**5. Molecular distillation**

456 Recent Trends for Enhancing the Diversity and Quality of Soybean Products

temperature using immobilized lipases have been also reported (Weber *et al.*, 2001). In this case neither organic solvent, nor water or any drying reagent such as molecular sieves, are used. This and others studies (Shimada, 2000) showed that in the process of esterification of sterols with free fatty acids, the best results are obtained with *Candida rugosa* lipase and *Pseudomonas* sp. However, enzymatic conversion of FFAs to FAMEs or FAEEs is carry out frequently in the presence of *Candida antactica* lipase or *Alcalygenes* sp. (Torres, 2007), (Nagao

In the following paragraphs some examples of methodologies using enzymes in the pretreatment of SODD are described. Most of them will be further developed in following

Shimada and coworkers (Shimada, 2000) converted sterols from SODD to fatty acid sterol esters and completely hydrolyzed acylglycerols by applying lipase reactions (*Candida rugosa*  or *Pseudomonas* sp., at 35 ºC for 24 h) to the purification of tocopherols and sterols, resulting in an efficient fractionation of tocopherols and sterols as fatty acids steryl esters (FASEs) by short-path distillation. This process included the drawback that FFA and tocopherols were not efficiently fractionated because the boiling points of the two substances were close. This problem could be solved by conversion of the FFA to their corresponding methyl esters. An attempt to develop a reaction system in which the methyl esterification of FFA proceeded simultaneously with the conversion of sterols to FASEs and the hydrolysis of acylglycerols

Nagao and coworkers (Nagao, 2005) and Watanabe and coworkers (Watanabe, 2004) have applied a procedure based on using a lipase to promote the simultaneous esterification of sterols with free fatty acids and hydrolysis of acylglycerols before the esterification of the free fatty acids with methanol. These authors use *Candida rugosa* lipase for the purification of tocopherol in SODD. Watanabe and coworkers reported 80% conversion of the initial sterols to FASEs, complete hydrolysis of the acylglycerols, and a 78% decrease in the initial FFA content by methyl esterification in 40 h. Tocopherols did not change throughout the process. Distillation of the reaction mixture purified tocopherols to 76.4% (recovery, 89.6%) and sterols to 97.2% as FASEs (recovery, 86.3%). Nagao and coworkers reported a more effective sterols esterification, with a degree of esterification reached 95%. The second-step reaction was then conducted at 30 ºC for 20 h with *Alcaligenes* sp. lipase. 95% FFAs were converted to FAME, and steryl esters synthesized by the first-step reaction were not reconverted to free sterols. Finally, tocopherols and steryl esters were purified from the reaction mixture by short-path distillation. Tocopherols were purified to 72% (yield, 88%) and steryl esters were purified to 97% (yield, 97%). One of the main disadvantages of this method is that the

remaining free fatty acids are not completely separated from the tocopherols.

Lipase-catalyzed esterification of sterols and ethyl esterification simultaneously, are governed by the concentration of water present. The degree to which esterification of sterols occurs relative to ethyl esterification requires to attain a balance not always easy to achieve because the presence of an excess of water favours hydrolysis, whereas esterification predominates when a very limited amount of water is present (Marangoni & Rousseau, 1995). By appropriate choice of reaction conditions, however, it is possible to separate the sterol esterification and ethyl esterification in time or space. It is then possible to optimize each of these reactions independently, thereby minimizing costs or improving the yield of

This is precisely the procedure carried out by Torres and coworkers (Torres, 2007), who proposed a two-step enzymatic procedure to obtain FASEs, tocopherols, and fatty acid ethyl

*et al.*, 2005).

sections:

has been also reported (Watanabe *et al.*, 2004).

the desired final reaction products.

Most of the substances that are present in soybean deodorizer distillate are molecules of high molecular weight and thermally sensitive. These properties hinder the separation or purification of these compounds through traditional methods, because they are decomposed when subjected to high temperatures.

An alternative separation/purification procedure of such products is the use of molecular or short-path distillation. It consists of transferring molecules from the surface of an evaporating liquid to the cooled surface of a condenser through a short path, which is on the order of 2-5 cm. In this process, distillation of heat-sensitive materials is accompanied by only negligible thermal decomposition (Lutisan *et al.*, 2002) because materials, by using high vacuum, are submitted to relatively reduced temperatures, and short residence times (Lutisan, 2002) inside the equipment. Furthermore, this process has advantages over other techniques that use toxic or flammable solvents as the separating agent, avoiding toxicity and environmental problems.

The combination of a small distance between the evaporator and the condenser of only a few centimetres and a high vacuum in the distillation gap, results in a specific mass transfer mechanism with evaporation outputs as high as 20–40 gm−2 s−1 (Cvengros *et al.*, 2000). Under these conditions (e.g., short residence time and low temperature), distillation of heatsensitive materials is accomplished without or only negligible thermal decomposition. Therefore, molecular distillation shows potential in the separation, purification and/or concentration of natural products, usually constituted by complex and thermally sensitive molecules such as tocopherols.

In lipid chemistry, it has been used for the purification of monoacylglycerols (Szelag & Zwierzykowski, 1983), recovery of carotenoids from palm oil (Batistella & Wolf Maciel, 1998), fractionation of polyunsatured fatty acids from fish oils (Breivik *et al.*, 1997), recovery of squalene (Sun *et al.*, 1997), and recovery of tocopherols (Batistella *et al.*, 2002), among others.

Normally, SODD have a high content of FFA and acylglycerol. To increase the separation efficiency of the compounds of interest, esterification and/or transesterification reactions are usually carried out prior to molecular distillation. Free fatty acids and acylglycerols are converted to fatty acid methyl esters, which are more easily removed by vacuum distillation

Extraction and Enzymatic Modification of

FASEs (86.3% recovery).

(fourth) residue.

Functional Lipids from Soybean Oil Deodorizer Distillate 459

tocopherols to 76.4 wt% purity (89.6% recovery) and free phytosterols to 97.2 wt% purity as

Nagao and coworkers (Nagao, 2005) carried out similar steps of isolation of tocopherols and free phytosterols as FASEs from SODDTSC. SODDTSC were first treated with *Candida rugosa* lipase (250 U/g activity) to convert free phytosterols to FASEs at 40 ºC for 24 h, achieving about 95% conversion. Unreacted FFAs contained in the reaction mixture was then converted to FAMEs by *Alcaligenes* sp. lipase at 30 ºC for 20 h, achieving 95% conversion. Reaction mixture was then subjected to a four-stage molecular distillation (160 ºC and 0.2 mm Hg, 175 ºC and 0.2 mm Hg, 230 ºC and 0.02 mm Hg, and 240 ºC and 0.02 mm Hg, respectively) to isolate tocopherols (72.3 wt% purity, 87.6% recovery) in the third distillate fraction and free phytosterols as FASEs (97 wt% purity, 97% recovery) in the last

Jacobs (Jacobs, 2005) proposed a method for recovering tocotrienols from fatty acid distillate FAD, which initially contained 1 wt% tocopherols and 0.3 wt% free phytosterols, by stripping FFAs (condition: 0.5–1.5 mm Hg, 180–240 ºC, and 0.5–1.5 min) to form a first stripped product. Short path distillation of the first stripped product gave a first distillate. Saponifying the second stripped product resulted in a saponified product with all FFA converted to FAMEs. A second short path distillation of the saponified product generated a second distillate without FAMEs. Solvent wintering (via filtration) of the second distillate gave a stripped filtrate. The stripped filtrate from the previous step is subjected to a third short path distillation at a temperature of 180° and an absolute pressure of 0.01 mm Hg. The resulting final tocotrienol product, about 1% of the original feed, contains from about 50% tocotrienols, 1% sterols and 49% other unsaponifiables and unknowns. Additionally, the

Fizet (Fizet, 1996) esterified free phytosterols from deodorizer distillate with FFAs at either 180 ºC for 2.5 h or 250 ºC for 1.5 h. Esterification product was then distilled at 120–150 ºC and 0.08 mm Hg to obtain a residue containing mostly tocopherols and FASEs and a distillate containing mostly fatty acids. The residue was then distilled again at 200–220 ºC and 0.1 mbar to obtain a distillate containing mostly tocopherols and a residue containing mostly FASEs. Tocopherols enriched in distillate were then subjected to an ion exchange chromatography and FASEs enriched in residue were then subjected to an acid-catalyzed

Even so, some authors (Martins *et al.*, 2006) have been trying to achieve an efficient FFA separation from SODD with the lowest loss of tocopherols by molecular distillation, without preliminary steps. This separation is difficult to achieve although is technologically viable at least at lab scale, due to the differences between molecular weights and vapour pressures of FFA (MW 180 g·mol-1, VP at 200 ºC = 4 mm Hg) and tocopherols (MW 415 g·mol-1, VP at 200 ºC = 0.15 mm Hg). Martins and coworkers (Martins, 2006) employed molecular distillation at 160 ºC, 7.5×10-4 mm Hg, and 10.4 g/min feed flow rate to remove FFAs into the distillate fraction and obtain a residue fraction, which contained 6.4 wt% FFAs and 18.3 wt% tocopherols from a SODD feed which contained 57.8 wt% FFAs and 8.97 wt% tocopherols. They succeeded in removing 96.16% FFAs and recovering 81.23% tocopherols. Martins and coworkers (Martins *et al.*, 2005) reported the isolation of tocopherols by first converting acylglycerols in SODD into FFAs through saponification at 65 ºC followed by acidulation, and then submitting the unsaponifiable product to five stages of molecular distillation. They succeeded in enriching tocopherols (34.14% purity) by 5.8 times. The major disadvantage in

final product contains from about 15% to about 30% tocopherols.

transesterification with methanol to produce free phytosterols.

these cases is the residual free fatty acids in the tocopherol mixture.

due to their higher vapor pressure than those of the corresponding free fatty acids and acylglycerols. However, this step made the whole process more complicated and laborintensive when compared with the saponification process. Other problem is that the separation of tocopherols from phytosterols is difficult because they have similar molecular weights, boiling points and vapor pressure, and, consequently, they are distillated together (Ghosh & Bhattacharyya, 1996).

Different processes have been proposed in the literature to eliminate FFA by molecular distillation and purify tocopherols and phytosterols. Most of them include a preliminary chemical or enzymatic treatment step.

Ramamurthi and McCurdy (Ramamurthi & McCurdy, 1993) studied the pretreatment of deodorizer distillate using a lipase-catalyzed esterification reaction to convert FFA into methyl esters, followed by vacuum distillation (1-2 mm Hg) to remove them and concentrate tocopherols and sterols (recoveries were over 90%).

Hirota and coworkers (Hirota *et al.*, 2003) isolated naturally occurring Fatty Acid Steryl Esters (FASEs) from SODD. SODD was firstly subjected to molecular distillation at 250 ºC and 0.02 mm Hg to obtain a residue which was rich in DAGs and TAGs, and steryl esters. Enzymatic lipolysis was then conducted to specifically hydrolyze DAGs and TAGs at 35 ºC for 24 h, resulting in a mixture from which fatty acid steryl esters were later purified using a two-stage molecular distillation (180 ºC and 0.2 mm Hg, and 250 ºC and 0.02 mm Hg). The recovery and purity of FASEs were about 87.7% and 97.3 wt%, respectively.

Purification of tocopherols from SODD was carried out by Shimada and coworkers (Shimada, 2000). SODD was distilled using molecular distillation at 250 ºC and 0.02 mm Hg and the resulting distillate was used as a starting material. Sterols in SODD were converted to FA sterol esters and acylglycerols were completely hydrolyzed by applying lipase reactions. FASEs were recovered as residue from the reaction mixture via molecular distillation at 250 ºC and 0.2 mm Hg. However, the last stage of molecular distillation failed to separate FFAs and tocopherols. A second esterification of free phytosterols was applied at 35 ºC for 24 h, followed by another four-stage molecular distillation (160 ºC and 0.2 mm Hg, 200 ºC and 0.2 mm Hg, 230 ºC and 0.04 mm Hg, and 255 ºC and 0.03 mm Hg) which yielded tocopherols with purity and recovery of about 65.3 wt% and 54.6%, respectively.

Watanabe and coworkers (Watanabe, 2004) isolated tocopherols and free phytosterols as their esters from SODD tocopherol/sterol concentrate (SODDTSC). SODDTSC was obtained via molecular distillation at 240 ºC and 0.02 mm Hg, resulting in a distillate rich in tocopherols and free phytosterols (SODDTSC), and a residue rich in FASEs, DAGs, and TAGs. SODDTSC, which contained also MAGs, DAGs, FFAs, and unidentified hydrocarbons, were then subjected to a two-step in situ enzymatic reaction. SODDTSC were treated with *Candida rugosa* lipase (200 U/g activity) to convert free phytosterols to FASEs, acylglycerols (MAGs and DAGs) to FFAs, and FFAs to FAMEs at 30 ºC for 40 h, achieving 80% conversion of the initial sterols to FA steryl esters, complete hydrolysis of the acylglycerols, and a 78% decrease in the initial FFA content by methyl esterification. Tocopherols did not change throughout the process. To enhance degree of steryl and methyl esterification, FASEs and FAMEs enriched in the reaction product were then removed by a two-step molecular distillation. In the first step, FAMEs was removed in the distillate (160 ºC and 0.2 mm Hg). In the second step (240 ºC, 0.2 mm Hg), FASEs was isolated in the residue and the distillate containing tocopherols, free phytosterols, and FFAs were treated again with lipase. A three-step molecular distillation of the reaction mixture purified

due to their higher vapor pressure than those of the corresponding free fatty acids and acylglycerols. However, this step made the whole process more complicated and laborintensive when compared with the saponification process. Other problem is that the separation of tocopherols from phytosterols is difficult because they have similar molecular weights, boiling points and vapor pressure, and, consequently, they are distillated together

Different processes have been proposed in the literature to eliminate FFA by molecular distillation and purify tocopherols and phytosterols. Most of them include a preliminary

Ramamurthi and McCurdy (Ramamurthi & McCurdy, 1993) studied the pretreatment of deodorizer distillate using a lipase-catalyzed esterification reaction to convert FFA into methyl esters, followed by vacuum distillation (1-2 mm Hg) to remove them and

Hirota and coworkers (Hirota *et al.*, 2003) isolated naturally occurring Fatty Acid Steryl Esters (FASEs) from SODD. SODD was firstly subjected to molecular distillation at 250 ºC and 0.02 mm Hg to obtain a residue which was rich in DAGs and TAGs, and steryl esters. Enzymatic lipolysis was then conducted to specifically hydrolyze DAGs and TAGs at 35 ºC for 24 h, resulting in a mixture from which fatty acid steryl esters were later purified using a two-stage molecular distillation (180 ºC and 0.2 mm Hg, and 250 ºC and 0.02 mm Hg). The

Purification of tocopherols from SODD was carried out by Shimada and coworkers (Shimada, 2000). SODD was distilled using molecular distillation at 250 ºC and 0.02 mm Hg and the resulting distillate was used as a starting material. Sterols in SODD were converted to FA sterol esters and acylglycerols were completely hydrolyzed by applying lipase reactions. FASEs were recovered as residue from the reaction mixture via molecular distillation at 250 ºC and 0.2 mm Hg. However, the last stage of molecular distillation failed to separate FFAs and tocopherols. A second esterification of free phytosterols was applied at 35 ºC for 24 h, followed by another four-stage molecular distillation (160 ºC and 0.2 mm Hg, 200 ºC and 0.2 mm Hg, 230 ºC and 0.04 mm Hg, and 255 ºC and 0.03 mm Hg) which yielded

Watanabe and coworkers (Watanabe, 2004) isolated tocopherols and free phytosterols as their esters from SODD tocopherol/sterol concentrate (SODDTSC). SODDTSC was obtained via molecular distillation at 240 ºC and 0.02 mm Hg, resulting in a distillate rich in tocopherols and free phytosterols (SODDTSC), and a residue rich in FASEs, DAGs, and TAGs. SODDTSC, which contained also MAGs, DAGs, FFAs, and unidentified hydrocarbons, were then subjected to a two-step in situ enzymatic reaction. SODDTSC were treated with *Candida rugosa* lipase (200 U/g activity) to convert free phytosterols to FASEs, acylglycerols (MAGs and DAGs) to FFAs, and FFAs to FAMEs at 30 ºC for 40 h, achieving 80% conversion of the initial sterols to FA steryl esters, complete hydrolysis of the acylglycerols, and a 78% decrease in the initial FFA content by methyl esterification. Tocopherols did not change throughout the process. To enhance degree of steryl and methyl esterification, FASEs and FAMEs enriched in the reaction product were then removed by a two-step molecular distillation. In the first step, FAMEs was removed in the distillate (160 ºC and 0.2 mm Hg). In the second step (240 ºC, 0.2 mm Hg), FASEs was isolated in the residue and the distillate containing tocopherols, free phytosterols, and FFAs were treated again with lipase. A three-step molecular distillation of the reaction mixture purified

(Ghosh & Bhattacharyya, 1996).

chemical or enzymatic treatment step.

concentrate tocopherols and sterols (recoveries were over 90%).

recovery and purity of FASEs were about 87.7% and 97.3 wt%, respectively.

tocopherols with purity and recovery of about 65.3 wt% and 54.6%, respectively.

tocopherols to 76.4 wt% purity (89.6% recovery) and free phytosterols to 97.2 wt% purity as FASEs (86.3% recovery).

Nagao and coworkers (Nagao, 2005) carried out similar steps of isolation of tocopherols and free phytosterols as FASEs from SODDTSC. SODDTSC were first treated with *Candida rugosa* lipase (250 U/g activity) to convert free phytosterols to FASEs at 40 ºC for 24 h, achieving about 95% conversion. Unreacted FFAs contained in the reaction mixture was then converted to FAMEs by *Alcaligenes* sp. lipase at 30 ºC for 20 h, achieving 95% conversion. Reaction mixture was then subjected to a four-stage molecular distillation (160 ºC and 0.2 mm Hg, 175 ºC and 0.2 mm Hg, 230 ºC and 0.02 mm Hg, and 240 ºC and 0.02 mm Hg, respectively) to isolate tocopherols (72.3 wt% purity, 87.6% recovery) in the third distillate fraction and free phytosterols as FASEs (97 wt% purity, 97% recovery) in the last (fourth) residue.

Jacobs (Jacobs, 2005) proposed a method for recovering tocotrienols from fatty acid distillate FAD, which initially contained 1 wt% tocopherols and 0.3 wt% free phytosterols, by stripping FFAs (condition: 0.5–1.5 mm Hg, 180–240 ºC, and 0.5–1.5 min) to form a first stripped product. Short path distillation of the first stripped product gave a first distillate. Saponifying the second stripped product resulted in a saponified product with all FFA converted to FAMEs. A second short path distillation of the saponified product generated a second distillate without FAMEs. Solvent wintering (via filtration) of the second distillate gave a stripped filtrate. The stripped filtrate from the previous step is subjected to a third short path distillation at a temperature of 180° and an absolute pressure of 0.01 mm Hg. The resulting final tocotrienol product, about 1% of the original feed, contains from about 50% tocotrienols, 1% sterols and 49% other unsaponifiables and unknowns. Additionally, the final product contains from about 15% to about 30% tocopherols.

Fizet (Fizet, 1996) esterified free phytosterols from deodorizer distillate with FFAs at either 180 ºC for 2.5 h or 250 ºC for 1.5 h. Esterification product was then distilled at 120–150 ºC and 0.08 mm Hg to obtain a residue containing mostly tocopherols and FASEs and a distillate containing mostly fatty acids. The residue was then distilled again at 200–220 ºC and 0.1 mbar to obtain a distillate containing mostly tocopherols and a residue containing mostly FASEs. Tocopherols enriched in distillate were then subjected to an ion exchange chromatography and FASEs enriched in residue were then subjected to an acid-catalyzed transesterification with methanol to produce free phytosterols.

Even so, some authors (Martins *et al.*, 2006) have been trying to achieve an efficient FFA separation from SODD with the lowest loss of tocopherols by molecular distillation, without preliminary steps. This separation is difficult to achieve although is technologically viable at least at lab scale, due to the differences between molecular weights and vapour pressures of FFA (MW 180 g·mol-1, VP at 200 ºC = 4 mm Hg) and tocopherols (MW 415 g·mol-1, VP at 200 ºC = 0.15 mm Hg). Martins and coworkers (Martins, 2006) employed molecular distillation at 160 ºC, 7.5×10-4 mm Hg, and 10.4 g/min feed flow rate to remove FFAs into the distillate fraction and obtain a residue fraction, which contained 6.4 wt% FFAs and 18.3 wt% tocopherols from a SODD feed which contained 57.8 wt% FFAs and 8.97 wt% tocopherols. They succeeded in removing 96.16% FFAs and recovering 81.23% tocopherols. Martins and coworkers (Martins *et al.*, 2005) reported the isolation of tocopherols by first converting acylglycerols in SODD into FFAs through saponification at 65 ºC followed by acidulation, and then submitting the unsaponifiable product to five stages of molecular distillation. They succeeded in enriching tocopherols (34.14% purity) by 5.8 times. The major disadvantage in these cases is the residual free fatty acids in the tocopherol mixture.

Extraction and Enzymatic Modification of

2002).

column.

Functional Lipids from Soybean Oil Deodorizer Distillate 461

recycling the solvent does not endanger the viability of the process. The value of the rate of return on investment and time of return on investment for the process that does not recycle the carbon dioxide is higher than those of recycling the solvent. This is due to the compression cost that represents more than 59% of the total cost of the production (Mendes,

 SODD as such will not be feasible to work with SC-CO2 for the tocopherol enrichment, owing to its poor SC-CO2 solubility. So, to concentrate tocopherols from SODD, pretreatment of the raw material, including the esterification of free fatty acids and the removal of sterols with alcohol recrystallization, is needed to obtain the primary tocopherols concentrate with improve solubility in SC-CO2 (Shishikura *et al.*, 1988). For that, triglycerides and FFAs which constitute a major component in SODD have to be chemically

Several researchers have tried to concentrate tocopherols from SODD by supercritical CO2 (Lee, 1991), (Brunner *et al.*, 1991), (Brunner, 1994b), (Zhao *et al.*, 2000), (Nagesha, G. K. *et al.*, 2003), (Fang *et al.*, 2007), but the operation parameters, especially pressure, differ from author to author. Moreover, in all the cases, tocopherols content of the extract depended on

The interest in the tocopherol concentration using supercritical fluid extraction started with Lee and coworkers (Lee, 1991) followed by Brunner and coworkers (Brunner, 1991) and Brunner (Brunner, 1994b). The operational conditions used varied from 35 to 90 °C and from 200 to 400 bar using extractors or countercurrent columns. Lee and coworkers (Lee, 1991) attempted to modify soybean sludge chemically, to improve the solubility in SC-CO2. A simple batch process was utilized to recover tocopherols at 40% concentration at a pressure of 400 bar from the esterified soybean sludge which initially contained 13-14% tocopherols. The solubility of the esterified soybean sludge in supercritical carbon dioxide was more than 4 to 6 times higher than that of the sterols. Brunner and coworkers (Brunner, 1991) obtained a higher enrichment of tocopherols from SODD using supercritical carbon dioxide as a solvent compared to results obtained by Lee and coworkers This group recovered tocopherols from a model mixture of squalene, tocopherols, and sterols using two continuous countercurrent fractionation columns. Squalene was separated from the model mixture in the first column. Sterols were removed from the bottom of the second column, resulting in 85-95% concentration of tocopherol being obtained at the top of the second

These works concluded that the fatty acids are extracted initially and the tocopherols are enriched inside the extractor. The results also indicated that the solubility of the tocopherols

Chang and coworkers (Chang, 2000) worked on the separation of several SODD components. Their supercritical fluid extraction apparatus had a separation and an extraction unit. Free fatty acids, squalene and tocopherols were recovered in the extract and the sterols were recovered in the raffinates. The average tocopherol concentration factor was 1.38, which means that the mixture in the extract did not separate. However, the author mentioned that with the increase of CO2 volume, the separation factor can reach 1.7, but the poor increase in the concentration factor does not justify the raise in gas volume. The following research groups focused on the separation of the problematic pairs by supercritical fluids using synthetic mixtures: (Mendes, 2005), (Mendes *et al.*, 2000), (Wang *et* 

is intermediary when compared to the solubilities of squalene and stigmasterol.

modified to obtain free fatty acids and then their methyl esters by esterification.

the composition and properties of the natural matrix.

*al.*, 2004), (Nagesha, 2004) and (Nagesha, 2003).

Among the great variety of processes that have been patented for the purification of the compounds of the SODD, only the processes of esterification of fatty acids and acylglycerols with methanol or ethanol followed by high vacuum distillation, have been developed on a commercial scale for the concentration of tocopherols (Takagi & Kai, 1984), (Su-Min *et al.*, 1992), (Yong-Bo *et al.*, 1994), (Rohr & Trujillo-Quijano, 2002). These processes are the most time efficient and economical methods, however high purity of sterols or tocopherols cannot be achieved due to the similar boiling points of these two substances.

In the case of separation by distillation of unsaponifiable valuable products of SODD subjected to saponification, the difference between the boiling point of volatile products, such as unsaponifiable components, and the boiling point of the sodium and potassium organic acid soaps is so great that separation is theoretically possible at a high level of efficiency. However, a problem related to this separation technique is that the soaps have a very high melting point, close to the decomposition temperature of the sodium or potassium soaps (i.e. the sodium or potassium salts of fatty acids, rosin acids etc), and, when melted, these soaps produce an extremely viscous liquids. These two factors combined make industrial handling difficult. Furthermore, while at the high temperature necessary to maintain their flow, these soaps are in permanent decomposition, compromising the separation output and the quality of the final product, as many of the unsaponifiable valuable products are heat sensitive.

#### **6. Supercritical fluid extraction (SFE)**

Although the conventional methods, vacuum and molecular distillation, have been applied to commercial production of tocopherols from SODD, there are some drawbacks such as residual solvents, high temperature, large amounts of energy consumption, high production costs and the unreliable quality of the products that require further developments. Since thermal degradation of tocopherols is commonly caused by processing at high temperatures (de Lucas *et al.*, 2002), new alternative isolation techniques are desired.

Supercritical carbon dioxide extraction is a process where carbon dioxide passes through a mixture of interest at a certain temperature and pressure until it reaches an extractor. This process is used because supercritical carbon dioxide has a low viscosity, a high diffusivity and a low surface tension that provides selective extraction, fractionation and purification, allowing its penetration in micro- and macro-porous materials. Carbon dioxide is the most desirable supercritical fluid solvent for the separation of natural products used in foods and medicines because of its inertness, nontoxicity, low cost, and high volatility. The major advantage of this method is the easy post-reaction separation of the components by depressurization, so resultant extract does not contain solvent residue and hence naturalquality extracts can be obtained. Another advantage is the low temperatures used for the majority of the experimentations because carbon dioxide has a near-ambient critical temperature (31.1 °C), so is suitable for thermolabile natural products.

However, the use of high pressure conditions to concentrate tocopherols makes the system energetically expensive, but the industrial process can be economically viable using conditions of approximately 90 atm and 40 °C (Mendes *et al.*, 2002). At these specific conditions, only fatty acids are separated from tocopherol (Mendes *et al.*, 2005). An increase in pressure and temperature increases the oil extraction and tocopherol recovery, although different pressure–temperature systems need to be used in order to separate the different components (sterols, tocopherols, fatty acids and squalene). It is important to know that

Among the great variety of processes that have been patented for the purification of the compounds of the SODD, only the processes of esterification of fatty acids and acylglycerols with methanol or ethanol followed by high vacuum distillation, have been developed on a commercial scale for the concentration of tocopherols (Takagi & Kai, 1984), (Su-Min *et al.*, 1992), (Yong-Bo *et al.*, 1994), (Rohr & Trujillo-Quijano, 2002). These processes are the most time efficient and economical methods, however high purity of sterols or tocopherols cannot

In the case of separation by distillation of unsaponifiable valuable products of SODD subjected to saponification, the difference between the boiling point of volatile products, such as unsaponifiable components, and the boiling point of the sodium and potassium organic acid soaps is so great that separation is theoretically possible at a high level of efficiency. However, a problem related to this separation technique is that the soaps have a very high melting point, close to the decomposition temperature of the sodium or potassium soaps (i.e. the sodium or potassium salts of fatty acids, rosin acids etc), and, when melted, these soaps produce an extremely viscous liquids. These two factors combined make industrial handling difficult. Furthermore, while at the high temperature necessary to maintain their flow, these soaps are in permanent decomposition, compromising the separation output and the quality of the final product, as many of the unsaponifiable

Although the conventional methods, vacuum and molecular distillation, have been applied to commercial production of tocopherols from SODD, there are some drawbacks such as residual solvents, high temperature, large amounts of energy consumption, high production costs and the unreliable quality of the products that require further developments. Since thermal degradation of tocopherols is commonly caused by processing at high temperatures

Supercritical carbon dioxide extraction is a process where carbon dioxide passes through a mixture of interest at a certain temperature and pressure until it reaches an extractor. This process is used because supercritical carbon dioxide has a low viscosity, a high diffusivity and a low surface tension that provides selective extraction, fractionation and purification, allowing its penetration in micro- and macro-porous materials. Carbon dioxide is the most desirable supercritical fluid solvent for the separation of natural products used in foods and medicines because of its inertness, nontoxicity, low cost, and high volatility. The major advantage of this method is the easy post-reaction separation of the components by depressurization, so resultant extract does not contain solvent residue and hence naturalquality extracts can be obtained. Another advantage is the low temperatures used for the majority of the experimentations because carbon dioxide has a near-ambient critical

However, the use of high pressure conditions to concentrate tocopherols makes the system energetically expensive, but the industrial process can be economically viable using conditions of approximately 90 atm and 40 °C (Mendes *et al.*, 2002). At these specific conditions, only fatty acids are separated from tocopherol (Mendes *et al.*, 2005). An increase in pressure and temperature increases the oil extraction and tocopherol recovery, although different pressure–temperature systems need to be used in order to separate the different components (sterols, tocopherols, fatty acids and squalene). It is important to know that

be achieved due to the similar boiling points of these two substances.

(de Lucas *et al.*, 2002), new alternative isolation techniques are desired.

temperature (31.1 °C), so is suitable for thermolabile natural products.

valuable products are heat sensitive.

**6. Supercritical fluid extraction (SFE)** 

recycling the solvent does not endanger the viability of the process. The value of the rate of return on investment and time of return on investment for the process that does not recycle the carbon dioxide is higher than those of recycling the solvent. This is due to the compression cost that represents more than 59% of the total cost of the production (Mendes, 2002).

 SODD as such will not be feasible to work with SC-CO2 for the tocopherol enrichment, owing to its poor SC-CO2 solubility. So, to concentrate tocopherols from SODD, pretreatment of the raw material, including the esterification of free fatty acids and the removal of sterols with alcohol recrystallization, is needed to obtain the primary tocopherols concentrate with improve solubility in SC-CO2 (Shishikura *et al.*, 1988). For that, triglycerides and FFAs which constitute a major component in SODD have to be chemically modified to obtain free fatty acids and then their methyl esters by esterification.

Several researchers have tried to concentrate tocopherols from SODD by supercritical CO2 (Lee, 1991), (Brunner *et al.*, 1991), (Brunner, 1994b), (Zhao *et al.*, 2000), (Nagesha, G. K. *et al.*, 2003), (Fang *et al.*, 2007), but the operation parameters, especially pressure, differ from author to author. Moreover, in all the cases, tocopherols content of the extract depended on the composition and properties of the natural matrix.

The interest in the tocopherol concentration using supercritical fluid extraction started with Lee and coworkers (Lee, 1991) followed by Brunner and coworkers (Brunner, 1991) and Brunner (Brunner, 1994b). The operational conditions used varied from 35 to 90 °C and from 200 to 400 bar using extractors or countercurrent columns. Lee and coworkers (Lee, 1991) attempted to modify soybean sludge chemically, to improve the solubility in SC-CO2. A simple batch process was utilized to recover tocopherols at 40% concentration at a pressure of 400 bar from the esterified soybean sludge which initially contained 13-14% tocopherols. The solubility of the esterified soybean sludge in supercritical carbon dioxide was more than 4 to 6 times higher than that of the sterols. Brunner and coworkers (Brunner, 1991) obtained a higher enrichment of tocopherols from SODD using supercritical carbon dioxide as a solvent compared to results obtained by Lee and coworkers This group recovered tocopherols from a model mixture of squalene, tocopherols, and sterols using two continuous countercurrent fractionation columns. Squalene was separated from the model mixture in the first column. Sterols were removed from the bottom of the second column, resulting in 85-95% concentration of tocopherol being obtained at the top of the second column.

These works concluded that the fatty acids are extracted initially and the tocopherols are enriched inside the extractor. The results also indicated that the solubility of the tocopherols is intermediary when compared to the solubilities of squalene and stigmasterol.

Chang and coworkers (Chang, 2000) worked on the separation of several SODD components. Their supercritical fluid extraction apparatus had a separation and an extraction unit. Free fatty acids, squalene and tocopherols were recovered in the extract and the sterols were recovered in the raffinates. The average tocopherol concentration factor was 1.38, which means that the mixture in the extract did not separate. However, the author mentioned that with the increase of CO2 volume, the separation factor can reach 1.7, but the poor increase in the concentration factor does not justify the raise in gas volume. The following research groups focused on the separation of the problematic pairs by supercritical fluids using synthetic mixtures: (Mendes, 2005), (Mendes *et al.*, 2000), (Wang *et al.*, 2004), (Nagesha, 2004) and (Nagesha, 2003).

Extraction and Enzymatic Modification of

tocopherol recovery (about 80%).

Functional Lipids from Soybean Oil Deodorizer Distillate 463

initial pressure, feed location, temperature gradient, and ratio of CO2 to ME-DOD were optimized for separating FAMEs. For the following tocopherol concentration step, a final pressure of 200 bar resulted in the greatest average tocopherol content (>50%) and

The important step in concentrating natural tocopherols from these systems is to remove the FAMEs. FAMEs are important chemical materials in biofuel, metal-cutting oil, and cleaning agent production, as well as in the synthesis of other fatty acid products (Swern, 1986). Some works on phase equilibrium for the realistic system of modified esterification SODD/supercritical CO2 (Fang *et al.*, 2005) established that the separation factor 1 between tocopherols and FAMEs was always smaller than unity in the range investigated. This indicates that when supercritical CO2 is used as the separation solvent, tocopherols, unlike FAMEs, tend to enrich in the liquid phase. In particular, the separation factors at pressures lower than 200 bar were relatively small. At 40 ºC, for instance, the separation factor remained lower than 0.2 for all pressures lower than 150 bar. As pressure increased, the separation factor increased greatly, reaching 0.35 at 200 bar. The influence of temperature was contrary to that of pressure, with an increase in temperature leading to a decrease in separation factor. Low pressure and high temperature result in high selectivity, indicated by a low separation factor, which is advantageous in the separation of FAMEs from tocopherols with supercritical CO2. King and coworkers (King *et al.*, 1996) combined supercritical fluid extraction (SFE) with supercritical fluid chromatography (SFC) for concentrating tocopherols and the optimized conditions were 250 bar/80 ºC for SFE and 250 bar/40 ºC for SFC. Approximately 60% of the available tocopherols in soyflakes can be recovered in the SFE step, yielding enrichment factors of 1.83-4.33 for the four tocopherol species found in soybean oil. Additional enrichment of tocopherol species can be realized in the SFC stage, with enrichment factors 2

Starting with a feed containing 48.3 wt% tocopherols, Gast and coworkers (Gast *et al.*, 2005) were able to obtain tocopherols with a purity of 94.4 wt% in bottom phase by supercritical CO2 extraction at 230 bar and 80 ºC, with a solvent-to-feed ratio of 110 and a reflux ratio of

Torres and coworkers (Torres, 2007) proposed a two-step enzymatic procedure to obtain FASEs, tocopherols, and fatty acid ethyl esters (FAEEs) from SODD, together with minor amounts of squalene, free fatty acids, free sterols and triacylglycerols. The final product obtained was used as starting material to purify FASEs, tocopherols, and FAEEs via supercritical CO2 extraction The phytosterol esters were then purified from this mixture using supercritical carbon dioxide (Torres, 2009). Experimental extractions were carried out in an isothermal countercurrent column (without reflux), with pressures ranging from 200 to 280 bar, temperatures of 45-55 °C and solvent-to-feed ratios from 15 to 35 kg/kg. Using these extraction conditions, the fatty acid esters were completely extracted and, thus, the fractionation of tocopherols and phytosterol esters was studied. At 250 bar, 55 °C and a

1 The separation factor represents the process selectivity for separating methyl oleate from tocopherol. In detail, a lower value indicates higher selectivity, whereas a higher value indicates that it is more difficult to separate the two compounds under certain conditions. Furthermore, when the separation factor equals unity, the composition in the gas phase is similar to that in the liquid, and the supercritical

2 The enrichment factors were the ratio of individual tocopherols in extracts versus the same tocopherol

ranging from 30.8 for delta-tocopherol to 2.41 for beta-tocopherol.

4.6. Squalene was completely recovered in top phase.

CO2 process cannot separate methyl oleate from tocopherol.

content initially found in the soyflakes.

There are some interesting relationships and conclusions that can be deduced from the results obtained by Mendes (Mendes, 2005, Mendes, 2000) and Chang (Chang, 2000) regarding the yield and concentration factor of pairs of compounds at different conditions of pressure and temperature. The binary mixture of tocopherol and squalene cannot be separated at low pressure conditions. An acceptable separation needs a raise in pressure to almost 203 bar ((Mendes, 2005) and (Mendes, 2000)). However, a recovery of 90% and a purity of 60% of α-tocopherol has been achieved using a pressure swing adsorption (PSA) device, that is a widely used process in the separation of gas mixtures for air-drying, oxygen and nitrogen separation of air, hydrogen purification, and various other separations. The PSA process is based on the regeneration of adsorber by the difference in adsorbed amounts of gas solute as a function of pressure. In the case of a two-bed process, one bed is in the adsorption step, while the other is simultaneously in the desorption step. The adsorption and desorption steps had pressure conditions of 160 and300 bar, respectively (Wang, 2004).

The ternary mixture of tocopherol, fatty acids and squalene behaved differently from the tocopherol–fatty acid binary mixture. For the same conditions of pressure (160 and 300 bar), the binary mixture had a total separation while the ternary mixture did not achieve any separation. Squalene and stigmasterol mixtures are also very difficult to separate. At low pressure conditions, the yield is less than 10% but at higher pressure, the yield is 76%. It is important to note that low temperatures were used in these studies.

Results of these authors suggest that the supercritical-CO2 process could be used for the separation of squalene, fatty acids and tocopherols. In order to enhance the squalene, fatty acids and tocopherols separation, deodorizer distillate mixtures should be processed several times in supercritical-CO2 at different temperature and pressure conditions. However, there is no data reported regarding the total extraction time of the mixture and this makes it impossible to estimate the operational cost. The major advantage of this method is the total removal of free fatty acids from the mixture.

Zhao and coworkers (Zhao, 2000) concentrated tocopherols up to 75% at 120 bar using a fractionation column with a gradient of temperature from 30–80 °C in a pilot plant scale.

Nagesha and coworkers (Nagesha, 2003) performed chemical modification of SODD containing about 2.9 wt% of tocopherols, as well as triglycerides (56 wt%), free fatty acids (25.3 wt%), sterols (7.8 wt%), hydrocarbons (0.6 wt%), and unsaponifiables (6.4 wt%) apart from tocopherols. Chemical modification of SODD included saponification and esterification steps to result in fatty acid methyl esters from free fatty acids, so as to improve the solubility of SODD in SC-CO2 extraction. Reactions were conducted in dark with continuous flushing of N2 and 1.0 wt% of pyrogallol was added to prevent the oxidation of tocopherols. After chemical modification, esterified SODD contained about 3.7 wt% of tocopherols. Tocopherols concentrates of about 36% was obtained by SC-CO2 extraction at the pressure 180 bar and temperature 60 ºC.

Fang and coworkers (Fang, 2007) carried out a pretreatment of methyl esterification and methanolysis reactions, which converted most of free fatty acids and glycerides to fatty acid methyl esters (FAMEs), respectively, to simplify the composition of SODD and improve his solubility in supercritical CO2 extraction. The mixture was held at 3 ºC in a refrigerator for 12 h, as a result most of sterols were crystallized and removed by filtering under a reduced pressure. Supercritical CO2 fractionation was employed to concentrate tocopherols from Methyl Ester Oil Deodorizer Distillate (ME-DOD) product, mainly contained FAMEs (65– 80 wt.%), tocopherols (10–15 wt.%), and impurities (such as residual sterols, glycerides, squalene, pigments, and long chain paraffins, comprising in total about 10–15 wt.%). The

There are some interesting relationships and conclusions that can be deduced from the results obtained by Mendes (Mendes, 2005, Mendes, 2000) and Chang (Chang, 2000) regarding the yield and concentration factor of pairs of compounds at different conditions of pressure and temperature. The binary mixture of tocopherol and squalene cannot be separated at low pressure conditions. An acceptable separation needs a raise in pressure to almost 203 bar ((Mendes, 2005) and (Mendes, 2000)). However, a recovery of 90% and a purity of 60% of α-tocopherol has been achieved using a pressure swing adsorption (PSA) device, that is a widely used process in the separation of gas mixtures for air-drying, oxygen and nitrogen separation of air, hydrogen purification, and various other separations. The PSA process is based on the regeneration of adsorber by the difference in adsorbed amounts of gas solute as a function of pressure. In the case of a two-bed process, one bed is in the adsorption step, while the other is simultaneously in the desorption step. The adsorption and desorption steps had pressure conditions of 160 and300 bar, respectively (Wang, 2004). The ternary mixture of tocopherol, fatty acids and squalene behaved differently from the tocopherol–fatty acid binary mixture. For the same conditions of pressure (160 and 300 bar), the binary mixture had a total separation while the ternary mixture did not achieve any separation. Squalene and stigmasterol mixtures are also very difficult to separate. At low pressure conditions, the yield is less than 10% but at higher pressure, the yield is 76%. It is

Results of these authors suggest that the supercritical-CO2 process could be used for the separation of squalene, fatty acids and tocopherols. In order to enhance the squalene, fatty acids and tocopherols separation, deodorizer distillate mixtures should be processed several times in supercritical-CO2 at different temperature and pressure conditions. However, there is no data reported regarding the total extraction time of the mixture and this makes it impossible to estimate the operational cost. The major advantage of this method is the total

Zhao and coworkers (Zhao, 2000) concentrated tocopherols up to 75% at 120 bar using a fractionation column with a gradient of temperature from 30–80 °C in a pilot plant scale. Nagesha and coworkers (Nagesha, 2003) performed chemical modification of SODD containing about 2.9 wt% of tocopherols, as well as triglycerides (56 wt%), free fatty acids (25.3 wt%), sterols (7.8 wt%), hydrocarbons (0.6 wt%), and unsaponifiables (6.4 wt%) apart from tocopherols. Chemical modification of SODD included saponification and esterification steps to result in fatty acid methyl esters from free fatty acids, so as to improve the solubility of SODD in SC-CO2 extraction. Reactions were conducted in dark with continuous flushing of N2 and 1.0 wt% of pyrogallol was added to prevent the oxidation of tocopherols. After chemical modification, esterified SODD contained about 3.7 wt% of tocopherols. Tocopherols concentrates of about 36% was obtained by SC-CO2 extraction at

Fang and coworkers (Fang, 2007) carried out a pretreatment of methyl esterification and methanolysis reactions, which converted most of free fatty acids and glycerides to fatty acid methyl esters (FAMEs), respectively, to simplify the composition of SODD and improve his solubility in supercritical CO2 extraction. The mixture was held at 3 ºC in a refrigerator for 12 h, as a result most of sterols were crystallized and removed by filtering under a reduced pressure. Supercritical CO2 fractionation was employed to concentrate tocopherols from Methyl Ester Oil Deodorizer Distillate (ME-DOD) product, mainly contained FAMEs (65– 80 wt.%), tocopherols (10–15 wt.%), and impurities (such as residual sterols, glycerides, squalene, pigments, and long chain paraffins, comprising in total about 10–15 wt.%). The

important to note that low temperatures were used in these studies.

removal of free fatty acids from the mixture.

the pressure 180 bar and temperature 60 ºC.

initial pressure, feed location, temperature gradient, and ratio of CO2 to ME-DOD were optimized for separating FAMEs. For the following tocopherol concentration step, a final pressure of 200 bar resulted in the greatest average tocopherol content (>50%) and tocopherol recovery (about 80%).

The important step in concentrating natural tocopherols from these systems is to remove the FAMEs. FAMEs are important chemical materials in biofuel, metal-cutting oil, and cleaning agent production, as well as in the synthesis of other fatty acid products (Swern, 1986).

Some works on phase equilibrium for the realistic system of modified esterification SODD/supercritical CO2 (Fang *et al.*, 2005) established that the separation factor 1 between tocopherols and FAMEs was always smaller than unity in the range investigated. This indicates that when supercritical CO2 is used as the separation solvent, tocopherols, unlike FAMEs, tend to enrich in the liquid phase. In particular, the separation factors at pressures lower than 200 bar were relatively small. At 40 ºC, for instance, the separation factor remained lower than 0.2 for all pressures lower than 150 bar. As pressure increased, the separation factor increased greatly, reaching 0.35 at 200 bar. The influence of temperature was contrary to that of pressure, with an increase in temperature leading to a decrease in separation factor. Low pressure and high temperature result in high selectivity, indicated by a low separation factor, which is advantageous in the separation of FAMEs from tocopherols with supercritical CO2.

King and coworkers (King *et al.*, 1996) combined supercritical fluid extraction (SFE) with supercritical fluid chromatography (SFC) for concentrating tocopherols and the optimized conditions were 250 bar/80 ºC for SFE and 250 bar/40 ºC for SFC. Approximately 60% of the available tocopherols in soyflakes can be recovered in the SFE step, yielding enrichment factors of 1.83-4.33 for the four tocopherol species found in soybean oil. Additional enrichment of tocopherol species can be realized in the SFC stage, with enrichment factors 2 ranging from 30.8 for delta-tocopherol to 2.41 for beta-tocopherol.

Starting with a feed containing 48.3 wt% tocopherols, Gast and coworkers (Gast *et al.*, 2005) were able to obtain tocopherols with a purity of 94.4 wt% in bottom phase by supercritical CO2 extraction at 230 bar and 80 ºC, with a solvent-to-feed ratio of 110 and a reflux ratio of 4.6. Squalene was completely recovered in top phase.

Torres and coworkers (Torres, 2007) proposed a two-step enzymatic procedure to obtain FASEs, tocopherols, and fatty acid ethyl esters (FAEEs) from SODD, together with minor amounts of squalene, free fatty acids, free sterols and triacylglycerols. The final product obtained was used as starting material to purify FASEs, tocopherols, and FAEEs via supercritical CO2 extraction The phytosterol esters were then purified from this mixture using supercritical carbon dioxide (Torres, 2009). Experimental extractions were carried out in an isothermal countercurrent column (without reflux), with pressures ranging from 200 to 280 bar, temperatures of 45-55 °C and solvent-to-feed ratios from 15 to 35 kg/kg. Using these extraction conditions, the fatty acid esters were completely extracted and, thus, the fractionation of tocopherols and phytosterol esters was studied. At 250 bar, 55 °C and a

 1 The separation factor represents the process selectivity for separating methyl oleate from tocopherol. In detail, a lower value indicates higher selectivity, whereas a higher value indicates that it is more difficult to separate the two compounds under certain conditions. Furthermore, when the separation factor equals unity, the composition in the gas phase is similar to that in the liquid, and the supercritical CO2 process cannot separate methyl oleate from tocopherol.

<sup>2</sup> The enrichment factors were the ratio of individual tocopherols in extracts versus the same tocopherol content initially found in the soyflakes.

Extraction and Enzymatic Modification of

or complement other methods at industrial scale.

**8.1 Influence of refining on phytosterols** 

**8. Degradation and oxidation of functional lipids from SODD** 

collected after chemical and physical refining processes (Dowd, 1998).

esterified sterols has been studied by (Verleyen, 2002b).

loss in the soapstock (Gutfinger & Letan, 1974).

*al.*, 1994).

Functional Lipids from Soybean Oil Deodorizer Distillate 465

These methods do not present a significant advance regarding the most frequently utilized methods and probably their application to production scale would be little profitable. However, they can be used to isolate tocopherols and sterols from SODD at laboratory scale

During refining of edible fats and oils, the content of total sterols decreases due to degradation and formation of products through isomerization (D5 to D7-sterol), dehydration, polymerization, and formation of hydrocarbons or sterenes and sterol oxidation products (Dutta, 2006). These qualitative and quantitative changes in sterols can be traced in the refined oil and in by-products such as soapstocks and distillate fractions

Acid hydrolysis of steryl esters may occur upon bleaching with an acid activated bleaching earth. The slight reduction of the total sterol content is due to the formation of steradienes and disteryl ethers. A gradual reduction in the total sterol content is observed at increasing deodorization temperature due to distillation and steradiene formation. Increasing the temperature from 220 °C to 260 °C resulted in a gradual reduction of the total sterol recovery from 90.4% to 67.7% in physical refining and from 93% to 62.7% in chemical refining. However in physical refining, an increase of 40% in the steryl ester fraction is observed due to an esterification reaction, promoted by high temperature between a sterol and a fatty acid. Due to the absence of free fatty acids in the chemical refining their esterification did not occur (Verleyen *et al.*, 2001b). The influence of refining on free and

Phytosterols are progressively lost during refining while continuously altering the ratio of free and esterified sterols (Kochhar, 1983). During chemical neutralization, the free sterol content is significantly reduced especially upon addition of weak caustic solution due to the

Bleaching effects on phytosterols are generally minor and mainly limited to the formation of some nonpolar dehydration products (Ferrari, 1996) and partial hydrolysis of sterol esters (Homberg & Bielefeld, 1982). Steradienes and disteryl ether dehydration products (Figure 1) are formed during bleaching step by the bleaching temperature and the degree of acid activation of the bleaching earth, while during the deodorization, the degree of sterol dehydration is mainly influenced by deodorization temperature giving rise to a concentration of the steradienes in the distillate (Verleyen, 2002b, Verleyen, 2001c). The presence of steradienes can also be used as a marker for the presence of refined oils (Grob *et* 

Whenever applied, hydrogenation has a tremendous effect on sterol structures, including hydrogenation of double bonds, opening of cyclopropane rings, and positional

A part of a multinational EU research project (FOOD-CT2004-007020) was to carry out qualitative and quantitative assessment of sterols and sterol oxidation products in samples of by-products from chemical and physical refining of edible fats and oils collected from various locations in Europe. To the best of our knowledge, this is the first report on the contents of oxidized sterols in soapstock and distillate fractions from edible oil refining processes. The levels of sterol oxidation products were higher in acid oil obtained from

isomerization of side chain unsaturation (Strocchi & Marascio, 1993).

solvent-to-feed ratio of 35, the phytosterol esters were concentrated in the raffinate up to 82.4 wt-% with satisfactory yield (72%).

Other supercritical fluids have been explored but unsuccessfully for the separation of different pairs of components. An attempt at using similar methodology to (Mendes, 2000) and (Mendes, 2005) but using liquid gas petroleum instead of carbon dioxide did not change the poor concentration factor between the critical pairs of components (Buczenko *et al.*, 2003). Buczenko and coworkers (Buczenko, 2003) performed the saponification of the raw material and the extraction of unsaponifiable matter as pre-treatment of VODD.

As discussed above, there are a lot of experimental studies proving the efficiency of the supercritical extraction to concentrate the vitamin E from different raw material or in some cases, from synthetic mixtures representing the deodorizer distillate, but the extraction of sterols using supercritical fluid from the deodorizer distillate was not described in the literature.

On the other hand, due to the low content of squalene in SODD, specific extraction processes of squalene using supercritical fluid from SODD was not described in the literature. Existing studies are models of fractionation of artificial mixtures such as those mentioned above (Chang, 2000). For example, Brunner (Brunner, 1994a) studied the phase equilibrium for recovering α-tocopherol from a mixture of squalene, tocopherol, and campesterol. He concluded that the separation factor for squalene/α-tocopherol varied between a value of 4 at low squalene concentrations (0.5 wt %), to a value of 1 at high squalene concentrations (85 wt %), at pressures ranging from 200 to 300 bar and temperatures ranging from 70 to 100 ºC.

Bondioli and coworkers (Bondioli *et al.*, 1993) esterified FFAs into their corresponding glycerides and then applied a supercritical carbon dioxide extraction to produce a squaleneenriched fraction (purity 90.0%, yield 91.1%), but from olive oil deodorizer distillates.
