Open Access
Add to BibSonomyAbstract
We introduce a new type of electroplasmonic interfacing component to electrically generate surface plasmons. Specifically, an electron-fed optical tunneling gap antenna is integrated on a plasmonic waveguiding platform. When electrical charges are injected in the tunneling barrier of the gap antenna, a broad-band radiation is emitted from the feed area by a process identified as a thermal emission of hot electrons. Part of the emitted photons couples to surface plasmon modes sustained by the waveguide geometry. The transducing optical antenna is thus acting as a localized electrical source of surface plasmon polaritons. The integration of electrically–activated optical antennas into a plasmonic architecture mitigates the need for complex coupling scheme and proposes a solution for realizing nanoscale units at the interface between nano-electronics and photonics.
© 2016 Optical Society of America
1. Introduction
The large size difference between electronic and photonic components drastically hinders the optimal co-integration of these information processing technologies on the same chip. Photonic devices have typical dimensions comparable to or larger than the operating wavelength, while electronic units have sizes approaching the 10 nm node. In the last decade, plasmonics emerged as a way to design integrated optical components at subwavelength scales [1,2]. Metal-based plasmonic devices offer a unique technological asset enabling the simultaneous transport of electrical and electromagnetic signals on the same physical support, facilitating thus the merging of photon-management functionalities with electronic circuits [3, 4]. Undoubtedly, a hallmark component marking this interface is the integration of an electrical source of surface plasmon polaritons (SPPs). Excitation of surface plasmons was confirmed with electroluminescent semiconductors. Light emitted from nano-crystals embedded in metal-insulator-metal membranes [5] and single carbon nanotube transistors [6] may couple to surface plasmon modes existing near the active site. In these devices, photon emission results from an impact ionization, which typically involves large static biases (> 10 V) and low external quantum efficiency (10−6). Higher conversion yields and better surface plasmon coupling efficiencies are found from light-emitting diodes [7–10]. However, the technology requires complex fabrication steps involving molecular-beam epitaxy and focused-ion beam milling with device footprints exceeding several μm2. In this context, contacted GaAs nanowires provide an elegant route to realize more compact electrically-driven surface plasmon generators [11].
In a 1957 prediction, Ritchie suggested that accelerated electrons impinging on a thin metal foil lose energy via the excitation of surface plasmons [12]. High-energy electrons have since been used to drive modern plasmonic structures [13–15]. Along the same vein, electrons tunneling through a potential barrier, such as the tip of a scanning tunneling microscope, have a probability to inelastically scatter their energies by releasing photons [16]. These photons couple to underlying localized surface plasmon modes [17, 18], but may also launch propagating surface plasmon modes when available [19–22]. Co-planar light-emitting devices based on inelastic electron tunneling are now emerging to tackle the challenge of integration [23–25], but no attempt has been made to couple the light to an on-chip plateform.
In this paper, we present a novel integrated electrical source of SPPs formed by interfacing an electrically-contacted tunneling gap nanoantenna with a surface plasmon waveguide. Compared to alternative routes, a tunneling antenna provides an ultracompact photon source (<100 nm2) with low activation voltage that is interfaced to a surface plasmon waveguide. The plasmon waveguides used in this work are dielectric-loaded surface plasmon structures (DL-SPPW). They consist of subwavelength dielectric ridges deposited on a metal film. DLSPPW are generally used to increase the lateral confinement of the plasmon mode [26] and to provide active plasmonic plateforms through an external modification of the dielectric properties of the loading material [27, 28].
When electrons are injected in the feedgap of a tunneling antenna, photons may be emitted through two different processes. For relatively large in-plane tunneling gaps, emission is observed by inelastically transported electrons through the barrier [23, 24]. In this radiation process the energy of the photons hν relates to the applied static bias Vbias through the relation hν ≤ eVbias, where h is Planck’s constant, e is the electron’s charge and ν the frequency of the light. The applied bias is thus dictacting the highest photon energy emitted by the junction.
A second emission mechanism, violating the quantum limit, is at play for tunneling antennas characterized by large zero-bias conductance and sustaining a large driving current. A portion of the electrical power may be transferred to the electron subsystem. This energy exchange creates an out-of-equilibrium electronic distribution that may spontaneously decay into photons [29]. The emitted spectrum is no longer limited by eVbias but by the electronic temperature of the carriers. We show in the following that the emission of such electron-fed antennas can excite waveguided plasmonic modes, turning the light-emitting antenna into an integrated electrical source of surface plasmons.
2. Fabrication details
The complete platform is realized using a multistep lithography process followed by a controlled formation of a tunneling gap. Dielectric-loaded surface plasmon polariton waveguides are produced by standard electron-beam lithography and lift-off processes. The waveguides consist of a thermally evaporated 3 nm Cr adhesion layer and a ∼60 nm thick Au strip. The strip is 10 μm long and is either 0.8 μm or 2 μm wide to mitigate the role of the edges [30]. The DLSPPW is finalized in a second fabrication step by depositing a 500 nm wide and 60 nm thick SiOx ridge centered on the Au strip. The width of the dielectric layer is chosen to insure single mode operation for the wavelength range considered [31]. We choose SiOx over standard polymeric materials [32] because of its robustness to the subsequent fabrication steps. The composition of the SiOx is determined by ellipsometry to be two thirds SiO2 and one third of SiO, for an optical index of n=1.6 for wavelengths around 800 nm. The structure of the optical antenna together with the microscopic contacts are realized in a third electron-beam lithography step. The antenna itself is formed from a gold nanowire with a length of 1.1 μm and a width of 160 nm. The nanowire is connected on both ends to a set of electrodes. The thickness of the nanowire and of the electrodes is 80 nm, including a 5 nm Cr adhesion layer. The Au and SiOx areas defining the DLSPPW are extended under the contacting electrodes to insure the electrical continuity of the topmost Au layer. A final step of ultraviolet optical lithography is then used to define a series of large Au leads connecting to the microscopic electrodes and the nanowire.
Our strategy to create in-situ a tunneling optical gap antenna relies on the controlled electro-migration of the nanowire [29] and proceeds as follow: A voltage V = Vbias + Vmod cos(2πFt) is applied to the electrodes, with Vmod=20 mV and F=10 kHz. The differential conductance dG of the nanowire is constantly monitored by a lock-in amplifier (Zürich Instruments) synced to F. Vbias is monotonously increased (up to a few Volts) until dG starts to drop, which marks the start of the electromigration. Vbias is then slowly decreased during the electromigration to avoid a catastrophic rupture of the nanowire. The electromigration stops when the conductance of the nanowire falls below the quantum conductance G0 = 2e2/h indicating the formation of a gap in the nanowire [29]. Colorized scanning electron micrographs of complete devices are illustrated in Fig. 1 for the two kind of Au strips fabricated. The yellow areas are the different gold layers defining the DLSPPW, the nanowire and the contacting source and drain electrodes. The blue areas indicate the SiOx layer. The glass substrate is left in gray tone. A close-up view of the nanowire after the electromigration process points the location of the tunneling section forming the feedgap of the optical gap antenna. Note that the images were taken after servicing the devices and the actual gap sizes do not reflect the operating conditions discussed below. The freshly formed junction has the electrical output characteristics of a tunnel barrier, as displayed in Fig. 1(b). With this manual control of the electromigration, we obtain tunneling gaps sustaining large currents for moderate static biases at room temperature. The junctions are stable during measurements for applied voltage |Vbias| ≤ 1 V. Higher voltages irreversibly degrade the junction resulting in a sharp drop of the tunnel current. In general, the gap on the nanowire forms near the source of electrons. The precise location is however determined by the grain boundaries inside the polycrystalline gold nanowire [33] and the gap may therefore appear at any points along the nanowire [34].
Fig. 1 (a) Scanning electron micrograph of an electromigrated gold nanowire with its connected electrodes fabricated on top of a dielectric-loaded surface plasmon waveguide (DL-SPPW). The DLSPPW is constituted of an underlying 0.8 μm wide Au strip coated with a 500 nm wide SiOx waveguide. The gold is colored in yellow and the SiOx in blue (the glass substrate is kept in gray). When electrically activated, the electromigrated tunnel junction acts as an electron-fed optical gap antenna. (b) Output characteristics of the tunneling gap antenna measured just after electromigration. The size of the gap after electromigration is initially commensurate to sustain a large tunnel current at room temperature. Under large bias polarity, the gap eventually degrades after continuous operation as shown in the inset of (a). (c) and (d) are the electron micrograph and corresponding output characteristics for a 2 μm wide DLSPPW base, respectively. The insert in (c) represents a cross-sectional description of the different layers of the device.
3. Coupling of the antenna emission into surface plasmon modes
We demonstrate in this section the electrical excitation of the surface plasmon mode in a DL-SPPW realized on a 2 μm-wide Au strip. The colorized image of the device is reproduced in Fig. 2(a) for comparison purposes. Light emitted by the tunnel junction and signals indicative of surface plasmon excitation are collected with an inverted optical microscope (Nikon, Eclipse) equipped with a 100×, 1.49 numerical aperture (N.A.) objective and two Charged Coupled Device (CCD) cameras. A first CCD (Andor, Luca EM S 658M) records the object plane of the microscope. The second camera (Andor, iKon-M 934) measures a plane conjugate to the Fourier plane. The object plane provides information about the two-dimensional distribution of the light in the device, while the Fourier plane gives information about the angular distribution of the light emitted into the glass substrate [35]. The largest projected angle experimentally measured is limited by the N.A. of the objective, here 1.49.
Fig. 2 (a) SEM image of the entire DLSPPW placed on a 2 μm-wide Au strip. The connected electrodes and antenna feedgap are readily observed on the right hand side of the device. (b) Distribution of the light in the device when the antenna is polarized with a bias of 1 V and a driving current of 50 μA. Emission at the feedgap and various surface plasmon signatures are detected including edge and DLSPPW end-face scattering. This picture was taken using an electron multiplier gain of 255 for the CCD camera. (c) Angular distribution of the radiation emitted in the glass substrate. The essential of the emission is concentrated in the signature of the plasmon mode. kx/k0 and ky/k0 are the projected wavevectors along the x and y directions respectively, with ki/k0=n sinθ ≤N.A, where i = x, y, n is the refractive index of the substrate, θ is the angle with respect to the normal of the interface, and N.A. is the numerical aperture of the objective.
The optical antenna is electrically powered by applying a bias of 1 V during 60 s. This time does not correspond to the lifetime of the device, although some devices may burn shortly after applying the bias. 60 s is a conservative period to maintain the device in constant operation. Increasing the voltage to larger values may lead to a degradation of the tunnel characteristics. In general, junctions are powered for no more than a few minutes during measurements. Particularly stable junctions on glass substrate have however been operated for several hours.
Figure 2(b) shows a false-color image of the light distributed in the device. We observe a strong luminous response emitted by the electron-fed antenna. Under this operation condition (i.e. large current, moderate bias), emission originates from the radiation of a hot electron distribution created near the antenna feedgap [29]. The dynamic range of the color scale is adjusted to reveal remote contributions. The junction area is thus appearing as a saturated intensity region. Light is however emitted from a point-like source and is convoluted by the point-spread function of the microscope. The concentric fringe pattern near the source of radiation results from this convolution. We have not estimated the efficiency of the electron-to-photon conversion. The number of electrons required to produce a photon in the present experiment is hidden in different parameters imposed by the geometry like the finite transmission through the multilayer system and the coupling yield of the source to the various modes discussed below. However, for electromigrated junctions fabricated on a glass substrate, thermal emission has external quantum yields ranging from device to device between 10−11 to 10−7 for light covering the visible spectrum [29]. This estimation of the yield does not take into account the dominant energy emitted in the infrared part of the spectrum where silicon detectors are blind. For comparison purposes, the external conversion yield matches up other conversion mechanisms at play in nanoscale light sources, like individual quantum dots. Indeed, with a single quantum dot, extracting 103 photons/s requires a few tens of nW of optical power. The external yield is thus approximately 10−8 to 10−6 depending on the configuration [36]. Even for advanced structures like electrically pumped quantum dots laser operating at cryogenic temperature, the wall-plug efficiency remains low at 10−6 [37]. Furthermore, the efficiency of inelastic electron tunneling is comparatively several orders of magnitude larger than the spontaneous emission discussed here [23, 24]. Yet, the predominant emission process of the source in Fig. 2(b) is the one occurring with the lowest yield. The detected spectral range (energy higher than the energy provided by the bias voltage) is discriminating inelastic electron tunneling to favor the thermal nature of the emission.
Despite the lower electron-to-photon conversion efficiency, operating tunnel junctions in a thermal emission regime provides additional benefits. With electromigrated junctions working in ambient conditions, the flow of tunneling electrons is affected by the presence of impurities and the mobility of the Au atoms. Large current bursts are thus frequently observed for Au junctions characterized by a low zero-bias conductance (large tunneling gap) [38]. These current fluctuations eventually lead to a rapid degradation of the device. For large zero-bias conductance, a prerequisite for a spontaneous hot electron emission process, the current in the device is considerably more stable. At voltages around 500 mV, current fluctuations are contained within 1% and are slightly increasing to 6% at 1 V.
Examining Fig. 2(b), light is manifestly not restricted to the antenna feedgap but is also emanating from the physical contour of the geometry. This signal suggests the existence of surface plasmons in the Au layer. When the plasmons reach the edges of the Au layer, they couple to free-space photons. We did not measure the polarization state of the source, but from the angular spread of the plasmon signatures, we argue that there is no preferred emission polarization as already observed in [29]. In this context, the corresponding Fourier plane shown in Fig. 2(c) helps to identify the different modal excitations. Because of its extremely local character, the photons released by the source are widely distributed in the Fourier plane and constitute the diffuse background of the image. The scattering at the upper and lower Au edges also gives a weak signal. The lateral dispersion of light along the edges suggests that the source is not directive and couples its radiation to plasmons propagating with a large directional spread. This scattered photons are thus contained within a large wave-vector spectrum in the Fourier plane. Their distribution can be found on either side of the ±ky/k0 following an arc of circle of radius given by the effective index of the SPP Au/air interface. k0 is the vacuum wavevector taken at the maximum of the spectral emission.
The light scattered out at the left extremity of the dielectric ridge in Fig. 2(b) indicates the presence of a waveguided surface plasmon mode supported by the DLSPPW, and is thus the mode of interest. In the Fourier plane, the surface plasmon diffracted by the physical termination of a waveguide is unidirectional and bears the signature of the guided mode [31, 39, 40]. Note that it is difficult to fully discriminate between the leakage of the guided mode around +kx/k0=1.1 and its scattering at the waveguide extremity that occurs near the critical angle +kx/k0 ≃1.0. However, since no mode leakage is visible in the image plane, the luminous line located near +kx/k0=1.0 is therefore attributed to the scattering of the mode by the end of the waveguide [39]. This contribution is the brightest feature in the Fourier plane because of its narrow angular distribution.
Finally, Fig. 2(b) shows a scattered contribution emitted at the right-hand side of the junction. The dielectric environment near this edge suggests that a mode is propagating in the −x direction and is scattered out by an area of the sample that was not completely lifted off after fabrication. This scattering edge is visible in the SEM image of Fig. 2(a). In the Fourier plane, the corresponding angular signature is spread near −kx/k0=1.0.
We carried out the experiments on a substrate containing five demonstrators. The excitation of surface plasmons by the emission of the junction obtained on this device was qualitatively reproduced with another similar structure as well as on two devices featuring a 0.8 μm-wide Au strip as the one shown in Fig. 1(a). We successfully obtained tunnel junctions for all the devices, with values for the tunnel current of at least several μA. We observe the DLSPPW guided plasmon mode with four of those tunnel junctions, which corresponds to a success rate of 80%. We have a better return of experience concerning the electromigration of nanowires simply deposited on glass, where 90% of the junctions do emit light. Electrical excitation of the surface plasmon guided modes is not restricted to the light emission process discussed here. Photons released by inelastic electron tunneling events as in [23–25] are sharing similar essential characteristics such as the extreme localisation of the emission and are thus expected to couple to the existing plasmon modes. Reports of such modal excitation can be found in the literature for STM configuration [19, 41].
In this series of measurements, the device is operated with a constant voltage. At the maximum operated bias (1V), the electrical power required to generate and hold the optical signal is 50 μW. For comparison purposes, ring modulators in an on-chip photonic network platform requires 30 μW of static power to maintain the optical signal constant [42].
4. Finite-element simulations of the architecture
To better understand the response of the waveguide geometry, we performed a three-dimensional simulation using finite-element computation (Comsol). Light emitted by inelastically tunneling electrons is well approximated by a dipole oscillating normal to the gap [19,23,43]. For a hot electron emission process, a dipole-like pattern is also measured [29], where the electromigrated nanowire acts as a redirecting antenna. With the above considerations, the source is modeled as a dipole emitting at a wavelength of 850 nm inside the antenna feedgap at a height of 15 nm from the underlying dielectric ridge. The dipole radiates a constant power in all three dimensions of space. We compared several simulations by including either a single dipole oriented along the y-axis or a sum of three dipoles oriented along x, y and z to reflect the fact that the emission is not polarized. We obtained qualitatively similar results in both cases. The dimensions of the calculated geometry follow the experimental device.
Figures 3(a) and 3(b) show the distribution of the radiation in the device at a x, y plane corresponding to the middle of the DLSPPW. The color is in logarithmic scale to reveal the different modal contributions existing in the structure. According to the simulation, the main mode excited by the junction is a gap plasmon mode confined in the SiOx layer between the nanowire and the DLSPPW gold strip. Figure 3(c) shows the two-dimensional map of the gap plasmon mode. At 800 nm the effective index extracted from the two-dimensional simulation of the gap mode is 2.2. It is thus outside the detection window of the objective limited at N.A.=1.49, and explains why this mode is not observed experimentally. Despite a large effective index difference with the DLSPPW mode, which has an effective index of 1.1 at 800 nm according to the two-dimensional simulation, the gap plasmon mode couples to the DLSPPW mode. Figure 3(d) displays the calculated lateral distribution of the plasmon sustained by the dielectric load. Because of the non optimal SiOx thickness, the mode significantly extends in the air in accordance with the weak confinement experimentally observed in the Fourier plane.
Fig. 3 (a) 3D Finite-element simulation of the intensity of the electrical field inside the structure. (b) Sectional view of the 3D simulation taken in the middle of the SiOx layer. A dipole emitting at 850 nm mimics the electron-fed antenna. Logarithmic color scale. (c) Two-dimensional distribution of the gap plasmon mode confined by the Au interfaces in the SiOx layer. (d) Two-dimensional distribution of the mode supported by the DLSPPW. For (c) and (d) the color scales are linear.
The mismatch of the effective indices gives rise to a reflection of the gap plasmon mode within the finite length of the upper Au layer constituting the contact electrodes and the nanowire. This reflection forms interference fringes along the structure akin to a Fabry–Pérot cavity. Such cavity-like behavior is evidently dictating the spectrum of the light coupled into the dielectric-loaded surface plasmon waveguide. A somewhat similar phenomenon was observed by Gruber et al. [44] when studying plasmonic silver nanowires excited by quantum dots. The authors found a quantum dot’s emission spectrum modulated by resonances of the plasmonic modes. The modulated spectrum in our device is discussed in the following.
5. Spectral characteristics
We analyze the spectral distribution of the light emitted by the optical antenna and compare it to the numerical results discussed above. Spectra are acquired by dispersing the collected light in a spectrometer (Andor, Shamrock 303i) equipped with a CCD camera (Andor, Newton EMCCD). By placing the DLSPPW in a direction parallel to the slit of the spectrometer, we simultaneously record spectra emitted at different points along the structure. The experimental solid red line in Fig. 4(a) shows the spectral content radiated by the electron-fed antenna for a bias of 1 V. The spectrum confirms the assumption about the physical origin of the light emission. The energy of the released photons exceeds the energy provided by the bias, and thus violates the quantum limit. We thus discard an electroluminescent process involving inelastic electron tunneling. Coherent two-electron interactions may lead to above-threshold radiation [40, 45, 46]. However, we did not observed the expected sharp kink in the spectra at the onset of overbias emission. The experimental similarities with the mechanism discussed in [29] suggests that the spontaneous emission of an electrically–pumped hot electron gas is the dominant process at play. At the drain electrode, tunneling electrons give an extra thermal energy to the electron distribution via electron-electron scattering. Electron-phonon interaction is reduced in this confined geometry because the characteristic length is shorter than the inelastic scattering mean-free path. Part of the energy of the hot electron gas dissipates by colliding with the boundaries of the structure [47] and emits a blackbody radiation dictated by the temperature of the electrons. The red curve of Fig. 4(a) reflects the tail of this radiation. The solid blue line in Fig. 4(a) is a spectrum acquired at the entrance of the DLSPPW. It features a number of well-defined resonance peaks that are absent on the emission spectrum of the antenna. Based upon the numerical results discussed above, we attribute these peaks to Fabry-Pérot resonances of the gap plasmon mode coupling to the DLSPPW. To compare spectra recorded at different locations along the DLSPPW, it is important to calculate the wavelength-dependent propagation distance of the surface plasmon mode. Such dependence is shown in Fig. 4(b). As expected from a Au plasmonic waveguide, near-infrared wavelengths are generally less attenuated than bluer components. The optimum propagation length at ∼930 nm originates from the finite size of the Au strip [31]. We use the dispersion of the propagation distance to interpret the spectrum emitted near the end of the waveguide. Using the trend computed in Fig. 4(b) and the radiation measured at the entrance of the DLSPPW, we estimate the spectrum near the extremity of the DLSPPW considering a propagation length of 4 μm. The dotted green line in Fig. 4(a) is the expected curve, and is in fair agreement with the corresponding experimental spectrum (solid green line).
Fig. 4 (a) The red curve is the spectrum emitted by the electron-fed antenna. The blue curve is acquired at the entrance of the DLSPPW and the green curve near the end of the DLSPPW. The dotted green curve is the spectrum measured at the beginning of the waveguide multiplied by the plasmon attenuation resulting from the propagation in the DLSPPW (see Fig. 4(b)). All spectra are normalized by the quantum efficiency of the detector and are corrected for etalon effect. An electron multiplier gain of 255 was used for the CCD camera equipping the spectrometer. (b) Calculated wavelength-dependent propagation length of the DLSPPW mode. The finite width of the Au strip limits the propagation length for longer wavelengths. (c) The red curve is the normalized spectrum of the plasmonic mode calculated by dividing the curve measured at the entrance of the DLSPPW by the wavelength content emitted by the antenna. This spectrum features a number of resonances peaks resulting from Fabry-Pérot interferences. The blue curve represents the calculated transmission of a Fabry-Pérot cavity with an effective index equal to the calculated effective index of the gap mode (see Fig. 4(d)) and an effective length of 920 nm (corresponding to the distance between the antenna feedgap and the drain electrode). (d) Calculated effective indices of the gap mode (red curve) and the DLSPPW mode (blue curve).
The solid red line on Fig. 4(c) represents the normalized spectrum of the plasmonic mode calculated by dividing the spectral content measured at the entrance of the DLSPPW by the one radiated from the antenna. Fabry-Pérot resonances are clearly visible. We compare in the following this spectrum to the transmission of a cavity calculated using the classical formula of a Fabry-Pérot resonator:
where is calculated using the effective optical index of the gap mode and takes into account an effective length of the cavity Leff. The reflection coefficient R at the interface between the gap mode of effective index and the waveguided plasmonic mode of effective index is given by:The dispersions of the effective indices are calculated using Comsol and are shown on Fig. 4(d). The blue curve in Fig. 4(c) is the best agreement between the experimental data and this simple model. It is obtained for a Fabry-Pérot cavity with an effective length of 920 nm. This value corresponds to a cavity length comprised between the antenna feedgap on one side and the drain electrode on the other side (including the tapered section between the nanowire and the electrode). The effective index of the gap mode changes radically once it is no longer confined under the nanowire. However, the mode extends under the tapered electrode as illustrated in the computational intensity maps in Fig. 4(a). Moreover, electromigrated gaps are notoriously irregular with a complex geometry (see for instance the inset of Fig. 1). The ill-defined interstice could therefore be wide enough at some location to affect the propagating gap mode and causes a reflection. So, it is likely that the physical edge of the drain electrode and the antenna feedgap constitute the effective limits of the Fabry-Pérot resonator. While this cavity model is extremely simple, it provides a general understanding of the spectral characteristics of the electrically excited modes propagating in the plasmonic waveguide.
6. Conclusion
We propose in this work an alternative integrated technology to electrically excite surface plasmons in a waveguide geometry. The electrical signal is transduced to a propagating surface plasmon mode by the mediation of a light-emitting electron-fed optical gap antenna. The antenna radiates a broadband emission that couples to the surface plasmon modes of the structure. The spectral content of the plasmon exiting the guiding platform is dictated by design geometry, offering thus a route to tune the transmission and optimize the device for a specific operating wavelength. The discussion is focused on the spectral content localized in the visible to near-infrared region. However, the emitted spectrum covers a larger bandwidth extending over the telecom bands. In this infrared spectral region, light emitted from inelastically tunneling electrons coexists with the maximum of thermal emission and may favorably contribute at increasing the external quantum yield. We restricted our study to the electrical excitation of surface plasmon modes sustained in dielectric-loaded surface plasmon waveguides. However, this approach can be generalized to other plasmonic geometries and integrated photonic platforms. Specifically, metal nanowires and slot waveguides may offer a favorable alternative to reduce the size mismatch between electronic and optical interconnects. Efforts at embedding an electrical excitation of photonic modes in TiO2 waveguides are currently underway. The similarity of the electrical light source introduced here with optical rectennas discussed in [48] offers a strategy for developing a common approach seamlessly integrating on-chip the electrical excitation and the detection of light signals [49].
Acknowledgments
The research leading to these results has received fundings from the European Research Council under the European Community’s Seventh Framework Program FP7/2007–2013 Grant Agreement n°306772, the Labex ACTION (contract ANR-11-LABX-01-01) and the regional program PARI. We thank J.-C. Weeber for suggesting to us the role of gap plasmon mode, and S. Sauvage from IEF for giving us more information about the external quantum efficiency of quantum dots. The work of A. V. Uskov was financially supported in part by the Government of the Russian Federation (Grant 074-U01) through ITMO Visiting Professorship program.
References and links
1. W. L. Barnes, A. Dereux, and T. W. Ebbesen, “Surface plasmon subwavelength optics,” Nature 424(6950), 824–830 (2003). [CrossRef] [PubMed]
2. M. Dragoman and D. Dragoman, “Plasmonics: applications to nanoscale terahertz and optical devices,” Prog. Quantum Electron. 32(1), 1–41 (2008). [CrossRef]
3. E. Ozbay, “Plasmonics: merging photonics and electronics at nanoscale dimensions,” Science 311(5758), 189–193 (2006). [CrossRef] [PubMed]
4. R. Zia, J. A. Schuller, A. Chandran, and M. L. Brongersma, “Plasmonics: the next chip-scale technology,” Mater. Today 9(7–8), 20–27 (2006). [CrossRef]
5. R. J. Walters, R. V. A. Van Loon, I. Brunets, J. Schmitz, and A. Polman, “A silicon based electrical source of surface plasmon polaritons,” Nature Mater. 9(1), 21–25 (2010). [CrossRef]
6. P. Rai, N. Hartmann, J. Berthelot, J. Arocas, G. Colas des Francs, A. Hartschuh, and A. Bouhelier, “Electrical excitation of surface plasmons by an individual carbon nanotube transistor,” Phys. Rev. Lett. 111(2), 026804 (2013). [CrossRef] [PubMed]
7. D. M. Koller, A. Hohenau, H. Ditlbacher, N. Galler, F. Reil, F. R. Aussenegg, A. Leitner, E. J. W. List, and J. R. Krenn, “Organic plasmon-emitting diode,” Nature Photon. 2(11), 684–687 (2008). [CrossRef]
8. P. Neutens, L. Lagae, G. Borghs, and P. Van Dorpe, “Electrical excitation of confined surface plasmon polaritons in metallic slot waveguides,” Nano Lett. 10(4), 1429–1432 (2010). [CrossRef] [PubMed]
9. L. Wang, T. Li, L. Li, W. Xia, X. G. Xu, and S. N. Zhu, “Electrically generated unidirectional surface plasmon source,” Opt. Express 20(8), 8710–8717 (2012). [CrossRef] [PubMed]
10. K. C. Y. Huang, M. K. Seo, T. Sarmiento, Y. Huo, J. S. Harris, and M. L. Brongersma, “Electrically driven subwavelength optical nanocircuits,” Nature Photon. 8(3), 244–249 (2014). [CrossRef]
11. P. Y. Fan, C. Colombo, K. C. Y. Huang, P. Krogstrup, J. Nygard, A. F. I. Morral, and M. L. Brongersma, “An electrically-driven GaAs nanowire surface plasmon source,” Nano Lett. 12(9), 4943–4947 (2012). [CrossRef] [PubMed]
12. R. H. Ritchie, “Plasma losses by fast electrons in thin films,” Phys. Rev. 106(5), 874–881 (1957). [CrossRef]
13. M. V. Bashevoy, F. Jonsson, A. V. Krasavin, N. I. Zheludev, Y. Chen, and M. I. Stockman, “Generation of traveling surface plasmon waves by free-electron impact,” Nano Lett. 6(6), 1113–1115 (2006). [CrossRef] [PubMed]
14. E. J. R. Vesseur, R. de Waele, H. J. Lezec, H. A. Atwater, F. J. G. de Abajo, and A. Polman, “Surface plasmon polariton modes in a single-crystal Au nanoresonator fabricated using focused-ion-beam milling,” Appl. Phys. Lett. 92(8), 083110 (2008). [CrossRef]
15. W. Cai, R. Sainidou, J. J. Xu, A. Polman, and F. J. G. de Abajo, “Efficient generation of propagating plasmons by electron beams,” Nano Lett. 9(3), 1176–1181 (2009). [CrossRef] [PubMed]
16. J. Lambe and S. L. McCarthy, “Light emission from inelastic electron tunneling,” Phys. Rev. Lett. 37(14), 923–925 (1976). [CrossRef]
17. R. Berndt and J. K. Gimzewski, “Inelastic tunneling excitation of tip-induced plasmon modes on noble-metal surfaces,” Phys. Rev. Lett. 67(27), 3796–3799 (1991). [CrossRef] [PubMed]
18. Z. C. Dong, X. L. Zhang, H. Y. Gao, Y. Luo, C. Zhang, L. G. Chen, R. Zhang, X. Tao, Y. Zhang, J. L. Yang, and J. G. Hou, “Generation of molecular hot electroluminescence by resonant nanocavity plasmons,” Nature Photon. 4(1), 50–54 (2010). [CrossRef]
19. P. Bharadwaj, A. Bouhelier, and L. Novotny, “Electrical excitation of surface plasmons,” Phys. Rev. Lett. 106(22), 226802 (2011). [CrossRef] [PubMed]
20. T. Wang, E. Boer-Duchemin, Y. Zhang, G. Comtet, and G. Dujardin, “Excitation of propagating surface plasmons with a scanning tunnelling microscope,” Nanotech. 22(17), 175201 (2011). [CrossRef]
21. Z. Dong, H. S. Chu, D. Zhu, W. Du, Y. A. Akimov, W. P. Goh, T. Wang, K. E. J. Goh, C. Troadec, C. A. Nijhuis, and J. K. W. Yang, “Electrically-excited surface plasmon polaritons with directionality control,” ACS Photon. 2(3), 385–391 (2015). [CrossRef]
22. C. Grosse, A. Kabakchiev, T. Lutz, R. Froidevaux, F. Schramm, M. Ruben, M. Etzkorn, U. Schlickum, K. Kuhnke, and K. Kern, “Dynamic control of plasmon generation by an individual quantum system,” Nano Lett. 14(10), 5693–5697 (2014). [CrossRef] [PubMed]
23. J. Kern, R. Kullock, J. Prangsma, M. Emmerling, M. Kamp, and B. Hecht, “Electrically driven optical antennas,” Nature Photon. 9(9), 582 (2015). [CrossRef]
24. M. Parzefall, P. Bharadwaj, A. Jain, T. Taniguchi, K. Watanabe, and L. Novotny, “Antenna-coupled photon emission from hexagonal boron nitride tunnel junctions,” Nature Nanotech. 10(10), 1058 (2015). [CrossRef]
25. Y. Vardi, E. Cohen-Hoshen, G. Shalem, and I. Bar-Joseph, “Fano resonance in an electrically driven plasmonic device,” Nano Lett. 16(1), 748–752 (2016). [CrossRef]
26. T. Holmgaard and S. I. Bozhevolnyi, “Theoretical analysis of dielectric-loaded surface plasmon-polariton waveguides,” Phys. Rev. B 75(24), 245405 (2007). [CrossRef]
27. K. Hassan, J.C. Weeber, L. Markey, and A. Dereux, “Thermo-optical control of dielectric loaded plasmonic racetrack resonators,” J. Appl. Phys. 110(2), 023106 (2011). [CrossRef]
28. J. Gosciniak and S. I. Bozhevolnyi, “Performance of thermo-optic components based on dielectric-loaded surface plasmon polariton waveguides,” Sci. Rep. 3, 1803 (2013). [CrossRef]
29. 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(9), 5811–5818 (2015). [CrossRef] [PubMed]
30. J. Grandidier, G. Colas des Francs, L. Markey, A. Bouhelier, S. Massenot, J. C. Weeber, and A. Dereux, “Dielectric-loaded surface plasmon polariton waveguides on a finite-width metal strip,” Appl. Phys. Lett. 96(6), 063105 (2010). [CrossRef]
31. J. Grandidier, S. Massenot, G. Colas des Francs, A. Bouhelier, J. C. Weeber, L. Markey, A. Dereux, J. Renger, M. U. Gonzalez, and R. Quidant, “Dielectric-loaded surface plasmon polariton waveguides: figures of merit and mode characterization by image and Fourier plane leakage microscopy,” Phys. Rev. B 78(24), 245419 (2008). [CrossRef]
32. S. Massenot, J. Grandidier, A. Bouhelier, G. Colas des Francs, L. Markey, J.-C. Weeber, A. Dereux, J. Renger, M. U. Gonzàlez, and R. Quidant, “Theoretical analysis of dielectric-loaded surface plasmon-polariton waveguides,” Appl. Phys. Lett. 91(24), 243102 (2007). [CrossRef]
33. S. Girod, J. L. Bubendorff, F. Montaigne, L. Simon, D. Lacour, and M. Hehn, “Real time atomic force microscopy imaging during nanogap formation by electromigration,” Nanotech. 23(36), 365302 (2012). [CrossRef]
34. M.-M. Mennemanteuil, J. Dellinger, M. Buret, G. Colas des Francs, and A. Bouhelier, “Pre-determining the location of electromigrated gaps by nonlinear optical imaging,” Appl. Phys. Lett. 105(2), 021101 (2014). [CrossRef]
35. J. Berthelot, F. Tantussi, P. Rai, G. Colas des Francs, J. C. Weeber, A. Dereux, F. Fuso, M. Allegrini, and A. Bouhelier, “Determinant role of the edges in defining surface plasmon propagation in stripe waveguides and tapered concentrators,” J. Opt. Soc. Am. B. 29(2), 226 (2012). [CrossRef]
36. M. E. Reimer, M. P. van Kouwen, M. Barkelid, M. Hocevar, M. H. M. van Weert, R. E. Algra, E. P. A. M. Bakkers, M. T. Bjork, H. Schmid, H. Riel, L. P. Kouwenhoven, and V. Zwillera, “Single photon emission and detection at the nanoscale utilizing semiconductor nanowires,” J. Nanophoton. 5(1), 053502 (2011). [CrossRef]
37. B. Ellis, M. A. Mayer, G. Shambat, T. Sarmiento, J. Harris, E. E. Haller, and J. Vuckovic, “Ultralow–threshold electrically pumped quantum–dot photonic–crystal nanocavity laser,” Nature Photon. 5(5), 297–300 (2011). [CrossRef]
38. M. Frimmer, G. Puebla-Hellmann, A. Wallraff, and L. Novotny, “The role of titanium in electromigrated tunnel junctions,” Appl. Phys. Lett. 105(22), 221118 (2014). [CrossRef]
39. T. Shegai, V. D. Miljkovic, K. Bao, H. Xu, P. Nordlander, P. Johansson, and M. Kall, “Unidirectional broadband light emission from supported plasmonic nanowires,” Nano Lett. 11(2), 706–711 (2011). [CrossRef] [PubMed]
40. Z. Wang, H. Wei, D. Pan, and H. Xu, “Controlling the radiation direction of propagating surface plasmons on silver nanowires,” Laser Photon. Rev. 8(4), 596–601 (2014). [CrossRef]
41. B. Rogez, R. Horeis, E. Le Moal, J. Christoffers, K. Al-Shamery, G. Dujardin, and E. Boer-Duchemin, “Optical and electrical excitation of hybrid guided modes in an organic nanofiber-gold film system,” J. Phys. Chem. C 119(38), 22217–22224 (2015). [CrossRef]
42. K. Bergman, L. P. Carloni, A. Biberman, J. Chan, and G. Hendry, Photonic network-on-chip design (Springer, 2013).
43. E. Le Moal, S. Marguet, B. Rogez, S. Mukherjee, P. Dos Santos, E. Boer-Duchemin, G. Comtet, and G. Dujardin, “An electrically excited nanoscale light source with active angular control of the emitted light,” Nano Lett. 13(9), 4198–4205 (2013). [CrossRef] [PubMed]
44. C. Gruber, A. Trugler, A. Hohenau, U. Hohenester, and J. R. Krenn, “Spectral modifications and polarization dependent coupling in tailored assemblies of quantum dots and plasmonic nanowires,” Nano Lett. 13(9), 4257–4262 (2013). [CrossRef] [PubMed]
45. C. Schneider, A. Rahimi-Iman, N. Y. Kim, J. Fischer, I. G. Savenko, M. Amthor, M. Lermer, A. Wolf, L. Worschech, V. D. Kulakovskii, I. A. Shelykh, M. Kamp, S. Reitzenstein, A. Forchel, Y. Yamamoto, and S. Hofling, “An electrically pumped polariton laser,” Nature 497(7449), 348–352 (2013). [CrossRef] [PubMed]
46. K. Kaasbjerg and A. Nitzan, “Theory of light emission from quantum noise in plasmonic contacts: above-threshold emission from higher-order electron-plasmon scattering,” Phys. Rev. Lett. 114(12), 126803 (2015). [CrossRef] [PubMed]
47. R. D. Fedorovich, A. G. Naumovets, and P. M. Tomchuk, “Electron and light emission from island metal films and generation of hot electrons in nanoparticles,” Phys. Rep. 328(2–3), 74–179 (2000). [CrossRef]
48. 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(5), 2330–2338 (2014). [CrossRef] [PubMed]
49. Z. Han, I. P. Radko, N. Mazurski, B. Desiatov, J. Beermann, O. Albrektsen, U. Levy, and S. I. Bozhevolnyi, “On-chip detection of radiation guided by dielectric-loaded plasmonic waveguides,” Nano Lett. 15(1), 476–480 (2015). [CrossRef]
