Abstract: Over the last five decades, the advancement of information technology and computing has transformed our daily lives, and this may alter even more of our lives in the future. The continuation of this improvement directly relates to the downsizing of CMOS chips. However, further downscaling is facing severe challenges due to the increased power density and limited performance of existing computer architecture (because much energy and time are consumed for the movement of data between the processing unit and memory). To circumvent these limits researchers are looking to Neuromorphic Computing (NC), which is inspired by the human brain. As the memory and computation are co-localized in this system, the NC system changes the computing paradigm. In a human brain, computation is done in networks of synapses and neurons; likewise, in NC, information is processed and stored in memory. Thus, this system removes the energy burdens and latency of the conventional computing system. In the NC system, different resistive switching devices can be used as synapses such as Phase-change-memory (PCM), Magnetic-Random Access-Memory (MRAM), Ferroelectric-Random-Access-Memory (FeRAM), and Resistive Random Access Memory (RRAM). Compared to all other memory, RRAM devices provide lower footprints, high switching speed (less than 100ps), non-volatility, large endurance, and finally small power consumption and this has inspired me to explore the use of RRAM as synapses. RRAM devices store information in their conductance/resistance state and are composed of two metal electrode layers (Typically Pt, Al, Cu, Au) and an oxide layer (TiOx, TaOx, HfOx, other chalcogenides) that acts as a barrier layer. In the past decades, researchers have been investigating different materials for RRAM performance improvement. Very recently, Li-ion based RRAM has drawn the attention of the research community since Li-ion has the highest ionic conductivity. Among different Li-based materials (such as LiFePO4, LiCoO2, LiNbOx, LiFe5O8, LixTiyOz, and LiPON) LiPON (Lithium Phosphorous Oxynitrate) has proved more promising as a barrier layer in RRAM due to its superior conductivity. All these factors motivate me to fabricate Li ion-based RRAM synapses to use for NC. Optimized growth condition of LiPON layer has been done by XPS characterization, and highest ionic conductivity of the layer has been found. The micrometer size devices have already been fabricated and characterized. Initial results of devices with LiPON layer on top of TiOx are stunning as they reduced the switching current by a factor of ten. To obtain a standard device, several device structures need to be studied, and the progress of this study is continuing. Also, to mimic the brain like behaviour, plasticity of the device will be explored. Thus, how RRAM synapses might pave the way for the ‘brain on a chip’ will be studied. Since my project aim is to investigate new materials for optimized RRAM device performance and construct synapses with an array of RRAM devices, it is inherently elevating the status of nanotechnology.