Drying Characteristics and Quality Attributes Affected by a Fluidized-Bed Drying Assisted with Swirling Compressed-Air for Preparing Instant Red Jasmine Rice

A new process for the production of instant red jasmine rice was investigated using fluidized bed drying with the aid of swirling compressed air. Drying characteristics were evaluated using the operating parameters of fluidizing air temperature (90–120 °C) and pressure of swirling compressed air (4–6 bar). Appropriate air pressure was determined based on the highest value of model parameters from the semi-empirical Page equation and effective diffusivity. Influences of supply time of swirling compressed air (2–10 min) and drying temperature of 90–120 °C were investigated and optimized based on the quality attributes using response surface methodology. Drying at 120 °C and compressed air pressure of 6 bar gave the highest rate constant and effective diffusion coefficient. Drying at 120 °C combined with injecting swirling air for 2 min was the most suitable approach, while drying at 90 °C and supplying compressed air for 10 min was the best choice to preserve antioxidant properties. Air temperature of 98.5 °C with 2 min supply of swirling compressed air suitably provided high physical and rehydration properties and retained high health benefits of antioxidant compounds. Finally, after rehydration in warm water at 70 °C for 10 min, the textural properties of the rehydrated rice sample were comparable to conventionally cooked rice.


Introduction
Pigmented rice is now increasingly consumed as a source of daily calories and bioactive compounds, such as polyphenols, that reduce or prevent cell damage caused by free radicals. After harvesting, its high moisture content (18-22%wb) is normally reduced to approximately 14%wb by means of drying for safe storage [1][2][3]. Among pigmented rice varieties, red jasmine rice in Northeastern Thailand (also called 'Mali dang') possesses the highest total phenolic content [4]. Red jasmine rice is consumed for its health benefits and impressive aroma. However, the long cooking time and gummy texture of the cooked rice are not preferred by consumers [5]. Therefore, instant red jasmine rice would be a popular alternative choice as a convenient product to match modern consumer lifestyles [6,7].
Drying plays the most important role in producing instant rice, as this affects rehydration and organoleptic properties [6][7][8][9]. Copious research has investigated quality improvements of instant rice products [5,7,8]. Freeze drying, known as an ideal drying process, was used to produce porous structural instant rice. Due to its high porosity with remaining high nutrients, rehydrated freeze-dried rice were comparable with freshly cooked rice [9][10][11]. However, alternative drying techniques needed to be explored due to the high operation cost and long drying time of the freeze-drying method [11]. The conventional convective hot-air drying method was employed as it was simple and cheap operation Red jasmine rice was purchased from the community enterprise Gudkhaedon, Loeng Nok Tha, Yasothon Province, Thailand. The rice was sorted and cooked at a rice-to-water ratio of 1:3 by a domestic electric rice cooker and used as the control sample. To retain their original shape after cooking, the rice grains were cooked for 30 min and subsequently kept in the cooker for 10 min to ensure complete gelatinization. Moisture content was measured based on the standard AOAC method [20] and 200 g of cooked rice was subjected to the drying process. Textural and antioxidant properties of the control were also examined.

A Process of Fluidized Bed Drying Assisted with Swirling Compressed Air Stream
As shown in Figure 1, a lab-scale fluidized bed dryer assisted with swirling compressed air consisted of three main compartments as the air heating system, cylindrical drying chamber and swirling compressed air system. Ambient air was drawn by a 3-phase 1-hp blower (Mitsubishi Electric Automation, Co., Ltd., BangKok, Thailand), with velocity controlled by an inverter (Model H-3200 Series, Haitec Transmission Equipment Co., Ltd., Guangzhou, China). The air was drawn through a heating box equipped with ten 1-kW finned heaters. Fluidizing air temperature was controlled using a PID controller (Model MAC-3D, Shimax Co., Ltd., Akita, Japan). The hot fluidizing air was blown through a perforated plate from the bottom of a stainless-steel tube with diameter of 0.1 m and height of 1 m (see Figure 2).
The important part of the proposed process was the swirling compressed air system (shown in Figure 3), comprising the air stream controller and air compressor. This system successfully restricted the agglomeration of cooked rice kernels during the drying process using swirling air streams released from two air tubes inserted at the bed bottom with outlet size of 1 mm. The air was compressed by a 2-hp air compressor (Model PP-22, Puma Industrial Co., Ltd., Taiwan) equipped with a pressure regulator. The compressed air was supplied intermittently with time controlled by a PLC controller (Model Siemens Logo 6ED1052-1FB08-0BA0 Logic Module, Siemens, Munich, Germany).
Processes 2021, 9, x FOR PEER REVIEW 3 of 22 was controlled using a PID controller (Model MAC-3D, Shimax Co., Ltd., Japan). The hot fluidizing air was blown through a perforated plate from the bottom of a stainless-steel tube with diameter of 0.1 m and height of 1 m (see Figure 2). The important part of the proposed process was the swirling compressed air system (shown in Figure 3), comprising the air stream controller and air compressor. This system successfully restricted the agglomeration of cooked rice kernels during the drying process using swirling air streams released from two air tubes inserted at the bed bottom with outlet size of 1 mm. The air was compressed by a 2-hp air compressor (Model PP-22, Puma Industrial Co., Ltd., Taiwan) equipped with a pressure regulator. The compressed air was supplied intermittently with time controlled by a PLC controller (Model Siemens Logo 6ED1052-1FB08-0BA0 Logic Module, Siemens, Germany).   Processes 2021, 9, x FOR PEER REVIEW 3 of 22 was controlled using a PID controller (Model MAC-3D, Shimax Co., Ltd., Japan). The hot fluidizing air was blown through a perforated plate from the bottom of a stainless-steel tube with diameter of 0.1 m and height of 1 m (see Figure 2). The important part of the proposed process was the swirling compressed air system (shown in Figure 3), comprising the air stream controller and air compressor. This system successfully restricted the agglomeration of cooked rice kernels during the drying process using swirling air streams released from two air tubes inserted at the bed bottom with outlet size of 1 mm. The air was compressed by a 2-hp air compressor (Model PP-22, Puma Industrial Co., Ltd., Taiwan) equipped with a pressure regulator. The compressed air was supplied intermittently with time controlled by a PLC controller (Model Siemens Logo 6ED1052-1FB08-0BA0 Logic Module, Siemens, Germany).   Before drying, fluidizing air temperature was set and the apparatus was operated for 30 min to ensure that the condition was in a steady state. Two hundred grams of cooked red jasmine rice, corresponding to a static bed height of approximately 4 cm were loaded into a cylindrical drying chamber. All experiments were assessed under full factorial design with two factors of drying temperature (90, 105, and 120 • C) and compressed air pressure (4, 5, and 6 bar) to clarify drying kinetics. The optimal compressed air pressure was subsequently used with variations of drying temperature (90-120 • C) and swirling compressed air supply times of 2, 4, and 6 min to investigate the influence on instant red jasmine rice quality attributes.  Before drying, fluidizing air temperature was set and the apparatus was operated for 30 min to ensure that the condition was in a steady state. Two hundred grams of cooked red jasmine rice, corresponding to a static bed height of approximately 4 cm were loaded into a cylindrical drying chamber. All experiments were assessed under full factorial design with two factors of drying temperature (90, 105, and 120 °C) and compressed air pressure (4, 5, and 6 bar) to clarify drying kinetics. The optimal compressed air pressure was subsequently used with variations of drying temperature (90-120 °C) and swirling compressed air supply times of 2, 4, and 6 min to investigate the influence on instant red jasmine rice quality attributes.

Analysis of Drying Characteristics
Moisture content of the sample was normalized using a moisture ratio (MR) and expressed as Equation (1) where M denotes moisture content and subscripts i, t and e represent moisture content at initial drying time, at a certain time and at equilibrium, respectively. All values of moisture content were expressed as wet basis (%wb) throughout this paper. Data of MR as a function of drying time, obtained from nine experimental runs with variations of drying temperature (90-120 °C) and swirling air pressure (4-6 bar), were fitted to an empirical Page equation, as shown in Equation (2): where k and n denote drying rate constant and power constant, respectively. Model parameters appearing in Equation (2) were evaluated using a non-linear regression technique. For most biological materials, drying takes place at a falling-rate with moisture movement from inside to surface depending mainly on liquid diffusion [21]. Here, effective diffusivity was calculated in accordance with the first term of Fick's second law, developed for materials in a finite circular cylinder geometry, as expressed below.
( ) Equation (3) is a liquid diffusion model governed from Fick's law, where Deff is the effective diffusivity (m 2 s −1 ), λ1 is first root of the Bessel function (2.4048), β1 is shape ratio (=πr/2l), r is the radius, l is the length of a rice grain (m) and t is drying time (min). The nonlinear function of Equation (3) was converted into a natural logarithm, resulting in a linear relationship between ln(MR) and t. Deff was then approximated from the slope

Analysis of Drying Characteristics
Moisture content of the sample was normalized using a moisture ratio (MR) and expressed as Equation (1): where M denotes moisture content and subscripts i, t and e represent moisture content at initial drying time, at a certain time and at equilibrium, respectively. All values of moisture content were expressed as wet basis (%wb) throughout this paper. Data of MR as a function of drying time, obtained from nine experimental runs with variations of drying temperature (90-120 • C) and swirling air pressure (4-6 bar), were fitted to an empirical Page equation, as shown in Equation (2): where k and n denote drying rate constant and power constant, respectively. Model parameters appearing in Equation (2) were evaluated using a non-linear regression technique. For most biological materials, drying takes place at a falling-rate with moisture movement from inside to surface depending mainly on liquid diffusion [21]. Here, effective diffusivity was calculated in accordance with the first term of Fick's second law, developed for materials in a finite circular cylinder geometry, as expressed below.
Equation (3) is a liquid diffusion model governed from Fick's law, where D eff is the effective diffusivity (m 2 s −1 ), λ 1 is first root of the Bessel function (2.4048), β 1 is shape ratio (=πr/2l), r is the radius, l is the length of a rice grain (m) and t is drying time (min). The nonlinear function of Equation (3) was converted into a natural logarithm, resulting in a linear relationship between ln(MR) and t. D eff was then approximated from the slope ((5.7831 + β 1 )D eff /r 2 ) of a function. Because D eff varied with different drying air temperatures, the Arrhenius equation was used to describe effective diffusivity affected by fluidizing air temperature as: where D 0 is a pre-exponential factor (m 2 s −1 ), E a is the activation energy (kJ kmol −1 ), R is the universal gas constant (8.314 kJ kmol −1 K −1 ) and T is the absolute temperature of drying air (K). The proposed process was successfully employed to overcome the kernel agglomeration during drying; however, additional compressed air with high shear force increased the percentage of broken rice [19]. Therefore, physical properties including breakage and agglomerate percentages were evaluated using the following equations. %Breakage = Total weight of broken rice Total weight × 100 (5) %Agglomerate = Total weight of agglomerates Total weight × 100

Color in CIE System
The color of instant red jasmine rice was measured using a Minolta Colorimeter with color system of CIE (L* a* b*), where L*, a* and b* represent lightness, redness and yellowness, respectively. Color difference (∆E) was evaluated against uncooked rice grains and expressed as ∆E = (∆a * ) 2 + (b * ) 2 , where ∆a* and ∆b* were a color difference in redness and yellowness, respectively.

Shrinkage and Rehydration Ratio
Shrinkage of instant red jasmine rice, related to the rehydration properties of dried products was measured by the volume change between cooked and dried kernels. A sample volume was measured based on liquid replacement (Archimedes' theorem). Shrinkage was presented as a percentage and expressed by: where v i and v f represent sample volume before and after drying (m 3 ). Water is absorbed during rehydration, consequently, rehydrated materials swell. Ten grams of instant red jasmine rice was immersed in 100 mL of water at temperatures of 70 • C and 100 • C for 30 min. The water was rinsed and the sample was spread on a cotton cloth for 1 min to absorb excess water on the kernel surfaces. Rehydration ratio was calculated as the weight of the rehydrated sample (w r ) divided by the weight of instant rice (w i ) as shown below.
Rehydration ratio = w r w i (8)

Microstructure of Instant Red Jasmine Rice
Porosity of the sample is normally related to rehydration quality. Therefore, the porous structure of instant red jasmine rice prepared by different drying conditions was examined using scanning electron microscopy (SEM). Following the method proposed by Prasert and Suwannaporn [7] with slight modifications, the rice kernels were broken transversely and placed on a sputtering device. Before scanning under voltage of 15 kV and ×30 magnification, the sample was coated with gold in a vacuum chamber using a sputter coater.

Analysis of Textural Properties
Hardness and stickiness are commonly reported as they impact consumer acceptability. The ISO11747 standard protocol was used with a texture analyzer (Stable Micro System, Ta-XT2i, Surrey, UK). Seventeen grams of cooked or rehydrated rice were placed in an extrusion chamber. A plunger initially situated 55 mm above a perforated base moved downward at a test speed of 1 mm s −1 until reaching the 3 mm gap between the probe surface and the base. The sample was extruded through a perforated plate and compressive force was plotted against time. Mean force (kg) over the plateau region of the curve was determined and its value divided by the testing area of the extrusion plate was considered as the mean force required to extrude the sample (kg cm −2 ).

Antioxidant Properties Sample Preparation
Dried red jasmine rice was ground and then sorted using a wired mesh size of 150 microns. Extraction was conducted in accordance with the method proposed by Sutharut and Sudarat [22] with slight modifications. Briefly, 0.25 g of ground sample was placed into a 1.5 mL microtube. After adding 1 mL of methanol, the mixture was shaken continuously for 30 s, and then heated to 60 • C in a water bath for 20 min with shaking at 10 min intervals. The mixture was then centrifuged at 13,000 rpm for 10 min, and the supernatant volume was adjusted to 5 mL with methanol. The resulting extract was kept in a fridge under 4 • C.

Total Anthocyanin Content Analysis
The method proposed by Martynenko and Chen [23] was modified and used to determine the total anthocyanin content (TAC) of the rice samples. Briefly, the extract was diluted with pH 1.0 and pH 4.5 buffer solutions and then left for 20 min to reach equilibrium. Absorbance of both buffers at 510 and 700 nm was measured by a UV-VIS spectrophotometer (HITACHI Model U1900) with the use of distilled water as a blank. TAC was calculated in terms of cyanidin-3-glucoside (Equation (9)) and expressed as units of mg cyanidin-3-glucoside per 100 g dry matter.
where A diff is the difference of absorbance between two wavelengths (510 and 700 nm) for pH 1.0 and pH 4.5, MW is molecular weight of cyanidin-3-glucoside (449.2 g mol −1 ), ε is the molar absorbability (26,000 L mol −1 cm −1 ), l is cuvette width (1 cm) and DF is a dilution factor.

Total Phenolic Content Analysis
The protocol described by Butsat and Siriamornpun [24] was slightly modified to analyze total phenolic content of instant red jasmine rice dried under different conditions. The extract was mixed with 10% Folin-Ciocalteu reagent in portions of 0.1 mL and 0.5 mL, respectively. After shaking and then storing in a dark container for 1 min, sodium carbonate (7.5%, 1.5 mL) was added to the mixture and the volume was adjusted with distilled water to 4 mL. Before measuring the absorbance at wavelength 765 nm using a UV-VIS spectrophotometer, the mixture was kept under ambient conditions for 30 min. The standard curve of gallic acid was used to calculate total phenolic content (TPC), presented in terms of mg gallic acid equivalent (GAE) per 100 g dry sample.

Total Flavonoid Content Analysis
Total flavonoid content (TFC) was determined according to the method reported by Chang et al. [25]. The protocol with slight modification was as follows. A mixture was first prepared using distilled water (2.8 mL), methanol (1.5 mL), 10% aluminum chloride (0.1 mL), and 1M potassium acetate (0.1 mL), and then added with 0.5 mL of the extract. The resultant solution was kept at room temperature for 30 min before measuring absorbance at 415 nm. TFC (mg QE per 100 g dry matter) was determined using the standard curve of quercetin solution as a function of absorbance.

Analysis of DPPH Radical Scavenging Activity
Antioxidant activity was analyzed by the DPPH radical scavenging activity method following the protocol reported by Loypimai et al. [26]. DPPH solution at a concentration of 0.1 mmol was first prepared under darkness, and then 3.0 mL of this solution was added Processes 2021, 9, 1738 7 of 21 into 1 mL of the extract. After shaking and storing in a dark container for 30 min, the mixture was subjected to absorbance measurement at 517 nm using a spectrophotometer. Antioxidant activity of instant red jasmine rice affected by the drying process was reported as scavenging percentage and calculated by Equation (10): where A control and A sample are the absorbance values of the control and sample at wavelength of 517 nm, respectively.

Response Surface Methodology
To measure all responses (y), a full factorial design with two factors: fluidizing air temperature (T, • C) and swirling air supply time (ST, min) with three levels was employed. Responses tested included percentage agglomerate (%agglomerate), percentage breakage (%break), percentage shrinkage (%shrink), rehydration ratio (RR), total phenolic content (TPC), total flavonoid content (TFC), total anthocyanin content (TAC), and antioxidant activity expressed by DPPH radical scavenging (DPPH). All variables and their corresponding coded values are presented in Table 1. Quadratic regression, as expressed in Equation (11), was employed to predict the response variables [27]. In Equation (11), β 0 , β i and β ii represent the constant terms of intercept, the linear, quadratic and interaction term, respectively, while β ij is the effect of the ij interaction between factors x i and x j .
The desirability function was used to evaluate the responses by considering the following constraints: (a) minimum values of agglomerate, breakage and shrinkage percentages, and (b) maximum values of rehydration ratio and retention of all antioxidant properties.
In order to compare the mean values of the results presented in this work, one-way ANOVA was used with Tukey's test at 5% probability.

Drying Model
The semi-empirical Page model, widely used to describe the drying behavior of biological materials [28], was fitted to experimental moisture ratio (MR) with variations of swirling compressed air pressure and fluidizing air temperature. Important characteristic parameters of drying, including drying rate constant (k), power constant (n), effective diffusivity (D eff ) and activation energy (E a ) were estimated, as shown in Table 2.
The rate constant (k) indicates the drying performance, where a high drying rate is represented by a high value. This parameter was used to determine the most suitable condition, based on the highest k value, for further investigation. As shown in Table 2, the k values varied with different drying conditions and ranged from 0.09745 (at a swirling compressed air pressure of 5 bar and an air temperature of 90 • C) to 0.15072 (at a swirling compressed air pressure of 6 bar and an air temperature of 120 • C). At each drying temperature, the k values decreased when using higher pressure (from 4 bar to 5 bar). This finding concurred with our results reported previously [19], stating that decrease in the drying rate constant was attributed to interference of additional compressed air. However, at 6 bar the k values increased at all temperatures, possibly attributed to higher turbulence intensity and partial impingement scheme caused by the high velocity air jet releasing from two nozzles located at the bottom bed. Table 2. Drying characteristic parameters, effective diffusivity, and activation energy affected by swirling compressed air pressure and fluidizing air temperature.

SCAP (bar)
FAT In a fluidized bed system, convection plays an important role by influencing heat transfer performance. The hydrodynamics of gas-solid phases are closely related to heat transfer phenomena and can be enhanced by turbulence due to increased flow homogeneity and mixing efficiency [29,30]. At the high pressure of 6 bar, increased turbulence intensity was dominant and affected heat transfer phenomena compared to interference effects of additional ambient air streams at pressures of 4 and 5 bar. In this fluidization regime, solid circulation and contact with a gas phase could be improved, resulting in higher efficiency. Heat transfer enhancement due to increased turbulence flow field intensity has been achieved in many applications [29][30][31]. Another plausible explanation was the impingement effect from high pressure air streams at the dense bottom bed. Tangential gas jets removed thermal boundary layers on the grain surface, resulting in higher heat and mass transfer coefficients [32][33][34].
The Page model parameter influencing the drying mechanism depends on the material type and drying conditions [14]. Table 2 shows this exponential constant varying from 0.82108 to 0.89014, tending to increase with higher air temperature but not definitely observed with respect to compressed air pressure. Correlations of the constants (k and n) obtained from the Page drying model were evaluated. Both model constants correlated quadratically with fluidizing air temperature (T) and pressure of swirling compressed air (p), with R 2 values of 0.9059 and 0.6730 for the rate constant k and power constant n, respectively. Thus, the Page equation used to describe the drying characteristics of the proposed process was expressed as: where k = 0.39744 + 1.01 × 10 −3 T − 0.16458P − 4.92 × 10 −5 TP + 5.11 × 10 −7 T 2 + 0.01794P 2 (R 2 = 0.9059), and n = 0.44849 + 0.01467T − 0.17758P + 8.25 × 10 −5 TP − 6.62 × 10 −5 T 2 + 0.017668P 2 (R 2 = 0.6730). The Page equation (Equation (12)) and its model parameters were used to predict the drying time to obtain the desired moisture content (MC) of 10% (wb). Five drying runs with corresponding estimated drying times (see Table 3) were conducted in triplicate for model validation. The percentage of mean relative error (%error) between experimental moisture content (MC exp ) and desired moisture content (MC desir ) for each run showed reasonable Processes 2021, 9, 1738 9 of 21 discrepancy ranging 10.79-23.72%. The modulus of mean relative deviation (%p) was also calculated as the summation of all relative errors divided by the number of experimental runs. The %p of 17.89%, calculated by Equation (13), indicated good agreement between the experimental data and the predicted model results.

Effective Diffusivity
In a falling-rate drying process, moisture in food materials transfers to the surface and subsequently to ambient air by means of internal diffusion, called effective diffusivity (D eff ) [21]. Table 2 shows how this drying characteristic was affected by operating parameters. Effective diffusivity varied in the range 2.92-5.94 × 10 −8 m 2 s −1 , higher than the normally expected range of 10 −11-10 −9 m 2 s −1 for food materials [35]. D eff is commonly related to the drying rate constant; higher D eff means faster drying. At constant compressed air pressure, D eff tended to increase with higher air temperature. This is commonly observed in the drying process since heat and mass transfer are enhanced when higher driving force is provided by increasing thermal energy [36]. Zielinska and Michalska [21] reported that the D eff value increased with higher air temperature in convective hot air drying of blueberries. A similar result was also observed when mortiño (Vaccinium meridionale Swartz) was dried at 40-60 • C [36].
With variation of swirling air pressure, a decreasing trend was found when increasing pressure from 4 bar to 5 bar at each air temperature. This was consistent with our previous study [19] when interference of additional ambient compressed air led to reducing drying rate and subsequently reduced effective diffusivity. However, D eff slightly increased at the same temperature when compressed air pressure increased to 6 bar. As previously explained, high pressure air resulted in more turbulence intensity that enhanced heat transfer in the system [29][30][31]. Air impingement was also a plausible explanation for this observation. Tangential air jets with high velocity may remove or reduce the thermal boundary layers of tangentially fluidizing grains when passing nozzle outlets, and this enhances heat and mass transfer. At high pressure, high velocity air streams avoid case hardening, resulting in increased moisture evaporation from grain surfaces [32][33][34].

Activation Energy
In the drying process, moisture is transported from a solid matrix using activation energy. Table 2 also shows activation energy (E a ) values of 15.45, 13.66 and 13.81 kJ mol −1 at compressed air pressure of 4, 5, and 6 bar, respectively. These values were in the common range for dried food materials of 12.7 to 110 kJ mol −1 [37]. However, activation energy obtained here was not significantly different at varied pressures of swirling compressed air.
Based on the highest drying model parameters (k and n) and effective diffusivity (D eff ), swirling compressed air pressure of 6 bar was selected to prepare instant red jasmine rice.

Influence of Fluidizing Air Temperature Combined with Supply of Swirling Compressed Air
The influences of fluidizing air temperature and supply time of swirling compressed air were investigated using response surface methodology (RSM). Responses of these two input parameters were categorized into two groups including physical and rehydration properties and antioxidant properties.

Physical Properties and Rehydration Ratio
Experimental design and the corresponding response values evaluated using ANOVA are presented in Table 4. During the optimization process, response values of physical and rehydration properties were predicted by estimated coefficients for the actual functional components, as shown in Table 5. All responses were extremely significant compared to the model (p < 0.01) with regression coefficients higher than 0.7.   Table 5 show that the effects of monomial fluidizing air temperature and its quadratic term on percentage agglomeration were extremely significant at p < 0.01, while the other terms were insignificant (p > 0.05). The F value in Table 5 indicated fluidizing air temperature as the most important, while swirling time did not significantly affect the agglomerate percentage. For percentage breakage, the effects of monomial fluidizing air temperature and swirling time, as well as their interaction were extremely significant at p < 0.01, while all quadratic terms were insignificant (p > 0.05). Among these two factors, the higher F value showed that fluidizing air temperature was the most important. Similarly, the linear term of both input parameters greatly affected percentage shrinkage at p < 0.01, whereas the interaction term and quadratic fluidizing air temperature were not significant (p > 0.05). Results in Table 5 also showed that the quadratic swirling time was significant at the p < 0.05 level, and this input parameter was more important, as confirmed by the higher F value. Effects of monomial fluidizing air temperature and swirling time on rehydration ratio were also significant at p < 0.01 and p < 0.05, respectively, while the other terms were insignificant. According to the F value in Table 5, fluidizing air temperature was the most important, followed by the swirling time. Figure 4 shows three-dimensional plots of all responses regarding physical and rehydration properties as a function of fluidizing air temperature and swirling time. Fluidizing air temperature (T) had a negative influence on percentage agglomerate of the dried cooked rice sample (Figure 4a), while agglomeration did not change significantly with increasing swirling time (ST) at temperature higher than 90 • C. In our proposed process, cooked rice was prepared before subjection to the drying step. Starch gelatinization resulted in surface stickiness among cooked rice kernels, leading to problematic agglomeration during drying. At 90 • C, swirling compressed air time significantly affected the agglomeration percentage of the kernels (∼2.3-3.5%); the longer the lower. This result slightly differed from our previous study where percentage agglomeration of riceberry rice kernels was ∼0.8% at 4 bar [19]. The difference was attributed to longer time from the onset till the end of the drying process, while time spent supplying high pressure air streams here ranged between 2 and 10 min. At temperatures higher than 90 • C, agglomerates were not observed and the rice kernels dried quickly and were completely separated by the high shear force exerted by compressed air streams. Therefore, high fluidizing air temperature with the shortest swirling air time was the optimal condition with respect to minimal agglomeration. Figure 4b shows the influence of fluidizing air temperature and swirling time on percentage breakage. An obvious trend appeared when using higher temperature and longer time. Maximum value of percentage breakage at 75% was found when drying at 120 • C and supplying swirling air for 10 min. Swirling compressed air reduced surface stickiness during drying but also resulted in kernel breakage due to high shear force. Figure 4b shows the breakage percentage of rice kernels at varying fluidizing air temperatures and swirling compressed air times. At a fixed temperature, breakage increased with time of supplying swirling air, except at 90 • C when breakage slightly increased but not statistically significantly (p > 0.05). Fluidizing air temperature strongly affected kernel breakage, ranging 17.33-24.00%, 18.67-39.33% and 13.00-72.00% at 90 • C, 105 • C and 120 • C, respectively. Increased fragrance at higher temperature could be a possible explanation for this trend, while at the same temperature longer compressed air supply resulted in higher breakage. Consequently, 120 • C with compressed air supplied for 2 min was a suitable condition that retained perfectly shaped kernels without agglomerates.

Results in
During drying, solid shrinkage generally occurs due to collapse and disruption of the cells inside a material, especially at higher temperatures [38]. Solid shrinkage is normally influenced by variations of drying factors, as shown in Figure 4c. Percentage shrinkage ranged from 33.67 to 47.67%, depending on the fluidizing air temperature and compressed air time. Figure 4c shows that the swirling time had positive effect on percentage shrinkage, while negative influence was found for fluidizing air temperature. Percentage shrinkage slightly decreased with higher temperatures ranging 42.00-47.67%, 37.67-46.33% and 33.67-44.33% at 90 • C, 105 • C and 120 • C, respectively. During drying, water (moisture) evaporated and then moved from the inside toward the material surface due to the pressure imbalance resulting in rice kernel shrinkage. At higher temperatures, higher drying rate Processes 2021, 9, 1738 12 of 21 caused rapid heating and accumulation of vapor pressure in the rice kernels. This resulted in less volume shrinkage or even volume expansion. This so-called puffing effect impacted the microstructure, normally required to facilitate the rehydration process [7,13]. A slight decrease in shrinkage with higher drying temperature obtained here contradicted the results of Le and Jittanit [13] who used lower temperatures of 50-90 • C. ation during drying. At 90 °C, swirling compressed air time significantly affected glomeration percentage of the kernels (∼2.3-3.5%); the longer the lower. This slightly differed from our previous study where percentage agglomeration of ri rice kernels was ∼0.8% at 4 bar [19]. The difference was attributed to longer time fr onset till the end of the drying process, while time spent supplying high press streams here ranged between 2 and 10 min. At temperatures higher than 90 °C, a erates were not observed and the rice kernels dried quickly and were completel rated by the high shear force exerted by compressed air streams. Therefore, high ing air temperature with the shortest swirling air time was the optimal conditio respect to minimal agglomeration.  Figure 4b shows the influence of fluidizing air temperature and swirling t percentage breakage. An obvious trend appeared when using higher temperatu longer time. Maximum value of percentage breakage at 75% was found when dr 120 °C and supplying swirling air for 10 min. Swirling compressed air reduced stickiness during drying but also resulted in kernel breakage due to high shear for ure 4b shows the breakage percentage of rice kernels at varying fluidizing air te tures and swirling compressed air times. At a fixed temperature, breakage increase time of supplying swirling air, except at 90 °C when breakage slightly increased statistically significantly (p > 0.05). Fluidizing air temperature strongly affected breakage, ranging 17.33-24.00%, 18.67-39.33% and 13.00-72.00% at 90 °C, 105 °C a Shrinkage increased with compressed air time as a result of system interference by additional ambient air, consistent with our results reported previously [19]. Interestingly, interference of swirling time on shrinkage percentage dramatically increased with longer time due to additional ambient swirling air interfering in the drying system, resulting in lower drying rate. The moisture was removed slowly, and resultant cell collapse and disruption led to more solid shrinkage [38]. Therefore, in addition to the lowest agglomeration percentage and breakage percentage, fluidizing air temperature of 120 • C and swirling air time of 2 min were considered as the optimal drying condition that provided the lowest shrinkage.
Rehydration is a crucial characteristic of instant products that is normally related to material shrinkage [12]. Samples with less shrinkage have more pore structure and this facilitates water absorption during rehydration. Figure 4d shows the rehydration ratio (RR) of instant red jasmine rice after rehydration in boiling water for 30 min. Figure 4d shows that fluidizing air temperature had a positive influence on rehydration ratio, especially at the shortest swirling time, while the swirling time slightly affected rice rehydration. The RR varied with different fluidizing air temperatures and swirling air times, ranging from 0.90 to 1.35 (Table 4). High values indicated high capacity of water absorption during the rehydration process. At fixed air temperature, RR tended to decrease with longer supply of swirling air, possibly caused by more interference of lower temperature air supplied to the system with longer compressed air time. The colder fluidizing air led to slower drying, and subsequently compacted the microstructure inside the dried kernels. At each swirling air time, drying temperature significantly affected RR and increased with higher temperature. The porous structure could be a plausible explanation for higher RR. Maximum value of RR was obtained at 120 • C for 2 min. This condition provided high drying rate as moisture inside the cooked kernels moved quickly to the surface, and subsequently less cell shrinkage. These results were confirmed by SEM images as shown in Figure 5a-j. It is shown from this figure that porosity in the rice kernel dried under fluidizing air temperature of 120 • C and supply time of 2 min (Figure 5g) was reasonably comparable with that of freeze-dried sample, as shown in Figure 5j. Table 6 shows values of responses, including total phenolic content (TPC), total flavonoid content (TFC), total anthocyanin content (TAC) and antioxidant activity (DPPH) of red jasmine rice. Antioxidant properties changed with varied drying conditions. In the drying process, temperature generally plays the most important role as it accelerates reactions that cause oxidative decomposition or activities of bioactive compounds. Changes in antioxidant properties with various drying temperatures found here were consistent with previous studies but in different ways, decreasing [39][40][41][42] or increasing [43,44] at higher operating temperatures. Antioxidant compounds such as phenolics, flavonoids and anthocyanins are commonly abundant in pigmented rice as they are mainly linked with the grain pericarp [45]. The influence of heat in the proposed drying method on such properties is presented in Table 6. Freeze-drying is an ideal process that provides dry materials with high physical and chemical qualities [46]. Compared to freshly cooked rice, all antioxidant properties of freeze-dried samples slightly decreased. The highest reduction of 9.5% was determined for TFC, while TPC, TAC and DPPH decreased by only 2.7%, 5.3% and 3.3%, respectively. These slight reductions in antioxidant properties were consistent with previous research focusing on the influence of freeze-drying [47].  Table 6 shows values of responses, including total phenolic content (TPC), total vonoid content (TFC), total anthocyanin content (TAC) and antioxidant activity (DPP of red jasmine rice. Antioxidant properties changed with varied drying conditions. In drying process, temperature generally plays the most important role as it accelerates actions that cause oxidative decomposition or activities of bioactive compounds. Chan in antioxidant properties with various drying temperatures found here were consist with previous studies but in different ways, decreasing [39][40][41][42] or increasing [43,44 higher operating temperatures. Antioxidant compounds such as phenolics, flavonoids and anthocyanins are co monly abundant in pigmented rice as they are mainly linked with the grain pericarp [4 The influence of heat in the proposed drying method on such properties is presented Table 6. Freeze-drying is an ideal process that provides dry materials with high phys and chemical qualities [46]. Compared to freshly cooked rice, all antioxidant propertie freeze-dried samples slightly decreased. The highest reduction of 9.5% was determin for TFC, while TPC, TAC and DPPH decreased by only 2.7%, 5.3% and 3.3%, respectiv These slight reductions in antioxidant properties were consistent with previous resea focusing on the influence of freeze-drying [47].

Antioxidant Properties
In addition to values of physical and rehydration responses, Table 7 shows ANO results of the fitted model for antioxidant properties in fluidized-bed drying assisted w In addition to values of physical and rehydration responses, Table 7 shows ANOVA results of the fitted model for antioxidant properties in fluidized-bed drying assisted with swirling compressed air of instant red jasmine rice. Results in Table 7 show that the effects of monomial fluidizing air temperature on TPC were significant at the p < 0.01 level, while the other terms were insignificant. The F value confirmed that fluidizing air temperature was the most important input parameter. Considering the response of TFC, influences of monomial fluidizing air temperature and its quadratic term were significant at the p < 0.01 level and p < 0.05 level, respectively, whereas the other terms were insignificant. The F value of fluidizing air temperature was extremely higher than swirling time, implying a more dominant input parameter of the drying system. TAC was affected by fluidizing air temperature and swirling time at significance levels of 0.01 and 0.05, respectively, while the former was a more important factor as it gave a higher F value. Estimated coefficients of DPPH radical scavenging showed that only monomial fluidizing air temperature significantly (p < 0.01) affected the response, corresponding to higher F value and indicating that drying air temperature was more important (Table 7). Figure 6a-d shows three-dimensional plots of TPC, TFC, TAC, and DPPH as a function of both operating parameters. Fluidizing air temperature had a dominantly negative effect on antioxidant properties tested here, while swirling time had an insignificant influence. In Figure 6a, maximum retention of TPC at 485-489 mg GAE per 100 g was found at 90 • C when supplying swirling times in a test range. However, at the highest temperature, influence of swirling time became more significant as TPC increased at longer time due to interference of additional ambient air for a longer time, resulting in reduced thermal degradation of TPC. Similarly, fluidizing air temperature had a strong effect on TFC of dried cooked jasmine rice (Figure 6b). At each swirling time, TFC quadratically decreased with higher temperature. This observation was consistent with the ANOVA results of linear and quadratic term fluidizing air temperature (X 1 and X 2 1 , respectively), as shown in Table 7. Maximum TFC of approximately 27 mg QE per 100 g was found at 90 • C. TAC decreased with higher fluidizing air temperature in the same manner as found in TPC and TFC. Maximum TAC of 35 mg cy-3-glu per 100 g was obtained at 90 • C with swirling time of 2 min. Antioxidant activity represented by DPPH radical scavenging (%inhibition) as affected by input parameters is shown in Figure 6d. Again, fluidizing air temperature had a dominant influence on DPPH of the dried rice sample, while less significance was found for swirling time. Maximum value of 54% inhibition was obtained when drying at 90 • C.
After drying using the fluidized bed equipment assisted with compressed swirling air streams, all antioxidant properties tested significantly decreased with variations of operating parameters. Slight decrease was found for TPC (8-11%), TFC (23-29%), and DPPH (14-19%), while TAC dramatically decreased by 40-42%. Most antioxidant compounds are susceptible to environmental factors such as temperature, oxygen, and humidity [43], explaining these observations. Ratseewo et al. [48] reported that TPC, TFC, TAC, and DPPH reduced by 21%, 25%, 53% and 27.5%, respectively after red jasmine rice was subjected to hot air drying at 60 • C. These findings concurred with other applications, for instance, rice bran and husk [49], mulberry leaves [50], blueberry leather [51] and mortiño [36]. A drying temperature of 90 • C and all times of swirling air supply was the most suitable condition, giving the lowest reduction levels of all antioxidant properties, as shown in Figure 6. Conversely, severe operating temperature of 120 • C and 2 min supply of compressed air gave the highest percentage of reduction compared to the freshly cooked sample.

Optimization
Problems encountered with surface stickiness and preparation steps were reduced by our novel proposed drying technique to produce instant red jasmine rice. As described in previous sections, all operating parameters affected both physical and antioxidant properties in different ways. Therefore, correlations between drying factors (fluidizing air temperature, T and swirling air supply time, ST) and their responses (physical, rehydration and all antioxidant properties) were determined. Drying conditions were subsequently optimized based on criteria of minimum values of agglomerate, breakage and shrinkage percentages, maximum values of rehydration ratio, and all antioxidant properties. Figure 7 shows a contour plot of desirability function obtained from correlations be-tween input parameters and response of percentage agglomerate, percentage breakage, percentage shrinkage, rehydration ratio, TPC, TFC, TAC and DPPH as response functions. Table 8 indicates all optimal process conditions with desirability values ranging from 0.532 to 0.603. Based on the highest desirability value of 0.603, fluidizing air temperature and compressed air time at 98.5 • C and 2 min, respectively, were recommended as a guide for process optimization. swirling time of 2 min. Antioxidant activity represented by DPPH radical scaven (%inhibition) as affected by input parameters is shown in Figure 6d. Again, fluidizin temperature had a dominant influence on DPPH of the dried rice sample, while less nificance was found for swirling time. Maximum value of 54% inhibition was obta when drying at 90 °C. After drying using the fluidized bed equipment assisted with compressed swi air streams, all antioxidant properties tested significantly decreased with variations o erating parameters. Slight decrease was found for TPC (8-11%), TFC (23-29%), and D (14-19%), while TAC dramatically decreased by 40-42%. Most antioxidant compo are susceptible to environmental factors such as temperature, oxygen, and humidity explaining these observations. Ratseewo et al. [48] reported that TPC, TFC, TAC, DPPH reduced by 21%, 25%, 53% and 27.5%, respectively after red jasmine rice was jected to hot air drying at 60 °C. These findings concurred with other applications instance, rice bran and husk [49], mulberry leaves [50], blueberry leather [51] and mo [36]. A drying temperature of 90 °C and all times of swirling air supply was the suitable condition, giving the lowest reduction levels of all antioxidant properties, as shown in Figure 6. Conversely, severe operating temperature of 120 °C and 2 min supply of compressed air gave the highest percentage of reduction compared to the freshly cooked sample.

Optimization
Problems encountered with surface stickiness and preparation steps were reduced by our novel proposed drying technique to produce instant red jasmine rice. As described in previous sections, all operating parameters affected both physical and antioxidant properties in different ways. Therefore, correlations between drying factors (fluidizing air temperature, T and swirling air supply time, ST) and their responses (physical, rehydration and all antioxidant properties) were determined. Drying conditions were subsequently optimized based on criteria of minimum values of agglomerate, breakage and shrinkage percentages, maximum values of rehydration ratio, and all antioxidant properties. Figure 7 shows a contour plot of desirability function obtained from correlations between input parameters and response of percentage agglomerate, percentage breakage, percentage shrinkage, rehydration ratio, TPC, TFC, TAC and DPPH as response functions. Table 8 indicates all optimal process conditions with desirability values ranging from 0.532 to 0.603. Based on the highest desirability value of 0.603, fluidizing air temperature and compressed air time at 98.5 °C and 2 min, respectively, were recommended as a guide for process optimization.    T = fluidizing air temperature ( • C), ST = swirling air time (min), %Agglo = %agglomerate, %Broke = %breakage, %Shrink = %shrinkage, RR = rehydration ratio, TPC = total phenolic content (mg GAE/100g), TFC = total flavonoid content (mg QE/100g), TAC = total anthocyanin content (mg cy-3-glu/100g), DPPH = DPPH radical scavenging (%inhibition).

Suitable Rehydration Condition
Surface stickiness of rice kernels can be avoided using our proposed drying method, with high quality in terms of physical, rehydration and antioxidant properties. However, textural characteristics of rehydrated rice are important for consumer preference. Therefore, instant red jasmine rice prepared by the aforementioned optimal drying condition (98.5 • C with swirling air time 2 min) was rehydrated under different rehydration conditions and textural properties were compared with the control sample conventionally cooked by a domestic rice cooker, as shown in Table 9. The instant rice sample was suitably rehydrated in 70 • C water for 10 min. Under this condition, mean force of 13.18 ± 0.79 kg and mean extrusion force of 1.76 ± 0.10 kg cm −2 were reasonably comparable with the control (16.42 ± 0.36 kg and 2.19 ± 0.04 kg cm −2 , respectively), while the freeze-dried sample was extremely soft with low values of 6.60 ± 0.23 kg and 0.88 ± 0.03 kg cm −2 , respectively. Therefore, instant red jasmine rice prepared by our proposed novel drying method under optimal conditions should be rehydrated in water at 70 • C for 10 min.

Conclusions
Fluidized bed drying associated with swirling compressed air was proposed as an alternative process to produce instant red jasmine rice. The rate constant obtained from the drying Page equation increased with higher temperature but decreased with the pressure of compressed air up to 5 bar. Interestingly, a drying temperature of 120 • C associated with 6 bar compressed air pressure was the most suitable, probably due to higher turbulence intensity and/or partial impingement. Extending the time for supplying high velocity air jets resulted in less kernel agglomeration but an increase in grain breakage and shrinkage. With increasing temperature and shorter time for compressed air streams, a higher degree of rehydration was observed, as confirmed by the more porous structure shown by scanning electron microscopy. However, at more severe drying temperatures, deterioration of antioxidant properties dominated. The optimal operating parameters were determined based on a compromise between the physical properties, rehydration ratio, and all antioxidant properties. Response surface methodology provided the optimal fluidizing air temperature and compressed air duration of 98.5 • C and 2 min, respectively. Textual analysis of the rehydrated sample suggested that soaking instant red jasmine rice in warm water at 70 • C for 10 min was preferable compared to conventional cooked rice.