Glazed buildings with large transparent surfaces are becoming increasingly common in modern architecture. They provide attractive aesthetics while maximizing the use of natural daylight, thereby enhancing occupants' well-being. However, extensive glazing can also compromise both visual and thermal comfort by allowing excessive solar radiation to enter buildings, leading to glare and overheating. In warm climates and during summer months, these effects are typically mitigated through increased use of energy-intensive air-conditioning systems.

Smart materials have emerged as promising solutions to transform conventional windows into static or dynamic optical filters capable of regulating solar light transmission and heat gain, thereby reducing the energy required to maintain indoor thermal comfort. Nevertheless, their widespread adoption in commercial and residential buildings remains limited by high production costs, poor adaptability to different climatic conditions, limited durability, and challenges related to large-scale manufacturing.

Herein, I will present a new class of transparent nanodroplet-based composite materials designed to produce cost-effective, scalable photo- and thermoresponsive smart window films and coatings. These materials are capable of self-adapting to changing environmental conditions without requiring external power sources or integrated electronic circuitry. The composites consist of polymer matrices incorporating oil-core microcapsules or paraffin particles and exhibit strong modulation of both visible and near-infrared light, simultaneously improving visual comfort and thermal regulation.

The materials are fabricated through sustainable, low-energy processes (e.g., doctor blade coating) from water-based formulations. Energy-saving studies performed on model houses demonstrate a significant reduction in solar heat buildup when smart windows are used instead of conventional clear glass. Finally, the same nanodroplet-based approach is extended to fabricate efficient luminescent glasses that guide and concentrate emitted light toward their edges, functioning as luminescent solar concentrators. These multifunctional materials not only filter and manage incoming sunlight but also convert part of the absorbed solar energy into usable electricity.

Halide perovskites are regarded as leading candidates for next-generation light-emitting diodes, photodetectors and solar cells owing to their outstanding optoelectronic properties, including tuneable bandgaps, strong absorption and narrow emission linewidths. In this presentation, I will introduce a methodology for extracting the effective complex refractive index of buried perovskite layers and employing these realistic optical constants for the design of advanced optoelectronic devices. First, we employ optical interference and Tamm-plasmon resonances to design LEDs with controlled spectral and angular emission.[1] Then, I will show how we apply the same design principles to resonant-cavity perovskite photodetectors, enabling efficient and spectrally tuneable narrowband detection.[2]  Finally, we model and optimise monolithic all-perovskite triple-junction solar cells, where current matching is maximised by design.[3]

1. Z. Y. Ooi, …, G. Vega, et al., Nature Communications 15, 5802 (2024).

2. Z. Y. Ooi, S. Nie, G. Vega, et al., ACS Photonics 12, 8, 4119–4129 (2025).

3. T. C.-J. Yang, …, G. Vega, et al., EES Sol. 1 (1): 41–55 (2025).

Upconversion of near-infrared (NIR) photons via quantum dot (QD)-sensitized triplet-triplet annihilation (TTA-UC) offers a promising route toward improved night vision, photovoltaic, and bioimaging technologies. Practical deployment requires efficient solid-state architectures, and bulk heterojunction (BHJ) thin films have emerged as a leading platform. Our group recently demonstrated a ternary BHJ system comprising surface-modified PbS QDs as sensitizer, TES-ADT as annihilator, and DBP as emitter — representing the current efficiency and spectral frontier of this architecture, with anti-Stokes shifts up to 0.75 eV [1]. A central challenge, however, is the tendency of TES-ADT to crystallize in the solid state, causing rapid degradation of upconversion efficiency over time. Here, we investigate the spatiotemporal evolution of upconverted photoluminescence in BHJ thin films using NIR laser excitation coupled to a motorized microstage and optical microscope, identifying the microscopic features that initiate crystallization and tracking its propagation across the active layer. Motivated by the need for greater photostability, we investigate the incorporation of polystyrene as an amorphous host matrix, suppressing crystallization and enhancing long-term stability. Together, these advances chart a clear path toward efficient, stable, and processable solid-state upconversion devices operating under incoherent illumination.

Colloidal quantum dot (QD) color-conversion films for next-generation micro-LED and wide-color-gamut displays face a fundamental photonic bottleneck; even when the intrinsic photoluminescence quantum yield approaches unity, roughly 80% of generated photons remain trapped within the film stack due to total internal reflection (TIR) at the high-index composite/air interface. Embedding TiO₂ scattering particles is a standard industrial strategy to disrupt TIR and extend the pump optical path length, but the competing effects of scattering anisotropy, reabsorption, and surface morphology make empirical optimization slow and unreliable.

We present a wavelength-dependent 2D Monte Carlo ray-tracing simulation that decouples these competing mechanisms in perovskite QD/PMMA composites embedded with TiO₂ scatterers. The model integrates full Mie theory (via miepython), Maxwell-Garnett effective medium theory for the QD-loaded matrix, and a dependent-scattering correction for dense particle ensembles, going beyond the commonly used Henyey-Greenstein approximation, which systematically misses coherent backscattering resonances critical to accurate efficiency prediction.

Three quantitative findings emerge. First, TiO₂ particles with radius ~100 nm yield a 3-fold enhancement in average pump absorption path length (512 nm to 1552 nm) and increase total pump absorption by 76.3% in the low-concentration regime, converting an otherwise optically transparent 1 µm film into an efficient absorber without increasing QD loading. Second, the same scatterers suppress parasitic photon recycling: in high-concentration films the reabsorption rate falls from 101.5% to 23% upon introducing scatterers at optimal loading, simultaneously boosting extraction efficiency and preserving the emission color point. We confirm this spectral stabilization experimentally via angle-resolved PL measurements on FAPbBr₃/PMMA/TiO₂ films, which show negligible spectral redshift out to 70° viewing angle. Third, the simulation reveals a critical role for surface topology: while volumetric scattering alone eliminates waveguide side-losses above 2% volume fraction, specular films hit a hard limit (~16.8% top EQE) set by pump back-scattering. Coupling a Lambertian top surface with a TiO₂ volume fraction of 20% pushes top EQE to 19.6%, a relative gain of 16.7%, while simultaneously broadening the process window so that >90% of peak performance is maintained over a wide range of scatterer concentrations.

These results provide a quantitative design blueprint, optimal TiO₂ radius, volume fraction, and surface texture, for scalable, chromatically stable QD color-conversion architectures for display applications.

Semiconductor nanostructures exhibit remarkable optical and physical properties, including high absorption cross sections, narrow emission full-width at half-maximum (FWHM), highly tunable emission profiles, high photoluminescence quantum yields (PLQYs), and superior photochemical stability [1–5]. Due to quantum confinement effects, these properties become strongly size- and dimension-dependent when the physical dimensions of the nanostructure are smaller than the exciton Bohr radius of the bulk semiconductor. Incorporating alloyed compositions further enhances spectral tunability across a broader wavelength range; for instance, the emission of CdZnSe-based nanocrystal heterostructures (NCHs) can be tuned from 470 to 650 nm simply by varying the core's Cd/Zn ratio. Alternatively, the emission spectral range can be expanded using multiple confined nanocrystalline heterostructures. Dual quantum-confined NCHs featuring a wide-bandgap barrier have attracted significant interest due to their unique dual-band emission under ambient excitation. To exploit this, researchers have engineered several spherical quantum dot/barrier/quantum well (QD/barrier/QWell) NCHs—such as CdSe/ZnS/CdSe and PbS/CdS/CdSe core/inner-shell/outer-shell nanocrystals. In these architectures, both the QD core (CdSe or PbS) and the QWell outer shell (CdSe) exhibit distinct emission bands that can be independently modulated by adjusting the core diameter and outer-shell thickness, respectively. Furthermore, Mews et al. developed colloidal double-quantum wells based on CdS–HgS systems. Using a surface-replacement strategy to insert HgS layers between CdS barriers, the authors observed a significant red-shift in the absorption spectra of the complex nanocrystals, indicating strong electronic coupling between the two quantum wells.

However, spectral tuning in these QD–QWell dual-domain NCHs remains challenging due to inherent limitations in epitaxial growth. Depositing uniform QW layers onto the highly curved, multifaceted surfaces of QD cores often results in broad emission bandwidths and unpredictable intensities. This bottleneck can be overcome by synthesizing quantum wires (QWs) onto the QDs. For instance, Lim et al. reported QD/quantum rod (QR) NCHs that achieved dual-band emission by growing CdSe nanorods over CdSe/CdS QDs separated by a ZnS barrier layer. They utilized a photoetching technique to precisely tune the quantum confinement dimensions (such as QR width) simply by adjusting the excitation wavelength.

Here, we report a simple, conventional bottom-up colloidal growth strategy to synthesize sophisticated quantum dot–quantum wire (QD–QW) heterostructures with targeted optical properties. Anisotropic CdZnSe/ZnS core QDs were fabricated first, followed by the growth of multiple quantum wires on their surfaces. These NCHs exhibit precise, continuously tunable dual- and triple-band emissions. Through this multi-step colloidal route, we nanoengineered CdZnSe/ZnS/CdSe structures featuring a QD core, a ZnS shell, and CdSe nanorods to form dual- and triple-quantum-confined NCHs. We provide a detailed mechanism for the formation of these CdSe-based quantum wires. Crucially, the quantum wire thickness depends strictly on the reaction temperature and is independent of the synthesis duration. Core QDs of three distinct sizes (6.0 nm, 7.5 nm, and 8.5 nm) were synthesized by varying the reaction time and precursor concentration. Nanowires were then grown by exposing these core QDs to a subsequent round of CdSe precursors at various temperatures, yielding QD–QW NCHs with two distinct emission bands. To further enhance spectral tunability, the QD–QW structures were exposed to an additional round of CdSe precursors at a lower temperature. This step produced triple-quantum-confined QD–QW1–QW2 NCHs exhibiting three emission bands. To achieve white-light emission, blue-emitting core QDs were integrated with primary and secondary QWs. Finally, these sophisticated nanostructures were fabricated into white-light-emitting LEDs, demonstrating an external quantum efficiency (EQE) of 0.5%, which is highly comparable to the current literature benchmark of 0.66% for CdSe-based nanostructures.

This talk will discuss new materials for very-large-scale photonic structures for photovoltaics, agrivoltaics, and radiative cooling. Emerging technologies for electricity generation and cooling are increasingly moving toward dual-use sites, such as building-integration and agricultural sites. This raises new challenges for light management, as strategies are needed to optimally harvest, split, and direct sunlight over broad spectral and angular ranges. This talk will discuss the development of new materials to capture sunlight over large areas. First, any light harvesting strategy based on metamaterials will require very large-area, high resolution patterning strategies that are sustainable and cost-effective. I will discuss an all-additive, roll-to-roll fabrication route to produce metasurfaces over m2 with highly tailored optical response. The second part will discuss the use of spectrum shifting films of nanocrystals embedded in polymers for agrivoltaic applications such as greenhouse windows. These films can be used either as luminescent solar concentrators for co-generation of electricity, or as stand-alone light filters to tailor the solar spectrum. I will discuss the selection of nanocrystal emitters and polymers, as well as the design tradeoffs in quantum yield and absorption, the role of luminescence, and the impact of location. Finally I will discuss some new materials for large-area radiative cooling based on polymer nanostructuring.

Engineered disorder is emerging as a powerful paradigm in modern photonics, enabling unconventional mechanisms for light transport, localization, and enhanced light–matter interaction beyond the limits of ordered photonic architectures [1–3]. In this framework, fractal silicon nanowire networks represent an ideal platform for multifunctional silicon optoelectronics, combining scalable fabrication, large surface-to-volume ratio, and broadband optical response with silicon technological compatibility.

In this invited contribution, we discuss recent advances in the development of two-dimensional fractal arrays of ultrathin silicon nanowires fabricated by metal-assisted chemical etching. The correlated fractal morphology promotes strong multiple scattering and efficient light trapping over broad spectral ranges, leading to extremely low reflectance, enhanced Raman scattering, and room-temperature photoluminescence [4–6]. Real-space and Fourier-space optical imaging reveal coherent and weakly localized Raman transport together with directional Raman beaming phenomena, providing direct visualization of photon diffusion and localization processes in disordered photonic media [5,6].

Beyond fundamental light transport physics, fractal silicon nanowire platforms enable several emerging applications in silicon optoelectronics and energy technologies. The strong enhancement of light–matter interaction supports highly sensitive optical biosensing, including amplification-free detection of viral targets [7], while the ultrahigh nanowire density and large electrochemically active surface area make these architectures promising candidates for energy-storage devices such as high-performance supercapacitors [8]. Additional functionalities are achieved through hybrid photonic configurations, including rare-earth active coatings for enhanced optical emission and telecom photonics applications.

These results demonstrate how engineered fractal disorder can transform silicon nanostructures into multifunctional photonic platforms for advanced light management, optical sensing, and sustainable optoelectronic technologies.

Focusing light with lenses played a pivotal role in the development of modern science, particularly in physics, biology and astronomy. Passive (non-absorbing) lenses are only able to focus direct (collimated) light and therefore cannot concentrate light from the thermodynamic perspective: any decrease in spatial extent must be compensated by an increase in the angular extent of the light beam to satisfy the second law of thermodynamics. This makes it impossible to use a lens to increase the brightness (photons Sr^-1m^-2s^-1eV^-1) of a beam. This is fundamentally one of the reasons lasers are so useful: they provide a high brightness light source. Lasing, however, is not the only mechanism to concentrate light. In principle light absorbing (active) lenses are capable of not only focusing but also concentrating light, increasing the brightness of a source. In order to satisfy thermodynamics, these lenses must compensate the increase in brightness (optical concentration) with a decrease in the photon energy (Stokes shift).[1] Although this principle has been experimentally validated, the performance of these luminescent concentrators falls remarkably short of their thermodynamic potential: all demonstrations of light concentration beyond 4X can improve theoretically by at least a factor of a million. This enormous gap between the thermodynamic and practical limit of light concentration stems from the requirement of emitted light to travel long distances through a strongly absorbing waveguide. Higher concentration factors require longer distances, which unavoidably magnify reabsorption and scattering losses. We propose a novel approach using carrier funneling and luminescent collimation, which decouples concentration from emission propagation distance, breaking the major practical limitation of luminescent concentrators. Finite-difference time-domain, transport and recombination simulations combined with realistic material properties of mixed halide perovskite film/microlens arrays demonstrate concentration factors above 290x, reaching more than 14% of the thermodynamic limit.

Perovskite Solar Cells (PSCs) have recently emerged as a highly promising alternative to conventional silicon photovoltaics. However, the toxicity and safety risks associated with the lead content of perovskites remain critical barriers to large-scale industrial mass production. Despite extensive research, entirely replacing lead without compromising power conversion efficiency (PCE) remains a severe challenge; consequently, minimizing the volume of the absorber layer presents a highly viable alternative strategy.[1]

In this work, we present an advanced wave-optical light-trapping (LT) architecture that enables cutting the lead content by half without compromising macroscopic device performance. This is achieved via the integration of front-contact nanophotonic voids that provide gradual effective refractive index matching at the air-cell interface, leading to broadband anti-reflection and intense forward light scattering into the underlying absorber film.[2] This electrically harmless LT scheme dramatically enhances short-wavelength photon absorption within thin (<200 nm) perovskite layers, yielding a substantial increase in short-circuit current density without introducing parasitic recombination or deteriorating other photovoltaic parameters.[3] The high-index nanophotonic features were fabricated using a highly scalable colloidal lithography technique, where key processing parameters (including microsphere diameter, reactive ion etching times, ITO thickness, and annealing profiles) were exhaustively optimized. The optimized nanostructured design produced a 20% increase in optical photocurrent density for a 200 nm thick perovskite cell, successfully matching the optoelectronic performance of a conventional, thick (400 nm) reference device.

Ultimately, this lead mitigation strategy provides a viable, post-process engineering path to circumvent toxicity regulations, acting as a crucial catalyst for the accelerated market deployment and commercialization of sustainable perovskite photovoltaics.

Single-junction organic photovoltaics (OPVs) nowadays have reached promising power conversion efficiencies, around 20%. Besides new materials, going beyond the current efficiencies could, in principle, be achieved by multi-junction devices, which promise a reduction in thermalization and absorption losses [1]. In this talk, we will present a multi-junction in-plane spectral splitting geometry that we call Rainbow solar cells and that aims at overcoming the limitations of stacked solution processed devices [2]. In the Rainbow geometry, a series of sub-cells are placed next to each other laterally, and illuminated through an optical component that splits the incoming white beam into its spectral components, thus matching local spectrum and absorption for each sub-cell. The fabricated n-terminal devices are capable of extracting the maximum power of each sub-cell without the need for current matching nor processing challenges.

We demonstrate the concept for a high and low band-gap sub-cells, obtaining an efficiency increase of around 30% of the Rainbow geometry with respect to our best single junction device [2]. Then, we show that ternary mixing can further enhance the overall efficiency in this type of device by providing means to tune the V_oc in narrow gap cells [3]. Monolithically integrated lateral cells have been fabricated using meniscus guided blade coating, and the corresponding PCE improves from 12.9% in single-junction devices to 15.9% in 2-junction devices (16.4% in simulations) and 17.3% in 3-junction devices (17.7% in simulations). Detailed balance analysis indicates that the potential of this geometry can be very high provided that high efficiency wide bandgap (2–2.5 eV) materials become available [3]. Finally, we use simulations based on real EQE and JV values to evaluate the potential of this technology for organic, silicon and perovskite solar cells and their combinations.

Perovskite-based tandem solar cells surpass the efficiency limits of single-junction devices. By stacking a wide-bandgap perovskite top cell with a low-bandgap bottom cell—such as silicon or a narrow-bandgap perovskite—these architectures enable more efficient utilization of the solar spectrum. In addition to optoelectronic enhancements, achieving high power conversion efficiencies requires precise light management to maximize photo absorption at minimized losses. In my presentation, I will discuss our advancements for different perovskite-based tandem technologies to reduce parasitic absorption and reflection losses across the tandem structure. These strategies utilize both the development of novel, optimized contact materials and advanced light management strategies such as textured interfaces, anti-reflective coatings and optical spacers.

In the most used architecture, all-perovskite tandem solar cells suffer from parasitic absorption losses in the recombination layer of ultrathin gold and PEDOT:PSS, strongly reducing their optical potential. Replacing these layers is therefore of high importance to access their full optical potential. Moreover, all-perovskite tandem solar cells -being a full thin-film technology- suffer from thin-film interference losses, which can be strongly suppressed with the use of nanostructured interfaces. With these optimizations, we were able to increase the photogeneration in each subcell by 1 mA/cm² and in my presentation, I will give an outlook of how the photogeneration in all-perovskite tandem solar cells can be further increased.

Achieving high photovoltaics conversion efficiencies in perovskite-silicon multi-junction solar cells requires a precisely tailored distribution of the solar spectrum between subcells. It calls for simultaneous optimization of front-side incoupling, forward and backwards scattering at the cell interfaces, reducing parasitic absorption losses, and at the same time optimizing the angular redistribution of scattered light to enhance light trapping. This requires a multi-scale approach in determining the optimum combination of layers and textures combining ray optical and wave optical methods.

In this work, we investigate light coupling and trapping in silicon–perovskite tandem solar cells with a focus on light trapping in the silicon subcell. We study different nano- and micro-texture types and their impact on reflection, transmission, light trapping and parasitic absorption losses in full tandem stacks. Using an angular scattering matrix formalism, we compute redistribution matrices of textured interfaces, allowing us to quantify the redirection and distribution of light in each of the absorber layers.

We analyze multiple scattering interactions between the front and rear interfaces in silicon and silicon-perovskite model systems featuring a random pyramidal front texture and a nanostructured metallo-dielectric rear diffraction grating that was recently used to achieve a record 36.1% efficiency for Si-based solar cells using a Si/III-V triple-junction geometry.^1,2 We then optimize the design of the rear diffractive grating, taking into account the angular redistribution of light refracted from the front interface into polar angles 23.3° and 64.8°. The detailed understanding of scattering channels arising from transmission and reflection at the front interface enables the identification of key design constraints for the optimization of the rear back-reflector, including minimization of plasmonic scattering losses and front-interface outcoupling. We establish a step-by-step optimization strategy based on controlling diffraction orders, angular selectivity and parasitic absorption at the rear interface. This combined understanding of front- and rear-interface scattering enables targeted design of rear nanostructures for increased optical path length in the silicon subcell and improved light management in silicon-perovskite and other multi-junction solar cells.

The implementation of textured interfaces in solar cells has been proven as an outstanding strategy to reduce optical losses and optimize the performance of solar cells. In perovskite-based solar cells, the application of solution processing deposition methods on textures - particularly on textures with micrometer dimensions - has been considered challenging for a long time due to the difficulties in growing a conformal perovskite layer on the textures. Recently, perovskite-based tandem solar cells with micro-textured interfaces were fabricated using standard solution processing methods [1-3] and showed a lot of potential.

This contribution will show the results on the implementation of periodic inverted micro-pyramidal texture on silicon heterojunction (SHJ) solar cells and single junction perovskite solar cells. Periodic inverted pyramidal textures have been discussed to have the potential to reach or even surpass Lambertian limit of light trapping. [4, 5] We show that the implementation of inverted pyramidal texture on SHJ solar cells, obtained by etching of silicon with potassium hydroxide (KOH) solution, reduces reflection losses by up to 4 mA/cm² in short circuit current density (J_SC) compared to its planar counterparts. Moreover, solution processing of perovskite single junction solar cells on this texture resulted in a fully textured solar cell interfaces, shown by atomic force microscopy (AFM) and scanning electron microscopy (SEM) images. The fully textured perovskite solar cell interfaces also reduce reflection losses of up to 1.2 mA/cm² in J_SC. [6] This contribution also aims at exploring complementary characterization methods for the textured perovskite absorber to understand differences in material properties of the film and on sub-micrometer scale, such as perovskite composition, crystallization, and electronic properties. The characterizations include X-ray diffraction (XRD) in Bragg-Brentano configuration, photoluminescence (PL) microscopy, SEM with energy-dispersive X-ray spectroscopy (EDX), and kelvin probe force microscopy (KPFM) of the top and buried interfaces of the perovskite. PL microscopy with bandpass filters revealed an indication that the perovskite grown on the texture’s valleys are rich in iodine, while the perovskite grown on the texture’s hills are rich in bromine. To prove this, spatially resolved elemental analysis with SEM-EDX was done on the perovskite absorber. SEM-EDX measurement on multiple sites confirmed that at the valleys of the pyramids, there are brighter perovskite domains with higher iodine signal and lower bromine signal compared to the surrounding darker domains. These measurements revealed that fully textured perovskite absorber deposited on periodic inverted micro-pyramids might have slightly different composition, morphology, crystallographic properties, and electronic properties depending on the spatial position within the texture, creating inhomogeneity within the perovskite absorber which might affect the resulting solar cells negatively. Furthermore, from the studies conducted on SHJ and single junction perovskite solar cells based on periodic inverted pyramidal texture, fully textured solution processed perovskite-silicon tandem solar cells based on this texture with improved optical performance was demonstrated.

This study explores how periodic inverted micro-pyramidal textures can be utilized for silicon and perovskite based solar cells to improve their optical performances. It was also revealed that while this texture allows solution processing of perovskite absorber to result in a fully textured interface, the resulting perovskite might differ in its composition and electronic properties depending on the spatial position within the texture. The results of this study lay the foundation for the development of optically and electronically optimized perovskite-silicon tandem solar cells.

While luminescence is ubiquitous in photovoltaic research, be it as a diagnostic tool or fingerprint of efficient device operation, the emissive process is seldom treated on equal footing with (re-)absorption in the simulation-based analysis of photovoltaic device operation. On the other hand, there is an increasing variety of opto-electronic devices that rely on the re-absorption of internal emission: not only do the most efficient solar cells benefit from photon recycling (PR) and luminescent coupling (LC), but radiative energy transfer also plays a vital role in spectral conversion devices, laser power converters, and thermophotonic (TPX) applications.

In our contribution, we review our comprehensive approach to the simulation of re-absorption effects in photovoltaic devices, both on the level of the optical detailed balance limit [1] and integrated with full drift-diffusion charge transport [2]. After revisiting key results of the literature in the light of our theory – such as the enhancement of V_OC by PR [1,2] and the role of PR and LC in tandems [3,4], we demonstrate the versatility of the framework by application to the modelling of UV downshifting encapsulations for silicon solar cells, III-V multi-junction laser power converters, and near-field TPX devices.     

Abstract:  
The power conversion efficiency (PCE) of single-junction solar cells is fundamentally limited to approximately 33% by intrinsic energy loss mechanisms, including thermalization, below-bandgap, and Boltzmann losses. Among these, Boltzmann losses are of particular significance because their mitigation through optical confinement can theoretically increase the efficiency limit up to 43%. Perovskite solar cells (PSCs) are especially well suited for addressing these losses owing to their high photoluminescence quantum yield (PLQY). While most of the previous studies have focused on suppressing non-radiative recombination through materials and interface engineering, comparatively little attention has been given to reducing the unavoidable radiative losses caused by the angular mismatch between absorbed and emitted photons. In this work, we designed, modelled, and experimentally realized the photonic structures to restrict the optical emission cone, and minimizing these intrinsic losses. As a result of which inverted PSCs incorporating the proposed photonic structure exhibited an increase in PCE from 20% to 20.7%, accompanied by a 17mV enhancement in open-circuit voltage (Voc), primarily attributed to the suppression of unavoidable radiative losses. These findings establish photonic engineering as an effective strategy for bringing PSCs closer to their fundamental efficiency limit by addressing a loss mechanism that cannot be eliminated through conventional materials or interface optimization alone.

Keywords: Perovskite, Photonic Strucures, Boltzmann Losses, Radiative Losses, Photoluminescence Quantum Yield