Kinetic modeling of multi‐component crystallization of industrial‐grade oils and fats

2015 
Transient crystallization kinetics is investigated for complex, industrial-grade vegetable oils consisting of more than ten triacylglycerols (TAG). The classical nucleation model has been used to describe primary nucleation, while secondary nucleation has been described by a semi-empirical approach. Growth is modeled using a modified Burton-Cabrera-Frank (BCF) model. Surface tensions and growth constants have been determined using focused-beam-reflectance measurements (FBRM). The required adjustable parameters in the model have been fitted to overall crystallization curves obtained by solid-fat content (SFC) measurements for a given oil at different cooling rates and degrees of dilution. The developed model can accommodate more polymorphs simultaneously and performs well with respect to predicting crystallization onset, rate of crystallization and final SFC value. It can also qualitatively describe how higher cooling rates lead to formation of more meta-stable crystals and smaller mean-crystal sizes. Practical applications: The model provides a good starting point for developing more realistic, transient models for TAG crystallization with the ability to accommodate processing conditions and complex chemical compositions. Such a predictive model may provide a powerful tool to screen and optimize oil formulations in industrial processes and allow product developers to evaluate recrystallization events. An advanced transient, mathematical model describing the course of crystallization in natural, multi-component, multi-phase vegetable oils and fats has been developed. The model is based on well described nucleation and growth kinetics and accommodates the presence of more polymorphic crystal forms. The result is an industry-relevant model aiding understanding and predictability of oil and fat crystallization with respect to processing conditions and chemical compositions.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    49
    References
    5
    Citations
    NaN
    KQI
    []