ELSEVIER
Aqueous and enzymatic processes for edible oil extraction A. Rosenthal,
D. L. Pyle, and K. Niranjan
of Food Science United Kingdom
Department Reading,
and Technology,
University of Reading,
Whiteknights,
Industrial processes for the extraction of edible oil from oilseeds generally involve a solvent extraction step which may or may not be preceded by pressing. Hexane is the preferred solvent; hexane-based processes have been in commercial operation for a long time. For such processes, it is possible to achieve oil yields in excess of 95% with a solvent recovery of over 95%. In the past, the main concern of this process has been the safety implications surrounding the use of hexane. This prompted attempts to develop processes based on the use of aqueous extraction media which were unsuccessful mainly due to low oil yields. The scenario at present appears to be changing. Interest in aqueous extraction processes has been revived by increasing environmental concern. An aqueous process is looked upon as an environmentally cleaner alternative technology for oil extraction. Organic solvents such as hexane, in particular, can contribute to the industrial emissions of volatile organic compounds (VOCs). The production of VOCs in the conventional process is particularly worrisome since these can react in the atmosphere with other pollutants to produce ozone and other photochemical oxidants which can be hazardous to human health and can cause damage to crops. Besides this, the VOCs are themselves “greenhouse gases”; some are carcinogenic and have toxic properties. Other advantages of the aqueous process compared with solvent-based processes include: (I) simultaneous production of edible oil and protein isolate or concentration in the same process, (2) lower protein damage during extraction, and (3) improved process safety due to the lower risk of fire and explosion. It is also reported that aqueous extraction processes may be more cost effective since the solvent recovery step is eliminated. The main limitations of this process appear to be: (I) lower eflciency of oil extraction as evident in earlier studies, (2) demulsification requirements to recover oil when emulsions are formed, and (3) treatment of the resulting aqueous efluent. With the objective of improving the yield of aqueous processes, enzymes have been used to facilitate oil release. Selected enzymes have been tried on different types of oilseeds, resulting in extraction yields much higher than the original aqueous process (in some cases of over 90%). These enzymes mainly hydrolyze the structural polysaccharides which form the cell wall of oilseeds or the proteins which form the cell and lipid body membrane. This article aims to review aqueous and enzyme-based processes and discuss related issues. Keywords:
Aqueous
oil extraction;
enzymatic
treatment
of oilseeds; oil recovery;
Introduction Aqueous enzymatic oil extraction is undoubtedly an emerging technology in the fats and oil industry since it offers many advantages compared to conventional extraction. For instance, it eliminates solvent consumption which reportedly may also lower investment costs172 and energy require-
Address reprint requests to Dr. K. Niranjan, Food Science and Technology, University of Reading, P.O. Box 226, Whiteknights. Reading, Berkshire RG6 6AP, United Kingdom Received 29 June 1995; revised 22 January 1996; accepted 31 January 1996
Enzyme and Microbial Technology 19:402-420, 1996 0 1996 by Elsevier Science inc. 655 Avenue of the Americas, New York, NY 10010
protein recovery
merits.‘’ Also, it enables simultaneous recovery of oil and protein from most oilseeds and the process yields oil of good quality complying with Codex specifications. The need for further degumming operations is eliminated and the process allows ready removal of some toxins or antinutritional compounds from certain oilseeds. In this sense, some of the needs triggering technology innovation in the oil extraction sector such as cost savings, environmental and safety concerns, and nutrition issues seem to be achievable by successful development of aqueous enzyme-based processes. Over the last four decades, several studies have been carried out on aqueous processing of oilseeds. Although
0141-0229/96/$15.00 PII SOI41-0229(96)00053-I
Edible
the concept appears potentially attractive compared to the conventional hexane-based process, the comparatively low oil yield and relatively high content of oil in the residue (and in some cases also in the protein isolate) have discouraged its commercial application. The problem of low extraction efficiency of aqueous processes may be overcome by the use of hydrolytic enzymes which help release oil and increase the yield as some studies have shown.g-” Besides this, environmental issues, especially the increasing concern about volatile organic compounds (VOCs) caused by solvent emissions, have also sharpened the focus on this process. A recently published reviewI gave an interesting account of enzymatic pretreatments applied to different oil seeds and fruits. The main objective of this review is: to discuss key advantages and disadvantages of aqueous enzymatic processes vis u vis conventional solvent-based processes p~icul~ly with respect to environments and economics aspects and the quality of the resulting products; to analyze published information on processes employing aqueous extraction media which may or may not use enzymes; to discuss the oilseed structure and the mechanisms by which enzymes act on oilseed which would lead to rational selection of enzymes for the aqueous or conventional process; to discuss the role of operational parameters on process optimization; and finally to discuss the formation and stability of oil-in-water emulsions and alternative strategies for downs~eam processing for oil and protein recovery.
Oleaginous crop structure and oil extraction processes To gain a better understanding of the process and possible role of enzymes, it is essential to investigate the structure of oil-bearing materials. The main feature of oilseed cotyledon cells is the existence of discrete cellular organelles called lipid and protein bodies which contain, respectively, most of the oil and protein in the grain. Figure I shows the microscopic structure of soybean which is similar to that of many other oilseeds. Protein bodies (or aleurone grains) vary in size depending on the oilseed and also over a wide range inside each oilseed. In the case of soya which is also similar to peanuts,14 the average size of the protein body is 8-10 pm, but variation from 2-20 pm has been reported.‘“-‘* These protein bodies contain approximately 60-70% of the total protein present in oilseeds.” Lipid bodies (also known as spherosomes and oleosomes) are the principal repository sites of li id reserves not only in oilseeds but also in oleaginous fruits. P9320Their sizes are often in the range of l-2 pm in diameter for most oilseeds’J.‘8.‘9 although they can also vary from 0.2-0.5 pm in the case of soybean’8’2’ to as large as 4 pm in cottonseeds.‘” On the other hand, in the case of oleogenic fruits such as olive, avocado, and oil aim, the storage lipid bodies are often in excess of 20 p,rn.2! In these oleogenic fruits, it is normally the mesocarp tissue which accumulates most of the storage lipid.*’ Scanning electron microscopy (SEM) analysis has shown that lipid bodies from oilseeds like soybeans23 and
oil extraction:
A. Rosenthal
et al.
Cuticle Palisade Cells Hourglass Cells
Parenchyma
Endssperm
-
Alsurone Cells Compressed Cells Cotyledon Epidermis
Palisade-like Cell
- Protein Body
- i&id Body
Figure 1 Cotyledon and endosperm microscopic soybean grain13 (reproduced with permission)
structure
of
peanuts” are enmeshed in a kind of cytoplasmic network presumably composed of proteins.” The spaces between protein bodies in cotyledon cells are then filled with the lipid body and cytoplasmic network.“*” Unlike the cytoplasmic features which are characterized by the presence of protein and lipid, the walls which surround the cell are primarily composed of cellulose, hemicellulose, and lignin in addition to pectin.‘* In the usual solvent-based process, the grain is flaked, thereby causing the cell wall to rupture; this exposes the oil located inside the cell and also facilitates the percolation of solvents into which oil can diffuse. Taking soybean as an example, the usual thickness obtained after flaking is about 0.25 mm.74 Considering that the soybean cotyledon cells are about 1% 20 pm in diameter and 70-80 km long, the resulting flake size allows the rupture of a high propo~ion of the cells: therefore, in solvent extraction, the oil diffuses and is extracted through the solvent while protein remains in the meal with the fibers and carbohydrates. Aqueous processes also involve the use of comminuted material which serves the same purpose of exposing and releasing the oil and protein more easily from the material. In this case, however, while the soluble components diffuse into water, the released oil forms a separate liquid phase or is partially emulsified with water. The role of most hy~olytic enzymes such as cellulases, hemicellulases, and pectinases in these processes is to break the structure of cotyledon cell walls. Figure 2 shows the effect of enzyme treatment on a previously ground oilseed. It can be seen from Figure 2 that enzyme action makes the structure more permeable; the extent depends on particle size. Proteolytic enzymes mainly hydrolyze the proteins in the cell membranes as well as inside the cytoplasm. It is noteworthy that seed oil bodies contain abundant
Enzyme Microb. Technol.,
1996, vol. 19, November
1
403
Review
A:
Very
fine
(mechanical
8:
Ffnc
(enzyme
C:
Coarse
Figure 2 Effect of milling and enzymatic treatment cell structurez5 (reproduced with permission)
millfng actfc.01
milling action)
milllng
on oilseed
proteins termed oleosins which seem to play a major role in stabilizing these bodies.20.26 The structure of oleosin is generally the same for most oilseeds; it consists of low molecular weight proteins in the range of 15,000-26,000. The structure can be divided into three basic domains: (1) an amphipathic domain present at the amino terminus of oleosins which are most probably associated with the oil body surface; (2) a central hydrophobic domain which contains around 70 nonpolar amino acids in succession that may strongly interact with the triacylglycerol matrix of the lipid bodies; and (3) an amphipathic domain at or near the carboxy1 terminus which interacts with the surface of a phospholipic monolayer surrounding the triacylglycerol matrix. Due to their peculiar structure, the oleosins seem to be decisive in maintaining the integrity of the lipid bodies during desiccation that accompanies seed maturation by preventing interaction and possible coalesce.20*26 Proteolytic enzymes can potentially hydrolyze the lipid body membranes. Bair and Snyder”’ and Tzen and Huang” isolated lipid bodies from soybean and mature maize, respectively, and carried out tryptic hydrolysis. As a result, some breakdown of the lipid body membranes due to hydrolysis of the oleosins occurred and coalescence between some of the bodies was evident in both cases.““’ It can therefore be concluded that the released oil can be more easily separated from the cotyledon cells by an aqueous medium or an organic solvent following proteolytic action. Proteolytic enzymes can also affect the cytoplasmic network which is largely composed of proteins in the case of soybean and some other oilseeds, thereby making the inner structure less tightly bound and compact and thus enabling easier removal of protein and lipid from the cell. Unlike oilseeds, the lipid bodies of fruits possess a relatively insignificant amount of oleosins.*’ This seems to be due to the fact that the mesocarp does not undergo desiccation or germination and hence does not require either small stable oil bodies or oleosins. The proteolytic enzymes, therefore, do not appear to be useful for extraction from most oleaginous fruits. Furthermore, the oil-in-water emulsion resulting from the aqueous process tends to be less 404
Enzyme Microb. Technol.,
stable in the case of fruits which makes oil separation easier when compared to oilseeds. Considering the specificity of each carbohydrase, a rational choice of the enzyme for a given oilseed or fruit can only be made after gaining an in-depth understanding of the complex arrangement of polysaccharides in the cell wall. Primary cell walls of a variety of higher plants have many features in common and a very similar structure consisting of cellulose fibers to which strands of hemicelluloses are attached.‘8s3” The fibers are embedded in a matrix of pectic substances linked to a structural protein as shown in Figure 3. This suggests that enzyme preparations capable of attacking cell walls must necessarily contain a mixture of cellulases, hemicellulases, pectinases, and even proteases30 However, commercial preparations containing these enzymes are often unable to hydrolyze specific components and further development may be required to degrade, for instance, complex polysaccharides found in the cell wall of unlignified plant material.“’ The differences in the oilseed composition determine the choice of enzymes to be used for each oilseed or fruit. The typical fiber composition of four oil-bearing materials is given in Table I. While rapeseed has a significant content of pectin, relatively lower levels are found in coconut, and corn germ. For instance, an enzymatic mixture which is tailored to hydrolyze the different substances should be used in order to obtain highest possible oil yields. Particularly in the case of soybean, the higher protein and lower oil content requires the use of proteolytic enzymes for high yields.” The main operations and the potential use of enzymes in the solvent-based extraction process are discussed next.
Conventional oil extraction processes Process description
and technical evolution
Historically, the three most common processes for recovering oil from seeds are hydraulic pressing, expeller pressing, and solvent extraction.“4*3’ Hydraulic pressing, the earliest of the processes, is said
h
Protein
(structural)
hamnogslrcruro
Figure 3 Structure of primary cell wal13’ based on Keegstra ef a/.” (reproduced with permission)
1996, vol. 19, November
1
Edible oil extraction: Table 1
Typical composition
of cell wall polysaccharides Soybean
Polysaccharide
components
Pectic substances Mannan Galactomannan Arabinogalactans Cellulose Hemicellulose xyloglucan arabioxylan Other
contentz4 (%)
of some oleaginous
A. Rosenthal
et al.
crops
Rapeseed content3’ (%)
Coconut kernel conten?’ (%)
Corn germ content3’ (%)
39
-
8 22
61 26 some 13
29 -
-
30 * * * 20 50 * * *
2
39
* 40 10
*Not specified -Negligible
in Europe in 1795.24 Being labor intensive, its use declined over the years and is no longer used.24 Continuous screwpresses such as expellers have replaced the original hydraulic equipment and are still used on a wide variety of oleaginous materials. For materials having relatively high oil contents, a two-step process is carried out which consists of a continuous prepress stage followed by solvent extraction. The main advantage of prepressing is that it allows solvent extraction to be applied to oleaginous materials that would be quite difficult to process by direct extraction methods. Also, solvent requirements are lowered considerably.94 This combined process is generally used on high oil sources (with an oil content above 35%) such as flaxseed, safflower, sunflower, tung nuts, cottonseed, and corn germ. For soybeans, a single solvent extraction is employed24.32 due to a relatively lower oil content as evident from Table 2. Solvent extraction originated as a batch process in Europe in 1870. Technological advances shortly after World War I led to the development of continuous solventextraction systems which proved excellent for processing oleaginous materials to a meal with a very low oil content.‘4 The modern solvent-based process usually consists of extraction by successive countercurrent washes with hexane of the previously cracked, flaked, ground, or pressed oleaginous material. The extracted meal is then carried by a sealed conveyor for solvent recovery in enclosed vessels by using jacket or sparged steam. Hexane is removed from the oil in rising film evaporators and finally, by vacuum distillation.‘3 A basic flow chart of a combined pressing and solventextraction process which is typical for oilseeds with a high oil content is presented in Figure 4. to have originated
Following extraction, the crude oil is submitted to a refining process for removing oil-soluble and oil-insoluble impurities.34 The refining process includes several operations: the degumming step removes phosphatides and mucilaginous gums that, when hydrated, become insoluble in the oil; free fatty acids, color bodies, and metallic prooxidants are removed to a varying degree in the alkalirefining step; treatment of the refined oil with a bleaching absorbent removes more pigments and residual soaps and improves the taste of the oil; and remaining n#$‘avors are removed by high-vacuum steam distillation in the deodorization step.”
CRACKING/GRINDING
MEAL-CONDITIONING (moisturr/tcmpclrtllm)
1 PRESSMG 1 CAKE CRUSHING
1
PUIUFICATION SOLVENT
EXTRACTION
1 PILTRATION
Table 2 material
Typical content of oil and protein of some oil-bearing MEAL+SOLVENT (residual)
Oil material Soybean”’ Rapeseed3’ Coconut kerne13’ (desicated coconut) Corn germs’ Flax seed3’
Oil (%)
Protein (%)
20 40 70 35 45
40 20 6.5 20 25
De%mlvelltizer taster
Oil Stripper
MEAL
CRUDEOIL
Figure 4 Conventional oil extraction of oilseed combining solvent extraction and pressing (modified from Mustaka@’ and Bernardini3?. All the steps may not be necessary for all oilseeds
Enzyme Microb. Technol.,
1996, vol. 19, November
1
405
Review Use of enzymes in solvent-based pressing extraction
and
The possibility of partially releasing the oil from a full fat, heat-treated soya meal by using a hydrolytic enzyme was initially investigated by Sherba et al.35 as reported by AdlerNissen. 36 The oil was more easily recovered by extraction with petroleum ether. This approach was also followed by Fullbrook, lo who investigated aqueous hydrolysis of oilseeds followed by solvent extraction and also tried hydrolysis in the presence of a solvent to simultaneously extract the released oil. It was observed that yields could be considerably improved if hydrolysis of the finely ground flour of soybean and rapeseed was carried out in the presence of solvent. In the case of rapeseed, 50% more oil was thus obtained. For soybeans, the increase in yield was even higher, thereby resulting in a net extraction of about 90% of the original extractable oil; an enzymatic mixture obtained from Aspergillus niger at a concentration of 3% was used.” Olsen3’ also described aqueous hydrolysis of dehulled rapeseed followed by extraction of the residual oil with petroleum ether. The hydrolysis mixture included three cell walldegrading enzymes: pectinase, cellulase, and hemicellulase. This partial hydrolysis increased the permeability of the cell wall, thus allowing more efficient percolation of the solvent and a more efficient extraction of the oil. It was also found that an enzyme mixture containing all three active groups together gave the most extensive hydrolysis. Sosulski et al.” evaluated the effect of different carbohydrases on the extraction time and yield of canola oil. The enzyme reaction was carried out on previously autoclaved and moistureadjusted canola flakes, and followed by drying and hexane extraction. The enzyme efficiency based on oil yield enhancement was: mixed activity enzyme > B-glucanase > pectinase > hemicellulase > cellulase. Enzymatic treatment before Soxhlet extraction for a given time gave 45% higher yields. Furthermore, the time to extract the total extractable oil comparatively decreased indicating increased rates of extraction. The maximum extraction yield obtained for the control in a Goldfish apparatus was approximately 5% lower compared to the ones obtained with enzyme action.’ ’ Similarly, Dominguez et al” reported an increase in soybean extractibility by 8-10% of the extractable oil and up to 4% in the case of sunflower oi13* after enzymatic treatment with different commercial enzymes. The resulting higher extraction rates enabled shorter o eration to obtain a given percentage of the extractable oil. P, Bhatnagar and Johari4’ also verified that treatment with enzymes originating from thermophilus molds increased the amount of Soxhlet extractable oil in the case of some oilseeds. The recovery of cotton oil increased by up to 5% with previous enzymatic action while sunflower oil recovery increased by 4.2% after using another enzyme originating from mold. In the same study, the oil yield was greater when the enzymatic treatment was carried out in the presence of a solvent instead of an aqueous medium; this was attributed to the greater solubilization in organic solvents of plant tissues and proteins to which oil may remain bound even after the normal extraction procedures. Tano-Debrah and 0hta41*42 obtained an increase of around 20% in the solvent
406
Enzyme Microb.
Technol.,
oil extraction yield by pretreating kernels of shea tree with a mixture of protease and carbohydrases. In relation to the pressing process, Cheah et aI.J’ extracted 97.7% of the oil from the pectinase-treated mesocarp of palm by using a hydraulic press compared to 91.1% obtained from the untreated material. In the same way, Bouvier and Entressangles44 used a cellulase preparation to reduce by 3% and 18%, respectively, the palm oil losses from press fibers and crude juice during clarification compared with the traditional process without use of enzymes. Smith et a1.45 enhanced the yield of expelled oil from soybean by up to 2.78% of moisture-free oil with a previous Aspergillusfimigatus mixed-activity crude enzymatic treatment which corresponded to an increase of about 11.7% in oil recovery; this was equivalent to 63.5% of the total extractable oil.
Aqueous process Process description
and unit operations
The aqueous extraction process (AEP; i.e., extraction using water alone as medium) was originally suggested as an alternative to the solvent oil extraction process in the 1950s. It was thought to be safer and cheaper. Moreover, the simultaneous recovery of oil and protein (concentrates or isolates) from oil-bearing materials was possible.46 AEP uses a very distinctive principle compared to the usual solvent extraction processes which are based on the capacity of the oil to dissolve and be extracted by the solvent. In AEP, the oil does not have a high chemical affinity for the extraction medium and, therefore, there is no chemical potential for oil dissolution. Although solvent extraction theory cannot explain AEP, it certainly helps to understand the mechanisms involved in the process. It is well known that the spontaneous dissolution of any solute in a solvent is always accompanied by a negative charge in free energy (AG)47 which is related by the Gibbs equation to enthalpy (W, absolute temperature (r), and entropy (A,‘?) as follows: AG = AH-TAS
(1)
During dissolution, thermal energy is consumed to separate solute molecules and dissociate the solvent molecules, and it is released when the dispersed solute molecules interact with neighboring solvent molecules. The overall enthalpy change will be more negative (exothermic) if the energy released during solute-solute and solvent-solvent interactions is greater than the energy absorbed in solute-solvent interactions; however, if the solute molecules are strongly bound to each other as in water, the solute dissolves well only if dissolution results in stronger solute-solvent interactions. This is not the case for triglyceride-water interactions which are weak and cannot offset the large amount of energy required to break hydrogen bonds in water. On the other hand, solubility of oil in hexane is high because the stronger solute-solvent interactions compensate the energy losses4’ Extraction of oil by AEP is, consequently, based more on the insolubility of oil in water than on the dissolution of oiL4’ In this case, the water soluble components of oil crops
1996, vol. 19, November
1
Edible
diffuse in the water rather than in oil, thereby releasing the oil which was previously bound in the original structure. The aqueous extraction process can, in general, be improved by any treatment that enhances the dissolution of the other components, for instance, by using enzymes or increasing the temperature. AEP uses the same principle as hot water flotation. The process involves mixing ground and dehulled oilseeds in vats of hot water and skimming off the oil rising to the surface.4” The hot water flotation method for edible oil extraction from oilseed kernels is a traditional process used in the rural areas of many developing coun~es.49 The process is accomplished in five basic steps: (1) heat conditioning of the seed, (2) grinding, (3) extraction by boiling, (4) oil recovery, and (5) drying. The grinding is traditionally carried out by pounding using a pestle and mortar. The ground seed is boiled in water, thereby liberating oil which floats to the surface. The oil is then carefully scooped from the water surface using a dish and dried to remove residual moisture.“” In the modem aqueous process, centrifuges have replaced gravity separation and the emphasis is on operating conditions which result in the highest oil and protein recov-
CONCENTRATE
oif extraction:
OIL EMULSION
1
PROCEDURE
ISOLATE PROCEDURE
Alcali extraction
1
Residual solids
et al.
ery and the least damage to the nut~tional value of food proteins.50.5’ Application of aqueous processing the different oilseeds requires, to some extent, changing specific conditions (like pH and temperature of extraction) because of the differing chemical compositions and physical structure of the seeds.’ The unit operations used in AEP may also vary with different oleaginous material, but generally consist of grinding, solid-li uid separation, cent~fugation, demulsi~cation, and drying. 2*46 The general flowchart for the aqueous process is shown in Figure 5. ~en~fugation divides the aqueous extract into oil, solid, and aqueous phases. After the separation step, the bulk of the proteins may be recovered as concentrate in the solid phase or as isolate in the aqueous phase depending on the pH of the extraction medium, i.e., whether it is acidic or basic, respectively.” In the case of the protein isolate, final recovery is achieved by isoelectric precipitation.2.46 On the other hand, the oil recovered after breaking the emulsion and separating the phases is usually of a high quality: the water-washed oil requires little further treatment except possibly removal of residual water before eventual refining-37.46
Acid Extraction
1
A. Rosenthal
WHEY
OIL EMULSION
I
I
demulsify
WHEY SOLIDS
Residual solids
--I
LIQUID EXTRACT
Acid Precipitate Centrifuge I PRQTEilPIIn8CKATE
(protein + carbohydrates) Figure 5 Steps involved in the aqueous extraction process (AEPI based on Lusas and Jividen. 48 Two alternatives results in the production of protein concentrate and the other in the production of protein isolate
Enzyme Microb.
Technol.,
are given. One
1996, vol. 19, November
1
407
Review Another possible alternative to recovering protein directly from oilseed flour extracts and avoiding generation of large quantities of process effluent is the utilization of the membrane isolation process (MIP) combined with AEP. The combination process yields food products with different protein and oil contents from oilseeds.2,57353 In the AEPMIP procedure, ultrafiltration (UF) membranes are used to recover protein directly from aqueous oilseed flour extracts, thereby avoiding generation of whey resulting from acid precipitation. Reverse osmosis (RO) membranes are employed to process the UF permeate in order to recover a secondary product and render the effluent water suitable for reuse.52.53 The process for the AEP-MIP procedure using alkaline extraction and avoiding isoelectric precipitation is shown in Figure 6. Besides the production of low-strength effluent, it is claimed that MIP offers advantages of increased isolate yield since whey proteins are recovered along with protein normally precipitated, products with enhanced nitrogen solubility, greatly reduced process water requirements, and products with more desirable functional properties.2,5’S53 Finally, another potential alternative to the MIP-AEP
Recycle Water
Ground
combined process consists of the usual steps leading to protein recovery by concentrate or isolate procedure followed by whey concentration and fractionation by membrane technology to produce protein and carbohydrate fractions. The water is recycled in the extraction step.
Factors affecting oil and protein recovey
yields
Grinding. As already mentioned, the critical step in the aqueous extraction process which affects the oil and protein yields is the grinding operation which determines the oilseed particle size. Efficient grinding which breaks down the walls of the oil-containing cell is considered essential.“6.“9 Smaller particle size allows not only easier diffusion of water-soluble components, thereby disintegrating the original structure and facilitating oil release, but also enhances enzyme diffusion rates which can then more easily act on the substrates. The grinding operation may be carried out either wet or dry depending on the oilseed. The choice between wet or dry grinding would be governed by several factors such as
Beans
Recycle
,L
UF 1 Membrane I
Water
1 I
Ext’n. Oil Emulsion
UF Concentrate
T
UF Permeate
RO
1M:yfjlbra”e jP,Jrmeate RO Concentrate Residue Product to Dryer (mainly fiber)
Figure 6 Diagram of aqueous extraction precipitation of proteins’
408
Enzyme Microb. Technol.,
+
1
Low-fat Dry Product (protein t carbohydrate)
RO Product to Dryer (protein t carbohydrate)
process combined with membrane
1996, vol. 19, November
1
isolation (AEP-MIP).52
This process eliminates
isoelectric
Edible the initial moisture content, chemical composition, and the physical structure of the oilseeds. In the case of materials with a high moisture level like coconut, wet grinding is more appropriate, thus avoiding an additions intensive drying step before g~nding. On the other hand, in the case of materials with a low initial moisture content like peanuts. rapeseed, and soybean, dry grinding is considered more suitable. Wet grinding is sometimes considered better for rupturing cells mainly because the water softens the cell walls; however, with some materials like peanuts, wet grinding can produce a stable emulsion which has to be broken to recover oil.‘6 Conflicting effects of wet grinding are thus evident. Excessive grinding favors cell rupture and increases the efficiency of oil and protein extraction; however, it also produces smaller oil globules which makes demulsification more difficult. Insufficient grinding, on the other hand, results in unacceptable losses of oil in the residue.J6*“0 In the case of coconut, wet-milling and grinding which conditions the coconut so that only less than 5% of the oil remains with the residue results in the fo~ation of oil globules about 10 p,rn in diameter. This generally corresponds to the formation of a stable emulsion similar to milk. It is evident that a stable emulsion is the price one must pay for efficient wet grinding of oil from fresh coconut.“”
Extraction. The extraction step basically consists of dispersing the ground seeds in water and then agitating the dispersion to enhance oil and protein extraction. Optimum extraction conditions vary according to the oilseed composition and structure.’ Factors which influence the efficiency of extraction include solid:water ratio, particle size, pH time, temperature. degree of agitation,2346 and the number of extraction stages. Concerning the power dissipation level, simple stirring is sometimes sufficient to obtain high yields as in the case of peanut? and sunflower?“; however, high power dissipation levels may be required occasionally which can be achieved by vigorously stirring the dispersion. High-shear stirring also achieves further disintegration of the individual cells and eases the release oil from cornminuted seeds. Increasing the blending time, in general, improves the oil extraction yield but, as mentioned earlier, results in the formation of an emulsion with greater stability which can adversely affect the total yield after subsequent separation steps.5’ In the case of soybean, we found that the yield obtained after repeated consecutive low agitatio~power input extractions using fresh aqueous medium each time was equivalent to that obtained through a single high speed/high energy input operation. It therefore appears necessary to optimize agitation speeds and power dissipation levels by considering not only the process yield but also the stability of the resulting dispersion. One must also take into account the fact that there is a close relationship between the process variables. In this sense, the optimum solid:water ratio and extraction time depend very much on the energy input level on the system which in turn depends on the size and shape of the vessel (extractor). Unfortunately, few published works specify power input levels and relevant scaledependent engineering parameters. As with power dissipation, there is an optimum extrac-
oii extraction:
A. Rosenthal
et al.
tion pH which. as one might expect, varies quite sensitively with the oil-bearing material, It seems that there is a close relationship between the oil and protein extraction mechanisms in some oilseeds since the conditions for the highest oil extraction yields generally coincide with those for the highest protein yields; therefore, the highest oil extraction yield occurs at pH values corresponding to maximum protein extractibility in the aqueous system and the lowest yield is obtained when the protein solubility is at its lowest, i.e., around the isoelectric point. Thus, aqueous extraction of oil from certain seeds can be regarded as a process primarily aimed at solubilizing proteins which results in the release of the oil.“’ Hence, in the case of soybean,5’.“h sunflower,5” and coconut,3’.“o the pH values corresponding to the least oil extraction which also results in the smallest proportion of oil free after the downstream separation step. i.e., around 4.5 for soybean, 5.0 for sunflower. and 4.0 for coconut. were also found to be the overall isoelectric point-of most proteins in the grains which suggests that protein binds oil considerably better in the isoelectric pH range.31.s0 In general for most oilseeds, aqueous extraction for maximizing oil and protein yields is carried out at a pH not near the isoelectric point. If a protein isolate is desired. the pH will be higher than that at the isoelectric point and if protein concentrate is desired, the pH will be much lower.“+‘x Optimum pH conditions for oil extraction found in other studies were 4.0, 7.0, or 10.0 for peanut.“-h.s7 10.0 for sunflower?“ and 6.6 for rapeseed.” Yields of oil and protein recovery reported for some oil-bearing materials through nonenzymatic aqueous processes are shown in Table 3. The temperature also seems to have a considerable effect on yields. Although some authors consider that it is important in the hot water flotation process to ensure that boiling water is used for the extraction, they also state that prolonged boiling may not affect yields for certain oilseeds.‘y However, Lusas ef ~1.’ observed that the extraction temperature was not critical for protein recovery but was critical to oil extraction yields for soybeans. The maximum oil recovery in this case occurred between 40-60°C. Various other temperatures have been employed in aqueous extraction process with other oilseeds: room temperature for the extraction of sunflower oil? 60-65°C for the extraction of peanut oil,“.” 80°C for the extraction of coconut oil,“” and 70°C for the extraction of rapeseed oi15’ The highest possible solid:water ratio is desirable in the extraction step so as to obtain less stable emulsions and generate less effluent; however, to obtain the highest extraction rates and extraction yields, it is usually necessary to use large quantities of water. Solid:water ratios recommended in the literature for peanut oil extraction vary from 1:s to 1: 12;+“.6,59.M’ Hagenmaie? used 1: 10 for sunflower while Lusas et af.’ found 1: 12 as the optimum solid: water ratio for soybeans. Embong and Jelen ’ reported an optimum ratio varying between I :2.S and 13.5 for rapeseed oil extraction. The fact that some liquid may get occluded to the oilseed must also be considered in deciding the solid:liquid ratio. The time required to reach a desired extraction level depends on the oilseed as well as the process variables
Enzyme Microb. Technol.,
1996, vol. 19, November
1
409
Review Table 3
Oil and
protein yields obtained
by aqueous extraction
of different oleaginous
materials
Extraction conditions Oleaginous
material
Coconut Palm Peanut Peanut Rapeseed Sunflower Lupin seed
Temp (“C)
Time (min)
Ratio oiiseed:water
+
*
45 * *
*
I:4 * *
75 * 65
6L 45 40
I:3 730 I:15
pH
Oil yield= (%)
Protein yieldb (%)
Reference
7.0 7.0 4.0 7.0 6.6 10.0 8.0-9.0
93 84 89 86 90 86 36-37
91 *
46 8 46 46 51 54 58
92 89 * 851
*Data not reported ‘Based on total extractable oil in the seed bBased on the total protein content of the oilseed “Time for wet grinding, blending, or stirring was not included
above; however, more stable emulsions are likely to be formed in a prolonged extraction as pointed out earlier. According to Lusas et cd.,* a period of 40 min was sufficient for extracting both soybean oil and protein while Embong and Jelen’l employed blending up to 15 min for rapeseed oil extraction followed by a prolonged period of l-4 h of stirring. Apart for the studies mentioned above, there is a lack of information on the kinetics of aqueous extraction from different oilseeds which has impeded rational process design and development. mentioned
Downstream
processes for oil and protein recovery
The processes involved in the separation and recovery of protein and oil include a demuls~~cation step followed by separation of the aqueous and oil phases by centrifugation. Demulsification has been carried out in a number of different ways. The aim of demulsification is to promote or accelerate (in a thermodynamic sense) the mechanisms responsible for separation of the components (Figure 7). The tendency of emulsified components to separate into distinctive phases can be understood from the Second Law of The~~ynamics as explained by Tadros6’ The free energy for the formation of an emulsification is given by AGfo’=’ = yO,Aa - TASconf
Enzyme
Microb. Technol.,
2gr*Ap
(3
v=X-
where v is the creaming or sedimentation velocity, r is the radius of the oil droplet, Ap is the difference in density between the continuous and dispersed phase, and q is the dynamic viscosity. Some of the other breakdown processes are also related to the oil droplet size and the emulsion viscosity such as Ostwald ripening, which represents the growth of large droplets at the expense of smaller ones caused by excess (Laplace) pressure and coalescence.
L.__l\
(2)
where yoWis the interfacial tension and Aa is the increase in interfacial area. During emulsification, the interfacial area a between the oil and water phases is greatly increased as a result of the oil droplets subdividing into much smaller units; this is accompanied by increased interfacial energy yoWAa. The formation of a large number of droplets is accompanied by an increase in the total entropy of the system. This increase in entropy facilitates emu~si~cation although its value is relatively small compared to the change in the interfacial free energy. Hence, YowAa> TAS and therefore AGfo” is large and positive. Thus, the process of emulsification is not spontaneous so that, with time, the droplets tend to aggregate and/or coalesce to reduce the total energy of the system. On the other hand, when surface active agents (surfactam) are present in the system, they tend to get adsorbed at the interface which decreases the interfacial tension and 410
results in a much more stable emulsion. This is precisely what happens in the aqueous extraction process; the proteins present in high concentration may act as surfactants. Among the various breakdown mechanisms which occur in emulsions (Figure 7), creaming or sedimentation of oil droplets due to the effect of gravity follows Stoke’s law:
crkhnins
phase / inversion
Figure 7 emulsion
1996, vol. ‘19, November
1
sedirnthation
floculhon
coale~cence’Ostwald ripening I
A schematic representation of mechanisms leading to breakdown” (reproduced with permission)
Edible Demulsification operations like centrifugation can promote coalescence and creaming, but sometimes a more complete separation is obtained through phase inversion. Temperature changes can help make the emulsion less stable, thereby facilitating separation. Earlier studies have addressed demulsification in an aqueous process. In coconut,62 attempts to break the protein stabilized oil-in-water emulsion included heating and centrifugation63; freezing and thawing@!; enzyme action followed by freezing and thawing; and chilling and thawing the coconut cream obtained after centrifugation. The last method requires significantly lower energy compared to when freezing alone was employed. In this last process, the emulsion was centrifuged before chilling and thawing to obtain better packing of the coconut oil globules. The centrifugation step allowed the packing of cream oil globules which crystallized on lowering the temperature. On thawing, the oil globules lost their spherical shape and coalesced to form large droplets of varying sizes. Preliminary centrifugation, besides helping to pack the oil before breaking the emulsion, can also be used to remove undissolved solids after the extraction. The removal of solids present in high percentages in the dispersion of some oilseeds like coconuts and peanuts is essential for efficient recovery of oil by centrifugation.46 The solids are mainly fibrous materials, carbohydrates, and proteins depending upon the pH employed in the extraction step. An alternative method for solid-liquid separation besides centrifugation is reported to be filtration through vibrating or pressure-type screens. Like demulsification, the selection of the most appropriate method of solid-liquid separation depends on the physical nature of the material in the dispersion as well as the operating costs.“h.51 Hagenmaier et al.7.50 used shearing to promote phase inversion and breakage of the coconut emulsion. Another method of achieving phase inversion originally described by Sugarman ” involves the addition of clear oil to the emulsion. With the aid of high shear and temperature, almost all the oil can be freed from the emulsion.46 This is possible due to the reduction in the level of moisture below a threshold value above which the emulsion cannot be broken.J6 Barrios er ul. I increased coconut oil yield from 69.379.4% by aqueous enzymatic extraction simply by adding extra water after the extraction to separate the cream through flotation followed by centrifugation; recycling the water which carries the enzyme; and draining and pressing the coconut meal after the extraction. This procedure considerably decreased the amount of water required for the process and the effluent to be treated and disposed; it also permitted the possibility of enzyme reuse which reduced the costs. An alternative method proposed for increasing the amount of free oil was by stirring the emulsion after extraction.5’ Appropriate levels of stirring promote coalescence of oil droplets which results in the formation of a greater amount of free oil after the separation step. These conclusions were reported for rapeseed, peanut,6357 and sunflower oil.54 After demulsification, the separation of the two or three phases (in case solid removal is not complete) is carried out by centrifugation which is one of the key steps in the aque-
oil extraction:
A. Rosenthal
et al.
ous process. Depending on the material and extraction conditions, oil can either be recovered as free oil or as an oil-in-water emulsion.46 Eapen et aL5 verified that for the alkaline extraction process, moisture conditioning prior to flaking and dispersing peanuts greatly influenced the separation of oil and protein by centrifugation. In a study of rapeseed oil extraction, Embong and Jeler?’ reported that, above a given critical rotational speed, the time required for the centrifuge to attain speed had a significant effect on oil release. Indeed, the initial acceleration had a greater effect than the final speed itself. The greater the acceleration to a given speed, the greater the yield. The authors attributed this observation to the prevalence of different flow patterns in the system under different conditions with turbulent flow patterns possibly contributing to more oil release from the solid fibrous particle matrix at higher acceleration rates.5’ A more conclusive explanation cannot be drawn from this study since precise values of acceleration and related data are not reported. After the separation step, depending on the process employed, most of the protein is either present in the solid phase or the aqueous phase and can be recovered by isoelectric precipitation’ as mentioned above. The aqueous phase can eventually be concentrated and fractionated by membrane technology (ultrafiltration and reverse osmosis). The resulting protein and carbohydrate concentrated solutions can then be dried by spray or freeze drying. Optimization of the complete process should not only be based on obtaining the highest possible oil yield, but should also consider the ease with which the resultant emulsion can be broken. Table 4 lists relevant process parameters and response variables which must be considered to optimize aqueous extraction processes; these parameters are equally relevant to enzymatic extraction processes which are considered later. In particular, some parameters like oil droplet size and viscosity of the bulk emulsion are very useful in evaluating emulsion stability.h’
Use of enzymes in aqueous extraction process Cell-wall degrading enzymes can be used to extract oil by solubilizing the structural cell wall components of the oilseed. This concept has already been commercialized for the production of olive oil and has also been investigated for other oil-bearing materials.‘0 Table 5 shows the oil extraction yields obtained with different materials and by using different types of enzymes. As evident from Table 5, enzyme mixtures with combined activity give, in general, better results than individual enzymes. Besides carbohydrases, proteolytic enzymes were also found to improve yields of oil and protein by hydrolyzing the structural fibrous protein in which fat globules are embedded.s6,67 Thus, Yoon et al.56 reported an improvement in lipid extraction from soybean using only proteolytic enzymes which resulted in a final yield of 86% compared to 62% in the process carried out without enzymes. In the same study, the simultaneous extraction of protein increased from 62 to 89%. In the same way Olsen,68 according to Adler-Nissen,36 carried out a proteolytic hydrolysis on an acid-washed full-fat soya flour.
Enzyme Microb. Technol.,
1996, vol. 19, November
1
411
Review Table 4
Aqueous
and enzymatic
extraction
Process
process optimization ----Main
steps, parameters,
Parameter/operation
Aqueous
extraction
and response variables Response variable
-PH -particle size -power input during mixing extraction -extraction time
-total oil yield -free oil yield -emulsion stability (oil droplet size: low shear stree viscosity; rate of cream~ng/phase separation) -operational costs
-simultaneous or separate reaction/extraction -individual or mixture of enzymes; concentration or level of activity of enzymes -degree of substrate hydrolysis (for different pHs, time, temperature of reaction)
-total oil yield -free oil yield -emulsion stability toif droplet size; low shear stree viscosity; rate of creaming/phase separation -0perafional costs (energy, enzymes, etc.)
-Demulsification
-freezing and thawing -chilling and thawing ~entrifugat~on; oil addition (phase inversion) -high shear stress (phase inversion)
-free oil/protein yield -operational costs
-Final Oil and Protein Recovery
-centrifugation -dehydration ~oncentration/separation -water recycling/enzyme
Enzymatic
process
Downstream
process
(f.i. membrane reutilization
The initial acid wash allowed the recovery of about onethird of the oil originally present in the flour. The hydrolysis of the washed flour (approximately to a degree of 10%) followed by a ~entrifugat~on resulted in an oil yield of 60%. The sludge originating in the centrifuge was then hydrolyzed once more in a similar way, thereby releasing about 99% of the oil into the liquid phase which was partially emulsified with protein hydrolysate.‘6 It is very difficult to compare the results reported in different studies in order to select the best enzyme combinations for a given oilseed. As evident from Table 5, different studies suggest different combinations of enzymes for improving oil extraction. It may be noted that most of these studies employ different extraction conditions such as different pH values, power input, particle size, and reaction temperature which render the results highly specific to the experimental conditions. In the case of coconut, McGlone et a1.67 and Barrios et al. ’ substantially increased the extraction yield of coconut oil by combined treatment of polygalacturonase, IX--amylase, and protease in an aqueous system, thereby obtaining final yields as high as 80%; however, Christensen30 reported oil extraction yields above 90% using a galactomannase combined with a polysaccharide-enzyme degrading complex. All the studies agreed that different enzymes are required to degrade components of the structural cell wa1130,67.69which include mannan, galactomannan, arabinoxylogactan, and cellulose.30 The enzymes did not influence the emulsion stability during coconut oil extraction and the emulsion formed was very unstable resulting in rapid oif separation.67 Barrios et al.’ also reported the significant effect of pectinolytic enzymes on the extraction (attributed to the mannase activity of the commercial enzyme) but pointed out the increased cost of extraction at higher concentrations of pectinase. Table 6 shows the effect of enzyme concen~ation and reaction time on the extraction yield. It 412
Enzyme Microb. Technol.,
technology)
can be seen that higher amounts of enzyme and higher reaction times, in general, correspond to an increase in the extraction yield but with diminishing returns at higher concen~ations. The effect of pectinase concentration on extraction yield and the corresponding cost of the final product is given in Table 7. It can be seen that the increase in oil yield is propo~ionally smaller than the increase in cost which suggests the need to optimize the process in terms of enzyme costs as previously pointed out in Table 4. Reducing enzyme costs by recycling the liquid phase after the extraction, as previously considered, is clearly impo~ant. In the case of avocado, Buenrostro66 obtained better extraction yields by using cx-amylase alone which resulted in an extraction of 75% of the original oil content compared to 65% with the triple enzymatic mixture of polyg~a~turonase, a-amylase, and protease. Freitas et al.” were also able to improve the avocado oil yield by using mixtures of commercial prep~ations with different c~bohy~ase activities. In the case of olives, most research studies have concluded that pectinase and cellulase are the most effective enzymes for increasing oil yields.7’-80 Fantozzi7’ showed that treatment of ground olive paste with a pectinolytic/ cellulolytic enzyme after single pressing extraction could increase the oil yield by up to 20% depending on the olive variety. The olive water originating from milling could be used as a fermentation medium for microorganism growth and pectic enzyme production.8’-85 The enzyme produced could then be concentrated and directly used for mechanics extraction of olive oil, thereby allowing a reduction in the pollution load, a decrease in process costs, and an increase in process yield.” Cheah et al.“” extracted 57% of the palm oil in an aqueous process after treating the palm mesocarp with a cellulase preparation. The untreated material and the one treated with a pectinase resulted in similar yields of about 28%.
1996, vol. 19, November
1
Edible Table 5
Enzymatic
Oilseed Rapeseed
Soybean
coconut
Avocado
Sunflower
Peanut
aqueous extraction
for different oil-bearing
materials
in comparison
Enzyme Control (aqueous without enzyme) Pectinase (Pectinex ultra-sp) Cellulase Multi-carbohydrases Wiscotyme IZOL! Pectinase (NovoZyme 249) Cellulase (NovoZyme 465) Pectinase (NovoZyme 249) + Cellulase (NovoZvme
465)
Control (aqueous without enzyme) Protease (Alcalase) Protease (Sigma) Control (aqueous without enzyme) Pectinase (Clarex) + alpha-amylase (Tanase) + protease (HT-proteolytic) Pectinase (Irgazyme) + alpha-amvlase (Tanase) + protease (HT-proteolytic) Pectinase (Petcimex) + pectinase (Clarex) + alpha-amylase (Tanase) + protease (HT-proteol~ic) Pectinase (Clearzyme) + pectinase (Clarex) + alpha-amylase (Tanase) + protease (HT-proteolytic) Pectinase (Rohapec) + alpha-amylase (Tanase) + protease (HT-proteolytic) Beta-glucanase (brew-n-zyme) Beta-Glucanase (brew-n-zyme) + pectinase Klarex) + alpha-amylase (Tanase) + protease (HT-proteol~ic) Control (aqueous without enzyme) Alpha-amylase (Tanase) Alpha-amylase + protease Alpha-amylase + cellulase Cellulase + protease + alpha-amylase Control (aqueous without enzyme) Cellulase (CGA) Alpha I,4 galacturonide glican~hydrolase Cellulase (CGA1 + alpha I,4 galacturonide glicano-hydrolase (Ultrazym)
(Ultrazym)
Control (aqueous without enzyme) Protease (pepsin-Merck) Cellulase (CGA) Alpha I,4 galacturonide glicano-hydrolase (Ultrazym) Protease (pepsin-Merck} + cellulase (CGA) Protease (pepsin-Merck) + alpha I,4 galacturonide glicano-hydrolase (Ultrazym) Cellulase (CGA) + alpha 1,4 galacturonide glicano-hydrolase (Ultrazym) Protease (pepsin-Merck) + cellulase (CGA) + alpha I,4 gaiacturonide glicano-hydrolase KJltrazym)
Lanzani’ obtained peanut oil yields of 74-78% by aqueous extraction using protease, cellulase, and a-1,4-galacturonide glycano hydrolase either separately or in combination; the extraction without enzymes resulted in a 72% yield. In the same study, greater recoveries of oil from sunflower resulted by using cellulase and CY-1,4-galacturonide glycano hydrolase enzymes which gave a final yield of around 52%. That value is significantly higher than 30% obtained without enzymes.’ Dominguez et al.86 obtained an increase in the sunflower aqueous extraction yield up to 30% of the total oil by using mixtures of cellulase and pectinase. Bocevska et al.” evaluated a group of commercial enzymes for aqueous extraction of corn germ oil and con-
oil extraction:
A. Rosenthal
et al.
to control
Concentration or activity
Oil yield (%)
Reference
2% 300 unit 2.5% 0.2% 0.9% 0.4:0.1%
53.9 71.4 55.4 71.3 70.0 54.2 80.2
65 65 65 65 65
0.2% 0.2%
62.0 84.0 86.0
56 56 56
0.1:0.1:0.1%
12.0 80.0
1 1
0.1:0.1:0.1%
89.3
1
0.1:0.1:0.1:0.1%
87.6
1
0.1:0.1:0.1:0.1%
89.4
1
0.1:0.1:0.1%
83.5
1
0.3% 0.1:0.1:0.1:0.1%
14.4 93.8
1 1
?.O% 1.0% 1 .O% 1.0%
2.0 70.0 67.0 67.0 62.0
66 66 66 66 66
3% 3% 1.5:1.5%
30.0 44.0 44.0 52.0
9 9 9 9
3% 3% 3% 1.5:1.5% 1.5:1.5%
72.0 78.0 75.0 74.0 78.0 76.0
9 9 9 9 9 9
1.5:1.5%
74.0
9
1.0:1.0:1.0
78.0
9
cluded that a carbohydrase (mainly cellulase) complex from Trichoderma reesei was most effective; it released 84.7% of the total oil, 74.3% of which appeared as free oil after cent~fugation. In the case of rapeseed, an extraction yield of 78% was obtained with the pair of protease and or-1,4_galacturonide glycano hydrolase enzymes.’ A comparable result in the same study was obtained with the use of only a protease. 01sen3’ described an aqueous extraction process in which cell wall-degrading enzymes (pectinase, cellulase, and hemicellulase combinations) were used to degrade rapeseed. Marek et ai8* reported an aqueous process resulting in 35% of the original oil content in the residue by using only cellulolytic enzymes on a previously steamed rapeseed
Enzyme Microb. Technol.,
1996, vol. 19, November
1
413
Review Table 6 Effect of enzyme concentration coconut oil extraction yield
Pectinase 0.1 0.075 0.05 0.05 0.075 0.0375 According
and reaction time on
Protease
Amylase
Reaction time (min)
Extraction yield (%)
0.1 0.05 0.05 0.05 0.05 0.025
0.1 0.05 0.05 0.05 0.05 0.025
30 60 60 90 90 90
79.3 69.9 49.6 74.2 76.4 62.4
to Barrios et al.’
flour. Deng et a1.65obtained a yield of 80% in rapeseed oil extraction by using a commercial carbohydrase complex in the aqueous process and reported little effect of a single cellulolytic enzyme on the extraction yield. Besides the types and dosages of the enzymes, the degree of grinding can also affect oil yields. Deng et al.65 reported that the main parameters which affected the enzymatic aqueous extraction from rapeseed were the degree of grinding, pH of dispersion, incubation temperature and time and centrifugation conditions. These conclusions confirm that different results in the oil yield obtained with different enzymes for the same oilseed not only reflect the differences in the enzyme efficiency with respect to oil release, but also the effect of other parameters on the overall efficiency of the process. It seems that the main parameters affecting the aqueous extraction process (see above) coincide with the ones that affect the enzymatic aqueous process. Some of these are related to the extraction itself and others to the optimum conditions for enzyme activity. Thus, the main task in the development of a robust process is to simultaneously consider the parameters which satisfy both steps, the enzyme reaction and extraction, whether carried out separately or in one operation. Another important aspect to be considered is the difference between the role of carbohydrases (pectinase, cellulase, and hemicelluloses) and the proteolytic enzymes. The action of the carbohydrases are specific to cell wall hydrolysis which allows a higher release of oil into the aqueous medium. On the other hand, the action of proteolytic enzymes seem to be related not only to the hydrolysis of the membranes surrounding lipid bodies and the cytoplasmic protein (see below), but also to the modification of one
Table 7 Effect of pectinase concentration tion yield and oil cost increment
Concentration 0.10% 0.15% 0.20%
on coconut oil extrac-
Relative extraction yield (%I
Pectinase costa ($U.S I-’ of extracted oil)
100 109 116
0.20 0.275 0.346
According to Barrios et al.’ (costs reported in 1990) ‘Based on a pectinase cost of U.S. $7.67 kg-’
414
Enzyme Microb. Technol.,
essential functional property of the protein relating to the aqueous process ---its emulsifying capacity. Many studies have shown that the emulsifying capacity of some proteins increases during enzyme proteolysis until a certain degree of hydrolysis is achieved. The emulsifying capacity then starts to decrease; however, the op osite happens to the $ stability of the resulting emulsion.8 -9J Thus, in contrast with the other hydrolytic enzymes used in the aqueous processes, proteolytic enzymes can potentially have a negative or positive effect on the process depending on the degree of hydrolysis of the protein. Although more oil can be released from the lipid bodies as the proteolytic action is developed, the increase in the emulsifying capacity can lower extraction levels of free oil. Hence, the extent of proteolytic action must be optimized both to obtain a higher oil yield and also to obtain a less stable oil-in-water emulsion.
Solvent extraction vis a vis aqueous extraction processes The main features of aqueous extraction and solvent extraction are compared and contrasted in Table 8 and can be discussed in terms of environmental, economic, and quality aspects.
Environmental
and economics aspects
The main environmental concern relating to the conventional solvent-based oil extraction is hexane loss and associated pollution problems (Figure 8). Hexane, like other VOCs, can react with pollutants, principally oxides of nitrogen, in the presence of sunlight to form ozone (0,) and other species collectively known as photochemical oxidants. Although ozone is essential in the upper atmosphere to shield against the UV radiation from the sun, excess ozone is undesirable at ground leve1.95 The emission of VOCs by various industrial sectors within the UK has been reviewed and discussed by Finlayson-Pitts.96 Global background levels of ozone which used to be only about 10 ppb around a century ago today are typically in the 20-50 ppb range (1 ppb = 1 part in lo9 parts of air). It is feared that even at relatively low levels approaching 100 ppb, ozone may have significant adverse health effects. Furthermore at even lower concentrations, it can affect sensitive agricultural crops and forest ecosystems. As a result, increasing levels of ozone are of rising concern in urban and “downwind” suburban/rural areas throughout the world.96 Although hexane has a medium-low photochemical ozone creation potential (POCP) and despite recent developments in the techniques for minimizing VOC emissions, the food industry is still responsible for about 7.5% of the mass emissions from stationary sources.97 Among the food industries, the vegetable oil sector is principally responsible for the high VOC emission level. For example, according to Mustakas,‘4 solvent losses in American soybean oil plants did not exceed 4-8 I-’ ton of seed processed. It was also reported that a well-designed and operated plant in Canada lost about 1.5 1-l ton of processed rapeseed. Considering that the annual level of soybean processed for oil extraction in the U.S. is about 56 million tons, the annual level of
1996, vol. 19, November
1
Edible oil extraction: A. Rosenthal et al. Table 8
Aqueous
extraction
vis a vis solvent extraction
and possible alternatives
Disadvantages
Advantages
Aspects
Simultaneous
Economic
for aqueous process improvement
oil and protein
recovery No energy required for organic solvent stripping No high investment required for VOCs emission monitor and control
Environmental
No VOCs emission and ozone formation
Safety
Less fire and explosion
Product quality
High quality without degumming requirement
Possible improvements
Lower oil yield
Enzyme utilization
High energy required for water removal and high effluent generation
Water recycling
Enzyme costs Demulsification
requirement
Process optimization
Membrane technology untilization Optimization and use of alternative methods of demulsi~catjon
hazard
Less denatured protein with higher biological value Products usually without antinutritional factors
hexane emission from the soybean oil industries could alone be as high as 210-430 million 1 assuming that emission levels have been maintained the same as recorded by MustakasZ4 in 1980. In the U.K., the edible oil sector was estimated to groduce VOC emissions of greater than 1Oktia in 1991. 5 It may become necessary for the vegetable oil industry to find an alternative extraction method which eliminates or drastically reduces the use of all volatile organic compounds or at least reduces their emission levels significantly. It is therefore appropriate that this sector should be considerin 95 development of new environmentally clean processes. The success of aqueous-based extraction processes can achieve this objective.”
-“g~eenhiwse gases6
Volatile Organic Compound
_toxic -carcinogenic
(VOCs, e.g. hexane) +
at ground~lewl!: -heaM! hezard -wps_ damage
Figure 8
Environmental pound emissions
consequences
of volatile organic com-
Besides environmental problems, the use of hexane is also worrying from the safety point of view. Hexane is highly inflammable and in spite of elaborate precautions which have been developed to avoid fire and explosion hazards, there is still the danger of severe accidents.24 Aqueous processes eliminate the problem of solvent safety, thereby resulting in lower fire hazard and less operational danger. 2.9.46.4X.51-53.99 Aqueous processes can potentially be more cost effective since investments relating to solvent recovery, process safety, and solvent loss control systems will be much lower. Moreover, such a process seems to allow smaller installations which offers significant economical advantages. 1?,9.46.48,99 Today it appears that the aqueous process may not be as efficient as the solvent extraction process for most oleaginous materials.‘,5’.53.99Furthermore, it must also be recognized that some of the savings mentioned above can be offset by costs relating to demulsification, water removal from the final product, and hygiene requirements of a tiler process.99 There seems to be a distinct possibility for developing industrial-scale aqueous processes as a result of technical developments relating to the use of enzymes and the increasing importance of environmental requirements. The possibilities are more encouraging for those oils with high commercial prices such as olive and avocado oil. There are few detailed economic analyses but Hagenmaier et ~1.‘~’ developed an economic analysis for a coconut oil aqueous extraction plant in the Philippines which would process 250 tons day-’ of dehusked coconuts resulting in 8,700 tons of oil. The project was assumed to be a high-risk venture requiring a high return on investment within a short time. The analysis was carried out with costs prevailing in 1974. In order to minimize the costs, it was assumed that fuel and electricity were bought and that wet by-products (residue and insoluble proteins) were disposed at the factory gate.
Enzyme Microb. Technol.,
1996, vol. 19, November
1
415
Review The estimated fixed capital investment for the plant (U.S. $1.9 million in 1974) was much lower than the annual operating expenses (U.S. $3.1 million); the annual revenues (approximately U.S. $3.8 million) included oil and coconut skim milk sales. The final rate of return (preincome tax) was 19.7% with a return on investment (internal rate of return pattern, preincome tax) at 22.1% which was close to the value of 20% considered to be the minimum acceptable return for new ventures. Sensitivity analysis showed that the selling price of the coconut milk solids was the most critical parameter and the oil yield (assumed to be 91% in the project) was also a key aspect. The effect of enzymes on the economics of the aqueous process will depend on the oilseed and the balance of costs/ benefits resulting from enzyme utilization. Barrios et al.’ in another study on coconut aqueous extraction reported that the use of an enzyme mixture consisting on 0.075% (w/v) pectinase and 0.05% (w/v) each of protease and amylase resulted in oil extraction yields as high as 76.4% when compared with a nonenzymatic process yield of less than 20%. The implied additional enzyme costs was U.S. $620ton-’ oil (in 1990 prices) which decreased to U.S. $107.5 ton-’ oil by recycling the enzyme. It is difficult to compare this study with that of Hagenmaier et ~1.‘~~because of the high discrepancy in the extraction yields; however, the enzyme costs assumed by Barrios et al.’ would be expected to result in an increase in the process costs by approximately U.S. $1 million (based on 1990 costs) which would render the process unfeasible. The use of enzymes in aqueous processes will certainly depend on the ability of enzymes to increase the yields without turning the process costs prohibitive. Table 9 illustrates how income can be enhanced in the case of soybeans and olives by using enzymes in an aqueous process. Both Protease A and B increase oil and protein extraction yields from soybean; however, protease B is about 44 times more expensive than Protease A mainly due to a much higher
Table 9
Additional
income from sales that can be obtained
degree of purity. It is evident from Table 4, that the additional income which can be obtained with Protease A is much higher than the enzyme costs. It has been assumed that the product sells at current market rates. On the other hand in the case of Protease B, the additional enzyme costs are almost twice the additional income that can be obtained from the sales in comparison with the aqueous process. Thus, the enzyme cost is an essential factor to turn the process using Protease A into a more profitable one (with a net profit as high as 35% compared to the aqueous process) and also to turn the Protease B a less profitable process (about 53% in terms of net profit) in comparison with the aqueous process. The use of a multiactive enzyme complex consisting of pectinolytic, hemicellulolytic, and polysaccharidase for olive oil extraction could result in an increase in the oil yield by about 10% (or 2 kg oil 100 kg-’ fruit) but could not result in a positive net profit compared to the aqueous process as evident in Table 9: it possibly demands a better choice of enzymes. The economic comparison between aqueous processes (with or without enzymes) with solvent-based processes depends on specifying details relating to the aqueous processes such as the type of separation steps (e.g., using membranes) and water and enzyme recycle and reutilization which must be considered and improved in order to turn the aqueous process into a more competitive technology. Membrane separation and water recycle will directly influence water requirements for the process and thereby determine the energy required for further water removal and effluent treatment. In the absence of information relating to these alternatives, it is difficult to make a detailed economic assessment for all oilseeds at this stage.
Oil and protein quality Processes currently employed are primarily targeted to produce edible oil. As a result, very little attention has been
using enzymes
in aqueous process Olive
Soybean Enzyme
Protease A
Protease B
Enzyme price Enzyme requirement (kg) Reported increase in oil producedC (kg) Reported increase in protein producedC (kg) Assumed price of oil (kg-‘) Assumed price of protein (kg-‘) Additional income from sales’ Oil Protein Total Net profit”
f l;:5$ekg-‘a
f68.40 100 g-lb 0.05= 3.8= 6.0= f0.40 f2.00
Basis “Cost bCost ‘lover dSale “Sale
416
4.6== 6.4= f0.40 f2.00
f1.50d f12.00” f13.50 -f20.70 (-53.53%)
f 1 .84d f12.80” f 14.64 + f 13.09 (+35.04%)
Multiactivity
enzyme
f30.75 kg-‘” 0.25 2.0 f2.00 f4.00d f4.00 -f3.70 (-9.25%)
is 100 kg soybean or olive based on quotation with the manufacturer company; may vary with quantity based on catalog; may vary with quantity nonenzymic process” prices based on personal communication obtained from a trade company; may vary with amount prices based on personal communication obtained from a trade company; may vary significantly with amount and purity
Enzyme Microb. Technol.,
1996, vol. 19, November
1
Edible given to the quality of the protein residues in the context of human consumption. The residue (or flakes) from solvent extraction processes must be heated to remove solvent which requires a high amount of thermal energy not only to strip extraction solvent from micellia but also for improving extractibility prior to the extraction operation.24*46 In contrast, it is well known that the drastic thermal treatment of oilseeds reduces the quality of extracted oil and proteinaceous materials.56 The main nutritional effect resulting from excessive heating in the conventional process is a decrease in the nutritional availability of some essential amino acids (mainly lysine) which results from the Maillard reaction between free amino groups from proteins and carbonyl groups from reducing sugars. Since the overall quality of the protein is determined by the biological availability of each of the essential amino acids which forms the protein, the Maillard reaction causes not only a negative nutritional effect in the essential amino acids involved in the reaction, but also on the biological value of the whole protein. A similar effect occurs as a result of the hi h temperatures generated by expellers during the pressing. 9 rT6* In contrast, aqueous processes avoid serious damage to the proteins of the seed which allows the production of food-grade instead of feed-grade protein products.46 Aqueous processes also allow inactivation or removal of antinutritional factors and other undesirable substances which are present in some oilseeds and can reduce the overall quality of the products or nutritive value of their proteins or can be toxic to humans.2*46352,53 In this context, it has been demonstrated that caffeic and chlorogenic acids, which cause sunflower seed meal to turn dark green or brown when wetted, can be removed by aqueous extraction.86*‘“’ In the same way, toxic sulfur compounds (goitrogens) present in rag:seed can be successfully removed by aqueous extraction. In the case of peanuts, the addition of some chemicals such as hydrogen peroxide and sodium hypochlorite in the aqueous extraction medium has also proven to be effective for destroying aflatoxins.lo3 Aqueous processes are also effective in removing the bitter, poisonous alkaloids from lupin seeds58 and also the nongossip01 pigments from cottonseed.48 It is also believed that the application of aqueous processing to soybean might provide an opportunity for the removal of sugars and other compounds which cause the problems of flatulence and offflavor.46 With regard to oil quality, it is reported that during aqueous processing, phospholipids are separated from the oil so
Table 10
Quality characterization
Quality parameter Acid value Peroxide value Saponification value Iodine value Phospholipids
oiVo4
A. Rosenthal
et al.
there is no need for degumming.5”87 The quality evaluation of different oils obtained by aqueous enzymatic processes in comparison to the ones extracted by hexane is shown in Table IO. In general, most of the studies carried out to evaluate the oil quality do not report significant differences between aqueous and solvent-based processes. Gunetileke and Laurentius6* compared coconut oil obtained by aqueous process with a commercial one and found no significant differences in the percentage of free fatty acids, refractive index, color on Lovibond scale, saponification value, iodine value, Reichert value, Polenske value Kirschner value, and the fatty acid composition of the oil. Furthermore, the product which was obtained in a pilot plant resulted in oil of superior quality compared to the one obtained by pressing the dried meal in a conventional screwtype expeller. Hagenmaier et al.’ obtained an aqueous extracted coconut oil with 0.4% free fatty acids which was considered superior to oil obtained by crushing copra which has a typical free fatty acid content of approximately 5%. McGlone et al.67 compared the quality of coconut oil obtained by aqueous enzymatic process with the Official Mexico Standards. Without any further purification, the oil obtained after reaction and centrifugation was reported to be an excellent quality and with a simple deodorization process it could be readily used in existing food applications. Embong and Jelen” compared rapeseed oil obtained by laboratory batch aqueous process to Soxhlet-extracted oil and industrial crude oil. The aqueous extracted product was slightly superior to the Soxhlet-extracted oil, but much superior to the industrial oil in terms of the sulfur content and peroxide value. There were no differences in the nonsaponificable matter in any of the oils investigated in the study. Kim and Yoon’04 compared iodine, acid, and saponification values; nonsaponificable matters; iron; sterol, tocopherol, and phosphorus contents; and fatty acid composition of soybean crude oils obtained by aqueous and hexane extraction and did not find any significant differences in the values of these parameters. The oxidative stability of the oils obtained by hexane and aqueous extraction were also not significantly different during ten weeks of storage; however, Yoon et a1.56concluded that aqueous-extracted soybean oil was slightly darker than the hexane-extracted one after comparing Lovibond units of red yellow colors for both products. In contrast, Bocevska et al.*’ obtained a corn germ oil with a lighter yellow color using an enzymatic aqueous process when compared to the commercial product; he concluded that the aqueous product may be refined with less bleaching earth. It was also suggested that a thermal discoloration may that
of oils obtained by aqueous enzymatic extraction Soybean
oil extraction:
in comparison
Corn germ 0iV
with standard industrial processes Rapeseed oils5
Hexane extraction
AEP
Expeller extraction
AEP
Hexane extraction
AEP
0.5 195 136 0.7
0.4 195 133 0.6
1.11 1.10 1.40
1.50 0.00 0.02
1.92 2.53 176 108 -
2.06 4.18 173 107 -
Enzyme Microb. Technol.,
1996, vol. 19, November
1
417
Review
be useful. In the same study, it was also shown that the quality of corn germ oil obtained by an aqueous process had free fatty acids content and primary and secondary oxidation product levels in the same range of values as obtained for oil by solvent extraction. The aqueous product also showed good stability especially when extraction was preceded by hydrothermal pretreatment to inactivate the lipase in the grain.
10. Il.
12.
13.
Concluding remarks Processes for the extraction of edible oils based on aqueous extraction media with or without enzymes offer several potential advantages over solvent-based processes currently used. These advantages relate to environmental, safety, and possibly economic aspects; however, further studies leading to the development of more effective enzymes specific for each oilseed have to be carried out. Detailed investigations are also needed to select, in a rational way, the downstream process operations which permit ready separation of the released oil. Other aspects relating to water recycling and enzyme reutilization must also be considered in order to render the process more attractive. For process optimization, one must not only consider the yield obtained in the extraction step, but also the stability of the resulting oil-inwater emulsion and its implications on downstream processing. Although basic information on aqueous processes is available, it is restricted to laboratory-scale studies. Systematic process engineering investigations and economic evaluation of the processes are necessary before considering scale-up.
The support of Biotechnology and Biological Sciences Research Council (BBSRC Grant 45/FOO700) and Brazilian Company for Agricultural Research (EMBRAPA) are gratefully acknowledged. The authors also wish to thank Prof. D.L. Pyle, Dr. K. Niranjan, and Mr. A. Rosenthal for their permission to reproduce Figure 1 in this article.
3. 4.
5.
6.
7.
8. 9.
418
Barrios, V. A., Olmos, D. A., Noyola, R. A., and Lopez-Munguia, C. A. Optimization of an enzymatic process for coconut oil extraction. Oleagineux 1990, 45, 3542 Lusas, E. W., Lawhon, J. T., and Rhee, K. C. Producing edible oil and protein from oilseeds by aqueous processing. Oil Mill Gazer 1982, 4,28-34 Caragay, A. B. Pacing technologies in the fats and oils industry. J. Am. Oil Chem. Sot. 1983, 60, 1641-1644 Subrahmanyan, J., Bhatia, D. S., Kalbag, S. S., and Subramanian, N. Integrated processing of peanut for the separation of major constituents. J. Am. Oil Chem. Sot. 1959, 36,66-70 Eapen, K. E., Kalbag, S. S., and Subrahmanyan, V. Operations in the wet-rendering of peanut for the separation of protein, oil, and starch. J. Am. Oil Chem. Sot. 1966, 43,585-589 Rhee, K. C., Cater, C. M., and Mattil, K. F. Simultaneous recovery of protein and oil from raw peanuts in an aqueous system. J. Food Sci. 1972, 37, 90-93 Hagenmaier, R. D., Cater, C. M., and Mattil, K. F. Aqueous processing of fresh coconuts for recovery of oil and coconut skim milk. J. Food Sci. 1973, 38, 516-518 Kim, H. K. Aqueous extraction of oil from palm kernel. J. Food Sci. 1989, 54,491492 Lanzani, A., Petrini, M. C., Cozzoli, O., Gallavresi, P., Carola, C., and Jacini, G. On the use of enzymes for vegetable-oil extraction.
Enzyme Microb.
16. 17.
18.
19.
20. 21. 22.
24.
25.
26.
References
2.
15.
23.
Acknowledgments
1.
14.
Technol.,
1996, vol. 19, November
27. 28.
29.
30.
31.
32.
33. 34.
1
A preliminary report. Riv. Ital. Sostunze Grasse 1975. Lll, X6229 Fullbrook, P. D. The use of enzymes in the processing of oilsceds. J. Am. Oil Chem. Sot. 1983, 60,428A_430A Sosulski. K.. Sosulski. F. W., and Coxworth, E. Carbohydrase hydrolysis of canola to enhance oil extraction with hexane. J. Am. Oil Chem. Sec. 1988. 65, 357-361 Dominguez, H., Nunez, M. J., and Lema, J. M. Enzymatic pretreatment to enhance oil extraction from fruits and oilseeds: A review. Food Chem. 1994, 49, 27 l-286 Bair, C. W. Microscopy of soybean seeds: Cellular and subcellular structure during germination, development, and processing with emphasis on lipid bodies. Ph.D. thesis, Iowa State University, Iowa, 1979 Young, C. T. and Schadel, W. E. Microstructure of peanut seed: A review. Food Struct. 1990, 9, 317-328 Tombs, M. P. Protein bodies of the soybean. Plant Physiol. 1967, 42, 797-8 13 Wolf, W. J. Scanning electron microscopy of soybean protein bodies. J. Am. Oil Chem. Sot. 1970, 107-108 Wolf. W. J. and Baker, F. L. Scanning electron microscopy of soybeans and soybean protein products. In: Scanning EZectron Microscopy (Wolf, W. J., and Baker, F. L.. Eds.). AMF O’Hare, 1981. 239-252 Snyder. H. E. and Kwon, T. W. Morphology and composition. In: Soybean Utilization (Snyder, H. E., and Kwon, T. W., Eds.). Van Nostrand Reinhold Co. Inc., New York, 1987, 19-70 Jacks, T. J., Hensarling, T. P., Neucere, J. N., Yatsu, L. Y., and Barker, R. H. Isolation and physicochemical characterization of the half-unit membranes of oilseed lipid bodies. J. Am. Oil Chem. Sot. 1990, 67, 353-361 Murphy, D. J. Structure, function, and biogenesis of storage lipid bodies and oleosins in plants. Prog. Lipid Res. 1993, 32, 247-280 Bair, C. W. and Snyder, H. E. Electron microscopy of soybean lipid bodies. J. Am. Oil Chem. Sot. 1980, 279-282 Yatsu, L. Y., Jacks, T. J., and Hensarling, T. P. Isolation of spherosomes (oleosomes) from onion. cabbage, and cottonseed tissues. Plant Physiol. 1971, 48, 675-682 Wolf, W. J. and Baker, F. L. Scanning electron microscopy of soybeans, soy flours. protein concentrates, and protein isolates. Cereal Chem. 1975, 52, 387-396 Mustakas. G. C. Recovery of oil from soybeans. In: Handbook of Soy Oil Processing and Utilization (Erickson, D. R.. Ed.). American Soybean Association and American Oil Chemists’ Society, St. Louis. 1980, 49-65 Kofod. L. V. Aqueous enzymatic extraction of rapeseed oil. M. SC. thesis, Institute of Biotechnology, Technical University of Denmark, 1988 Huang, A. H. C. Structure of plant seed oil bodies. Curr. Opinion Struct. Biol. 1994, 4, 494498 Tzen, J. T. C. and Huang, A. H. C. Surface structure and properties of plant seed oil bodies. J. Biol. Chem. 1992, 117, 327-335 Talmadge, K. W., Keegstra, K., Bauer, W. D., and Albersheim, P. The structure of plant cell walls I. The macromolecular components of the walls of suspension-cultured sycamore cells with a detailed analysis of the pectic polysaccharides. Phnt Physiol. 1973, 51, 158-173 Keegstra, K., Talmadge, K. W., Bauer, W. D., and Albersheim, P. The structure of the wall of suspension-cultured sycamore cells based on the interconnections of the macromolecular components. Plant Physiol. 1973, 51, 188-196 Christensen, F. M. Enzyme technology versus engineering technology in the food industry. Biotechnol. Appl. Biochem. 1989,11, 249-265 Olsen, H. S. Aqueous enzymatic extraction of oil from seeds. In: Asian Food Conference Proceedings Bankgok, Thailand, 1988. Reprinted by: Novo Industry A/S, pp A-0604la Bemardini, E. Processing of oilseeds: Rapeseed. In: The New Oil and Fat Technology (Bemardini, E., Ed.). Technologie, Rome, 1973, Part II, 73 Serrate, A. Cl. Extraction of oil from soybeans. J. Am. Oil Chem. Sot. 1981, 3, 157-159 Brekke. 0. L. Edible oil processing. An introduction. In: Handbook of Soy Oil Processing and Utilization (Erickson, D. R., Ed.).
Edible
35.
36.
37. 38.
39.
40. 41.
42.
43.
44.
45.
46.
47. 48.
49.
50.
51.
52.
53.
54. 55. 56.
57.
58. 59. 60.
American Soybean Association and American Oil Chemists Society, St. Louis, 1980 Sherba. S. E., Steigerwalt, R. B., Faith, W. T., Smythe Jr., C. V. Soybean fractionation employing a protease. U.S. patent 3,640,725; 1972 Adler-Nissen, J. A review of food protein hydrolysis-specific areas. In: Enzymic Hydrolysis of Food Proteins (Adler-Nissen, J. Eds.). Elsevier Applied Science Publishers, London and New York, 1986, 78-79 Dominguez, H., Nunez, M. J., and Lema, J. M. Enzyme-assisted hexane extraction of soya bean oil. Food Chem. 1995,54,223-23 1 Dominguez, H., Nunez, M. J., and Lema, J. M. Oil extractibility from enzymatically treated soybean and sunflower: Range of operational variables. Food Chem. 1993, 46, 277-284 Dominguez, H., Nunez, M. J., and Lema, J. M. Enzymatic treatment of soybean simultaneous with the hexane extraction. S&h Eur. Congr. Eiotechnol. Proc. Vol. 1, Firenze, Italy, 1993, MO153 Bhatnagar, S., and Johari, B. N. Microbial enzymes in the processing of oilseeds. Curr. Sci. 1987, 56, 775-776 Tano-Debrah, K., and Ohta, Y. Enzyme-assisted aqueous extraction of fat from kernels of the shea tree Butyrospermum parkii. J. Am. Oil Chem. Sot. 1994, 71, 979-983 Tano-Deb&, K. and Ohta, Y. Enzyme-assisted aqueous extraction of shea fat: A rural approach. J. Am. Oil Chem. Sot. 1995. 72, 251-256 Cheah, S. C., Augustin, M. A., and Ooi, L. C. L. Enzymic extraction of palm oil. Palm Oil Res. Bull. Malaysia Bull. 1990. 20, 30-36 Bouvier, F. and Entressangles, B. Utilization of cellulases and pectinases in the extraction of palm oil. Revue Fruncaise des Corps Grus 1992, 39, 245-252 Smith, D. D., Agrawal, Y. C., Sarkar, B. C., and Singh, B. N. P. Enzymatic hydrolysis pretreatment for mechanical expelling of soybeans. J. Am. Oil Chem. Sot. 1993, 70, 885-890 Cater, C. M., Rhee, K. C., Hagenmaier, R. D., and Mattil. K. F. Aqueous extraction-an alternative oilseed milling. J. Am. Oil Chem. Sot. 1974, 51, 137-141 Johnson, L. A. and Lusas, E. W. Comparison of alternative solvents for oils extraction. J. Am. Oil Chem. Sot. 1983, 60,229-242 Lusas, E. W. and Jividen, Cl. M. Glandless cottonseed: A review of the first 25 years of processing and utilization research. J. Am. Oil Chem. Sot. 1987, 64, 839-854 Southwell, K. H. and Harris. R. V. Extraction of oil from oilseeds using the hot water flotation method. Tropical Sci. 1991, 32, 217221 Hagenmaier. R., Cater, C. M., and Mattil, K. F. Critical unit operations of the aqueous processing of fresh coconuts. J. Am. Oil Chem. Sot. 1972, 49, 178-181 Embong, M. B. and Jelen. P. Technical feasibility of aqueous extraction of rapeseed oil-a laboratory study. J. Inst. Can. Sci. Technol. Aliment. 1977, 10, 239-243 Lawhon, J. T., Manak, L. J., Rhee, K. C., Rhee, K. S., and Lusas, E. W. Combining aqueous extraction and membrane isolation techniques to recover protein and oil from soybeans. J. Food Sri. 1981, 46,912-916 Lawhon, J. T., Manak, L. J., Rhee, K. C., and Lusas, E. W. Production of oil and protein food products from raw peanuts by aqueous extraction and ultrafiltration. J. Food Sci. 1981, 46, 391395 Hagenmaier, R. D. Aqueous processing of full-fat sunflower seeds: Yields of oil and protein. J. Am. Oil Chem. Sot. 1974,51,470-471 Rosenthal, A., Pyle, D. L., and Niranjan, K. To be forwarded 1996 Yoon, S.-H., Kim, L-H., Kim, S.-H., and Kwon, T.-W. Effects of enzyme treatments and ultrasonification on extraction yields of lipids of protein from soybean by aqueous process. Korean J. Food Sci. Technol. 1991, 23, 673-676 Rhee, K. C., Cater, C. M., and Mattil, K. F. Effect of processing pH on the properties of peanut protein isolates and oil. Cereal Chem. 1973, 50, 395404 Aguilera, I. M., Gemgross, M. F., and Lusas, E. W. Aqueous processing of lupin seed. J. Food Technol. 1983, 18, 327-333 Sugarman. N. Processing for extraction of oil and protein simultaneously from oil-bearing materials. U.S. patent 2,762,820; 1956 Bhatia, D. S., Parpia, M. A. B., and Baliga, B. P. Peanut protein
61. 62.
63.
64. 65.
66. 67.
68.
69. 70.
71.
72.
73.
74.
75. 76.
77.
78.
79.
80.
81.
82.
83.
oil extraction:
A. Rosenthal
et al.
isolate-production. properties. J. Food Sci. Technol. (India) 1966, 3, 2 Tadros, T. F. Colloidal aspects of pesticidal and pharmaceutical formulation-an overview. Pest. Sci. 1989, 26, 51-77 Gunetileke, K. G. and Laurentius, S. F. Conditions for the separation of oil and protein from coconut milk emulsion. J. Food Sci. 1974, 39, 230-233 Rajasekharan, N. and Sreenivasan, A. The use of coconut preparations as a protein supplement in child feedings. L Food Sci. Technol. (India) 1967, 4, 59 Roxas. P. G. Recovering oils from oleaginous meats of nuts, beans. and seeds. U.S. patent 3.083.365: 1963 Deng. Y., Pyle, D. L., and Niranjan, K. Studies of aqueous enzymatic extraction of oil from rapeseed. Agri. Eng. Rural Dev. I Co@ Proc. Vol. 1 (Zhang, W., Giro, P. W., Zhang, S. W.. Ed%). International Academic Publishers. Beijing, 1992, 190-195 Buenrostro, M. and Lopez-Munguia. C. A. Enzymatic extraction of avocado oil. Biotechnol. Left. 1986. 8, 505-506 McGlone, 0. C., Lopez-Munguia, C. A., and Cater. J. V. Coconut oil extraction by a new enzymatic process. ./. Food Sci. 1986. 51, 695697 Olsen, H. S. Method of producing soy protein hydrolysate from fat-containing soy material, and soy protein hydrolysate. U.S. patent 4.324.805; 1982 Sihotang. D.. Kandiah. and Balasubramaniam. Recovering coconut protein and oil. U.S. patent 2,005,713A; 1979 Freitas, S. P.. Lago, R. C. A., Jablonka, F. H.. and Hartman, L. Aqueous enzymatic extraction of avocado oil from fresh pulp. Revue Francaise des Corps Gras 1993. 40, 365-371 Fantozzi. P., Petruccioli. G.. and Montedoro, G. Enzymatic treatment of olive pastes after single pressing extraction. Effect of cultivar, harvesting time, and storage. Rils. Ital. Sosfanze Grusse 1977. 54, 381-388 Leone, A. M., Lamparelli, F.. Notte. E. L.. Liuzzi. V. A.. and Padula. M. The use of pectolytic-cellulolytic enzyme system in olive oil production. Product yield and quality. Rii,. Itul. Sostun:e Grasse 1977. 54, 5 14-530 Montedoro. G. Results of some biochemical treatments of olive pastes for centrifugal extraction, Riv. Ital. Sostanze Grasse 1977. 54,45 IL458 Montedoro, G. and Petruccioli. G. Biochemical treatments of whole and pitted olive pulps designed to aid mechanical oil extraction Riv. Ital. Sosfanze Grusse 1978. 5.5, 150-159 Santos Antunes. A. F. D. Use of enzymes in extraction of olive oil. Biol. Inst. Azeite Prod. Oleaginosos 1978. 6, 39-52 Montedoro. G. and Petruccioli. G. Optimization of extraction of olive oil by pressing or by percolation with reference to the use of enzymic preparations and other adjuvants. Medirerrun Zone Symp. Comission Intemazionale des Industries Agricoles et Alimentares, 1978 Duarte Costa, H. M. P. and Sameiro. M. E. M. Quality of olive oil extracted using enzyme preparations. Bol. Inst. A,-cite Prod. O/eaginosos 1978, 6, 25-38 Manachini, P. L., Parini. C., and Fortina, M. G. Pectic substances from Aureobasidium pullulans LV 10. Enzyme Microb. Technol. 1988, 10, 682-685 Giovacchino, L. D. and Mascolo, A. The influence of working techniques on olive oil extraction by continuous system. II. Riv. Ital. Sosfanze Grasse 1988. 65, 283-289 Ranalli, A. and Martinelli. N. Extraction of the oil from olive pastes by biological and not conventional industrial techniques. Industrie Alimentari 1994. 33, 1073-1083 Servili, M.. Montedoro, G. F., Begliomini, A. L., Petruccioli, M.. and Federici, F. Use of endopolygalacturonase produced by Cryptococcus albidus var. albidus for olive oil extraction. lndustrie Alimentari 1989, 28, 1075-1078 Petruccioli, M., Servili, M., Montedoro. G. F., and Federici, F. Development of a recycle procedure for the utilization of vegetation waters in the olive-oil extraction process. Biotechnol, Left. 1988, 10,55-60 Federici, F.. Petruccioli, M., and Montedoro, G. F. Combined physicochemical and biological treatment of olive vegetation waters in pilot-plant scale. Meded. Fat. Landbouwwet. Rijksuniv. Gent. 1991. 56, 1573-1577
Enzyme Microb. Technol.,
1996, vol. 19, November
1
419
Review 84.
85.
86.
87.
88.
89.
90.
91. 92.
93.
420
Servili, M., Begliomini, A. L., Montedoro, G., Petruccioli, M., and Federici, F. Utilization of a yeast pectinase in olive oil extraction and red wine making processes. J. Sri. Food Agric. 1992. 58, 253-260 Montedoro, G. F., Begliomini, A. L., Servili, M., Petruccioli, M., and Federici, F. Pectinase production from olive vegetation waters and its use in the mechanical olive oil extraction process to increase oil yield and improve quality. Ital. J. Food Sci. 1993. 5, 355-362 Dominguez, H., Nunez, M. J., and Lema, J. M. Aqueous processing of sunflower kernels with enzymatic technology. Food Chem. 1995,53,427-434 Bocevska, M., Karlovic, D., Turkulov, .I., and Pericin, D. Quality of corn germ oil obtained by aqueous enzymatic extraction. J. Am. Oil Chem. Six. 1993.70, 1273-1277 Marek, E., Schalinatus, E., Weigelt, E., Mieth, G., Kerns, G., and Kude, J. Interbitech ‘8%Mathematical Modelling Biotechnol. Vol. 6 (Blazej, A. and Ottova, A., Eds.). Elsevier, Amsterdam, 1989,471474 DuBois, M. W., Anglemier, A., Montgomery, M. W., and Davidson, W. D. Effet of proteolysis on the em~si~cation characteristics of bovine skeletal muscle. J. Food Sci. 1972, 37, 27-28 Kuehler, C. A. and Stine, C. M. Effect of enzymatic hydrolysis on some functional properties of whey protein. J. Food Sci. 1974,39, 379-382 P&i, G. Modification of functional properties of soy proteins by proteolytic enzyme treatment. Cereal Chem. 1975,53,650-&S Zakaria, F. and McFeeters, R. F. improvement of the emulsification properties of soy protein by limited pepsin hydrolysis. Lebensm. Win. Technol. 1978, 11,424 Smith, D. M. and Brekke, C. J. Functional properties of enzymatitally modified beef heart protein. L Food Sci. 1984, 49, 152% 1528
Enzyme Microb. Technol.,
94.
Hamada. .I. S. and Marshall, W. E. Preparation and functional properties of enzymatically deamidated soy proteins. J, Scrod Sci. 1989.54,598-X%% Marlowe, I. T., Giddings, T. J., Richardson, S. J., Stentiford, A. U.K. industry and ozone pollution from volatile organic compound emissions. II. Update to October 1991. Warres Spring Laboratory-The Environmental Technology Executive Agency of the Department of Trade and Industry, London, 1991, Report 878 (PA) Finlayson-Pitts, B. J. and Pitts, Jr., J. N. Volatile organic compounds: Ozone formation, alternative fuels, and toxics. Chem. Ind. (London) 1993, 20,796800 Valentin, F. H. H. Volatile organic compounds and odors. Environ. Health 1992, Aug., 224-227 Lawhon, J. T. Alternative processes for extracting soybean oil. Food Res. Dev. Center Annu. Rep. Texas Engineering Experiment Station, Texas, 1984 Erickson, D. R. Soybean oil: Update on number one. J. Am. Oil Chem. Sot. 1983, 60, 303A-308A Hagenmaier, R. D., Carter, C. M., and Mattil, K. F. Aqueous processing of coconuts: Economic analysis. J. Am. Oil Chem. Sot. 1975, 52,s9 Sosulski, F. W. and McCleary, C. W. Diffusion extraction of chlorogenic acid from sunflower kernels. J. Food Sci. 1972, 37, 253256 Sosulski, F. W., Soliman, F. S., and Bhatty, R. S. Diffusion extraction of glucosinates from rapeseed. Can. Inst. Food Sci. Technol. J, 1972.5, IOl-104 Rhee, K. C., Natarajan, K. R., Cater, C. M., and Mattil, K. F. Processing edible peanut protein concentrates and isolates to inactivate aflatoxins. J. Am. Oil Chem. Sot. 1977, 54, 245A-249A Kim, I.-H. and Yoon, S. H. Effect of extraction solvents on oxidative stability of crude soybean oil. J. Am. Oil Cbem. Sot. 1990, 67, 165-167
95.
96.
97. 98.
99. loo.
101.
102.
103.
104.
1996, vol. 19, November
1