Thermoelectric phenomena are a way to directly convert thermal energy into electrical energy; they thus are at the heart of several researches in the field of energy conversion. The optimization of the thermoelectric generators includes materials improvement but a reflection on their working conditions is also mandatory. The contribution of the thermal contacts between the generator and the heat reservoirs is a factor that will change the optimum operating conditions of the generator. Using the concept of convective heat flow, developed by Thomson more than 150 years ago, we generalize the classical expression of maximum power conditions. Moreover, we note that these conditions may be reduced to impedance matching conditions, both thermal and electrical. In addition to its practical interest, the thermoelectric generator is also an ideal model system to study the theory of coupled transport and of irreversible phenomena. Using the description of this system given by Ioffe, we show that the maximum power efficiency, a coefficient of performance at the heart of finite time thermodynamics, expressed as a simple function of the system parameters. The novelty of this work is based on a proper consideration of internal dissipation associated with the energy conversion process. The results are then generalized to other thermal engines such as the Feynman ratchet.
