The effect of harvest conditions and drying temperature on drying kinetics of two popcorn genotypes

The present work aimed to assess the effect of the initial moisture content of grains (at harvest), velocity of the harvester cylinder and air temperature on the diffusion coefficient, as well as to achieve the activation energy of the drying process of two popcorn cultivars. Grains of popcorn from the cultivars Zélia and CMS 43 were mechanically and manually harvested and thrashed. The rotations of the threshing cylinder were fixed at 500, 600 and 700 rpm, with initial moisture contents of 0.235 and 0.175 db. They were dried with drying air temperatures of 40, 50 and 60oC, until reaching the final moisture content of 0.137 db. The diffusion coefficient increased with increased air temperature, presenting average values from 6.345 × 10 to 3.075 × 10 ms for the temperature range studied, depending on the initial moisture content, cultivar and level of mechanical damage. The cultivar CMS 43 (C2) is more resistant to the mechanical damage caused by the threshing cylinder, and the structural changes caused by it increased the diffusion coefficient. The relation between diffusion coefficient and temperature can be described by the Arrhenius expression, which presents activation energy for liquid diffusion in the grains of popcorn from 1.50 to 45.92 kJ mol.


INTRODUCTION
Popcorn cultivation is increasing throughout the years in several countries, including Brazil, due to the consumption increment of popcorn, especially after microwave use.Furthermore, this culture has some favorable aspects when compared to conventional corn grain, such as: High mechanization potential at all production processes; lower susceptibility to plague and diseases during plant development.Besides, its commercial value is, at least, two times superior to traditional corn grain.
Moisture content reduction is essential in order to diminish the material biological activity.In addition, chemical and physical occurs due to drying procedure, which consequently affects storage of the product.The moisture content reduction involves heat and mass transfer processes, which can substantially change the grain's quality.
Water transport in agricultural products demands a 14.90 mm Exit opening between cylinder and concave (cultivar Zélia) 7.65 mm Entrance opening between cylinder and concave (cultivar CMS 43) 17.08 mm Exit opening between cylinder and concave (cultivar CMS 43) 8.89 mm driving force, either by adsorption or desorption, that is, a concentration gradient between their surface and the inner part (Botelho, 2009).Although many mechanisms have been proposed in the available literature, the primary means of water transportation is liquid diffusion (Fortes and Okos, 1980;Geankoplis, 1983).Diffusion occurs in solids with thin structure and in capillaries, pores and small holes full of steam.However, the diffusion theory disregards shrinkage, hardening of the shell and sorption isotherms (Barbosa-Cánovas and Veja-Mercado, 2000).
The liquid diffusion mechanism is very complex due to the diversity of the chemical composition and physical structure of products.The data available in literature presented very different values, due to the complexity of the products, and the function of the different estimation methods, type of material, moisture content, drying process and methodology used.
The effective diffusion coefficient is an important property in the transportation of water.Its dependence on temperature is frequently reported in classical literature (Gely and Giner, 2007;Addo et al., 2006;Goneli et al., 2007).According to Giner and Mascheroni (2002), the value of this coefficient also depends on the initial moisture content.Besides, Botelho (2009) considers that the physical integrity of a product is another fact that significantly affects this value.
Mechanical damage in grains and seeds during harvest, thrashing and processing are extremely harmful to their quality, causing breakage, cracks, cuts and abrasion, which may reduce their physiological quality (Araújo et al., 2002).In addition, Goneli et al. (2005) found that the effect of mechanical damage on the physiological quality of seeds of popcorn during storage depends on the level of damage caused by mechanical impact.
The present work aims to aid the producers and companies related to popcorn agribusiness to perform harvest decision due to the impact of different procedures during harvest and the consequences for popcorn quality.Therefore, investigation of the effect of the initial moisture contents of the grains (at harvest), velocity of the harvester cylinder and drying air temperature on the diffusion coefficient, as well the activation energy of the drying process in two popcorn cultivars were accomplished.

MATERIALS AND METHODS
Grains of popcorn from the cultivars Zélia and CMS 43 were used in this work.The planting was performed in the experimental field of the EMBRAPA-Centro Nacional de Pesquisa de Milho e Sorgo, located in Sete Lagoas-MG.The harvest was carried out when maize reached the moisture contents of 0.235 and 0.175 db.The hot air drying method at 105 ± 3ºC for 24 h was employed to determine the moisture content (Brazil, 1992).
Harvest and thrashing were performed in two ways: Manually and with the aid of a harvester machine (pull-type harvester, model "Double Master", axial flow system), with harvester cylinder rotations fixed at 500, 600 and 700 rpm.Internal characteristics of harvesting machine are presented at Table 1.
After thrashing, the samples were manually cleaned and then placed into two low-density polyethylene bags, properly sealed and identified, which were put into a refrigerator at 4ºC for 24 h, to maintain the initial moisture content of the harvest.Then, they were transferred in styrofoam boxes to the Laboratory of Pre-Processing of Agricultural Products of the Department of Agricultural Engineering of the Universidade Federal de Viçosa, located in Viçosa-MG, where they were stored in a refrigerator at 4ºC and 80 ± 5% of RH, until the beginning of the experiment.
Drying was carried out with the aid of a laboratory dryer (Ethik Technology, model 440/DE), in thin layer, with controlled temperature and air flow.Drying air velocity was maintained constant at 1 m s -1 and monitored by a digital anemometer with rotating blades.The air flow was provided by an axial fan that conducted the air to the plenum and from there to three individualized chambers with removable trays with screened bottom.Drying air temperatures of 40, 50 and 60ºC were used and measured by a mercury thermometer installed just below the layer of grains.The temperature and air relative humidity were recorded, with the use of a thermohygrograph (RENE GRAF).It was determined that drying would finish when the samples reached the pre-established moisture content of 0.137 db.A semi-analytical balance (Gehaka,model BK 400,Brazil) with precision of ± 0.01 g was used to monitor drying, and the samples were weighed at every five-minute interval by retrieving the drying trays from the dryer.Weighting procedure lasted about 30 s.
Thin layer drying is defined as the one with thickness of only one grain.The thin layer drying equation, added to representative equations of other physical properties specific to the product under study, forms a set of mathematical relations that help in the calculations and understanding of the thick layer drying process.It is considered that a thick layer of grains is composed of successive overlapped thin layers.For Chittenden and Hustrulid (1966), the theory used to describe the drying phenomenon may be based on the principle that resistance to moisture transport is concentrated in the surface of grains, according to the following differential equation: ( ) Where U and U e are the moisture content of the product for a certain time and the equilibrium moisture content (kg/kg), respectively; K is the drying constant (s -1 ); t is the drying time (s).
By integrating this equation between the U 0 limits in the beginning of the process, and U t at any time of drying, t, we have: Where RU is the humidity ratio, dimensionless; U 0 is the moisture content of the product at time zero (kg/kg); t is the drying time (s).
The equilibrium moisture content of popcorn, for the conditions of temperature and drying air relative humidity, was calculated by the Sigma-Copace equation (Corrêa et al., 1998).
Where T is the drying air temperature, ºC; UR is the drying air relative humidity, decimal; a, b and c are the specific parameter of the product (a = 1.2658, b = 0.0053 and c = 0.7879).
According to Brooker et al. (1992), in the liquid diffusion theory, the second Law of Fick has been used to establish water diffusion according to the concentration gradient.
Where D ef is the liquid diffusion coefficient, m 2 s -1 ; x is the distance from a reference point, m.
The variation of the moisture content according to the time of drying, for homogenous materials with constant diffusion coefficients, is represented by the following equation: Where r is the radial distance or thickness, m; c is 0 for plane bodies, 1 for cylindrical bodies and 2 for spherical bodies.Brooker et al. (1992) present the following analytical solution for the spherical geometric shape: Where n is the number of terms; R is the equivalent radius, m.
The diffusion coefficient was achieved by the Equation ( 6).The analytical solution of this equation is presented as an infinite series and the finite number of terms (n) in the truncation can determine the accuracy of the results.
The equivalent radius, used in the liquid diffusion model, is defined as the radius of a sphere with the same volume as the grain.It is determined by the measurement of the three orthogonal axes (length, width and thickness), as proposed by Mohsenin (1986), in fifteen grains, with the aid of a digital caliper.The volume of each grain, considered a spheroid, was achieved by the expression: Where a is the largest grain axis, mm; b is the average grain axis, mm; c is the smallest grain axis.
The dependence of effective moisture diffusivity (D ef ) on drying temperature has been shown to follow an Arrhenius relationship (Bai et al., 2013;Xiao et al., 2012) as shown in Equation ( 8).
Where D 0 is the pre-exponential factor, m 2 s -1 ; E a is the activation energy, kJ mol -1 ; R is the universal gas constant, kJ mol -1 K -1 ; T abs is the absolute temperature, K.

RESULTS AND DISCUSSION
Table 2 shows the values of the diffusion coefficient, using the initial equivalent radius of popcorn, related to the cultivars, initial moisture contents of the product at harvest, velocity of the harvester cylinder and drying air temperature.
The analysis of the results demonstrates that the diffusion coefficient increased with increased drying air temperature.It agrees with several researchers (Goneli et al., 2007;Côrrea et al., 2006a;Campos et al., 2009) and suggests that, for lower temperatures, popcorn grains offer more internal resistance to water transport, which leads to lower diffusion coefficients.Thus, increased drying air temperature indicates higher intensity of the phenomenon of water transport from the inner part to the periphery of the grain.Campos et al. (2009) affirm that water viscosity decreases with increased temperature.Since it is a measure of resistance of the fluid to flow, variations of this property result in changes in water diffusion in the capillaries of grains, thus favoring the movement of this fluid in the product.Similarly, the level of molecular vibrations of water molecules is also intensified, which contributes for a more efficient diffusion.
The values of diffusion coefficient ranged from 0.30756 to 63.45711 × 10 -12 m 2 s -1 .Xiao et al. (2010a) found this parameter values from 1.82 to 5.84 × 10 −10 m s −2 for grapes whilst Xiao et al. (2010b) reported values from 0.265 to 1.052 × 10 −9 m 2 s -1 for carrot cubes.These different values can be explained by differences in the chemical composition, physical structure, and geometry of the products, in addition of different drying methods.
It can be observed that, for the initial moisture content (U1), the diffusion was higher than (U2), since the higher the water concentration gradient between the surface and the inner part of a product, the higher the diffusion.This fact was also observed by Giner and Mascheroni (2002), who reported increased diffusion coefficient for initial moisture content values of 0.2694; 0.2396; 0.2133 and 0.1891 db for wheat grains.Besides, on average, the cultivar CMS 43 (C2) is the most resistant to the mechanical damage of the cylinder, because it presented lower values for diffusion coefficient.Higher values of diffusion coefficient were found for elevated thrashing velocity (higher mechanical damage), on average, when compared to the values of this parameter for grains with lower thrashing velocity (lower mechanical damage).The same phenomenon was observed by Botelho (2009), while studying water absorption by maize with different levels of mechanical damage, and by Campos et al. (2009), when they assessed coffee grain drying at different stages of wet processing.However, it did not occur for the combination C2U1M because a preferential way for gas exchange with the environment may have been formed, while mechanical damage caused changes in the intracellular configuration, bringing together the cells previously apart from each other.
As an example, Figure 1 presents the graphic of correspondence among the values observed and estimated according to humidity for the combination C1U1V1.This figure shows the appropriate adjustment of the moisture diffusion model for the kinetic description of popcorn grain drying.
The dependence of the diffusion coefficient on drying air temperature has been satisfactorily described by the Arrhenius equation.Figure 2 shows the Arrhenius representation for the diffusion coefficients according to the drying air temperatures for the combination C1U1V1.
The slope of the curve of the Arrhenius representation provides the E a /R relation, while its intersection with the Y-axis indicates the value of D 0 .According to Sharma and Prasad (2004), thermodynamically, the activation energy represents the energy necessary for the disruption of the barrier that the water molecules face during the drying process, when they migrate from the interior to the surface of the product, and the lower it is, the higher the water diffusivity in the product.Reduced activation energy of a process causes increased average energy of the  molecules that compose it.The values of activation energy and pre-exponential factor found in this work (Table 3) corroborate those found in literature (Campos et al., 2009).
It is observed that, on average, for the cultivar Zélia (C1), the higher the moisture content, the lower the activation energy needed, as found by Xiao et al. (2012).However, for the cultivar CMS 43 (C2), the same trend is not observed, since it presents higher resistance to the impacts and abrasive mechanical efforts that generate fissures and cracks, caused by the thrasher cylinder, which overlaps the moisture content factor.
Values of activation energy ranged from 1.50 to 45.92 kJ mol -1 with median value of 26.79 kJ mol -1 .Researchers reported different values of activation energy, such as 67.29 and 20.17 kJ mol -1 , respectively by Xiao et al. (2010a) and Xiao et al. (2010b).This trend indicates the need of the study of different agricultural products, due to variety of physical and chemical properties, which interferes at the drying process.

Conclusions
1.The liquid diffusion model satisfactorily represents the kinetics of the drying of popcorn grains, for the experimental conditions.2. The diffusion coefficient increases with increased air temperature, presenting average values from 6.345 × 10 - 11 to 3.075 × 10 -13 m 2 s -1 for the temperature range studied, and it depended on the initial moisture content, cultivar and level of mechanical damage.
3. The cultivar CMS 43 (C2) is more resistant than the cultivar Zélia to the mechanical damage caused by the thrasher cylinder.4. The structural changes caused by the thrasher cylinder led to increased diffusion coefficient, mainly due to the formation of fissures and cracks which are preferential paths to water diffusion.5.The relation between diffusion coefficient and temperature can be described by the Arrhenius expression, which presents activation energy for liquid diffusion of 1.50 to 45.92 kJ mol -1 in popcorn grains.

Figure 1 .
Figure 1.Observed and estimated values of humidity for the combination C1U1V1.

Figure 2 .
Figure 2. Arrhenius representation for the diffusion coefficients according to the drying air temperatures for the combination C1U1V1.

Table 1 .
Harvesting machine characteristics for each cultivar.

Table 2 .
Diffusion coefficient using the initial equivalent radius of two popcorn cultivars, Zélia and CMS 43, moisture content of the product at harvest, velocity of the harvester cylinder and drying air temperature.

Table 3 .
Calculated values of the activation energy and pre-exponential factor of water diffusion in popcorn grains for the temperature range under study.C1 and C2: Cultivars Zélia and CMS 43, respectively; U1 and U2: Initial moisture content of the product at harvest, 0.235 and 0.175 (d.b.), respectively; M: Manual harvest and thrashing; V1, V2 and V3: Mechanical harvest and thrashing with thrasher cylinder velocity of 500, 600 and 700 rpm, respectively. *