Plasmon assisted photoelectric hot electron collection in a metal-semiconductor junction can allow for sub-bandgap optical to electrical energy conversion. Here we report hot electron collection by wafer-scale Au/TiO2 metallic-semiconductor photonic crystals (MSPhC), with a broadband photoresponse below the bandgap of TiO2. Multiple absorption modes supported by the 2D nano-cavity structure of the MSPhC extend the photon-metal interaction time and fulfill a broadband light absorption. The surface plasmon absorption mode provides access to enhanced electric field oscillation and hot electron generation at the interface between Au and TiO2. A broadband sub-bandgap photoresponse centered at 590 nm was achieved due to surface plasmon absorption. Gold nanorods were deposited on the surface of MSPhC to study localized surface plasmon (LSP) mode absorption and subsequent injection to the TiO2 catalyst at different wavelengths. Applications of these results could lead to low-cost and robust photo-electrochemical applications such as more efficient solar water splitting.
© 2016 Optical Society of America
The fundamental limitation on spectral range and efficiency of solar energy harvesting for photoelectrochemistry applications lies in the large bandgap of stable semiconductor catalysts, which significantly reduces the efficiencies of applications such as water splitting [1, 2]. The large bandgap of electrochemically stable semiconductor catalysts commonly used, for example rutile TiO2 with a bandgap of about 3.0 eV , severely restricts solar-to-fuel efficiency with very poor solar energy absorption and conversion at visible spectrum.
Adding a second layer of metal, forming a Schottky barrier at metal/semiconductor interface, provides a promising solution to enhance the photoelectric energy harvesting efficiency at sub-bandgap spectrum . Electrons in the metal could be excited from the Fermi level to form so called “hot electrons” . Since the Schottky barrier at metal/semiconductor interface is always lower than the bandgap of semiconductor, sub-bandgap photoelectric energy conversion could be achieved by effectively injecting hot electrons which could overcome the Schottky barrier.
Two possible hot electron generation mechanisms in metal are direct photoexcitation and plasmon decay . Direct photoexcitation mainly happens through interband transition of electrons from large-density bands, such as d-band in Au, to higher bands close to the Fermi surface, which do not have enough energy to overcome the Schottky barrier . On the other hand, surface plasmon polariton (SPP) and localized surface plasmon resonance (LSPR) could form electron oscillation with strongly enhanced electromagnetic field at the interface between metal and semiconductor . Through non-radiative decay of surface plasmon resonance, hot electrons with energy level higher than the Schottky barrier could be generated. Because the d-bands of noble metals such as Au and Ag locate 2.4 eV and 4.0 eV below the Fermi level , hot electrons with high energy levels are mainly generated through intraband transition in surface plasmon decay. Recent breakthroughs in plasmonic energy conversion have shown that surface plasmon resonance could largely improve the efficiency of hot electron generation and injection [5, 9]. Devices fabricated with e-beam lithography [10, 11], nanoparticles and nano-imprint lithography  show application potentials for solar energy harvesting and photo detection [12–14]. However, the absorption properties of these reported hot carrier devices are not optimized, especially for sub-bandgap range. Meanwhile the nanoscale size of metallic structures used for plasmonic hot electron generation severely limits the large scale production and increase the fabrication costs.
In order to design a broadband sub-bandgap hot electron collection device, three major challenges need to be achieved: (1) broadband sub-bandgap absorption (2) surface plasmon resonance below the bandgap of semiconductor and (3) large scale production. The authors’ previous research in wafer-scale metal-dielectric photonic crystal device (MDPhC) demonstrated a broadband absorption in thin metal layers over the whole solar spectrum . The multiple absorption modes supported by the 2D nano-cavity arrays increases the interaction time between photon and metal, improving the absorption over most of the solar spectrum. Here we report the wafer-scale Au/TiO2 metallic-semiconductor photonic crystals device achieving all the three challenges above. Major experimental data of this paper was presented at the 42nd IEEE PVSC conference with a view that SPP at the interface between Au and TiO2 significantly contributed to the generation and injection of hot electrons. This observation inspires us to induce multiple SPP at difference wavelength in our MSPhC device for multi-band hot electron collection.
2. Design and optimization
Figure 1(a) shows the top view of a 1 cm by 1cm MSPhC chip. Multiple layers are deposited on the Al2O3 nano-cavity arrays we fabricated for solar absorbers before [15, 16]. 200 nm Au and Ti layer are used as probe contact at the two ends of the chip. The widths of region I, III and IV are approximately 2 mm. Figure 1(b) shows the basic structure and operation design of MSPhC of the II and III region in Fig. 1(a), which uses a thin Au layer as the optical absorber, TiO2 as the n-type semiconductor layer, and indium tin oxide (ITO) as the back transparent electrical contact to reduce the series resistance. As shown in Fig. 1(c), the cavity structure could support three light absorption modes: cavity, waveguide and SPP modes. These multiple modes will increase the total light absorption of the device, especially by the Au layer. Figures 1(d) and 1(e) shows the Au/TiO2/ITO interfaces and band diagram, and how hot electron is generated and collected. After being excited at or below the Fermi level, hot electrons with enough energy could transfer through the interface between Au and TiO2, be injected to and collected at the conduction band of TiO2.
In order to optimize the photoelectric performance, we investigated the optical absorption properties of MSPhC with different structural parameters. The Au layer is an absorber layer to generate hot electrons. Thus, we want to trap the incident light in the cavity and increase the interaction time between light and Au. The dependence of total absorption by the device on the radius and depth of the nano-cavity arrays were studied through finite difference time domain (FDTD) simulation method. We averaged the total absorption over the wavelength range from 200 nm to 2 µm with and without solar spectrum (AM 1.5) weighting. Generally, the trade-off between the light-trapping ability and numbers of supported absorption mode determines the absorption of MSPhC. A larger and deeper cavity can support more absorption modes but lose the confinement of incident light. The cavity structure used in simulation has 13 nm Au, 75 nm TiO2, 50 nm Al2O3 and 30 nm ITO. We varied the radius from 167 nm to 1,000 nm and depth from 250 nm to 1,500 nm. As shown in Fig. 2(a), FDTD simulation results show that the maximum total absorption of MSPhC happens when radius are 250 nm for solar-weighted cases and 500 nm for non-solar-weighted cases. For depth dependent absorption, as shown in Fig. 2(b), a deeper cavity can hold more cavity modes and increase the total absorption. For the fabrication convenience, to form a uniform deposition over the cavity structure, we choose 1 µm as the cavity depth.
FDTD simulation method was also used to analyze the dependence of absorption on Au layer thickness. Absorption per unit volume was estimated by the expression Pabs = 0.5ω|E|2imag(ε), where ω is the frequency, E is the electric field, and ε is the permittivity of the material. The radius and depth of cavity used in simulation is 250 nm and 1 µm. Thickness of Al2O3, ITO, TiO2 are the same as above. Figure 3 shows that with the decrease of Au thickness, from 52 nm to 26 nm to 13 nm, the absorption spectrum is broadened into IR spectrum. With 52 nm, a clear cut-off mode of 740 nm is shown where light at longer wavelengths is reflected and no longer absorbed. With 13 nm Au layer, the cut-off of absorption even reaches above 1,500 nm. Thin Au layer and the Al2O3 cavity structure allows light to penetrate and be coupled within the metal/insulator/metal (M/I/M) structure, which increase the light-metal interaction time and the total absorption. The waveguide and cavity modes concentrate the electric field oscillation in Al2O3 and Air in the cavity. Because the imaginary part of permittivity of air, Al2O3 and TiO2 are pretty small at visible spectrum, the light trapped in these layers will experience low-loss oscillating and finally be absorbed thorough the interaction with Au layer . In the perspective of hot electron injection, thin Au layer could reduce the hot electron transfer distance, which can improve the possibility of hot electron injection before they are scattered and thermalized . The thickness effect on the hot electron transfer is measured and discussed below. In summary, our simulations suggest that MSPhC structures with thinner Au layers allow for increased broadband absorption which may be beneficial for hot electron generation and injection.
Through absorption optimization by FDTD, the final structure we fabricated is as following. The nano-cavity array of Al2O3 with a depth of d = 1 µm, an inner radius of r = 250 nm and distance between centers of two nearest cavities of 840 nm is fabricated through a wafer-scale and CMOS compatible sidewall lithography process, which is described in details in our previous work [15, 16]. Basically, the cavities were patterned via large wafer-scale stepper lithography. Within a single dye of 1 cm X 1 cm, the cavity yield is very high with an estimated yield of >99%. Small variations within each cavity may exist such as connections between the cavities, roughness on the cavity top surface, and sidewall roughness. Across the 6” wafer the variation from the center of the wafer to the edge are within the standard wafer scale processing limits. For our applications these parameters do not have an effect on our overall device performance. Then, 30 nm of indium tin oxide (ITO) is deposited through sputtering and annealed at 450°C for 2 hour as back electrical contact. About 75 nm of TiO2 is deposited through atomic layer deposition (ALD), following by annealing in air at 450°C for 1 hour. Finally, a thin Au is deposited through sputtering. To verify the effect of Au layer thickness on the absorption and photoelectric conversion characteristic of MSPhC, Au layer of three different thickness of 10, 20 and 30 nm were sputtered. Finally, as shown in Fig. 1(a), 200 nm of Au and Ti are deposited on two ends of the MSPhC chip as electrical contact for device test. The thick Ti layer forms ohmic contact with TiO2. Even though we deposited the lateral contact layer of ITO to reduce the series resistance of the device, the TiO2 in between of Ti and ITO will increase the series resistance and reduce the photoelectrical conversion efficiency. From measured I-V curve, the estimated series resistance of our device is about 39 kΩ . Figure 4 shows the photos of MSPhC by SEM and focused ion beam (FIB) milling. In Fig. 4(c), the compositions of different layers are shown on the cross-section view of a single nano-cavity. Pt layer on top and bottom of the nano-cavity serves as protective layer for the purpose of imaging. The particle-like structure on the inside wall of the cavity is formed due to Pt deposition.
4.1 Optical characterization
Since MSPhC is fabricated on Silicon substrate, it is difficult to directly get the absorption spectrum to compare against our calculations. In order to evaluate the absorption ability of it, we measure the UV-Vis reflection spectrum of MSPhC with Au layer of 10 nm and compared it with FDTD simulation of MSPhC with 10 nm Au. The actual thickness of the Au layer is picked from the images of MSPhC. As shown in Fig. 5, both the experimental result and the simulated reflection result shows a low reflection over the range from 200 nm to 1500 nm. Due to the thin Au film, a fraction of the incident light is able to transmit through the Au, and thus a reflectivity of only 34% is shown at 1500 nm. Differences between the simulation and experiment can be attributed to the non-uniform coverage of the Au deposition and fabrication geometrical variations averaged out over a large measurement area. The real Au thickness is not constant and varies as the Au goes down the sidewall of the cavity, which may cause the frequency shift in reflection spectrum between simulation and measurement. But both of them show the overall low reflectivity due to the MSPhC structure.
The broadband absorption property of MSPhC comes from the multiple absorption modes that the nano-cavity structure can support. In these modes, incident light is coupled to the MSPhC structure, and forms strong electric field oscillation. In order to understand the optical absorption in MSPhC, we did modal analysis to decompose the contributions of different modes. Generally speaking, the MSPhC can hold four basic modes. (i) Cavity mode , in which the electric field of incident light is trapped inside the cavity. (ii) Gap mode, in which light is coupled to the gaps between cavities. (iii) Waveguide mode , in which the light is coupled inside MIM waveguide-like structure. (iv) Surface plasmon polariton (SPP) mode , in which light is confined at the interface between Au and TiO2.
FDTD simulation method was used to verify the existence of different modes in MSPhC. In order to stimulate different resonance modes in MSPhC, 9 broadband dipole source were placed randomly inside the cavity structure. Then 9 time monitors placed randomly at different positions in MSPhC recorded the time domain of electric field at these positions. The electric field oscillation is then analyzed through discrete Fourier transform. As shown in Fig. 6, two major peaks appear at 500 nm and 600 nm. With the increase of Au layer thickness, the amplitude of electric field oscillation decreased. By plotting the electric field distribution inside MSPhC we could identify the resonance mode at 500 nm and 600 nm. As shown in Fig. 7(a), at 500 nm, the electric field concentrates at the Al2O3 insulator layer, which is a waveguide resonance mode. Due to the low imaginary part of permittivity of Al2O3, waveguide resonance does not enhance the total optical absorption. Besides, because electric field is confined in the insulator layer, the oscillation may not contribute to the generation of hot electrons in Au layer. The second resonance peak near 600 nm is revealed to be a SPP resonance mode. As shown in Fig. 7(b), the electric field propagating at the interface between Au and TiO2. The FWHM of SPP mode is larger than that of waveguide mode, indicating that the SPP mode is a high-loss mode in which the plasmons are decayed non-radiatively into interfacial electron-hole pairs. Cavity resonance was also observed in the simulation, as shown in Fig. 7(c) at about 750 nm. However, its amplitude is several orders smaller compared with waveguide and SPP modes.
4.2 Photoresponse measurement
We measured the photoresponse of MSPhC with laser diodes at several specific wavelengths from 405 nm to 805 nm. Figure 8 shows the photoresponse of MSPhC with 10, 20 and 30 nm Au layer compared with a flat chip of 30 nm Au, 30 nm TiO2 and 30 nm ITO. Due to high series resistance resulted from the quality of deposited TiO2 film, the photoresponse of MSPhC near the bandgap edge of TiO2 is lower than the flat chip device. However, it shows a higher photoresponse at sub-bandgap photoelectric conversion efficiency, especially at 635 nm. From the measurement, we also demonstrate the influence of Au thickness on the photoelectric conversion efficiency. As shown in Fig. 8, with the decrease in Au thickness, the photocurrent increases. This results agrees with our simulation that thin Au film will improve the optical absorption, which can further increase the photoresponse.
Due to the quality of deposited TiO2, the highest photoresponse was got on a device of 20 nm Au, 75 nm TiO2 and without ITO contact layer. We measured the high resolution of photoresponse of this device with a 300 W Xenon arc lamp source monochromated by a holographic diffraction grating from 400 nm to 800 nm. The photoresponse normalized against the peak value located at 590 nm is shown in Fig. 9. The result shows a strong sub-bandgap photoresponse, with a broad full-width at half-maximum (FWHM) of 235 nm. However, the high series resistance introduced by the structured device lowers the overall measured photocurrent. The efficiency of the device could be improved by increasing the conductivity of the ITO back contact layer. In order to understand this wavelength dependent photoresponse enhancement, we further compare our result with Fowler’s law [21, 22]. According to Fowler’s law, the wavelength dependence internal photoemission at flat interface between a metal and semiconductor is following the rule of:
Whereis photoresponse, is a constant, is the Schottky barrier between the metal and semiconductor (Approximately 1.1 eV for Au and TiO2 [9, 23]). We fitted Fowler’s law to our result, as shown in Fig. 9. The blue fitting line has a similar shape as the flat chip measurement (green line) in Fig. 8. The deviation of photoresponse of MSPhC from Fowler’s law at the range of 440 nm to 780 nm suggests an enhanced sub-bandgap hot electron generation and injection due to the nano-cavity array structure.
The peak of photoresponse at 590 nm in Fig. 9 matches the SPP resonance mode in modal analysis in Fig. 6. In SPP mode, the coupling between conduction electrons and interfacial electric field oscillating will result in the generation of hot electron-hole pairs at the interface [7, 24–26]. The enhancement of electric field oscillation at surface plasmon frequency can largely promote the collection of hot electrons [27–29]. Experimental results shows that the coupling between SPP and LSPR in nanostructures will result in multiple bands of photoelectric hot electron collection [12, 14].
Even though we could achieve a broadband absorption from visible to IR spectrum, no photoresponse peak at wavelengths other than 590 nm implies that waveguide and cavity modes may not contribute to the hot electron generation. It appears likely that the enhancement of hot electron generation and collection in MSPhC only results from SPP absorption at the interface between Au and TiO2. Reported experimental results have only indicates that SPP and LSPR can improve the photoelectric conversion efficiency . Recent theoretical analysis further implies that, with the confinement effect introduced by metallic nanostructures, the photoexcitation of hot electrons in the metal near the metal/semiconductor interface dominates the photocurrent [6, 27]. Surface plasmon can significantly promotes hot electron generation by enhancing the electric field intensity in the metallic nanostructure [7, 30]. As shown in Fig. 7, only SPP mode generates electric field oscillation at the Au/TiO2 interface and enhances electric field intensity in the Au layer. Therefore, the photoresponse shows only a single enhancement peak at the plasmon frequency. Besides, in Fowler’s law, the generated hot electrons have an isotropic and uniform momentum distribution. Only electrons within the “escape cone” could be injected [4, 22]. The enhancement peak at 590 nm suggests a non-isotropic momentum distribution of hot electrons generatd by SPP mode. It has been approved that this effect could improve hot electron collection [7, 31, 32].
In our design, we seek to maximize absorption by generating different resonance modes in the structure. However, the results shows that only SPP modes directly contribute to the generation of hot electrons. This feature indicates that multiple band SPP absorption may be the key to promote photoelectric conversion at difference wavelength. Previous experimental results show that randomly patterned gold nano-island and nanoparticles on TiO2 could stimulate LSPR and enhance hot electron collection [23, 33]. In order to achieve multiple band SPP absorption in MSPhC, we deposited gold nanorods with diameter of 15 nm and length of 45 nm on MSPhC through electrophoretic deposition . Due to the shape of gold nanorods, transverse and longitudinal plasmon resonance could be stimulated at different wavelength . As shown in Fig. 10(a), the gold nanorods formed a random distribution on MSPhC with their longitudinal axis parallel to the metal/semiconductor interface. Figure 10(b) shows the UV-Vis absorption spectra of gold nanorods on a flat glass slide. The two peaks are corresponding to the transverse and longitudinal surface plasmon absorption at near 500 nm and 700 nm. In future work, we expect to control the surface plasmon mode frequencies in the photocurrent injection and to achieve a multiple peak photoelectric conversion by depositing gold nanorods on MSPhC.
The efficiency of plasmon assisted hot electron collection also depends on other factors. Injection rate of hot electrons before carrier recombination limits the efficiency . The electron scattering during hot electron transfer will cause the decrease in the amount of hot electrons that could reach the metal/dielectric interface. The momentum and energy distribution of generated hot electrons also determines the injection efficiency. Theoretical calculation results indicates that  due to the confinement effect introduced by nanoscale structure, hot electrons may not distribute uniformly and isotropicly in the momentum space. Only hot electrons locate in a momentum cone of Schottky barrier have the chance to be injected . Currently, work is underway to investigate the generation and injection of hot electrons through plasmon decay in gold nanorods.
In this paper, we present the design of a wafer-scale Au/TiO2 metallic-semiconductor photonic crystal (MSPhC) device for photoelectric hot electron generation and collection. Through FDTD simulation, optical absorption property was optimized by tuning structural parameters. Simulation results indicates that a thinner Au layer will result in higher absorption due to the coupling of incident with the MSPhC structure. The nano-cavity structure increases the light-metal interaction time, and up to 70% of solar spectrum could be absorbed in simulation of the MSPhC. Through microfabrication process, MSPhC devices were fabricated on 6” Silicon wafer, which shows a potential in large-scale and low-cost fabrication. The measured reflection spectrum of MSPhC matches well with simulation results for incident under 1.5 µm, suggesting a broadband absorption spectrum. Short-circuit photocurrent test indicates hot electron generation and injection below the bandgap of TiO2, with a peak at 590 nm due to SPP mode at the interface between Au and TiO2. Modal analysis by FDTD simulation confirms that only SPP absorption mode contribute significantly to hot electron collection, and other modes such as cavity and waveguide modes do not enhance or couple to hot electron generation in Au. Due to the non-ideal quality of deposited TiO2, the high resistance of MSPhC limits the photoelectric conversion efficiency. Normalized by the device resistance, photocurrent enhancement of up to 12 times is achieved on MSPhC at sub-bandgap range, compared with a flat chip device. To fulfill broadband hot electron collection, future work will focus on understanding the generation and injection of hot electrons in metal nanostructures and utilizing metal nanorods to stimulate multi-band SPP and LSPR in MSPhC.
References and links
2. W.-J. Yin, H. Tang, S.-H. Wei, M. M. Al-Jassim, J. Turner, and Y. Yan, “Band structure engineering of semiconductors for enhanced photoelectrochemical water splitting: the case of TiO 2,” Phys. Rev. B Condens. Matter 82(4), 045106 (2010). [CrossRef]
3. J. Nowotny, Oxide Semiconductors for Solar Energy Conversion: Titanium Dioxide (CRC, 2011).
4. C. Scales and P. Berini, “Thin-film schottky barrier photodetector models,” IEEE J. Quantum Electron. 46(5), 633–643 (2010). [CrossRef]
6. A. M. Brown, R. Sundararaman, P. Narang, W. A. Goddard, and H. A. Atwater, “Nonradiative plasmon decay and hot carrier dynamics: effects of phonons, surfaces, and geometry,” ACS Nano 10(1), 957–966 (2016). [PubMed]
7. A. O. Govorov, H. Zhang, and Y. K. Gun’ko, “Theory of photoinjection of hot plasmonic carriers from metal nanostructures into semiconductors and surface molecules,” J. Phys. Chem. C 117(32), 16616–16631 (2013). [CrossRef]
8. S. A. Maier, Plasmonics: Fundamentals and Applications (Springer Science and Business Media, 2007).
9. C. Clavero, “Plasmon-induced hot-electron generation at nanoparticle/metal-oxide interfaces for photovoltaic and photocatalytic devices,” Nat. Photonics 8(2), 95–103 (2014). [CrossRef]
10. Y. Nishijima, K. Ueno, Y. Yokota, K. Murakoshi, and H. Misawa, “Plasmon-assisted photocurrent generation from visible to near-infrared wavelength using a Au-nanorods/TiO2 electrode,” J. Phys. Chem. Lett. 1(13), 2031–2036 (2010). [CrossRef]
12. F. P. García de Arquer, A. Mihi, and G. Konstantatos, “Large-Area Plasmonic-Crystal–Hot-Electron-Based Photodetectors,” ACS Photonics 2(7), 950–957 (2015). [CrossRef]
15. J. B. Chou, Y. X. Yeng, Y. E. Lee, A. Lenert, V. Rinnerbauer, I. Celanovic, M. Soljačić, N. X. Fang, E. N. Wang, and S. G. Kim, “Enabling ideal selective solar absorption with 2D metallic dielectric photonic crystals,” Adv. Mater. 26(47), 8041–8045 (2014). [CrossRef] [PubMed]
16. J. B. Chou, Y. X. Yeng, A. Lenert, V. Rinnerbauer, I. Celanovic, M. Soljačić, E. N. Wang, and S.-G. Kim, “Design of wide-angle selective absorbers/emitters with dielectric filled metallic photonic crystals for energy applications,” Opt. Express 22(S1Suppl 1), A144–A154 (2014). [CrossRef] [PubMed]
17. J. B. Chou, D. P. Fenning, Y. Wang, M. A. M. Polanco, J. Hwang, A. El-Faer, F. Sammoura, J. Viegas, M. Rasras, and A. M. Kolpak, “Broadband photoelectric hot carrier collection with wafer-scale metallic-semiconductor photonic crystals,” in 2015 IEEE 42nd Photovoltaic Specialist Conference (PVSC) (IEEE, 2015), pp. 1–6. [CrossRef]
20. F. Lemarchand, A. Sentenac, E. Cambril, and H. Giovannini, “Study of the resonant behaviour of waveguide gratings: increasing the angular tolerance of guided-mode filters,” J. Opt. A, Pure Appl. Opt. 1(4), 545–551 (1999). [CrossRef]
22. R. H. Fowler, “The analysis of photoelectric sensitivity curves for clean metals at various temperatures,” Phys. Rev. 38(1), 45–56 (1931). [CrossRef]
23. Y. K. Lee, C. H. Jung, J. Park, H. Seo, G. A. Somorjai, and J. Y. Park, “Surface plasmon-driven hot electron flow probed with metal-semiconductor nanodiodes,” Nano Lett. 11(10), 4251–4255 (2011). [CrossRef] [PubMed]
24. M. Bernardi, J. Mustafa, J. B. Neaton, and S. G. Louie, “Theory and computation of hot carriers generated by surface plasmon polaritons in noble metals,” Nat. Commun. 6, 7044 (2015). [CrossRef] [PubMed]
25. R. Sundararaman, P. Narang, A. S. Jermyn, W. A. Goddard 3rd, and H. A. Atwater, “Theoretical predictions for hot-carrier generation from surface plasmon decay,” Nat. Commun. 5, 5788 (2014). [CrossRef] [PubMed]
27. A. V. Uskov, I. E. Protsenko, R. S. Ikhsanov, V. E. Babicheva, S. V. Zhukovsky, A. V. Lavrinenko, E. P. O’Reilly, and H. Xu, “Internal photoemission from plasmonic nanoparticles: comparison between surface and volume photoelectric effects,” Nanoscale 6(9), 4716–4727 (2014). [CrossRef] [PubMed]
28. D. Jin, Q. Hu, D. Neuhauser, F. von Cube, Y. Yang, R. Sachan, T. S. Luk, D. C. Bell, and N. X. Fang, “Quantum-spillover-enhanced surface-plasmonic absorption at the interface of silver and high-index dielectrics,” Phys. Rev. Lett. 115(19), 193901 (2015). [CrossRef] [PubMed]
30. B. Y. Zheng, H. Zhao, A. Manjavacas, M. McClain, P. Nordlander, and N. J. Halas, “Distinguishing between plasmon-induced and photoexcited carriers in a device geometry,” Nat. Commun. 6, 7797 (2015). [CrossRef] [PubMed]
31. W. Li, Z. J. Coppens, L. V. Besteiro, W. Wang, A. O. Govorov, and J. Valentine, “Circularly polarized light detection with hot electrons in chiral plasmonic metamaterials,” Nat. Commun. 6, 8379 (2015). [CrossRef] [PubMed]
32. M. W. Knight, Y. Wang, A. S. Urban, A. Sobhani, B. Y. Zheng, P. Nordlander, and N. J. Halas, “Embedding plasmonic nanostructure diodes enhances hot electron emission,” Nano Lett. 13(4), 1687–1692 (2013). [CrossRef] [PubMed]
33. Y. Tian and T. Tatsuma, “Mechanisms and applications of plasmon-induced charge separation at TiO2 films loaded with gold nanoparticles,” J. Am. Chem. Soc. 127(20), 7632–7637 (2005). [CrossRef] [PubMed]
34. A. Elfaer, Y. Wang, X. Li, J. Chou, and S. Kim, “Gold nanorods coated metallic photonic crystal for enhanced hot electron transfer in electrochemical cells,” MRS Adv. 1(13), 831–837 (2016). [CrossRef]