Methodology for Minimizing Losses for the Harman Technique at High Temperatures

A high-temperature Harman technique for measuring material ZT, or thermo- electric figure of merit, with increased measurement accuracy is presented. Traditional Harman tests are sensitive to radiation heat losses at elevated temperature, and measurement errors are minimized by applying current in positive and negative polarities while thermally sinking the sample base to a constant temperature for both polarities (referred to here as bottom temperature match, BTM). Since the sample top temperature differs between polarities in BTM, the heat losses are not equivalent and still add error to the ZT measurement. A modification is presented in which the sample base temperature is adjusted until the sample top temperature is the same in both polarities (referred to as top temperature match, TTM). This ensures that heat losses from the top of the sample are nearly identical and cancel out of the ZT calculation. A temperature-controlled radiation guard maintained at the sample top temperature is employed to minimize radiation loss and increase ZT calculation accuracy. Finite-element analysis (FEA) models suggest that ZT errors less than 5% for Bi2Te3 alloys tested at 250°C are achievable, a 30% improvement over the conventional BTM approach. Experimental results support these trends.