Interference effects from coexisting fatty acids on elaidic acid separation by fractionating crystallization: A model study

A multi‐stage temperature‐programmed fractionating crystallization process was carried out to examine the effects of the presence of stearic acid (SA), oleic acid (OA), and linoleic acid (LA) on the separation of elaidic acid (EA). The results showed that the efficiency of fractionating crystallization of EA depended largely on the crystallization temperature, initial concentration of EA and presence of SA. The content of SA plays very important role for the fractionating performance. It was a characteristic observation that only when SA <2%, substantial crystallization of EA (>50% in stepwise crystal fractions) were obtained regardless of the initial concentration of SA. In general, SA induced crystallization of EA in earlier stage but delayed further crystallization of EA in later stage; the crystallization of EA was independent from co‐existing OA and LA. After reduction of EA content in solution to certain extent (7–10%, at ?20°C), further reduction of EA content requires much lower crystallization temperatures (trans‐fatty acids (TFA) from partially hydrogenated soybean oil (PHSO) is of high commercial interest. One of the strategies is to selectively release TFAs as free fatty acids from PHSO enzymatically. However, all commercially available enzymes are far away from qualified to selectively release TFAs, where there are always substantial non‐trans FAs hydrolyzed simultaneously. Therefore, developing post‐processing technology is requisite in order to recover those non‐trans fatty acids. Thus, this model system was designed based on FA composition characteristic of PHSO, which aimed to acquire some basic data and experience that lack in available literatures, so as to serve designing efficient practical process for removing trans‐fatty acid moieties from PHSO. The results from this work may be of general value to achieve a better understanding of fractionating crystallization behaviors of different FAs, relationship with individual molecular feature and property, and their interference effects, which might contribute to the design of practically feasible protocol to remove TFAs from PHSO and recover non‐trans FAs at the same time.