6. Case Studies

THERMOELECTRIC DEVICES are interesting candidates for a series of applications due to their capability of direct conversion of heat into electric energy. For a rigorous study, general device structures have to be considered, which demands the application of a general-purpose device simulator. For the following investigations, the device simulator MINIMOS-NT [268] has been applied. As a part of this work, the transport model has been revised and mechanisms for the visualization of thermoelectrically relevant quantities have been implemented.

In the sequel, the device behavior of two classes of devices is analyzed. First, the general behavior of classical thermoelectric devices is investigated with special attention payed to the influence of geometrical parameters as well as thermal environment and device parameters. The influence of material composition in alloys is illustrated for the example of silicon-germanium devices. Based on mixed-mode simulation [295], non-ideal thermal conditions are considered by an analog electric compact model.

For the simulation of lead telluride based devices, the model database has been extended by an according material class and model parameters. Furthermore, a set of models has been implemented in MINIMOS-NT in order to accurately account for the physical properties, such as the band gap and the doping dependent electric contribution to the thermal conductivity, as described in detail in Chapter 5. On the basis of a lead telluride device, an optimization strategy for stacked devices is discussed.

As a second thermoelectric structure, pn-junctions under a temperature gradient are rigorously investigated. In contrast to classical devices, carrier generation plays an important role within this structure. In order to sketch some possibilities for device optimization, the influence of several device parameters on its performance is investigated. Particularly, graded material alloys are investigated as a possibility to engineer the temperature profile within the device. Furthermore, the influence of additional traps on thermal carrier generation and thus the device performance is investigated and geometrical variations in order to improve transport within the device are considered. The complex interaction of several mechanisms are presented in an elaborate case study.

Subsections

• 6.1 Ideal Classical Thermoelectric Generators
• 6.2 Non-Ideal Devices and Thermal Environment
• 6.3 Reduced Thermal Conductivity by Alloys
• 6.4 Stacked Lead Telluride Thermoelectric Devices
• 6.5 Pn-Junctions as Thermoelectric Devices
  • 6.5.1 Temperature Control by Graded Material Alloys
  • 6.5.2 Generation Enhancement by Additional Traps
  • 6.5.3 Optimization of Device Characteristics

M. Wagner: Simulation of Thermoelectric Devices