APPLICATIONS OF MEMBRANE-BASED OPERATIONS IN SEPARATION OF NATURAL PHENOLIC COMPOUNDS

Abstract

With the growing global population, there is an increasing need to develop sustainable and green separation processes for high-value bioactive compounds. Naturally occurring phenolic compounds such as betanin and rutin have gained considerable importance in the food and pharmaceutical industries; however, existing separation methods often rely on solvent-intensive, energy-demanding, or non-scalable techniques that compromise product purity and biological activity. Although membrane-based processes have been explored for polyphenol recovery, there remains a lack of systematic, model-integrated, and scale-up-oriented studies that bridge laboratory experimentation with industrial feasibility, particularly for real feed streams. In this context, the present thesis makes a distinct contribution by establishing nanofiltration (NF) as a quantitatively optimised and industrially translatable platform for the selective concentration of two structurally different phenolic compounds, namely betanin and rutin. A self-assembled NF setup using HFT-NF 150 membranes is employed to generate a comprehensive experimental dataset, wherein the individual and interactive effects of pressure, feed concentration, and feed flow rate on permeate flux and solute rejection are rigorously evaluated. Unlike prior studies that report only empirical trends, this work integrates a three-parameter Spiegler-Kedem transport model to extract membrane reflection coefficients, solute permeability, and hydraulic permeability, thereby providing mechanistic insight into solute-membrane interactions and enabling predictive validation of experimental performance. A further aspect of this research is the application of variance-based global sensitivity analysis to membrane separations, allowing the quantitative ranking of operating and transport parameters by their influence on flux and rejection for phenolic solutes. This approach moves beyond conventional one-factor-at-a-time analyses and establishes a robust framework for rational process optimisation. Crucially, the thesis extends beyond model solute systems to address the separation of betanin from its natural matrix, beetroot juice, an area scarcely examined in existing literature. Comprehensive fouling studies using multiple theoretical fouling models are performed to elucidate viii dominant fouling mechanisms, quantify flux decline behaviour, and assess long-term operational stability under realistic feed conditions. These results provide actionable design insights to mitigate fouling and extend membrane longevity. Finally, scale-up simulations grounded in experimentally derived transport and sensitivity parameters demonstrate the technical feasibility of translating laboratory findings into an industrial-scale NF unit. Collectively, this thesis offers a unified experimental-modelling-scaling framework for phenolic compound recovery, delivering original contributions in mechanistic understanding, optimisation strategy, real-feed validation, and process design that advance the state of the art in sustainable membrane-based separations.