Optimization of microwave vacuum drying parameters for germinated lentils based on starch digestibility, antioxidant activity and total phenolic content

The aim of this study was to optimize the processing parameters of pulse mode microwave-vacuum drying of germinated green and red lentils (CDC Greenland and CDC Maxim) and investigate the changes in their total phenolic content (TPC), total antioxidant activity (TAA) and In-vitro starch digestibility (SD). The lentils were germinated for 5 days and dried by a pulse mode microwavevacuum method, using 2 s to 8 s out of 10 s pulsed mode at 2000W microwave power and varying the vacuum pressure level between 15 and 45 kPa. In-vitro starch digestibility increased significantly with increased microwave power level. The TPC and TAA appeared to vary distinctively in the two varieties of selected lentils. Vacuum pressure levels did not significantly (p>0.05) affect any responses. Green lentils could be dried at 8 s microwave power and 45 kPa vacuum pressure and red lentils could be dried at 5.5 s microwave power and 42.19 kPa vacuum pressure. The microwave-vacuum drying showed great potential for the drying of germinated lentils.


Introduction
Lentils (Lens culinaris), an important legume crop, are among the oldest food cultivated for human consumption (Iqbal, Khalil, Ateeq, & Khan, 2006). They are widely cultivated in the American and Asian subcontinents. They are consumed mainly in south-east Asian countries, middle-east, Africa and Latin America. Pulses and lentils are considered as an inexpensive and affordable source of protein and carbohydrates with other valuable nutriente (Giannakoula, Ilias, Maksimovic, Maksimovic, & Zivanovic, 2012;Gharachorloo, Tarzi, & Baharinia, 2013). In the context of living a healthy lifestyle, consump-2002). Among germinated legumes, lentils contain various functional components which upon consumption may reduce or prevent risks of diabetes, cardiovascular diseases and inflammation (Chung, Liu, Hoover, Warkentin, & Vandenberg, 2008;Caccialupi et al., 2010). In addition, consumption after germination is proven as one of the most effective methods for increasing the bioavailability of certain nutrients (Swieca, Baraniak, & Gawlik-Dziki, 2013). Many trials have been made to modify the functional chemical composition of lentils using germination processes by varying time, illumination and elicitation (Swieca et al., 2013). During the germination process, seeds are first soaked in water before being placed under warm and humid conditions which are also ideal and optimal condition for bacterial proliferation. The germination step is the main source of contamination in sprouts; hence, consumers are recommended to cook (thermal treatment) the sprouts thoroughly before consumption (FDA, 2004). On the other hand, incorporation of germinated lentils in our daily diet can be made through breakfast cereals, snacks and extruded products for example pasta and noodles. However, germinated lentils, having higher moisture content, should be dried to desirable moisture content before further processing and value addition. Among the electrotechnology-based heating technologies, viz., radio frequency (RF), microwave (MW), ohmic and infrared (IR), microwave heating shows more promise in drying of food items (Sham, Scaman, & Durance, 2001). Kadlec et al. (2001) reported that germination followed by microwave and hot air heat treatment of peas reduces α-galactooligosaccharides while improving their nutritive value. With the combination of vacuum with the microwave drying system, the materials can be dried in a shorter time and at low temperatures allowing better retention of sensory attributes such as colour, texture and essential nutrients (Mitra & Meda, 2009). Vacuum assisted microwave drying has been applied to reduce the moisture content of various plant materials, such as cranberries (Sunjka, Rennie, Beaudry, & Raghavan, 2004), tomatoes (Durance & Wang, 2002), garlic (Figiel, 2009), rosemary (Szumny, Figiel, Gutierrez-Ortiz, & Carbonell-Barrachina, 2010;Calin-Sanchez et al., 2011) Saskatoon berries (Mitra & Meda, 2009) and sour cherries (Wojdyo, Figiel, Lech, Nowicka, & Oszmianski, 2014). Hence, microwave assisted drying shows a potential method for drying of germinated seeds as the operation is rapid, more uniform and energy efficient as compared to conventional hot air drying. So far, little has been reported regarding the effects of drying of germinated lentils on antioxidants and In-vitro starch digestibility. Therefore, the objective of the current study was to optimize the processing parameters of microwave-vacuum drying of germinated lentils and investigate its changes on total phenolic content, total antioxidant activity and In-vitro starch digestibility.

Materials and Methods
Green (CDC Greenland) and Red (CDC Maxim) lentil varieties were supplied by the University of Saskatchewan's Crop Development Centre (CDC), Canada. Both varieties were germinated in aluminium containers at room temperature (25°C) for 5 days (120 h) under dark conditions after soaking in distilled water for 8 h. Sprouts were rinsed twice a day (every 12 h). The number of days of germination were selected based on preliminary trials that gave the highest antioxidant activity and starch digestibility (data not shown). The 2,2-diphenyl-1-picrylhydrazyl (DPPH), Folin-Ciocolteu reagent, gallic acid and 6-hydroxy-2,5,7,8-tetramethychromane-2carboxylic acid (Trolox) were purchased from Sigma-Aldrich Canada Ltd. (Oakville, Ontario, Canada).

Experimental design of
microwave-vacuum drying nated lentils. The dryer is capable of producing 2000 W microwave power continuously, however it can also be used in pulsed mode by supplying microwave power for 0-10 s out of 10 s, where 10 s out of 10 s served as continuous mode. We used pulsed mode microwave power of 2-8 s out of 10 s and constant vacuum pressure of 15-45 kPa as independent parameters, each in 5 levels. Therefore, a total of 13 experiments were conducted for each lentil variety. Samples of 300 g were dried to 10±2 % moisture content (wet basis) using a perforated polypropylene cylindrical container rotating at 10 rpm. The actual and coded values of experimental design are shown in Table 1. The input range was selected based on initial trials to avoid burning of samples. The samples were ground using a multipurpose grinding mill with 1 mm diameter sieve, and stored at -20°C in a sealable polypropylene bag for their chemical analysis. All the experiments were conducted in triplicate.
The following second order polynomial response surface model was fitted to each of the response for analysis, as below: where, Y is response, β 0 , β i , β i i, and β i i are regression coefficients of the constant, linear, quadratic and interaction terms, respectively, and X i and X j are codes of independent variables and k is the number of variables. X i and X j were replaced by A and B which signify microwave power and vacuum pressure, respectively. A sequential model sum of square (type I) was carried out to select the final model, where a quadratic model was suggested to fit the responses as cubic and higher order polynomial models were aliased. The goodness-of-fit of the model was measured by coefficient of determination (R 2 ). The adequacy of the quadratic models was confirmed by analysis of variance (ANOVA) using Fisher Test value (F-value) and lack of fit. The optimization of process variables was based on maximum values of all the responses i.e. total phenolic content (TPC), total antioxidant activity (TAA) and In-vitro starch digestibility (SD).
In the software, responses can be assigned to different importance levels, where the importance level varies from 1 to 5. Therefore, the process variables were optimized based on higher desir-ability of maximum TAA and SD (importance level: 5) as compare to maximum TPC (importance level: 3).

Total phenolic content and total antioxidant activity
Ground samples were extracted using 70 % aqueous ethanol (1/10 w/v) at 25°C. The supernatants were decanted into a glass vial after centrifuging at 3500 rpm for 10 minutes. This process was repeated three times to maximize the extraction. On third repetition, the supernatant was a clear solution. The extracts were stored at -20°C for further analysis.
The TPC was determined based on the colour reaction of Folin-Ciocolteu reagent with hydroxyl group (Swieca, Seczyk, & Gawlik-Dziki, 2014). The absorbance was measured using a spectrophotometer (Model: 6305, Jenway, Bibby Scientific Limited, Staffordshire, UK) at 765 nm and total phenolic content were expressed in terms of gallic acid equivalent (GAE) in mg/g DM. For TAA, a 2,2-diphenyl-1-picrylhydrazyl (DPPH) stock solution of 500 µ M was used. TAA was determined according to the procedures described in Mitra, Meda, and Green (2013). The absorbance was measured at 517 nm for both samples and blank. The spectrophotometer was calibrated using distilled water and 70 % methanol for TPC and TAA, respectively. Percentage DPPH inhibition was determined using following the equation: Abs = absorbance for both sample and blank. The level of DPPH radical scavenging capacity IJFS April 2017 Volume 6 pages 44-55 was expressed in terms of Trolox equivalent (Teq) mg per g dry matter by plotting a standard curve of DPPH inhibition against Trolox concentration.

In-vitro starch digestibility
In-vitro starch digestibility (% SD) was determined using Megazyme resistant starch assay kit (K-RSTAR, Megazyme International Ireland Limited, Wicklow, Ireland). Resistant starch (RS), non resistant starch (NRS) and Total Starch (TS) content were determined by hydrolyzing the sample using pancreatic α-amylase and amyloglucosidase for 16 h at 37°C, following the procedure described in the assay kit. Absorbances were measured using the spectrophotometer at 510 nm against reagent bank (sodium acetate buffer and glucose oxidase/peroxidase reagent only). Total starch and In-vitro starch digestibility was measured using following equation (Swieca et al., 2013): where T S is the total starch, RS is the Resistant starch and N RS is the Non-resistant starch. Kiwi (Kaya, Aydm, & Kolayli, 2010). For total antioxidant activity (TAA), minimum and maximum values were observed at similar processing parameters as for TPC for both varieties. the maximum value of TAA for germinated green lentils was 10.12 Teq mg/g DM and a minimum of 3.67 Teq mg/g DM, while 11.57 Teq mg/g DM and 6.75 Teq mg/g DM were found for red lentils. This showed that there is a good relationship between TPC and TAA. The reduction and a similar relationship between TPC and TAA were also observed in drying of Thai red curry powder (Inchuen, Narkrugsa, & Pornchaloempong, 2010), goldenberry (Izli, Yildiz, Unal, Isik, & Uylaser, 2014) and other fruits (Sultana, Anwar, Ashraf, & Saari, 2012). It can also be noted that the TAA of germinated red lentils was higher than green lentils at same processing conditions, except for runs 2, 4 and 6. While the TPC of green lentils was higher than for the red ones at same processing conditions, except for runs 8, 9 and 11. In the case of In-vitro starch digestibility (SD), the percent reduction from maximum to minimum in both the lentil varieties was much less, being 3.39 % and 4.99 % for germinated green and red lentils, respectively. Maximum SD was observed at 9.243 s MW power and 30 kPa vacuum pressure for green lentils and 8 s MW power and 45 kPa vacuum pressure for germinated red lentils. In contrast, minimum SD was observed at 0.757 s MW power and 30 kPa vacuum pressure for both varieties. Similar effects of microwave power were reported after drying of germinated peas (Kadlec et al., 2001) and drying of barley (Emami, Perera, Meda, & Tyler, 2012). In-vitro SD of germinated green lentils was higher than germinated red lentils after microwave drying.

Statistical analysis and model fitting using RSM
The values of TPC, TAA and SD at different microwave (MW) and vacuum pressure levels were fitted to a second order polynomial response surface model and their significance assessed using ANOVA. Estimated coefficients of process parameters, R 2 and adjusted R 2 of the models developed are shown in table 3. Lack of fit, Fvalue and p-value of individual processing parameters of microwave vacuum drying of germinated green and red lentils are listed in table 4 and 5, respectively. The models developed for all the responses were significant (p<0.05). The R 2 value, which determined the variation in fitting the model, was higher than 0.8 for all the models. In addition, the adjusted R 2 indicated the percentage of variation explained only by the independent variables that actually affected the dependent variable. The minor difference between R 2 and the adjusted R 2 is desirable since it means that the model's ability to explain data was not diminished with inclusion of a number of covariates/factors. Since the models developed were significant with higher R 2 values, they could be used to predict and explain the changes made during pulsed mode microwave-vacuum drying of the germinated lentils (within the range). The F-value and P-value of TPC of green lentil (Y 11 ), which indicate the significance of the model developed, were 14.49 and 0.0014 respectively, which implied that the model was significant. Similarly, other responses, TAA (Y 12 ) and SD (Y 13 ) of germinated green lentils with F-value of 8.24 and 5.61 respectively were significant (p<0.05). In Y 11 , only microwave power level (A) and A 2 were significant. This showed that TPC was significantly (p<0.05) influenced by microwave power only. Similarly, for Y 12 and Y 13 , only term A was significant and vacuum pressure was less effective in altering total antioxidant activity and In-vitro starch digestibility (SD) during pulsed mode microwave vacuum drying of germinated lentils.

Response surface analysis
Response surfaces, shown in Fig. 1, were plotted to help visualize the effects of microwave power, vacuum pressure and their interactions on TPC, TAA and SD of germinated green and red lentils. It was observed that TPC, TAA and SD of germinated green lentils increased with increased microwave power and vacuum pressure, and microwave power level was more significant for changes in the responses. In all the responses, the interaction terms were not affected significantly. However, for germinated red lentils, the TPC and TAA increased with increasing microwave power up to 5 s microwave power level and started to descend with increasing microwave power level. There was a gradual rise in TPC and TAA with increasing vacuum pressure level, but only the square of microwave pressure level was significant. Similar to the case of germinated green lentils, the interaction terms of microwave power and vacuum pressure level did not affect significantly (p>0.05) the TPC and TAA. The increase in TPC and TAA with increasing microwave and vacuum pressure level may have been due to shorter drying time and lower temperature of drying. Microwave energy causes volumetric heating resulting in heat transfer from inside to the surface while higher vacuum pres-sure helps lower evaporating temperature which ultimately results in faster drying. Prolonged exposure to this type of drying technique may result in loss of volatile material and oxidative reduction of phenolic compounds and these phenomena (Mitra et al., 2013). Lower power MW drying resulted in longer drying time and lower temperatures. At lower MW power, it is possible that oxidative enzymes such as polyphenoloxidases and peroxidases were not inactivated immediately, resulting in degradation of phenolic compounds during long exposure. Moreover, most of the antioxidants are phenolic compounds present in the seed coats of lentils, which are easily accessible to oxidative processes (Khan, Jacobsen, & Eggum, 1979). There are reports of positive relationships between phenolics and antioxidant activity in lentils (Cevallos-Casals & Cisneros-Zevallos, 2010;Swieca, 2015) and thus, the reduction of antioxidant activity can be explained by oxidative processes and long exposure of lentils during drying process. The difference in trends of TPC and TAA among the varieties studied may be due to differences in size, as the surface area for oxidation varied and also due to differences in the types and amount of phenolics present in them.
In-vitro starch digestibility of both varieties of germinated lentils also increased with the increase in microwave power similar and the findings matched those of Kadlec et al. (2001) in the drying of germinated peas and Emami et al. (2012) in the drying of barley. It has been reported that starch might degrade due to starch fragmentation (starch micronization) during irradiation resulting in increasing sugar content of legumes and cereal crops (Khan et al., 1979;Gonzalez & Perez, 2002). Emami, Meda, Pickard, and Tyler (2010) also reported that micronization increases the amount of readily digestible starch and decreases resistant starch and slowly digestible starch. This means the In-vitro digestibility of starch will increase with increased dose of microwave energy. In addition, higher microwave power may decrease the levels of antinutritional compounds leading to higher digestibility of starch. The optimum processing parameters generated based on maximizing TPC, TAA and SD with maximum importance given to TAA and SD are     shown in table 6. Samples were treated at optimum process conditions to validate the values of optimum processing parameters. The variations with the predicted and observed values were less than 10 % (Table 6). Therefore, the models generated could be employed for determining changes during drying of germinated green lentil (CDC Greenland) and red lentil (CDC Maxim) in the given processing range. Raw samples (before germination) had TPC values of 7.48 and 6.42 GAE mg/g DM for green and red lentils, respectively, while TAA values were 10.44 and 12.87 Teq mg/g DM, respectively for green and red lentils. These results show that the reduction in TPC and TAA of the lentils after processing at optimum conditions was about 16 % and 13 %, respectively. For the SD (raw sample having SD of 64.95 % and 62.18 % for green and red lentils respectively), there was an increase in starch digestibility up to about 40 %, showing that microwave vacuum drying after germination process increased starch digestibility of lentils with only a small reduction in TPC and TAA. Aguilera et al. (2009) reported that starch digestibility may be improved by initial soaking and thermal treatment due to gelatinization of starch granules and reduction of antinutritional compounds from the raw seed.

Conclusions
In drying of germinated lentils using pulsed mode microwave-vacuum drying technique, starch digestibility increased significantly with increased microwave power level. The TPC and TAA may have varied distinctively with variation in lentil variety. Vacuum pressure levels did not affect significantly (p>0.05) the responses that were stud-ied. For the optimum drying of 5 days germinated lentils in pulsed mode microwave vacuum condition, green lentils could be dried at a pulse rate of 8 s out of 10 s microwave power of 2000 W and 45 kPa constant vacuum pressure, whereas red lentils could be dried at a pulse rate of 5.5 s out of 10 s microwave power of 2000 W and 42.19 kPa constant vacuum pressure. The mathematical model generated for drying of germinated lentils could be used for scaling up the process. Germination followed by microwave vacuum drying could be a promising path to increase the digestibility of starch with minimum changes in phenolics and antioxidant activity. Moreover, this technology could also be employed for drying other pulses for further processing. The dried germinated lentils obtained could be treated as a raw material for further utilized as ingredients in bakery, extruded snacks and other value added healthy food products. However, further studies may be required for new product development.