Final Report for the GCEP Project: Ultra High Efficiency Thermo-Photovoltaic Solar Cells Using Metallic Photonic Crystals As Intermediate Absorber and Emitter

We have achieved significant accomplishments on the developments of nanophotonic absorber and emitter structures for solar thermal photovoltaic applications. The highlight include: (1) The first computational design of a realistic nanophotonic absorber and emitter structures that enable system efficiency beyond the ShockleyQueisser limit. We have provided the first computational design of a Tungsten photonic structures that enable all-angle, near-unity absorption for the entire solar spectrum. We have also designed the first Tungsten emitter structure that possesses sufficient spectral selectivity in order to achieve system efficiency beyond the Shockley-Queisser limit. (2) Experimental demonstrations of photonic crystal thermal emitters that operate at record high temperatures. To achieve high efficiency solar TPV system requires the absorber and emitter to operate at a high temperature (~2000K). This places severe constraint on the materials and structures that can be used in the intermediate. For example, while Tungsten nanostructures have been very widely used to tailor thermal emission, there has been substantial concern regarding the stability of such structures at high temperatures. We have discovered several template-based synthetic methodologies for forming materials with high thermal stabilities into threedimensional photonic crystals. Using these methodologies, we have demonstrated photonic crystal thermal emitters that can withstand a recordhigh temperature of 1800K without structural degradation. This is an important milestone towards overcoming one of the fundamental challenges that have prevented the success of solar TPV in the past. (3) The establishment of an emissometer at Stanford that enables accurate and absolute characterization of emissivity at elevated temperatures. This system provides absolute emissivity measurements at elevated temperature with high accuracy. We have established excellent consistency satisfying Kirkhoff’s law between the emissivity measurement obtained on this system, and direct measurement of the absorptivity. Obtaining such a high quality data at elevated temperature is important as it directly allows us to assess the feasibility of nanophotonic structures at the challenging conditions of solar thermophotovoltaic systems. (4) Theoretical proposal and experimental demonstration of the thermal extraction scheme. We show that the thermal emission of a finite-size blackbody emitter can be enhanced in a thermal extraction scheme, where one places the emitter in thermal contact with an extraction device consists of a transparent object. Our result indicates that one can get far more thermal emission from an emitter of a given size than previously anticipated, while staying within the constraints of the second law of thermodynamics. Our work points to the design of a new generation of thermal emitter with power output significantly beyond what has been previously thought to be possible. Introduction When a single-junction solar cell is illuminated by sunlight, its efficiency is subject to the Shockley-Queisser limit, which sets a fundamental upper bound on the efficiency. This limit arises from several intrinsic loss mechanisms: Solar photons below the band gap do not contribute to electrical current. For each photon above the band gap, the difference between the photon energy, and the output energy at approximately the band gap energy minus approximately 0.5eV (biased at the maximum power point), is dissipated as heat. As a result, the theoretical maximum efficiency of an ideal single junction cell maintained at a room temperature of 300K cannot exceed 41% under solar illumination and maximum concentration. In the absence of concentration, this limit is 31%. Figure 1. The concept of solar thermophotovoltaics. In thermo-photovoltaic (TPV) solar energy conversion, the solar cell is not directly exposed to solar radiation. Instead, an intermediate element is heated by absorbing solar radiation (Figure 1). The emitted thermal radiation from the intermediary, whose spectrum can be very different from that of sunlight, is then converted into electrical energy by a solar cell. The TPV solar cell concept can be highly efficient, because the intermediate absorber/emitter tailors the spectrum of the light incident on the solar cell. The ideal intermediary should provide broadband absorption of sunlight, as well as a narrow band emission with a wavelength tuned to the band gap of the solar cell. When exposed to such a narrow band emission, the performance of a single junction solar cell approaches the thermodynamic limit of the Carnot efficiency. Taking into account the radiative thermal transfer between the sun and the intermediate absorber, the overall maximum efficiency of a TPV cell isη = 1−Ts 4 /Ti 4 ( )× 1−Ti /Tc ( ) , where Ts and Tc are the temperatures of the sun and the solar cell, at 6000K and 300K, respectively. The maximum efficiency of the TPV cell is 85.4% when the temperature of the intermediary is maintained atTi = 2544K . This efficiency is very close to the thermodynamic limit (86.8%) in any reciprocal system. With a single-junction solar cell, a solar TPV system therefore has a theoretical efficiency limit that is more than double that of the case where the same single-junction cell is directly exposed to sunlight. In spite of its conceptual appeal, there has not been any experimental demonstration of a solar TPV system with efficiencies beyond standard solar cells. The objective of our project is to systematically remove some of these basic roadblocks that prevents the demonstration of high-efficiency solar TPV systems. In particular: (1) Prior to our work, there has never been an actual design of a nanophotonic structure that enables system efficiency beyond the Shockley-Queisser limit. Instead, the experimental demonstration has focused on flat tungsten surface, which we know from theoretical analysis can never enable superior performance. We have computationally designed realistic absorbers and emitters that can improve solar TPV system efficiency beyond Shockley-Queisser limit, and moreover developed a novel thermal extraction scheme that can significantly improve the output power of thermal emitters. (2) While nanostructured narrow-band thermal emitters have been widely studied in the past decade, the material systems of these emitters are such that the nanostructure does not survive at the kind of temperature (>2000K) required in a solar TPV system. Also, these nanostructure emitters have not been optimized for solar TPV system. In fact, our own system analysis indicates that the existing emitters are inadequate to enable high efficiency performance. A significant accomplishment of our project is the development of photonic crystal emitter structures that are stable at high temperature relevant for solar TPV applications. The accomplishments of our project will be discussed in more details in the Results Section. Background The control of thermal radiation through nanophotonic structures represents a very active direction within the general area of nanophotonics. Within the period while our project is on-going, significant progresses have been made on the design of new nanophotonic structures to tailor thermal emission. Control of the spectral properties of thermal emission has been demonstrated using arrays of metallic thermal antennas. Spatial shaping of thermal emission has also been demonstrated in a metallic surface with concentric metallic grating. There has also been significant works exploring single thermal emitters, where absorption cross-section beyond its geometric cross-section has been demonstrated. Since the Kirkhoff law relates the absorption and the thermal emission properties of any object, such a demonstration then indicates the possibilities of these structures for enhancing total thermal emission power. Finally, a group at MIT has demonstrated the integration of miniaturized thermophotovoltaic systems integrating TPV emitters with solar cells. However, very few of these works have directly tackled the significant challenges that are inherent in the solar thermophotovoltaic systems. Most high temperature TPV studies have focused on tungsten-based structures. Tungsten is both refractory (melting point of 3422 °C) and it exhibits an intrinsically preferential thermal emission for TPV applications (higher emissivity at optical frequencies, and lower emissivity in the IR). While other materials have been considered, the vast majority of the reports focus on tungsten due to these advantageous properties. High quality 2D structured thermal emitters, formed via conventional lithography tools, have been of interest for TPV applications for some time. 3D structured emitters offer the potential to manipulate emission to a level not possible via a 2D structured surface. However, patterning tungsten at the small length scales into thermally stable structures, which are required to effectively manipulate emission at wavelengths matching high efficiency PV cells, remains challenging. Previous reports of 3D tungsten structures include those formed by chemical vapor deposition, sol gel processing, and electrodeposition. While these methods were able to form tungstenbased structures with the appropriate characteristic length scales for TPV emitters, their thermal stability was limited, at least in part by the quality and density of the infilled material. Through our work, many of these limitations have now been overcome.