Optical concentration can lower the cost of solar energy conversion by reducing photovoltaic cell area and increasing photovoltaic efficiency. Luminescent solar concentrators offer an attractive approach to combined spectral and spatial concentration of both specular and diffuse light without tracking, but they have been plagued by luminophore self-absorption losses when employed on practical size scales. Here, we introduce doped semiconductor nanocrystals as a new class of phosphors for use in luminescent solar concentrators. In proof-of-concept experiments, visibly transparent, ultraviolet-selective luminescent solar concentrators have been prepared using colloidal Mn(2+)-doped ZnSe nanocrystals that show no luminescence reabsorption. Optical quantum efficiencies of 37% are measured, yielding a maximum projected energy concentration of ∼6× and flux gain for a-Si photovoltaics of 15.6 in the large-area limit, for the first time bounded not by luminophore self-absorption but by the transparency of the waveguide itself. Future directions in the use of colloidal doped nanocrystals as robust, processable spectrum-shifting phosphors for luminescent solar concentration on the large scales required for practical application of this technology are discussed.

A diffusion-based synthesis of doped colloidal semiconductor nanocrystals is demonstrated. This approach involves thermodynamically controlled addition of both impurity cations and host anions to preformed seed nanocrystals under equilibrium conditions, rather than kinetically controlled doping during growth. This chemistry allows thermodynamic crystal compositions to be prepared without sacrificing other kinetically trapped properties such as shape, size, or crystallographic phase. This doping chemistry thus shares some similarities with cation-exchange reactions, but proceeds without the loss of host cations and excels at the introduction of relatively unreactive impurity ions that have not been previously accessible using cation exchange. Specifically, we demonstrate the preparation of Cd(1-x)Mn(x)Se (0 ≤ x ≤ ∼0.2) nanocrystals with narrow size distribution, unprecedentedly high Mn(2+) content, and very large magneto-optical effects by diffusion of Mn(2+) into seed CdSe nanocrystals grown by hot injection. Controlling the solution and lattice chemical potentials of Cd(2+) and Mn(2+) allows Mn(2+) diffusion into the internal volumes of the CdSe nanocrystals with negligible Ostwald ripening, while retaining the crystallographic phase (wurtzite or zinc blende), shape anisotropy, and ensemble size uniformity of the seed nanocrystals. Experimental results for diffusion doping of other nanocrystals with other cations are also presented that indicate this method may be generalized, providing access to a variety of new doped semiconductor nanostructures not previously attainable by kinetic routes or cation exchange.

Colloidal halide perovskite nanocrystals of CsPbCl3 doped with Yb3+ have demonstrated remarkably high sensitized photoluminescence quantum yields (PLQYs), approaching 200%, attributed to a picosecond quantum-cutting process in which one photon absorbed by the nanocrystal generates two photons emitted by the Yb3+ dopants. This quantum-cutting process is thought to involve a charge-neutral defect cluster within the nanocrystal’s internal volume. We demonstrate that Yb3+-doped CsPbCl3 nanocrystals can be converted postsynthetically to Yb3+-doped CsPb­(Cl1–x Br x )3 nanocrystals without compromising the desired high PLQYs. Nanocrystal energy gaps can be tuned continuously from E g ≈ 3.06 eV (405 nm) in CsPbCl3 down to E g ≈ 2.53 eV (∼490 nm) in CsPb­(Cl0.25Br0.75)3 while retaining a constant PLQY above 100%. Reducing E g further causes a rapid drop in PLQY, interpreted as reflecting an energy threshold for quantum cutting at approximately twice the energy of the Yb3+ 2F7/2 → 2F5/2 absorption threshold. These data demonstrate that very high quantum-cutting energy efficiencies can be achieved in Yb3+-doped CsPb­(Cl1–x Br x )3 nanocrystals, offering the possibility to circumvent thermalization losses in conventional solar technologies. The presence of water during anion exchange is found to have a deleterious effect on the Yb3+ PLQYs but does not affect the nanocrystal shapes or morphologies, or even reduce the excitonic PLQYs of analogous undoped CsPb­(Cl1–x Br x )3 nanocrystals. These results provide valuable information relevant to the development and application of these unique materials for spectral-shifting solar energy conversion technologies.

Luminescent solar concentrators (LSCs) use down-converting luminophores embedded in a waveguide to absorb sunlight and deliver high irradiance, narrowband output light for driving photovoltaic and other solar energy conversion devices. Achieving a technologically useful level of optical gain requires bright, broadly absorbing, large-Stokes-shift luminophores incorporated into low-loss waveguides, a combination that has long posed a challenge to the development of practical LSCs. The recent introduction of giant effective Stokes shift semiconductor nanocrystal (NC) phosphors for LSC applications has led to significant performance improvements by increasing solar absorption while reducing escape cone and nonradiative losses compounded by reabsorption, placing increased emphasis on the importance of minimizing parasitic waveguide losses caused by scattering from NC aggregates and optical imperfections. Here, we report a detailed analysis of optical losses in polymer–NC composite waveguide LSCs based on CuInS2/CdS NC phosphors, which have been shown to provide best-in-class performance in large-area, semitransparent concentrators. A comprehensive analytical optical model is introduced enabling quantification of parasitic waveguide, scattering, escape cone, and nonradiative relaxation losses on the basis of distance-dependent edge-emission measurements. By examining the effect of NC loading, we show that NC clustering in polymer composite waveguides leads to light scattering losses that ultimately limit efficiency at large geometric gain. By optimizing NC concentration, optical power efficiencies up to 5.7% under AM1.5 illumination are demonstrated for devices having a geometric gain G = 6.7×, with limiting achievable efficiencies predicted to exceed 10%.

Yb 3+ -doped halide perovskites have recently emerged as extraordinarily promising materials for solar spectral downconversion applications because of their extremely high photoluminescence quantum yields of nearly 200%, attributable to a highly efficient picosecond quantum-cutting process. One of the major roadblocks to widespread application of these materials is their photoluminescence saturation under modest photoexcitation fluences. In this study, we examine the excitation-fluence dependence of Yb 3+ -doped CsPb(Cl 1−x Br x ) 3 nanocrystal photoluminescence to develop a quantitative understanding of this saturation. Facile saturation is observed across a multitude of halide and Yb 3+ compositions, with specific trends that provide insight into the microscopic mechanism behind this saturation. We show that the data can be simulated well by a kinetic model that introduces a specific new Auger-type cross-relaxation process involving nonradiative energy transfer from photoexcited nanocrystals to Yb 3+ ions that are already in their luminescent 2 F 5/2 excited state from a previous photoexcitation event. This cross relaxation occurs with a subnanosecond rate constant, allowing it to compete with picosecond quantum cutting when excited-state Yb 3+ is accumulated. These results point to specific strategies for ameliorating photoluminescence saturation in this class of materials, one of which is demonstrated experimentally. The proposed strategies provide guidance for future materials development and application efforts involving these materials.