Speaker: Professor Rene van Roij (Utrecht University)
Due to the irreversible mixing of fresh and salty water an enormous free-energy dissipation of about 2kJ per liter river water takes place in river mouths, equivalent to a waterfall of 200m. This source of energy, which is renewable and emission-free, is currently untapped but could globally account for several percent of the energy demand (and for much more locally in river deltas). Harvesting this so-called ‘blue energy’ has become possible in recent years due to the development of nanostructured devices, either based on ion-selective membranes or on nanoporous solid-state electrodes that can form supercapacitors with internal surfaces of the order of a km2/kg. In this talk we will mostly focus on cyclic charging and discharging processes of these supercapacitors immersed in salty and fresh water. We will perform a thermodynamic analysis of this cycle, mapping the ion flow and electric voltage-charge work of this ‘blue engine’ onto the heat flow and the mechanical pressure-volume work of Stirling’s heat engine. Our mapping naturally leads to the prediction of the most efficient ‘blue cycle’ as an analogue of the Carnot cycle, which harvests the full 2kJ/liter fully reversibly. Interestingly, running the Carnot-like cycle backward is the basis for the thermodynamically cheapest desalination process, where brackish water is separated into fresh and salty water at the expense of a minimum energy input. Microscopically, on the nanometer scale, these devices are governed by electric double layers at the electrode-electrolyte interface, where ionic packing, hydration, and polarisation affect the voltage-charge relation and the capacitance, which we will briefly discuss. Finally, we will focus on some recent and ongoing work on maximum power conditions (involving non-equilibrium (dis-)charging processes on the RC-time scale), temperature effects (involving cold sea water and warm fresh water for more efficient blue-energy harvesting using industrial waste-heat), and generalisations to run these devices on other chemical gradients (such as CO2 in clean air and in combustion gases).