Degree Name

Doctor of Philosophy (PhD)


School of Engineering Physics - Faculty of Engineering


Thermionic refrigeration using semiconductor heterostructures is examined theoretically and experimentally. A theory of single-barrier devices is first developed where two classes of single-barrier devices are defined and compared. So-called class 1 devices are found to always perform better. A theory of multiple-barrier devices based on class 1 barriers is then developed using a numerical solution. Experimentally, three generations of 10-barrier devices based on A1(subscript x)Ga(subscript 1-x)As-GaAs heterostructures were made and electrically characterised. This material is by no means ideal (as will be discussed) but was used to availability and because, at the commencement of this work, had never been used for this purpose before. Thermal measurements were made to determine if any cooling occurred at room temperature. No cooling was observed but the electrical characteristics allowed for examination of the models developed. It was found that the earlier models used did not accurately model the I-V characteristics of the devices. This was attributed to the fact that the initial models did not take space-charge into account. A more robust numerical model is developed in which the I-V characteristics of devices are predicted much more accurately. This model is then used to design new generations of devices. The work concludes by recommending a next generation design in which substantially more cooling is expected compared to the samples examined here. The probability of cooling being observed in the future is therefore increased. The types of devices described here will always be hindered because of heat conduction. Other methods incorporating thermionic emission, such as an opto-thermionic system in which removed heat is given off as light, may ultimately prove to be the best solution. This aside, it is hoped that the work presented here will contribute to the understanding of the field.