Over the last century, quantum theories have revolutionized our understanding of material properties. One of the most striking quantum phenomena occurring in heterogeneous media is the quantum tunneling effect, where carriers can tunnel through potential barriers even if the barrier height exceeds the carrier energy. Interestingly, the tunneling process can be accompanied by the absorption or emission of light. In most tunneling junctions made of noble metal electrodes, these optical phenomena are governed by plasmonic modes, i.e., light-driven collective oscillations of surface electrons. In the emission process, plasmon excitation via inelastic tunneling electrons can improve the efficiency of photon generation, resulting in bright nanoscale optical sources. On the other hand, the incident light can affect the tunneling behavior of plasmonic junctions as well, leading to phenomena such as optical rectification and induced photocurrent. Thus, plasmonic tunneling junctions provide a rich platform for investigating light–matter interactions, paving the way for various applications, including nanoscale light sources, sensors, and chemical reactors. In this paper, we will introduce recent research progress and promising applications based on plasmonic tunneling junctions.

In quantum theory, a particle with energy E can penetrate a potential barrier whose potential is higher than its own energy E. This is called the tunneling effect. According to the WKB approximation, its transmission probability is1 

T=e2γ,
(1)

where γ=1/0apxdx, a is the width of the barrier, and p is the momentum of the particle. An obvious consequence of this dependence is that the transmission probability decreases with the width and height of the barrier. One of the most frequently encountered cases of tunneling effects is the electron tunneling between two materials. Typically, electrons can migrate from one electrode to another even if they are separated by a thin insulating layer. Due to the Fermi–Dirac distribution of the electron velocities inside the electrodes, the tunneling current can be obtained by integrating the product of the transmission probability [Eq. (1)] with the electron velocity distribution over the velocity range. The tunneling current density can be written as2,3

J=J0φeAφφ+eVeAφ+eV,
(2)

where φ is the average potential barrier height, J0=e/2πhβΔs2, A=4πβΔs/h2m, β is a correction factor, Δs is the width of the barrier at the starting point of the tunneling, and h is the Planck constant. The current density includes two exponential terms, which account for two opposite electron flows—the reflection and the transmission. Thus, the high bias voltages lower the barrier height, and the thinner insulators can be advantageous to realize tunneling, resulting in larger current densities. In most practical situations, the thickness of insulator layers should be on the order of a few nanometers (<5nm).

In many nanophotonic applications, tunneling junctions consist of noble metal electrodes, which can support the generation of surface plasmons, typically in the form of gap modes. The excitation of strongly confined modes leads to a strong field enhancement due to the small mode volume, making them fascinating platforms to study light–matter interactions. The gap plasmons have been employed in many applications,4 such as strong plasmon-exciton coupling,5,6 enhanced Raman scattering,7–9 fluorescence modifications,10 and nonlinear light generation.11,12 Combining plasmonic gap modes with tunneling junctions opens up a myriad of possibilities for exploring and tailoring light–matter interactions, as we discuss below.

Plasmonic gap modes can be excited by inelastic tunneling electrons as a by-product of tunneling current generation in the nanometric junctions. The excited plasmons can couple to free-space modes, resulting in the emission of light. Thus, a tunneling junction can be treated as a nanoscale light source whose properties are determined by its materials’ properties. Besides the traditional metal–insulator–metal tunneling junctions,3,13–18 inelastic tunneling and light emission have been realized in metal–insulator–semiconductor junctions,19,20 metal–2D material junctions,21–24 and molecular junctions.25,26

A larger tunneling current typically provides a higher number of emitted photons. According to Eq. (2), a thinner insulating layer and a higher voltage could be beneficial for generating strong light emission. The limiting factors for such devices are related to the stability of the junction under the high voltages and the physical limitations of fabricating tunnel gaps with decreasing thicknesses. Furthermore, the plasmonic properties start to diverge from the classical predictions in ultrathin gaps because of quantum effects. Specifically, the charge transfer effects diminish the field enhancement in the gaps due to the quenching of the near fields.7,27,28 Although these effects on light generation in such small gaps are still not completely understood, a recent study has shown that the light emission is indeed related to the charge-transfer modes in the ultrathin gaps.29 However, these effects can usually be ignored because most experimental conditions involve tunneling junctions with thick enough insulator layers (>1 nm) that are not dominated by the charge-transfer modes.

The ability to tailor the properties of light-emitting plasmonic tunneling junctions is critical for achieving realistic applications. To that end, various configurations and geometric parameters have been explored. For instance, interference and diffraction of the propagating surface plasmon polaritons can be achieved in the tunneling junctions.30,31 Aperiodic groove arrays have been successfully used to control the surface plasmon polariton (SPP) propagation.32 The spatial control of light emissions was realized in the alternating current (ac)-driven Al–AlOx–Cu tunneling junctions with changing barrier thicknesses.33 Furthermore, in order to increase the light emission efficiency, considerable efforts have been dedicated to the optimization of the electron-to-photon conversion efficiency, including nanostructured optical antennas,3,18,21 rough electrode surfaces,34 and resonant tunneling junctions.35,36

On the other hand, incident light or excited plasmons can affect the electronic properties of tunneling junctions as well. Photon-assisted tunneling and optical rectification have been realized in many experiments.37–43 Similar to electric rectifiers, optical rectifiers can convert ac signals into direct current (dc) signals. This transformation provides a method to detect ultrafast ac signals at optical frequencies, which are challenging to be measured directly due to their high frequencies. These phenomena that involve interactions between light and electronic excitations can be used to characterize properties of sub-nanometric gaps, such as near-field enhancement and temperature.38,44,45

Besides direct applications of optically-active tunneling junctions based on light emission and absorption, these systems can be attractive for various other indirect applications. The emitted photons can be used to monitor the chemical reactions,46,47 even at single-molecule levels.47 Tunneling junctions encompassing molecules can work as solid-state molecular devices controlling charge transport.48,49 Finally, the electronic properties of the tunneling junctions can be treated as sensors and monitors to reveal physical processes.42,43,50–52

In this perspective article, we will first discuss the light emission in junctions based on inelastic tunneling, including its basic mechanisms, generation of plasmonic modes, optical antennas, and above-threshold light emission. Next, photon-assisted tunneling will be introduced, focusing on optical rectification, thermal effects, and photoemission. Finally, novel chemical applications based on the tunneling junctions will be reviewed.

A nanoscale plasmonic light source can be realized using junctions, which support inelastic electron tunneling. In general, the electroluminescence in the tunneling junctions can originate from two possible processes—inelastic tunneling (IET) and electron–hole recombination—depending on the junctions’ properties, such as their materials, geometries, and bias voltages.19 In particular, electroluminescence can originate from electron–hole recombination in junctions, including semiconductors such as Si. According to Ref. 19, for Si/SiO2/Au junctions, electroluminescence is mainly dominated by the recombination of electrons and holes for Si with a low doping level and by the quantum tunneling mechanism for Si with a high doping level. Here, we will focus on the former process in the quantum tunneling regime.

Most common tunneling junctions consist of one or two metallic electrodes, typically noble metals such as gold. These metal structures can support plasmonic modes, including gap modes and propagating modes. With the bias voltage applied on the junction, quantum tunneling can occur between two electrodes, whose current–voltage (I–V) characteristics can be described by the Simmons model.2 There are two kinds of tunneling electrons generated in this process—elastic tunneling electrons and inelastic tunneling electrons, as shown in Fig. 1(a). Due to energy conservation, it is obvious that elastic tunneling electrons do not lose energy and will not couple to the local density of optical states (LDOS), which is related to the density of electromagnetic modes in the volume and is proportional to the decay rate of a system.53 In contrast, the inelastic tunneling electrons can be coupled to LDOS, leading to the excitation of gap plasmons. Subsequently, gap plasmons can couple to free-space radiation through some processes [Fig. 1(b)], which are dependent on the details of junctions. When the bias voltage exceeds the threshold that, which usually is a turning point in the Fowler−Nordheim representation,3 the tunneling regime will be established in the field emission, leading to photon emission.3,39 Additionally, photons can be excited by inelastic tunneling electrons directly as well; however, the efficiency of this mechanism is extremely low because of the momentum mismatch between photons and electrons.14 

FIG. 1.

(a) Schematic of the tunneling process (b) Schematic of the electroluminescence processes in the tunneling junctions.

FIG. 1.

(a) Schematic of the tunneling process (b) Schematic of the electroluminescence processes in the tunneling junctions.

Close modal

Parzefall and Novotny have previously summarized the inelastic tunneling mechanism and the efficiency of the electroluminescence, providing a systematic guideline for these processes.54,55 Here, we briefly introduce this framework as a basis for the following discussions. Inelastic tunneling is typically discussed in the context of three main theoretical approaches—the energy-loss model, the current fluctuation model, and the spontaneous emission model.54,55 The latter model offers a connection between IET rate and LDOS ρopt and a suitable description of plasmon excitation in nanogaps.54,55 The IET rate can be calculated by integrating the spectral inelastic tunneling rate γinelω over all energies Γinel=0γinel(ω)dω, where ℏω is the mode energy. In most cases, Γel ≫ Γinel, therefore, without considering the optical properties of nanogaps the efficiency of IET in vacuum can be approximated as ηele=γinelω,ρopt=ρ0/Γel where ρ0 is the vacuum LDOS, which is called the source efficiency. This quantity is determined by the bias voltage, the gap distance, and the barrier height; however, it is not dependent on the optical properties of the system. Its efficiency is on the order of 10−8 eV−1, which is an extremely weak process.54 

The source efficiency of IET in a vacuum is only dependent on the electronic properties of the nanogaps. However, IET needs to couple to the optical modes to excite plasmons in the tunneling gaps. Thus, the optical properties of the system play a significant role in the total source efficiency of IET due to the high LDOS in the tunneling junctions. The total electroluminescence efficiency ηtot includes three parts,54 

ηtot=ηele×ηopt×ηem,
(3)

where ηele is the ratio of the inelastic and elastic tunneling, ηopt = ρopt/ρ0 is the ratio of the LDOS of the junction and the vacuum LDOS, and ηem is the outcoupling efficiency of the gap plasmons to surface plasmon polariton (SPP) or photon conversion. The second part is only dependent on the optical properties of the system. There are several contributions to the LDOS in the nanogaps, including gap plasmon modes, propagating plasmon modes (SPPs), and free photon modes. However, the LDOS of gap modes is several orders of magnitude higher than other contributions,54 which means that gap modes dominate the IET excitation process.54,55 Therefore, the free-space modes or the propagating modes are unlikely to be directly excited via IET. Equation (3) implies that inelastic electrons are coupled to gap modes and omits the direct coupling of the inelastic tunneling electrons to free-space modes because the probability of the latter is extremely low. Most research studies focus on these three main factors to optimize the electroluminescence at the tunneling junctions.

Although optically excited plasmons are more common in recent literature, electron-excited modes trace their origins to the first observations of plasmonic phenomena in the 1950s.56 This was followed by the observation of light emission based on inelastic tunneling two decades later.13 More recently, this field has attracted renewed attention due to its promising applications as nanoscale optical devices.15,25,26

As mentioned above, exciting gap modes is the necessary and crucial intermediate process to induce surface plasmon emission. In Ref. 14, a tunneling junction consisting of a smooth 20 nm gold film and a gold STM tip was used to observe the excitation of SPPs on a metal/air interface. Thus, the propagating modes were excited by a dipole-like gap plasmon formed between the sharp tip and the film. However, the emission ring corresponding to the omnidirectional propagation of SPPs was replaced with a uniform emission pattern when the gold film was replaced by a 5 nm granular gold film, indicating that the increased surface roughness of the film leads to the disappearance of the propagating modes and their replacement with localized modes. It shows that gap modes excited by IET are indeed unable to outcouple to free space directly. Furthermore, a gold nanowire was used to separate the photon emission region from the tunneling junction region, as shown in Fig. 2(a). Photon emissions are detected successfully at the end of the nanowire, as shown in Fig. 2(b). Notwithstanding the low efficiency of only 5% due to the non-optimal tunneling junction structure, these experiments demonstrated a clear pathway for energy conversion from electrons to photons and the potential for the development of electrically-driven nanoscale photon sources.14,57

FIG. 2.

(a) Diagram of the experimental setup. (b) Photon emission map overlapped with the scanning electron microscope (SEM) image; Inset: emission intensity along the nanowire. Reproduced with permission from Bharadwaj et al., Phys. Rev. Lett. 106, 226802 (2011). Copyright 2011, American Physical Society.

FIG. 2.

(a) Diagram of the experimental setup. (b) Photon emission map overlapped with the scanning electron microscope (SEM) image; Inset: emission intensity along the nanowire. Reproduced with permission from Bharadwaj et al., Phys. Rev. Lett. 106, 226802 (2011). Copyright 2011, American Physical Society.

Close modal

However, there is not only one pathway for converting plasmons to free-space photons after the gap mode is excited. Reference 34 shows three main paths to outcouple the gap modes in tunneling junctions: first, gap mode plasmons can be scattered into photons at the junction interfaces; second, gap modes can be coupled to propagating modes at the metal–dielectric interfaces; and third, gap modes can be coupled to photon modes and propagating modes at the junction edge, as shown in Figs. 3(a) and 3(b). For example, the outcoupling route in Ref. 14 can be attributed to pathway 2, where gap plasmons are converted to propagating modes.

FIG. 3.

(a) and (b) Three outcoupling paths of gap plasmons in a Al–Al2O3–Au junction. Reproduced with permission from Makarenko et al., Adv. Sci. 7, 1900291 (2020). Copyright 2020 Author(s), licensed under a Creative Commons Attribution 4.0 License. (c) Sketch of the source and detector tunneling junctions. (d) Response current densities as a function of time with different bias voltages. Reproduced with permission from Du et al., Nat. Photonics 11, 623–627 (2017). Copyright 2017 Author(s), licensed under a Creative Commons Attribution 4.0 License.

FIG. 3.

(a) and (b) Three outcoupling paths of gap plasmons in a Al–Al2O3–Au junction. Reproduced with permission from Makarenko et al., Adv. Sci. 7, 1900291 (2020). Copyright 2020 Author(s), licensed under a Creative Commons Attribution 4.0 License. (c) Sketch of the source and detector tunneling junctions. (d) Response current densities as a function of time with different bias voltages. Reproduced with permission from Du et al., Nat. Photonics 11, 623–627 (2017). Copyright 2017 Author(s), licensed under a Creative Commons Attribution 4.0 License.

Close modal

In addition, research has shown that the roughness of the gap has an important effect on its plasmonic properties.58 The outcoupling efficiencies of tunnel junctions can be improved by the gap roughness, especially for pathway 1, because of the roughness-induced momentum matching.34 In Ref. 34, the outcoupling efficiency of Al–Al2O3–Au junctions is successfully raised to 0.62% by optimizing the electrode thickness and roughening the junction areas.

Most electron-excited plasmons are detected by their accompanying light emissions in the tunneling junctions. However, they can also be sensed directly by tunneling currents without the plasmon-to-photon conversion. As shown in Fig. 3(c), this can be done using two adjacent tunneling junctions.15 One of them is treated as a plasmonic source, and another one is treated as a detector. Similar to other tunneling mechanisms, the gap plasmons are excited in the source junction first, and then the gap mode couples to the propagating mode, which can cross over the gap between the two junctions. Finally, the propagating plasmons are detected by the right tunneling junction via the optical rectification effect. The tunneling current of the detector junction can change in the presence of propagating plasmons, as shown Fig. 3(d).15 

Compared to the plasmonic source, the direct light emission source3,16–18,21–23 is a much more attractive potential application based on the tunneling junction because of its compact size, ultrafast excitation time, and widely tunable emission wavelength. According to recent studies, three factors can affect the light emission rate—the electronic properties, the LDOS of the junction, and the output efficiency.54,55

Although roughening the junctions mentioned above can improve the outcoupling efficiency, this approach will lead to saturation with increasing roughness.54 Another method to increase the outcoupling efficiency is to optimize the geometry of electrodes in the junctions by using optical antennas. Instead of using planar electrodes, gratings,59 nanoparticles,23 and nano antennas3,16,18 can be employed as the electrodes in the tunneling gaps to improve the efficiencies by increasing both the LDOS and the outcoupling efficiency.

For example, a grating structure—a gold periodical antenna—was used as the top-layer electrode in an Al–Al2O3–Au tunneling junction,59 as shown in Fig. 4(a). Its electron-to-photon efficiency can reach 1.6 × 10−6 photons per electron, which is 2700 times higher than the efficiency of the planar junction. Due to modified scattering cross sections, the antenna array significantly modifies the electroluminescence (EL) spectra compared to planar junctions, as shown in Fig. 4(b).

FIG. 4.

(a) Optical microscope and SEM images of a junction with grating structures. (b) Left: Experimental normalized EL spectra of the periodical antenna junctions with different bias voltages; Right: Theoretical normalized EL spectra of the periodical antenna junctions and the planar junction with different bias voltages. Reproduced with permission from Zhang et al., Nat. Commun. 10, 4949 (2019). Copyright 2019 Author(s), licensed under a Creative Commons Attribution 4.0 License.

FIG. 4.

(a) Optical microscope and SEM images of a junction with grating structures. (b) Left: Experimental normalized EL spectra of the periodical antenna junctions with different bias voltages; Right: Theoretical normalized EL spectra of the periodical antenna junctions and the planar junction with different bias voltages. Reproduced with permission from Zhang et al., Nat. Commun. 10, 4949 (2019). Copyright 2019 Author(s), licensed under a Creative Commons Attribution 4.0 License.

Close modal

Similar to the grating nanostructures, a nano antenna with a nanoparticle in the gap was successfully used as a tunneling junction to emit light, as shown in Fig. 5(a).3 This structure not only can improve the electron-to-photon efficiency but can also provide a platform to modify the EL spectra by providing the gap modes needed to couple to the free-space modes (photon modes) via scattering. As shown in Fig. 5(b), the scattering spectrum profiles with different geometries match the EL profiles well. Thus, the scattering cross sections of nano antennas can modify the EL spectra by changing the outcoupling efficiency. Furthermore, the cutoff frequencies of the EL spectra are limited by the bias voltages, which is evidence of the underlying inelastic tunneling mechanism. This nanostructure is utilized to realize nanoscale Yagi-Uda antennas,18 as shown in Fig. 5(c). These nanoscale antennas can provide highly unidirectional emission in the optical regime with reflectors and directors, as shown in Fig. 5(d).

FIG. 5.

(a) SEM image of tunneling junctions. Nano antennas with different shapes are shown in (b), left: SEM images; right: EL (open circles) and scattering spectra (solid lines). Reproduced with permission from Kern et al., Nat. Photonics 9, 582–586 (2015). Copyright 2015, Nature Publishing Group. (c) SEM image of a Yagi-Uda antenna. (d) The highly unidirectional emission based on the simulation of the Yagi-Uda antenna. Reproduced with permission from Kullock et al., Nat. Commun. 11, 115 (2020). Copyright 2020 Author(s), licensed under a Creative Commons Attribution 4.0 License.

FIG. 5.

(a) SEM image of tunneling junctions. Nano antennas with different shapes are shown in (b), left: SEM images; right: EL (open circles) and scattering spectra (solid lines). Reproduced with permission from Kern et al., Nat. Photonics 9, 582–586 (2015). Copyright 2015, Nature Publishing Group. (c) SEM image of a Yagi-Uda antenna. (d) The highly unidirectional emission based on the simulation of the Yagi-Uda antenna. Reproduced with permission from Kullock et al., Nat. Commun. 11, 115 (2020). Copyright 2020 Author(s), licensed under a Creative Commons Attribution 4.0 License.

Close modal

The optical antennas’ output efficiency is mostly dependent on their scattering cross sections. In the gaps, however, the LDOS sometimes does not exactly overlap with the scattering cross sections. This mismatch can lead to low total output efficiency. As shown in Figs. 6(a) and 6(b), an optical antenna with two edge-to-edge silver cubes provides (1.8 ± 0.2) × 10−3 external quantum efficiency when the scattering cross section is not matched with the LDOS.16 With geometric optimizations, the external quantum efficiency finally increases by a factor of 10 and approaches 2% after the two are matched with each other, as shown in Fig. 6(c).

FIG. 6.

(a) Sketch of an edge-to-edge tunneling juncntion. (b) TEM image of the tunneling junction. (c) Simulated and experimental external quantum efficiency (EQE) of different nano antennas with different aspect ratios and heights (aspect ratio b/a, height c). Reproduced with permission from Qian et al., Nat. Photonics 12, 485–488 (2018). Copyright 2018 Author(s), licensed under a Creative Commons Attribution 4.0 License.

FIG. 6.

(a) Sketch of an edge-to-edge tunneling juncntion. (b) TEM image of the tunneling junction. (c) Simulated and experimental external quantum efficiency (EQE) of different nano antennas with different aspect ratios and heights (aspect ratio b/a, height c). Reproduced with permission from Qian et al., Nat. Photonics 12, 485–488 (2018). Copyright 2018 Author(s), licensed under a Creative Commons Attribution 4.0 License.

Close modal

Furthermore, 2D materials, including van der Waals materials, are widely used in tunneling optical antennas due to their special properties. For instance, hexagonal boron nitride (hBN), with its high potential barrier, can be used as a robust insulator layer to support stable tunneling. In Ref. 21, the tunneling junctions are realized using Au-hBN-Au structures, as shown in Fig. 7(a). The nanoslots are fabricated on one of the Au electrons with four separated sections to support different plasmonic modes. The emission profiles are determined by the transmittance spectra of these nanoslots, and the intensities are improved due to the high outcoupling efficiencies, as shown in Fig. 7(b). Simultaneously, fast modulation of light emission based on inelastic tunneling was demonstrated in this work. In addition to the insulator, 2D materials can also be used as electrodes in the tunneling junctions. For example, graphene is a good conductor, which is routinely used as a top electrode with an Au bottom electrode and a hBN insulator, Fig. 7(c).23 In such structures, the plasmonic modes are not supported by the graphene electrode but rather by the silver cubes on top of the graphene coupled to the bottom Au layer. Similarly, the emission profiles are modified by the plasmonic modes.

FIG. 7.

(a) SEM image of bottom electrodes with different slots for the tunneling junctions. (b) Emission spectra (open circle) and transmission enhancements (solid line) of these junctions. Reproduced with permission from Parzefall et al., Nat. Nanotechnol. 10, 1058–1063 (2015). Copyright 2015, Nature Publishing Group (c) Left: sketch of the tunneling junction consisting of the gold electrode, the hBN insulator layer, the graphene electrode, and the silver nanocube. Right-top: spatial distribution of emitted light from a nanocube; right-bottom: emission spectra from a nanocube.23 Copyright 2019 Author(s), licensed under a Creative Commons Attribution 4.0 License.

FIG. 7.

(a) SEM image of bottom electrodes with different slots for the tunneling junctions. (b) Emission spectra (open circle) and transmission enhancements (solid line) of these junctions. Reproduced with permission from Parzefall et al., Nat. Nanotechnol. 10, 1058–1063 (2015). Copyright 2015, Nature Publishing Group (c) Left: sketch of the tunneling junction consisting of the gold electrode, the hBN insulator layer, the graphene electrode, and the silver nanocube. Right-top: spatial distribution of emitted light from a nanocube; right-bottom: emission spectra from a nanocube.23 Copyright 2019 Author(s), licensed under a Creative Commons Attribution 4.0 License.

Close modal

Besides the optimization of the LDOS and the output efficiency, the first term in Eq. (3)—the electronic properties of the junction—can be used to enhance the electron-to-photon conversion efficiency as well, e.g., by using resonant tunneling structures. Recently, Qian et al. used metallic quantum well (MQW)-based tunnel junctions to realize the 30% external quantum efficiency.35 This offers a step forward in the development of tunneling junctions as compact plasmon/light sources.

So far, we have discussed light emission in tunneling junctions with photon energies below the applied bias threshold. However, above-threshold light emission can be observed in tunneling junctions as well. The cut-off frequencies of the light emission excited by inelastic tunneling discussed above are limited by the bias voltages applied on the junction due to the energy conversion. However, the cut-off frequencies of the above-threshold light emission can be much higher than the energy of the inelastic electrons. The above-threshold light emission has been observed in the break junctions,60–62 where the electromigration method is used to fabricate ultranarrow gaps supporting tunneling.17,39,60,61,63,64 Buret et al. explain this emission behavior by invoking spontaneous blackbody emission from hot electron gases thermalized by electron collisions.62 Recently, other mechanisms based on hot carriers generated by nonradiative plasmonic decay60 and electrons heated by plasmon decay at source electrodes20 have been used to describe above-threshold light emission processes.

As shown in Fig. 8(a), a metallic electromigration break junction is used to support tunneling and emit photons.60Figure 8(b) shows the Au/Cr junction’s emission spectra, clearly showing that huge parts of the emitted light have energies higher than the maximum energy loss of the inelastic electrons. The pure gold junctions have the highest photon yield efficiency compared to other lossy metals, such as Cr and Pd. Thus, it was concluded that the light emission is indeed influenced by the plasmonic decay rate. The lossy metal can decrease the number of plasmons excited by the inelastic electrons and reduce the hot-carrier generation rate. In Fig. 8(c), it also shows that the electron effective temperature Teff is linearly increased with the bias voltages, TeffV, which is evidence of the plasmon-induced hot carrier generation.60 In this mechanism, plasmons excited by inelastic tunneling electrons can induce hot electrons and hot holes via nonradiative decay, and recombination of these hot electrons and hot holes leads to above-threshold light emission, as shown in Fig. 8(d).

FIG. 8.

(a) Sketches of the electromigration break tunneling junction and the setup. (b) Emission spectra of the Au/Cr junction with different bias voltages. (c) Effective temperature as a function of the bias voltage for different junctions. (d) Sketch of the above-threshold light emission mechanism, reproduced with permission from Cui et al., Nano Lett. 20, 6067–6075 (2020). Copyright 2020 American Chemical Society.

FIG. 8.

(a) Sketches of the electromigration break tunneling junction and the setup. (b) Emission spectra of the Au/Cr junction with different bias voltages. (c) Effective temperature as a function of the bias voltage for different junctions. (d) Sketch of the above-threshold light emission mechanism, reproduced with permission from Cui et al., Nano Lett. 20, 6067–6075 (2020). Copyright 2020 American Chemical Society.

Close modal

In Ref. 20, a metal−insulator−semiconductor structure is used to realize above-threshold emission, as shown in Fig. 9(a). Because the electrons flow from the metal to the semiconductor, this structure can rule out the possibility of heating electrons by tunneling or by plasmons in the drain. Comparing the emission spectra to the bias voltages, it is clear that parts of the emitted photons have energies higher than the threshold (eVbias), which are shaded in red in Fig. 9(b). To analyze the mechanism behind this phenomenon, this paper extracted the electron temperature from the spectra by fitting a Fermi−Dirac distribution. In Fig. 9(c), the effective temperature increases linearly with the applied bias when the voltage is lower than the plasmonic frequency. However, the increase slows down when the voltage is higher than the plasmonic frequency. The cut-off frequencies have similar trends. Thus, the plasmon indeed plays an important role in the above-threshold light emission, and its mechanism is related to hot electrons based on the plasmon-electron interactions in the metal source electrode. Similarly, Shalem et al. claim that the energy of an electron bath is proportional to the plasmon energy.20 Although both of these studies involve the excitation of hot electrons via plasmon decay,20,60 their light emission processes are considerably different from each other. In Cui et al., the above-threshold photons are generated via the recombination of hot electrons and hot holes.60 In Shalem et al., however, the mechanism involves photon excitation through the inelastic tunneling of hot electrons.

FIG. 9.

(a) Sketch of a metal−insulator−semiconductor (Au/Al2O3/p-type Si) tunneling junction. (b) Emssion spectra for different bias voltages. (c) Left axis: electron temperature as a function of the bias voltages for two devices with different plasmonic frequencies; right axis: cut-off energy as a function of the bias voltages. Reproduced with permission from Shalem et al., Nano Lett. 21, 1282–1287 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 License. (d) Photon intensity as a function of photon energy with the light and electronic excitation. Reproduced with permission from Cui et al., Nano Lett. 21, 2658–2665 (2021). Copyright 2021, American Chemical Society.

FIG. 9.

(a) Sketch of a metal−insulator−semiconductor (Au/Al2O3/p-type Si) tunneling junction. (b) Emssion spectra for different bias voltages. (c) Left axis: electron temperature as a function of the bias voltages for two devices with different plasmonic frequencies; right axis: cut-off energy as a function of the bias voltages. Reproduced with permission from Shalem et al., Nano Lett. 21, 1282–1287 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 License. (d) Photon intensity as a function of photon energy with the light and electronic excitation. Reproduced with permission from Cui et al., Nano Lett. 21, 2658–2665 (2021). Copyright 2021, American Chemical Society.

Close modal

The mechanism of the above-threshold light emission in plasmonic tunneling junctions is still not completely understood. However, it offers a novel approach for controlling light emission in tunneling junctions. Recently, for example, the above-threshold light emission was shown to increase by a thousand times when the electromigration break junction was excited by both the tunneling electrons and the laser at the same time, as shown in Fig. 9(d).61 

After considering light emission from plasmonic tunneling junctions in Sec. II, here we will discuss a related phenomenon of photon-assisted tunneling, where the energy conversion is reversed, resulting in current generation upon light absorption in tunnel junctions. The experimental observation and theoretical description of photon-assisted tunneling were first developed in the 1960s, and the following research has generated considerable attention for decades, especially in quantum dot systems and molecular junctions.65,66 Recently, the plasmonic properties of the junctions have also successfully been used to assist tunneling due to their extremely large near fields and increased interaction cross sections.37–39 

When the tunneling junction is illuminated by a laser, the ac voltage induced on the junction via the optical fields with frequency ω can be expressed as Voptcosωt. Thus, its time-average tunneling current can be written as37–39,67

Itot=n=Jn2eVoptωIdcVdc+nωe,
(4)

where Jn is a Bessel function of the nth order, Vdc is the dc bias voltage, and Vopt is the amplitude of the optical voltage caused by the laser or plasmonic fields. According to Refs. 3739 and 67, if the variance of the junction conductance is much slower than that of the optical voltage, the dc tunneling current can be approximated as

ItotIdcVdc,Vopt=0+14Vopt22IdcVdc2,
(5)

by omitting the higher orders in Eq. (4). The first term of Eq. (5) is the regular tunneling current under the applied dc bias, and the second term is the photon-induced tunneling current, which is typically called photocurrent. This mechanism is called optical rectification, which has been observed in plasmonic tunneling junctions.37–39,68 The photocurrent is proportional to the nonlinear conductance and the intensity of the incident light.

Ac tunneling caused by incident light accompanies dc tunneling in many experimental configurations. However, the sensitivity and speed limitations of the measurement devices make it extremely challenging to detect the instantaneous ac quantum tunneling currents at optical frequencies. The time-average rectified dc tunneling currents can provide a chance to characterize the photon-induced tunneling behaviors.69 The standard lock-in techniques are widely used in investigations of optical rectification to extract relationships between the photocurrent and the nonlinear conductance.37–39,41

A low capacitance junction can decrease the response time of the tunneling, which can be beneficial to the realization of the optical rectification. The electromigration method mentioned in the last section is a good approach for fabricating small tunneling junctions supporting optical rectification.37,39,70 As shown in the left inset of Fig. 10(a), the electromigration break Au junctions were successfully employed to detect optical rectification.37 When the laser is incident on the gap area, the plasmons are excited, leading to the appearance of a photocurrent. The signals extracted from the lock-in setups show that the photocurrent is almost the same as the nonlinear conductance term with the varying dc bias voltages in Fig. 10(a). Thus, the mechanism of photocurrent generation can be identified as optical rectification. Simultaneously, as Fig. 10(b) shows, the photocurrent linearly increases with the incident laser power, which matches with the second term of Eq. (5) and is also evidence of the optical rectification.

FIG. 10.

(a) Photocurrent (red line) and 1/4Vopt22Idc/Vdc2 (black line) as a function of the dc bias voltage; Left inset: SEM image of the tunneling gap; Right inset: conductance as a function of the dc bias voltage. (b) Photocurrent as a function of incident laser power and the linear fitting line (red). Reproduced with permission from Ward et al., Nat. Nanotechnol. 5, 732–736 (2010). Copyright 2010, Nature Publishing Group. (c) Measured field enhancement in the tunneling gap as a function of the gap size; Inset: diagram of the tunneling gap. Reproduced with permission from Arielly et al., Nano Lett. 11, 2968–2972 (2011). Copyright 2011, American Chemical Society. (d) Top: laser signal and photocurrent as a function of time; bottom: geometry of tunneling junctions. Reproduced with permission from Kos et al., ACS Nano 15, 14535–14543 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 License.

FIG. 10.

(a) Photocurrent (red line) and 1/4Vopt22Idc/Vdc2 (black line) as a function of the dc bias voltage; Left inset: SEM image of the tunneling gap; Right inset: conductance as a function of the dc bias voltage. (b) Photocurrent as a function of incident laser power and the linear fitting line (red). Reproduced with permission from Ward et al., Nat. Nanotechnol. 5, 732–736 (2010). Copyright 2010, Nature Publishing Group. (c) Measured field enhancement in the tunneling gap as a function of the gap size; Inset: diagram of the tunneling gap. Reproduced with permission from Arielly et al., Nano Lett. 11, 2968–2972 (2011). Copyright 2011, American Chemical Society. (d) Top: laser signal and photocurrent as a function of time; bottom: geometry of tunneling junctions. Reproduced with permission from Kos et al., ACS Nano 15, 14535–14543 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 License.

Close modal

Molecule tunneling junctions are not only a good choice to realize rectification71–74 but are also suitable for the observation of the effect in the optical regime, because of their well-defined and controllable gap sizes.38,41 Similar to the surface-enhanced Raman spectroscopy (SERS) used to probe molecules in the plasmonic gaps,7,8,10,75,76 the relationship between the photocurrent and the ac voltage can provide a method for detecting the plasmonic near-fields in the gaps. In Ref. 38, the Au plasmonic tunneling junctions with the embedded monolayer 1-octanethiol (C8), 1-decanethiol (C10), and 1-dodecanethiol (C12) molecules are employed to realize optical rectification, as shown in the inset of Fig. 10(c). In this case, the gap distance is determined by the length of the molecules. Their I–V characteristics show that the photocurrent is proportional to the nonlinear conductance, proving that the process is governed by optical rectification. The optical ac voltages Vopt can be extracted from the measurements, which are a measure of the near-field enhancement in the plasmonic gaps compared to the incident free space power. Figure 10(c) shows that the field enhancement extracted from the photocurrent measurements decreases with the increasing gap size, which matches the simulation predictions. Similarly, a self-assembled molecular monolayer (SAM) junction with gold electrodes can be used for the optical rectification, with photocurrent being modulated by the incident light as shown in Fig. 10(d).41 

According to Eq. (5), the photocurrent generation is expected to exhibit linear dependence on the incident power. However, this behavior can break down for powers above a certain threshold. This scenario is explored in Ref. 39, where with the modest incident power, the photocurrent, the differential conductance, and the nonlinear conductance are enhanced along with a strong SHG signal when the laser illuminates the gap [Fig. 11(c)]. This behavior can again be described in the framework of the optical rectification phenomenon. As shown in Fig. 11(a), the photocurrent initially shows linear behavior. However, for larger powers, the dependence switches to the third power, and finally to the fifth power for even larger excitation powers. According to DFT calculations, these nonlinear photon-assisted tunneling effects can be attributed to d-band electrons of Au, whose density of states and transition processes are shown in Fig. 11(b). When the excitation power is large enough, d-band electrons can be excited to states near Fermi level, leading to nonlinear photocurrent via multiphoton absorptions.

FIG. 11.

(a) Photocurrent as a function of laser intensity for different dc bias voltages. Inset: fitting photocurrent and laser intensity by power law. (b) Calculated densities of states and transition processes. (c) Scanning maps for SHG, photocurrent, conductance, and nonlinear conductance with 900 mV dc bias voltage. Reproduced with permission from Stolz et al., Nano Lett. 14, 2330–2338 (2014). Copyright 2014, American Chemical Society.

FIG. 11.

(a) Photocurrent as a function of laser intensity for different dc bias voltages. Inset: fitting photocurrent and laser intensity by power law. (b) Calculated densities of states and transition processes. (c) Scanning maps for SHG, photocurrent, conductance, and nonlinear conductance with 900 mV dc bias voltage. Reproduced with permission from Stolz et al., Nano Lett. 14, 2330–2338 (2014). Copyright 2014, American Chemical Society.

Close modal

Besides optical rectification, there are other phenomena induced by incident light, such as thermal effects and photoemission currents, that can contribute to current generation in plasmonic tunneling junctions. Here, we will give a brief introduction to these interesting phenomena, which can provide additional avenues for manipulating the tunneling currents.

Although many investigations of optical rectification claim that thermal effects can be ignored due to their thick and symmetric plasmonic tunneling junctions,37–39 it should be noted that thermal effects, such as thermal voltage and thermal expansion, caused by the incident light can also contribute to photon-assisted tunneling in some special conditions.45,77–79 For example, due to the thermal expansion, the resistance of the junction can decrease for a suitable incident power, which causes the tunneling currents to increase.77 

Reference 39 observes the thermovoltage in the tunneling gaps when the incident power reaches 4 MW/cm2, which is much larger than the power (hundreds of kilowatts per square centimeter) used to realize nonlinear photo-assisted tunneling. Figure 12(a) shows a photocurrent with different signs between two sides of the gap. The thermovoltage changes its sign when the laser illuminates the different sides of the gap, leading to a temperature difference. Similar results are observed in Ref. 78, as shown in Fig. 12(b). When light is incident on a tunneling gap, a voltage difference is observed across the gap. It is claimed that the laser-induced voltage is caused by the electron energy-distribution differences between two sides of the gap, which is different from the conventional thermoelectric effect. Recently, Mennemanteuil et al. found that both optical rectification and thermal effects contribute to the tunneling current.70 Similarly, the changing I–V characteristics of the molecule junction can be used to detect the electronic and local lattice temperatures.45 

FIG. 12.

(a) SHG and photocurrent maps with a 4 MW/cm2 laser. Reproduced with permission from Stolz et al., Nano Lett. 14, 2330–2338 (2014). Copyright 2014 American Chemical Society. (b) Photovoltage map of Ti-free Au nanogap. Reproduced with permission from Zolotavin et al., J. Phys. Chem. Lett. 8, 1739–1744 (2017). Copyright 2017 American Chemical Society (c) Top: SEM images of the nanojunctions before and after laser ablation. Bottom: Photoemission current as a function of incident intensity. Reproduced with permission from Zimmermann et al., Nano Lett. 19, 1172–1178 (2019). Copyright 2019 American Chemical Society (d) Photocurrent and Keldysh parameters as functions of the incident power. Reproduced with permission from Zhou et al., Adv. Sci. 8, 2101572 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 License. (e) Sketch of hot electrons contributing to the current in a nanojunction. Reproduced with permission from Chalabi et al., Nano Lett. 14, 1374–1380 (2014). Copyright 2014 American Chemical Society.

FIG. 12.

(a) SHG and photocurrent maps with a 4 MW/cm2 laser. Reproduced with permission from Stolz et al., Nano Lett. 14, 2330–2338 (2014). Copyright 2014 American Chemical Society. (b) Photovoltage map of Ti-free Au nanogap. Reproduced with permission from Zolotavin et al., J. Phys. Chem. Lett. 8, 1739–1744 (2017). Copyright 2017 American Chemical Society (c) Top: SEM images of the nanojunctions before and after laser ablation. Bottom: Photoemission current as a function of incident intensity. Reproduced with permission from Zimmermann et al., Nano Lett. 19, 1172–1178 (2019). Copyright 2019 American Chemical Society (d) Photocurrent and Keldysh parameters as functions of the incident power. Reproduced with permission from Zhou et al., Adv. Sci. 8, 2101572 (2021). Copyright 2021 Author(s), licensed under a Creative Commons Attribution 4.0 License. (e) Sketch of hot electrons contributing to the current in a nanojunction. Reproduced with permission from Chalabi et al., Nano Lett. 14, 1374–1380 (2014). Copyright 2014 American Chemical Society.

Close modal

Photoemission caused by multiphoton absorption is another mechanism generating current in plasmonic tunneling junctions, although this cannot be classified as a “real” tunneling process as the electron energy can be higher than the potential barrier. As shown in Fig. 12(c), the tunneling junctions are typically formed by laser ablation.40 Zimmermann et al. find that the photocurrent is proportional to the 3.55 power of the squared laser field without bias voltage, which is characteristic for the photoemission process. According to the Keldysh theory, multiphoton absorption is believed to dominate the photoemission current in a weak laser field.40 For the strong laser field, their calculations show that the mechanism can turn into a tunneling process.40 In Ref. 80, three steps of photocurrent are observed with the increasing laser power, as shown in Fig. 12(d). For the lower power range, the photocurrent is proportional to the third power order of the incident low power, which is believed to be caused by a hybrid mechanism consisting of two-photon photoemission, one-photon-assisted tunneling, and direct tunneling.80 For the medium power range, the photocurrent increases slowly with the power and even decreases, which can be attributed to the disturbed Schottky barrier. The final process shows a strong nonlinear behavior, which is believed to be caused by the optical field-driven tunneling. Additionally, hot electrons excited by the laser can also generate a current in the plasmonic gap without the tunneling process, whose mechanism is described in Fig. 12(e).81 In conclusion, a plasmonic tunneling gap is a suitable platform to realize and monitor light–matter interactions due to its strong dependence on near-field enhancement and sensitive current.

In previous sections, we have discussed two main categories of applications. The first was nanoscale light sources based on the inelastic tunneling. Second, applications based on the photo-assisted tunneling, such as nanoscale sensors, which can detect certain physical properties in the extremely small volumes with fast response times were discussed. As a final example, in this section we will introduce how plasmonic tunneling junctions can be used in chemical reactions.

Figure 13(a) shows a plasmonic tunneling junction consisting of Au rods, Al2O3, a monolayer of poly-L-histidine (PLH), Ga2O3, and liquid eutectic gallium indium (EGaIn), where PLH is used as both a tunnel barrier and a reactant.47 This tunneling junction can support both oxidation and reduction reactions related to O2 and H2, facilitated by hot electrons in the Au rods.47 The interaction between O2 molecules and hot electrons can accelerate the generation of transient negative O2 , which finally becomes O atoms to oxidize Au. For the reduction reactions, the increased temperature via the relaxation of hot electrons can facilitate the reduction of Au oxides with H2. Simultaneously, due to hot electrons, H2 molecules can be dissociated into H atoms, which can strongly reduce Au oxides. The hot electrons can be excited by both the elastic tunneling and the light excitation, although they have different effects on the reactions due to their different excitation power densities. With the low optical excitation power, plasmon-induced hot electrons cannot increase the temperature to facilitate the reduction of Au oxides. On the contrary, the tunneling-induced hot electrons can raise the temperature due to higher electric excitation power. Figure 13(b) shows that the dc bias voltages have a strong influence on the reactions related to both O2 and H2, with higher voltages producing faster reaction rates. The chemical reactions driven by hot electrons are shown in Fig. 13(c), and oxidation of electrodes and barriers can change the electronic and optical properties of the junction, leading to varying tunneling currents and emitted light. Thus, this junction can be used not only as a nanoreactor but also as a sensor.

FIG. 13.

(a) Sketch of the tunneling gap used for chemical reactions. (b) Emission intensity changes as a function of bias voltages without incident light. (c) Processes of the chemical reactions in the tunneling junction. Reproduced with permission from Wang et al., Nat. Nanotechnol. 13, 159–164 (2018). Copyright 2017 Author(s), licensed under a Creative Commons Attribution 4.0 License. (d) Illustration of the real-time observation of chemical reactions. (e) Top: STM images; Bottom: Tunneling current and bias voltage as a function of time with incident light. (f) Top: STM images; Bottom: Tunneling current and bias voltage as a function of time without incident light. Red arrow: rotation; Green arrow: dissociation. Reproduced with permission from Kazuma et al., Science 360, 521–525 (2018). Copyright 2018 Author(s), licensed under a Creative Commons Attribution 4.0 License..

FIG. 13.

(a) Sketch of the tunneling gap used for chemical reactions. (b) Emission intensity changes as a function of bias voltages without incident light. (c) Processes of the chemical reactions in the tunneling junction. Reproduced with permission from Wang et al., Nat. Nanotechnol. 13, 159–164 (2018). Copyright 2017 Author(s), licensed under a Creative Commons Attribution 4.0 License. (d) Illustration of the real-time observation of chemical reactions. (e) Top: STM images; Bottom: Tunneling current and bias voltage as a function of time with incident light. (f) Top: STM images; Bottom: Tunneling current and bias voltage as a function of time without incident light. Red arrow: rotation; Green arrow: dissociation. Reproduced with permission from Kazuma et al., Science 360, 521–525 (2018). Copyright 2018 Author(s), licensed under a Creative Commons Attribution 4.0 License..

Close modal

Kazuma et al. used tunneling currents and tunneling electrons in a scanning tunneling microscope (STM) to excite and monitor single-molecule chemical reactions in real time.46 The dissociation reactions of dimethyl disulfide (CH3S)2 can be induced by a strong field due to LSPR or by injecting tunneling electrons. Due to the sensitivity of tunneling currents to the gap size, the tunneling current can reveal the changes in (CH3S)2 in the gap, as illustrated in Fig. 13(d). As shown in Fig. 13(e), the tunneling current with a low bias voltage, where the tunneling electrons with low energy cannot induce any reactions, shows the dissociation of (CH3S)2 when the incident light exciting the plasmon leads to this chemical reaction. When the bias voltage increases, the energy of tunneling electrons is high enough to induce rotation and even dissociation reactions without light excitation. In this case, these processes can also be detected by the tunneling current in real time, as shown in Fig. 13(f). In summary, plasmonic tunneling gaps offer a promising functionality to control and observe chemical reactions on the nanoscale.

In this perspective article, we introduce the recent research and developments of plasmonic tunneling junctions, encompassing three main sections. First, we discussed the light emission via inelastic tunneling electrons in plasmonic tunneling gaps, which is an attractive photon source on the nanoscale. Due to the ultrafast process of tunneling, this light source can provide femtosecond response times. Various parameters of the tunneling junctions, including geometries, plasmonic properties, electronic properties, and materials, can be modified to make the emitted light satisfy various experimental requirements. Furthermore, above-threshold light emission can also be realized in plasmonic tunneling junctions, providing a wider spectral range of emitted photons. Second, we focused on the light-induced currents in plasmonic tunneling junctions, which show how the light–matter interactions in the nanogaps affect the tunneling phenomena, including optical rectification, thermal effects, and photoemission. These effects can be employed as detectors to reveal the physical properties of materials and processes within nanogaps. Finally, nanoscale reactors and sensors based on plasmonic tunneling junctions were introduced. These can serve as promising platforms for chemical applications. Additionally, many related attractive phenomena and applications are not mentioned in this paper, such as the nonlinear light generation in tunneling gaps,82 energy harvesting based on optical rectifications,83 and tunneling switches.84 In conclusion, all of these interesting phenomena and unique characteristics make plasmonic tunneling junctions a promising field to study light–matter interaction on the nanoscale and to realize potential applications in the future.

This work was supported by the National Science Foundation (Grant No. CBET–2231857).

The authors have no conflicts to disclose.

Yuankai Tang: Conceptualization (equal); Writing – original draft (lead). Hayk Harutyunyan: Conceptualization (equal); Writing – review & editing (lead).

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

1.
D. J.
Griffiths
and
D. F.
Schroeter
,
Introduction to Quantum Mechanics
(
Cambridge University Press
,
2018
).
2.
J. G.
Simmons
, “
Generalized formula for the electric tunnel effect between similar electrodes separated by a thin insulating film
,”
J. Appl. Phys.
34
,
1793
1803
(
1963
).
3.
J.
Kern
,
R.
Kullock
,
J.
Prangsma
,
M.
Emmerling
,
M.
Kamp
, and
B.
Hecht
, “
Electrically driven optical antennas
,”
Nat. Photonics
9
,
582
586
(
2015
).
4.
J. J.
Baumberg
,
J.
Aizpurua
,
M. H.
Mikkelsen
, and
D. R.
Smith
, “
Extreme nanophotonics from ultrathin metallic gaps
,”
Nat. Mater.
18
,
668
678
(
2019
).
5.
R.
Chikkaraddy
,
B.
De Nijs
,
F.
Benz
,
S. J.
Barrow
,
O. A.
Scherman
,
E.
Rosta
,
A.
Demetriadou
,
P.
Fox
,
O.
Hess
, and
J. J.
Baumberg
, “
Single-molecule strong coupling at room temperature in plasmonic nanocavities
,”
Nature
535
,
127
130
(
2016
).
6.
K.
Santhosh
,
O.
Bitton
,
L.
Chuntonov
, and
G.
Haran
, “
Vacuum Rabi splitting in a plasmonic cavity at the single quantum emitter limit
,”
Nat. Commun.
7
,
11823
(
2016
).
7.
W.
Zhu
and
K. B.
Crozier
, “
Quantum mechanical limit to plasmonic enhancement as observed by surface-enhanced Raman scattering
,”
Nat. Commun.
5
,
5228
(
2014
).
8.
S.-Y.
Ding
,
J.
Yi
,
J.-F.
Li
,
B.
Ren
,
D.-Y.
Wu
,
R.
Panneerselvam
, and
Z.-Q.
Tian
, “
Nanostructure-based plasmon-enhanced Raman spectroscopy for surface analysis of materials
,”
Nat. Rev. Mater.
1
,
16021
(
2016
).
9.
D. O.
Sigle
,
S.
Kasera
,
L. O.
Herrmann
,
A.
Palma
,
B.
De Nijs
,
F.
Benz
,
S.
Mahajan
,
J. J.
Baumberg
, and
O. A.
Scherman
, “
Observing single molecules complexing with cucurbit[7]uril through nanogap surface-enhanced Raman spectroscopy
,”
J. Phys. Chem. Lett.
7
,
704
710
(
2016
).
10.
Y.
Zhang
,
W.
Chen
,
T.
Fu
,
J.
Sun
,
D.
Zhang
,
Y.
Li
,
S.
Zhang
, and
H.
Xu
, “
Simultaneous surface-enhanced resonant Raman and fluorescence spectroscopy of monolayer MoSe2: Determination of ultrafast decay rates in nanometer dimension
,”
Nano Lett.
19
,
6284
6291
(
2019
).
11.
J. B.
Lassiter
,
X.
Chen
,
X.
Liu
,
C.
Ciracì
,
T. B.
Hoang
,
S.
Larouche
,
S.-H.
Oh
,
M. H.
Mikkelsen
, and
D. R.
Smith
, “
Third-harmonic generation enhancement by film-coupled plasmonic stripe resonators
,”
ACS Photonics
1
,
1212
1217
(
2014
).
12.
F.
Wang
,
A. B. F.
Martinson
, and
H.
Harutyunyan
, “
Efficient nonlinear metasurface based on nonplanar plasmonic nanocavities
,”
ACS Photonics
4
,
1188
1194
(
2017
).
13.
J.
Lambe
and
S. L.
Mccarthy
, “
Light emission from inelastic electron tunneling
,”
Phys. Rev. Lett.
37
,
923
925
(
1976
).
14.
P.
Bharadwaj
,
A.
Bouhelier
, and
L.
Novotny
, “
Electrical excitation of surface plasmons
,”
Phys. Rev. Lett.
106
,
226802
(
2011
).
15.
W.
Du
,
T.
Wang
,
H.-S.
Chu
, and
C. A.
Nijhuis
, “
Highly efficient on-chip direct electronic–plasmonic transducers
,”
Nat. Photonics
11
,
623
627
(
2017
).
16.
H.
Qian
,
S.-W.
Hsu
,
K.
Gurunatha
,
C. T.
Riley
,
J.
Zhao
,
D.
Lu
,
A. R.
Tao
, and
Z.
Liu
, “
Efficient light generation from enhanced inelastic electron tunnelling
,”
Nat. Photonics
12
,
485
488
(
2018
).
17.
J.
Qin
,
Y.
Liu
,
H.
Luo
,
Z.
Jiang
,
W.
Cai
, and
L.
Wang
, “
Tunable light emission by electrically excited plasmonic antenna
,”
ACS Photonics
6
,
2392
2396
(
2019
).
18.
R.
Kullock
,
M.
Ochs
,
P.
Grimm
,
M.
Emmerling
, and
B.
Hecht
, “
Electrically-driven Yagi-Uda antennas for light
,”
Nat. Commun.
11
,
115
(
2020
).
19.
F.
Wang
,
Y.
Liu
,
T. J.
Duffin
,
V.
Kalathingal
,
S.
Gao
,
W.
Hu
,
Y.
Guo
,
S.-J.
Chua
, and
C. A.
Nijhuis
, “
Silicon-based quantum mechanical tunnel junction for plasmon excitation from low-energy electron tunneling
,”
ACS Photonics
8
,
1951
1960
(
2021
).
20.
G.
Shalem
,
O.
Erez-Cohen
,
D.
Mahalu
, and
I.
Bar-Joseph
, “
Light emission in metal–semiconductor tunnel junctions: Direct evidence for electron heating by plasmon decay
,”
Nano Lett.
21
,
1282
1287
(
2021
).
21.
M.
Parzefall
,
P.
Bharadwaj
,
A.
Jain
,
T.
Taniguchi
,
K.
Watanabe
, and
L.
Novotny
, “
Antenna-coupled photon emission from hexagonal boron nitride tunnel junctions
,”
Nat. Nanotechnol.
10
,
1058
1063
(
2015
).
22.
S.
Namgung
,
D. A.
Mohr
,
D.
Yoo
,
P.
Bharadwaj
,
S. J.
Koester
, and
S.-H.
Oh
, “
Ultrasmall plasmonic single nanoparticle light source driven by a graphene tunnel junction
,”
ACS Nano
12
,
2780
2788
(
2018
).
23.
M.
Parzefall
,
Á.
Szabó
,
T.
Taniguchi
,
K.
Watanabe
,
M.
Luisier
, and
L.
Novotny
, “
Light from van der Waals quantum tunneling devices
,”
Nat. Commun.
10
,
292
(
2019
).
24.
Z.
Wang
,
V.
Kalathingal
,
T. X.
Hoang
,
H.-S.
Chu
, and
C. A.
Nijhuis
, “
Optical anisotropy in van der Waals materials: Impact on direct excitation of plasmons and photons by quantum tunneling
,”
Light: Sci. Appl.
10
,
230
(
2021
).
25.
X.
He
,
J.
Tang
,
H.
Hu
,
J.
Shi
,
Z.
Guan
,
S.
Zhang
, and
H.
Xu
, “
Electrically driven highly tunable cavity plasmons
,”
ACS Photonics
6
,
823
829
(
2019
).
26.
W.
Du
,
T.
Wang
,
H.-S.
Chu
,
L.
Wu
,
R.
Liu
,
S.
Sun
,
W. K.
Phua
,
L.
Wang
,
N.
Tomczak
, and
C. A.
Nijhuis
, “
On-chip molecular electronic plasmon sources based on self-assembled monolayer tunnel junctions
,”
Nat. Photonics
10
,
274
280
(
2016
).
27.
W.
Zhu
,
R.
Esteban
,
A. G.
Borisov
,
J. J.
Baumberg
,
P.
Nordlander
,
H. J.
Lezec
,
J.
Aizpurua
, and
K. B.
Crozier
, “
Quantum mechanical effects in plasmonic structures with subnanometre gaps
,”
Nat. Commun.
7
,
11495
(
2016
).
28.
G.
Hajisalem
,
M. S.
Nezami
, and
R.
Gordon
, “
Probing the quantum tunneling limit of plasmonic enhancement by third harmonic generation
,”
Nano Lett.
14
,
6651
6654
(
2014
).
29.
A. L.
Paoletta
,
E.-D.
Fung
, and
L.
Venkataraman
, “
Gap size-dependent plasmonic enhancement in electroluminescent tunnel junctions
,”
ACS Photonics
9
,
688
693
(
2022
).
30.
A.
Radulescu
,
V.
Kalathingal
,
Z.
Wang
, and
C. A.
Nijhuis
, “
Coherence between different propagating surface plasmon polariton modes excited by quantum mechanical tunnel junctions
,”
Adv. Opt. Mater.
10
,
2101804
(
2022
).
31.
A.
Radulescu
,
V.
Kalathingal
, and
C. A.
Nijhuis
, “
Geometric control over the edge diffraction of electrically excited surface plasmon polaritons by tunnel junctions
,”
ACS Photonics
8
,
3591
3598
(
2021
).
32.
Y.
Lin
,
T. X.
Hoang
,
H.-S.
Chu
, and
C. A.
Nijhuis
, “
Directional launching of surface plasmon polaritons by electrically driven aperiodic groove array reflectors
,”
Nanophotonics
10
,
1145
1154
(
2021
).
33.
F.
Wang
,
T. X.
Hoang
,
H. S.
Chu
, and
C. A.
Nijhuis
, “
Spatial control over stable light‐emission from AC‐driven CMOS‐compatible quantum mechanical tunnel junctions
,”
Laser Photonics Rev.
16
,
2100419
(
2022
).
34.
K. S.
Makarenko
,
T. X.
Hoang
,
T. J.
Duffin
,
A.
Radulescu
,
V.
Kalathingal
,
H. J.
Lezec
,
H. S.
Chu
, and
C. A.
Nijhuis
, “
Efficient surface plasmon polariton excitation and control over outcoupling mechanisms in metal–insulator–metal tunneling junctions
,”
Adv. Sci.
7
,
1900291
(
2020
).
35.
H.
Qian
,
S.
Li
,
S.-W.
Hsu
,
C.-F.
Chen
,
F.
Tian
,
A. R.
Tao
, and
Z.
Liu
, “
Highly-efficient electrically-driven localized surface plasmon source enabled by resonant inelastic electron tunneling
,”
Nat. Commun.
12
,
3111
(
2021
).
36.
A. V.
Uskov
,
J. B.
Khurgin
,
I. E.
Protsenko
,
I. V.
Smetanin
, and
A.
Bouhelier
, “
Excitation of plasmonic nanoantennas by nonresonant and resonant electron tunnelling
,”
Nanoscale
8
,
14573
14579
(
2016
).
37.
D. R.
Ward
,
F.
Hüser
,
F.
Pauly
,
J. C.
Cuevas
, and
D.
Natelson
, “
Optical rectification and field enhancement in a plasmonic nanogap
,”
Nat. Nanotechnol.
5
,
732
736
(
2010
).
38.
R.
Arielly
,
A.
Ofarim
,
G.
Noy
, and
Y.
Selzer
, “
Accurate determination of plasmonic fields in molecular junctions by current rectification at optical frequencies
,”
Nano Lett.
11
,
2968
2972
(
2011
).
39.
A.
Stolz
,
J.
Berthelot
,
M.-M.
Mennemanteuil
,
G.
Colas Des Francs
,
L.
Markey
,
V.
Meunier
, and
A.
Bouhelier
, “
Nonlinear photon-assisted tunneling transport in optical gap antennas
,”
Nano Lett.
14
,
2330
2338
(
2014
).
40.
P.
Zimmermann
,
A.
Hötger
,
N.
Fernandez
,
A.
Nolinder
,
K.
Müller
,
J. J.
Finley
, and
A. W.
Holleitner
, “
Toward plasmonic tunnel gaps for nanoscale photoemission currents by on-chip laser ablation
,”
Nano Lett.
19
,
1172
1178
(
2019
).
41.
D.
Kos
,
D. R.
Assumpcao
,
C.
Guo
, and
J. J.
Baumberg
, “
Quantum tunneling induced optical rectification and plasmon-enhanced photocurrent in nanocavity molecular junctions
,”
ACS Nano
15
,
14535
14543
(
2021
).
42.
Q.
Zou
,
X.
Chen
,
Y.
Zhou
,
X.
Jin
,
Z.
Zhang
,
J.
Qiu
,
R.
Wang
,
W.
Hong
,
J.
Su
,
D.-H.
Qu
, and
H.
Tian
, “
Photoconductance from the bent-to-planar photocycle between ground and excited states in single-molecule junctions
,”
J. Am. Chem. Soc.
144
,
10042
10052
(
2022
).
43.
Y.
Selzer
, “
Elucidating the contributions of plasmon-induced excitons and hot carriers to the photocurrent of molecular junctions
,”
J. Phys. Chem. C
124
,
8680
8688
(
2020
).
44.
H.
Bi
,
C.
Jing
,
P.
Hasch
,
Y.
Gong
,
D.
Gerster
,
J. V.
Barth
, and
J.
Reichert
, “
Single molecules in strong optical fields: A variable-temperature molecular junction spectroscopy setup
,”
Anal. Chem.
93
,
9853
9859
(
2021
).
45.
N.
Nachman
and
Y.
Selzer
, “
Thermometry of plasmonic heating by inelastic electron tunneling spectroscopy (IETS)
,”
Nano Lett.
17
,
5855
5861
(
2017
).
46.
E.
Kazuma
,
J.
Jung
,
H.
Ueba
,
M.
Trenary
, and
Y.
Kim
, “
Real-space and real-time observation of a plasmon-induced chemical reaction of a single molecule
,”
Science
360
,
521
525
(
2018
).
47.
P.
Wang
,
A. V.
Krasavin
,
M. E.
Nasir
,
W.
Dickson
, and
A. V.
Zayats
, “
Reactive tunnel junctions in electrically driven plasmonic nanorod metamaterials
,”
Nat. Nanotechnol.
13
,
159
164
(
2018
).
48.
B.
Han
,
Y.
Li
,
X.
Ji
,
X.
Song
,
S.
Ding
,
B.
Li
,
H.
Khalid
,
Y.
Zhang
,
X.
Xu
,
L.
Tian
,
H.
Dong
,
X.
Yu
, and
W.
Hu
, “
Systematic modulation of charge transport in molecular devices through facile control of molecule–electrode coupling using a double self-assembled monolayer nanowire junction
,”
J. Am. Chem. Soc.
142
(
2020
).
49.
R.
Gupta
,
P.
Jash
,
P.
Sachan
,
A.
Bayat
,
V.
Singh
, and
P. C.
Mondal
, “
Electrochemical potential‐driven high‐throughput molecular electronic and spintronic devices: From molecules to applications
,”
Angew. Chem., Int. Ed.
60
,
26904
26921
(
2021
).
50.
X.
Chen
and
C. A.
Nijhuis
, “
The unusual dielectric response of large area molecular tunnel junctions probed with impedance spectroscopy
,”
Adv. Electron. Mater.
8
,
2100495
(
2022
).
51.
P.
Gehring
,
J. M.
Thijssen
, and
H. S. J.
Van Der Zant
, “
Single-molecule quantum-transport phenomena in break junctions
,”
Nat. Rev. Phys.
1
,
381
396
(
2019
).
52.
L.
Xue
,
H.
Yamazaki
,
R.
Ren
,
M.
Wanunu
,
A. P.
Ivanov
, and
J. B.
Edel
, “
Solid-state nanopore sensors
,”
Nat. Rev. Mater.
5
,
931
951
(
2020
).
53.
L.
Novotny
and
B.
Hecht
,
Principles of Nano-Optics
(
Cambridge University Press
,
2012
).
54.
M.
Parzefall
and
L.
Novotny
, “
Light at the end of the tunnel
,”
ACS Photonics
5
,
4195
4202
(
2018
).
55.
M.
Parzefall
and
L.
Novotny
, “
Optical antennas driven by quantum tunneling: A key issues review
,”
Rep. Prog. Phys.
82
,
112401
(
2019
).
56.
S. A.
Maier
,
Plasmonics: Fundamentals and Applications
(
Springer Science & Business Media
,
2007
).
57.
T.
Wang
,
E.
Boer-Duchemin
,
Y.
Zhang
,
G.
Comtet
, and
G.
Dujardin
, “
Excitation of propagating surface plasmons with a scanning tunnelling microscope
,”
Nanotechnology
22
,
175201
(
2011
).
58.
C.
Ciracì
,
F.
Vidal-Codina
,
D.
Yoo
,
J.
Peraire
,
S.-H.
Oh
, and
D. R.
Smith
, “
Impact of surface roughness in nanogap plasmonic systems
,”
ACS Photonics
7
,
908
913
(
2020
).
59.
C.
Zhang
,
J. P.
Hugonin
,
A. L.
Coutrot
,
C.
Sauvan
,
F.
Marquier
, and
J. J.
Greffet
, “
Antenna surface plasmon emission by inelastic tunneling
,”
Nat. Commun.
10
,
4949
(
2019
).
60.
L.
Cui
,
Y.
Zhu
,
M.
Abbasi
,
A.
Ahmadivand
,
B.
Gerislioglu
,
P.
Nordlander
, and
D.
Natelson
, “
Electrically driven hot-carrier generation and above-threshold light emission in plasmonic tunnel junctions
,”
Nano Lett.
20
,
6067
6075
(
2020
).
61.
L.
Cui
,
Y.
Zhu
,
P.
Nordlander
,
M.
Di Ventra
, and
D.
Natelson
, “
Thousand-fold increase in plasmonic light emission via combined electronic and optical excitations
,”
Nano Lett.
21
,
2658
2665
(
2021
).
62.
M.
Buret
,
A. V.
Uskov
,
J.
Dellinger
,
N.
Cazier
,
M.-M.
Mennemanteuil
,
J.
Berthelot
,
I. V.
Smetanin
,
I. E.
Protsenko
,
G.
Colas-Des-Francs
, and
A.
Bouhelier
, “
Spontaneous hot-electron light emission from electron-fed optical antennas
,”
Nano Lett.
15
,
5811
5818
(
2015
).
63.
H.
Park
,
A. K. L.
Lim
,
A. P.
Alivisatos
,
J.
Park
, and
P. L.
Mceuen
, “
Fabrication of metallic electrodes with nanometer separation by electromigration
,”
Appl. Phys. Lett.
75
,
301
303
(
1999
).
64.
J. B.
Herzog
,
M. W.
Knight
,
Y.
Li
,
K. M.
Evans
,
N. J.
Halas
, and
D.
Natelson
, “
Dark plasmons in hot spot generation and polarization in interelectrode nanoscale junctions
,”
Nano Lett.
13
,
1359
1364
(
2013
).
65.
G.
Platero
and
R.
Aguado
, “
Photon-assisted transport in semiconductor nanostructures
,”
Phys. Rep.
395
,
1
157
(
2004
).
66.
S.
Kohler
,
J.
Lehmann
, and
P.
Hanggi
, “
Driven quantum transport on the nanoscale
,”
Phys. Rep.
406
,
379
443
(
2005
).
67.
J.
Tucker
, “
Quantum limited detection in tunnel junction mixers
,”
IEEE J. Quantum Electron.
15
,
1234
1258
(
1979
).
68.
M. M. A.
Yajadda
,
K.-H.
Müller
,
D. I.
Farrant
, and
K.
Ostrikov
, “
Partial rectification of the plasmon-induced electrical tunnel current in discontinuous thin gold film at optical frequency
,”
Appl. Phys. Lett.
100
,
211105
(
2012
).
69.
A. V.
Bragas
,
S. M.
Landi
, and
O. E.
Martínez
, “
Laser field enhancement at the scanning tunneling microscope junction measured by optical rectification
,”
Appl. Phys. Lett.
72
,
2075
2077
(
1998
).
70.
M. M.
Mennemanteuil
,
M.
Buret
,
G.
Colas-Des-Francs
, and
A.
Bouhelier
, “
Optical rectification and thermal currents in optical tunneling gap antennas
,”
Nanophotonics
11
,
4197
4208
(
2022
).
71.
X.
Chen
,
M.
Roemer
,
L.
Yuan
,
W.
Du
,
D.
Thompson
,
E.
Del Barco
, and
C. A.
Nijhuis
, “
Molecular diodes with rectification ratios exceeding 105 driven by electrostatic interactions
,”
Nat. Nanotechnol.
12
,
797
803
(
2017
).
72.
D.
Thompson
and
C. A.
Nijhuis
, “
Even the odd numbers help: Failure modes of SAM-based tunnel junctions probed via odd-even effects revealed in synchrotrons and supercomputers
,”
Acc. Chem. Res.
49
,
2061
2069
(
2016
).
73.
C. A.
Nijhuis
,
W. F.
Reus
,
J. R.
Barber
,
M. D.
Dickey
, and
G. M.
Whitesides
, “
Charge transport and rectification in arrays of SAM-based tunneling junctions
,”
Nano Lett.
10
,
3611
3619
(
2010
).
74.
Q.
Van Nguyen
, “
Controlling rectification in metal–molecules–metal junctions based on 11-(ferrocenyl) undecanethiol: Effects of the electronic coupling strength
,”
J. Phys. Chem. C
126
,
6405
6412
(
2022
).
75.
A.
Lombardi
,
A.
Demetriadou
,
L.
Weller
,
P.
Andrae
,
F.
Benz
,
R.
Chikkaraddy
,
J.
Aizpurua
, and
J. J.
Baumberg
, “
Anomalous spectral shift of near- and far-field plasmonic resonances in nanogaps
,”
ACS Photonics
3
,
471
477
(
2016
).
76.
M.
Yang
,
M. S.
Mattei
,
C. R.
Cherqui
,
X.
Chen
,
R. P.
Van Duyne
, and
G. C.
Schatz
, “
Tip-enhanced Raman excitation spectroscopy (TERES): Direct spectral characterization of the gap-mode plasmon
,”
Nano Lett.
19
,
7309
7316
(
2019
).
77.
M. M. A.
Yajadda
,
K.-H.
Müller
, and
K.
Ostrikov
, “
Effect of Coulomb blockade, gold resistance, and thermal expansion on the electrical resistance of ultrathin gold films
,”
Phys. Rev. B
84
,
235431
(
2011
).
78.
P.
Zolotavin
,
C.
Evans
, and
D.
Natelson
, “
Photothermoelectric effects and large photovoltages in plasmonic Au nanowires with nanogaps
,”
J. Phys. Chem. Lett.
8
,
1739
1744
(
2017
).
79.
M.-M.
Mennemanteuil
,
G.
Colas-Des-Francs
,
M.
Buret
,
A.
Dasgupta
,
A.
Cuadrado
,
J.
Alda
, and
A.
Bouhelier
, “
Laser-induced thermoelectric effects in electrically biased nanoscale constrictions
,”
Nanophotonics
7
,
1917
1927
(
2018
).
80.
S.
Zhou
,
X.
Guo
,
K.
Chen
,
M. T.
Cole
,
X.
Wang
,
Z.
Li
,
J.
Dai
,
C.
Li
, and
Q.
Dai
, “
Optical‐field‐driven electron tunneling in metal–insulator–metal nanojunction
,”
Adv. Sci.
8
,
2101572
(
2021
).
81.
H.
Chalabi
,
D.
Schoen
, and
M. L.
Brongersma
, “
Hot-electron photodetection with a plasmonic nanostripe antenna
,”
Nano Lett.
14
,
1374
1380
(
2014
).
82.
G.
Aguirregabiria
,
D. C.
Marinica
,
R.
Esteban
,
A. K.
Kazansky
,
J.
Aizpurua
, and
A. G.
Borisov
, “
Role of electron tunneling in the nonlinear response of plasmonic nanogaps
,”
Phys. Rev. B
97
,
115430
(
2018
).
83.
C. A.
Reynaud
,
D.
Duché
,
J.-J.
Simon
,
E.
Sanchez-Adaime
,
O.
Margeat
,
J.
Ackermann
,
V.
Jangid
,
C.
Lebouin
,
D.
Brunel
,
F.
Dumur
,
D.
Gigmes
,
G.
Berginc
,
C. A.
Nijhuis
, and
L.
Escoubas
, “
Rectifying antennas for energy harvesting from the microwaves to visible light: A review
,”
Prog. Quantum Electron.
72
,
100265
(
2020
).
84.
F.
Niroui
,
A. I.
Wang
,
E. M.
Sletten
,
Y.
Song
,
J.
Kong
,
E.
Yablonovitch
,
T. M.
Swager
,
J. H.
Lang
, and
V.
Bulović
, “
Tunneling nanoelectromechanical switches based on compressible molecular thin films
,”
ACS Nano
9
,
7886
7894
(
2015
).