Properties of quantum dots coupled to plasmons and optical cavities

The fields of photophysics and photochemistry have generally focused on the use of photons to promote molecules to a metastable excited state that accesses interesting relaxation pathways including intersystem crossing, charge transfers,^1–3 bond making/breaking,^2,4 or isomerizations.⁵ Quantum electrodynamics (QED), in which a quantized electric field generated either by an optical cavity or by a plasmonic excitation interacts with a photonic emitter such as a molecule or a quantum dot (QD), is emerging as a subfield of these areas. When this interaction is sufficiently strong, it generates a new potential energy surface for the hybridized system. QED is therefore a way to achieve unprecedented control over photochemical reaction pathways and to access new types of photochemical reactivities based on, for example, creation of conical intersections and enhancement of interemitter energy transfer pathways.^6,7

There exists a range of coupling strengths between a cavity-mediated quantized electric field and a photonic emitter from “weak” to “ultrastrong.” In the weak coupling regime [Fig. 1(a)], two possible excited states are present in the system, one in which the emitter is in its excited electronic state (purple trace) and the cavity contains zero photons, and the other in which the emitter is in its ground state (blue trace) and the cavity contains a photon (orange trace). As the coupling between the cavity and the emitter increases, the two excited states hybridize to form polaritonic states [Figs. 1(b) and 1(c)] with potential energy surfaces dependent upon the strength of the coupling, which can be modulated by the properties and geometry of the field and emitter. Theoretical treatment of a molecular photonic emitter coupled strongly to a quantized electric field predicts unprecedented selectivity for a single stereochemical product of a photodriven reaction.^8,9 The quantum yield for such a reaction may be greater than 1 due to the ability of multiple individual emitters to couple collectively to one photonic excitation.⁹ 

Zero-dimensional semiconductor nanocrystals, or QDs, which can be deposited from solution or grown epitaxially within a cavity or near plasmonic structures, have some advantages over molecules for use as emitters in QED systems. QDs are exceptionally photostable upon prolonged exposure to high energy and high intensity light sources.¹⁰ Spherical QDs have isotropic transition dipole moments¹¹ such that their orientation in space does not affect their ability to couple to a quantized electric field. Anisotropic nanorods and nanoplatelets, which depending on their size are quantum-confined in between one and three dimensions,^12–14 must be aligned within a cavity for maximum coupling but have oscillator strengths that are a factor of ten or more higher than analogous spherical particles.¹⁵ Previous work has compared QDs to atoms for light-matter coupled systems.¹⁶ 

In this perspective, we explore the potential of epitaxial and colloidal QDs to serve as emitters within QED systems and highlight some advantages of QD-cavity systems for applications such as photochemistry. This perspective outlines the fundamental physics of these systems, provides a general overview of the field as it currently stands, and shares our outlook on new avenues for the field to explore but does not comprehensively review the field. We first discuss the characteristic spectral changes for an emitter in the weak, intermediate, and strong emitter-field coupling regimes and highlight applications in photochemistry and quantum information. We then discuss experimental realizations of strongly coupled polaritonic systems with QD emitters and discuss the application of such systems to biexciton extraction from QDs, generation of entangled photon pairs, and photochemistry. For the reader’s convenience, we have compiled a glossary of common terms in Table I.

The coupling of a QD (or any emitter) to a cavity can be described by a coupled-oscillator model.¹⁷ Consider the on-resonance (or zero detuning) case, where the energy of the cavity mode is equal to the energy of the exciton mode of the QD (E_C = E_X = E₀). At resonance, the energy of the coupled modes, E_1,2, is defined by the following equation:

where γ_C is the full-width-at-half-maximum (FWHM) of the cavity mode, γ_X is the FWHM of the exciton mode, and g is the exciton-photon coupling parameter (Fig. 2).^17,g is the scalar product of the transition matrix element of the exciton’s dipole moment and the value of the electric field at the QD’s position, either within a microcavity or with respect to a proximate plasmonic structure. The threshold for the strong coupling of a QD and a cavity is defined as $g>γC−γX4$⁠.¹⁷ 

A semiclassical model of a plasmonic metal nanoparticle interacting with an excitonic material suggests that the only type of coupled system in which the coupling constant is independent of the magnitude of the electric field at the position of the excitonic material is one in which the plasmonic metal nanoparticle is homogenously and infinitely coated with the excitonic material.¹⁸ Since a single QD cannot homogenously and infinitely encompass the electromagnetic field of a cavity, the QD should be placed in the antinode of the field in all three dimensions to maximize the available field strength. With this geometry, the coupling parameter is optimized by maximizing the product $fVm12Q$⁠, where f is the QD oscillator strength, V_m is the mode volume, and Q is the quality of the cavity. Q is defined as E_C/γ_C, the ratio of the energy of the cavity to the FWHM of the cavity mode.¹⁷ While higher Q values are desirable for increasing the coupling strength, the mode volume of the cavity must also be considered. Ultrahigh quality (Q > 10⁸) cavities have been fabricated,¹⁹ but their microscale geometry results in larger cavity mode volumes, which limit the maximum possible coupling strength. In fact, Q factors of just 10²–10³ may result in higher coupling strengths if the cavity is of nanoscale.^20,21 The strong coupling threshold therefore may be achieved by increasing the oscillator strength of the relevant QD excitonic transitions (by, for instance, using semiconductor nanoplatelets or nanorods), decreasing the mode volume for cavities of high quality, or increasing the quality factor, Q, for cavities with small mode volumes.

We proceed below by defining the various exciton-photon coupling regimes (summarized in Table II). We specify both the weak and intermediate regimes, which fall within the mathematically defined weak-coupling limit (⁠$g≤γC−γX4$⁠), as well as the strong and ultrastrong regimes, which are within the mathematically defined strong-coupling limit (⁠$g>γC−γX4$⁠). It is possible to observe multiple coupling regimes within the same system (Fig. 3).^22,23

In the weak coupling regime, the energies of the coupled modes, E_1,2, are degenerate (i.e., zero splitting)¹⁷ and the QD and cavity spectral components generally resemble their isolated states.^24,25 Weak coupling between a QD and a cavity results in a peak shift and line broadening of the absorption spectrum of the QD, although both the multipeak absorption spectrum and the single-peak emission spectrum retain the shape of those of an isolated QD. The rate of irreversible spontaneous emission from the QD is, however, enhanced (when on-resonance with the cavity mode) or suppressed (upon detuning) due to the increase of the density of states of radiation that occurs upon confinement within a cavity, as dictated by Fermi’s golden rule.^24,26 The magnitude of the enhancement resulting from this phenomenon, known as the Purcell effect,^27,28 can be calculated using cavity properties Q and V_m, and is quantified by the Purcell factor, F_P,

where λ is the wavelength of the transition within the cavity. Equation (2) assumes that (i) the QD is located in the antinode of the cavity’s electromagnetic field in all three directions, (ii) the linewidth of the QD is narrower than the linewidth of the cavity mode, (iii) the dipole of the QD is aligned with the polarization of the cavity mode, and (iv) the cavity mode is in resonance with the frequency of the excitonic transition (i.e., zero detuning).²⁹ Although it may be expected that the linewidth of a single QD would be larger than that of a cavity mode at 300 K, an epitaxially grown InAs QD located within a microcavitiy has been shown to have a smaller emitter linewidth compared to the cavity mode.³⁰ Such a condition has also been fulfilled with epitaxially grown CdSe/ZnSe QDs located between SiO₂/TiO₂ Bragg reflectors.³¹ 

As the coupling strength increases, but remains below the strong-coupling limit, an asymmetric shape in the absorption spectrum (or scattering spectrum) of the emitter develops due to quantum interference between the discrete exciton mode and the continuum of states to which it couples.^22,23,32,33 The emission spectrum, however, retains a single peak. The appearance of the asymmetric shape in the optical spectrum of the emitter is known as the Fano effect.³³ 

In the strong coupling regime, the energies of the coupled modes (E_1,2) formed upon hybridization result in two resolvable peaks with different energies but identical linewidths. The presence of clearly resolvable doublets in the spontaneous emission of a QED system is known as vacuum Rabi splitting,³⁴ and the magnitude of the splitting increases with the coupling strength.³⁵ In the strong coupling regime, coherent energy exchange is faster than energy dissipation due to decay and/or decoherence;^8,26,28 therefore, coherent oscillations and quantum effects resulting from coupling are observed.²⁵ The nondegeneracy of the eigenstates gives rise to avoided level crossings, a key indication of a strongly interacting quantum system.³⁵ 

The so-called “ultrastrong coupling” occurs when the coupling strength, g, is greater than the dipole transition energy of the emitter. In the ultrastrong coupling regime, unlike weaker coupling regimes, the ground state of the coupled system is modified, and previously neglected terms of the Hamiltonian, such as contributions from the carriers of the cavity mode and counter-rotating terms typically neglected under the rotating wave approximation, become significant.^25,26,36,37 Accessing this regime may result in asymmetry in the vacuum Rabi splitting, photon generation, and new transition pathways.³⁷ 

Strong coupling has been achieved in a variety of QED structures where either colloidal or epitaxial QDs are the emitters (Table III). Initial QED systems used epitaxially grown QDs and optical microcavities; these cavities are diffraction-limited, where the mode volume, V_m, is dictated by the wavelength of incident light.³⁸ More recently, QED devices were fabricated with plasmonic cavities, which overcome the diffraction limit.

In early examples of strong coupling between quantum-confined semiconductors and optical microcavities, the cavity devices were fabricated using either molecular beam epitaxy^39,40 or chemical vapor deposition.⁴¹ The emitters in these devices were typically In_xGa_1−xAs or GaAs quantum wells (QWs). The three most common geometries of these devices include micropillar devices, microdisk devices, and photonic crystal cavities, as shown in Fig. 4.

Micropillar devices [Fig. 4(a)] consist of one or multiple layers of QWs sandwiched between Bragg reflectors, alternating layers of AlAs and AlGaAs.^39–41 The Bragg reflectors act as mirrors and optically confine light in the axial direction such that light from micropillar devices is emitted in this direction. Q values for the cavity modes of micropillar devices can reach 5000–10 000.³⁴ In microdisk devices [Fig. 4(b)], QWs are grown between slabs of a dielectric material.⁴² Electron beam lithography and chemical etching are then utilized to create the microdisk structures.⁴² Light is confined in all three dimensions by total internal reflection, which in some cases results in cavity modes with Q-factors greater than 10 000.³⁴ In photonic crystal cavity devices [Fig. 4(c)], QWs are grown epitaxially in dielectric slabs, and lithography and etching are then utilized to cut holes into the dielectric slabs; these holes serve as the cavities.⁴³ The geometry of the holes in the photonic crystal layer has a large impact on the Q value of the cavity mode; GaAs-based photonic crystal cavities are known to achieve Q factors of 20 000.⁴⁴ 

With the development of increasingly precise fabrication techniques, use of epitaxial QDs as emitters became possible. QDs offer two primary advantages over QWs: narrower linewidths under weak excitation conditions²⁹ and an enhanced rate of spontaneous emission into the cavity mode.^30,45

The main limitation of these epitaxially grown, optical microcavity devices is the V_m. The mode volumes are large because they are dictated by the wavelength of the incident light, which limits the Q-factor of the cavity and results in Rabi splitting energies of <1 meV.⁴⁶ As a result of the large V_m of the cavities, the emitters in the cavities must have large dipole moments in order for the system to achieve strong coupling;³⁴ this condition limits the types of quantum-confined semiconductors that can be used as emitters. One can straightforwardly increase the oscillator strength of semiconductor emitters by moving from spherical QDs to anisotropic nanocrystals such as nanorods, nanoplatelets, and nanosheets. This strategy was successfully employed in the development of a QED architecture that exhibited strong-coupling between the exciton mode of a CdSe/ZnS nanorod and a photon mode of a dielectric polystyrene microsphere.⁴⁷ 

Cavities composed of plasmonic materials can have cavity-mode volumes that are smaller than the diffraction limit because surface plasmon polaritons (SPPs) can localize light fields to nanoscale dimensions.^24,46 In experimental realizations of plasmonic QED systems, the QD exciton interacts with (i) single isolated surface plasmons, (ii) multiple coupled plasmons, or (iii) the collective plasmon of the cavity.

The earliest experimental examples of plasmonic QED systems demonstrated coupling between QDs and the surface plasmons of planar Ag films.^18,48,49 These devices were prepared by spin coating colloidal suspensions of CdSe QDs onto thermally evaporated Ag films on glass substrates. In one example, a QED device was fabricated by lithographically placing CdSeTe/ZnS QDs along a plasmonic silver wire [Fig. 5(a)].⁵⁰ This device construction enables the QD excitons to interact with the localized surface plasmon resonances of the Ag nanowire.⁵⁰ One strategy to increase the coupling strength in a QED system is to decease the size of the nanoantenna; e.g., in the case of spheroid nanoantenna, the size of the minor axis affects the coupling strength between the nanoantenna and QDs.⁵¹ Small nanoantennae, however, suffer from high absorption losses.⁵¹ 

One promising QED architecture for modulating the electric field generated by localized surface plasmons is the metal nanogap structure, where the electric field of the cavity is produced by coupled plasmons of proximate nanostructures. Such systems have been studied extensively since the theoretical prediction that strong coupling is achievable between a single quantum emitter, e.g., a QD, that is placed in a nanoscale gap between two metal nanoparticles and the electric field in that gap.³⁸ A variety of methods have been utilized to create metal nanogap structures. For example, interfacial capillary forces were exploited to install CdSe/ZnS QDs in the gap between two Ag nanoparticles arranged in a “bowtie” shape [Fig. 5(b)].⁵² The resonance of the cavity was tuned by varying the side-to-side length of the Ag bowties and the distance between the bowties.⁵² In the same year, strong coupling was reported between colloidal suspensions of mixtures of Ag nanoshells on SiO₂ cores and CdSe/ZnS QDs.⁵³ Because the Ag nanoshells were synthesized using a seed-mediated method, their surfaces were notably roughened with Ag nanoparticles. The electric field of the nanoshell under laser excitation was enhanced, especially between neighboring nanoparticles, and, as such, strong coupling was achievable with QDs.⁵³ 

More recently, surface chemistry of colloidal QDs was used to covalently link Au nanoparticles to CdSe/CdS QDs through the ligands on both the Au particles and the QDs.²² These linked structures were then drop-cast onto Ag films, and an Au nanoparticle and the Ag film acted as the cavity.²² Weak, intermediate, and strong coupling were observed between plasmonic cavities and QDs in the system due to the nonuniformity of the local structure.²² Another method, explored theoretically, for achieving such devices used Ag nanotweezers to optically trap QDs in the cavity.⁵⁴ A major limitation of the metal nanogap structure strategy for achieving strong coupling between a QD and a photon is that, as mentioned above, the QD emitter must be localized to the antinode of the electric field of the cavity.

There are several examples of QED devices where the QD interacts with the collective plasmon mode of the cavity. Such a QED device has been fabricated with a drop-cast layer of colloidal CdSe QDs and an Au nanohole array [Fig. 5(c)].⁵⁵ The coupling strength between the QDs and the Au nanoarray was altered by changing the spacing of the array.⁵⁵ Recently, a slit-like nanoresonator on the corner of an Ag flake served as a probe for nanometer-precise raster scanning of a film of isolated CdSeTe/ZnS QDs embedded in a poly(methylmethacrylate) film.⁵⁶ The coupling strength of the QED system was easily modifiable by changing the distance between the QD and the nanoresonator.⁵⁶ Although challenging to fabricate, plasmonic nanoresonators are attractive for cavities because their resonance wavelength is easily tunable. The electric field produced by the resonators, however, is weaker than the one produced in the “hot spot” of the nanogap structures.

As discussed in the Definitions of Coupling Regimes, the coupling of QDs to photonic cavities or surface plasmon polaritons (SPPs) can greatly affect the absorption and emission properties of the system. Even weakly coupled systems can increase the rate of spontaneous emission by the Purcell effect, as observed, for example, through time-resolved fluorescence spectroscopy [Fig. 6(a)].⁵⁷ In addition, coupling a QD to a cavity can dramatically decrease the probability of multiphoton emission. The deterministic generation of single photons by QD-microcavity systems is a first step toward realization of photonic qubits, where the quantum properties of light such as polarization, momentum, and energy can be used for the storage and processing of quantum information.⁵⁸ 

The directionality and polarization of the QD emission can also be controlled by an anisotropic electric field that encourages spontaneous emission in a preferred direction.²⁴ This anisotropy is obtained with plasmonic materials by altering the shape of the metal nanostructure and the position of the QD.^{24,50,59–61} For example, Fig. 6(b) shows that placing a single QD emitter in a Yagi-Uda antenna array results in unidirectional and strongly polarized emission.⁶² Similarly, in 2009, it was reported that the rate of directional emission for CdS-ZnS core/shell QDs was enhanced when coupled to an Au plasmonic nanodisk system.⁵⁹ Using time-resolved emission, the authors observed that the radiative decay of the system is sensitive to the polarization of the incident light and the collection angle, suggesting that the enhancement of the directional emission requires both spectral and spatial overlap. This directionality can potentially be applied in lighting, optical sensing, quantum emitters, and microscopy.

While QD-cavity systems in the weak coupling regime show perturbations of the QD emission, systems in the strong coupling regime demonstrate properties unique to either of the separate components. As discussed in the Definitions of Coupling Regimes, strong coupling occurs when the emitter is able to exchange energy with the cavity or SPP, resulting in the formation of hybrid exciton/electric field states.³⁴ The observation of these new upper polariton (UP) and lower polariton (LP) states and the Rabi splitting between them is the definitive indication that the system is strongly coupled.⁶³ Optical excitation from the QD ground state to the UP and LP states has been predicted theoretically⁶⁴ and experimentally measured through steady-state and time-resolved absorbance spectroscopy.^40,55 Similarly, Rabi splitting has been observed in scattering spectra.^18,52 For example, in 2010, room-temperature coupling between CdSe QDs and a Ag plasmonic film was reported with a Rabi splitting of ∼112 meV.⁴⁸ The authors proposed that, in the strong coupling regime, QD-based materials have the potential to be used in next-generation all-optical nonlinear devices, threshold-less laser operation, and single photon transistors. One consideration in identifying the strong coupling regime, however, is that similar splitting in the scattering spectra can be caused by Fano-like interference in the intermediate coupling regime.^22,64 The distinction between Fano-splitting and Rabi splitting can be resolved by examining the photoluminescence of the QD, as shown in Figs. 3(a)–3(c). Photoluminescence, unlike scattering, is an incoherent process and thus is unaffected by Fano interferences. For these reasons, fluorescence spectroscopy is recognized as the preferred characterization technique for identifying and quantifying Rabi splitting.^22,34

For more than a decade, it was predicted theoretically that strong coupling of a QD and a metal nanoparticle could result in splitting of the emission spectrum of the QD.^65–67 More recently, such splitting has also been measured experimentally for QDs coupled with SPPs^{22,49,52,53,56,68} and microcavities.^17,44,69 Particularly with SPPs, much work has been dedicated to understanding how the geometry of the metal nanoparticle can be optimized to maximize coupling while keeping the size of the system small. In 2010, it was predicted that Rabi splitting should be observed for a single QD emitter such as CdS or ZnS in the center of a Ag dimer nanoantenna, a system with dimensions smaller than 36 nm.³⁸ As mentioned in the Experimental architectures section, “bowtie” structures like the example shown in Fig. 7(a) are known to create a “hotspot” region between the two prisms with an electric field that is enhanced compared to that created by the single nanostructure. In one study, placing CdSe/ZnS QDs in this hotspot resulted in a 140-meV Rabi splitting.^52,68 The utilization of SPPs to produce focused electromagnetic fields provides an effective approach for quantum logic operations on a scale compatible with compact electronic devices.

The transition between a polariton state and a lasing state has been observed in a self-assembled InAs QD-photonic crystal microcavity (Fig. 8).⁷⁰ The emission spectrum of the system changes from double-peaked to single-peaked upon increasing the power of the light source, suggesting that lasing is possible even in the strongly coupled regime. Similarly, in 2014, it was determined computationally that the dark plasmon in a QD-SPP system can transfer its energy to the QD to achieve emission without population inversion. This gain-without-inversion effect can be used for QD lasing.⁷¹ 

A further method of identifying strong coupling is to observe the mixing of QD electronic excited states and photonic states to form the new UP and LP modes as a function of coupling strength. The strength of coupling can be changed by manipulating the physical location of the QD relative to the plasmonic material^41,56,72 or by “detuning:” bringing the energies of the excitonic and optical modes in and out of resonance.²⁴ Early experiments with photonic cavities demonstrated detuning by changing the temperature of the system.^16,17,42,44

For planar systems, the angle of incidence can similarly be used to vary the energy of the plasmonic state and, therefore, the degree of detuning.^39,48,49,73 As shown in Figs. 7(b) and 7(c), a dispersion curve can be generated by mapping the UP and LP states as a function of the angle of incidence. The characteristics of the UP and LP states switch from primarily excitonic to primarily plasmonic depending on the angle of incidence. For example, the SPP state is expected to have a shorter lifetime than the exciton state.⁷⁴ Time-resolved reflectivity measurements revealed a decrease in the decay rate of the LP mode between 45° and 65°, which indicates that the relative composition of the hybrid state changes from primarily plasmonic to primarily excitonic.⁴⁹ 

QDs are good emitters for QED systems, as they have narrow photoluminescence linewidths (20–30 nm at 298 K) and are often more photostable than molecular emitters. QDs also offer an advantage over molecules when it comes to biexciton extraction and the generation of entangled photon pairs. A recent study showed that CdSe/ZnSe QDs in the strong coupling regime with an optical microcavity have exciton-photon couplings on the order of 20–30 meV, which is on the same order of the biexciton binding energy in these materials and therefore suggests that biexciton extraction and generation of entangled photon pairs is possible for colloidal QD-based structures.⁷⁴ 

QD-based systems still suffer from several limiting factors. First, QDs are not “artificial atoms” as they are often modeled. The quasicontinuum of transitions of the QD can overlap with the cavity resonance, increasing the difficulty of accurately predicting their light-matter interactions.^75,76 QDs are, in general, photostable, but for a strongly coupled light-matter system, continuous incoherent pumping averages out Rabi oscillations and broadens emission lines.^77,78 In addition to homogeneous broadening, there is also dispersity in QD sizes and, therefore, bandgaps.⁷⁸ Even high-quality synthetic batches of QDs have particles with radii that vary by 5%–10% within the ensemble; such heterogeneity complicates the determination of Rabi splitting.⁷³ Realizing strong coupling with single QD-cavity systems alleviates this issue, but it is also challenging to measure Rabi splitting, the magnitude of which is proportional to $N$⁠, where N is the number of emitters, with isolated individual QDs .^52,56,79 Multiemitter ensembles additionally have the advantage that they can give rise to the so-called “cavity protection,” where the polariton is decoupled from the nonemissive states that result in its relaxation.⁷⁸ 

Cavity-coupled photochemistry, in which coupling to a cavity modifies the potential energy surface of the reaction, has been demonstrated with organic dye molecules in a number of theoretical and experimental studies. For example, in 2016, theoretical prediction showed that the rate of electron transfer from a molecular excited state coupled to the cavity mode to a molecular excited state decoupled from the cavity mode can be significantly enhanced within an optical cavity. This increase was attributed to (i) a reduced energy gap between donor and acceptor orbitals due to Rabi splitting and (ii) the suppression of the rearrangement of nuclei (reduced reorganization energy) during charge transfer due to strong cavity-molecule coupling.⁸⁰ In 2012, Rabi splitting was observed in the absorbance of spyropyran in a silver mirror cavity. This formation of hybridized light-matter states leads to a decrease in the rate of photoinduced ring-opening of spyropyran.⁷ A study from 2016 reported a decrease in the rate of the deprotection of 1-phenyl-2-trimethylsilylacetylene when the Si-C vibrational mode couples with the vacuum electromagnetic field.⁸¹ In 2017, a system was reported that used J-aggregate cyanine dyes as energy donors and acceptors spatially separated by poly(methylmethacrylate) in an optical cavity of two parallel silver mirrors. The energy transfer between dyes in the strongly coupled regime occurred with an efficiency of 37% when the distance between the donor and the acceptor was larger than 100 nm, which is beyond the characteristic Förster radius of the system, due to formation of hybrid polaritonic states.⁸² Thus, we see that cavity-coupled organic dye molecules may undergo photochemical processes that are highly sensitive to this coupling.

We propose that a next step in cavity-coupled photochemistry is to use QDs as the photosensitizer or photocatalyst for a substrate molecule and thereby investigate the emergent photochemical properties of cavity-coupled QD systems. In order to achieve such a system, we may take advantage of extensive previous work with (i) QD-coupled QED systems in both the weak and strong coupling regime, (ii) QED photochemistry achieved with molecular emitters, and (iii) studies of QDs as photocatalysts for a variety of excited state and redox reactions.⁸³ 

The confluence of these three research avenues provides a foundation for the future exploration of QD-based QED photochemistry. For example, the ability of QD photocatalysts to perform electron or energy transfer to molecular acceptors may be significantly enhanced when coupling with a plasmonic material or an optical cavity. Such architectures may also enable chiral photocatalysis, without the use of bulky substituents, through the incorporation of a chiral emitter that also serves as the photocatalyst. QDs can be transformed into chiral emitters through functionalization with chiral ligands⁸⁴ or coupling with nanophotonic waveguides.^85,86

The photochemical QED systems that have been realized use dye molecules as both the emitter and the reagent, such that reaction pathways access newly formed hybrid polaritonic states of the emitter-cavity system. With QD-based QED systems, the QD emitter is generally not the reagent in a photochemical reaction but rather the photosensitizer or the photocatalyst. The fact that the QDs in such a system would be intermediate chromophores may pose a challenge to the modification of the potential surface of a cavity-coupled photochemical reaction. We propose that a promising direction to explore to answer this challenge is to link the QDs to catalytic substrate molecules through so-called “exciton-delocalizing ligands” such as phenyldithiocarbamate and its derivatives.^87–89 When capped with exciton-delocalizing ligands, the wavefunction of the QD extends into the ligand capping layer through the generation of interfacial orbitals at the QD surface, resulting in a relaxation of exciton confinement and strong coupling with the substrate molecule. The creation of such interfacial states should result in the entire QD-delocalizing linker-substrate system acting as essentially one species that is coupled to the cavity. We therefore suspect that the QD and its ligand shell can serve as an intermediary between the substrate and the cavity and provide enough coupling to change the potential energy surface for the photocatalytic reaction.⁸⁹ 

Experimental systems demonstrating strong-coupling between a single QD or multiple QDs and a quantized electric field are few in number compared to the theoretical studies on such systems. The few experimental QD-based polaritonic systems in the literature suggest that more extensive research is warranted to fully understand how to optimize QED systems with QDs for applications such as quantum information, lasing, and new photochemical reactivity.

The authors thank the National Science Foundation (Grant No. CHE-1664184) and the Air Force Office for Scientific Research (Grant No. FA9550-17-1-0271) for support of this work. K.P.M. acknowledges a graduate fellowship through the National Science Foundation (Award No. DGE-1324585). K.A.P. was supported by the Department of Defense (DoD) through the National Defense Science and Engineering Graduate Fellowship (NDSEG) Program. J.C.S. acknowledges a graduate fellowship through the National Science Foundation (Award No. DGE-1324585).

The authors declare no competing financial interests.