xu, J. Phys. Chem. C, 2011, 115, 9629 CrossRef CAS.
  99. L. Kang, P. Xu, B. Zhang, H. Tsai, X. Han and H.-L. Wang, Chem. Commun., 2013, 49, 3389 RSC.
  100. D.-Y. Wu, X.-M. Liu, Y.-F. Huang, B. Ren, X. Xu and Z.-Q. Tian, J. Phys. Chem. C, 2009, 113, 18212 CrossRef CAS.
  101. E. Kazuma, J. Jung, H. Ueba, M. Trenary and Y. Kim, Prog. Surf. Sci., 2018, 93, 163 CrossRef CAS.
  102. J. Lee, K. T. Crampton, N. Tallarida and V. A. Apkarian, Nature, 2019, 568, 78 CrossRef CAS PubMed.
  103. L. Gross, F. Mohn, N. Moll, P. Liljeroth and G. Meyer, Science, 2009, 325, 1110 CrossRef CAS PubMed.
  104. L. Gross, F. Mohn, N. Moll, B. Schuler, A. Criado, E. Guitián, D. Peña, A. Gourdon and G. Meyer, Science, 2012, 337, 1326 CrossRef CAS PubMed.
  105. Z.-C. Zeng, S.-C. Huang, D.-Y. Wu, L.-Y. Meng, M.-H. Li, T.-X. Huang, J.-H. Zhong, X. Wang, Z.-L. Yang and B. Ren, J. Am. Chem. Soc., 2015, 137, 11928 CrossRef CAS PubMed.
  106. Y. Yokota, N. Hayazawa, B. Yang, E. Kazuma, F. C. I. Catalan and Y. Kim, J. Phys. Chem. C, 2019, 123, 2953 CrossRef CAS.
  107. G. Goubert, X. Chen, S. Jiang and R. P. Van Duyne, J. Phys. Chem. Lett., 2018, 9, 3825 CrossRef CAS PubMed.

This journal is © the Owner Societies 2019