Introduction
Few inventions have been as ubiquitous as the incandescent light bulb. Edison’s 1880 patent was born from over 40 years of effort by multiple researchers to optimize the filament materials, the bulb atmosphere, and the socket interface. The invention revolutionized society, bringing light into homes and the workplace, lengthening work and leisure time, and laying the foundation for an interconnected world. It is a technology that underscores the importance of light and light-based technologies to our world and highlights the crucial role of materials research in the optical sciences. Today, efforts continue to make lighting technologies brighter and more energy efficient.
Controlling light extends far beyond the lighting industry. It underpins the Internet, which relies on optical signal propagation, modulation, and detection for fast and long-ranging data communication. It enables new imaging technologies, from digital single-lens reflex cameras for photography enthusiasts to advanced biomedical endoscopy and microscopy. It is foundational to solar energy generation and storage—from natural photosynthesis to engineered photovoltaic and photocatalytic cells.
Whereas the diffraction limit would seemingly prohibit the development of desktop optical computers, flat camera zoom lenses, nanoscale endoscopes, or ultra-efficient nanostructured solar cells, plasmons offer a potentially transformative way to control light and enable new light-based technologies. Because plasmons can tailor light propagation from the molecular to the macroscale, they promise almost complete control of photons into and out of materials and devices. To highlight just a few exciting applications, plasmons have enabled brighter light-emitting diodes, smartphone-compatible molecular sensors, photothermal cancer therapies, solar-driven water purification, nanoscale lithography, subwavelength lasers, and high-density data storage and holographic displays.
Plasmonic technologies might be as revolutionary to nanoscale optics as the light bulb was for large-scale lighting. However, as for the light bulb, considerable research remains to optimize materials and performance before plasmonic devices become widespread. This article highlights the most promising societal and industrial applications of plasmons, identifies challenges in their industrial integration, and proposes potential solutions toward technology transfer. We focus on the energy, biomedical, and information-technology sectors, as these sectors not only represent some of the largest industries today, but are also industries most likely to be affected by plasmonic technologies. We speculate that the field is poised for significant industrial impact, provided that a few key materials and engineering challenges can be overcome.
Plasmonic nanoparticles: Light concentrators and converters
Plasmons are quanta of collective oscillations of free electrons. They exist in conducting materials, including metals and doped semiconductors, and can be excited by light or electrons. Whether persisting in nanoparticles, in nanowires, or on the surfaces of films, plasmons can concentrate incident radiation to subwavelength volumes. We focus in this article on resonant plasmonic materials—and, in particular, on nanoparticles—to complement a number of outstanding articles and texts discussing the properties and applications of plasmons.1–7 As illustrated in Figure 1 , the excitation of localized plasmon resonances gives rise to intense near fields that are sensitive to changes in the charge density and refractive index of the particle or embedding medium. Plasmonic nanoparticles can therefore be used to enhance photon absorption in adjacent materials (Figure 1a) or to accurately sense changes in the environment at the nano- and molecular-scales (Figure 1b). The locally elevated temperatures of nanoparticles can also be used in photothermally driven processes, including catalysis, medical treatments, and heat-assisted data recording (Figure 1c). Furthermore, plasmon resonances can quickly decay into hot carriers, with energies much higher than the ambient temperature provides. These carriers can subsequently be harvested and used to enable new photochemical reactions and improved energy conversion for photovoltaic and solar fuel generation (Figure 1d). Reference Brongersma, Halas and Nordlander8–Reference Manjavacas, Liu, Kulkarni and Nordlander11
Figure 1. Four modalities of localized surface plasmon resonances (LSPRs). (a) The intense electric fields at the surface of an illuminated plasmonic nanoparticle can increase the absorption cross section of an adjacent semiconductor or dielectric. (b) LSPRs are extremely sensitive to small changes in the dielectric environment, allowing use of small shifts in their spectral positions (denoted as Δλ) to detect processes such as protein binding and ion intercalation. (c) Surface plasmons can decay nonradiatively, producing heat at the surface of a nanoparticle. (d) Surface plasmons can decay into “hot” electron–hole pairs that can be harvested to conduct nonequilibrium chemical processes at the surface of the plasmonic nanoparticles, opening new avenues for heterogeneous catalysis.
Apart from their versatile physical properties, plasmonic nanoparticles can be readily synthesized in large quantities and with very high quality using colloidal techniques. Skilled chemists can precisely control nanoparticle size, shape, and morphology, as well as ligand coating and molecular functionalization. The bottom-up synthesis of nanoparticles is routine in laboratory settings and forms the basis for a number of current companies, including nanoComposix, Nanopartz, SkySpring Nanomaterials, Sienna Labs, and LamdaGen. The relevance of nanoparticles is further highlighted by their inclusion in the catalogues of many major chemical suppliers such as Sigma-Aldrich, Alfa Aesar, and Perkin-Elmer. The well-controlled, cost-effective colloidal synthesis of nanoparticles makes them particularly suited for industrial-scale technologies.
Plasmonic nanoparticles in energy conversion
Catalysis and chemical energy conversion
It is estimated that catalysis contributes to nearly 35% of the global gross domestic product. Reference Armor12 Improving catalytic conversion rates is therefore among the most important industrial pursuits and the subject of intense investigation in the plasmonics community. Recently, visible-light excitation of reaction mixtures containing plasmonic nanoparticles was shown to drive catalytic conversions at lower temperatures and even enable reactions that cannot be activated using conventional thermal processes. Reference Qiu and Wei13,Reference Christopher, Xin, Marimuthu and Linic14 For example, Linic and colleagues demonstrated that silver nanoparticles could help drive a variety of commercially important oxidation reactions, including ethylene epoxidation, CO oxidation, and NH3 oxidation upon illumination with low-intensity visible photons. Reference Christopher, Xin and Linic15 The energetic electrons formed through excitation of surface plasmons were transferred to absorbed molecular O2, inducing nuclear oxygen-atom vibration and allowing activation of the oxygen–oxygen bond at reduced temperatures. A related study indicated that plasmonic hot carriers from gold nanoparticles can dissociate H2 molecules. Reference Mukherjee, Libisch, Large, Neumann, Brown, Cheng, Lassiter, Carter, Nordlander and Halas16 Further, the excitation of localized plasmon resonances on copper nanoparticles was used to reduce their surface oxide layer and increase their selectivity toward the epoxidation of propylene. Reference Marimuthu, Zhang and Linic17 Together, these studies demonstrate that plasmonic nanoparticles are a new family of photocatalysts that not only reduce the energy budget of reactions, but also promise to enhance catalyst stability and product selectivity.
Beyond heterogeneous catalysis, the possibility of harvesting hot carriers in metal–semiconductor nanoparticle junctions is exciting for the conversion of sunlight into electricity and solar fuels. In analogy to the organic dyes used in dye-sensitized solar cells, metal nanoparticles can be used as efficient, tunable, and photochemically stable light absorbers in conjunction with a semiconductor. This effect was first demonstrated by Tian and Tatsuma, who reported an incident-photon-to-current conversion efficiency of 26% for a composite of porous TiO2 and gold nanoparticles. Reference Tian and Tatsuma18,Reference Tian and Tatsuma19 Recently, Mubeen et al. demonstrated a water-splitting device consisting of gold nanorods coated with TiO2 and decorated with cobalt-based oxygen-evolving catalysts. Importantly, almost all charge carriers used in the water-splitting reaction came from hot carriers generated by plasmon resonance excitation. Reference Mubeen, Lee, Singh, Krämer, Stucky and Moskovits20
Apart from supporting hot-carrier generation and injection, the intense, localized electric fields of plasmonic nanostructures can also increase the absorption efficiency of semiconducting photocatalysts. This enhancement allows for the use of thinner absorbing layers, potentially reducing costs and mitigating any inherent material shortcomings, such as short exciton diffusion lengths or sluggish charge transport. For example, the introduction of gold nanoparticles embedded in or on top of hematite (Fe2O3) has been shown to enhance water oxidation efficiency by over 7% under AM1.5 simulated solar illumination. Reference Thimsen, Formal, Grätzel and Warren21,Reference Thomann, Pinaud, Chen, Clemens, Jaramillo and Brongersma22 As illustrated in Figure 2a, Reference Brown, Suteewong, Santosh Kumar, D’Innocenzo, Petrozza, Lee, Wiesner and Snaith23,Reference Ferry, Verschuuren, Li, Verhagen, Walters, Schropp, Atwater and Polman24 such enhancements arise from increased absorption in the active layer, resulting in increased photocurrent. In another study, doped titania (TiO2), which absorbs only weakly in the visible spectrum, was coated with silver nanospheres, resulting in a tenfold enhancement in photocurrent generation under broadband, visible irradiation. Reference Ingram and Linic25
Figure 2. Plasmon-enhanced energy conversion. (a) Measured photocurrent-enhancement (red symbols) and simulated absorption-enhancement (solid blue line) spectra of a 100-nm thin Fe2O3 photoelectrode layer containing silica-coated gold nanoparticles (Au NPs). (b) Current–voltage characteristic for solar cells sensitized with only N719 ruthenium–organic dye (blue, open circles) and with the additional incorporation of Au@SiO2 (15-nm Au core with 3-nm SiO2 shell) nanoparticles under AM1.5 illumination. (c) Photograph of an array of 88 plasmonic amorphous silicon solar cells; each colored square is a separate cell with varying plasmonic particle diameter and interparticle separation. The inset shows an electron micrograph of the cross section of one solar cell. Note: a-Si:H, hydrogenated amorphous silicon; ITO, indium tin oxide; TCO, transparent conductive oxide; ZnO:Al, aluminum-doped zinc oxide. (a) Adapted with permission from Reference Reference Thomann, Pinaud, Chen, Clemens, Jaramillo and Brongersma22. © 2011 American Chemical Society. (b) Adapted with permission from Reference Reference Brown, Suteewong, Santosh Kumar, D’Innocenzo, Petrozza, Lee, Wiesner and Snaith23. © 2011 American Chemical Society. (c) Adapted with permission from Reference Reference Ferry, Verschuuren, Li, Verhagen, Walters, Schropp, Atwater and Polman24. © 2010 Optical Society of America.
Plasmon resonances have further enabled photothermal vapor generation at relatively low illumination intensities, including outdoor solar irradiation. For example, aqueous core–shell Au@SiO2 nanoparticles (consisting of a silica core encapsulated in gold) exposed to concentrated ambient daylight undergo a nonequilibrium process to form and release high-temperature steam within seconds of illumination. Reference Neumann, Urban, Day, Lal, Nordlander and Halas26 This vaporization phenomenon could potentially be employed in water purification and instrument sterilization in remote locations. Reference Neumann, Feronti, Neumann, Dong, Schell, Lu, Kim, Quinn, Thompson, Grady, Nordlander, Oden and Halas27 In addition, this process can achieve improved distillation ability to generate 99% ethanol vapor from a water–vapor mixture, surpassing the 95% azeotropic concentration that conventional techniques cannot exceed. Reference Neumann, Urban, Day, Lal, Nordlander and Halas26
Plasmonic photovoltaics
Plasmonic light concentration and redirection have shown potential to improve the performance of thin-film photovoltaics, which have historically been limited by light absorption in the active layer (see Figure 2b–c). Reference Ferry, Verschuuren, Li, Verhagen, Walters, Schropp, Atwater and Polman24,Reference Ferry, Munday and Atwater28 Three primary geometries have been proposed for plasmonic-structure inclusion: (1) deposition of nanostructures at the top surface that preferentially scatter light into the solar cell, (2) patterning of the back electrode to redirect the incident light and create propagating waveguide modes in the absorbing layer, and (3) embedding nanoparticles within or near the active layer to make use of the generated high-intensity near fields.
These techniques have been used to improve the performance of thin-film silicon, Reference Pala, White, Barnard, Liu and Brongersma29 organic, Reference Gan, Bartoli and Kafafi30 and dye-sensitized Reference Brown, Suteewong, Santosh Kumar, D’Innocenzo, Petrozza, Lee, Wiesner and Snaith23 cells, enabling greater light absorption in the thin active layer and increasing the short-circuit current. For example, plasmonic inclusions can enable a 30-fold reduction in silicon wafer thickness while maintaining 85% of the original efficiency. Reference Zhang, Stokes, Jia, Fan and Gu31
Plasmonic materials can additionally be used in the creation of next-generation electrodes to replace the traditional transparent conducting oxides. Solution-processed silver nanowires have been shown to form fused, interconnected networks through light-induced plasmonic nanowelding from localized field concentration and heating. Reference Garnett, Cai, Cha, Mahmood, Connor, Christoforo, Cui, McGehee and Brongersma32 This electrode fabrication strategy requires only a tungsten-halogen lamp white-light source (30 W/cm2), avoiding the high energy requirements of full-system, high-temperature baking.
Plasmonic nanoparticles in health and medicine
Photothermal therapy
Light-controlled localized heating of nanoparticles has been used for a number of biomedical applications, including targeted tumor ablation, Reference Baffou and Quidant33 targeted drug delivery, Reference Pissuwan, Niidome and Cortie34 selective bacterial killing, Reference Zharov, Mercer, Galitovskaya and Smeltzer35 and single-cell nanosurgery. Reference Urban, Pfeiffer, Fedoruk, Lutich and Feldmann36
In the context of tumor ablation, plasmonic nanoparticles can be used to locally heat malignant cells without damaging the surrounding healthy tissues, as seen in Figure 3 . Reference Stern, Stanfield, Kabbani, Hsieh and Cadeddu37,Reference Paasonena, Sipilä, Subrizi, Laurinmäki, Butcher, Rappolt, Yaghmur, Urtti and Yliperttula38 Tumors typically exhibit unusually porous blood vessels, which allows nanoparticles to passively accumulate in the cancerous region; once there, they remain lodged because of the diseased region’s diminished lymphatic drainage. Alternatively, the surface of the nanoparticles can be functionalized with ligands that bind to receptors that are more commonly found on cancer cells, such as an epidermal growth factor receptor. Reference Qian, Peng, Ansari, Yin-Goen, Chen, Shin, Yang, Young, Wang and Nie39 Nanoshells Reference Bardhan, Lal, Joshi and Halas40 and nanorods Reference Huang, El-Sayed, Qian and El-Sayed41 have been extensively explored, in part because their near-infrared resonances match the transparency window of biological tissues. Reference Wang, Black, Luehmann, Li, Zhang, Cai, Wan, Liu, Li, Kim, Li, Wang, Liu and Xia42
Figure 3. Cancer therapy and molecule release using plasmon resonances. (a) Top: Schematic of cancerous tumor ablation. Plasmonic nanoparticles (gray spheres) are localized in cancerous tissue. Upon irradiation, the nanoparticles release heat to the surrounding tissue, leading to cell death. Bottom: Photographs taken before and after photothermal ablation of a subcutaneous tumor, assisted by plasmon resonances in 110-nm gold nanoshells. The tumor was irradiated with a 810-nm laser at 4 W/cm2 for 3 min. (b) Top: Schematic showing thermally induced release of small molecules from a liposome. Plasmonic nanoparticles (gray spheres) are embedded in the phospholipid membrane of the liposome. Upon irradiation, the nanoparticles produce heat, inducing a phase transition in the membrane that increases its permeability; small molecules (such as drugs or, here, the fluorescent molecule calcein) can then pass through the membrane. Bottom: UV-light-induced calcein release from liposomes at constant temperature (37°C) with nanogold-loaded nanoparticles. (a) Photographs reproduced with permission from Reference Reference Stern, Stanfield, Kabbani, Hsieh and Cadeddu37. © 2008 Elsevier. (b) Reproduced with permission from Reference Reference Paasonena, Sipilä, Subrizi, Laurinmäki, Butcher, Rappolt, Yaghmur, Urtti and Yliperttula38. © 2010 Elsevier.
Photothermal effects can also be used to mechanically ablate cancer cells. For example, Wagner et al. used local heating of gold nanoparticles to generate vapor nanobubbles that expand and collapse within nanoseconds, creating a localized mechanical impact that leads to cell death. Reference Wagner, Delk, Lukianova-Hleb, Hafner, Farach-Carson and Lapotko43 Further, plasmonic resonances can be used to enhance the efficiency of photodynamic therapy, in which cell death is triggered by the release of cytotoxic reactive oxygen species by photosensitizer molecules. Reference Ogilby44
Beyond cancer therapeutics, photothermal heating can be used for controlled, local dosing, as illustrated in Figure 3b. Plasmonic nanoparticles have been used to trigger the release of drugs from various materials, including liposomes Reference Paasonen, Laaksonen, Johans, Yliperttula, Kontturi and Urtti45 and hydrogels, Reference Sershen, Westcott, Halas and West46 as well as to transport and release DNA for targeted gene therapy. Reference Takahashi, Niidome and Yamada47 Combined diagnostics and therapeutics is made possible by using nanoparticles as contrast agents in imaging, including magnetic resonance imaging, Reference Larson, Bankson, Aaron and Sokolov48 optical coherence tomography, Reference Gobin, Lee, Halas, James, Drezek and West49 and photoacoustic tomographic imaging. Reference Jin, Jia, Huang, O’Donnell and Gao50
Biosensing
Plasmonic nanostructures can serve as ultraresponsive sensors, revealing slight changes in their own electronic structure, density, geometry, and temperature, or that of their surroundings, through the energies and widths of their plasmonic resonance peaks. Reference Larsson, Syrenova and Langhammer51,Reference Langhammer, Larsson, Kasemo and Zoric52 This property can be used for biosensing, and indeed, one of the first commercial applications of plasmonics was a home pregnancy test for detecting elevated concentrations of human chorionic gonadotropin hormone. Reference Stockman53
Recently, large-area plasmonic sensors have been used to detect exosomes (vesicles released by cells into bodily fluids) for the diagnosis of some cancers, as shown in Figure 4a. Reference Im, Shao, Park, Peterson, Castro, Weissleder and Lee54,Reference Ayas, Cupallari, Ekiz, Kaya and Dana55 Additionally, solution-phase plasmonic sensors have been developed for the label-free detection of protein binding in real time. Reference Wu, Henzie, Lin, Rhodes, Li, Sartorel, Thorner, Yang and Groves56 Most recently, plasmonic sensing has achieved monolayer protein sensitivity. For example, Altug and co-workers reported a hand-held device incorporating plasmonic arrays that sense binding of a monolayer of the bacteria-derived fusion protein A/G with immunoglobulin G (IgG) antibodies. Reference Cetin, Coskun, Galarreta, Huang, Herman, Ozcan and Altug57 The hand-held sensor could be integrated with cell phones, as illustrated in Figure 4b, providing high-throughput and low-cost refractive-index sensing of bacteria and proteins in blood or saliva to benefit medical diagnosis in underdeveloped areas. Reference Zhu, Sencan, Wong, Dimitrov, Tseng, Nagashima and Ozcan58
Figure 4. Plasmonic sensors. (a) Solid-state localized surface plasmon resonance-based sensor using gold hole arrays to detect binding of exosomes (vesicles released by cells). Left: The arrays are functionalized with affinity ligands for different exosomal protein markers. Upon binding, spectral shifts or intensity changes proportional to the levels of target marker proteins are observed. Right: A representative schematic of changes in transmission spectra showing exosome detection. Compared to conventional methods, this technology offers sensitive and label-free exosome analyses and enables continuous, real-time monitoring of molecular binding. (b) Hand-held surface-enhanced Raman spectrometer. This particular sensor is based on large-area silver nanoparticle films that can, in principle, detect single-molecule motion. Raman spectra are collected on the smartphone by converting the camera into a low-resolution spectrometer through the inclusion of a collimator and a grating. Note: PEG, poly(ethylene glycol). (a) Adapted with permission from Reference Reference Im, Shao, Park, Peterson, Castro, Weissleder and Lee54. © 2014 Nature Publishing Group. (b) Adapted with permission from Reference Reference Ayas, Cupallari, Ekiz, Kaya and Dana55. © 2014 American Chemical Society.
Rather than relying on specific binding interactions (e.g., with antibodies), chemical identification can also be achieved using mid-infrared molecular vibrational resonances, which serve as “molecular fingerprints.” Surface-enhanced infrared absorption spectroscopy of octadecanethiol and hemoglobin recently reached detection sensitivities in the attomolar and zeptomolar ranges, respectively. Reference Brown, Zhao, King, Sobhani, Nordlander and Halas59 Likewise, surface-enhanced Raman spectroscopy (SERS) using plasmonic substrates has achieved signal enhancements of up to seven orders of magnitude. Hand-held SERS systems Reference Ayas, Cupallari, Ekiz, Kaya and Dana55 have been demonstrated for intraoperative detection of malignant tumors, Reference Mohs, Mancini, Singhal, Provenzale, Leyland-Jones, Wang and Nie60 detection of food-contamination-related molecules, Reference Luo, Sivashanmugan, Liao, Yao and Peng61 and sensing bacteria with nanomolar sensitivity. Reference Cowcher, Xu and Goodacre62 The combination of a gold-nanoparticle SERS platform with an exponential amplification reaction enabled detection of microRNAs in lung cancer cells with a sensitivity to 0.5 fM. Reference Ye, Hu, Liang and Zhang63 This technique therefore provided a six-order-of-magnitude improvement in sensitivity compared to existing SERS-based direct assays and a nine-order-of-magnitude improvement compared to northern blot methods for microRNA detection. Platforms need not be solid-state, and liquid-state substrate-free Raman spectroscopy is an emerging technology. Reference Kim, Han, Choi, Lee, Hong, Suh, Lee and Kang64
By using the strongly enhanced field between two metals, tip-enhanced Raman spectroscopy (TERS) promises sub-1-nm spectral mapping. Reference Zhang, Zhang, Dong, Jiang, Zhang, Chen, Zhang, Liao, Aizpurua, Luo, Yang and Hou65 At low temperature and high vacuum, the tip-to-surface distance is precisely controlled to match the nanocavity plasmon resonance with the energy of molecular vibrations, which significantly enhances the Raman signal. This remarkable technique has achieved single-molecule resolution with combined subnanometer imaging. In more recent studies, Belkin and co-workers have adapted TERS to also sense the mechanical force that molecular vibrations exert on the tip. Reference Lu, Jin and Belkin66 This technique circumvents the need for mid-infrared detectors and promises sensitivity down to a few tens of molecules.
Plasmonic materials for information technology
Modulators and lasers
Optical components promise faster, smarter, and more energy-efficient information technologies. In 2012, IBM announced a technology breakthrough with monolithic integration of optical modulators, photodetectors, and multiplexers into a 90-nm-base high-performance computing chip. Reference Assefa, Shank, Green, Khater, Kiewra, Reinholm, Kamlapurkar, Rylyakov, Schow, Horst, Pan, Topuria, Rice, Gill, Rosenberg, Barwicz, Yang, Proesel, Hofrichter, Offrein, Gu, Haensch, Ellis-Monaghan and Vlasov67 However, integration of electronics and photonics requires further miniaturization of optical components, efficient electrical or optical tuning of light propagation, dynamic access to light sources, and isolation of optical signals analogous to that in electronics. Reference Davoyan and Engheta68 One possible integrated optoelectronic circuit design is illustrated in Figure 5a, including an artist’s depiction of nanophotonic sources, waveguides, modulators, diodes, filters, and photodetectors. Reference Dionne, Sweatlock, Sheldon, Alivisatos and Atwater69 Localized surface plasmons are critical to many of these components.
Figure 5. Plasmonics for computation and communications. (a) Schematic of a photonic and plasmonic circuit, consisting of (0) light-incoupling structures, (1) color demultiplexing in a “Z” add/drop filter, (2) bends and tapers in nanophotonic waveguides, (3) all-optical preprocessing logic, (4) integrated photodetection, (5) optical clock, (6) nano-optical subcircuit (on-chip nanoscale laser, spaser, or light-emitting diode; plasmonic modulators and diodes; and integrated photodetection), (7) collection of light by “photon sorting,” and (8) integrated plasmonic filtering and beam shaping. (b) Photograph of a plasmonic complementary metal–oxide–semiconductor (CMOS) image sensor. (c) Photograph taken with the CMOS sensor, demonstrating full-color, high-resolution plasmonic imaging. (a) Reproduced with permission from Reference Reference Dionne, Sweatlock, Sheldon, Alivisatos and Atwater69. © 2010 Institute of Electrical and Electronics Engineers. (b–c) Reproduced with permission from Reference Reference Burgos, Yokogawa and Atwater92. © 2013 American Chemical Society.
Nanophotonic modulators based on plasmon resonances can switch optical signals with high modulation ratios and subwavelength device footprints, even when traditionally weak nonlinear media are used. For example, Dionne et al. demonstrated a silicon-based plasmonic modulator with a 10 dB/2 μm extinction ratio; Reference Dionne, Diest, Sweatlock and Atwater70 Volker et al. demonstrated an indium tin oxide modulator with a 1 dB/1 μm extinction ratio; Reference Volker, Lanzillotti-Kimura, Ren-Min and Xiang71 and Zhao and Lu demonstrated a silicon nitride modulator with a 7.7 dB/400 nm extinction ratio. Reference Zhao and Lu72 Submicron-scale electro-optic modulators have also been demonstrated using the metal–insulator phase transition in vanadium oxide; Reference Sweatlock and Diest73,Reference Kruger, Joushaghani and Poon74 the ferroelectric transition in bismuth ferrite; Reference Babicheva, Zhukovsky and Lavrinenko75 and the tunability of novel materials such as graphene, Reference Gosciniak and Tan76,Reference Bao and Loh77 nonlinear polymers, Reference Melikyan, Alloatti, Muslija, Hillerkuss, Schindler, Li, Palmer, Korn, Muehlbrandt, Van Thourhout, Chen, Dinu, Sommer, Koos, Kohl, Freude and Leuthold78 and thermo-optic polymers. Reference Gosciniak, Bozhevolnyi, Andersen, Volkov, Kjelstrup-Hansen, Markey and Dereux79
In parallel, localized plasmon resonances have enabled coherent light generation in submicron structures, with on-chip footprints that are comparable to those of electronic devices. Reference Hill80–Reference Walther, Scalari, Amanti, Beck and Faist86 For example, thresholdless continuous-wave lasing has been achieved using nanoscale plasmonic coaxial cavities, Reference Khajavikhan, Simic, Katz, Lee, Slutsky, Mizrahi, Lomakin and Fainman84 and ultrafast (800-fs) pulsed emission was observed from hybrid plasmon–semiconducting nanowire lasers. Reference Sidiropoulos, Röder, Geburt, Hess, Maier, Ronning and Oulton87 Further, spasers Reference Stockman88 promise to be next-generation sources of intense, localized, optical fields on the nanoscale. Unlike conventional lasers that emit photons, spasers are sources of coherent surface plasmons. Because the feedback necessary for stimulated emission is provided by surface plasmon modes instead of diffraction-limited photonic modes, the footprint can be just a few nanometers, that is, comparable to the size of the surface-plasmon wavelength.
Data storage and displays
An additional area of information processing facilitated by plasmons is data storage. For example, the intense near fields of plasmonic particles have been used in heat-assisted magnetic recording (HAMR), a magnetic data-storage technology for achieving extremely high data densities. HAMR relies on local heating of a single memory bit, with typical dimensions of 30 nm. By integrating plasmonic nanoparticles onto a magnetic write head, it is possible to address a single memory bit through photothermal effects and, hence, to write data with areal densities exceeding the 1Tb/in. Reference Willets and Van Duyne2 limit of conventional magnetic recording techniques. Reference Challener, Peng, Itagi, Karns, Peng, Peng, Yang, Zhu, Gokemeijer, Hsia, Ju, Rottmayer, Seigler and Gage89,Reference Zhou, Xu, Hammack, Stipe, Gao, Scholz and Gage90 Direct storage of information in photons has also been augmented by plasmonics. Whereas compact discs (CDs) store information in two-dimensional (2D) space using a light-sensitive glass, plasmons have been used to store optical bits in five-dimensional space: Reference Zijlstra, Chon and Gu91 three dimensions of space, one dimension of wavelength, and yet another dimension of polarization. In prototype demonstrations, plasmonic optical recording was achieved through light-dependent reshaping of gold nanorods suspended in a dielectric host. With sufficient input optical power, the rods are heated and change shape; this shape change stores information in a nonvolatile manner and is read out using low-intensity light, as with conventional CDs.
Mobile and static display technologies could also be revolutionized by plasmonics. For example, as seen in Figure 5, resonant plasmonic hole arrays have been integrated into 2D complementary metal oxide semiconductor (CMOS) image sensors, promising reduced fabrication complexity and cost compared to traditional dye-filter technologies (i.e., liquid-crystal displays). Reference Burgos, Yokogawa and Atwater92 Further, tunable plasmonic color filters with sizes of just a few hundred nanometers promise high-resolution displays and hyperspectral imaging. Reference Diest, Dionne, Spain and Atwater93,Reference Xu, Wu, Luo and Guo94
Beyond two dimensions, holography is an emerging technology for 3D displays. Unfortunately, rendering a 3D image with the same resolution users have come to expect from traditional 2D displays requires a monumental increase in pixel density; pixel size currently limits both the resolution and viewing angles of holograms. Recently, plasmonic metasurfaces created from arrays of gold nanorods were used to create large-field-of-view and high-density holograms. Each pixel was composed of a single 150-nm-long, 75-nm-wide nanorod, with the holographic information encoded in the nanorod’s orientation. Reference Huang, Chen, Mühlenbernd, Zhang, Chen, Bai, Tan, Jin, Cheah, Qiu, Li, Zentgraf and Zhang95 The phase interference of an incoming circularly polarized beam was used to create a 3D image (330 μm × 232 μm × 48.2 μm) across a broad wavelength range. Whereas this nanorod holographic display is passive, reconfigurable holograms are under active investigation.
Emerging optical information processing: Directional and quantum plasmonics
Directionality and isolation are critical features of integrated electronic circuits, but they are challenging to realize with optical devices. Reference Miller96 Recently, one-way transmission was demonstrated in nonlinear passive microrings Reference Fan, Wang, Varghese, Shen, Niu, Xuan, Weiner and Qi97 and in more exotic parity–time-symmetric microtoroids Reference Nazari, Bender, Ramezani, Moravvej-Farshi, Christodoulides and Kottos98 and optical waveguides. Reference Feng, Xu, Fegadolli, Lu, Oliveira, Almeida, Chen and Scherer99,Reference Feng, Ayache, Huang, Xu, Lu, Chen, Fainman and Scherer100 Plasmonic components can enable the same effects on smaller scales, by using highly localized magnetic circulating currents, Reference Davoyan and Engheta68 directional polarization conversion from two-layer metamaterials, Reference Naruse, Hori, Ishii, Drezet, Huant, Hoga, Ohyagi, Matsumoto, Tate and Ohtsu101 and plasmonic parity–time-symmetric components. Reference Alaeian and Dionne102–Reference Baum, Alaeian and Dionne104
Plasmonics has also had an impact on an emerging field of information processing: quantum computing. Single-photon sources, such as nitrogen-vacancy centers in diamond, are particularly well-suited for room-temperature quantum photonics, except for their low rates of photon emission. Plasmonics provides a means to enhance the emission of such single-photon sources without the use of bulky photonic cavities. Akimov et al. Reference Akimov, Mukherjee, Yu, Chang, Zibrov, Hemmer, Park and Lukin105 demonstrated enhanced emission from quantum dots coupled to silver nanowires, and Choy et al. Reference Choy, Hausmann, Babinec, Bulu, Khan, Maletinsky, Yacoby and Lončar106 extended the concept to diamond nitrogen-vacancy centers coupled to silver disks. Larger enhancements in emission rate can be achieved by using plasmonic hyperbolic metamaterials, which provide an enormous increase in the local density of states and, consequently, more decay channels. Reference Shalaginov, Vorobyov, Liu, Ferrera, Akimov, Lagutchev, Smolyaninov, Klimov, Irudayaraj, Kildishev, Boltasseva and Shalaev107 Emission-rate enhancements from plasmonic elements can also enhance detection rates—an effect that could improve the sensitivity of single-photon photodetectors.
Outlook
The field of plasmonics is undoubtedly poised to impact the energy, biomedical, and information-technology industries. Start-up companies based on new plasmonic materials, sensors, and probes are thriving. For example, Biacore offers label-free sensing platforms for the pharmaceutical, biotechnology, and diagnostic markets; Nano-Meta Technologies develops new HAMR and thermophotovoltaic technologies; and Nanospectra commercializes plasmonic-nanoparticle-based medical therapies and is currently conducting clinical trials on photothermal ablation of head, neck, and metastatic lung tumors.
For increased industrial applicability, there is strong impetus to solve challenges related to plasmonic materials, processing, and integration. In particular, choosing materials that are CMOS-compatible and exhibit lower loss will be important for many information-technology applications. Reference West, Ishii, Naik, Emani, Shalaev and Boltasseva108 High-temperature applications (e.g., in catalysis, thermophotovoltaics, photothermal treatments, or heat-assisted memory) might require plasmonic materials that are more thermally stable than noble metals. Moreover, fabrication methods beyond electron-beam lithography will be required for large-area, cost-effective device architectures. Plasmonic technologies will also have to confront the constraints imposed by current infrastructure and policy and the economies of joint cost and scale en route toward rapid commercialization.
Solutions to the technical challenges are within reach. Plasmonic devices can now be built from CMOS-compatible materials such as aluminum, copper, and titanium nitride. Reference Emboras, Najar, Nambiar, Grosse, Augendre, Leroux, de Salvo and de Lamaestre109–Reference Zheng, Wang, Nordlander and Halas113 Several new plasmonic materials, including transition-metal nitrides, promise high-temperature stability. Finally, self-assembly and imprint techniques are gaining momentum for large-area, engineered materials and devices with nanometer-scale features. Reference Sheikholeslami, Alaeian, Koh and Dionne114–Reference Henzie, Lee and Odom116 It will be exciting to see what the coming decades hold for plasmonic materials and applications, both in academia and on the market. Undoubtedly, the field will be successful if it can hold true to the mindset of one of the world’s greatest optical-device inventors, Thomas Edison: “Find out what the world needs. Then … go ahead and try to invent it.”
Acknowledgments
The authors greatly appreciate useful feedback from all Dionne group members. Funding from a Presidential Early Career Award administered through the US Air Force Office of Scientific Research is gratefully acknowledged (Grant FA9550-15-1-0006), as are funds from a National Science Foundation CAREER Award (Grant DMR-1151231), the Gordon and Betty Moore Foundation, and Northrop Grumman. This work was additionally supported by the US Department of Energy “Light–Material Interactions in Energy Conversion” Energy Frontier Research Center under Grant DE-SC0001293 and by the research program “Fellowships for Young Energy Scientists” (YES!) of the Foundation for Fundamental Research on Matter (FOM), which is financially supported by the Netherlands Organization for Scientific Research (NWO).
