Nanophotonic designs for luminescent solar concentrators
In recent years, the concept of light management for photovoltaics has emerged as a viable option to enhance the performance of solar cells. Although photovoltaic development has traditionally focused on producing high-quality, inexpensive materials with excellent electronic properties, manipulating the optical properties of materials is a new approach to solar cell design. Integrating photonic concepts, or structures with wavelength- or sub-wavelength-scale structuring that influence light propagation, can enhance the performance of photovoltaics through improvements in light absorption, carrier collection, anti-reflection coatings, and up/down converters or downshifters.1, 2 The challenge for photonic designs for photovoltaics is to operate effectively over a large spectral bandwidth and at all angles of incidence.
Some photonic designs enable radically new types of solar cells by rethinking the entire architecture, whereas others seek to integrate new physical concepts into existing device architectures to minimize production barriers. Luminescent solar concentrators (LSCs), which harness luminescent materials to downshift the incident solar spectrum and concentrate it onto a high efficiency solar cell, are of the latter type and have been explored for more than 30 years.3, 4 A typical LSC consists of a plastic plate with embedded lumophores (see Figure 1). The lumophores absorb sunlight and re-emit it at longer wavelengths, a process known as the Stokes shift. The re-emitted light is directed onto an adjacent solar cell through total internal reflection and, as the plate is many times larger than the solar cell, this serves to concentrate solar flux. The LSC operates under both direct and diffuse incidence, eliminating the need for tracking devices and lenses. In principle, LSCs can achieve very high efficiencies, but performances have been limited to date.5 Inside the LSC, the emitted photons reflect off the interfaces multiple times before reaching the solar cell, and are subject to escape cone losses as well as re-absorption by other embedded lumophores. If the overlap between the absorption and emission of the lumophore is too high and the quantum yield too low, these re-absorption events will substantially reduce the conversion efficiency.
Photonic designs could help address these limitations. Concepts for photonic LSCs include the use of aligned lumophores, wavelength-selective mirrors, and photonic crystals, among others.6–10 Typical LSCs use emitting dye molecules or inorganic nanocrystals, which have the disadvantage of overlapping absorption and emission bands. In recent years, a new class of emitting material based on semiconductor nanoparticles with large Stokes shifts has been proposed for use with LSCs.11–14 One option is core-shell nanoparticles, where the Stokes shift can be tailored through control of the core size and shell thickness. From a nanophotonics perspective, there is another benefit to using these new materials: the narrower emission spectrum of these core-shell nanoparticles compared with dye molecules relaxes the constraint on integrating photonic designs. Instead of requiring operation over the broader spectrum of the dye, photonic designs can now be optimized over a narrower spectral region, whereby periodic nanostructures and layers can result in stronger light/matter interactions.
Colleagues and I recently demonstrated this concept by constructing LSCs using dot-in-rod heterostructures comprising cadmium selenide (CdSe) cores and cadmium sulfide (CdS) shells.11 The CdSe seed diameter was held constant at 2.5nm. We tuned the length of the CdS nanorod, producing a red-shift in the emission spectrum maximum from 570nm to 610nm for the longest nanorods. These nanorods were incorporated into a polymer sheet containing an embedded silicon solar cell. We found that the nanorods with the largest Stokes shift produced LSCs with lower re-absorption and, hence, the longest propagation lengths of luminesced photons.
We also used a ray-tracing Monte-Carlo method to accurately reproduce the measured concentration factors using known physical parameters such as the absorption and emission spectra, nanoparticle loading, and edge reflectivity. We then used this numerical model to explore the effect of integrating a simple photonic structure consisting of a 1D Bragg reflector on the top of the LSC (see Figure 1). This mirror is designed to reflect light over the emission band but not the absorbed wavelengths, thereby trapping the re-emitted light within the LSC where it eventually reaches the solar cell. For effective operation, the reflection band of the photonic mirror from normal incidence up to the total internal reflection angle must include the full emission spectrum of the lumophore (see Figure 2). An emission spectrum from a reference dye molecule is included in Figure 2 for comparison, showing that the narrower inorganic nanorod emission spectrum enables reflection by the mirror across the entire emission band. This light-trapping approach is found to be most useful for the nanorods with large Stokes shifts, which reduce reabsorption losses. With further optimization of the design, we predict that photonic structures could enable concentration factors exceeding 100.
In summary, we have shown that photonic designs can be integrated with tunable emitting nanocrystals to produce highly efficient LSCs. In future work, we shall focus on improving the match between the lumophore and solar cell, improving the quantum yield of the materials with the highest Stokes shifts, and exploring more complex photonic designs.
The author is grateful to Noah Bronstein and A. Paul Alivisatos.
University of Minnesota – Twin Cities
Vivian E. Ferry is an assistant professor. Her research interests include fundamentals of plasmonics, nanophotonics, and metamaterials, and their applications to photovoltaics, optoelectronic devices, and sensing.