Plasmonic metasurfaces based on ensembles of distributed metallic nanostructures can absorb, scatter, and in other ways shape light at the nanoscale. Forming hybrid plasmonic metasurfaces by combination with other materials opens up for new research directions and novel applications. This perspective highlights some of the recent advancements in this vibrant research field. Particular emphasis is put on hybrid plasmonic metasurfaces comprising organic materials and on concepts related to switchable surfaces, light-to-heat conversion, and hybridized light-matter states based on strong coupling.
Light interacts strongly with metal nanostructures via resonant excitation of plasmons, which are collective oscillations of the conduction electrons in the nanostructures.1 The nanostructures are typically on the order of tens to hundreds of nanometers in size and effectively act as antennas for light. So-called plasmonic metasurfaces are 2D arrays of such optical nanoantennas, distributed on a substrate to provide properties beyond the mere average response of the materials they are made from.2,3 The recent large interest in plasmonic metasurfaces and nanoantennas is related to their ability to control light at the nanoscale.4 They can focus optical fields to nanoscale “hot spots”5–7 and abruptly modify optical wavefronts.3,8,9 Furthermore, they can generate vibrant structural colors10–15 or convert light to heat in nanoscopic volumes.16–21 As a result of all these functionalities, plasmonic metasurfaces have found use in an exceptionally wide range of areas, including biosensing,22–38 energy conversion,33,39–47 display technologies,14,48–51 photocatalysis,52–55 ultrathin optical components,8,10,56–59 and many more.60–72
Focus of this perspective is put on recent research that combines plasmonic metasurfaces with complementary materials and concepts, forming so-called “hybrid plasmonic” metasurfaces. Such hybrid combinations can provide novel properties and functions that are difficult or even not possible to achieve by the original components when used separately. Potential applications range from new types of displays14,48, 73–77 and sensor systems78–81 to different energy conversion concepts,39,45,46,54,82 including devices that can harvest energy from light fluctuations.83 An additional branch of hybrid plasmonics relates to coupling of molecules to the vacuum field of plasmonic nanocavities, which forms hybrid light-matter energy states and hybrid systems with a whole new set of additional interesting possibilities.69,70,84 This perspective covers by no means all exciting studies on hybrid plasmonics, but primarily uses our recent contributions as examples to present the current state of the field, and as a basis to discuss new possibilities and future directions. For an in-depth overview of hybrid plasmonics including inorganic systems, see, for example, the recent review by Jiang et al.85 We focus particularly on systems that utilize organic functional materials and on three directions: dynamic systems, heat management, and strong coupling.
II. HYBRID SYSTEMS FOR DYNAMIC PLASMONIC METASURFACES AND REFLECTIVE DISPLAYS
The use of plasmonic nanostructures for color generation dates back to at least the 4th century86 and currently forms an emerging technology for ink-free color production.13,15,87–91 Advantages include possibility for high resolution and chromaticity, excellent stability over time, and environmental friendliness due to low consumption of materials compared with traditional coloration based on dyes or pigments.15,49,88,92 The mechanism of plasmonic coloration relates to the nature of the plasmon excitation itself, which is a resonant phenomenon that occurs preferentially for certain frequencies of light. The plasmon resonance frequency, and hence color, depends on many factors, including shape, size, and distribution of the nanostructures, as well as on the complex permittivity of the metal and the permittivity of the surrounding . In order to illustrate the resonant nature of plasmonic interactions and their dependence on different factors, we present the polarizability for a single ellipsoidal metal nanoparticle. In the quasistatic approximation for particles with dimensions much smaller than the wavelength, we have93
where V is the particle volume and is an axis-specific geometrical factor, fulfilling .94 The optical extinction upon excitation along axis i is then given by , where k is the wave number of the incident light.93 For metals with low imaginary permittivity, resonance (vanishing denominator) occurs approximately at , where we have introduced . is 2 for a sphere and can vary significantly for other shapes but remains positive. These features illustrate that plasmon resonances can be tuned by geometry and also why plasmon excitation occurs specifically for metals, which can provide the negative real permittivity needed to fulfil the resonance criterion. In addition, arrays of nanoparticles and other more complex systems provide further degrees of freedom to locally control resonances and colors. This was utilized to reproduce color photographs with extreme resolution,15 as well as to create polarization-dependent color images.90,95
While plasmonic metasurfaces can be designed to shine in more or less any color, it is more challenging to change their properties after fabrication. The main reason is that the materials that the plasmonic structures are based on, often gold or silver, have well-defined optical properties that are not easily modified. Besides fixed permittivity, the most common plasmonic materials and structures also do not support modulation of particle shape back and forth. On the other hand, being able to tune plasmonic metasurfaces in situ and dynamically control optical fields and interactions at the nanoscale could lead to entirely new scientific directions and novel applications.96 For example, ultrathin flat metasurface lenses based on abrupt wavefront engineering could find many additional uses if they could be dynamically tuned, as recently demonstrated for a stretchable metasurface featuring dielectric nanoantennas.97
Substantial efforts have recently been focused on achieving dynamic control of the optical response of plasmonic metasurfaces.96 Some interesting approaches utilized permittivity-modulation in materials such as ultrathin gold films,98 transparent conducting oxides,99–101 or graphene,102–104 or tuning based on more exotic materials like polycyclic aromatic hydrocarbons105,106 and phase change materials.107,108 Our group also recently contributed by introducing organic conductive polymers as new plasmonic materials with redox-tunable properties.109 Other routes focused on varying geometrical factors, such as periodicity of nanostructures in plasmonic arrays110,111 or gap size for particle dimers.112 In addition to stretching using elastomeric substrates,110 researchers have also explored electromechanical actuation for reconfigurable plasmonics.113 Yet further directions for dynamic control of plasmonic systems have involved liquid crystals,14,114–117 magnetoplasmonic systems,71,72,118,119 and electrovariable nanoplasmonics at liquid-liquid interfaces.67 More information on dynamic plasmonic systems and active metasurfaces can be found in the recent review by Shaltout et al.96
One important application area for dynamic plasmonic metasurfaces is reflective displays (electronic paper, or e-paper).48 Electronic displays are already responsible for a substantial fraction of our energy consumption, and the global use of displays is inevitably increasing. Since the energy needed to drive emissive displays cannot be reduced indefinitely, we need alternative and complementary types of systems. Reflective displays can save energy by not emitting light but instead controlling how ambient light (sun light, indoor lighting, etc.) is reflected to produce text and images. Besides energy savings, reflective displays also come with additional advantages, such that they can be used in sunny conditions. Plasmonic metasurfaces are interesting for reflective displays since they can provide control of reflected colors while maintaining high absolute reflection.48 This is important because reflective displays are limited to using only the amount of light that is available from natural lighting. Motivated by this, we have explored switchable hybrid plasmonic metasurfaces for reflective displays in color. These systems combine colorful high-reflective plasmonic metasurfaces with switchable electrochromic conducting polymer materials. The optical properties of conducting polymers can be controlled electrochemically via the redox state of the material. In brief, the redox state determines the density of charge carriers along the backbone of the conjugated polymer, which affects both the electrical conductivity and optical transparency of the materials.120–123 Combined with advantages such as low-cost, sustainability, easy processing, and patterning, this has made electrochromic polymers popular for reflective labels and displays.124–128 One limitation, however, is that these electrochromic materials typically lack control of color and primarily enable monochromic tuning. Recent research circumvented this by combining electrochromic polymers with colorful plasmonic metasurfaces, which, thereby, could be turned into reflective or transmissive red-green-blue (RGB) pixels.73,75–77 Such hybrid systems have potential to enable energy-efficient e-papers in color. To facilitate low-cost sustainable devices compatible with large-scale use, we developed colored plasmonic metasurfaces based on aluminum and copper instead of the more commonly used materials such as gold and silver.77 The main structure comprised an aluminum mirror and a plasmonic copper nanohole film, separated by an aluminum oxide spacer layer [see Fig. 1(a)]. Resonance positions and color of the optical nanocavity could be controlled by varying the spacer thickness, enabling red, green, and blue metasurfaces [Fig. 1(b)], and the possibility to accurately reproduce color images. The addition of nanoholes in the top mirror of the optical nanocavity allowed an enhancement of the coloration (except for red pixels), exploiting the resonant plasmonic excitation of the nanostructures. While these metasurface images are static, we introduced dynamic switching by screen-printing a thin film of the electrochromic polymer PEDOT:PSS (poly[3,4-ethylenedioxythiophene] doped with polystyrenesulfonate) on top of the metasurfaces. Based on its redox-tunable optical transmission, the electrochromic layer allowed the reflection from the metasurface to be repeatedly turned on and off [see Figs. 1(c)–1(e)]. The electrochromic polymer provides bistability (relatively stable in both its transparent and opaque states), making the hybrid plasmonic system promising for reflective color displays requiring ultralow energy consumption, for use in applications ranging from billboards to smart labels and packaging. Future systems may benefit from other types of polymers and metasurface designs,73,120 as well innovative means of production.129 There are also interesting systems that utilize classical optical microcavities instead of plasmonic metasurfaces, including tunable devices based on microelectromechanical systems130 and phase-changing mirrors.131
Brightness, contrast, chromaticity, viewing angles, and power consumption are some of the key parameters for color displays.48 While a wide range of techniques to obtain structural color are available, the tunability of those structures remains challenging. In particular, response times might be long (especially for electrochemical, chemical, and phase change methods) or there could be issues with long-term stability. Brightness is also a crucial aspect, because reflective displays are restricted to working with the incident light. Many systems have the disadvantage of having a relatively low overall reflection efficiency. This is particularly problematic in light of RGB subpixel systems, for which the reflected intensity for a given color cannot exceed 33% of the incident light. Single pixels that are tunable throughout the entire visible range could be a promising solution to this issue.
III. BRING UP THE HEAT WITH HYBRID PLASMONICS
Plasmonic metasurfaces can be used as light-triggered nanoscale heat sources, which has been studied extensively and found use in many applications, ranging from photothermal therapy and solar autoclaving to energy harvesting and plasmon-driven biomolecular thermophoresis.17,19,24,132–136 Other examples include heat management of windows and novel routes for ski goggles with antifogging properties.137 In fact, it is challenging to avoid heat losses in plasmonic systems, which has triggered exploration of alternative low-loss dielectric nanoantennas for use in applications where losses pose a problem.138 The field of thermoplasmonics, instead, utilizes plasmonic heating favorably in various novel concepts and applications.139–142 The phenomenon is related to Joule heating from the optically induced current in the metal, with local heat power density q at arbitrary position x inside the metal given by
where is the electric field generated by the plasmonic excitation and is the complex amplitude of the electronic current density. This expression can be modified to
where is the vacuum permittivity.17,136 The above equation illustrates that the heat power is proportional to the imaginary component of the metal's permittivity and to the square of the electric field generated inside the metal nanostructure. The heat generation is due to nonradiative decay of plasmons via absorption, and the total heat generation, or heat source Q, for a nanostructure is given by
where is the absorption cross section of the nanostructure and I is the incident light irradiance.
Notably, heat generation is different from the temperature increase. The increase in temperature for a system upon plasmonic excitation is not determined only by the generated heat, but governed by the balance between the generated heat and heat dissipated to the surrounding environment. Hence, one should consider the complete system, including the surrounding environment and nearby nanostructures, when designing thermoplasmonic systems. For example, a plasmonic nanodisk typically generates less heat (hence, absorb less light) compared with a single plasmonic nanohole in a metal film, but illuminating the nanodisk can still result in larger local temperature increase, because the nanohole system more efficiently dissipates the heat through the continuous metal film.143 To further illustrate the importance of accounting for the whole system, the situation can be the opposite for arrays of nanostructures, for which metal nanohole arrays can provide superior heating and temperature increase over nanodisk arrays, primarily because the metal film no longer acts as an effective heat sink.144 For plasmonic arrays, the presence of nearby nanostructures highly affects the temperature increase of each single structure. In fact, the major contribution to the temperature increase for any given plasmonic particle may come from collective heating from neighboring particles rather than from heat generated by the particle itself.16,145 For more information on plasmonic heating, we refer to recent review articles on the topic.17,136,146
Plasmonic metasurfaces can be made transparent in the visible spectral region while generating heat from absorption of the near-infrared tail of the solar spectrum.147 Such systems form novel routes for thermal management in buildings via transparent windows with optimized energy transmission. Xu et al. recently combined this concept with electrochromic materials to create hybrid plasmonic tunable windows, showing that the photothermal properties could also reduce the tendency of the windows being attacked by micro-organisms.224 This forms an example of the many concepts and applications enabled by “hybrid” thermoplasmonic systems.148–150
Hybrid thermoplasmonics allows for novel concepts for energy harvesting and radiation sensing.80,81,83,144,151,152 Our group has explored different directions within this area, including the combination of plasmonic heating with ionic thermoelectrics144,153 and with organic pyroelectrics.79,83 Based on the latter, we designed a hybrid plasmonic metasurface for harvesting of energy from light fluctuations.83 The concept utilized a gold nanodisk array as the thermoplasmonic metasurface, combined with a thin organic layer with pyroelectric properties (see Fig. 2). If the molecules in the organic layer are properly aligned (polarized), the pyroelectric material [here poly[vinylidenefluoride-co-trifluoroethylene] (P[VDF-TRFE])] could translate temporal thermal fluctuations to electrical signals over the thin film. The phenomenon is related to the temperature-dependence of the permanent dipole moment over the thin film. Changes in temperature modulate the charge density on the material interfaces, in turn inducing a compensating current through an external circuit. For our hybrid thermoplasmonic device, the thermal fluctuations resulted from fluctuations in illumination intensity and corresponding thermoplasmonic heat generation. The devices could harvest energy and produce electricity from fluctuating illumination produced by leaves swinging in the wind.83
Hybrid thermoplasmonic metasurfaces also enable new types of self-powered light and heat sensors, which could be suitable for electronic skin applications in robotics and health care. Our recent approach combined plasmonic heating (of a gold nanohole array) with a new concept we called thermodiffusion-assisted pyroelectrics.79 A pyroelectric film provided rapid transient signals upon changes in temperature (induced by direct heating or by plasmon-induced heating upon irradiation), while a thermoionic gel,154 capacitively coupled to the pyroelectric part, contributed by also providing stable signals at equilibrium. The response was not only rapid, but the stable signal was also found significantly enhanced compared with the value expected from the pure thermoelectric response.
The hybrid thermoplasmonic concepts presented above exemplify that optical losses in plasmonic systems can be advantageous and enable novel applications and solutions when combined with other materials in hybrid systems. There are also challenges remaining that need further work, not least regarding improving efficiency of the thermoplasmonic-based energy harvesting concept in order to move from proof-of-concept to more practical useful devices. In that regard, sensor applications have already shown promise in terms of performance. Compared with light-induced heating by nonplasmonic materials, we believe that future work on hybrid thermoplasmonics will benefit from further utilizing the spectral tunability of plasmonic systems, such as designing transparent devices heated by the infrared tail of the sun. The ultralow thickness of plasmonic metasurfaces forms another strength that could be particularly valuable in low-weight applications, for example, in certain robotics and space applications.
IV. HYBRID SYSTEMS BASED ON STRONG COUPLING
The examples of hybrid plasmonic metasurfaces above combine features provided by a plasmonic metasurface (e.g., coloration or light-induced heating) with functionalities provided by a second system (e.g., modulation of transparency or generation of electricity), thereby enabling novel devices and applications. Another promising research direction instead focuses on how molecules and plasmonic cavities affect each other on the fundamental level due to their mere proximity. Placing molecules close to a plasmonic structure or other optical cavity forms an exciting route to control the functions of molecules without changing their structure, enabling exotic concepts such as long-range energy transfer,155–157 low-threshold polariton lasing,158 Bose–Einstein condensation,159 and superfluidity.160 The concept is based on coupling between molecules (or other entities with strong transition dipole moments) and a plasmonic cavity via spontaneous exchange of energy between molecular transitions and cavity resonances. If the coupling is sufficiently strong, the optical and molecular resonances hybridize to new light-matter states. This regime is referred to as strong coupling, which splits the initial molecular transition into two new polariton states that are separated by the vacuum Rabi splitting [, see illustration in Fig. 3(a)]. The vacuum Rabi splitting is determined by the number of molecules contributing to the coupling , the molecular transition dipole moment , and the vacuum electric field of the plasmonic cavity ,161
The fact that is proportional to d makes organic materials favorable for strong coupling since they typically have strong transition dipole moments.162 Introducing the expression for the vacuum electric field further gives69,163
where we have assumed that the molecules and the cavity are properly aligned for maximum coupling. V is the mode volume of the optical cavity, is the resonance energy, is the vacuum permittivity, is the relative permittivity of the surrounding, and is the reduced Planck's constant. The fact that increases with decreasing V makes plasmonic systems particularly suitable for strong coupling due to their capability to squeeze optical fields into ultrasmall mode volumes. Indeed, Chikkaraddy et al. recently utilized a plasmonic nanocavity to achieve strong coupling even for single molecules at room temperature.164 For systems that instead involve many molecules , we note that the coupling process not only produces the two polariton states, but also dark states that do not couple to light and that remain at the original energy level of the molecules [see Fig. 3(a)].165 For many practical applications that involve a large number of coupled molecules, these dark subradiant states, thus, heavily outnumber the two radiant upper and lower polariton states. The roles of dark states in strong coupling applications are not yet fully understood and form an interesting area for further research.69 Recent reports, for example, suggest that also the dark states may possess polaritonic properties such as a delocalized character.166 We also note that Eq. (6) does not contain the intensity of any light source, illustrating that the polariton formation originates from coupling of the molecules to the vacuum electromagnetic field of the cavity.84 The phenomenon can thereby be exploited for nonoptical applications as well. To mention some interesting examples, researchers have explored influence of strong coupling on chemical reactivity,167–170 ground state thermodynamics,171,172 work functions,173 and long-range transport of charges.64
As mentioned above, single plasmonic nanocavities can enable coupling of few or even single molecules owing to their small mode volumes. Plasmonic metasurfaces composed of ensembles of metal nanostructures can instead provide resonances that are delocalized over larger areas and thereby be interesting for different macroscopic applications of strong coupling.70,174 Compared to more commonly used Fabry Perot cavities, such plasmonic metasurfaces are not closed by mirrors and thereby provide physical access to the strongly coupled molecules. On the other hand, plasmonic metasurfaces can also have features that for some applications may be undesirable, such as low quality factors and vacuum electric fields that are localized at metal interfaces and less uniform compared with more conventional microcavities. Metal nanohole arrays form a class of plasmonic metasurfaces that can provide delocalized plasmonic modes, via excitation of surface plasmon polaritons (SPPs) propagating at the interfaces of the metal film.175–177 We have studied plasmonic nanohole arrays for various applications,79,144,178 and recently also explored hybrid systems based on metal nanohole films coupled to organic J-aggregates.179–181 Based on clear anticrossing behavior and Rabi splitting proportional to the square root of molecular concentration, we could conclude that the hybrid system was in the strong coupling regime, with Rabi splitting reaching several hundred millielectron volts.179 Furthermore, the polariton modes inherited the delocalized nature of the original plasmons of the nanohole films. Interestingly, we also found that the spectral positions of the upper and lower polaritons (P+ and P−) did not match peaks appearing in optical extinction spectra [compare the blue curves in Fig. 3(b)]. Compared with the polariton positions, as determined by integrating sphere absorption measurements, the extinction peaks (hence, transmission dips) were all significantly redshifted. This result pinpoints an important aspect of systems comprising both resonant states and nonresonant continuum states, which leads to Fano interference that modifies the transmission and reflection spectra.177,182–184 In that sense, we note that the same behavior occurs also for bare (nonhybrid) plasmonic nanohole metasurfaces.184 Depending on details like hole dimensions and film thickness, resonances can be centered in the middle between transmission peaks and dips, highlighting the importance of using absorption measurements (e.g., using integrating sphere) to identify resonances in these systems.
Plasmonic nanoparticle arrays form another interesting class of metasurfaces that can provide optical modes with delocalized nature. In contrast to metal nanohole arrays, a metal particle array naturally does not contain continuous metal interfaces for propagation of conventional SPPs. However, certain nanoparticle arrays can still sustain delocalized surface lattice resonances (SLRs) through localized plasmon resonances that are diffractively coupled to the array.174,185,186 Proper choice of nanoparticle size, shape, and periodicity can lead to extremely high quality factors and correspondingly narrow resonances compared with those of other plasmonic systems.174,187 Interestingly, SLRs share characteristics of surface plasmon polaritons on metal-dielectric interfaces and can propagate over several periods in the array.188 Regarding hybrid systems, the coupling of SLRs to molecular excitons189,190 can lead to long-range spatial coherence lengths of micrometers also for the strongly coupled regime, indicating delocalized nature of the hybrid polariton states.191 Compared with plasmonic nanohole arrays, nanoparticle arrays contain much less metal per area (assuming same thickness) and, therefore, show lower material absorption and higher direct transmission. As a result, SLR systems show only minimal Fano interference effects, and resonances can be determined by the positions of dips observed in the transmission spectra.184,192 Nanohole arrays instead provide advantages such as superior light-to-heat conversion and capability to act as electrodes.144
Hybrid plasmonic metasurfaces with delocalized polariton states have been explored for exotic macroscopic optical and nonoptical phenomena, where the manipulation of transport in organic thin films forms an interesting example. Orgiu et al. studied aromatic diimide-based organic semiconducting polymers coupled with plasmonic nanohole arrays and reported enhancement of charge transport due to strong coupling.64 Long-range polariton transport facilitated by strong coupling has been reported for nonplasmonic optical cavity systems,193 and the enhancement of both exciton transport194 and charge transport195 has been addressed theoretically. Other reports did not observe enhancement in charge transport upon strong coupling, including for systems based on microcavities coupled to p-type organic polymers,196 carbon nanotubes,197 or high mobility ambipolar polymers,198 as well as for hybrid plasmonic metasurfaces based on SLRs coupled with ambipolar polymers.199 Hence, it is currently not known whether the enhancement of charge transport is limited to only certain types of strongly coupled systems or if the concept is more general and possible to utilize more widely. In turn, this makes effects of strong coupling on charge transport an interesting topic for further studies, including the investigation of possible roles of dark states and their potential delocalized nature.69,166 Overall, hybrid strongly coupled plasmonic metasurfaces form an intriguing research area with several interesting unexplored directions and phenomena that are not yet fully understood, while also holding great potential for important (room temperature) applications, including electrical pumping,197 low-threshold polariton lasers,200,201 optical logic circuits,202 and quantum polaritonic devices.203,204
We hope that this perspective illustrates how hybrid plasmonic metasurfaces enable unique studies and applications beyond what is feasible with nonhybrid systems. While focused on three prime directions—somewhat biased to our own work—other plasmonic functionalities like hot electron production are also highly relevant for hybrid plasmonic metasurfaces.205,206 The perspective also focuses on systems where plasmonic metasurfaces are combined with organic materials, while interesting current and future research directions also feature other materials, such as transition metal dichalcogenides,207 MXenes,208,209 perovskites,210 as well as inorganic electrochromic materials211 and semiconductors.41,85,212 Likewise, not only plasmonic, but also dielectric metasurfaces based on high-index optical nanoantennas hold great promise for contributing to the field of hybrid metasurfaces, in particular for applications requiring low losses.97,138,213,214 Future studies and applications of hybrid metasurfaces may also cover wavelength ranges beyond the visible or near-infrared part of the spectrum, including the thermal emissivity range in the far infrared66,126,215 as well as the Terahertz range.102,216 These applications and studies could benefit particularly from the use of nontraditional plasmonic materials, including transparent conducting oxides,100,217–219 graphene220–222 and other 2D materials,209, 223 and even organic systems like redox-tunable conductive polymers.109
The authors acknowledge funding from the Wenner-Gren Foundations, the Swedish Research Council, the Swedish Foundation for Strategic Research, and Swedish Government Strategic Research Area in Materials Science on Functional Materials at Linköping University (Faculty Grant No. SFO-Mat-LiU o 2009 00971).