A theoretical and experimental study on broadband asymmetric light interfaces to reduce top surface losses in luminescent solar concentrators

Abstract

A shift from traditional energy sources to renewable energy sources, such as photovoltaics (PV), is occurring across the planet. PV installations and usage has increased rapidly in recent years and there is becoming even more of a focus on building-integrated photovoltaics (BIPV). A plethora of research can now be found on new ways to incorporate renewable systems into the fabric of buildings themselves. Luminescent solar concentrators (LSCs) are concentrating devices that focus sunlight through total internal reflection that were first conceptualized in the 70’s. LSCs have not been able to find widespread commercial success due to their low efficiency, mostly attributed to losses via the escape cone of the top surface. A number of possible solutions to this problem have been proposed and researched, however there is not currently a broadband effective solution. Asymmetric light transmission (ALT) is a phenomenon first observed over a century ago, but is only now becoming more popular in the realm of optical physics. Most research for ALT revolves around creating optical diodes for future computers, however this concept can be applied to LSCs to enable a top surface that allows more light in than it allows to exit and therefore create a broadband solution to the escape cone losses. This thesis proposes a pyramidal nanostructure that exhibits ALT effects as a modification to the top surface of LSCs. This nanostructure was optimized and numerically tested in COMSOL Multiphysics for a wavelength range ideal for silicon solar cells. Over a wavelength range between 400nm-1200nm and an incident angle between 0-80°, this ALT surface was found to have a directionally averaged spectral transmissivity difference of 50% comparing the transmissivity of the forward and backward directions. To illustrate the potential of the COMSOL results to enhance the optical efficiency of LSCs, the data was integrated into a ray-tracing Monte Carlo code that predicts LSC performance. The method to integrate COMSOL data with the Monte Carlo code was developed, rigorously tested using COMSOL data for a plain LSC interface, and validated against other numerical solutions before testing the efficacy of a nanostructured interface. This code predicts an LSC with an ALT top surface will have an optical efficiency 82% higher than a plain interface LSC. Beyond these various numerical solutions, a procedure was developed for a simple, cost-effective fabrication process of ALT interfaces. ALT interface samples were developed in a nanofabrication facility and experimentally tested. Spectrometer testing for nanostructures with periodicities of 800nm and 900nm showed a “reduced” transmissivity difference of approximately 21% and 10%, respectively. The transmissivity value is “reduced” compared to the total transmissivity value because the spectrometer can only capture some of the scattered diffraction orders caused by the nanostructure. These trends were compared and validated through COMSOL simulations. The results contained in this thesis provide a proof-of-concept solution to reduce LSC top surface losses. Numerical results show that this solution is broadband and effective over all realistic incident angles. Experimental results compare well to numerical expectations and showcase the ease in manufacturing of a nanopattern the is flexible with geometric variation and surface defects.