The reduction of the thermal conductivity in nanostructures opens up the possibility of exploiting for thermoelectric purposes also materials such as silicon, which are cheap, available and sustainable but with a high thermal conductivity in their bulk form. The development of thermoelectric devices based on these innovative materials requires reliable techniques for the measurement of thermal conductivity on a nanometric scale. The approximations introduced by conventional techniques for thermal conductivity measurements can lead to unreliable results when applied to nanostructures, because heaters and temperature sensors needed for the measurement cannot have a negligible size, and therefore perturb the result. In this paper we focus on the 3$omega$ technique, applied to the thermal conductivity measurement of suspended silicon nanomembranes. To overcome the approximations introduced by conventional analytical models used for the interpretation of the 3$omega$ data, we propose to use a numerical solution, performed by means of finite element modeling, of the thermal and electrical transport equations. An excellent fit of the experimental data will be presented, discussed, and compared with an analytical model.
Download