Solid Fat Content Determination by Ultrasonic Velocimetry

In this work an ultrasonic velocimetry technique was compared to direct pulsed nuclear magnetic resonance (pNMR) spectroscopy for the determination of the solid fat content (SFC) of anhydrous milk fat (AMF), cocoa butter (CB) and 70% (w/w) blends of AMF and CB with canola oil. In situ measurements of ultrasonic velocity were carried out during cooling (50–5 °C) and heating (5–50 °C) of the fat at 1 °C/min. SFCs were also determined in situ by pNMR, while polarized light microscopy (PLM) was used to image the crystallization process at the same time-temperature combinations as for the pNMR and ultrasonic velocimetry experiments. Peak melting temperatures determined by differential scanning calorimetry were used as an indicator of the polymorphic state of the different fats and fat blends. Our results suggest that ultrasonic velocity was highly dependent on the microstructure and polymorphism of the solid fat, and was higher for fats in more stable polymorphic forms. A high signal attenuation was observed in milkfat and cocoa butter at lower temperatures, particularly after 24 h of storage. SFC results obtained by pNMR did not have these problems and agreed quite closely with PLM results. Ultrasonic velocity was linearly correlated to SFC in AMF up to 40% solids during both crystallization and melting. The Miles equation could be used to model this relationship, however large deviations were observed above 20% solids. Ultrasonic velocity – SFC patterns for CB, on the other hand, were different for each CB dilution, thus making ultrasonic velocimetry not suitable for SFC determination in the CB systems. Ultrasonic velocity was lineary correlated with the box counting dimension (Db) in AMF during both crystallization and melting, while for CB, ultrasonic velocity was negatively correlated to Db upon melting only.