As part of the global energy transition, large amounts of hydrogen will be needed to provide fuel for heavy-duty transport, heat for industrial processes, and as a feedstock in chemical processes. To be environmentally sustainable, this hydrogen has to be generated using electricity via water electrolysis. With an increase in hydrogen production, however, the additional demand for freshwater in water electrolysis will start to compete with other freshwater consumers, especially humans. One approach to lessen the freshwater impact of water electrolysis is to use seawater as the process feed and perform in-situ water purification. Among others, this purification can be done by using forward osmosis membranes. However, previous research on this topic has only been performed at low current densities, while simultaneously demonstrating high cell overpotentials. Therefore, the overall aim of this thesis is to quantify the attainable current density of the forward osmosis-based purification method, and to optimise the cell performance through use of a zero-gap design. In addition, an evaluation of performance and material degradation is pursued. To this end, tests were conducted to quantify the water flux over the forward osmosis membrane within the given system. Next, a variety of long-term electrochemical tests were performed, both with and without water purification, and at different current densities. Furthermore, the cell voltages before and after testing were broken down into anode, cathode, and membrane contributions to gauge performance degradation and locate its origins. Finally, a variety of testing methods such as SEM, EDX, and XPS were used to evaluate the observed material and performance degradation and its causes. The maximum current density attainable within the designed system is found to to amount to 62.39 mA/cm2 . Furthermore, an initial performance improvement is found for the zero-gap cell design. This performance improvement is attributed to lower concentration and bubble overpotentials. However, the cell design exhibits strong material and performance degradation under all investigated testing parameters. The performance degradation is caused by increasing anode and membrane overpotentials. The anode overpotential increase is hypothesised to be due to the dissolution of a catalytically-active iron and nickel-rich anode top layer. The membrane overpotential is hypothesised to be due to membrane blockage by material deposits, and sodium-proton substitution in the membrane. In contrast, stable, or even improving, cathode overpotentials are found. This observation is attributed to the formation of catalytically superior metal phosphates and nickel hydroxide on the cathode surface. Lastly, the root cause of all degradative processes is found to be the use of a selective membrane, which enables a gradual acidification of the anolyte and thus creates a highly corrosive environment at the stainless steel anode. The results of this thesis allow for a better understanding of the inherent limitations connected to the use of FO membranes in electrolysis cells. It also offers a wide range of opportunities for improving the system and cell design in future research. Specifically, it enables a more careful membrane selection for future cell designs. By highlighting the limited achievable water influx with the given system, it also emphasises the need for larger FO membranes, and thus a different system architecture, or alternatively the need for a switch to a different draw solution or electrolyte.