Solid-state batteries is one of the main contenders for domination of the future battery marked. Thin film technology is important in the development of these batteries. In this work, we have shown that amorphous thin film FePO4 with a thickness around 10 nm, deposited by atomic layer deposition (ALD), can reach a specific power above 1 MW/kg and approach theoretical capacity at lower currents. The 10 nm thin film also shows very good cycling stability at elevated currents and can retain 70 % of peak capacity after 8000 cycles at 80 µA (40C). The material also shows a significant self-enhancing mechanism leading to an increase in capacity during early cycling stages. We observed a capacity increase of 90 % for 10 nm after 100 cycles at 80 µA. In this study, we used quartz crystal microbalance (QCM) analysis to establish a stable ALD process for depositing amorphous thin films from the Fe-P-O system. By varying the pulsing ratio between the precursors, we obtained films with different compositions and chose to study Fe4(P2O7)3 and FePO4 more in detail. The films were uniform and flat with an RMS roughness below 1 nm. As the FePO4 films proved to be significantly better than the Fe4(P2O7)3, we focused mainly on FePO4. We used galvanostatic cycling (GC), cyclic voltammetry (CV) and electrochemical impedance spectroscopy (EIS) to characterize the electrochemical properties of the thin films. An important part of this study was to develop a good baseline for testing, including the use of reference batteries. In this work, we confirmed that LiClO4 is a better choice than LiPF6 as electrolyte for testing thin film cathodes, because of minimal side reactions with the steel casing. The FePO4 thin films show a combination of capacitive and redox behavior where both contribute to the capacity. In this study, we have tried to separate the two contributions and find their thickness and current dependency. In an attempt to increase the area capacity of the cathodes without increasing the film thickness, we created soot substrates with high surface 3D structures of carbon, deposited from the flames of a candle. We managed to maintain the structure and evenly coat it with FePO4. Despite the increase in mass, we obtained no higher capacity or better battery performance from the soot batteries.