2.1.2. Drying Equipment

A pilot-scale hot air convection dryer was used to dry lemongrass slices. The dryer consisted of two drying rooms, in which each room contained 10 wire mesh trays. The drying tray size was 80 × 80 × 3 cm and the mesh size was 1 × 1 cm (Figure 2). Moist air from the external environment at 30.5 ± 1.0 ◦C and 70% RH was fed to the dryer by exhaust fans where it was heated to a maximum temperature of 65 ◦C. The heated air then came into contact with materials, absorbing moisture and exiting the dryer.

**Figure 2.** Convection dryer (**A**) and tray (**B**) used in this study.

## 2.1.3. Drying Procedure

For each sample, a thin layer (1000 g) was put in the tray dryer. In this study, the air-drying temperature was set at 50, 55, 60 and 65 ◦C. The sample weight was recorded during drying at 10-min interval. Experimental data was used to evaluate mathematical models of drying curve as well as calculate the effective moisture diffusivity and activation energy.

A drying temperature range of 50–65 ◦C is generally feasible for a wide number of medicinal plants [14] and was selected in this study for a number of reasons. First, it is suggested that, to obtain the highest essential oil content, drying of lemongrass should be conducted in an oven at 45 ◦C for 7h[15]. In another study, it is suggested that temperature of drying should range from 50 to 70 ◦C to achieve maximum oil retention and that a temperature of below 30 ◦C was not advisable due to

the promotion of fungi [16].Second, citral content, which determines the quality of the lemongrass products, was shown to be insensitive to thermal treatment in the range of 50–65 ◦C. Another study attempting drying of lemon myrtle leaves suggested that the optimal drying temperature for maximum citral retention was 50 ◦C [17] and that slightly higher drying temperature was not detrimental to citral content. This is due to the protective effects of the partially dried surface layers formed by high temperature, limiting the mass transfer and the diffusion of volatile components to the surface. This result is supported by Rocha et al., (2000) who stated that both the oil yield and composition of citronella (*Cymbopogon winterianus*) were at optimal levels at a drying temperature of 60 ◦C [18].

#### *2.2. Determination of Mathematical Components*

#### 2.2.1. Moisture Content

Moisture content (g water/g dry matter) was calculated by the following equation [19].

$$\mathbf{M} = \frac{\mathbf{m}\_{\rm W}}{\mathbf{m}\_{\rm dm}},$$

where M is the moisture content (g water/g dry matter), m w is the mass of water in sample (g), and mdm is the mass of dry matter in sample (g)

## 2.2.2. Drying Rate

Drying rate (DR) is defined as the amount of evaporated moisture over time. The drying rate (g water/g dry matter/min) during the process of drying lemongrass was determined using the following equation:

$$\text{DR} = \frac{\text{M}\_{\text{t}} - \text{M}\_{\text{t}+\text{dt}}}{\text{dt}},\tag{2}$$

where Mt is the moisture content at t time (g water/g dry matter), Mt+dt is the moisture content at t + dt time (g water/g dry matter), and dt is drying time (min).

#### 2.2.3. Mathematical Modeling of Drying Curves

Moisture ratio is defined as follows [20].

$$\text{MR} = \frac{\text{M}\_{\text{t}} - \text{M}\_{\text{e}}}{\text{M}\_{\text{0}} - \text{M}\_{\text{e}}},\tag{3}$$

where Mt is the moisture content (g water/g dry matter). The subscripts t, 0, and e denotes time t, initial, and equilibrium, respectively.

To identify the suitable mathematical model for lemongrass drying, the experimental data were fitted to different thin-layer drying models (Table 1).

Evaluation of mathematical models was performed based on coefficient of determination (R2), root mean square error (RMSE) and chi-squared (*x*2). The criteria for a suitable model included high a R<sup>2</sup> value, and low *x*2 and RMSE values. The statistical values were defined as follows [21]:

$$\text{The coefficient of determination}: \text{ R}^2 = 1 - \frac{\sum\_{i=1}^{\text{N}} \left( MR\_{\text{exp},i} - MR\_{pre,i} \right)^2}{\sum\_{i=1}^{\text{N}} \left( \overline{MR\_{\text{exp},i}} - MR\_{pre,i} \right)^2} \tag{4}$$

$$\text{The Root Mean Square Error}: \text{ RMSE} = \sqrt{\frac{1}{N} \sum\_{i=1}^{N} \left( MR\_{\text{exp},i} - MR\_{pr;i} \right)^2} \tag{5}$$

$$\text{Chi}-\text{squared}:\ \chi^2 = \frac{\sum\_{i=1}^{\text{N}} \left( MR\_{\text{exp},i} - MR\_{pre,i} \right)^2}{\text{N}-Z} \tag{6}$$

where MRexp is the experimental dimensionless moisture ratio, MRpre is the predicted dimensionless moisture ratio, MRexp is the mean value of the experimental dimensionless moisture ratio, N is the number of observations, and Z is the number of constants in the mathematical model.

#### 2.2.4. Estimation of the Effective Moisture Diffusivity

To determine the moisture diffusion of a material, the Fick diffusion model was applied as follows [21].

$$\frac{\partial \mathbf{M}}{\partial \mathbf{t}} = \mathbf{D}\_{\mathrm{eff}} \frac{\partial^2 \mathbf{M}}{\partial \mathbf{x}^2} \,' \,\tag{7}$$

where M is the moisture content (g water/g dry matter), Deff is the effective moisture diffusivity (m2/s), and *x* is position with the dimensions of length (m).

Crank (1975) has provided a solution that deals with different shapes [22]:

$$\text{MR} = \frac{8}{\pi^2} \sum\_{n=0}^{\infty} \frac{1}{\left(2n+1\right)^2} \exp\left(-\left(2n+1\right)^2 \pi^2 \frac{\text{D}\_{eff}}{4\text{L}^2} \mathbf{t}\right),\tag{8}$$

where MR is the dimensionless moisture ratio, L is the half-thickness (m), n is the term in series expansion, and t is time (s).

When drying for a long time, the above equation can be reduced to:

$$\ln(\text{MR}) = \ln\left(\frac{8}{\pi^2}\right) - \left(\pi^2 \frac{\text{D}\_{\text{eff}}}{4\text{L}^2} \text{t}\right). \tag{9}$$

The effective moisture diffusivity could be described by empirical data using the graph of ln(MR) versus time (t) and the slope of straight line from the plot as − π<sup>2</sup> Deff 4L<sup>2</sup> t.

#### 2.2.5. Estimation of Activation Energy

The activation energy is calculated by the Arrhenius equation [20,23]:

$$\mathbf{D\_{eff}} = \mathbf{D\_0} \exp\left(-\frac{\mathbf{E\_a}}{\mathbf{RT}}\right),\tag{10}$$

where Ea is the activation energy (kJ/mol), R is the ideal gas constant (8.3143 kJ/mol), T is the absolute temperature (K), and D0 is the pre-exponential factor (m2/s).

## 2.2.6. Data Analysis

Microsoft Excel software was used to calculate moisture content, moisture ratio, and determine the effective moisture diffusivity and activation energy of the drying process. MATLAB 2014R software was used to find the best model fit and the mathematical model coefficients.
