Thermoelectric energy harvesting from day–night temperature swings with latent heat storage: Enhancing the efficiency by combining natural and Marangoni convection

Natural energy sources are a solution to power low-consumption electronic devices, such as sensors, in environments where batteries are impractical. Among these sources, thermoelectric conversion stands out for its ability to generate power from temperature fluctuations. However, its efficiency is severely constrained by the small temperature differences typically seen during natural day–night cycles, which limits its usability when relying on ambient thermal gradients. Through realistic physical modeling and 3D numerical simulations, we demonstrate that coupling a thermoelectric generator with a latent heat storage unit significantly enhances the conversion of natural day–night temperature swings into electricity. This enhancement is achieved by combining natural and Marangoni convective heat transfer. We utilize a standard thermoelectric module (Seebeck coefficient of α = 0.027 ) paired with a heat storage unit containing the phase change material hexadecane, which has a Prandtl number of 45.5 and configured with a Bond number of 8. Using temperature profiles representative of Western Europe, Eastern Europe, and Brazil, we illustrate the practical and broad application of these enhanced micro-energy harvesters to power environmental sensors. Over a 24-hour period, the combined effects of buoyancy and thermocapillarity in a 16 cm 3 heat storage unit yield harvested energies (average power densities) of 2.6 J ( 29.7 μ W / cm 2 ), 1.4 J ( 16.4 μ W / cm 2 ), and 2.4 J ( 27.2 μ W / cm 2 ) for the temperature profiles of Central Europe, Western Europe, and Brazil, respectively. Notably, even with weak thermocapillary effects at this Bond number, Marangoni convection doubles the harvested energy and average power density for the Central and Western Europe profiles compared to natural convection alone. The harvested energy is sufficient to uninterruptly power low-consumption sensors monitoring humidity, pressure, and ambient temperature, along with the necessary accompanying electronics. Importantly, this micro-energy harvester leverages fundamental physical properties of liquids: density variation with temperature (natural convection) and surface tension variation with temperature (Marangoni convection). The robustness of these results provides a foundation for further enhancements under more complex configurations. • Natural and Marangoni convection to enhance efficiency in thermoelectric conversion. • Micro-energy harvesting from day–night temperature swings in three distant world regions. • Realistic 3D modeling of TEGs coupled to latent heat storage. • Enhancement of thermoelectric efficiency based upon fundamental physical properties of liquids. • Examples of applications for powering temperature, humidity, pressor sensors.