Prediction of water and soluble solids concentration during osmotic dehydration of mango

• Published in: Food and bioproducts processing 2008-03, Vol.86 (1), p.7-13
• Main Authors: Khan, M.A.M., Ahrné, L., Oliveira, J.C., Oliveira, F.A.R.
• Format: Article
• Language: English
• Subjects: Arrhenius model; Biological and medical sciences; dried fruit; drying; Food Engineering; Food industries; Fruit and vegetable industries; Fundamental and applied biological sciences. Psychology; Livsmedelsteknik; mangoes; Mass transfer; Mathematical modelling; mathematical models; model validation; osmotic dehydration; osmotic treatment; Process optimisation; reaction kinetics; shrinkage; soluble solids; sucrose; temperature; temporal variation; water content
• ISSN: 0960-3085; 1744-3571; 1744-3571
• DOI: 10.1016/j.fbp.2007.10.012

The objective of this work was to develop a mathematical model to predict the kinetics of the change in water and soluble solids fractions in mango (cv. Haden) osmotically dehydrated in a sucrose solution. A full factorial design at three levels was used, varying temperature ( T) and concentration of soluble solids in the osmotic solution (SSC). The models based on the Weibull distribution were built up in two steps: (i) primary models to determine the kinetic parameters at constant T and SSC, (ii) secondary models to further include the influence of T and SSC on the parameters of the primary model. The Weibull model can successfully describe both water and sugar fractions during osmotic dehydration ( R 2 = 0.98 and 0.96, respectively for water and sugar models). The time constant ( τ) for both models followed an Arrhenius-type relationship with temperature, with the reference time constant ( τ ref) at the average T and increasing linearly with SSC. The shape factor ( β) was constant. The prediction accuracy of the models to predict water and sugar fraction was tested by cross validation and using a third set of experimental data, showing very good results with shrinkage values below 4.6% and errors on predictions lower than 1.6%.