A new technique, principally applicable to metals and alloys, has been developed for high resolution absolute thermoelectric power (TEP) and differential thermoelectric power measurements. It utilizes two continuously monitoring dc nanovolt amplifiers, one for the sample couple thermoelectric voltage and the other for a thermocouple voltage proportional to temperature differences developed across the samples. The outputs of these amplifiers are attached to the channels of an X-Y recorder, thus yielding direct slope-proportional data. By using heat from a low-mass, electrostatically guarded heater to develop transient temperature gradients across samples, it is possible to dynamically determine the absolute TEP of a sample versus lead (Pb) or the differential TEP of two nearly similar samples, one of whose composition or history of treatment has been slightly altered. Reducing thermal time constants with an independent heat leak permits the completion of a TEP measurement in a few seconds, thus yielding high resolution determinations under adverse experimental conditions. The use of a special sample necking and ``heat sinking'' techniques, as well as a low melting point, low TEP, ternary eutectic indium alloy solder, make it possible to obtain results with a range of high resolutions. The experimental resolution attainable with these two dynamic techniques can be summarized as ±[10−3)(ΔT)−1(μV/K)+1.0%] for absolute TEP and ±[ΔT)−2+Sabsolute2]1/2 10−3(μV/K) for differential TEP, where ΔT is the range 0.01–1.0 K increment of differential temperature and Sabsolute is the approximate value of the thermopower of one of the samples. Data utilizing this newly developed technique are presented on Chromel-P, Wood's metal solder, and the ternary eutectic InSnCd solder alloy, as well as differential TEP measurements of two opposing Chromel-P samples over the temperature range of T = 1.5–300 K.
Read full abstract