UC Santa Cruz

Solar energy has long been lauded as an inexhaustible fuel source with more energy reaching the earth's surface in one hour than the global consumption for a year. Although capable of satisfying the world's energy requirements, solar energy remains an expensive technology that has yet to attain grid parity. Another drawback is that existing solar farms require large quantities of land in order to generate power at useful rates. In this work, we look to luminescent solar concentrator systems and quantum dot technology as viable solutions to lowering the cost of solar electricity production with the flexibility to integrate such technologies into buildings to achieve dual land use.

Luminescent solar concentrator (LSC) windows with front-facing photovoltaic (PV) cells were built and their gain and power efficiency were investigated. Conventional LSCs employ a photovoltaic (PV) cell that is placed on the edge of the LSC, facing inward. This work describes a new design with the PV cells on the front-face allowing them to receive both direct solar irradiation and wave-guided photons emitted from a dye embedded in an acrylic sheet, which is optically coupled to the PV cells. Parameters investigated include the thickness of the waveguide, edge treatment of the window, cell width, and cell placement. The data allowed us to make projections that aided in designing windows for maximized overall efficiency. A gain in power of 2.2x over the PV cells alone was obtained with PV cell coverage of 5%, and a power conversion efficiency as high as 6.8% was obtained with a PV cell coverage of 31%. Balancing the trade-offs between gain and efficiency, the design with the lowest cost per watt attained a power efficiency of 3.8% and a gain of 1.6x.

With the viability of the LSC demonstrated, a weighted Monte-Carlo Ray Tracing program was developed to study the transport of photons and loss mechanisms in the LSC to aid in design optimization. The program imports measured absorption/emission spectra of an organic luminescent dye (LR305), the transmission coefficient and refractive index of acrylic as parameters that describe the system. Simulations suggest that for LR305, 8-10cm of luminescent material surrounding the PV cell yields the highest increase in power gain per unit area of LSC added, thereby determining the ideal spacing between PV cells in the panel. The model also predicts that for different PV cell dimensions, there exists an optimum waveguide thickness which efficiently transports photon collected by the waveguide to the PV cell with minimal loss, and maintains an even distribution of photons across the cell surface. For the case of the 12.5 by 1cm rectangular cells used in this work, the calculated waveguide thickness is 3mm. For larger cells, every 1cm increment in PV cell width should be accompanied by a 0.75mm increase in waveguide thickness to preserve peak performance.

In line with the goal of pushing for cost competitive photovoltaics, the last part of this work shifts to the study of colloidal quantum dot solar cells. A combination of low temperature, highly scalable fabrication process and reduced material usage for thin films offers us a means to produce flexible and cheap solar cells. Tagging on to existing work already performed on germanium quantum dot solar cells, additional work was carried out to further characterize the material. The effect of film thickness, nano-particle surface conditions and thermal anneal were investigated. There is evidence to suggest that the quantum dot devices contain high levels of parasitic resistances. Short circuit current densities increase by up to two times with two spin-cast layers compared to four, leading to the conjecture that charge carrier life time is low with high levels of recombination. Annealing to improve carrier mobility produces devices with current densities up to 301µA, a fourfold increase, but output voltages saw a sharp decrease from 0.12V to 0.015V.

In tandem with the work on germanium, experiments on silicon quantum dots were also carried out to investigate their viability for use as photovoltaic devices. The stronger bonds formed by silicon hindered the ligand exchange process. Schottky diodes were made via drop casting and displayed a clear photovoltaic effect albeit with very low current densities. Interestingly, an open circuit voltage was observed even when not under illumination and further investigations are ongoing.