(173f) Coupling Experiments and Computational Tools to Estimate Thermal Properties for Product Design in the Fat-Based Food Industry

Product design has been evaluated as one of the challenges that chemical engineering has to face to help industry meet market demands for value-added products with higher performance [1]. In the food and pharmaceutical industry, the design of fat-based products is of great importance, since many foods have their desired properties related to a complex phase behavior involving lipids and many cosmetic and pharmacological products use vegetable oils as a lipidic medium in which active principles and fine chemicals are dispersed. Attempting to use Process Systems Engineering tools (namely thermodynamic modeling and computational/numerical approach), for product design, the present work coupled experimental studies and computational simulations to access the thermal properties of different scales of triacylglycerol mixtures, ranging from simple binary mixtures to ternary blends of vegetables oils.

In order to test the developed tools in raw materials used by industry, different blends formed by palm stearin (PS), fully hydrogenated soybean oil (FHSO) and canola oil were studied. The solid fat content at different temperatures, melting points and changes in heat capacity due to solid-liquid transitions are covered by the present work, both experimentally and computationally. Computational simulations and experimental results are compared and discussed. The following experiments were conducted: fatty acids composition identification, stereo distribution of fatty acids in the glycerol structure, softening point and thermal analysis (differential scanning calorimetry). The computational simulations were performed by the following steps: prediction of triacylglycerol composition of the blends by using statistical methods (random and non-random distribution of fatty acids in the glycerol structure); predictions of the distribution of molecules among solid and liquid phases by direct minimization of Gibbs free energy function, using a Generalized Reduced Gradient Method (CONOPT 3 GAMS solver). The optimization algorithm was then coupled with the main program written in FORTRAN 90, which handles the calculation of interaction parameters, melting temperature and melting enthalpy and the generation of triacylglycerols from fatty acids data [2]. The Solid-liquid equilibrium problem was then solved over a temperature range to be compared with experimental data in different temperatures and at different blend ratios (single vegetable oil, binary blends and ternary blends).

As the demand for more sophisticated fat products increases and the simple physical mixture (blending) is not sufficient for matching the desired properties, new technologies have emerged. The chemical interesterification reaction is one important technique that changes the triacylglycerol profile of the blends and, thus, modifies their physical properties. This reaction uses catalysts (e.g. sodium methoxide) and heat to promote a random distribution of fatty acids among the three positions of glycerol [3]. In order to predict how the reaction affects the final solid fat content of the blends, this work simulates the effect of the reaction by using random distribution of fatty acid in the glycerol chain and computing the melting profile of the reaction product.

REFERENCES

[1] Hill, M. (2009). "Chemical product engineering — the third paradigm." Comput. Chem. Eng.33, 947–953.

[2] Teles dos Santos, M.; Gerbaud, V.; le Roux, G.A.C. (2013). "Modeling and simulation of melting curves and chemical interesterification of binary blends of vegetable oils." Chem. Eng. Sci. 87, 14–22.

[3] Rodrigues-Ract, J.N.; Cotting, L.N.; Poltronieri, T.P.; da Silva, R.C.; Gioielli, L.A. (2010). "Crystallization behavior of structured lipids by chemical interesterification of milkfat and sunflower oil." Food Sci. Techn. 30, 258-267.