Optimization has played a fundamental role in the evolution of chemical engineering, enabling systematic decision-making under technical, economic, and environmental constraints. This review presents a structured and comparative analysis of the historical development and current state of optimization methodologies applied to chemical engineering, covering the transition from early linear and nonlinear programming approaches to advanced data-driven and artificial intelligence-based frameworks. A systematic literature review was conducted following the PRISMA guidelines, through which a total of 101 articles were retained for analysis. The results indicate that mixed-integer programming and decomposition-based methods remain widely adopted for structured industrial problems, while metaheuristic and hybrid data-driven approaches have experienced significant growth in recent years. In particular, a clear trend toward the integration of machine learning and surrogate modeling techniques is observed, driven by the need to address large-scale, non-convex, and highly nonlinear systems. The analysis reveals a clear methodological shift from classical linear optimization frameworks toward hybrid optimization strategies capable of addressing large-scale, non-convex, and highly nonlinear problems. Finally, current challenges and future research directions are identified, emphasizing the need for robust hybrid approaches that combine mathematical programming and intelligent algorithms to effectively manage complexity in next-generation chemical systems.