. Dehydration kinetics
A series of drying curves, exemplified by Fig. 1, exhibit the
change in the Moisture ratio (MR) of sample with time under
different drying conditions. As seen in Fig. 1, MR generally
decreased with time to an asymptotic value of 0. The equilibrium
time for drying was clearly subjected to the drying
method used: hot air drying was the slowest method, taking
up to 40min; microwave drying was faster, taking up to 25min
and combined drying was faster again, requiring up to just
15min reaching a constant moisture level (Table 1). The moisture
content of dehydrated samples was also dependent on
the drying method, ranging from 7.0% (d.b., for MW= 560W,
T = 70 ◦C) to 11.6% (d.b., for MW= 0W, T = 70 ◦C) (Table 1). Compared
with convective drying and microwave drying alone,
combined drying allowed for a more efficient drying process
and resultant lower final moisture content. Given the absence
of a constant rate period and the asymptotic shape of the
dehydration curves (Fig. 1), the semi-empirical drying models
described below (Eqs. (6)–(9)) were all fitted to the drying
data to discriminate which one was best suited to describing
the dehydration data. Following nonlinear regression of
each model on the rice data set, the pooled standard error
(SE) and Akaike information criterion were estimated (Table 2).
The Akaike information criterion (AIC) is defined as: AIC =
−2 × log lik + 2 × npar. Therefore, the AIC can be used to compare
models in their suitability in describing datasets, taking
into account numbers of model parameters (npar). The Page
model (Eq. (9)) resulted in the lowest SE and AIC, and was
therefore the best suited to describe the drying kinetics of
instant rice for the purpose of simulation and scale up of the
process.