Sustainable electricity generation by the plant microbial fuel cell Fossil fuels are currently the main source of electricity production. Combustion of fossil fuels causes air pollution severely affecting human health and nature. This results in an increasing demand for renewable electricity sources. One of the emerging renewable electricity technologies is the plant microbial fuel cell (PMFC) as explained in chapter 1. PMFC generates electricity from the rhizodeposits of living plants. Naturally occurring electrochemically active microorganisms oxidize the rhizodeposits producing electrons at the anode of the PMFC. The electrons flow from the anode, via an external circuit where the electricity is harvested, to the cathode. At the cathode, the electrons reduce oxygen to water. PMFC is based on naturally occurring sustainable and renewable processes without net emissions and competition for arable land or nature. Large scale application of the PMFC is preferred in wetlands because a large waterlogged area is required. Prior to application, the cathode limitations of the PMFC have to be solved. Oxygen reduction at the cathode is slow, limiting the current and power output of the PMFC. An unsustainable chemical cathode is often used in PMFC research to overcome the cathode limitations. The sustainable oxygen reducing cathode has to be catalyzed when integrated in the PMFC. Most chemical catalyst are expensive and prohibit the commercial use in the PMFC. Oxygen reduction can also be biologically catalyzed by cheap and self-replenishing microorganisms. Next to the biocathode, also a suitable design of the PMFC has to be developed before application in wetlands. A tubular design was previously developed which can be invisibly integrated in wetlands. However, this design still used a chemical cathode and energy intensive pumping. The oxygen reducing biocathode should be integrated in the tubular design and oxygen should be passively supplied in the cathode. The objective of this thesis is to apply PMFC in wetlands with a sustainable biocathode. First, the biocathode is integrated in a lab scale PMFC. Afterwards, the PMFC is installed in wetlands using an improved tubular design with an integrated biocathode and passive oxygen supply. Lab scale experiments: integration of the biocathode and electricity localization in the bioanode of the PMFC In chapter 2, the oxygen reducing biocathode is integrated in a flat plate lab scale PMFC replacing the chemical ferricyanide cathode. The PMFC operated as a completely biocatalyzed system for 151 days. The sustainable PMFC with a biocathode was able to generate more power than the PMFC with a chemical cathode. The long term power generation of the lab scale PMFC improved from 155 mW m-2 plant growth area (PGA) to a record of 240 mW m-2 PGA. This record was reached due to the higher redox potential of oxygen reduction compared to ferricyanide reduction. Oxygen reduction was effectively catalyzed by microorganisms lowering the voltage losses at the cathode. As a result, the PMFC with a biocathode operated at a 127 mV higher cathode potential than a similar PMFC with a chemical ferricyanide cathode. The long term current generation of both PMFCs was 0.4 A m-2 PGA. The current generation was likely limited by the substrate availability in the anode of the PMFC. In chapter 3, the biocathode is further investigated. This chapter shows that the oxygen reducing biocathode can also catalyze the reversible reaction, water oxidation. Water is the most abundant electron donor available for electrochemical fuel production like the reduction of protons to hydrogen and the reduction of carbon dioxide to hydrocarbons. However, the water oxidation reaction is currently hampering the development of large scale water oxidation technologies. A bioanode containing electrochemically active microorganisms was able to reach a current density of 0.93 A m-2 at 0.7 V overpotential with a 22 % Coulombic efficiency linked to water oxidation. An optimized system could be used to produce fuels on a large scale. The flat plate PMFC of chapter 2 was also used to localize the electricity generation in the PMFC (chapter 4). In this experiment, the anode was partitioned in 30 separate small anodes at different width and depths. The current generation of each anode was analyzed over time and linked to the plant roots. The results show that after a start-up period of 70 days, significantly higher current was generated at anodes close to the plant roots due to rhizodeposition. Besides rhizodeposition (i.e. electron donors), the plant roots also excrete oxygen which is an electron acceptor lowering the current generation of the PFMC. Also oxygen was measured at the anodes close to the plant roots. This likely resulted in internal currents in the PMFC. Current was likely generated both from living and death roots. The electrons in the PMFC were probably transferred via mediators to locations without roots as mediators were present also at locations without plant roots. These mediators were likely excreted by plants and/or microorganisms in the anode. Electrons were likely not transferred over centimeter distance through conductive microorganism on the plant roots in the PMFC. Installation of the tubular PMFC with an integrated biocathode in wetlands After the successful integration of the biocathode in the PMFC, the focus of the research changed to application in wetlands. Two wetlands with an abundant occurrence in the Netherlands were investigated in this research. The first wetland was a Phragmites australis dominated fen peat soil, a large perennial grass. The peat soil in this research was collected in national park Alde Feanen in the north of the Netherlands. The second investigated wetland was a Spartina anglica dominated salt marsh. Spartina anglica is a perennial grass found in coastlines spread over the world. The salt marsh was collected in the Oosterschelde tidal basin in the southwest of the Netherlands. The first experiment in the wetlands was conducted to investigate the spatial and temporal differences in current and power generation in and between wetlands (chapter 5). PMFCs in the salt marsh were able to generate more than 10 times more power than the same PMFCs in the peat soil (18 vs 1.3 mW m-2 PGA on a long term). The higher power generation is mainly explained by the high ionic conductivity of the salt marsh and the presence of sulfide which is also oxidized next to the rhizodeposits at the anode of the PMFC. The top layer of the salt marsh generated most power due to the presence of the plants and tidal advection. In the peat soil, there was no significant difference in power generation over depth. Even though, in the top layer more living roots were present. Also the dead roots and organics in peat can be oxidized by the PMFC. In chapter 5, also the maximum current and power output of the wetlands was predicted based on rhizodeposition of the investigated plants and microbial processes in these wetlands. The calculations showed that the potential current generation of PMFC in the salt marsh is 0.21-0.48 A m-2 PGA and in peat soil 0.15-0.86 A m-2. In the peat soil, the PMFC is potentially able to generate a power density up to 0.52 W m-2 PGA. The second experiment in the wetland was the installation of a tubular PMFC with an in situ started oxygen reducing biocathode and passive oxygen supply into the cathode (chapter 6). The anode was the outside of the tube and placed directly between the plant roots. The oxygen reducing biocathode was located inside the tube. A silicone gas diffusion tube was placed in the cathode compartment to passively supply the required oxygen. The tubular PMFC with biocathode was successfully installed and started in the peat soil reaching a maximum daily average power generation of 22 mW m-2 PGA. In the salt marsh, the tubular biocathode PMFC only started while supplying pure oxygen in the gas diffusion tube. Air diffusion did not result in the start-up of the biocathode, likely because the oxygen was directly reduced via internal currents and therefore more oxygen was required. Once started with pure oxygen, the tubular PMFC was able to generate 82 mW m-2 PGA which was again higher than the peat soil. Completely biocatalyzed tubular PMFC were installed in both wetlands with natural occurring microorganisms in the anode and cathode. The power generation can be further increased by improving the PMFC design limiting crossover of oxygen and substrate. Future outlook: application of the PMFC in wetlands In chapter 5, the potential power generation of the two investigated wetlands was calculated. In chapter 7, these calculations were extended to a worldwide scale. PMFC applied in all wetlands could generate 0.67 to 1.35 TW and could cover 30 to 60 % of the global electricity consumption. 70 % of all the potential power could be generated in the tropics. Worldwide, 1.1 billion people have insufficient access to electricity from which 88 % lives in the tropics (i.e. Sub-Saharan Africa and South Asia). PMFC could be used to reach universal access of electricity in these locations and decrease the amount of premature deaths due to air pollution. PMFC can be applied with passive or active oxygen supply from the outside air into the silicone tube. The used tubular PMFC with passive oxygen supply can have a maximum length of less than one meter. Active supply of oxygen reduces the net power output of the PMFC, but allowing installation of long tubular PMFC. However, in both cases the material costs should be significantly reduced for economically feasible application at large scale. The costs of the material should be decreased to less than 1 % of the current PMFC costs to have a payback time of 50 years in the Dutch electricity market for only the tubular PMFC. Further cost reduction is required when also the current collectors, electricity transmission, production and installation costs are included. Application of PMFC in remote locations increases the economic feasibility of the PMFC as the PMFC could be applied independent from the grid reducing the transmission costs and avoiding the regular electricity network charges. Application of the PMFC in the total area of Spartina anglica salt marsh in the Oosterschelde, the location were the plants were collected, could produce a total of 11.6 GWh yr-1. The Oosterschelde could produce the electricity consumption of 8,360 persons and as such produce the electricity need of an average village directly located at the tidal basin. The Phragmites australis peat soil in the Alde Feanen national park could produce 2.5 GWh yr-1. The electricity could be directly used for ecotourism purposes, for example for the use of electric boats and a holiday park.