The effect of low melting oils on the crystallisation of confectionery fats

This thesis concentrates on gaining a fundamental understanding of the crystallisation, phase behaviour, and melting in relatively simple fat/oil blends. This is the first reported study of hot stage microscopy (HSM) experiments on tripalmitin (PPP)/triolein (OOO), 1,3-dipalmitoyl-2-oleoylglycerol (POP)/OOO, and cocoa butter (CB)/hazelnut oil (HZ) systems. The HSM technique allows the visualisation of the initial crystallisation, polymorphic transformations, and melting of fat crystals; melting points can also help identify polymorphic form. Supporting experiments were also performed using differential scanning calorimetry (DSC), nuclear magnetic resonance (NMR), and X-ray diffraction (XRD). In PPP-OOO samples, HSM visualised for the first time a melt-mediated transformation from β′ to β across a small liquid gap between the untransformed (β′) and transformed (β) material. This behaviour was not seen with pure PPP. Melting points obtained by HSM for the PPP/OOO system were above those predicted by the Hildebrand equation, but this is attributed to the non-equilibration of concentration gradients within the system. This was evidenced by the fact that a rapid cooling rate (to produce a finer microstructure with smaller crystals, and hence reduce diffusion distances) combined with a slower remelt rate enabled samples to melt close to ideal. Indeed, final melting points obtained via HSM were consistently higher than DSC results across all systems; convection and the more three-dimensional system in DSC (as compared with the two-dimensional HSM system) may have aided melting. The POP/OOO system displayed complex remelting behaviour, especially when warmed at a relatively slow rate; this also resulted in a higher production of β. Liquid oil content was shown to not only be important in aiding transformation of lower forms to β, but also reduced the number of polymorphs observed upon remelting, as compared with pure fat samples. Liquid oil content was also shown to be crucial for transformation to β in the CB/HZ experiments; very few β crystals were seen in pure CB samples. Experiments carried out on DSC for both CB and CB-HZ cooling at 1°C/min or faster produced both α and β′ crystals, but for CB-HZ this also then led to some transformation from α to β. The β polymorph was not observed when only β′ was formed at slower cooling rates. A curious result was that the effect was stronger the longer the samples were held at 0°C before rewarming, with the α becoming more resilient against transforming to βʹ and instead transforming directly to β. More extensive transformation to β occurred if the sample was held for 30 min at 18°C or 22°C during the rewarming step. As well as showing differences in melting temperatures, the HSM and DSC results also did not always match with respect to polymorphic form. Small quantities of β crystals were seen in HSM samples that were not always seen in thermograms of equivalent DSC samples. This either highlights the limitations of DSC or suggests that polymorphic behaviour in the more fluid DSC system differs to that in HSM, or both. Growth rate analysis of PPP (in OOO) showed that both reduced supersaturation and supercooling can be correlated with the growth rates of β′ and β. Growth rates of β that occurred via the melting of β′ were well correlated with driving forces that took into account that the concentration of PPP in the liquid gap between β′ and β (from which the β crystal was growing) was limited by the solubility of β′ at the sample temperature. Whilst temperature is often seen as a key driver in governing polymorphic transformation, part of the temperature effect may be an indirect effect via the extra amount of liquid content at higher temperatures. The ability of oil to aid transformation to β may be relevant to food systems where this higher polymorphic form is the preferred type, such as chocolate. Potential applications could include producing novel fat blends with relatively stable fat network structures at a lower overall saturate level, or developing blends with bloom inhibiting properties.