Machine learning enhanced ultra-high vacuum system for predicting field emission performance in graphene reinforced...

The field emission performance of aluminium-based metal matrix composites reinforced with graphene (AlGr-MMCs) has garnered significant attention due to their potential applications in advanced electronics and in materials-based cathode systems. The field emission performance plays a crucial role in the high-power micro-wave tube devices and in energy applications, where material composition significantly influences emission stability and efficiency. This research work explores the impact of graphene incorporation into aluminum-based metal matrix composites (AlGr -MMCs) on field emission characteristics. By leveraging machine learning (ML) models, we predict the trends of emission current density (J) as a function of the applied electric field(E) and the emission current stability (I) over time(t) for Aluminium-Graphene (AlGr) composites with varying graphene weight% (wt%) greater than 1 and less than 2 (1.25, 1.5, 1.75, and 2.0). A two-stage machine learning framework was implemented. In Stage 1, datasets for pure aluminum, 0.5 wt% and 1.0 wt% graphene reinforced aluminium composites were used to train various ML models, categorized into five baskets: Decision tree-based, Support Vector models, Neural networks, Bayesian Models and Statistical Models. Model evaluation was conducted based on R²(R-squared), RMSE (Root Mean Squared Error), and Adjusted R² scores. In stage 2, the top models were further refined using advanced techniques, including Gradient-Based Methods and Ensemble Methods. Among the studied compositions, AlGr 2, containing 2 wt% graphene, exhibits the lowest turn-on electric field, whereas other compositions, including 1.25, 1.5, and 1.75 wt%, show comparatively higher values. This remarkable performance of AlGr2 arises from a delicate balance between conductive network formation, field enhancement and minimal agglomeration. The superior field emission performance of AlGr2 can be attributed to its optimal dispersion and percolation of graphene within the aluminium matrix. The findings demonstrate the efficacy of machine learning in accurately predicting field emission behavior, providing valuable insights for optimizing metal matrix composites in high-performance applications.