▎ 摘　　要

This paper presents a process metamodel-based artificial neural network full factorial experimental design and analysis to study the yield of lengthy hexagonal graphene grown by chemical vapor deposition. All of the process variables of chemical vapor deposition such as temperature, pressure, and gas flow rate under the study played a role in influencing hexagonal graphene length; the current study investigated their main effects and interactions. The metamodel-based analysis demonstrates that the hydrocarbon flow rate and the pressure are the most statistically significant factors that influence the length of hexagonal graphene. In particular, minimum and maximum values of the chamber pressure are not significant in terms of the concentrating effect they may have on the flowing mixture of gases with very small flow rate, i.e. 50 sccm. At the highest flow rate of 400 sccm, the chamber pressure stepped up to 764 Torr, which can support the growth reaction to the extent that the resultant hexagonal graphene length of 900 mu m can be achieved. However, the two level effect of the flow rate can optimize the length to 990 mu m and approximate to 1390 mu m at 700Torr and 764Torr, respectively. In addition, the response surface graph confirms the factors of significance and adds that higher flow with lower pressure will consistently yield tall hexagonal graphene. We found that gas flow rate is the most significant of the control variables and only the optimum value of the gas flow rate of 225 sccm can ensure the growth of tall hexagonal graphene. We also found that the interaction of flow rate with temperature of the gases in the chamber is extremely significant to the quality of output. Outcomes of this investigation are beneficial for moving close to producing hexagonal graphene on production scale for future applications.