Experimental Procedures
Materials
Degummed, neutralized and bleached soybean oils, rapeseed oils, maize oils and sunflower seed oils were used for this study. Crude oils were obtained from solvent extracted or screw pressed oil from Wilmar Group, and the crude oils conformed the internal standards of TFA≤0.10% were selected. All chemicals used in this study for analyses were of analytical grade and purchased from Sigma-Aldrich (St. Louis, MO, USA).
Conventional deodorization
Neutralized and bleached oils were deodorized in an industrial deodorizer soft column, which was designed by Alfa Laval and has been widely used in industrial vegetable oil refining in recent years. The incoming bleached oils were firstly deaerated and then heated by heat exchanger and high-pressure steam to the final operating temperature of 248-250°C under reduced pressure (<2 mbar). Steam was applied as the stripping gas (approximately 1% of the oil weight). From the final heater the oils flowed by gravity into the deodorizer, through the deodorizer and out to be cooled by heat exchange with incoming bleaching oil. During flow through the deodorizer, the oil firstly underwent a continuous stage for fast FFA removal in a structured packing unit for a short time (5-10 min), and then a long time (90-120 min) holding stage for heat bleaching and deodorization.
Optimization of deodorization temperature
Deodorization temperatures were preliminarily optimized between 220-240°C in the soft column or tray column, in order to investigate the relationship between temperatures and TFA generation. Meanwhile deodorization temperatures were optimized between 225-235°C in a packed column, to investigate the relationship between temperatures and TFA generation. Other parameters (deodorization time, vacuum, stripping steam etc.) were not simultaneously adjusted as variable and remained at their best according to the actual conditions in each plant, to obtain deodorized oils with low FFA, light color and few odors.
Dual columns with dual temperatures deodorization
A new DCDT deodorizer was designed, upgraded and implemented in a large number of plants of Wilmar Group. This new deodorizer contained dual independent columns (tray column and packed column). Some plants combined independent tray column and packed column together, and some others firstly separated the soft column into a packed column and tray column, and then strung them in order (Fig. 1). During vacuum deodorization, oil from bleaching was heated by heat exchange with oil out from packed column to a moderate temperature (205°C for soybean oil and rapeseed oil, 210°C for maize oil and sunflower seed oil), after which it first passed through the tray column to undergo heat bleaching and deodorization for a long time (60-90 min). When the oil came out from the tray column, it was heated again by high-pressure steam to a higher temperature (225°C for soybean oil and rapeseed oil, 230°C for maize oil and sunflower seed oil) and then passed through the packed column for a short time (5-10 min) to undergo final deacidification and deodorization. The deodorized oil was then cooled by heat exchange with bleached oil to below 60°C.
Trans fatty acids analysis
TFA content was quantitatively analyzed by gas chromatography (GC) according to the national standard of China (GB 5009.257-2016).
Free fatty acids and acid value determination
Free fatty acids contents of oils were calculated as acid value divided by 1.99 according to the AOCS Official Method 940.28, by titration. Acid values were determined according to the AOCS Official Methods Cd 3d-63.
Peroxide values determination
Peroxide values were determined according to the AOCS Official Methods Cd 8-53.
COLOR analysis
Lovibond color was measured using Lovibond Tintometer (Model F, Lovibond, England) according to the ISO method 15305:1998. The R value was calculated as follows: R= (R+Y/10)/2, where R and Y represent the red unit and yellow unit, respectively. The type of cuvette used to determine the color has a path length of 133.4 mm.
Sensory evaluation
Ten trained panelists sniffed deodorized oil samples (150 g), which were filled in opaque bottles (220 ml) to avoid the influence of the color in the decision. These samples included the deodorized oils obtained at different deodorization conditions. Bottles were capped for more than 15 min prior to being tested. After smelling each sample, panelists were asked to score the samples according to the following four basic options of scale: “it does not smell”, “it smells, or seems to smell of refined oil”, “it slightly smells of undesirable odors”, and “it smells of undesirable odors” (coded with numbers from 1 to 4, respectively).