Food emulsions stabilised by blends of plant and dairy proteins

Many food products contain oil-in-water (O/W) emulsions, i.e., droplets of oil in a water phase. Due to the thermodynamic incompatibility between the two liquids, the droplets need to be covered by emulsifiers to ensure physical stability. The most widely used food emulsifiers are dairy proteins which are present in e.g., beverages, infant formula or dressings. However, their production has a large impact on the environment, therefore, plant proteins are currently considered as promising alternatives. Yet, the full replacement of dairy proteins by plant proteins brings along a number of technological challenges (e.g., functionality, product taste, pre-treatments). It can therefore be advantageous to rather consider partial replacement, but the properties of dairy-plant protein blends with regard to food systems’ stabilisation had been largely unexplored. In the present work, we aimed to achieve a rational compromise between technical functionality of such blends, and product quality. We investigated formation and stability of protein blend-stabilised emulsions and focussed on the relevant interfacial phenomena. To do so, we used a multiscale approach based on characterisations at different length- and time scales. 

In the first chapters (Chapter 2-4), we focussed on the long term interfacial phenomena when using protein blends to formulate emulsions. We assessed the related properties of blends of pea protein isolate (PPI) with either whey protein isolate (WPI) or with sodium caseinate (SC) in Chapter 2. We showed a synergistic behaviour in terms of physical stability of the emulsions, when the blends were used. The blend-stabilised emulsions had higher surface loads compared to the individual protein-stabilised emulsions, which showed that more proteins were needed to stabilise the interface. Furthermore, compositional rearrangements at the interface were observed over days. More specifically, after emulsion formation, whey proteins were able to partly displace pea proteins from the interface, which were themselves able to displace caseins. Such considerations are usually not considered in food emulsion formulation, even though they are very relevant, as the interfacial layer protects emulsions droplets against physical destabilisation.

As a next step, we studied the interfacial behaviour of the same proteins and their blends using model air-water and oil-water interfaces in Chapter 3. We showed that the rheological response of the blend-stabilised interfaces deviated from what could be expected from averaging those of the individual proteins. The layer’s strength decreased when WPI was blended with PPI. SC formed the weakest interfacial layers, but blending it with PPI improved the mechanical strength of the layer at both the air-water and oil-water interface. In general, higher elastic moduli and more rigid interfacial layers were formed at the air-water interface compared to the oil-water interface, except for pure PPI.

Even though the targeted applications of blends of proteins of different biological origins are still lacking, many traditional or emergent emulsion products contain mixtures of proteins (for example, caseins and whey proteins naturally co-exist in dairy products), resulting in complex, non-equilibrated interfacial structures (as discussed in Chapter 2). Therefore, in Chapter 4, we aimed to further evaluate and understand the interfacial rearrangements over time in plant-dairy protein blend-stabilised emulsions. We notably found that the whey proteins were able to displace pre-adsorbed pea proteins. Using model interfaces, we were able to determine that protein-protein interactions at the interface were the driving force for such a displacement, rather than a decrease in interfacial tension. These outcomes could be instrumental in defining new strategies for plant-animal protein “hybrid” products, but also for any protein blend-stabilised emulsion.

In Chapter 5 and 6, we focussed on the interfacial film formation at very short time scales (<1 s), which is relevant when considering the time scales involved in conventional emulsification processes. In Chapter 5, we used tailor-made microfluidic chips to probe the rheological properties of the interfacial films of freshly prepared droplets, and the stability of the droplets against coalescence under flow. We showed that these microfluidic tools are useful to assess the rheological properties of protein-stabilised droplet interfaces within 1 second after droplet formation. For instance, we determined that the PPI-stabilised interfaces had weaker in-plane interactions compared to WPI-stabilised ones, which resulted in less stable droplets. Although the blend-stabilised interfaces showed high connectivity between the adsorbed proteins, this could not prevent droplet coalescence probably due to structural heterogeneity of the droplet surface.

In Chapter 6, we proceeded with using PPI to stabilise emulsion droplets formed in a microfluidic device, and recorded coalescence stability after droplet formation for proteins exposed to metal-catalysed oxidation. Protein oxidation led to a strong loss in protein solubility, which damages the overall ingredient functionality. Yet, in the fraction that remained soluble, it led to the formation of low molecular weight fragments that were able to form more homogenous interfaces compared to the non-oxidised proteins (as confirmed by the Langmuir-Blodgett films), and increased coalescence stability. Interfacial films that were structurally more heterogeneous were therefore more prone to rupture, as was hypothesised in Chapter 5. Furthermore, it became clear that the emulsifying properties of pea proteins are strongly dependent on their chemical status, and on the associated structural properties at the molecular and supramolecular levels.

In Chapter 7, we zoomed in further and used fluorescence spectroscopy to gain insight in the tertiary structure of the proteins in solution and at the interface during a secondment at INRAE (Nantes, France). It became evident that the ‘soluble’ fraction of the commercial pea proteins was present as ‘soluble aggregates’, which thus behave as small particles. This is probably a direct result of the harsh processing applied to obtain those ingredients. In contrast, the isolates purified in-house by a mild procedure contained proteins that were less aggregated, and able to rearrange at the oil-water interface.

Plant protein ingredients generally contain a substantial insoluble protein fraction, of which the properties necessarily differ from those of the soluble fraction (Chapter 8). We showed that when the full fraction was used to formulate emulsions, the soluble and insoluble parts competed for interfacial localisation, resulting in physically unstable emulsions. In contrast, when used solely, the insoluble fraction could physically stabilize the emulsions, and in case of a high internal phase emulsion, even led to around 10 times higher viscosity than their whey protein-based counterparts. These results confirm that the constituents of commercial pea protein isolates behave very differently, which should be used as a starting point for designing stable plant protein-based emulsion.

Finally, in Chapter 9, we extensively reviewed the established and developing methods to measure interfacial displacement, which, in our view, is paramount for the production of physically stable food emulsions. It is crucial to take this complex phenomenon into account when characterising the emulsifying and interfacial properties of protein blends, and also evaluate the various fractions, as done within this thesis for pea proteins. We expect that both soluble plant proteins and insoluble fractions can be used to stabilise food emulsions, albeit through different mechanisms: either via the classical mechanisms involving amphiphilic polymers, or via a Pickering stabilisation mechanism, respectively. To be able to probe protein functionality in conditions relevant to emulsion processing and subsequent storage, we recommend the development and use of tailor-designed microfluidic tools to characterise in-depth the mechanical properties of interfacial films under flow over a broad range of time scales.