Process Parameters Optimization for Continuous Infrared Rice Bran Stabilizer and Assessment of Their Impact on Quality and Shelf Life

Rice is being cultivated on over 44 million hectares in India with productivity of 2,400 kg per hectare and produced about 110 million tonnes of rice during 2016-17 (SEA, 2017). In the rice milling industry, one of the valuable by-products is rice bran. The bran contains a substantial amount of protein, fat and dietary fiber in addition to minerals (such as magnesium and potassium) and vitamins (such as thiamine, riboflavin, niacin and pyridoxine) (Esa et al., 2013). Rice bran oil fatty acid comprises of 41% monounsaturated, 36% polyunsaturated and 19% saturated (Kahlon et al., 1992).  When the bran layer separated from the endosperm during milling, the lipase enzyme is activated, resulting in the breakdown of fat into free fatty acids (FFA) and glycerol. The FFA level rises immediately after milling and bran oil turns unsafe for consumption due to its decreased pH, rancid flavor and soapy taste (Rosniyana et al., 2009). Therefore, adopting an appropriate stabilization technology capable of inhibiting rancidity and microbiological activity is essential to preserving rice bran after milling and enhancing its quality and shelf life (Ju and Vali, 2005).
       
Different methods employed for rice bran stabilization are dry heat treatment (Yu et al., 2019), microwave treatment (Patil et al., 2016), ohmic heating (Dhingra et al., 2012), extrusion (Sharma et al., 2004), infrared radiations Wang et al., (2017) γ-irradiation, parboiling (Pradeep et al., 2014); (Thanonkaew et al., 2012) and toasting (Silva et al., 2006). Moist heat treatments like steam retorting are costly, while extrusion cooking is very expensive in terms of operating, initial and maintenance costs (Dhingra et al., 2012). Since the thermal energy of infrared radiation (IR) is directly absorbed by food products without warming the surrounding air, it has been regarded as a promising method for food processing (Skjöldebrand, 2001). It offers numerous benefits over traditional heating technology, such as uniform heating, adaptability, simple equipment, quick heating time, minimal quality losses and low energy usage (Irakli, et al., 2018; Kathiravan et al., 2007). Yan et al., (2020) demonstrated that IR processing may adequately ensure rice bran stability in terms of desired fatty acid compositions, appropriate enzyme activities and aroma attributes. This process may be designed using appropriate equipment that is simple, easy to control, automatable and safe (Sakai and Hanzawa 1994) while ensuring low processing time, energy consumption and capital expenses (Yilmaz, 2016). Most scientists have used the batch-type infrared stabilizer for rice bran stabilization, which requires more process time with limited capacity. Hence, the objective of the study is to develop a continuous rice bran stabilizer and optimization of infrared heating operational parameters (moisture content, thickness, power density and time) with respect to maximizing the capacity and minimizing the FFA.

Rice bran
 
Freshly milled, full-fatted raw rice bran (Long grain Variety: MTU-2716) was obtained from a rice mill. Prior to stabilization, the raw bran was sieved using an 18-mesh sieve to remove impurities such as rice husk, clay, wood, sawdust, etc. Cleaned rice bran was sealed in an aluminium foil bag and refrigerated at -18°C till the experiment was completed.
 
Stabilization of rice bran     
 
A prototype continuous infrared rice bran stabilizer (Fig 1) has been developed and used for fresh rice bran stabilization. The designed continuous infrared rice bran stabilizer is constructed of a galvanized angular iron frame measuring 15×27×70 cm in size. A 15×20 cm Teflon belt rides on 5 cm diameter and 22 cm long rollers and is driven by a 110 cm pulley with a 2.5 cm driver pulley from a 2 HP variable frequency drive motor to achieve a 44:1 speed ratio. A 1.1m short wave infrared emitter is mounted at a constant height of 15 cm from the bed. In an arc roof construction, a stabilizer chamber of 125×95 cm aluminium sheet with a thickness of 1mm has been used to offer the greatest reflection of infrared radiation on the belt. The radiation intensity of the emitter was altered by changing the voltage using a continuous autotransformer and measured by a pyranometer. A stainless steel feeding chute with dimensions of 50×18×6 cm and a stainless steel product delivery chute with dimensions of 40×12×8 cm are welded to the frame. Before the bran enters the stabilization chamber, the thickness of the bran on the belt will be adjusted using a shutter plate on the feeding chute. Exposer time of rice bran to infrared radiation was adjusted by controlling the belt speed. Treated rice bran was collected through the delivery chute.
 

 
Design of experiments and statistical analysis
 
The performance of continuous rice bran stabilizer was studied for processing at different treatments with different operational parameters namely moisture content (X₁), thickness (X₂), power density (X₃) and time (X₄) to study their effect on different parameters such as FFA and capacity. To reduce many experiments, with four independent variables, a central composite rotatable design (CCRD) and Response Surface Methodology (RSM) has been successfully applied to optimize operational parameters (Table 1). About thirty trials were carried out in accordance with the CCRD (Table 2) and their combined effects from Response Surface Methodology (RSM) were examined using Design Expert-11 software, which yielded optimal values based on the criteria shown in Table 3.
 

 

 

 
Quality analysis of rice bran
 
Oxidative rancidity deterioration is caused by a reaction between lipids and molecular oxygen. The process of inactivating deteriorative enzymes in fresh rice bran is known as stabilization and its shelf life is evaluated in terms of free fatty acids.
 
Moisture content
 
Moisture content was determined by the oven-dry method as the loss in weight due to evaporation from the sample at a temperature of 105°C. The weight loss is represented as the amount of moisture present in the sample (AOAC, 1990).
 
Determination of temperature
 
Treated bran was collected in a paper cup and promptly, the temperature of the bran was measured with non-contact infrared thermometer (RayTemp, UK), which measures temperature over the range of -60°C to 500°C with an assured accuracy of ±1°C.
 
Determination of fat content
 
The total fat content of the rice bran was estimated using Soxhlet extraction (SOCS Plus, SCS06 ASDLS) for 4 h with n-hexane (AOAC, 1984). 
 
Estimation of free fatty acid
 
About 1-10 g of the oil sample was mixed with 100 mL of neutralized alcohol and kept on a hot plate for 10-15 minutes. Two drops of phenolphthalein indicator were added and titrated against 0.1N KOH until consistent pale pink color was obtained (Ermosele, 1994).
   
Determination of energy consumption and power density
 
An electrical meter (Watt-hour meter, Power tech measurement system, Delhi, India) was used to calculate energy use (Power Density). The energy consumption in kilowatt-hours is calculated by continually monitoring the instantaneous voltage and amperes. The IR emitter’s and electric motor’s combined energy consumption was measured in kWh kg^-1 rice bran.
 
Determination of capacity of the IR stabilizer
 
The capacity of the continuous rice bran stabilizer was determined by collecting the rice bran from the collection chute for 5 min and capacity was calculated by following formula.
   
Packaging and storage of treated samples
 
Treated rice bran samples were placed in polyethylene zip-lock bags and the moisture content and FFA have been determined prior to storage at 4°C in the refrigerator. FA levels were taken at 10-day intervals for the best sample.

Statistical analysis Table 2 summarizes the experimental results of capacity, energy demand and FFA content under various treatment conditions. Statistical analysis showed that the proposed model was valid with acceptable R² values for all the responses and non-significant lack of fit. The R² values for capacity, energy demand and FFA content were 0.989, 0.917 and 0.916, respectively. The empirical model more accurately represents the real data when the R² value is higher. The lower the R² value, the less relevant the dependent variables in the model must explain variation in behavior (Little and Hills, 1982; Mendenhall, 1975). All regression models had probability (p) values less than 0.000, indicating that there was no lack of fit.
 
Effect of exposure time and bed thickness on rice bran stabilizer capacity
 
The capacity increased as the thickness of the rice bran bed increased at constant power density and moisture content, as seen predicted response surface plot (Fig 2). The capacity was minimum (6.7 kg/h) at 0.5 cm thickness and 5 min exposer time, whereas it was maximum (40.2 kg/h) at 1 cm bed thickness and 3 min exposer time. The quantity of material carried on the conveying belt at every turn of the belt increases as the thickness of the rice bran bed rises, resulting in an increase in the capacity of the stabilizer. It has been indicated that under constant power density and moisture, capacity declined with infrared exposure time. Infrared exposure time is proportional to the speed of the conveyor belt, i.e., as belt speed increases, infrared exposure time reduces. Since the amount of material released per unit time increases, the speed of the conveying system increases and the capacity of the stabilizer increases. As a result, reducing the exposure time enhances the continuous infrared stabilizer’s capacity.
 

       
The experimental data can be adequately fitted using quadratic model (p-0.001). F-value (201.14) revealed that the capacity had been significantly impacted by the linear terms of independent variables (thickness, time) and their interaction terms. The second-order nonlinear regression model has been developed based on the actual values of the independent variables moisture content (X₁), thickness (X₂), power density (X₃) and time(X₄) for the dependent variable capacity. Equation-(1) provides the derived correlation with real values (after the non-significant components have been eliminated).  
Predicted R² of 0.9695 and adjusted R² of 0.9898 are reasonably in agreement; that is, the difference between the two values is less than 0.2, indicating that the derived model is quite well-fitted. The values of the CV (4.07) and APR (52.749) indicate that the experiment and model had appropriate accuracy and consistency.
 
Energy requirement of stabiliser as a function of exposure time and power density
 
For the interaction of independent variables on energy demand, the model-predicted response surfaces are shown in Fig 3. It has been demonstrated that the energy requirement increased as power density and exposure duration increased at constant bed thickness and moisture content. While power density and exposure duration increased, energy consumption increased as well, peaking at higher power densities and longer exposure times. The energy demand varied between 0.006-0.015 kWhkg^-1. The most often used method of stabilizing rice bran in the literature, extrusion at 130°C for a short period, followed by holding the bran for three minutes at 97-99°C before cooling, was estimated to consume 0.076 kWhkg^-1 of energy. Additionally, it was claimed that extrusion processing for stabilizing rice bran requires a significant capital investment as well as high operational and equipment maintenance expenses, rendering the method unprofitable (Malekian et al., 2000). Therefore, it was revealed that the energy consumption of continuous IR stabilization was identical to that of extrusion. It can be inferred that IR stabilization of rice bran is appropriate for industrial use in terms of energy efficiency, even if the energy consumption of IR stabilization relies on the type and quantity of IR emitters, the bran feeding capacity and the dimensions of the belt.
 

       
The experimental values can be effectively fitted by the quadratic model (p-0.001). The capacity had been significantly impacted by the linear terms of the independent variables, power density, thickness and time, as shown by the F-value (23.86) and the interaction terms between the squares of the variables are significant (p-0.001).
   
The adjusted R² of 0.9169 agrees well with the predicted R² of 0.7890. The APR (17.551) confirmed that the model has sufficient accuracy and reliability.
 
Free fatty acid response to independent variables
 
Fig 4 and 5 depict the model-predicted response surfaces for independent variables on FFA. The FFA decreased while rising power density and time at the constant thickness and moisture content, however, the trend of the graph is nonlinear as seen in Fig 4. Better lipase inactivation resulted through rising power density and infrared exposure time, but treated bran exhibits undesirable visual and sensory changes. To prevent undesirable changes in rice bran, exposure time can be increased at low radiation intensities or decreased at elevated radiation intensities. Yilmaz et al., (2014) reported that stabilization between 200-400 W infrared radiation for 10 min is not enough to inhibit hydrolytic rancidity, stabilization at these powers levels may take longer time to achieve better results. Short process time of 1 min was not sufficient to inactivate lipases even at high radiation intensities, considering stabilization between 800 and 900 W is an unacceptable strategy. Moreover, process durations longer than 1 min generated unpleasant sensory and visual changes in the bran.
 

 

 
At constant power density and time, FFA increased with increase in thickness and moisture content as shown in Fig 5. The FFA content was low at lower bed thickness and moisture content and found to increase relatively at a slower rate as thickness and moisture content rises to 0.8 cm and 15% respectively, further rise in bed thickness and moisture content caused rapid increase in the rate of FFA content. Since infrared radiation has a lower penetration depth, increasing the thickness of the bed causes uneven exposure of the rice bran along the bed thickness, resulting variation in the rice bran-free fatty content. Sandu (1986) reported that, even with short wavelength infrared radiation, the depth of penetration reported is relatively low, with depths seldom exceeding a few thousandths of an inch.
       
According to the model (F-value: 23.56), the FFA content had been substantially influenced by power density, moisture content, thickness and time and the interaction terms of the squares of the variables are significant (p 0.001). The nonlinear second-order regression equation is illustrated below:
   
The adjusted R² of 0.9159 and the predicted R2 of 0.7496 are reasonably compatible. The APR (14.625) greater than 4 indicates the experiment’s and model’s sufficient accuracy and reliability.
 
Optimum conditions
 
To ensure maximum capacity, efficiency and minimum energy demand and FFA, optimal conditions for continuous infrared rice bran stabilizer were established. The second-order polynomial regression equations were solved in Design Expert 11 using sequential quadratic programming. The optimum values obtained by substituting the respective coded values are 600W, 12% moisture content, 0.5cm thickness and 3 min exposure time. At these optimum conditions, capacity, efficiency, energy demand and FFA were 17.85 kg/h, 20.12%, 0.006kW-h/kg and 5.01% respectively.
 
Changes in FFA of infrared stabilized rice bran during the storage
 
The best sample obtained in optimization (600W, 12% moisture content, 0.5 cm thickness and 3 min time) was kept for storage in zip lock polyethylene packs at 4°C. During storage, the FFA content of rice bran was studied at 10 days intervals for 30 days.
       
The FFA content of raw rice bran increased from 3.32% initiallyto 22.08% at the end of the month during storage at 4°C. The FFA content of rice bran IR stabilized at different treatment conditions was below 6% after treatment (Table 4). However, the FFA level of rice bran stabilized at 450 Wm^-2 for 4 min and 700 W/m², 15% mc, 0.75 cm was above 6% after a month of storage. Also, the FFA content of rice bran stabilized at 600 W/m², 12% mc, 0.5 cm thickness for 3 min was 5.56% after 30 days of storage. Considering the initial FFA level of 3.32%, it can be stated that IR stabilization is effective in terms of preventing hydrolytic rancidity and that, by optimizing the operational parameters of stabilization; the shelf life of rice bran can be extended. Literature data on the FFA content of rice bran are highly variable. In raw bran, FFA increased rapidly throughout the course of 4 weeks of storage at 25°C, according to (Ramezanzadeh, 1999). Malekian et al., (2000) reported that the raw bran held in zip-lock bags for eight weeks had a rise in FFA content from 3.7 to 22.2%. In raw rice bran, FFA content was found to be 9.5% initially, raised to 96.8% over 345 days of storage (Mujahid, 2005).
 

A continuous infrared rice bran stabilizer has been developed and response surface methodology was effectively applied to determine optimal infrared radiation treatment conditions for stabilizing rice bran while using capacity, efficiency, energy demand and FFA content as responses. The results indicated that the 600 W power, 12 per cent moisture content, 3 mm thickness and 3.0 minutes exposure time were the ideal treatment conditions for stabilizing rice bran. The capacity, efficiency, energy demand and FFA under these optimum conditions were 17.85 kg/h, 20.12 per cent, 0.006 kW-h/kg and 5.01 per cent, respectively. After 30 days of storage, the FFA content of raw rice bran increased from 3.32 per cent to 5.56 per cent.