Commercial silicon solar cells have achieved a record efficiency of 26.8%. Due to their indirect bandgap nature, silicon cells require a thickness exceeding 150 μm for sufficient photon absorption. To build more cost-effective alternatives, the photovoltaic (PV) community has devoted considerable efforts to identifying emerging semiconductors capable of efficient light absorption in much thinner films. Perovskites, known for their large absorption coefficients, fulfill this requirement by enabling effective photon absorption within submicrometer-thin films. Over the past decade, we have witnessed a dramatic increase in the efficiency of perovskite solar cells, with the current record standing at 26.1%. This impressive progress is largely attributed to the advancements of carrier management in FAPbI3-based cells. FAPbI3 outperforms its perovskite counterparts because of its narrowest bandgap among lead-based options, thus maximizing solar spectrum absorption. However, the room in this avenue is increasingly narrow because FA cation has reached the limit of the Goldschmidt tolerance factor and iodide already presents the narrowest bandgap among halide anions. This fact motivates us to explore the potential of the other alternative direction—resonant photonics management.
Recently, Prof. Yi Hou's team (Jiangang Feng and Xi Wang et al.) at National University of Singapore leveraged resonant photonics to enhance light absorption and photocurrents in perovskite solar cells. Firstly, we assessed the potential for improving light management by comparing state-of-the-art perovskite solar cells with the theoretical limits set by the Shockley-Queisser (SQ) limit. Using a perovskite composition FAPbI3 with a bandgap of 1.48 eV, we found that the SQ limit for the short-circuit current (Jsc) is 29.7 mA/cm². This figure exceeds achieved Jsc values in existing perovskite cells by over 3 mA/cm², indicating a significant deficiency in light absorption. This shortfall is further evidenced by a wider PV bandgap observed in the external quantum efficiency (EQE) spectrum. Typically, the PV bandgap, which is closely related to light-harvesting efficiency of a solar cell, is 40-80 meV wider than the material optical bandgap in perovskites. This suggests that light absorption is particularly inefficient near the band edge, where the absorption coefficient is significantly lower than at shorter wavelengths.
Given the significant potential for further efficiency improvement, we begun to explore possible routes for effective photon management. Traditional photon management strategies in solar cells, such as anti-reflection coatings and textured structures on silicon surfaces, operate within the ray-optics regime. These ray-optics methods are constrained by a theoretical upper limit for light-absorption enhancement, denoted as 4n2 (where n is the refractive index), as established by Eli Yablonovitch in 1982. The Yablonovitch limit well explains the modest efficiency gains seen in perovskite solar cells that employ textured structures and diffraction gratings. Beyond ray-optics, we introduced resonant photonic structures that operate in the wave-optics regime, aiming for stronger light-matter interactions in perovskite solar cells. These resonant photonic structures, such as photonic crystals, resonant gratings, and metasurfaces, have laid the foundation for modern photonics and offer promising applications in miniaturized lasers, integrated photonic circuits, nonlinear and quantum optics. However, the integration of resonant photonics into perovskite solar cells remains a challenging endeavor.
This paper was the first to integrate resonant photonics with perovskite solar cells. By designing subwavelength resonant gratings, we introduced strong guided-mode resonances (GMRs) near the perovskite band edge, facilitating momentum matching between guided modes and free-space light. These gratings allow normally incident light to couple into GMRs, enabling efficient light absorption through transverse light propagation via total internal reflection (Figure 1). One limitation of resonant photonics is its narrowband response, which hampers efforts to achieve broad-spectrum light absorption. To overcome this, we introduced the concept of Brillouin Zone (BZ) folding through the design of a supercell grating. By extending a single unit cell to a supercell, we achieved BZ folding in momentum space, causing high-momentum photonic bands to fold back to the BZ center. This results in an additional resonance mode in both orthogonal polarizations, providing new channels for strong light confinement (Figure 1c). In contrast to accidental discoveries of multiple photonic bands, this BZ folding strategy offers a deterministic method to increase the photonic density of states. Thanks to these resonances, we observed strong light absorption near the perovskite band edge with low absorption coefficients.
To experimentally validate this optical design, we employed a nanoimprinting method that transfers nanostructures from a silicon template to perovskite materials. Completing within just 3 minutes, this process enables the scalable fabrication of nanophotonic structures over a 1 cm2 area. The reusability of the template adds to the cost-effectiveness of this method. Through a series of characterizations, we established a link between enhanced absorption and optical resonances, confirming that resonant structures have no adverse impact on perovskite quality in terms of crystallinity and defect density. Consequently, we successfully narrowed the PV bandgap by 35 meV, achieving an 18 nm redshift of the band edge and an EQE-integrated Jsc of up to 26.0 mA/cm2 in resonant solar cells (Figure 2). The strong absorption at band edge, together with efficient carrier transport in submicrometer thin perovskites, produces an efficiency of 24.4% for solar cells.
This paper pioneers a new avenue by leveraging resonant photonics to enhance perovskite solar cells. The approach is not only effective for controlling light absorption—beneficial for solar cells and photodetectors—but it is also adaptable for light-emitting applications such as LEDs and electrically pumped lasers. We provided evidence that these resonant solar cells also show enhanced electroluminescence, an effect that can be attributed to the time-reversal symmetry between light absorption and emission. Therefore, we forecast a bright future for optoelectronic devices augmented by resonant photonics.