;Graphene is a two dimension thin film consisted with carbon atoms in honeycomb ordered. Due to its unique band structure, graphene as unique electronic and material properties. Therefore graphene is expected to have great application potential in the future. However, it is still challenging to produce large amount and perfect graphene which is suitable for the application. This shortage limited the ceiling of the potential of graphene. Among all possible solution to produce large amount and defect-free graphene, chemical vapor deposition seems like to be the possible way to fabricate industry scale graphene. Three of the key issues in chemical vapor deposition is the time, cost and quality of graphene. First, nowadays it is the guarantee way to form large and perfect graphene in lower pressure chemical vapor deposition system. However, such a grain needs long growth time which is impossible to meet the needs of industry scale. Also, the long heating and cooling time for furnace decrease the throughput of graphene. Second, hot wall furnace waste unwanted heat into the environment which is not helpful for the chemical vapor deposition process. Third, the quality of graphene of CVD is comparing poorly with mechanical exfoliation. In order to suit above issues, rapid thermal chemical vapor deposition is considered. Although rapid thermal chemical vapor deposition is a low cost and fast production way to grow graphene, the graphene grain is small due to the non-equilibrium heating process. Recently this issue is solved by growing graphene on copper oxide. By exposing the oxygen on the defect on copper, rapid thermal chemical vapor deposition is able to grow large single crystal graphene. However, the underlying mechanism is still the shortage. In this work, we investigate the role of oxygen in graphene chemical vapor deposition on copper oxide. We find out the mechanism of the nucleation and growth process with oxygen exposure by extending the JMAK model into a non-equilibrium region to explain the initial situation of CVD process. The extending JMAK model is able to explain the increasing in nucleation rate. In addition, a correlation function analysis in traditional condensed matter physic is workable to quantify the spatial distribution and uniformity of Graphene Island. This analysis also points out the transition from carbon forming new grain at local nucleation site to joining larger cluster and coalescence after oxygen exposure.
