Modelling of drying kinetics and comparison of two processes: Forced convection drying and microwave drying of celery leaves (Apium graveolens L.)

The Annals of the University Dunarea de Jos of Galati Fascicle VI – Food Technology (2019), 43(2), 48-69 https://doi.org/10.35219/foodtechnology.2019.2.04 ORIGINAL RESEARCH PAPER MODELLING OF DRYING KINETICS AND COMPARISON OF TWO PROCESSES: FORCED CONVECTION DRYING AND MICROWAVE DRYING OF CELERY LEAVES (Apium graveolens L.) KHOKHA MOUHOUBI1, LILA BOULEKBACHE-MAKHLOUF1*, NAIMA GUENDOUZEBOUCHEFA1, MOHAMED LAMINE FREIDJA1,2, ALBERTO ROMERO3, KHODIR MADANI1,4 1Laboratoire de Biomathématiques, Biophysique, Biochimie et Scientométrie, Faculté des Sciences de la Nature et de la Vie, Université de Bejaia, 06000 Bejaia, Algérie 2Department of Biochemistry and Microbiology, Faculty of Sciences, University of M'sila , 28000 M'sila, Algeria 3Department of Chemical Engineering, Universidad de Sevilla, Sevilla, Spain 4Centre de recherche en technologie agro-alimentaire, Route de targua-ouzemour, 06000, Bejaia, Algérie *Corresponding author: [email protected] Received on 25th April 2019 Revised on 26th August 2019 The purpose of this work is to compare two processes: forced convection drying and microwave drying of celery leaves (Apium graveolens L.). This comparison is based on kinetical parameters, moisture diffusivity, variation of the drying rate and energy consumption calculation of both processes. The drying experiments were carried out at different air temperatures (40-120 °C) and at different microwave powers (100-1000 W).Twenty-two empirical models were used to simulate the thinlayer drying kinetics of celery leaves and the best models were selected using three statistical criteria (R2, χ2 and RMSE). Sledz’s model proved to be the best for celery leaves drying kinetics description with 0.9962 ≤ R2 ≤ 0.9992, 0.000065 ≤ χ2 ≤ 0.000284 and 0.007979 ≤ RMSE ≤ 0.016683 for all the studied temperatures and 0.9971 ≤ R2 ≤ 0.9989, 0.000124 ≤ χ2 ≤ 0.000291 and 0.010910 ≤ RMSE ≤ 0.016914 for all the used powers. Moisture effective diffusivity ranges from 2.22×10-12 to 6.40×10-11 for convective drying and from 1.18×10-11 to 3.13×10-10 m2/s for microwave drying. While in the same order, the activation energies were 36.09 kJ/mol and 77.3 W/g. Regarding the energy consumption, the Specific Electrical Energy Consumption decreased with decreasing temperature or power levels, whereas the opposite was observed with Energy Efficiency. It is clear that many advantages are attributed to microwave drying, including reduced drying time, high drying rate and high moisture diffusivity, low energy consumption and significant drying efficiency, especially when power levels are high. Keywords: Apium graveolens L., drying kinetics, simulation, moisture diffusivity, activation energy, energy consumption, drying rate Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 49 Introduction Aromatic plants represent an important resource for human food, due to their valuable properties including medicinal benefits (Soran et al., 2014). They find their applications in various fields: food, cosmetic and medical fields. Celery (Apium graveolens L.) is one of these plants; it belongs to Umbelliferae family. Endemic to Southern Europe, it is now grown worldwide for human consumption (Kurobayashi et al., 2014). It is a good source of vitamins, minerals and many active ingredients (Madamba and Liboon, 2007), and it is considered among the most desired products due to its taste, texture and low calorie content (González-Buesa et al., 2014). However, one of the main quality attributes of celery is it’s crispy nature which depend on degree of lignification of fibrous tissue, dehydration and the degree of alteration of pectic substances (Viña and Chaves, 2003). Therefore, celery is usually dried and used as a soup enhancer because of its characteristic odor, to enhance the taste and aroma of the stew while having therapeutic properties (Demirhan and Özbek, 2011). Drying is one of the oldest methods of food preservation as well as an important food processing stage (Kayisoglu and Ertekin, 2011). It is also a common industrial preservation method applicable to a wide range of agricultural and industrial products (Zhang et al., 2016). Defined as the removal of moisture from a product by heat, that confer to a product an acceptable moisture level in order to prevent marketing deterioration within a certain period of time, safe storage, processing, or transportation (Nwakuba et al., 2017). It aims to prevent microbial spoilage and chemical alterations thus prolonging shelf-life while realizing space and weight saving (Haksever and Moralar, 2017). Convective drying is the most universal dehydration method due to its simplicity and easiness of control. However, a long convective process is highly energy consuming, which leads to cell structure collapse during drying and results in inferior quality of dried herbs (Sledz et al., 2017). That is why, the use of microwave rays in the drying of products has become widespread because it minimizes the decline in quality and provides rapid and effective heat distribution in the material as well. Furthermore, high quality product is obtained via microwave drying in addition to the decline in drying period and energy conservation during drying (Alibas and Kacar, 2016). Given the very large variability and diversity of food and biological products, the best way to characterize the drying behavior of a product is to experimentally measure its drying kinetics (Bonazzi et al., 2008) which are the most important data required for the design and simulation of dryers (Ratti, 2008). Simulation or mathematical modelling of the drying process and equipment is an important aspect of drying technology in post-harvest processing of agricultural materials (Harish et al., 2014). Numerous models are created to describe the history of the product during the drying process, using parameters such as time, temperature, water content and quality index (Bonazzi and Dumoulin, 2011). The purpose of modelling is to allow the engineers to choose the most appropriate method of drying Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 50 for a given product as well as to choose suitable operating conditions. Full-scale experimentation for different products and system configurations is sometimes costly and not possible. Therefore, the prediction of drying kinetics of specific crops under various conditions is very useful in the design and optimization of dryers. The use of a simulation model is also a valuable tool for the prediction of drying systems performance (Khazaei and Daneshmandi, 2007). Previous studies on the modelling of the microwave drying process of celery leaves were reported by Demirhan and Özbek (2011) and Alibas (2014). However, these authors have not studied the information on energy consumption that allows the identification of the best drying process in terms of energy efficiency. Such information will make it easier for the industry to effectively manage drying techniques and thus avoid energy misuse. Moreover, no comparative study between convection drying and microwave drying has been performed on celery leaves, which is the case of the present study whose objectives were: (i) to study the kinetics of celery leaves drying by two methods (forced convection drying and microwave drying), (ii) designate the most appropriate kinetics model, (iii) calculate effective diffusivities, (iv) evaluate the evolution of drying rate kinetics and its variation depending on the moisture content, and finally (v) determine the activation and consumption energies. Materials and methods Materials Fresh celery (Apium graveolens L.) was purchased from a local market (Bejaia- Algeria) in November 2016. The leaves were separated from the stems. Before drying experiments, their initial moisture content was determined according to the protocol of Boulekbache-Makhlouf et al. (2013) and was 88.16 ±1.14%, their thickness was measured with a Vernier calliper, and the average was found to be 0.4 ±0.07 mm. Drying equipment and drying procedure The celery-leaf drying experiments were carried out using two different techniques: hot air drying and microwave drying. Hot air drying (40, 60, 80, 100 and 120 °C) was carried out using a ventilated oven (Memmert, Germany). The microwave drying process (100, 200, 300, 400, 500, 600, 700, 800, 900 and 1000 W) was carried out using a domestic microwave (Maxipower, Model MASMO23S, China). During the drying experiments, 10 g of celery leaves were uniformly spread as a thin layer and the moisture losses were periodically measured using an external analytical accuracy of 0.01 g (RADWAG, WPS 600/C/2, Poland). Three repetitions were performed for each temperature and power, and the data provided were an average of these results. The drying process was applied until the weight of the sample was constant, followed by the determination of the final moisture content of the powders, which is estimated at 7.96 ± 0.11%. Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 51 Mathematical modelling of drying curves Various mathematical models were reported to study the modelling of thin-layer drying kinetics. In this study, twenty-two (22) different expressions were examined (Table 1). Table 1. Mathematical models given by various authors for drying curves Name Mathematical equation References Newton MR=exp(-kt) Maskan (2000) Henderson and Pabis MR=a.exp(-kt) Nema et al. (2013) Logarithmic MR=a.exp(-kt)+c Nema et al. (2013) Page MR=exp(-ktn) Nema et al. (2013) Modified Page 1 MR=exp(-(kt)n) Nema et al. (2013) Modified Page 2 MR=exp(-kt)n Demirhan and ÖZbek (2009) Midilli et al. MR=a.exp(-ktn)+bt Nema et al. (2013) Two terms MR=a.exp(-kt)+b.exp(-k1t) Nema et al. (2013) Two terms exponential MR=a.exp(-kt )+(1-a).exp(-kat) Nema et al. (2013) Approximation of diffusion MR=a.exp(-kt )+(1-a).exp(-kbt) Nema et al. (2013) Verma et al. MR=a.exp(-kt )+(1-a).exp(-k1t) Nema et al. (2013) Modified Henderson and Pabis MR=a.exp(-kt)+b.exp(- k1t)+c.exp(-k2t) Vega-Gálvez et al. (2014) Parabolic MR=a+b.t+c.t2 Taghian Dinani et al. (2014) Wang and Singh MR=1+a.t+b.t2 Nema et al. (2013) Chavez-Mendez et al. MR=(1-(1-L2)L1t)1/(1-L2) Nema et al. (2013) Logistic MR=b/(1+a.exp(k.t)) Sledz et al. (2016) Sledz et al. MR=b.exp(-kt)/(1+a.exp(k1.t)) Sledz et al. (2016) Simplified Fick’s diffusion equation MR=a.exp(-k(t/L2)) Aghbashlo et al. (2009) Weibull MR=exp(-(t/a)b) Aghbashlo et al. (2009) Demir et al. MR=a.exp(-kt)n+b Amiri Chayjan and Shadidi (2014) Taghian Dinani et al. MR=a.exp(-((t-b)/a)2) Taghian Dinani et al. (2014) Fernando and Amarasinghe MR=(1+a.t+b.t2)/(1+c.t) Fernando and Amarasinghe (2016) Note: k, k1, k2 – drying coefficients (1/min); a, b, c, L1, L2 − coefficients of the equations; n − exponent; t– time (min); L – half of thickness (m). These models estimate that the drying curves respond to an equation of the moisture ratio as a function of drying time. A dimensionless moisture ratio (MR) was calculated using the following equation: 0 t e e M M MR M M   (1) Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 52 where: MR is the moisture ratio, Mt is the moisture content at a specific time t, and M0 and Me are the initial and equilibrium moisture content, respectively. However, Me is relatively small compared with Mt or M0, hence the error involved in the simplification by assuming that Me is equal to zero is negligible (H. Darvishi et al., 2016). For this purpose, the moisture ratio can be written in a more simplified form, as follows: 0 t M MR M  (2) Drying rate The drying rate (DR) was calculated using equation 3: t dt t M M DR dt    (3) where DR is the drying rate (kg water/kg dry matter (dm) min), Mt+dt and Mt are moisture content at t+dt (kg water/kg dm) and t (kg water/kg dm), respectively, and dt is the time difference (min). The graphical plot of the drying rate with respect to time makes it possible to determine the drying behaviour (Aral and Bese, 2016). Determination of effective moisture diffusivity One of the most important parameters in drying kinetics is the effective moisture diffusivity (Deff), which describes the transport of moisture from the material to the environment during the period of falling rates (Aral and Bese, 2016; Motevali et al., 2016; Murthy and Manohar, 2014). For this purpose, the second simplified Fick's equation for diffusion was used to estimate the effective moisture diffusivity (Deff) of celery leaves during their drying, considering constant moisture diffusivity, infinite slab geometry and uniform initial moisture distribution: 2 2 2 (2 1) ( ) 4 2 2 0 8 1 (2 1) eff n D t L n MR e n        (4) where Deff is the effective moisture diffusivity (m2/s), L is the half thickness (drying from both sides) of celery leaves (L=0.0002 m), n is the positive integer and t is the drying time (s). For a long drying period, the above-mentioned equation can be simplified to the first term of the series and Eq. (4) can be expressed in a logarithmic form as Eq. (5) by taking the natural logarithm of both sides (Madhava Naidu et al., 2016; Zhang et al., 2016):   2 2 2 8 ln ln 4 eff D t MR L     (5) This is evaluated numerically for the Fourier number, F0 = Deff ·t/4L2 (Hosain Darvishi et al., 2014). To this end, Eq. (5) can be rewritten as:   2 0 2 8 ln ln MR F     (6) Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 53 Thus:   0 0.101ln 0.0213 F MR   (7) The effective moisture diffusivity was calculated using Eq. (8):   0 2 2 0.101ln 0.0213 4 4 eff MR F Dt t L L                 (8) Activation energy Activation energy is often described using a simple Arrhenius exponential relationship to study the effect of temperature. For this purpose, in the case of hot air drying: equation 9 (Eq. (9)) linking temperature to the effective moisture diffusivity was used (Murthy and Manohar, 2014; Ben Haj Said et al., 2015). However, in the case of microwave drying, the process is not isothermal and temperature cannot be measured (Dadalı et al., 2007; Özbek and Dadali, 2007). Therefore, a modified Arrhenius equation (Eq. (10)), used in previous studies (Hosain Darvishi et al., 2014; Alibas and Kacar, 2016), was applied in this study to calculate the activation energy with STATISTICA software (v.8.0). 0 a E RT eff D D e         (9) 0' a E m P eff D D e         (10) where Deff is the effective diffusivity (m2/s), D0 and D0’ are the pre-exponential factor (m2/s) in oven and microwave, respectively, R is the universal gas constant (8.314 kJ/mol K), T is the temperature (K), m is the mass of raw sample (g), P is the power (W) and 𝐸 is the activation energy expressed in (kJ/mol) for hot air drying and in (W/g) for microwave drying. Energy consumption Energy consumption can be evaluated using two different efficiency indices, namely specific electrical energy consumption (Eq. (11)) or energy efficiency (Eq. (12)) (Tsotsas and Mujumdar, 2014): 3600 (X ) e S i f E SEC M X  (11) where SECe is the Specific Electrical Energy Consumption (MJ/kg water), E is the total electrical energy consumption (kWh), Xi and Xf refer to the initial and final moisture contents (dm), respectively, and Ms is the mass of dry solid matter (in kg).   .100 3600 v s i f h EE M X X E    (12) Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 54 where EE is the Electrical Energy (%) and Δhv is the evaporation enthalpy of water (2257 kJ/kg, at 100°C). Statistical analysis In order to model the experimental data of the drying kinetics, a nonlinear regression analysis was carried out using the STATISTICA software (v.8.0). Three statistical parameters, i.e., determination coefficient (R2), chi-square (χ2) and root mean square error (RMSE), were used to determine the ability of the tested mathematical models to represent the experimental data. The best model had the highest R2 value and the lowest χ2 and RMSE values (Chayjan et al., 2013). These parameters were calculated as follows: 2 , exp,i 21 ( ) N pre i i MR MR N z      (13) 2 , exp,i 1 ( ) N pre i i MR MR RMSE N    (14) where MRexp,i is the experimental moisture ratio, MRpre,i is the predicted moisture ratio, N is the number of observations and z is the number of constant parameters in the model equation. Concerning the differences between Deff, SECe and EE values, they were tested by the univariate ANOVA test. The significance of the differences between these values was determined by Tukey test at P ≤ 0.05. The calculations were performed using Minitab 17 software. Results and discussion Drying curves The variation of the moisture content as a function of the celery-leaf drying time at different temperatures (40-120 °C) and at different microwave powers (100-1000 W) are shown in Figure 1 (a) and (b), respectively. As shown in figure 1 (a) and (b), the drying process is characterized by a gradual decrease in the moisture content as drying time evolves. This decrease is fast at the beginning of the drying process and then slows down gradually as the drying process progresses. This can be explained by the fact that during the first phase of the drying process, the water content in the matrix is high, which leads to a large diffusion of moisture from the product (Alibas and Kacar, 2016). However, at an advanced drying process stage, we observed a reduction of the energy absorbed by the product surface. This is due to the lower water content of the partially dried product and to the dry surface, which prevents the penetration of heat and the migration of water (Ben Haj Said et al., 2015). Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 55 It is also noted that temperature and microwave power levels have a significant effect on the moisture content. Indeed, it is clear from the curves that the decrease in the moisture content becomes faster with the increase of both temperature and power levels. Thus, the drying time required to achieve the lowest moisture content was reduced from 338 to 10.83 min for temperatures ranging from 40 to 120 °C and from 48.5 to 2.66 min for powers ranging from 100 to 1000 W. These results have shown that an increase of temperature or power levels leads to a reduction of drying time, which are in accordance with reported observations on convective drying (Faal et al., 2015; İZlİ, 2017; Nguyen et al., 2015; Xiao et al., 2015) as well as those on microwave drying (Amiri Chayjan et al., 2015; H. Darvishi et al., 2016) of different food products other than celery. When comparing processes in terms of reduced drying time, a significant reduction has been observed when applying the microwave drying process, especially when applying high power levels; this could be explained by the fact that one of the advantages of microwave drying is a short processing time (Amiri Chayjan et al., 2015). (a) (b) Figure 1. Moisture ratio profiles of celery leaves during convective drying at various temperatures (a) and microwave drying at various power levels (b). 0 0.2 0.4 0.6 0.8 1 0 50 100 150 200 250 300 350 Moisture ratio (MR) Drying time (min) 40°C 60°C 80°C 100°C 120°C 0 0.2 0.4 0.6 0.8 1 1 10 Moisture ratio (MR) Drying time (min) 100W 200W 300W 400W 500W 600W 700W 800W 900W 1000W Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 56 Modelling of the experimental data The modelling of drying kinetics allows predicting the drying kinetics of many agricultural and food materials and to evaluate the end point of the process. Several models were reported in the literature. However, in recent years new models have been proposed by several authors, such as the models proposed by Taghian Dinani et al. (2014), Fernando and Amarasinghe (2016) and Sledz et al. (2016). To better estimate the model that interprets the kinetics of drying celery leaves, twenty-two (22) models (Table 1) were tested and the results were evaluated on the basis of the results of the following statistical parameters: coefficient of determination (R2), chisquare (χ2) and root mean square error (RMSE). Of the 22 models tested (Table 1), five models describe the drying behavior of celery leaves in the case of forced convection process (Table 2), and four models describe their behavior in the case of microwave drying (Table 3). Indeed, all these models confirmed the goodness of fit with the R2 values closest to unity and the lowest χ2 and RMSE values (Table 2 and 3). However, among these models, the highest R2 values and the lowest χ2 and RMSE values were obtained with the Sledz et al. (2016) model with 0.9962 ≤R2≤ 0.9992, 0.000065≤ χ2≤ 0.000284 and 0.007979≤ RMSE ≤0, 016683 and 0.9971≤ R2≤ 0.9989, 0.000124≤ χ2≤ 0.000291 and 0.010910≤ RMSE≤ 0.016914 for both forced convection and microwave drying, respectively. Consequently, this model was chosen as the optimal model for its satisfactory description of the drying characteristics of the celery leaves. This model was also chosen in previous studies to describe the drying behaviour of parsley leaves (Sledz et al., 2016) and basil leaves (Sledz et al., 2017). Effective moisture diffusivity The effective moisture diffusivities (Deff) were estimated using Eq. (8); the values are summarized in Table 4. These values show a variation ranging from 2.22×10-12 to 6.40×10-11 m2/s and from 1.18×10-11 to 3.13×10-10 m2/s for forced convection and microwave drying, respectively, shifting from the lowest temperatures and power levels to the highest ones. This indicates that, the higher the temperature or power levels, the higher Deff values. In the case of convective drying, the increase in Deff values in parallel with high temperature levels can be explained by the fact that the rise in temperature strongly activates water molecules, accelerating their transfer to the surface of the matrix (Amami et al., 2017; Aral and Bese, 2016). However, its increase in the case of microwave drying, is related to the energy absorbed by moisture, which increases with the increment of microwave power level (Zhu et al., 2015). The Deff values obtained in this study fall within the general range defined by Azeez et al. (2017), which is 10−8 to 10−12 m2/s for food products. These values are also included in the range of values obtained for food sheets in other studies: for example, 2.55×10-12 to 8.83×10-12 m2/s for Allium roseum leaves at temperatures ranging from 40 to 60 °C (Ben Haj Said et al., 2015), 0.2×10-12 to 9.4×10-12 m2/s for rosemary leaves at temperatures ranging from 40 to 60 °C (Bensebia and Allia, 2015), 3.982×10-11 to 2.073×10-10 m2/s for the drying of mint leaves at power levels ranging from 180 to 900 W (Özbek and Dadali, 2007), 2.168×10-10 to 7.899×10-10 m2/s for Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 63 okra (Dadalı et al., 2007), 14.67 W/g for pepper (Hosain Darvishi et al., 2014), 11.41 W/g for basil leaves (Demirhan and ÖZbek, 2009) and 24.7 W/g for the yam of elephant foot yam (Harish et al., 2014). However, these values are particularly comparable to the Ea reported by Zhu et al. (2015), which is about 77.0485 W/g for Ximeng lignite. Energy consumption The calculation results for SECe and EE are illustrated in Table 5. Table 5. Result of SECe and EE calculations T(°C) Convection drying P(W) Microwave drying SECe (MJ/kg H 2 O) EE (%) SECe (MJ/kg H 2 O) EE (%) 40°C 2.61×10+09± 9.11×10+07aA 9.13×10-05± 3.22×10-06eL 100W 3.00×10+08± 1.00×10+07aC 7.53×10-04± 2.51×10-05jJK 200W 1.02×10+08± 6.42×10+06bEF 2.21×10-03± 1.38×10-04iI 60°C 4.88×10+08± 1.25×10+07bB 4.88×10-04± 1.23×10-05dKL 300W 6.47×10+07± 8.87×10+05cEFGH 3.49×10-03± 4.79×10-05hH 400W 5.24×10+07± 2.66×10+06dEFGH 4.31×10-03± 2.18×10-04gG 80°C 1.95×10+08± 6.64×10+06cD 1.22×10-03± 4.18×10-05cJ 500W 4.16×10+07± 1.42×10+06deFGH 5.43×10-03± 1.81×10-04fF 600W 3.56×10+07± 1.52×10+06efEFGH 6.34×10-03± 2.68×10-04eE 100°C 1.15×10+08± 6.01×10+06cE 2.06×10-03± 1.04×10-04bI 700W 3.10×10+07± 8.81×10+06efgFGH 7.29×10-03± 2.11×10-04dD 800W 2.76×10+07± 1.05×10+06fghGH 8.19×10-03± 3.05×10-04cC 120°C 9.27×10+07± 2.27×10+06cEFG 2.57×10-03± 6.23×10-05aI 900W 2.25×10+07± 1.06×10+06ghGH 1.01×10-02± 4.74×10-04bB 1000W 1.72×10+07± 2.85×10+06hH 1.31×10-02± 2.15×10-04aA Table shows mean values with standard deviation of the mean. a,b - same index letters indicate that the mean values are not significantly different at a confidence level of 95 % (p ≤0.05)for the same process. A,B - same index letters indicate that the mean values are not significantly different at a confidence level of 95 % (p ≤0.05) between the two process. As it can be observed, SECe decreases when moving from a low temperature or power level to higher temperatures or power levels for convection and microwave drying, respectively. Indeed, as indicated in Table 5, SECe ranges from 9.29×10+7 to 2.61×10+9 MJ/kg in the case of convective drying and from 2.24×10+7 to 3.00×10+8 MJ/kg for microwave drying. The statistical analysis showed significant differences between SECe values obtained at 40, 60 and 80 °C. In the case of microwave drying, the use of low power levels (100-400 W) produced significant differences within the obtained SECe values. When applying powers greater than or equal to 400 W, with significant differences between SECe values, overlaps were observed for some values. It should be noted that the decline in SECe values has been reported by many authors, such as Amiri Chayjan and Shadidi (2014), who found in their study on bean Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 64 drying, that SECe values ranged from 13.36×10+6 to 64.30×10+6 kJ/kg for the temperature range of 40 to 70°C. Chayjan et al. (2013) reported in their study on squash seed values ranging from 2.303×10+6 to 0.783×10+6 kJ/kg for 50-80 °C and H. Darvishi et al. (2016) reported values of 7.79 - 16.20 MJ/kg for kiwi slices drying. Concerning EE variations (Table 5), an opposite behavior to SECe was observed, i.e., EE values increased when moving from low temperatures or power levels to high temperatures or power levels. The values obtained for this parameter range from 9.13×10-5 to 2.57×10-3, from 40 to 120 °C, and from 7.53×10-4 to 1.31×10-2 for power levels ranging from 100 to 1000 W. The statistical analysis of these results showed significant differences when moving from one level to another for both temperatures and power levels. It should be noted that the opposite trends for SECe and EE were also observed by Tsotsas and Mujumdar (2014). Statistical analysis of the SECe values obtained for both drying processes showed that the most important SECe values are obtained at 40, 60, 80 °C and 100 W which are statistically different. The SECe values of the other temperatures (100 and 120 °C) and powers between 200 and 700 W do not show any statistical difference, but they are different from those obtained at powers between 800 and 1000 W, which in turn do not show a statistical difference from the SECe values of powers between 300 and 700 W. The statistical analysis of the EE values obtained for both drying processes showed that the lowest values are induced at low temperatures (40 and 60 °C) which are not statistically different from each other and the highest values are obtained at powers ranging from 300 to 1000 W and which are also statistically different from each other. However, the absence of significant differences in EE values is observed at temperatures of 60 to 120 °C overlapping the EE values obtained at 100 and 200 W. By comparing both drying processes it appears that in general, the energy consumption (most important SECe value) was obtained for the convection drying and the energy efficiency (most important EE value) was obtained for microwave drying. This suggests that the energy consumption is proportionally related to the drying time while the efficiency is inversely related to it. Conclusions The characteristics of thin-layer drying kinetics of celery leaves were studied by two methods: convection and microwave drying. In order to simulate the experimental data of the drying kinetics, twenty-two (22) mathematical models were tested and the results allowed choosing Sledz et al. model as the best model to describe the drying behaviour of celery leaves for both methods reinforced by the highest R2 values and the lowest χ2 and RMSE values. A comparison between both methods was made on the basis of drying parameters such as drying time, variation in drying rate as a function of time and as a function of moisture content, calculation of diffusivities and energy consumption. The results concluded that increasing the drying temperature or microwave power level resulted in a reduction in drying time and that this reduction is more significant when applying microwave drying. Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 65 As far as the effective moisture diffusivity is concerned, it rises with increasing temperature or power levels and the most important diffusivities are induced when applying high levels of microwave power. The variation in drying rate with time and its variation with moisture content increases with increasing temperature or power level and the highest drying rates are also obtained with high levels of microwave power. With regard to energy consumption and drying efficiency, calculations have shown that energy consumption is low and drying efficiency is high when applying high convection temperatures or microwave power levels. The results have also shown that energy consumption is higher when applying oven drying while drying efficiency is obtained for the microwave drying process. Based on the obtained results, it can be concluded that the microwave drying technique was better for celery leaves drying, given the many advantages reported in this study, including reduced drying time, high drying rates, high diffusivities, low energy consumption and drying efficiency. Acknowledgments We wish to acknowledge the Ministry of Higher Education and Scientific Research of Algeria. References Aghbashlo, M., Kianmehr, M., Khani, S., Ghasemi, M. 2009. Mathematical modelling of thin-layer drying of carrot. International Agrophysics, 23(4), 313-317. Alaei, B., Chayjan, R.A. 2015. Modelling of nectarine drying under near infrared - Vacuum conditions. Acta Scientiarum Polonorum Technologia Alimentaria, 14(1), 15-27. Alibas, I. 2014. Mathematical modeling of microwave dried celery leaves and determination of the effective moisture diffusivities and activation energy. Food Science and Technology, 34(2), 394-401. Alibas, I., Kacar, O. 2016. Microwave Drying Kinetics, Hypericin Content, Effective Moisture Diffusivity and Activation Energy ofHypericum perforatumL. Journal of Essential Oil Bearing Plants, 19(2), 454-465. Amami, E., Khezami, W., Mezrigui, S., Badwaik, L.S., Bejar, A.K., Perez, C.T., Kechaou, N. 2017. Effect of ultrasound-assisted osmotic dehydration pretreatment on the convective drying of strawberry. Ultrason Sonochem, 36, 286-300. Amiri Chayjan, R., Kaveh, M., Khayati, S. 2015. Modeling Drying Characteristics of Hawthorn Fruit under Microwave-Convective Conditions. Journal of Food Processing and Preservation, 39(3), 239-253. Amiri Chayjan, R., Shadidi, B. 2014. Modeling High-Moisture Faba Bean Drying in Fixed and Semi-Fluidized Bed Conditions. Journal of Food Processing and Preservation, 38(1), 200-211. Aral, S., Bese, A.V. 2016. Convective drying of hawthorn fruit (Crataegus spp.): Effect of experimental parameters on drying kinetics, color, shrinkage, and rehydration capacity. Food Chemistry, 210, 577-584. Azeez, L., Adebisi, S.A., Oyedeji, A.O., Adetoro, R.O., Tijani, K.O. 2017. Bioactive compounds’ contents, drying kinetics and mathematical modelling of tomato slices Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 66 influenced by drying temperatures and time. Journal of the Saudi Society of Agricultural Sciences. Ben Haj Said, L., Najjaa, H., Farhat, A., Neffati, M., Bellagha, S. 2015. Thin layer convective air drying of wild edible plant (Allium roseum) leaves: experimental kinetics, modeling and quality. Journal of Food Science and Technology, 52(6), 3739-3749. Bensebia, O., Allia, K. 2015. Drying and Extraction Kinetics of Rosemary Leaves: Experiments and Modeling. Journal of Essential Oil Bearing Plants, 18(1), 99-111. Bonazzi, C., Dumoulin, E. 2011. Quality changes in food materials as influenced by drying processes. Modern drying technology, volume 3: Product quality and formulation, 1-20. Bonazzi, C., Dumoulin, E., Bimbenet, J. 2008. Le séchage des produits alimentaires. Industrie Alimentaire Agricole, 125(03-04), 12-22. Boulekbache-Makhlouf, L., Slimani, S., Madani, K. 2013. Total phenolic content, antioxidant and antibacterial activities of fruits of Eucalyptus globulus cultivated in Algeria. Industrial Crops and Products, 41, 85-89. Chayjan, R.A., Salari, K., Abedi, Q., Sabziparvar, A.A. 2013. Modeling moisture diffusivity, activation energy and specific energy consumption of squash seeds in a semi fluidized and fluidized bed drying. Journal of Food Science and Technology, 50(4), 667-677. Cruz, A.C., Guiné, R.P.F., Gonçalves, J.C. 2014. Drying Kinetics and Product Quality for Convective Drying of Apples (cvs. Golden Delicious and Granny Smith). International Journal of Fruit Science, 15(1), 54-78. Dadalı, G., Kılıç Apar, D., Özbek, B. 2007. Microwave Drying Kinetics of Okra. Drying Technology, 25(5), 917-924. Darvishi, H., Asl, A.R., Asghari, A., Azadbakht, M., Najafi, G., Khodaei, J. 2014. Study of the drying kinetics of pepper. Journal of the Saudi Society of Agricultural Sciences, 13(2), 130-138. Darvishi, H., Zarein, M., Farhudi, Z. 2016. Energetic and exergetic performance analysis and modeling of drying kinetics of kiwi slices. Journal of Food Science and Technology, 53(5), 2317-2333. Demirhan, E., ÖZbek, B. 2009. Microwave-Drying Characteristics of Basil. Journal of Food Processing and Preservation, 34(3), 476-494. Demirhan, E., Özbek, B. 2011. Thin-Layer Drying Characteristics and Modeling of Celery Leaves Undergoing Microwave Treatment. Chemical Engineering Communications, 198(7), 957-975. Doymaz, I., Kipcak, A.S., Piskin, S. 2016. Microwave drying of green bean slices: drying kinetics and physical quality. Czech Journal of Food Sciences, 33(No. 4), 367-376. Faal, S., Tavakoli, T., Ghobadian, B. 2015. Mathematical modelling of thin layer hot air drying of apricot with combined heat and power dryer. Journal of Food Science and Technology, 52(5), 2950-2957. Fernando, J.A., Amarasinghe, A.D. 2016. Drying kinetics and mathematical modeling of hot air drying of coconut coir pith. Springerplus, 5(1), 807. Golpour, I., Nejad, M.Z., Chayjan, R.A., Nikbakht, A.M., Guiné, R.P.F., Dowlati, M. 2016. Investigating shrinkage and moisture diffusivity of melon seed in a microwave assisted thin layer fluidized bed dryer. Journal of Food Measurement and Characterization, 11(1), 1-11. Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 67 González-Buesa, J., Page, N., Kaminski, C., Ryser, E.T., Beaudry, R., Almenar, E. 2014. Effect of non-conventional atmospheres and bio-based packaging on the quality and safety of Listeria monocytogenes-inoculated fresh-cut celery (Apium graveolens L.) during storage. Postharvest Biology and Technology, 93, 29-37. Guiné, R.P.F., Pinho, S., Barroca, M.J. 2011. Study of the convective drying of pumpkin (Cucurbita maxima). Food and Bioproducts Processing, 89(4), 422-428. Haksever, A., Moralar, A. 2017. The Effects Of Microwave Energy To The Drying Of Apple (Gala) Slices. Karaelmas Fen ve Mühendislik Dergisi, 7(1), 228-236. Harish, A., Rashmi, M., Krishna Murthy, T., Blessy, B., Ananda, S. 2014. Mathematical modeling of thin layer microwave drying kinetics of elephant foot yam (Amorphophallus paeoniifolius). International Food Research Journal, 21(3), 1081-1087. İZlİ, G. 2017. Total phenolics, antioxidant capacity, colour and drying characteristics of date fruit dried with different methods. Food Science and Technology (Campinas), 37(1), 139- 147. Kaveh, M., Amiri Chayjan, R. 2017. Modeling Thin-Layer Drying of Turnip Slices Under Semi-Industrial Continuous Band Dryer. Journal of Food Processing and Preservation, 41(2), e12778. Kayisoglu, S., Ertekin, C. 2011. Vacuum drying kinetics of Barbunya bean (Phaseolus vulgaris L. elipticus Mart.). The Philippine Agricultural Scientist, 94(3). Khazaei, J., Daneshmandi, S. 2007. Modeling of thin-layer drying kinetics of sesame seeds: mathematical and neural networks modeling. International Agrophysics, 21(4), 335. Kumar, N., Sarkar, B.C., Sharma, H.K. 2012. Mathematical modelling of thin layer hot air drying of carrot pomace. Journal of Food Science and Technology, 49(1), 33-41. Kurobayashi, Y., Kouno, E., Fujita, A., Morimitsu, Y., Kubota, K. 2014. Potent Odorants Characterize the Aroma Quality of Leaves and Stalks in Raw and Boiled Celery. Bioscience, Biotechnology, and Biochemistry, 70(4), 958-965. Madamba, P.S., Liboon, F.A. 2007. Optimization of the Vacuum Dehydration of Celery (Apium Graveolens) Using the Response Surface Methodology. Drying Technology, 19(3-4), 611-626. Madhava Naidu, M., Vedashree, M., Satapathy, P., Khanum, H., Ramsamy, R., Hebbar, H.U. 2016. Effect of drying methods on the quality characteristics of dill (Anethum graveolens) greens. Food Chem, 192, 849-856. Maskan, M. 2000. Microwave/air and microwave finish drying of banana. Journal of Food Engineering, 44(2), 71-78. Minaei, S., Motevali, A., Ahmadi, E., Azizi, M. 2011. Mathematical models of drying pomegranate arils in vacuum and microwave dryers. Journal of Agricultural Science and Technology, 14(2), 311-325. Motevali, A., Minaei, S., Banakar, A., Ghobadian, B., Darvishi, H. 2016. Energy analyses and drying kinetics of chamomile leaves in microwave-convective dryer. Journal of the Saudi Society of Agricultural Sciences, 15(2), 179-187. Murthy, T.P., Manohar, B. 2014. Hot air drying characteristics of mango ginger: Prediction of drying kinetics by mathematical modeling and artificial neural network. J Food Sci Technol, 51(12), 3712-3721. Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 68 Nema, P.K., Mohapatra, D., Daniel, A., Mishra, S. 2013. Modeling pulse microwave drying kinetics of ginger (Zingiber officinale R.). Journal of Food Research and Technology| October-December, 1(2), 46-58. Nguyen, V.T., Van Vuong, Q., Bowyer, M.C., Van Altena, I.A., Scarlett, C.J. 2015. Effects of Different Drying Methods on Bioactive Compound Yield and Antioxidant Capacity of Phyllanthus amarus. Drying Technology, 33(8), 1006-1017. Nwakuba, N., Chukwuezie, O., Osuchukwu, L. 2017. Modeling of Drying Process and Energy Consumption of Onion (Ex-gidankwanoSpp.) Slices in a Hybrid Crop Dryer. Olawoye, B.T., Kadiri, O., Babalola, T.R., Yildiz, F. 2017. Modelling of thin-layer drying characteristic of unripe Cardaba banana (Musa ABB) slices. Cogent Food & Agriculture, 3(1). Özbek, B., Dadali, G. 2007. Thin-layer drying characteristics and modelling of mint leaves undergoing microwave treatment. Journal of Food Engineering, 83(4), 541-549. Ratti, C. 2008. Advances in food dehydration: CRC Press. Sarimeseli, A. 2011. Microwave drying characteristics of coriander (Coriandrum sativum L.) leaves. Energy Conversion and Management, 52(2), 1449-1453. Sledz, M., Wiktor, A., Nowacka, M., Witrowa-Rajchert, D. 2017. Drying Kinetics, Microstructure and Antioxidant Properties of Basil Treated by Ultrasound. Journal of Food Process Engineering, 40(1), e12271. Sledz, M., Wiktor, A., Rybak, K., Nowacka, M., Witrowa-Rajchert, D. 2016. The impact of ultrasound and steam blanching pre-treatments on the drying kinetics, energy consumption and selected properties of parsley leaves. Applied Acoustics, 103, 148-156. Soran, M.L., Stan, M., Niinemets, U., Copolovici, L. 2014. Influence of microwave frequency electromagnetic radiation on terpene emission and content in aromatic plants. J Plant Physiol, 171(15), 1436-1443. Taghian Dinani, S., Hamdami, N., Shahedi, M., Havet, M. 2014. Mathematical modeling of hot air/electrohydrodynamic (EHD) drying kinetics of mushroom slices. Energy Conversion and Management, 86, 70-80. Tasirin, S.M., Puspasari, I., Lun, A.W., Chai, P.V., Lee, W.T. 2014. Drying of kaffir lime leaves in a fluidized bed dryer with inert particles: Kinetics and quality determination. Industrial Crops and Products, 61, 193-201. Tsotsas, E., Mujumdar, A.S. 2014. Modern Drying Technology, Volume 5: Process Intensification: 2014, 300 pp Vega-Gálvez, A., Puente-Díaz, L., Lemus-Mondaca, R., Miranda, M., Torres, M.J. 2014. Mathematical Modeling of Thin-Layer Drying Kinetics of Cape Gooseberry (Physalis peruviana L.). Journal of Food Processing and Preservation, 38(2), 728-736. Viña, S.Z., Chaves, A.R. 2003. Texture changes in fresh cut celery during refrigerated storage. Journal of the Science of Food and Agriculture, 83(13), 1308-1314. Xiao, H.-W., Bai, J.-W., Xie, L., Sun, D.-W., Gao, Z.-J. 2015. Thin-layer air impingement drying enhances drying rate of American ginseng (Panax quinquefolium L.) slices with quality attributes considered. Food and Bioproducts Processing, 94, 581-591. Zhang, Q.-A., Song, Y., Wang, X., Zhao, W.-Q., Fan, X.-H. 2016. Mathematical modeling of debittered apricot (Prunus armeniaca L.) kernels during thin-layer drying. CyTA - Journal of Food, 14(4) 1-9. Mouhoubi et al. / AUDJG – Food Technology (2019), 43(2), 48-69 69 Zhu, J.-F., Liu, J.-Z., Wu, J.-H., Cheng, J., Zhou, J.-H., Cen, K.-F. 2015. Thin-layer drying characteristics and modeling of Ximeng lignite under microwave irradiation. Fuel Processing Technology, 130, 62-70.