THE THERMOELECTRIC POWER OF LIQUID LITHIUM-SODIUM ALLOYS

Abstract

14easurements of the thermoelectric power of liquid Li-Na alloys have been performed throughout the entire concentration range and at temperatures up to 400'~. The results are in reasonable agreement with calculations within the diffraction model. In the alloy with critical composition (65 at.% Li), critical effects have been observed in a temperature range up to several degrees above the phase separation temperature. INTRODUCTION. The thermopower measurements presented in this paper form part of a systematic investigation of the electronic properties of liquid Li-na alloys. In an earlier paper (Feitsma et al. 1975a) measurements of the electrical resistivity are presented and discussed; the results could be explained reasonably well within the diffraction model. Also Knight shift measurements have been done (Feitsma et al. 1975b); both the 7 ~ i and the 23~a Knight shift show a linear dependence on the concentration. From a thermodynamical point of view the system Li-Na is rather exceptional, as it is the only alloy of Li with an alkali metal which exhibits miscibility for all concentrations above 3 0 0 ~ ~ . However neutron diffraction measurements in an alloy of nearly the critical composition showed that a preference for like nearest neighbours is still present up to 150 K above the critical temperature Tc (Ruppersberg and Knoll 1977), whereas close to Tc critical opalescence has been observed (Brumberger et al. 1967). Resistivity measurements (Schiirmann and Parks 1971) demonstrate that the increase of the concentration fluctuations close to T also influences the electronic properties: a strong increase of the temperature derivative of the electrical resistivity is observed when the temperature approaches T c. EXPERIMENTAL. The measurements were based upon the small-AT!' method described by Feitsma et al. (1978). A description of the apparatus used will be given in a separate publication. The thermovoltages of the alloys were measured against copper. The measurements were performed at several temperatures with intervals of approximately 15Oc, between the liquidus temperature and 4 0 0 ~ ~ . The error in the thermopower is estimated to be t0.4 u ~ / O ~ . RESULTS. Measurements on pure sodium and pure lithium provided agreement within 0.2 ~v/'c with previous measurements by Kendall (1968), Davies (1969) and Feitsma et al. (1978). In figure 1 the thermopower, S, is shown as a function of concentration FIG. 1. Thermopower, S, of liquid Li-Na alloys. -: 310°C; --a--. 400°C. at 3 1 0 ~ ~ and 400°C. S was found to depend almost linearly on the temperature for all concentrations. The temperature derivative of the thermopower, dS/ dt, at 355Oc is shown in figure 2. In the alloy of the critical composition (65.3 at.% Li), the temperature was lowered in steps of 0.5Oc close to T Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:19808132 FIG. 2. Temperature derivative, dS/dt, of the thermopower of liquid Li-Na alloys at 355'~. in order to observe possible critical effects. A sudden increase of S, starting at about 3Oc above Tc, has been found (figure 4) . FIG. 3. Thermopower, S, of liquid Li-Na alloys at . Toigo-Wood. experimental; -.-.o--.-. 310°c. -. ruff screening; ---A---: Hartree screening; 0 --: Toigo-Woodruff, structurefactors from Feitsma et al. (1975a). DISCUSSION. In this section the experimentally obtained results will be compared with the results of calculations within the diffraction model. Within this model the thermopower is related to the formfactors w (k ) and the partial structure factors q F S ( q ) , S (q) and SCC(q) by the following formuNN NC lae: where F(2kF9kF) F s = and r = (2)