Ultrastructure of emulsions - a comparative electron microscopy study

Fish oil contains the two long chained omega-3 fatty acids eicosapentaenoic acid and docosahexaenoic acid. These are essential fatty acids and are not effectively converted from their precursor fatty acids in the human body. Hence it is advised to obtain these through the diet. Despite the recent debate about the effect of omega-3 supplements, the omega-3 fatty acids are subject to much interest from consumers, the food industry and several research fields.

One route of fish oil administration is through fish oil enriched food products. It is possible to add fish oil to food products initially without off flavors. But the highly unsaturated nature of the fatty acids renders them especially subjective to oxidation; a process which creates undesired off flavors and deteriorates the nutritional value.

It is therefore of interest to protect the fish oil against oxidation. One method suggested to do so is the emulsification of the fish oil to oil in water emulsions. A number of studies have been carried out with the purpose of determining the effect of this strategy, but whether emulsification of the oil is advantageous or not seems to be dependent on the food matrix to which the emulsion is added.

The Nanomega project, which is a cooperation between the National Food Institute, the Center for Electron Nanoscopy and the Department of Mechanical Engineering, all at the Technical University of Denmark, has dealt mainly with pure oil in water emulsions to describe the oxidation without the effect of an external food matrix.

Two PhD projects were carried out in the Nanomega project. One investigated the effects of different features such as emulsifiers, oil content, pH values, droplet sizes and production methods on the oxidative stability of emulsions. The other, which is the work presented in this thesis, was concerned with the structures of the interfaces between oil droplets and water in these emulsions.

The structure of the interfaces are interesting because it is generally acknowledged that the oxidation of the fish oil is initiated at the interface where the oil is in close vicinity to the water phase which might contain prooxidants such as trace metals and free radicals. It is thus speculated that the structure and composition of this interface can affect the oxidative stability of fish oil emulsions and perhaps provide protection against oxidation of the fish oil.

Thepurpose of this thesis work was thus to identify and apply suitable methods for the characterization of these interfaces. In the Nanomega project the interfaces were stabilized by proteins and phospholipids from milk. They form interface layers which are expected to be very thin, i.e. between 2-10 nm.

For this reason, electron microscopy was the method of choice. However, electron microscopy is inherently performed in vacuum and the emulsions presented a challenge by being liquid systems which are sensitiveto changes in water vapor and temperature. Furthermore, to achieve resolution and contrast of structural differences in specimens, electron microscopy relies on the scattering of electrons and the emulsions contain only light elements with low mass contrast.

The objective of this thesis was two-fold. One was to identify and further develop sample preparation methods to enable observation of the emulsions in the vacuum of the electron microscopes. Having developed the necessary experimental methods, the second objective was to characterize the structures of a number of emulsions that were found to be interesting for further analysis.

We have tested chemical fixations in different matrices and with different concentrations of fixative, employed high pressure freezing followed by either freeze-fracture cryo-SEM or freeze substitutions with different substitution media, and subjected some of the resulting samples to HAADF STEM tomography and EDS analysis.

We have analyzed the impact of glutaraldehyde on nano-emulsions by comparing shrinkage as a consequence of glutaraldehyde concentration. Furthermore we have compared the chemical fixation quality of larger oil droplets also as a consequence of glutaraldehyde concentration. We have characterized the structures of emulsions containing either 10% oil and oil droplets in the size range of ~130 nm or 70% oil emulsions with droplet sizes ~10-20 µm.

We have characterized the structures of different emulsifiers, sodium caseinate, whey protein isolate and emulsifiers containing both proteins and phospholipids from milk, on the surfaces of oil droplets. Furthermore we have characterized a number of different emulsions with respect to the location and structural arrangement of excess emulsifiers in the water phase where they are expected to act as antioxidants.

And finally we have observed that the structure of one of these emulsifiers, sodium caseinate, changes its structure as a consequence of whether the emulsion has been processed at high pressure or not. The results suggest a structural similarity with the native casein micelle and a similar response to high pressure conditions, but diverging slightly in the resulting structure.

This deviation can probably be ascribed to missing calcium phosphate in the emulsifier.