Atomic layer deposition of thin films containing alkali metals

Abstract

This thesis presents experimental work on thin films of different compounds containing lithium, sodium or potassium has been synthesized by atomic layer deposition (ALD). The overall motivation for this work has been to develop materials and methods to improve lithium ion battery technology by using ALD. A cathode in a lithium ion battery should have a long operating life, be environmentally benign and have high capacity and power density. Vanadium oxides are popular as cathodes in lithium ion batteries due to their relative low price and potentially high capacity. Most studies of vanadium oxide cathodes shows relatively short lifetime of the cathode or relatively fast cathodes. In this work a high power thin film cathode of V₂O₅ for lithium ion batteries has been developed. The cathode is deposited by ALD using VO(thd)₂ and ozone, which displays a rather peculiar type of ALD-growth. This peculiar growth is studied in detail, and the optical properties of these films are investigated. The films have an unusually rough surface, and it was found that a 10nm thick film deposited at 235ºC consisted of individual nano particles. The 10 nm thick cathode has been shown to endure more than 4000 dischargecycles at 120C and almost 1600 cycles while staying within 80% of the original capacity. The same cathode was also shown to sustain discharge rates of 960C which corresponds to a discharge in 3.75s. The power density obtained in this work bridges the gap between super capacitors and batteries and the combination of long lifetime and high discharge rate is not found previously for thin film batteries of V₂O₅. ALD of lithium containing materials has attracted widespread interest the last few years. The number of known precursors for lithium has grown, but the complete picture is still not understood. Therefore lithium hexamethyldisilazane (LiHMDS) is explored as a precursor for ALD of lithium compounds. The precursor is shown successful in deposition of Li₃N, Li₂CO₃ and LiNbO₃. The deposition of Li3N may be an important step to deposit solid electrolytes and the deposition of Li₂CO₃ proved to be important for proving the growth of oxides using this precursor. When comparing the growth of Li₃N and Li₂CO₃ it was found significant difference in the surface chemistry. The LiNbO₃-films were shown to be ferroelectric with an unusually high coercive field. It proved possible to deposit epitaxial LiNbO₃ on single crystal substrates of LaAlO₃ and Al₂O₃ and the orientation of the films could be controlled by the orientation of the substrate. A milestone in atomic layer deposition of lithium compounds would be to deposit a full battery. In order to realize this, a lithiated cathode material must be deposited. The cathode material LiMn₂O₄ was also studied in this work. It was discovered that the amount of lithium in the deposited films is more or less independent of the number lithium cycles to manganese cycles. It is hypothesized that the ligand of the lithium precursor reduces the manganese and the lithium is intercalated into the manganese oxide. This is a new approach to ALD of lithium compounds and the term film body controlled lithium deposition is used to describe the mechanism. The use of LiHMDS is also attempted in deposition of LiMn₂O₄, with no success. Sodium and potassium are among the few elements in the periodic table which are not yet used in ALD. Sodium and potassium are relatively similar to lithium and exploring the deposition of these elements will hopefully shed new light on the deposition of lithium compounds. Many oxides of sodium and potassium also have piezo- and ferroelectric properties, and the sodium ion battery is predicted to be a way to combat lithium shortage. Atomic layer deposition of sodium and potassium oxides is reported for the first time in this thesis. Six different precursors are investigated and evaluated and precursors for sodium and potassium. The initial study was performed by depositing sodium and potassium aluminates, in order to evaluate the precursors. The process for the aluminates was found to scale up to the 200mm wafer scale. The precursors were found to work in a large temperature window and react with both water and ozone, thus proving to relatively flexible and possible to combine with most known ALD-processes. Further development into deposition of sodium based ferroelectrics then explored by deposition of sodium tantalate and sodium niobate.