| Summary: | <p>The introductory chapter commences with a brief perspective, continues with a description of the compound III-V semiconductors and concludes with a survey of solid-state microwave devices including current commercial and state-of-the-art results.</p> <p>The properties of indium phosphide are outlined in Chapter two commencing with the preparation and physical properties, continuing with a description of the band-structure and related electrical properties and including a survey of scattering processes. Inter-valley transfer is discussed leading to the three-level transfer system and the concept of negative-differential mobility. The consequential operation of devices is considered and the chapter concludes with an outline of the Boltzmann transport equation and the solutions thereof by various means and a summary of non-linear effects in the carrier transport theory.</p> <p>Chapter three describes measurement methods of the velocity-field characteristic, and in particular the working equations and analysis of the experiment to measure microwave-conductivity. Perturbing mechanisms of doping inhomogeneity and energy relaxation are considered and other work on these reviewed. Numerical approximation techniques in the experimental analysis are compared and developed, leading to a complete analysis of experimental data and criteria to handle space-charge suppression. Calculations of the effects of transient heating of samples are produced, together with the effects of temperature induced carrier-concentration variation which in the experiments was sciall. The chapter concludes with a description of the experimental apparatus, its calibration and predicts a microwave measurement accuracy of 5%.</p> <p>The results of a computer simulation of the experiment and analysis of theoretical data and derived results are presented in Chapter four, where the various computer programmes are described. It is demonstrated that the velocity-field characteristic may be accurately recovered from data derived from it via the experimental method, assumptions therein being justified numerically. This Chapter concludes with a presentation of the interaction by microwave signals with inductively mounted semiconducting obstacles in rectangular waveguide. Some numerical calculations are described and presented.</p> <p>Chapter five describes the calibration of the experimental system using gallium arsenide samples. The sample preparation, mounting and biassing methods are described. Results from these experiments yielded salient values of threshold field, E<sub>T</sub> = 4.9 kV cm<sup>-1</sup> peak velocity, Vp = 2.1 10<sup>7</sup> cms<sup>-1</sup>, peak-to-valley ratio, r = 2.3. These data are in reasonable agreement with results published elsewhere.</p> <p>The sixth chapter describes the experimental detail relating to the indium phosphide samples, discusses measurement of sample uniformity and work at liquid nitrogen temperatures. The experimental results are presented and variations with epilayer thickness, carrier density, conductivity and temperature are described, this together with possible space-charge growth reduction in thinner layers produces figures 50% above the average in the velocity field characteristic derived from 5μ layers, rather than 10-20μ layers. Carrier density increase from 7.2×10<sup>14</sup> cm<sup>-3</sup> to 2.4×10<sup>15</sup> cm<sup>-3</sup> led to threshold field .decrease from 10.7 kV cm<sup>-1</sup> to 9.5 kV cm<sup>-1</sup> and peak velocity decrease from 2.6×10<sup>7</sup> cms<sup>-1</sup> to 2.3×10<sup>7</sup> cms<sup>-1</sup>, the negative differential mobility being unaffected within the experimental standard deviation. The results at 100°K indicated a lower threshold field of 9.1 kV cm<sup>-1</sup> and a higher peak velocity of 4.4×10<sup>7</sup> cms<sup>-1</sup>.</p> <p>Assumptions are then reviewed and comparison with other published data made. Space-charge effects and relaxation effects are suggested to produce the largest errors, also the results depend strongly on material quality. No correction for relaxation effects is made since it would be comparable to the overall experimental error, which was of the order of ± 15%. Comparison with other published data is good, especially bearing in mind the experimental differences, particularly in the material used.</p> <p>The peak-to-valley ratio was calculated to be in the range 2.15 ≤ r ≤ 2 at 300°K. These results are used to make a calculation of the efficiency of a notional LSA device. A sinusoidal mode efficiency of 7.0% is predicted for operation at 300°K, squarewave efficiencies being some four times this figure. No estimate of frequency dependence is made.</p> <p>Chapter seven forms the work's conclusion presenting the results formally, briefly discussing perturbing effects and errors and considering the device efficiency calculations. The chapter concludes with comments on further investigation into the semiconductor and devices made using it. Work on material, fundamental experiments and device operation and characterisation being associated to bring better understanding of the features of the material and its application.</p>
|
|---|