Thin‐layer drying of parchment Arabica coffee by controlling temperature and relative humidity

Abstract This paper presents thin‐layer drying of parchment coffee (Coffea arabica). Thin‐layer drying of parchment coffee was conducted under controlled temperatures (50°C, 60°C, and 70°C) and relative humidities (10%–30%). The temperature of the drying air was important for drying at a high temperature, which results in the rapid removal of moisture and reduced time for drying. Nine thin‐layer drying models (Newton, Page, Henderson and Pabis, logarithmic, two‐term, modified Henderson and Pabis, two‐term exponential, approximation diffusion, and modified‐Midilli) were fitted to the experimental data for parchment coffee. The drying parameters of parchment coffee were related to temperature and relative humidity. The best model was the modified‐Midilli model, which can be used to design the optimal dryer. The effective moisture diffusivity of parchment coffee drying was determined by minimizing the sum of squares of the deviations between the experimental data for the moisture content and the predicted values of thin‐layer drying. The effective moisture diffusivity as a function of the temperature at each relative humidity was expressed by the Arrhenius‐type equation.

Many studies have investigated the coffee drying process using several drying methods, such as solar drying (Deeto, Thepa, Monyakul, & Songprakorp, 2018), convective drying (Burmester & Eggers, 2010;Muhidong Mursalim, & Rahman, 2013;Nilnont et al., 2012;Siqueire et al., 2017), and hot air-assisted microwave drying (Ghosh, & Venkatachalapathy, 2014). However, the purposes of this study are as follows: (a) to study the effect of temperature and relative humidity on the drying characteristics of parchment coffee, (b) to develop an appropriate thin-layer drying model of parchment coffees, and (c) to evaluate the effective moisture diffusivity of parchment coffee.

| Experimental study
The parchment Arabica coffee was cleaned from the mucilage and stored at a temperature of 5°C. The parchment coffee stayed at room temperature for 16 hr before starting the experiment to achieve equilibrium conditions. Thin-layer drying of the parchment coffees was conducted in a laboratory dryer under controlled temperature and relative humidity conditions. The schematic diagram of the laboratory dryer in Figure 1 was designed and developed by Guarte (1996). This laboratory dryer contains two main sections: a humidifier section and a drying section. The humidifier section consists of a ceramic-packed bed, a water heater, a water pump, and a humidity control unit. The drying section contains a drying chamber, an air heater, and an air blower. The blower forces air to pass through a humid, ceramic-packed bed, a porous media, which absorbs and transfers moisture to the air. Then, the air becomes saturated at the same temperature as the heater controls the water. Then, the saturated air temperature is controlled to reduce the relative humidity by the air heater, and the air passes parallel to the parchment coffee. The relative humidity and temperature of the drying process are manually controlled by a psychometric chart with adjustable power supplied by the air heater and water heater.
The dryer was operated before starting an experiment for 1 hr to achieve steady-state drying conditions. Every experiment was fixed at an air velocity of 1 m/s. The parchment coffee (approximately 100 g) was placed on a tray in a thin layer inside the laboratory dryer.
Thin-layer drying of the parchment coffee was conducted at a temperature and relative humidity in the drying range of 50-70°C and 10%-30%, respectively. Drying air temperature was measured every 5 min with a thermocouple (K type) and was recorded with a data logger. The mass of the parchment coffee was recorded by an electronic balance (accuracy ± .01 g) at an interval of 1 hr. The electronic balance was placed near the drying chamber in the laboratory. The drying process continued until the change in the sample mass was the lowest. All experimental drying conditions were conducted in three replicates.

| Mathematical modeling
Models of thin-layer drying that describe the drying characteris- The moisture content dry basis was defined as: where s the mass of dried product samples at instant t and d is the mass of the oven-dried product samples at 105°C (Helrich, 1990).
For all experimental drying, the final moisture content of the parchment coffee achieved equilibrium under constant conditions of temperature and relative humidity. The nine thin-layer drying models in Table 1 were selected as suitable models for explaining the drying process of parchment coffee.
The mathematical models were adjusted by nonlinear least square regression analysis. In choosing the model, specific values were considered: the coefficient of determination (r 2 ) and the root mean square error (RMSE). For the best fit, r 2 must be the highest, and RMSE must be the lowest. These parameters were defined as follows: where M obs,i : the observed moisture ratios; M pre,i: the predicted moisture ratios; and n: the number of observations.

| Modeling of parchment coffee
The moisture content based on the data obtained in the experiments was converted to the moisture ratio (MR) and then fitted to the nine thin-layer drying models in Table 1 to identify a suitable equation for describing and predicting coffee drying behavior during the drying process that controlled the temperature and relative humidity.
The estimated parameter values and statistical analysis values (r 2 and RMSE) are also shown in Table 2. The statistical analysis results showed that of the models, the modified-Midilli model had the  Table 2. The value of the drying constant "k" increased with increasing drying temperature. The constant values "k," "n," and "b" of the modified-Midilli model were regressed by nonlinear least square regression analysis with respect to the temperature and relative humidity of the drying air. These values can be calculated through the following expression in the form of a second-order polynomial, as shown in Table 3.

| Effective moisture diffusivity
According to the drying experiment results, the mechanism of diffusion could be analyzed through Fick's second law for cylindrical shapes. The effective moisture diffusivities of parchment coffee for all conditions are presented in Table 4, with values ranging between 7.7554 × 10 -10 and 1.4525 × 10 -9 m 2 /s. These values are within the range of diffusivities (10 −11 m 2 /s-10 −9 m 2 /s) for different crops using several drying methods (McMinn & Magee, 1999), such as the moisture diffusivity of coffee ranging from 7.17 × 10 −10 to 10.00 × 10 −10 m 2 /s for drying air temperatures ranging from 40 to 60°C (Nilnont et al., 2012;Varadharaju, Karunanidhi, & Kailappan, 2001). For convective drying, a high drying air temperature transfers heat to the product, resulting in a higher product temperature, and then, the water within the product moves by diffusion and evaporates into the drying air. Drying at a low relative humidity decreases the vapor pressure causing the water inside the product to evaporate rapidly in TA B L E 4 The effective moisture diffusivity of parchment coffee large quantities. This mass transfer process will stop when the water vapor pressure at the product surface becomes equal to the water vapor pressure of the drying air. Therefore, the drying air temperature and relative humidity have an essential effect on the drying of parchment coffee: the effective moisture diffusivity of parchment coffee increased with a decrease in the relative humidity and an increase in the temperature of drying air. For RH = 30% where T ab is the absolute drying air temperature.

| CON CLUS IONS
The results of the coffee drying experiment under different conditions showed that relative humidity affected reductions in moisture (9) D parchment coffee = 8.0 × 10 −7 e (−2179.9∕T ab ) , r 2 = .997 (10) D parchment coffee = 5.0 × 10 −9 e (−542.28∕T ab ) , r 2 = .9647 (11) D parchment coffee = 4.0 × 10 −9 e (−546.25∕T ab ) , r 2 = .9841 F I G U R E 6 Effective moisture diffusivity of parchment coffee as a function of the reciprocal of absolute drying air temperature (Tab) at different relative humidities (10%-30%) content and decreased at each of the drying air temperatures. The temperature of saturated water within the product was dependent on the vapor pressure of the drying air because vapor pressure affects the boiling point of water, which results in the water inside a product evaporating rapidly in large quantities. Thus, the mass transfer caused by the evaporation of water in the product to the drying air was greater under a relative humidity of 10% and a temperature of 70°C than under other conditions. The maximum effective moisture diffusivity of a drying temperature of 70°C and relative humidity of 10% was 1.4525 × 10 -9 m 2 /s, which was determined by minimizing the sum of squares. Nine thinlayer drying models were used to describe the process during parchment coffee drying. Drying parameters were found to be a function of drying air temperature and relative humidity. The agreement between the predicted and experimental data for parchment coffee in the modified-Midilli model was excellent for considering the drying behavior of parchment coffee, and this model was used optimize the dryer.

ACK N OWLED G M ENTS
The authors would like to thank Division of Energy Technology, School of Energy, Environment and Materials, King Mongkut'sUniversity of Technology Thonburi and Department of Mechanical Engineering, Faculty of Engineering and Architecture, Rajamangala University of Technology Suvarnabhumi, Nonthaburi Campus for supporting experimental apparatus.

CO N FLI C T O F I NTE R E S T
None declared.

E TH I C A L S TATEM ENT
This study does not involve any human nor animal testing.

I N FO R M E D CO N S E NT
None.