From these plots, the model constants W0, Wc, n and k were determined. The initial moisture content (W0) of the filmogenic solutions ranged from 85 to 92 kg kg−1 d.b., which favors long periods with a constant drying rate. The drying rate

in the constant rate period is fully governed by the rate of external heat and mass transfer, since a film of free water is always available at the evaporating surface ( Cui, Xu, & Sun, 2004). To verify the effect of independent variables (yam starch and glycerol concentrations and temperature) on the parameters obtained by the model and the values of Def, regression analysis was applied using the response surface method ( Table 3). Initial moisture content of the filmogenic solutions was influenced FK228 order only by the amount of yam starch, showing that

the relationship between these variables was linear. The parameter “n” represents the drying rate Pexidartinib supplier during the constant period, where the model that best fit this variable was the linear model with interaction, in which the interaction of yam starch and temperature was significant. As expected, the slope of the drying curves increases as the drying temperature increases, i.e., the drying rate (n) is higher, since at higher temperatures there is a greater amount of heat transferred from the air to the material and, consequently, an increase in migration velocity of water from the interior to the surface of the product. The same occurred with dehydration of tomato fruits, where greater temperatures developed shorter drying time ( Sanjinez-Argandoña, Branco, Bittencourt, & Munhoz, 2011). A quadratic model was fitted to critical moisture (Wc) in which yam starch concentration had a linear and quadratic influence, and temperature tuclazepam only a linear influence. Finally, the diffusion coefficient (Def) calculated from the drying parameter “k”, which represents the period with decreasing drying rate, was adjusted to the linear model with interaction, where there was significant interaction between yam starch content and temperature. Fig. 2 was constructed to better visualize these effects. The regression models were significant at 5% (P ≤ 0.05) and expressed in the form

of equations. Equations (5), (6), (7) and (8) represent the models for initial moisture content (W0), the parameter n, critical moisture content (Wc) and diffusivity coefficient (Def). equation(5) W0=96.05−1.10F;(R2=99.86%) equation(6) n=−41.89+0.65T+0.01FT;(R−aj=80.40%) equation(7) Wc=0.11+4.38F−0.43F2+0.42T;(R−aj=89.23%) equation(8) Def=−5.97+0.50T−0.01FT;(R−aj=83.45%)Where F and T is the influence of starch content and temperature, g 100 g−1 and °C; FT is the influence of the interaction between starch and temperature, g °C 100 g−1; R2 is the determination coefficient for linear model; and R-aj is coefficient of determination adjusted for other models. There was no significant interaction of glycerol with any drying parameters.