Deodorization temperature optimization
Besides the design of the deodorizer, deodorization temperature is the most critical factor for oil quality among the four deodorization influencing parameters (temperature, time, pressure, stripping steam). A high temperature is beneficial for an efficient stripping of FFA and heat bleaching, but it also results in a higher loss of tocopherols and sterols (Hamn et al., 2013). More importantly, a high deodorization temperature is the most critical parameter for TFA generation. Reported activation energy values for the thermal cis-trans-isomerization of linoleic and linolenic acid are rather low, which is an indication that TFA are relatively easily formed at an elevated deodorization temperature (Shahidi, 2005). The degree of reaction will vary with the degree of fatty acid polyunsaturation. The higher the degree of polyunsaturation, the greater the tendency of the cis-acid to trans-acid formation. Different studies showed that the relative isomerization rate can be expressed as follows: C18:3(100)≥C18:2(10)≥C18:1(1) (Hénon et al., 1999; León-Camacho et al., 2001). Consequently, oils with a relative high linolenic acid content, such as soybean oil and rapeseed oil, are more sensitive to cis-trans-isomerization than maize oil and sunflower seed oil during deodorization. Generally, trans-formation is negligible below 220°C, whereas it becomes significant between 220°C and 240°C, and exponential above 240°C (Shahidi, 2005). Industrial deodorizations are usually operated at temperatures between 240°C and 260°C, and approximately 1-3% of TFA are found in full refined oils.
However, limited lab-scale deodorization cannot precisely simulate industrial deodorization. In order to confirm the industrial deodorization parameters for zero-TFA vegetable oil production with good quality, dozens of pilot trials were firstly carried out with soybean oil among plants to optimize deodorization temperatures of existing deodorizers (tray column, packed column and soft column). Meanwhile, the other parameters (deodorization time, vacuum, stripping steam etc.) were not simultaneously adjusted as variable and remained as usual at their best according to the actual conditions in each plant. Fig.2 shows results from the optimization of deodorization temperature in packed columns (Fig. 2a) or tray columns & soft columns (Fig.2b), and the relationship between TFA content and deodorization temperature. Since the oil would stay a long time in tray columns as well as in soft columns, the results were combined into Fig.2b. In contrast the oil would stay in packed columns for a much shorter time. The trial temperature was variable among 220-240°C. Approximately one-third of trial batches achieved TFA≤0.3%. Interestingly, the majority of them were deodorized by packed column, while TFA were considerably more than 0.30% with the use of tray columns and soft columns, and increased rapidly to an average of 1.34% when the temperature was increased to 240°C. Therefore, to achieve the goal of TFA≤0.3%, the upper limit of deodorization temperature for packed columns and tray columns & soft columns were significantly different. The temperature for packed columns was suggested to be no more than 235°C, while the temperature for tray columns & soft columns should be lower than 220°C. This might be due to the significant differences between deodorization times in columns with trays compared to packed columns.
The deodorization temperature was furthermore optimized in this new DCDT system using soybean oil. The results showed that temperatures for packed columns and tray columns should be controlled to a maximum of 225°C and 205°C respectively (Fig.3). However, the deodorization temperature cannot be infinitely low. The volatility of a given components (FFA, tocopherols, sterols, etc.) is expressed by its vapor pressure, which increases with increasing temperature. The lower the vapor pressure, the lower the volatility and, thus, it is more difficult to remove the components from the oil. For example, a moderate temperature deodorization was operated in neutralized red palm oil but found that FFA and all volatile components of odor could not be reduced although the oil was deodorized for 2 hours at 150°C (Riyadi et al., 2016). Therefore, to balance low TFA and oil physicochemical quality, the deodorization temperature for soybean oil in the DCDT system was calculated to be 225°C for packed columns and around 205°C for tray columns. The temperatures for rapeseed oil, maize oil and sunflower seed oil were then respectively optimized and customized, which are shown in Table 1. Soybean oil and rapeseed oil have similar, but higher linolenic acid content (respectively 4.2-11% and 5-14%) than the other oils, so the temperatures for them were designed the same. Due to the lower content of linolenic acid in maize oil (≤2%) and sunflower seed oil (≤0.3%), the temperatures for these two oils can be set higher to further ensure the physicochemical quality of deodorized oils. Comparing the new zero-TFA oils with refined oils collected from the market ten years ago (internal unpublished data) it can be shown that TFA content decreased by more than 90% (Fig.4). Furthermore, comparing zero-TFA oils with currently available conventional refined oils an 80% decrease in TFA content is observed (Fig.5).