There is an increasing demand for high performance glass thin films (GTFs) for applications, such as microbatteries, electrochromic systems, photonics, biomaterials or protection against corrosion. In particular, lithium phosphorus oxynitride (LiPON) GTF is currently the commercial standard electrolyte for all-solid-state microbatteries, which are promising devices for a broad range of applications pertaining to communication, consumer electronics, products and people identification, traceability, security (bank transaction) as well as to smart environment and the internet of things. The major limitation of LiPON GTF is its limited Li+ conductivity, 3.3·10-6 S.cm-1 at 298 K, a value, which is 3 orders of magnitude lower than that of conventional Li-ion cells using liquid electrolytes. Recently it has been shown that the incorporation of a second former, such as SiO2, or sulfates in LiPON GTFs can dramatically enhance the ionic conductivity. Nevertheless, the composition space for these GTFs still needs to be explored and the rational improvement of the conductivity of these GTFs requires to better understand how these changes in the chemical composition affect the atomic-level structure and hence, the mechanism of Li+ conduction. The characterization of GTFs is challenging since they are amorphous, they contain multiple molecular patterns and their volume is small. This project aims at the rational improvement of the properties (ionic conductivities, chemical and thermal stabilities) of these innovative GTFs by determining the relationships between their chemical composition, their atomic-level structures and dynamics as well as their properties. We will explore the composition space of LiPON GTFs incorporating a second glass former, such as SiO2, or lithium sulfate. These GTFs will be prepared by radiofrequency (rf) magnetron sputtering. We will determine the effect of these composition changes on the local atomic-level structure and dynamics by developing and applying advanced solid-state Nuclear Magnetic Resonance (NMR) methods (small coils, high-field, paramagnetic doping…) suitable for the characterization of thin-films. Dynamic Nuclear Polarization (DNP) will also be employed to enhance the NMR signals of the surface nuclei and better understand the electrode/electrolyte interfacial phenomena. The medium-range positional order in the GTFs will be investigated by Transmission Electron Microscopy (TEM) and Pair Distribution Function (PDF) analysis. TEM and annular dark field scanning TEM (ADF-STEM) will be combined to image the structure of glass networks in glass ultra-thin films. PDF analysis will provide information about the bond length, the atom coordination numbers and the geometry. Finally the electrical and electrochemical properties of the GTF electrolytes, bare and integrated in microbatteries, will be measured. These properties will be correlated to the chemical composition and the atomic-scale structure and will be used to elaborate in a rational way GTF with optimized properties, including (i) Li+ conductivity > 10-5 S.cm-1, (ii) low electronic conductivity, (iii) low contribution to the overall cell impedance when integrated into microbatteries and (iv) good (electro)chemical and thermal stabilities, notably near the interface between GTF electrolyte and lithium metal electrodes. The ultimate long-term goals of the project are (i) to improve the performance of microbatteries and (ii) to change the way in which material scientists and chemists characterize GTFs used for various applications (electrolyte, coating…).