G. Smestad, “Luminescence as a predictor of quantum solar energy conversion”, Thesis No. 1263 (1994), The Swiss Federal Institute of Technology (EPFL), Lausanne, Switzerland, thesis advisor: Prof. Dr. Michael Graetzel.
This thesis presents experimental data which supports the idea that room temperature photoluminescence measurements can be used to estimate the maximum thermodynamically allowed chemical potential and solar conversion efficiency for quantum solar converters. Experimental photoluminescence spectra obtained for silicon, chlorophyll based photosynthesis and for Ruthenium dye sensitized TiO2 photoelectrochemical solar cells are compared to predictions made using optical absorption data. The absorption data includes measurements of transmission and reflection, as well as those using the action (or induced photo – product) spectrum. Predictions are made using the measured quantum absorptivity, the Generalized Planck equation and Detailed Balance techniques that relate the chemical potential of the excited state to the photoluminescent emission and photocurrent available for performing work. Although the absorptivity obtained from the optical measurements can be used to predict much of the photoluminescence curve, the predicted values may deviate from measurements at long wavelengths. The photoluminescence spectrum for silicon, the Ruthenium dye cell, and for photosynthesis may also be predicted from action spectra. Both techniques yield a reasonable fit to the experimental spectral distribution and the predicted chemical potentials and voltages are consistent with electrical measurements. Experimental data is also presented on porous silicon solar cells, vacuum photodiodes and silver oxide (and chloride) based photochemistry in relation to the use of photoluminescence signal strength as an indicator of voltage or free energy generation. The novelty of this thesis is the description of all these varied types of quantum photoconverters within the same text and photoluminescence framework.
In chapter one, an introduction is made to the various types of solar conversion systems and their relationship and relevance to luminescence studies. Energy band diagrams are presented for each system and a discussion is made of the recombination pathways that produce work, and the pathways which dissipate energy. At the end of this chapter, a general quantum solar converter is presented and related to thermodynamic principles such as the generation of chemical potential and to the creation of an excited state concentration gradient. This general quantum converter consists of a light activated “pump” which maintains the concentration gradients and chemical potentials.
In chapter two, a more detailed theoretical framework is presented which introduces the previously derived Generalized Planck equation and its relationship to absorptivity, emissivity and the generation of a chemical potential. This equation is connected to the photoluminescence efficiency of the light absorber and to the current -voltage characteristics of a quantum converter. This connection is possible using assumptions regarding the non-radiative dissipation (or recombination) pathways relative to the generation of luminescence. These assumptions are examined using conventional p-n junction solid state theory and the transport equation, and found to be similar to the assumptions used in other device analysis techniques. The relationship between the photoluminescence and the kinetics of the conversion process is then outlined. Specifically, the radiative lifetime and injection rate constants are described for a general photochemical converter and applied to the dye sensitized photoelectrochemical cell. Lastly, a brief theory of phototube operation is presented which allows for the understanding of synergistic effects between heat and light in a vacuum photodiode photoconverter. The thermal and photochemical reactions for the silver based materials used in the phototube photodiode are also presented.
In sections three and four, experimental methods and results are presented for the systems described in the previous chapters. The apparatus is described for the measurement of transmission, absorptivity, photoluminescence, photo-product spectra and current-voltage characteristics. The photoluminescence of silicon, Ruthenium dye cells and photosynthesis is found to be consistent with predictions based on the optical absorptivity and induced photocurrent, or photo-product spectra.
A novel concept introduced by this thesis is the use of the photo-product spectra to estimate the emissivity and absorptivity used in the Generalized Planck equation. Solar conversion efficiency predictions based on this equation are applied to these three conversion systems and are found to be consistent with the experimental performance. Specifically, the maximum allowed chemical potentials for the silicon solar cell are found to be 0.6-0.65 eV, while those for photosynthesis and the dye sensitized cell are found to be near 1.3 eV. The porous silicon solar cell and phototube photodiode results highlight the finding that the chemical potentials predicted from the analysis of photoluminescence are generally much higher than for the actual device. This is due to the converter configuration and kinetic properties such as resistivity. In the case of the vacuum photodiode, the configuration of the converter allows for heat and light to simultaneously be converted to work. Although this has been predicted using photoluminescence based models, this thesis presents the first clear demonstration of a synergistic effect in a “hot” electron device. In hopes of utilizing this effect in a photochemical system, experiments were conducted using the same metal oxide (Ag2O) as was used as the phototube photodiode absorber. It is found, however, that although the oxide powder reacts at low temperatures, the action of light does not conclusively change the reaction thermodynamics or kinetics.
In chapter five, a discussion of the results in the previous chapters is presented, as are ramifications for the application of the theoretical formalism to practical quantum solar converters. Separate conclusions at the end of the thesis summarize the knowledge gained by examining all the various types of converters together. A table of symbols is given before the curriculum vitae to facilitate the description of the concepts presented and discussed.
In sections three and four, experimental methods and results are presented for the systems described in the previous chapters. The apparatus is described for the measurement of transmission, absorptivity, photoluminescence, photo-product spectra and current-voltage characteristics. The photoluminescence of silicon, Ruthenium dye cells and photosynthesis is found to be consistent with predictions based on the optical absorptivity and induced photocurrent, or photo-product spectra. A novel concept introduced by this thesis is the use of the photo-product spectra to estimate the emissivity and absorptivity used in the Generalized Planck equation. Solar conversion efficiency predictions based on this equation are applied to these three conversion systems and are found to be consistent with the experimental performance. Specifically, the maximum allowed chemical potentials for the silicon solar cell are found to be 0.6-0.65 eV, while those for photosynthesis and the dye sensitized cell are found to be near 1.3 eV. The porous silicon solar cell and phototube photodiode results highlight the finding that the chemical potentials predicted from the analysis of photoluminescence are generally much higher than for the actual device. This is due to the converter configuration and kinetic properties such as resistivity. In the case of the vacuum photodiode, the configuration of the converter allows for heat and light to simultaneously be converted to work. Although this has been predicted using photoluminescence based models, this thesis presents the first clear demonstration of a synergistic effect in a “hot” electron device. In hopes of utilizing this effect in a photochemical system, experiments were conducted using the same metal oxide (Ag2O) as was used as the phototube photodiode absorber. It is found, however, that although the oxide powder reacts at low temperatures, the action of light does not conclusively change the reaction thermodynamics or kinetics.
In chapter five, a discussion of the results in the previous chapters is presented, as are ramifications for the application of the theoretical formalism to practical quantum solar converters. Separate conclusions at the end of the thesis summarize the knowledge gained by examining all the various types of converters together. A table of symbols is given before the curriculum vitae to facilitate the description of the concepts presented and discussed.
For a table of contents (table des matières) and additional information about the thesis, see the EPFL website.