Milk powders have been widely used for decades across multiple categories and applications, from infant formula to milk chocolate. In recent years, the demand for plant-based milk alternatives has strongly increased, and with it the interest in plant-based milk powders. In milk powder substitutes, milk fat, protein and carbohydrates are replaced by plant-based counterparts to form a milk-like emulsion with consecutive drying, aiming to achieve similar nutritional and organoleptic properties compared to dairy milk. In order to achieve the long shelf-life required for milk powder substitutes, present lipids and proteins need to be protected in the powder matrix against oxidation by environmental oxygen. The present work aims to understand and leverage the complex factors impacting oxidation stability in order to enable targeted product and process design for extended powder shelf-life.

In this work, the contribution and interrelations of three critical impact factors on the shelf-life stability of dried emulsions are investigated (see Fig. 1): Particle structure, matrix permeability and properties of the oil-matrix interface. To quantify the relevant impact factors and their synergistic effects for shelf-life determination, dried model emulsions with different properties are produced:

• Particle structure (size and porosity) will be varied through process parameters during drying of model emulsions, drying technology or post drying processing, such as fluid bed agglomeration.
• Matrix permeability to oxygen will be altered through the molecular weight of the model emulsions’ hydrophilic, glassy matrix, using maltodextrin with different dextrose equivalents.
• Oil-matrix interface will be tuned by leveraging and modifying material properties of the emulsifying system in order to alter the interfacial layer that poses an important barrier for lipid oxidation.

For good comparison of the different factors on oxidation stability, dried model emulsions aim to achieve a constant, high encapsulation efficiency (EE) in order to minimize free fat on the particle surface.

Based on experimental results, a predictive model will be developed, estimating oxidation of a lipid marker based on oxygen diffusion through the particle structure, taking into account the matrix’s oxygen permeability and properties of the lipid-matrix interface.