NeMo: A network model program for analyzing the thermoelectric properties of meso and nanostructured composite materials

We present our NeMo program package where we implemented a network model for calculating the thermoelectric properties, i.e. total electric and thermal reiststances and total Seebeck coefficient, of structured composite materials. The real structure of the composite sample is translated into a pixel grid where local transport properties are assigned to each pixel including bulk-like properties as well as interface properties to adjacent pixels which can be adjusted to account for real interfaces. The resulting networks are embedded between a left and a right contact, each at constant temperature and constant electrostatic potential. By using nodal analysis, the total thermal and electric resistances of the structure, as well as local voltages or temperature differences between the pixels are obtained. In an additional step, the local temperature differences are used to simulate local Seebeck voltages as voltage sources between the pixels in order to obtain the total Seebeck voltage and, thus, the Seebeck coefficient, of the structure. In a series of examples, we discuss the merits and limitations of the model for describing the thermoelectric properties of various composite materials of interest and demonstrate that, despite its simplicity, it may yield helpful insight into the correlation between mesoscopic structure and macroscopic thermoelectric properties.