|
|
|
Models for the Development of Sortpion Isothems: A Review
|
|
Sunday Ibe Iji, Uwem Ekwere Inyang, Benjamin Reuben Etuk
|
|
American Journal of Food Science and Technology. 2025 13 (2). doi: 10.12691/ajfst-13-2-2
|
|
Keywords: Sorption Isotherm, Water activity, Moisture Content, Equilibrium Moisture content , Sorption Model
|
Context: Moisture sorption isotherms are fundamental for understanding the relationship between water activity and moisture content in food materials, directly impacting the design and optimization of drying, storage, and preservation processes. This review comprehensively evaluates sorption models, categorizing them into empirical, semi-empirical, theoretical, statistical physics-based, hybrid, and machine learning-based approaches. Empirical models, such as BET and GAB, remain widely used due to their simplicity and broad applicability across various food classes. However, these models often lack mechanistic insights, limiting their accuracy for complex or heterogeneous foods. Theoretical and statistical physics-based models provide deeper understanding of molecular sorption mechanisms but involve increased complexity and parameterization challenges. Recent advancements in hybrid and machine learning models demonstrate the potential to integrate physical laws with data-driven approaches, improving predictive performance and adaptability, particularly for novel and composite food matrices. Key challenges in model application include accurate model selection, robust parameter estimation, accounting for temperature dependency, and multi-component sorption phenomena. Moreover, data quality and limited experimental datasets often constrain model reliability. Emerging trends suggest that integrating physics-informed machine learning, real-time sensor data, multi-scale modeling, and digital twin technologies will enhance model robustness and facilitate real-time process control. Recommendations highlight the importance of selecting sorption models aligned with specific food types and processing requirements, employing rigorous fitting techniques, and combining mechanistic and data-driven approaches for optimal balance between accuracy and interpretability. Continued advancements in sorption isotherm modeling are expected to support sustainable, energy-efficient food processing and improve product stability and quality, addressing growing industry demands.
|