Tempering time and air temperature relationships for real-scale paddy drying and their effects on the physical, physicochemical, and morphological qualities of polished rice

Abstract Tempering-drying time may affect the quality of polished rice. Hence, this study aims to evaluate the effect of drying-tempering time and air temperature relationships on the physical, physicochemical, and morphological qualities of polished rice. Herein, paddy with 20–24% (w.b.) moisture were subjected to drying and tempering with air temperatures of 75–115 °C, 115–130 °C, and 75–130 °C over 14 h to reduce their moisture content up to 12% (w.b.). With the increase in the paddy mass temperature to 49 °C, the whole grain yield and the percentages of protein, crude fiber, fat, and ash reduced; additionally, the morphology of starch and the composition of minerals were altered. The cumulative effects and prolonged time of the drying-tempering influenced the physical, physicochemical, and morphological qualities of the rice. In conclusion, paddy drying technologies with tempering periods should not be designed for temperatures not exceeding 75 °C and drying times not longer than 8 h for good system performance and rice quality.


Introduction
Rice are mainly composed of primarily comprise carbohydrates, starch, proteins, lipids, fiber, minerals, vitamins, and phenolic. Starch is the main component of the endosperm, formed primarily by amylose and amylopectin. Maintaining the quality and yield of whole grains can determine the economic value of the product and the fate of rice in the industry. [1] Thus, the post-harvest processes, including drying, are fundamentally important to reduce the moisture content of the grains for favorable conditions of storage and milling of rice without losses.
The industrial quality and physicochemical composition of rice are affected by drying time, drying technology, and temperature. Thus, continuous drying with heated air increases the operational capacity and reduces the processing time; however, it can cause fissures, cracks, and breaks in the grain, thus altering the final quality of the processed rice. The high air drying temperature can increase the grain mass temperature and cause variations in the moisture content, as well as, pre-gelatinization and reduction of the starch crystallinity. [2] Tempering drying technology is most used in the paddy drying. It is characterized by the recirculation of the grain mass in the dryer with rest periods in the flashing chamber. After passing through the drying chamber, the grains are subjected to a rest period to ensure that the moisture from the innermost layers of the grains migrates from the center to the periphery owing to the potential difference, without the direct action of heat. This happens owing to the temperature acquired by the grains themselves passing through the drying chamber. [3] Internal tensions are formed in the grain during the continuous drying step. [4] These stresses can cause physical damage and potential grain deterioration. The variation of the moisture promotes the formation of cracks in the paddy during continuous drying, making the process critical. [4] The tempering stage can minimize the abrupt transfer of moisture inside the grain to the periphery and, consequently, mitigate the effect of drying temperature, thus achieving better quality in the grains. [5] The relationship between drying time and drying air temperatures optimize the process and final product quality. In tempering drying, monitoring the moisture content of the grains, temperatures of the grain mass and air, and tempering period are essential to guarantee operational quality.
Wang et al. [2] evaluated the effect of drying approaches in conjunction with variable temperature and tempering on the physicochemical quality of rice. According to the glass transition curve, the 17% moisture content of the rice was used as the time point for variable temperature. The variable temperature and tempering treatment accelerated the diffusion of water within the rice, thereby lowering the development of fissures and increasing the head rice yield. The results showed that increasing the tempering treatment changed the physicochemical quality of rice, and that the different drying approaches with development of fissures, rice yield, taste quality, and textural properties. High temperatures resulted in partial gelatinization of the starch, destruction of the starch granules, and an increase in fissures. In addition, the drying approaches with temperature 35 C, variation of 45-50 C, and 45 C in combination with the tempering treatment exhibited a high overall physicochemical quality. According to Wang et al. [2] the combination of variable temperature and tempering is an effective drying approach for enhancing the physicochemical quality of rice. Nosrati et al. [4] studied the moisture variation and fissure formation of rice kernels during multi-stage intermittent drying were simulated and analyzed based on the experimental data obtained in a laboratory-scale infrared-assisted vibratory bed dryer. The variations of drying parameters included farinfrared radiation intensity (0, 1000 and 2000 W m À2 ), inlet drying air temperature (30, 40 and 50 C), drying duration (15, 30 and 40 min), and tempering ratio (0, 2, 4 and 6). Two-dimensional moisture distribution within individual kernel was predicted. By assessing cracked kernel and specific energy consumption and simulation results, it was found that the magnitude of 10% d.b. mm À1 was considered as an index for the critical value for moisture content gradient to achieve the suitable drying and tempering duration. It was further recommended that the intermittent drying duration at each stage should be selected in such a way that the moisture content gradient value does not exceed the critical level and the shortest possible tempering duration should be chosen in a manner that at least 40% of critical level of moisture content gradient be eliminated.
However, the effect of the drying-tempering time on the quality of the grains could be evaluated as the longer the time, the greater the number of times the product that must undergo drying in the moisture removal process. Thus, the use of high air temperatures and variables at different stages throughout the drying process can minimize the effects of the dryingtempering time on the physical, physicochemical and microstructural quality of rice, since, the polishing of rice, being an abrasive process can cause more significant cracks in the endosperm of the grain, thereby altering the physical, physicochemical, and morphological qualities of the rice. Hence, this study aimed to evaluate the effects of drying-tempering time, different drying air temperature in the stages of the process, and milling on the physical, physicochemical, and morphological qualities of polished rice for process optimization using X-ray diffraction (XRD) analysis techniques, near-infrared spectroscopy (NIRS), and scanning electron microscopy (SEM).

Characterization of the experiment
The batches of paddy (IRGA 424 variety, 29% amylose) from mechanized harvesting with grain moisture content between 20% and 24% (w.b.) were immediately submitted to pre-cleaning and drying operations. The treatments analyzed were drying and tempering in three stages, namely initial (0-5 h), intermediate For each cycle of the grain mass in the dryer, the product remained one hour in the drying chamber and one hour in the tempering chamber. The grain mass was rotated 14 times through the dryer until complete drying. The average speed of the drying air was 5 m s À1 . The cumulative effects of drying and tempering were evaluated in the grains sampled and submitted to physical quality of rice and the physicochemical quality, crystallinity, and morphology of the polished rice.

Drying and sampling of paddy
After undergoing the pre-cleaning process in the air machine and sieve, the paddy were dried in a tempering dryer ( Figure 1) until the moisture content of the grains was reduced to 12% (w.b.). The tempering dryer comprised a drying chamber, tempering chamber, unloading, heating and ventilation system, and an elevator to perform flashing. In this system, the product was not constantly exposed to the drying air but only when it passed through the drying chamber. The product circulation time outside the drying chamber was considered as equalization time (tempering time).
Grain samples were collected hourly in the tempering and drying chambers. The grain mass temperature from the samples collected was measured using a mercury thermometer. Simultaneously with the collection of the samples, the speed and temperature of the drying and exhaust air were measured at the air inlet of the dryer and the outlet of the hood, respectively, using a blade anemometer. The ambient air temperature and relative humidity were measured using a calibrated digital thermohygrometer. The moisture content was determined via the indirect method of electrical capacitance and measured with the standard method in an oven at 105 C and 24 h. [6] Rice milling The paddy were milled in a rice tester (Zaccaria, PAZ-1/DTA, Limeira, Brazil), which was regulated and treated according to the technical recommendations for rice milling in the equipment manufacturing industry. [6] First, 100 grams of grains in husk were weighed and gradually placed in the "cone" hopper at the entrance of the beneficiation equipment to obtain the polished rice. Next, polishing was performed by the abrasive stones of the hone. After milling, the samples were submitted for quality analyses. Physical classification of rice A 5.5 mm honeycomb separator cylinder (Zaccaria, Paz-1/DTA, Limeira, Brazil) was used for sorting the whole grains. The rotating separator cylinder contained a gravitational function, which separated the grains by moving the broken grains to the horizontal hopper, whereas the whole grains were trapped in the cylinder containers for future unloading. The processed samples were weighed on a precision balance (Marte Cient ıfica, model AD330, São Paulo, Brazil) for future forwarding to the classification of polished rice according to the Normative Instruction 6/2009 of the Ministry of Agriculture, Livestock and Supply. [6] O The following physical defects were classified: broken grains in classification (BGC), cracked grains (CG), fragmented grains (FG), healthy grains (HEA), burnt grains (BG), moldy grains (MO), plastered grains (PLA), and chopped and stained grains (CS).
Evaluation of the physicochemical quality of rice NIRS (Metrohm, DS2500 spectrometer, Herisau, Switzerland) was used to determine the starch (ST), crude protein (CP), fat (Fat), ash (AS), and crude fiber (CF) content in the rice. The samples were homogenized and placed in the sampling capsule. The analysis was based on illuminating a sample with near-infrared radiation; subsequently, the difference between the amount of energy emitted by the spectroscopy and that reflected by the sample to the detector was measured. This difference was measured in several bands, thereby creating a spectrum for each sample. The spectral data recordings in reflectance mode were obtained in triplicate in the spectral range of 400-2500 nm. The accuracy level of the NIRS data with respect to true methods used to determine the properties was 95%. [7] X-ray diffraction analysis (XRD) The diffractograms of the samples were obtained using a Rigaku X-ray diffractometer (model Miniflex 300), operating in step mode with a scan speed of 0.5 s and scan step of 0.03 at angles 5 -100 . The equipment had Cu Ka radiation (k ¼ 1.54184 Å), with power supply of 30 kV and 10 mA. The relative crystallinity (RC) of the starch granules was calculated using Eq. (1), where A c is the crystalline area, and A a is the amorphous area on the X-ray diffractograms.

Statistical analysis
From Table 1, the treatments including drying and tempering temperatures and times of duration were evaluated for the dependent variables in a heat map using the average Euclidean distance and the k-means clustering method to evaluate the polished rice. These analyses were performed using the "ggfortify" package of the free application R; the procedures recommended by Naldi et al. were followed. [8] Pearson's correlations were estimated to verify the association between the variables. The results were graphically expressed in the correlation network owing to the large number of variables evaluated. The proximity between the nodes (traces) was proportional to the absolute value of the correlation between these nodes. The thickness of the edges was controlled by applying a cutoff value of 0.60. Finally, the positive correlations were highlighted in green, whereas the negative correlations were represented in red.

Paddy drying
The results of the temperature and relative humidity of the ambient air, as well as the temperature and air velocity at the entrance of the dryer and the exhaust during drying are shown in Figure 2. Throughout the drying process, oscillations were observed in the drying temperatures and air velocities (Figure 2A and  2B), effecting the quality of the rice. Even with high values, the tempering ratio (1:1) ensured the grain mass temperature was below the drying air temperature. In addition, the moisture removal rate increased during drying, thus reducing the drying time. Thus, at the end of the drying process the grain mass reached the equilibrium moisture content, [9] ceasing the mass transfer between the grains and the drying air. [10] According to the study performed by Meneghetti et al. [11] , the tempering ratio modifies the operating time and hourly rate of drying of paddy. Scariot et al. [1] reported that the (1:1) tempering rate and the temperatures of 55 and 65 C decreased the whole grain yield and lipid contents but increased the ash contents. Bertotto et al. [12] evaluated the initial moisture content (15, 17, and 19%), drying air temperature (40, 50, and 60 C), and tempering drying time (40, 80, and 120 min) in rice grain yield. They concluded that to achieve a yield above 58% (required by the industry), drying must be performed at temperatures close to 40 C, with a rest period of 40 min and an initial moisture content of 19%. Thus, we observed that the tempering and drying time relationship, regardless of the drying air temperature and initial moisture content, influenced the quality of the grains, following the results obtained in the study.
In this study, the high drying speed caused the moisture migration to suffer the effects of internal compression and surface traction owing to the high moisture gradient between the interior and the grain surface. The reduction in grain moisture content ( Figure 2C) was not similar between drying and tempering steps. The grains moisture content decreased from 21.7 to 10.9% with variation in grain mass temperature from 29 to 49 C ( Figure 2D). In the drying chamber, the moisture content of the grains was reduced, and the grain mass temperature increased over time. However, in the tempering chamber the uniformity was observed in the process of moisture migration from the inner layers to the grain surface. Meantime, the recycle of the grain mass in the drying and tempering chambers caused heat accumulation and increase grain mass temperature. The moisture directly influenced the temperature variation because its specific heat was higher than that of the grain. Thus, the grain temperature increased more quickly for the same amount of heat supplied.

Physical quality of polished rice
The drying stage and time were significant at 1 and 5% probability on the evaluated physical quality of the polished rice (Table 2). Figure 3 shows the physical quality results of the polished rice as a function of drying stages and time ( Figure 3A-F). During drying, the physical quality was altered through the occurrence of cracks in the seed coat. The whole grain yield was lower at the beginning of drying, compared with to the end ( Figure 3C and 3D), with variations from 33.19 to 62.57%. Drying caused cracks in the endosperm of the grains, whereas tempering drying, in both cases, did not attenuate the effects of drying temperature. [10,13] A whole grain yield above 60% was obtained up to a drying time of 11 h. Elevations in the drying air temperature can cause starch breakdown and changes in composition. Donlao et al. [14] investigated the effects of drying in the with hot air at 40 C, 65 C, 90 C and 115 C on the total starch content. They concluded that increasing the drying temperature increases the concentration of starch in the grains, which can be explained by the partial gelatinization and retrogradation of amylose.    Table 2 presents the results of the analysis of variance of the physical classification of the polished rice submitted to drying and tempering stages and times, according to describe in the methodology. The drying time was significant for the parameters evaluated, whereas the drying step was not significant for the moldy grain, chopped, and broken grains. Grain drying was beneficial, with a reduction in moisture content and an increase in final weight and grain without any type of physical defect ( Figure 4A). At 10 h, the quality of the grains obtained the best results in healthy grains ( Figure 4B). In 8 h of drying, there was a complete reduction in the percentage of moldy grains ( Figure 4D). By contrast, the percentage of broken grains ( Figure 4C) and chopped and stained grains ( Figure 4F) were significant up to 8 h of drying time, followed by a decrease in the percentages thereafter, although the coefficients of determination were low.
The number of plastered grains ( Figure 4E) constantly increased throughout the drying process, starting after 2.5 h of drying. The physical quality of the rice sampled in the tempering stage was slightly higher than that in the drying stage; however, drying time was shown to be the most significant influencing factor of physical quality. [15] Iguaz et al. [16] verified that tempering drying affected the quality of rice with an increase in drying air temperature. They also found that quenching decreases the proportion of cracked grains, thus improving grain quality. However, the effect of the number of drying cycles also influenced the cracking of the grains. Table 2 presents the significant results at 1 and 5% probability of the variables of the physical-chemical quality of the polished rice as a function of the stage and drying time. The regression results of the physicochemical variables are shown in Figure 5. A reduction in moisture contents of the grains close to 12% (w.b.) was achieved ( Figure 5A) between 8 and 9 h of drying. Under these conditions, the best results were achieved for CP, ST, CF, Fat ( Figure 5B-E, respectively), while, AS results were not significant ( Figure  5F). The grains sampled in the drying stage presented results slightly superior to the tempering stage. However, the drying time was the most influential factor. The results indicated that the tempering, despite attenuating drying effects, increased the process completion time. The increase in drying time with the resting stage of the grains in the tempering chamber resulted in the accumulation of heat in the grain mass, thereby raising the temperature of the grains, which may have been critical for the physical and physicochemical quality losses of the polished rice. Although there were differences in rice quality as a function of the different degrees of polishing [17][18][19] , the effects of paddy drying and intermittency further influenced the quality of polished rice. Figure 6 shows the heat map demonstrating the relationship between the factors and variables analyzed for the quality of polished rice. Evidently, the treatment T1 (step of tempering and drying time of 1 h) was the one that provided the lowest averages for all the evaluated variables, constituting an isolated group from the others. Group 2 allocated treatments associated with the lowest drying time, and it stood out for the highest averages of moldy grains, fragmented grains, and broken grains in the milling. Group 3 gathered most of the treatments evaluated, particularly those related to longer drying times. This group was characterized by the highest averages for the variables, MC, TW, HEA, YD, PLA, BG, CG, AS, CP, CF, ST, FAT. Thus, it was verified that the initial  Similarly, a positive correlation was observed between the variables BG, CG, and FG (located on the lower right side). The MO variable was negatively correlated (thick red lines) with most of the variables in this group, being the main physical variable indicating quality in the initial drying stage.

X-ray diffractogram analysis
The XRD patterns and the relative crystallinity obtained from the polished rice samples are shown in Figure 8. The diffractogram patterns of both samples indicated a semi-crystalline structure. Regardless of the drying time, the diffraction peaks observed in 2theta presented values of 15 , 17 , 18 , 20 , and 23 , thus indicating that the rice starch granule exhibited a type A pattern, crystalline structure of starch, typically found in cereals. Timm et al. [20] reported a reduction in rice crystallinity with increasing drying temperature, and they attributed it to the decrease in amylopectin chains associated with the use of intense heat treatments or accelerated removal of moisture, thus resulting in starch retrogradation. Other studies found the presence of the same peaks in the diffractograms obtained from rice genotypes. Ramos et al. [21] for rice grains with red pericarp and Ziegler et al. [22] for black and red rice. The relative crystallinity values were similar; however, despite the peaks in the same regions, lower values of 19.59 and 25.2% were found in tempering drying by Ramos et al. [21] and Ziegler et al., [22] respectively. The changes in the structure of the amylopectin branch clusters altered the crystallinity of the starch and its degree of granule expansion. [23,24] Scanning electron microscopy analysis The SEM images of the polished rice at the initial, intermediate, and final drying times are shown in Figure 9. In the micrographs, the damage to the rice starch granules on the surface of the samples, corresponding to the initial and intermediate times, exhibited a rough surface and an altered structure. However, the final sample exhibited pericarp ruptures and deeper deformations in the grain. This can be attributed to the drying time and the exposure of the sample to the highest drying temperature.
The physicochemical quality was related predominantly to the contents of starch, protein, lipids, and ash. Thus, the effects of heat treatments on rice grains and starch were also verified by the SEM analysis ( Figure 9). Iguaz et al. [16] found that tempering drying altered starch quality. The effect of the number of cycles of the grain mass in the dryer influenced the structure of the starch. As the temperature increased, the drying air evaporation capacity and the cracks in the grains increased.
Aquerreta et al. [25] studied the effect of the drying cycle at different tempering periods on cracking and whole-grain rice yield. They concluded that the percentage of cracked grains decreases when drying is performed in up to three stages. Jiazheng et al. [26] studied models of drying and stress formation. They reported that during drying, two stress zones are formed in the grain; the grain surface receiving a tensile stress of 7 MPa at the end of drying and the center receiving a compression of 3.2 MPa simultaneously are responsible for crack formation. Donlao et al. [14] observed through microscopic images that the samples subjected to drying at high temperatures (115 C) had a greater occurrence of cracks perpendicular to the axis of the rice grains.
In addition to the yield, another associated defect was the gypsum of the grains, which made the rice endosperm opaque. The presence of gypsum in the grain affected the appearance quality of the processed rice, whereas air spaces between the starch granules in the plastered endosperms decreased the grain density and caused the grains to break. The drying caused a molecular breakdown between the starch and the proteins, thereby forming spaces between the molecules filled by air. [27] The plastered grains tended to break more easily due to their rougher texture. [18] According to Yuliang et al. [28] rice grains have a lower amylose content when the endosperm is more opaque, mainly because there are cavities in the rice starch granule. According to the authors, the amylose  content has a negative correlation with the level of transparency of the rice endosperm.
The chemical composition of micronutrients obtained through energy scattering X-ray spectroscopy is shown in Figure 10. Evidently, the percentages of micronutrients changed as the drying time increased. Associated with the drying time, the effects caused by milling, primarily due to polishing, on the physicochemical quality of the grains were observed. [29] Polishing the rice removed the layers where the bran and germ are, which are rich in protein, fiber, and fat, changing the appearance of the rice. [30] Thus, the polishing process provided reduced nutritional properties ( Figure 10).
Energy scattering X-ray spectroscopy revealed the presence of oxygen, potassium, and carbon in the composition of all treatments. The image revealed intense energy spikes attributed to carbon and oxygen energy transitions. Carbon and oxygen atoms are abundantly found in structures, such as cellulose, hemicellulose, and lignin. The presence of silica, magnesium, and phosphorus ( Figure 10B) was verified in the composition of the samples from the initial time ( Figure 10A). The milling of polished rice is responsible for the considerable loss of micro and macronutrients from rice as the majority of minerals concentrated in the aleurone layer were partially removed during the process. [31] Tong et al. [30] observed a considerable loss of iron and zinc owing to grain milling. Liu et al. [32] observed the loss of selenium, magnesium, manganese, and lead.

Conclusions
Tempering drying attenuated the effects of drying air temperature on rice quality. However, the tempering step, being the determining factor for heating the grain mass and the changes in the quality of grains, was responsible for extending the drying time of the grains. The cumulative effects of the drying-tempering time were more decisive than the drying temperature and influenced the physical, physicochemical, and morphological qualities of the polished rice. In conclusion, rice drying technologies with tempering periods should not be designed and dimensioned for temperatures not exceeding 75 C and drying times not longer than 8 h for good system performance and grain quality.