1. Introduction

Spectrally selectivity is critical in many optoelectronic applications, such as optical imaging [1,2], target recognition [3,4], chemical detection [3,5–7] and solar energy harvesting [8–11], but is difficult to achieve in traditional semiconductors which typically absorb at all energies above their band gaps. Specific examples of technologies requiring fine-tuned spectral responsivity include finite bandwidth photodetectors and materials with controlled transparency windows for multijunction and transparent photovoltaics [8,9]. Spectral-selectivity can be employed for heat management in solar cells or other optoelectronic devices by reflecting unwanted wavelengths that would otherwise be parasitically absorbed in the contacts or other device layers [12–15]. Common solutions to this problem include using external filters for photodetectors [5], which come with the cost of adding complexity to the system, and empirically controlling the thicknesses of each absorbing material in tandem solar cells to realize current matching, which often sacrifices photocurrent output [16–18]. Here, we propose a solution that achieves controlled spectral selectivity within the absorbing material itself, i.e. the photogenerative material that is responsible for energy conversion or transfer of the absorbed photon energy: using photonic crystals to engineer the photonic band structure in absorbing media to directly control the wavelength-dependent absorption, reflectivity and transmissivity.

Photonic crystals (PCs) are materials with periodic variations in their dielectric functions, potentially creating a photonic band gap, a range of frequencies in which photons are forbidden to propagate. This compelling mechanism enables PCs to be used to manipulate light flow in many applications including optical communications [19–21], computing [22–25], and optoelectronics [26–30]. In addition to artificial structures, many examples of PCs can be found in the natural world, enabling effects such as the structural colors of butterfly wings and beetles [31–33].

Most of these examples use PCs with photonic band gaps lying in the naturally non-absorbing range of the materials, i.e. below the electronic band gap where the material behaves like a simple dielectric, although photonic band structures can be straight-forwardly tuned in frequency by adjusting the length scale of the dielectric function periodicity (lattice constant) [34]. Positioning the photonic band gap of a PC in the absorbing region of a material presents complications due to absorption being viewed as a loss mechanism for many applications. However, optoelectronic applications such as photovoltaics or photodetectors rely on semiconductor absorption and photogeneration as vital operating mechanisms, and the possibility of using photonic band engineering within the absorbing region represents a potential new spectral tuning mechanism. The concept is illustrated in Fig. 1(a) and 1(b). Previous work on engineering photonic band gaps in lossy materials includes forming PCs from metals that have shown diminished reflection peaks with increasing absorption [35]. Initial work on forming PCs within the photoactive layers for solar cells has been proposed, focusing on utilizing the PC structures for spectrally-selective light trapping [36] and absorption enhancement [36–39] based on density of states modulation. For wavelength selective-absorption, band-pass absorbers made from dye-glass [40] have been widely implemented for decades, and in recent years, metamaterials based on periodic plasmonic structures have been demonstrated with very high absorbance within the visible range for solar-thermal applications [41]. These implementations typically focus on extending the spectral response in their respective systems by combining structures with engineered responses in different spectral ranges. However, rather than focusing on absorption alone, a comprehensive method for inducing spectral-selectivity that aims to enable wavelength-dependent absorption, reflectivity and transmittivity simultaneously, including not just absorption enhancement but also suppression through the use of controlled transparency windows, has yet to be demonstrated. Here, we describe how embedding PCs in photogenerative materials could offer a handle for controlling spectral features across multiple wavelength bands and dynamic ranges in complicated optoelectronic applications such as multi-junction solar cells.

 

Fig. 1 Schematic of a generic 2D slab photonic crystal illustrating the spectral tuning concept (left). The “in-plane” photonic band structure is used to generate spectrally-selective reflectivity, transmissivity and absorption for target optoelectronic applications. Broadband light (white in color) is incident on the slab, with specific frequency components strongly coupled to the resonance modes of the slab (yellow), resulting in spectrally-selective transmission (blue) and reflection (red). A hypothetical photonic band diagram for the generic slab structure (photonic bands are shown in yellow; the light line is shown in blue in the center panel) and “out-of-plane” transmittance (blue) and reflectance (red) spectra at normal incidence are sketched on the right side of the Fig. 1. The green stripes show direct correlations (coupling) between the sharp resonance features in the transmittance and reflectance spectra and the photonic band states at the $\text{γ}$-point. A Brillouin Zone diagram for the hypothetical structure is shown above the photonic band diagram sketch.

Download Full Size | PDF

Spectral-selectivity is particularly applicable for the design of color-tuned materials with controlled transparency windows for multijunction and transparent photovoltaics. Infrared (IR) sensitive materials, such as small-bandgap semiconductors, absorb strongly at all energies above their band gaps. Molecular materials, such as organic semiconductors [42], typically have finite-bandwidth absorption peaks in the visible and UV, but IR-only responsive materials are rare. To incorporate IR-sensitive materials into multijunction solar cells, they must be positioned at the back side of a standard solar cell device to permit the absorption of visible radiation by the front cell, and the thicknesses of both cells must be fine-tuned to achieve current-matching. An ideal material for incorporating multijunction functionality into current single junction photovoltaic technology would have an absorption profile that was both spectrally tunable and finite in bandwidth, thus offering flexibility for both current-matching and device design.

The concept of using the photonic band structure in a slab-type PC to control the reflection and transmission of external propagating fields is illustrated in Fig. 1. In Fig. 1, we present a system consisting of a 2D PC “slab” structure in which a periodic lattice of air holes is drilled in a semiconductor. The “out-of-plane” incident, reflected, and transmitted fields are highlighted and interact with the “in-plane” resonant field with properties determined by the PC structure. Here we use “in-plane” to describe physical structures or fields and states that are bound by or mostly concentrated in the slab itself, and “out-of-plane” to describe fields and waves that propagate indefinitely outside of the slab structure. The “in-plane” fields and states compose the band structure of the slab, and they can couple to the “out-of-plane” incident waves at the γ point of the Brillouin Zone, shown on the right side of the Fig.. We sketch hypothetical reflection and transmission spectra at normal incidence alongside a hypothetical (generic) in-plane band diagram for a structure such as that shown in Fig. 1 to illustrate the coupling between the in-plane photonic bands and the out-of-plane reflection and transmission profiles. The transmission and reflection spectra should consist of a smoothly varying background that resembles a Fabry-Perot interference spectrum [43], with sharp and asymmetric resonance features on top. As will be discussed in the next section, the coupling between incident waves and the photonic bands sharing a lateral wave-vector results in resonance features in the transmission and reflection spectra of the slab. This coupling gives rise to the potential for tuning the “out-of-plane” spectral selectivity by tailoring the band structure of the PC slab, to achieve desired absorption, reflection and transmission profiles.

In this work, we use finite-difference time-domain (FDTD) simulations and Fourier modal methods [44–46] to quantify the effect of material absorption on a slab PC with relevant photonic bands that fall above the electronic band gap of the semiconductor slab material. We then use the insights gained from the simulations to design a PC structure in a solution-processed semiconductor, based on a PbS colloidal quantum dot (CQD) thin film, that strongly absorbs in the infrared but transmits visible light more strongly than in the non-structured semiconductor. This type of material could enable visible-blind infrared photodetectors without external filters, and it could also allow for flexibility in current-matching in a tandem solar cell.

2. Simulations

To study the effect of material absorption on photonic band structure, we based our simulations on the well-studied 2D GaAs slab PC structure [26], which consists of a triangular lattice of air pillars in a semiconductor slab of finite thickness. In FDTD simulations, we are able to artificially adjust the strength of the absorption through control of the imaginary part of the dielectric constant (${ϵ}_I$) as long as we keep the real part (${ϵ}_R$) constant, which is equivalent to varying the real and imaginary parts of the refractive index (n, k). Dispersion is not explicitly considered in the test-case model since it results in difficulty in satisfying the Kramers-Kronig relations [47]. The completely non-absorbing control case uses a material that has ${ϵ}_R$ set to 13, meant to approximate the average value of n for GaAs across the relevant frequency range, and ${ϵ}_I$ set to zero. We then gradually increase the value of ${ϵ}_I$ in the simulations, keeping ${ϵ}_R$ constant, in order to systematically quantify the effect of dissipation on the photonic band structure.

We use FDTD simulations to calculate the frequencies of the modes that can exist in the structure beyond the initial transient phase in our artificial materials to reconstruct their photonic band structures. The broadband field profile of the excitation source is chosen to ensure that all modes of interest are excited. Randomly distributed time monitors collect the time-resolved field data. Destructive interference causes rapid decay of non-resonant fields, while the excited modes of the structure resonate with varying decay rates for the bulk of the simulation time. The frequencies of these modes are extracted via fast Fourier transform (FFT).

Details of the FDTD simulation method for band structure simulations [48] are as follows: The simulation volume consists of an integer number of unit cells of the 2D periodic structure, and the volume is extended in the z-direction symmetrically above and below the slab by approximately 10 lattice constants. Bloch boundary conditions are used for the x and y (in-plane) directions and perfectly matched layers (PMLs) with symmetric or anti-symmetric boundary conditions are used in the simulations corresponding to the even (TE-like) or odd (TM-like) mode polarizations, respectively, for the z-direction. Identical broadband dipole sources with random polarizations are used to excite the modes and randomly distributed throughout the simulation volume. Conformal meshing is used near material interfaces, and Maxwell’s integral Eqs are used to account for structural variations within a single mesh cell. We apodize the loaded time signals from each time monitor for each field component, with a Gaussian-shaped windowing function used to only consider the portion of the time signal following the source pulse injection and before the simulation is cut off. The resulting FFT of the apodized signal is a spectrum with peaks at the resonant mode frequencies, corresponding to allowed photonic bands. The energy spectra of Fourier-transformed time signals for each field component of every time monitor are summed to ensure that we identify all of the resonant frequencies even if some of the randomly placed time monitors are located at the node of a mode. The simulations are repeated for each Bloch vector value, K, and frequency peaks for each K that meet the threshold tolerance are retained.

In our simulations, we chose a semiconductor slab thickness of 125 nm and tuned the lattice constant (a = 250 nm) and radius (r = 60 nm) of the air holes to produce a number of relevant photonic bands within the visible regime, or above the electronic band gap of GaAs. The model structure is shown in the inset of Fig. 2(c). The simulated band structure for the non-absorbing control case with both even and odd modes is shown in the top left panel of Fig. 2(a), followed by a series of photonic band structures with increasing k. The photonic band structure for a GaAs PC slab including full dispersion [49] in the refractive index model is shown in Fig. 2(b).

 

Fig. 2 (a) FDTD-calculated photonic band diagrams for the structure shown in (c) with media loss (absorption) varying from ${ϵ}_{\text{I}}=0$ to ${ϵ}_{\text{I}}=3.61$ and constant ${ϵ}_{\text{R}}$ = 13, with corresponding imaginary part of the refractive index also indicated. The light lines are plotted in blue. The color scale is in arbitrary logarithmic units corresponding to the field intensity. (b) FDTD-calculated photonic band diagram for the same structure for a GaAs slab medium (the dielectric constant includes dispersion in this case). (c) Quality factor for 5 selected modes, indicated by the blue markings at the γ point in the top left panel of (a), as a function of loss in the material. Inset: model of the simulated structure, a triangular lattice of air holes in a semiconductor slab with 120 nm diameter, 250 nm lattice constant, and 125 nm slab thickness.

Download Full Size | PDF

As can be seen from Fig. 2(a), the frequencies of the photonic bands are almost unchanged as dissipation is added to the optical model, if ${ϵ}_I$ remains small compared to ${ϵ}_R$, although the relative band strengths are reduced with increasing material absorption. As material loss increases, the widths of the bands are broadened, and the clarity of the higher order bands decreases faster than that of the lower order bands, which is expected from the shorter absorption lengths at higher frequencies present in the model. The rough maintenance of the photonic band frequencies in the presence of weak material loss can be understood using perturbation theory applied to the PC master Eq [26]. Adding a small imaginary part to the dielectric function, $ϵ$, results in the addition of an imaginary part to the resonance frequency, ${\text{ω}}_0={\text{ω}}_0-\text{iγ}$, which consequently adds to the linewidth of the Lorentzian resonance profile and reduces the resonance peak height.

To quantitatively study the properties of photonic bands in dissipative structures, we calculated quality factors, Q, for the individual bands in the structure shown in Fig. 2: ${\text{Q}}_{\text{i}}={\text{ω}}_0/{\text{Γ}}_{\text{i}}$, where ${\text{Γ}}_{\text{i}}$ is the energy decay rate, or the band resonance linewidth, of the ith photonic band at the γ point of the periodic structure. Modes at the γ point are above the light line, and because of the finite thickness of the PC slab, they are radiating modes at that point that can couple to external propagating fields. Here, we chose five modes capable of coupling to plane waves at normal incidence angles, i.e. they can be excited by incident plane waves and radiate energy to reflected and transmitted plane waves. For this reason, these bands dissipate energy from the slab and have finite $\text{Q}$ even without the presence of material absorption. Figure 2(c) shows the quality factors for the five bands as a function of increasing ${ϵ}_I$ or $k$ in the material. All quality factors exhibit similar decays as a function of ${ϵ}_I$ and can be well fitted by the function ${\text{Q}}_{\text{i}}=\frac{{\text{Q}}_0{\text{Γ}}_0}{{\text{Γ}}_0+{\text{α}}_{\text{i}}{ϵ}_{\text{I}}}$, where ${\text{Q}}_0$ is the quality factor of the lossless structure, and ${\text{α}}_{\text{i}}$ is a constant that depends on the spatial distribution of the ith mode [26]. As loss in the material (${ϵ}_I$) increases, the differences in $\text{Q}$ between different modes decrease, corresponding to a “smearing” and overlapping of the photonic bands until they eventually become indistinct at the limit of very high material absorption.

Understanding the interactions of external propagating fields with the in-plane photonic band structure of a slab-like PC, specifically the spectral reflection and transmission of a PC thin film, is a critical step in using them for absorbing optoelectronic applications. These interactions involve power transfer from external fields to radiating modes within the slab and vice versa. Such systems can be modeled as resonators interacting with external ports using coupled-resonator theories [50]. In PC thin films of interest for optoelectronic applications, the periodicities are always smaller than the wavelengths of interest; therefore, for plane wave sources, no diffraction orders exist. Consequently, all fields and modes that interact with an incident wave are contained in the reflected and transmitted waves with the same in-plane wave vector as the incident wave, and the in-resonance slab radiating modes. Waves impingent at normal incidence on a slab-PC therefore compose a system that can be fully and concisely modeled.

Figure 3 shows transmission and reflection spectra for absorbing and non-absorbing PC structures calculated via Fourier modal methods [44–46] at a fixed in-plane wave vector with polarization along one of the reciprocal lattice vectors. The smoothly-varying background curve resembles the spectrum from a simple Fabry-Perot cavity consisting of a uniform continuous media sandwiched between two mirrors, with additional sharp resonance features added on top. The resonances occur at the same frequencies as the in-plane PC radiating modes and are asymmetric with negative and positive features due to the nature of Fano resonance behavior [50]; these features include dramatic increases in transmission on one side of the resonance, even in the case with material absorption. The transmitted and reflected fields directly couple to the incident field while simultaneously indirectly coupling to the radiating mode of the photonic crystal excited by the incident field. Such resonance phenomena are well explained by temporal-coupled wave theory [50–52]. We note that not all modes in the photonic band structure can be excited with incident plane waves, due to restrictions in symmetry and polarization [50].

 

Fig. 3 FMM-calculated transmission (solid lines) and reflection (dashed lines) spectra (bottom) for a triangular lattice slab photonic crystal with r = 0.24a, t = 0.5a and $ϵ=13$ (blue and yellow spectra) and $ϵ=13+0.3\text{i}$ (red and purple spectra). The incident field is perpendicular to the slab structure. The corresponding FDTD-calculated band structure for the $ϵ=13$ case is shown in the top panel (light line plotted in white). The resonance regions are highlighted and associated with the modes at the $\text{γ}$ point in the band structure.

Download Full Size | PDF

In the absence of absorption, the Fano resonance features are narrow and sharp, due to the intrinsic long lifetimes of the lossless radiating modes. When absorption is added to the model, the lifetimes of the radiating modes are reduced, resulting in broader and weaker resonance features, which originate from the same effects as the reduction in Q factor seen for the modes in the band structures. The resonance frequencies for the same structure with a non-zero ${ϵ}_I$, set to 2% of ${ϵ}_R$, are almost unchanged, in accordance with the FDTD photonic band structure simulations. The FDTD-calculated photonic band structure for the ${ϵ}_R=0$ case is plotted next to the reflection and transmission spectra in Fig. 3 to illustrate direct spectral correspondence between the out-of-plane Fano resonance features and the in-plane photonic crystal bands. The overall reflection and transmission are both slightly reduced in the case with absorption compared to the case without absorption, as would be expected for a uniform lossy slab.

Although the resonance features are less apparent in the transmission and reflection spectra of the absorptive structure, the interactions of the PC medium with the radiating resonance modes are not necessarily weakened. Applications such as photovoltaics and photocatalysis that depend on absorption of the photoactive material could benefit from this phenomenon of increased absorption of the resonance modes. The amplitude of the ith steady-state resonance mode in the presence of material absorption can be approximated from temporal-coupled wave theory as shown in Eq. (1) [51,53]:

which can be maximized at the resonance frequency ${\text{ω}}_0$, if the absorption strength is comparable to the radiative strength. In the presence of material absorption, the widths of the resonances are also broadened so that a larger range of frequencies is capable of inducing stronger absorption in the media. Additionally, the integrated absorbed power associated with a specific resonance (photonic band) always increases with increasing material absorption strength.

Based on this analysis, using 2D PC slabs in absorptive materials should enable spectral modulation, including absorption and transmission tuning, with careful control of the PC parameters. Broadband transmission and reflection selectivity is controlled primarily by the average optical properties of the slab, i.e. the effective refractive index. The smoothly varying background in the reflection spectra can be very accurately fitted assuming an optically uniform slab with slight dispersion in the dielectric constant that varies around the average dielectric constant of the PC slab. Broadband transmission and reflection tuning is therefore effectively only dependent on the volume ratio of the periodic voids (low-index inclusions), the slab thickness, and the properties of the high-index material. The spectral locations of the sharp resonant features corresponding to the photonic bands within the material, on the other hand, are highly dependent on the PC structural properties, such as the periods and shapes of the voids, not on the slab thickness and high-to-low index material volume ratio. Additionally, due to the optical scalability of PC structures in the absence of strong dispersion, specific spectral features can be easily shifted to preferred frequency ranges simply by scaling the structure accordingly. The addition of material absorption allows abrupt spectral features to be smoothed and broadened without significant shifts in frequency or decreases in the absorbed incident power, an additional tool that can be used to tailor transmission and reflection profiles.

3. Experimental demonstration of spectral tuning using PCs in strongly absorbing materials

Our simulation results indicate that the photonic band structure can be at least partially preserved in a PC fabricated in a strongly absorbing medium and that this in-plane band structure has a strong effect on the out-of-plane transmission and reflection spectra. We sought to experimentally demonstrate this spectral tuning mechanism by fabricating a proof-of-principle PC structure in a strongly absorbing material with response in the infrared portion of the spectrum.

We chose to use a PbS CQD thin film as the absorbing media because of the demonstrated infrared absorption, facile solution processability, and applicability in many optoelectronic applications, including photodetectors [54], LEDs [55] and solar cells [56,57]. We used nanosphere self-assembly [58,59] to construct a monolayer triangular lattice structure out of polystyrene beads, which served as the low-index material, and infiltrated them with PbS CQDs as the high-index absorbing material to form a photonic crystal-CQD (PC-CQD) structure. We optimized the PC-CQD film for transmittance in the visible regime using FDTD simulations. Our simulations aimed to mimic the realistic system by incorporating slight non-uniformity in the large-scale film thicknesses. We calculated the transmission by averaging the simulation results for 11 different film thicknesses for the PC-CQD case (250 ± 50 nm) and 3 different film thicknesses for the CQD control case (200 ± 10 nm), based on our experimental thickness measurements. As shown in Fig. 4(a)-4(b), the PC-CQD film consisting of 250 nm beads in a monolayer triangular lattice array infiltrated with PbS CQDs displays a slight enhancement in visible transparency and a slight decrease in near-infrared (NIR) transparency compared to the control CQD film in both the simulation and experimental spectra. The electrical field profiles at the transmittance peak and valley of the 250 nm PC-CQD film are shown in Fig. 4(c) and 4(d), respectively. At the transmittance peak, the field is mainly confined within the low-index dielectric material, whereas at the transmittance valley, the field interacts more with the high-index absorbing media and thus more energy is absorbed at that wavelength.

 

Fig. 4 (a) FDTD-calculated transmittance for a control CQD film and a PC-CQD film. The inset is the PC-CQD structure: a triangular lattice monolayer of polystyrene beads infiltrated with PbS CQDs. The control CQD film is 200 nm thick on average, and the PC-CQD film consists of 250 nm diameter beads in a triangular array with a lattice constant of 250 nm; the space around the beads is filled with CQDs to form a 250 nm thick film on average. The spectra are averaged over several thicknesses to simulate roughness. The PC-CQD film shows a slight enhancement in visible transparency compared to the control CQD film. (b) UV-Vis-NIR spectrophotometric transmittance spectra of the PC-CQD film and the control CQD film, showing qualitative agreement with the FDTD calculations. Absolute difference in transmittance can be attributed to large-area non-uniformities in the films. Inset: Top-view SEM image of the PC-CQD structure consisting of mildly-etched self-assembled polystyrene beads infiltrated with PbS CQDs. (c) FDTD-calculated cross-section of the spatial electric field profile at the transmittance peak (d) valley. (e) Top-view SEM image of the etched bead array before CQD infiltration. The inset is a photo of the 1 inch x 1 inch bead array on a glass substrate before CQD infiltration. Large-scale order can be inferred from the strong iridescence of the structure.

Download Full Size | PDF

To fabricate the PC-CQD film, we started by treating glass substrates with an O2 plasma to make them hydrophilic and spin-cast 50 μL of an aqueous solution of polystyrene beads at a concentration of 10% w/v purchased from Magsphere at a speed of 500 rpm for 10 seconds, followed by a 2-minute 700 rpm drying step. The beads self-assemble to form a close-packed triangular array with a lattice constant equal to the bead diameter. We then applied an O2 plasma etching step at 120 W for a few minutes to open up space between the beads, adjusting the time to control the ratio of the bead radius-to-lattice constant. A scanning electron microscope (SEM) image of the bead array after etching is shown Fig. 4(e), and the inset contains a photograph of the 1 inch x 1 inch sample, which shows strong iridescence from the beads, indicating large-scale order. We synthesized oleic-acid capped PbS CQDs following previously published methods [60] and deposited the control CQD films from octane at a concentration of 50 mg/mL via a layer-by-layer spin-casting and 3-mercaptopropionic acid (MPA) ligand exchange process [60] to build up a film of the desired thickness. We used a lower concentration CQD solution (10 mg/mL) while keeping the concentration of the MPA-in-methanol solution at 1% by volume, to promote infiltration of the CQDs into the bead array. An SEM image of the PC-CQD film is shown in the inset of Fig. 4(b), showing preservation of the bead array and infiltration of the CQDs.

Optical transmittance spectra of the PC-CQD and control films are shown in Fig. 4(b) and were measured by placing samples at the entrance of an integrating sphere in a UV-Vis-NIR spectrophotometer. The experimental data is in rough qualitative agreement with the FDTD simulation results: the PC-CQD film shows a slight enhancement in visible transparency with a peak visible in both the experimental and simulation spectra at approximately 630 nm. The peaks and valleys of the experimental spectra are broadened and reduced in intensity compared to the simulated spectra, most likely due to non-uniformity across the films and significant surface roughness that results in reduction of interference effects. Although preliminary, these results demonstrate that photonic structuring in strongly absorbing materials can result in significant modulation of the optical spectra which could be a useful tuning knob for optoelectronic applications. Future work will involve complete photonic band structure calculations to identify optimal structures that can be fabricated using CQD materials with targeted spectral properties for specific device applications.

4. Summary and outlook

We developed and analyzed a new strategy for tuning the spectral selectivity of optoelectronic thin films: using photonic band engineering in strongly absorbing materials in which in-plane photonic bands are used to control the spectral properties of the out-of-plane reflection and transmission spectra. We analyzed a model system composed of a semiconductor-based slab photonic crystal in which the photonic bands of interest are located in the absorbing region of the material. By artificially varying k in FDTD and FMM simulations, we were able to quantify the impact of absorption on the photonic band structure. Specifically, adding absorption had little impact on the frequency of the photonic bands, although the widths of the bands were broadened and the quality factors of the in-plane modes decreased and saturated with increasing material absorption. Our FMM analysis showed that coupling between the photonic bands at the $\gamma$ point and normal-incidence wave induces sharp resonance features over the smoothly varying background in the reflection and transmission spectra, which can lead to strong frequency-dependent variations in the reflectivity and transmissivity associated with Fano resonances, even in the presence of material absorption. These results indicate that PC structures in strongly absorbing media can be used to produce spectrally selective optoelectronic thin films for targeted applications by careful adjustment of the lattice parameters.

Experimentally, we demonstrated the use of photonic structuring to tune the transmission spectrum of a strongly absorbing material by fabricating a proof-of-principle structure consisting of a self-assembled polystyrene bead monolayer infiltrated with PbS CQDs. The PC-CQD structure showed both near-infrared absorption enhancement and visible transparency enhancement over a control homogeneous CQD film of the same thickness, qualitatively matching predictions.

Future work will focus on extending these results by calculating full photonic band structures for solution-processed systems and including realistic dispersion in the optical models. We will use the insights gained from this study to design spectrally-selective photoactive optoelectronic films for targeted applications such as narrow-band infrared photodetectors and infrared solar cell materials for multijunction photovoltaics. The platform described here should form the basis for a new way to think about using photonic band structure engineering to control the spectral selectivity of strongly absorbing materials.