Uncertainty quantification for first-principles simulations of thermoelectric properties

Thermoelectric materials can convert waste heat into electricity, dramatically improving energy efficiency across a wide range of technological applications, from handheld devices to transport to manufacturing processes, but current thermoelectric materials are not efficient enough to be cost-effective. The aims of this project are to develop predictive computational methods based on first-principles simulations, including the uncertainties arising from the methods employed. These methods will enable rapid screening of materials for thermoelectric applications, and reliable modelling of materials properties. Of particular interest are the computational discovery of new thermoelectric materials, and patterning materials at the nanoscale to improve thermoelectric performance.