Measurement of the high-temperature Seebeck coefficient of thin films by means of an epitaxially regrown thermometric reference material

Abstract

The Seebeck coefficient of a typical thermoelectric material, silicon-doped InGaAs lattice-matched to InP, is measured over a temperature range from 300 K to 550 K. By depositing and patterning a thermometric reference bar of silicon-doped InP adjacent to a bar of the material under test, temperature differences are measured directly. This is in contrast to conventional two-thermocouple techniques that subtract two large temperatures to yield a small temperature difference, a procedure prone to errors. The proposed technique retains the simple instrumentation of two-thermocouple techniques while eliminating the critical dependence of the latter on good thermal contact. The repeatability of the proposed technique is demonstrated to be ±2.6% over three temperature sweeps, while the repeatability of two-thermocouple measurements is about ±5%. The improved repeatability is significant for reliable reporting of the ZT figure of merit, which is proportional to the square of the Seebeck coefficient. The accuracy of the proposed technique depends on the accuracy with which the high-temperature Seebeck coefficient of the reference material may be computed or measured. In this work, the Seebeck coefficient of the reference material, n+ InP, is computed by rigorous solution of the Boltzmann transport equation. The accuracy and repeatability of the proposed technique can be systematically improved by scaling, and the method is easily extensible to other material systems currently being investigated for high thermoelectric energy conversion efficiency.

Thermoelectric Performance Enhancement of Film by Pulse Electric Field and Multi‐Nanocomposite Strategy

“…The error bar of resistivity was estimated from the four resistivities above, while that of Seebeck coefficient was approximately 5% from the two‐thermocouple measurements. [ 50 ] In addition, the measurement accuracy of Seebeck coefficient was 7% according to the commercial apparatus maker. The picosecond time‐domain thermoreflectance (TD‐TR) technique using a customized focused thermal analysis system based on PicoTR (PicoTherm) was utilized to measure the thermal conductivity of the samples in the cross‐plane direction.…”

Experimental setup for anisotropic thermoelectric transport measurements using MPMS

“…In the present setup copper-constantan type-T thermocouple system is used to measure the ΔT across the sample between position (2, 4) and/or equivalently (1,3). The use of two separate sensors to measure ΔT has also been reported in literature [29], however, this approach suffers from inherent errors due to subtraction of large numerical values to estimate small values [33]. For the present setup the two differential thermocouple tips are fabricated by welding a copper wire on the two ends of a constantan wire as shown in figure 3.…”

The impact of commonly used approximations on the computation of the Seebeck coefficient and mobility of polar semiconductors

“…While this is worse than the precision of typical room-temperature Seebeck coefficient measurements (62%), it is comparable to the precision of high-temperature measurements, 65%. 9 In contrast, the combined RTA-BTE truncation underestimates the RT Seebeck coefficient by $16%. It is to be noted that the Z figure of merit varies as the square of the Seebeck coefficient.…”

“…A recently proposed technique for hightemperature Seebeck coefficient measurements of thin-films addresses these challenges by using an epitaxially regrown thermometric reference material. 9 Accurate computation of the Seebeck coefficient of the reference material is critical to the accuracy of this measurement.…”