Takumi Yamada

Cooling materials with light has been a frontier challenge in physics. While Doppler cooling is the most well-known optical cooling method, it is not applicable to solids. Instead, Anti-Stokes (AS) cooling, which utilizes Anti-Stokes photoluminescence (ASPL), has been proposed. ASPL refers to photoluminescence (PL) in which the emission energy exceeds the excitation energy. If the external PL quantum efficiency reaches 100%, the material loses net energy through optical absorption followed by ASPL, resulting in optical cooling. This mechanism fundamentally differs from conventional solid-state cooling based on thermal conduction, as it does not require physical contact to reduce temperature. To date, AS cooling studies have primarily focused on crystals doped with rare-earth ions, which have demonstrated successful AS cooling down to 90 K [1]. However, these systems suffer from weak photoabsorption and intrinsic low-temperature limits. Semiconductors, which exhibit strong optical absorption, are promising candidates for achieving lower temperatures and higher cooling efficiencies. However, AS cooling requires near-unity PL quantum efficiency, which presents a significant challenge for semiconductor systems. In this context, semiconductor quantum dots (QDs) are particularly promising. For optical cooling, we propose a novel QD structure, “dot-in-crystal” perovskites, in which halide perovskite CsPbBr ₃ QDs are embedded in a transparent Cs ₄ PbBr ₆ host crystal [2,3]. This structure, hereafter referred to as CsPbBr ₃ /Cs ₄ PbBr ₆ , offers both excellent optical properties and improved photostability. In addition, halide perovskites possess strong electron–phonon coupling [4,5], which enables phonon-assisted photoabsorption and makes them suitable for efficient ASPL. In fact, strong ASPL has been observed in halide perovskites [2,3,5]. We first investigated the ASPL characteristics of CsPbBr ₃ /Cs ₄ PbBr ₆ . The PL spectrum remains unchanged regardless of the excitation photon energy. Under below-bandgap excitation, phonon-assisted photoabsorption occurs, enabling ASPL. From detailed analysis of the PL excitation spectra, we demonstrated that optical cooling is possible if the PL quantum efficiency exceeds 97% [3]. Since near-unity PL quantum efficiency has been reported for CsPbBr ₃ /Cs ₄ PbBr ₆ , this strongly supports the feasibility of optical cooling in this material. Although achieving the required PL efficiency is feasible, maximizing the cooling power faces another challenge. While stronger photoexcitation might, intuitively, seem effective, high-density excitation in semiconductors gives rise to nonradiative Auger recombination, which reduces PL efficiency. As a result, the cooling gain cannot increase indefinitely and therefore has an upper limit. Moreover, due to strong quantum confinement, Auger recombination is more prominent in QDs. We investigated this effect in CsPbBr ₃ /Cs ₄ PbBr ₆ using time-resolved PL spectroscopy [3]. From the excitation-intensity-dependent PL dynamics, we extracted the lifetimes of excitons, trions, and biexcitons, which reflect the influence of nonradiative Auger recombination. Based on these results, we estimated the upper limit of the cooling gain to be approximately 2 fW per QD at the optimized excitation intensity. Combining these results with the Stefan–Boltzmann law, we estimated that the lowest achievable temperature of CsPbBr ₃ /Cs ₄ PbBr ₆ under ideal thermal isolation is approximately 10 K below room temperature [3]. Building on the above insights, we carefully performed cooling experiments on CsPbBr ₃ /Cs ₄ PbBr ₆ under optimized excitation conditions [3]. Since non-contact temperature measurement is essential, we employed a newly developed PL thermometry technique, which estimates the temperature from the high-energy tail of the PL spectrum. This method provides reliable temperature estimates with an accuracy of about 1 K. We used CsPbBr ₃ /Cs ₄ PbBr ₆ microparticles placed on a mica substrate inside a vacuum chamber, and selected a highly luminescent particle. We observed a temperature reduction of approximately 9 K under continuous-wave photoexcitation. This provides the first clear evidence of optical cooling of perovskite QDs. Furthermore, by tuning the excitation intensity, we demonstrated a transition from cooling to heating, in good agreement with the model that incorporates Auger recombination. In the presentation, we will discuss the ASPL mechanism in halide perovskites, which arises from phonon-assisted photoabsorption. We will also highlight the current limitations of semiconductor optical cooling and explore possible strategies to overcome these challenges, along with future prospects for this emerging technology. Part of this work was supported by the Asahi Glass Foundation and JST-CREST (Grant No. JPMJCR21B4). References [1] D. Melgaard, et al., Sci. Rep., 6, 20380 (2016). [2] Y. Kajino, et al., Phys. Rev. Mater. 6, L043001 (2022). [3] Y. Yamada, et al., Nano Lett. 24, 11255 (2024). [4] Y. Yamada, et al., Phys. Rev. Lett. 126, 237401 (2021). [5] Y. Yamada and Y. Kanemitsu, NPG Asia Mater. 14, 48 (2022).

The development of hole-collecting materials is indispensable to improving the performance of perovskite solar cells (PSCs). To date, several anchorable molecules have been reported as effective hole-collecting monolayer (HCM) materials for p-i-n PSCs. However, their structures are limited to well-known electron-donating skeletons, such as carbazole, triarylamine, etc. In this work, we developed a series of squaraine derivatives that have a π-conjugated core composed of a squaric acid moiety connected to an indoline moiety. Thanks to the polar carbonyl group of squaric acid, all of the molecules were found to form hydrophilic monolayers after being chemisorbed on transparent conducting oxide surfaces, which is beneficial for the subsequent deposition of the perovskite layer. The effect of the substituents on the squaric acid moiety and the anchoring groups connected to the indoline moiety on the molecular electronic structure as well as the solar cell device's performance was elucidated. The p-i-n PSC devices fabricated by using these squaraine derivatives as hole-collecting monolayer materials exhibited high power conversion efficiencies of up to 22.1%, together with good stability. This work highlights the potential of a simple squaric acid skeleton as the building block for hole-collecting monolayer materials to realize high-efficiency and cost-effective PSCs.

High-performance and cost-effective hole-collecting materials (HCMs) are indispensable for commercially viable perovskite solar cells (PSCs). Here, we report an anchorable HCM composed of a triazatruxene core connected with three alkyl carboxylic acid groups (3CATAT-C3). In contrast to the phosphonic acid-containing tripodal analog (3PATAT-C3), 3CATAT-C3 molecules can form a hydrophilic monolayer on a transparent conducting oxide surface, which is beneficial for subsequent perovskite film deposition in the traditional layer-by-layer fabrication process. More importantly, the larger diffusion coefficient and higher surface energy make 3CATAT-C3 suitable for the simplified, cost-effective one-step co-deposition process in which 3CATAT-C3 was directly added as part of the perovskite precursor solution. 3CATAT-C3 is predominantly located at the perovskite bottom surface after spin-coating the mixed precursor solution, facilitating charge extraction. Devices with 3CATAT-C3 fabricated by this co-deposition method exhibit superior performance with a champion power conversion efficiency of over 23%. The unencapsulated devices showed good operational stability (retaining 90% of the initial output after 100 h), thermal durability (retaining 95% of the initial value after heating at 105 °C under air), and excellent storage stability (showing no drop in performance over 8000 h). Based on the results of time-of-flight secondary-ion mass spectroscopy (ToF-SIMS) and diffusion order nuclear magnetic resonance spectroscopy (DOSY), we elucidated the effect of anchoring groups on the performance of the tripodal HCMs in PSCs as well as the mechanism of the co-deposition fabrication process. Our findings provide valuable insights for the molecular design of multifunctional hole-collecting materials, further advancing the performance of perovskite solar cells.

Although fullerene bisadducts are promising electron-transporting materials for tin halide perovskite solar cells, they are generally synthesized as a mixture of isomeric products that require a complicated separation process. Here, we introduce a phenylene-bridged bis(pyrrolidino)fullerene, Bis-PC, which forms only a single isomer due to geometrical restriction. When used in a tin perovskite solar cell with a PEA_0.15FA_0.85SnI₃ (PEA: phenylethylammonium, FA: formamidinium) light absorption layer, the resulting open-circuit voltage (V_OC) was 0.78 V, a value higher than fullerene monoadducts and comparable to the commonly used indene-C₆₀ bisadduct (ICBA). The performance could be further improved by the composition engineering of perovskite, where the PEA_0.15(FA_0.87MA_0.13)_0.85SnI ₃ based device (MA: methylammonium) reached a photoelectric conversion efficiency of 12.3% with a V_OC of 0.86 V. The device with single-isomer Bis-PC shows superior stability to mixed-isomer ICBA, maintaining the initial performance after 3,000 h storage under an inert atmosphere.

The band gap energy of halide perovskite semiconductors is manipulated by controlling the halide composition, and mixed halide perovskites are receiving much attention as top cell materials for tandem solar cells. To understand dynamic aspects of photoinduced halide segregation in mixed-halide perovskite films, we use a hyperspectral imaging technique. We reveal the space- and time-resolved photoluminescence (PL) spectra of Cs_0.15FA_0.85PbI₂Br perovskite films during prolonged light illumination. Under applied electric fields, we observe photoinduced phase segregation at the excitation laser spot, with a line-shape I-rich region of low PL efficiency appearing near the anode electrode. This I-rich region moves from the anode to the cathode electrodes and stops at the laser excitation spot. We discuss the significant enhancement of halide ion migration under light illumination and the dynamical changes of photoinduced halide segregation.

Lead halide perovskite nanocrystals (NCs) have attracted much attention as materials for light-emitting diodes and quantum light sources. A deep understanding of exciton-phonon couplings is essential for obtaining a narrow emission line, weak phonon-sideband photoluminescence (PL), and a long exciton coherence time, which are especially useful for high-color-purity quantum-light-source applications. Here, we report the PL spectra of single CsPbBr₃ NCs at 5.5 K as a function of the applied electric field. The exciton peak energy shows an asymmetric parabolic shift for positive and negative biases, implying the presence of a spontaneously generated internal electric field in the NCs when no field is applied. Both the internal electric field and exciton-phonon couplings become larger in smaller NCs, and they have a positive correlation with each other. Our findings show that the exciton-phonon couplings can be manipulated with an electric field, which dominates the PL properties of perovskite NCs.

Lead halide perovskites have attracted much attention as a new class of optoelectronic device materials because of their outstanding optical, electronic, and transport properties [1,2]. In addition, halide perovskite nanocrystals (NCs), which can be synthesized by simple chemical methods, show near unity photoluminescence (PL) quantum yields, with keeping the superior optoelectronic properties of their bulk counterparts [3,4]. Because of the small orbital degeneracy of the band edge, PL properties of halide perovskite NCs are governed by the formation and recombination dynamics of excitons, trions, and biexcitons [5,6]. Since individual NCs show narrow PL linewidths at low temperatures, multipeak PL structures appear and their origins are an exciton, a trion, and a biexciton. Single dot spectroscopy is a powerful tool to investigate the exciton–exciton, exciton–electron, and exciton–phonon interactions in perovskite NCs. Here, we prepared three types of perovskite NC samples, CsPbBr₃, CsPbI₃ and FAPbBr₃, and studied their PL spectra of single NCs at room temperature and low temperatures. At room temperature, all NCs show PL blinking, and nonradiative Auger recombination of trions and biexcitons determines the PL dynamics [7–10]. At low temperatures, all NC samples show strong PL, and several PL peaks originating a trion, a biexciton, and longitudinal-optical-phonon side bands of an exciton were clearly observed in the low energy side of the strong exciton peak [11–13]. The PL spectra of all samples were very similar to each other. We clarified the NC size dependence of the trion and biexciton binding energies and the Huang–Rhys (HR) factors, i.e., the strength of the exciton–phonon coupling, in three different perovskites. The positive binding energies of trions and biexcitons show the attractive exciton–exciton Coulomb interactions in the perovskite NCs, and these binding energies increase with decreasing NC size [11,13]. In addition, these size dependences follow a universal scaling curve regardless of composition when the binding energy and the NC size are normalized by the bulk exciton binding energy and the exciton Bohr radius, respectively [13]. The PL linewidths of all inorganic CsPbBr₃ and CsPbI₃ NCs were narrower than those of organic-inorganic hybrid FAPbBr₃ NCs, suggesting the strong exciton–phonon coupling with large organic cation. In FAPbBr₃ NCs, the size dependence of the HR factors is similar to that of the Urbach energy [10,11]. In all samples, the LO-phonon energies are independent of the NC size, while the HR factors increase as the NC size decreases [11,12]. Our findings provide new insights into understanding of the optical responses of halide perovskite NCs and the design of the single photon sources. Part of this work was supported by JST-CREST (JPMJCR21B4) and JSPS KAKENHI (JP19H05465). References [1] T. Yamada et al., J. Phys. D: Appl. Phys. 54, 383001 (2021). [2] Y. Yamada and Y. Kanemitsu, NPG Asia Materials 14, 48 (2022). [3] Y. Yamada et al., J. Am. Chem. Soc. 136, 11610–11613 (2014). [4] Y. Yamada et al., J. Am. Chem. Soc. 137, 10456–10459 (2015). [5] Y. Kanemitsu, J. Chem. Phys. 151, 170902 (2019). [6] G. Yumoto and Y. Kanemitsu, Phys. Chem. Chem. Phys. 24, 22405–22425 (2022). [7] N. Yarita et al., J. Phys. Chem. Lett. 8, 1413–1418 (2017). [8] N. Yarita et al., J. Phys. Chem. Lett. 8, 6041–6047 (2017). [9] N. Yarita et al., Phys. Rev. Materials 2, 116003 (2018). [10] S. Masada et al., Nano Lett. 20, 4022–4028 (2020). [11] K. Cho et al., Nano Lett. 21, 7206–7212 (2021). [12] K. Cho et al., Nano Lett. 22, 7674–7681 (2022). [13] K. Cho et al., ACS Nano 18, 5723–5729 (2024).

Hole‐collecting monolayers have greatly advanced the development of positive‐intrinsic‐negative perovskite solar cells (p‐i‐n PSCs). To date, however, most of the anchoring groups in the reported monolayer materials are designed to bind to the transparent conductive oxide (TCO) surface, resulting in less availability for other functions such as tuning the wettability of the monolayer surface. In this work, we developed two anchorable molecules, 4PATTI‐C3 and 4PATTI‐C4, by employing a saddle‐like indole‐fused cyclooctatetraene as a π‐core with four phosphonic acid anchoring groups linked through propyl or butyl chains. Both molecules form monolayers on TCO substrates. Thanks to the saddle shape of a cyclooctatetraene skeleton, two of the four phosphonic acid anchoring groups were found to point upward, resulting in hydrophilic surfaces. Compared to the devices using 4PATTI‐C4 as the hole‐collecting monolayer, 4PATTI‐C3‐based devices exhibit a faster hole‐collection process, leading to higher power conversion efficiencies of up to 21.7 % and 21.4 % for a mini‐cell (0.1 cm²) and a mini‐module (1.62 cm²), respectively, together with good operational stability. This work represents how structural modification of multipodal molecules could substantially modulate the functions of the hole‐collecting monolayers after being adsorbed onto TCO substrates.

Hole‐collecting monolayers have greatly advanced the development of positive‐intrinsic‐negative perovskite solar cells (p‐i‐n PSCs). To date, however, most of the anchoring groups in the reported monolayer materials are designed to bind to the transparent conductive oxide (TCO) surface, resulting in less availability for other functions such as tuning the wettability of the monolayer surface. In this work, we developed two anchorable molecules, 4PATTI‐C3 and 4PATTI‐C4, by employing a saddle‐like indole‐fused cyclooctatetraene as a π‐core with four phosphonic acid anchoring groups linked through propyl or butyl chains. Both molecules form monolayers on TCO substrates. Thanks to the saddle shape of a cyclooctatetraene skeleton, two of the four phosphonic acid anchoring groups were found to point upward, resulting in hydrophilic surfaces. Compared to the devices using 4PATTI‐C4 as the hole‐collecting monolayer, 4PATTI‐C3‐based devices exhibit a faster hole‐collection process, leading to higher power conversion efficiencies of up to 21.7 % and 21.4 % for a mini‐cell (0.1 cm²) and a mini‐module (1.62 cm²), respectively, together with good operational stability. This work represents how structural modification of multipodal molecules could substantially modulate the functions of the hole‐collecting monolayers after being adsorbed onto TCO substrates.

Highly efficient anti-Stokes (AS) photoluminescence (PL) is observed from halide perovskite quantum dots (QDs) due to their strong electron-phonon interactions. The AS PL is particularly intriguing, as it suggests the potential for semiconductor optical cooling if the external quantum efficiency approaches 100%. However, the PL quantum efficiency in QDs is primarily dominated by multiparticle nonradiative Auger recombination processes under intense photoexcitation, which impose limits on the optical cooling gain. Here, we investigate the Auger recombination of dot-in-crystal perovskites. We quantitatively estimate the maximum optical cooling gain and the corresponding excitation intensity. We further conducted optical cooling experiments and demonstrate a maximum photocooling of approximately 9 K from room temperature. Additionally, we confirmed that increasing the excitation intensity leads to a transition from photocooling to photoheating. These observations are consistent with our time-resolved measurements, offering insights into the potential and limitations of optical cooling in semiconductor QDs.