Microwave assisted drying and extraction technique; kinetic modelling, energy consumption and influence on antioxidant compounds of fenugreek leaves

Fenugreek (Trigonella foenum graecum L.) is famous as seasonings and medicinal herb due to aromatic and functional compounds present in it. Its leaves are good source of minerals (iron, calcium, phosphorous), vitamins (vitamin C, riboflavin, niacin, thiamine, folic acid), carotene and flavonoids. These flavonoids are present as complex glycosides (C-glycosidic and O-glycosidic bonds) and among these major flavonoids are quercetin, vitexin, apigenin, etc (Nagulapalli Venkata et al., 2017). These compounds exhibit the potential for the treatment of various ailments e.g. diabetes (type I and type II), chronic cough, diarrhea, ulcer and rickets (Anoopkumar et al., 2020; Aylanc et al., 2020; El Sohaimy et al., 2015).


Introduction
Fenugreek (Trigonella foenum graecum L.) is famous as seasonings and medicinal herb due to aromatic and functional compounds present in it. Its leaves are good source of minerals (iron, calcium, phosphorous), vitamins (vitamin C, riboflavin, niacin, thiamine, folic acid), carotene and flavonoids. These flavonoids are present as complex glycosides (C-glycosidic and O-glycosidic bonds) and among these major flavonoids are quercetin, vitexin, apigenin, etc (Nagulapalli Venkata et al., 2017). These compounds exhibit the potential for the treatment of various ailments e.g. diabetes (type I and type II), chronic cough, diarrhea, ulcer and rickets (Anoopkumar et al., 2020;Aylanc et al., 2020;El Sohaimy et al., 2015).
These techniques have high efficiency in drying, but the major concerns are loss of water-soluble bioactive compounds (BACs) and higher processing time & cost (Ferreira et al., 2018;Rocha & Melo, 2011;Thamkaew et al., 2020). Additionally, conventional drying techniques significantly reduce the quality of the product in terms of colour and textural properties (Duc-Pham et al., 2019). Therefore, an emerging drying technique with the purpose of making process of dehydration more sustainable is needed. Microwave based technique (Ali et al., 2020;Khan et al., 2016;Monton et al., 2019a) has exhibited its potential for drying i.e. reduced drying time and cost. Thus, microwave method of drying is energy efficient and easy to use compared to other drying methods (Alvi et al., 2019;Gamboa-Santos & Campañone, 2019;Khan et al., 2016;Lv et al., 2019). However, loss of water-soluble bioactive compounds is still a major problem in microwave-based dehydration systems.
The superiority of microwave-based method urges to develop a novel method that may collect the vapours containing bioactive compounds as a superfluous improvement in addition to drying process. Therefore, present study focuses on the modification and optimization of microwave assisted drying and extraction (MADE) technique. In this regard, fenugreek leaves are used as model product for extraction and drying purposes. The extracted liquid obtained from microwave extraction method was compared

Microwave assisted drying and extraction technique; kinetic modelling, energy consumption and influence on antioxidant compounds of fenugreek leaves
with methanol-based method. While, the dried product was compared with fresh leaves of fenugreek to estimate the impact of microwaves on leaves.

Procurement of raw materials
Fresh fenugreek leaves were procured from farms of University of Agriculture, Faisalabad-Pakistan. The impurities & dust particles were removed from fenugreek leaves through washing with running water. Afterwards, leaves were placed on tissue paper for the removal of excess water from their surface. These leaves were used for further experiments. Additionally, chemicals were procured from well-known company Sigma Aldrich (Germany).

Drying and extraction process
Drying of fenugreek leaves was done through microwave assisted extraction and drying technique. In this regard, an experimental setup was built in lab facility which consist of a modified microwave oven HDG 236S (Homage, Korea) as shown in Figure 1. The samples of leaves (50 g) were placed in glass reactor for drying purpose to extract the bioactive compounds from these leaves. The oven was operated at different powers (30, 50, 80 and 100 W). The loss in moisture was investigated by measuring the weight of the samples at different time intervals (1, 2, 3, 4, ……22 minutes) with the help of weighing balance (Shimadzu, Japan). The water vapors containing bioactive compounds were liquified with the help of condenser ( Figure 1) in glass bottles and preserved at 4˚C for further analysis. Additionally, methanol-based extraction was performed for the comparison point of view. For this purpose, sample (5 g) homogenized in methanol solution (80%) for 45 minutes at 8000 rpm. The supernatant layer was removed from the rest of material in tube and utilized for the TFC, TPC and DPPH analysis (Justine et al., 2019).

Drying rate
The values of drying rate for fenugreek leaves were calculated by using Equation 1 and expressed as g water/100 g.min.
where, t is drying time in minutes, V t represents the moisture contents at time t, while V t+dt are the moisture contents at t + dt interval (Alvi et al., 2019;Yilmaz & Alibas, 2017).

Moisture ratio
The moisture ratio of a product demonstrates the relative removal of water contents and calculated as below (Equation 2). where, M t , M o , and M e represents moisture levels in leaves at time interval of t, initial and equilibrium, respectively. This equation for moisture ratio was further modified to M t /M o by some researchers for ease in calculations as described in the literature (Al-Harahsheh et al., 2009;Doymaz, 2005;Karacabey, 2016). Besides, the most important aspect of drying process is mathematical modelling that may help to design drying equipment with optimum drying of desired product. The modelling was based on mathematical equations that can portray the system of drying (Sunil et al., 2014). Therefore, different models were selected from the literature and compared with experimental data to observe which model is best fit (Agbede et al., 2020). Moreover, diffusion coefficient was determined by using Fick's law (Mahjoorian et al., 2016). For microwave based drying process, Fick's law can be transcribed as below: where, D is the diffusion coefficient (m 2 /s) in the sample having thickness L in meters. Here, the value of D was obtained by slope of a graph which was plotted between the values of ln MR and time (minutes). The slop of this graph (α) is described as: Furthermore, influence of various variables (time, power and moisture content) on the values of drying rate and moisture ratio were determined by applying box-Behnken design. The box-Behnken design was elaborated by quadratic equation by using JMP software (SAS Institute, USA).

Specific energy of MADE
Specific energy consumption of MADE process of fenugreek leaves at various powers can be determined by the following equation (Darvishi et al., 2013;Jahanbakhshi et al., 2020) and expressed in MJ/kg (Equation 11): where, M o and M t are initial and final moisture contents of fenugreek leaves and M s is the dry matter in kilograms. E microwave was calculated by Equation 12: where, t and p are the time in seconds and power in watts, respectively (Alvi et al., 2019;Khan et al., 2016).

Determination of DPPH activity
The antioxidant activities (DPPH activity) of the samples were determined by the method described in literature (Barkat et al., 2018;Curi et al., 2019). In this method, sample (0.5 mL) material in the form of extract obtained from Microwave and methanol extraction method added in methanol-based DPPH (4% w/v) solution; subsequently vigorous shaking of the mixture was carried out. Afterwards, the mixture was placed in darkness (overnight) and absorbance was measured at 515 nm using a spectrophotometer (IRMECO, Germany). The sample results of DPPH activity were reported in terms of percentage (%).

Total phenolic and flavonoid contents
Total phenolic compounds (TPC) were determined according to the method of Sun et al. (2005) with minor modifications (Aydar, 2020;Lu et al., 2011). In this method, Folin-Ciocalteu reagent was diluted with distilled water to prepare ten-time dilutions. Afterwards, a small portion of this solution (0.75 mL) was mixed with extracted sample (0.1 mL) and incubated for ten minutes. Subsequently, sodium carbonate solution (0.75 mL) was mixed with solution and placed in the dark conditions for forty-five minutes. Finally, absorbance reading of sample was measured at 765 nm with spectrophotometer (IRMECO, Germany). Besides, a calibration curve of absorbance was sketched against the different concentrations of gallic acid (0, 5, 10, 15, 20, 30 and 40 mg/L). The sample values were calculated and reported as gallic acid equivalent (mg/g).
For total flavonoid contents (TFC) determination, 1 mL of extract was mixed in 3 mL solution of methanol containing aluminum chloride and potassium acetate. This solution was further diluted by adding 5.6 mL of distilled water and placed at room temperature for half an hour. Afterwards, sample absorbance was determined with spectrophotometer (IRMECO, Germany) at 415 nm wavelength. The standard curve of quercetin was taken as reference and sample values were expressed in terms of quercetin equivalents (Ghafar et al., 2017).

Statistical analysis
All the experiments were performed in triplicates and their means along with standard errors were reported. The JMP R (SAS Institute, USA) was used to perform analysis of variance and Tukey's test that determine the level of significance and comparison between treatments. Moreover, fitness of the models was determined by the values of R 2 , RMSE, SSE, RPD and Chi square (Guiné, 2018;Mota et al., 2010;Roberts et al., 2008;Equation 13-16

Drying rate
Drying rates of fenugreek leaves were calculated with the help of Equation 1 and results have been shown in Figure 2. It was evident that drying rate initially increased swiftly and reached up to 4.5 g/100 g.min within two minutes when heated at 100 W power (Figure 2a, b). At this point, maximum amount of moisture had been removed from the product surface. Afterward, value of drying rate decreased significantly indicating the reduction in moisture contents of the samples. Thus, drying rate of fenugreek leaves followed the parabolic trend in microwave based drying system (Figure 2b, c). Initially, the higher values of drying rate were due to higher moisture levels of leaves that absorbed more energy of microwaves; resulting in a prompt temperature rise. It stimulates the rapid evaporation process from the product surface (Figure 2b, c). After 50% moisture removal, less moisture level was present in sample for evaporation, compared to initial time; thereby reduced rate of drying (Soysal et al., 2006;Therdthai & Zhou, 2009;Wang et al., 2007). This lowering of the drying rate as a function of time was attributed to the reduced availability of free moisture as previously discussed in literature (Alvi et al., 2019;Khan et al., 2016Khan et al., , 2019. Besides, effect of microwave power on drying rate of leaves was observed. It was noted that the drying rate of leaves was increased from 1.79 to 4.56 ± 0.05 g/(100 g.min) when power was increased from 30 to 100 W. It indicated that increase in power had significantly increased the drying rate. Thereby, reducing the processing time; ultimately the processing cost.
Moreover, ANOVA was performed to assess the influence of independent variables on drying rate as linear, interaction and quadratic and residual coefficients. It was obvious that a significant effect of variables on drying rate (Table 1) was evident especially in terms of linear and quadratic effects. Besides, the authenticity of second order polynomial was evaluated through the values of coefficient of determination (R 2 ) and F-test. The resultant value for R 2 (0.98) is close to one and indicated the less variation in response model. Thus, this model may be used to explain current experimentation. The second order polynomial model equation obtained by response surface methodology analysis for drying rate is as follows (Equation 17 where, P, T and MC represent the power in watts, time in minutes and moisture content in percentage. These model values were compared with experimental ones and experimental drying rate was good fit with the predicted values ( Figure 2d) as supported by the statistical values for predicted and actual drying rate (R 2 =0.895, p<0.0001).

Drying time and moisture ratio
During MADE processing, time and power levels were evaluated. It is evident from the results that maximum time (21 min) was required for drying at power 30 W of microwave (Table 2). By increasing the power, drying time was reduced significantly, and minimum time required for complete drying was 5 min at power 100 W. The difference in drying time may be attributed to variation in heating intensity among the different power levels (Coolong et al., 2008;Özdemir & Altan, 2011).
Moisture ratio (MR) of fenugreek was calculated according to Equation 2 and results are shown in Figure 3. The value of MR at different powers of microwave (30, 50, 80, & 100 W) were compared (Figure 3) and results indicated that MR rapidly decreased at 100 W. This rapid decrease in MR indicates the swift drying of leaves consequently reducing the processing time. Furthermore, the experimental moisture ratio is in good agreement with predicted moisture ratio (<0.0001) with good authenticity (R 2 =0.997) of predicted values (Figure 3d).
Additionally, MR was predicted by various model equations (Equations 3-8) and compared with experimental values. According to statistical analysis, R 2 value was highest for Midilli model (0.9988) followed by two term model (0.9958), which indicates the good fit of the model (Table 3)

Diffusivity coefficient
The diffusive coefficient was calculated by using Equation 9 and 10. The results obtained had shown that diffusion process was controlled by the falling rate period. In this study, two phases of falling rates were observed that occurred in a constant slope. This slope determined the values of effective diffusion individually. The cut off value between these two falling rate periods was 0.3 of moisture ration. The effective diffusivities in the first falling rate period were influenced by the microwave powers and the range of values were 0.67 to 8.13 (10 −12 m 2 /s). Likewise, the values of effective diffusion in the second phase were influenced by power and values varied from 0.15 to 1.64 (10 −12 m 2 /s). The values of second period were about eight times smaller than first falling rate period ( Table 2).
The average values of effective diffusion of fenugreek leaves during drying process at 30-100 W varied in the range of 0.73 to 7.67 (10 −14 m 2 /s) ( Table 2). The diffusion coefficient increased with the increase of microwave power and results were in agreement with the research reported in literature (Alvi et al., 2019;Mahjoorian et al., 2016).

Specific energy consumption
The energy consumption of microwave assisted drying and extraction method was calculated and shown in Table 2. It is evident from the results that 80, 50 and 30 W consumed more energy (2.33, 2.47 and 2.39 MJ/kg, respectively) compared to 100 W (1.8672 MJ/kg). This may be attributed to lowest processing time (5 minutes) at 100 W processing compared to    others (21 minutes @ 30 W); longer the processing time, higher the energy consumption. Thus, heating the food products at 100 W is sustainable in terms of time, cost and energy utilization.

DPPH, total phenolic and flavonoid contents
DPPH, TPC and TFC contents of MADE extract were compared with methanol-based extraction method (Table 4). The results indicated that liquid sample, obtained through MADE processing, exhibited better values of TFC, TPC and DPPH activity compared to methanol-based extraction method. These findings were comparable to the literature values (Cheng et al., 2013;Khan et al., 2016). Moreover, higher values of TPC, TFC and DPPH activity in comparison with methanol-based extraction method exhibited that the proposed technique (MADE) did not affect the activity of bioactive compounds of extract and this claim was supported by the findings of (Ferreira et al., 2018) who reported that microwave drying does not affect the bioactive compounds (BACs) values of the product.

Conclusion
In this study, MADE technique was evaluated for drying of fenugreek leave as well as the extraction of BACs from these leaves. The drying time of MADE process was significantly reduced (75%) at highest processing power (100 W) and it consumed less energy compared to other microwave powers. Likewise, the drying rate values were found to be very high (4.56 ± 0.05 g/100 g.min) at 100 W that reduced the moisture content rapidly and ease in swift drying. Similarly, the MR values decreased quickly at 100 W compared to other powers and lowest energy (1.8672 MJ/kg) was consumed. Thus, 100 W for 5 min considered to be the optimum for drying of fenugreek leaves. Moreover, BACs extracted through MADE technique exhibited better TPC, TFC, and DPPH values compared to methanol-based extraction process. Based on these findings, it can be concluded that microwave based drying and extraction is an efficient and sustainable process compared to both traditional methods of drying and extraction.