Degumming revisited

July 2010

Note: The following article is based on the address given by Albert J. Dijkstra, the 2010 Timothy L. Mounts Award winner. His address was given at the 101st AOCS Annual Meeting & Expo, held in Phoenix, Arizona, USA, May 16-19.

In 1992, during the AOCS World Conference in Budapest (Dijkstra, 1993), I launched the enzymatic degumming process EnzyMax^® on behalf of Lurgi, the company that had developed this process. I reported that I had measured low residual phosphorus and iron contents in the enzymatically degummed oil samples Lurgi had given me. So I could not but conclude that the oil would yield good quality oil on physical refining and that the EnzyMax process would therefore compete with my own acid refining process TOP (Dijkstra and Van Opstal, 1989). However, by repeating in this presentation what Lurgi had told me, I also misled the audience by stating that the citrate buffer was added to the oil as such, whereas in actual practice, a concentrated citric acid solution was added first and a dilute caustic soda solution was added quite some time later. In 1998, when presenting a paper on degumming at the AOCS Annual Meeting & Expo in Chicago, Illinois, I rectified this statement by describing the EnzyMax process as an "acid refining process in disguise."

Since that time, the porcine phospholipase (PLA2) used in the original EnzyMax process has been replaced by microbial enzymes such as the phospholipase A1 (PLA1) enzymes Lecitase^® Novo and Ultra (Novozymes, Bagsværd, Denmark) and the PLA2 enzyme Rohalase^® MPL (AB Enzymes, Darmstadt, Germany), and additional microbial enzymes have been developed. One of these is Purifine^®, a phospholipase C (PLC; Verenium Corp., Cambridge, Massachusetts, USA). As indicated in Figure 1, it catalyzes the hydrolysis of the phosphate/glyceryl bond in phosphatidylcholine (PC) and phosphatidylethanolamine (PE), the major phosphatides in most vegetable oils. This hydrolysis yields diglycerides and phosphate esters of choline and ethanolamine, respectively. These diglycerides are not removed during subsequent refining processes, and since the phosphate esters do not retain any triglyceride oil, the use of PLC as a degumming agent can increase the oil yield, especially for high-phosphatide oils such as soybean oil that has not yet been water-degummed. For these oils, the yield increase amounts to 1% of oil for every 500 ppm of phosphorus present in the crude oil. Purifine PLC does not catalyze the hydrolysis of other phosphatides than PC and PE, so it is not capable of providing an oil that is so low in residual phosphorus and iron that it can be physically refined.

Another enzyme that is used in degumming is LysoMax^®, a lipid acyl transferase with PLA2 activity (LAT; Danisco AS, Copenhagen, Denmark). It can abstract a fatty acid moiety from a phosphatide and transfer this to a free sterol or stanol present in the oil and convert this into a fatty acid ester. If no free sterols are present, the enzyme causes free fatty acids to be formed. Its specificity is not limited to PC and PE so that, given time, all phosphatides present in the aqueous enzyme solution end up as lysophosphatides just as when the PLA1 enzyme is used. These lysophosphatides retain less oil than the original phosphatides, so their use in degumming also leads to an oil yield improvement: Hydrolysis reduces the dry gum weight by about 35% and on a dry weight basis, the lysophosphatide gums retain only some 15 weight % of oil as opposed to the non-hydrolyzed gums, which retain 30-35 weight % of oil. In the case of the LAT enzyme, this yield is even further increased because the fatty acids that are attached to sterols are retained and the free sterol loss during deodorization is reduced.

When comparing the porcine PLA2 with the new microbial PLA1 (Lecitase^® Novo), Clausen (2001) described the standard Novozymes laboratory degumming procedure that  forms the basis of all their presentations, publications, and patent applications. It comprises a 30-minute pre-treatment of the oil with dilute (2.2% w/w) aqueous citric acid, followed by the addition of the amount of caustic soda necessary to reach the pH aimed for. Then the enzyme is added and the experiment is continued for 6 hours. During the entire experiment the oil (0.6 L) is circulated over the reaction vessel at a rate of 1.1 L per minute by a Silverson in-line mixer. This mixer maintains a certain degree of dispersion of the aqueous phase in the oil and thus permits the reactions taking place at the oil/water interface to continue. Consequently, these laboratory experiments show a slow but steady decrease in residual phosphorus content of the oil.

In this respect, the laboratory conditions differ fundamentally from what can be realized on an industrial scale. There, a reaction vessel can be filled with a fine dispersion by using a suitable high-shear in-line mixer, but once this dispersion has entered the vessel, the agitator in this vessel cannot maintain this dispersion. It coalesces and the rate of reaction at the interface decreases. For an acid refining process such as TOP, this coalescence does not matter since the reaction aimed for, the decomposition of the non-hydratable phosphatides (NHP), is very fast (less than 1 minute), but for the much slower enzyme-catalyzed NHP decomposition, this coalescence is fatal: It stops.

So, contrary to what is commonly taken for granted, industrial enzymatic degumming processes using PLA or LAT and aiming at complete NHP removal achieve this by treating the oil with a strong solution of citric acid that is subsequently neutralized to arrive at the right pH for the enzyme. They are in fact "acid refining processes in disguise." All the enzyme does is hydrolyze the phosphatides that are already in the aqueous phase so that they retain less oil. This raises the question: Would it not be easier to treat the gums rather than the oil with one or more enzymes and effectuate the gain in oil yield in a separate and much smaller unit? This treatment can be a low-temperature enzymatic or chemical treatment (Kellens and De Greyt, 2010), or it can be a high-temperature hydrolysis that does not require any catalyst (Naudet et al., 1954).

What degumming process should be used where? An oil mill selling partially degummed oil (P < 200 ppm) could do a number of things. It could continue to degum its crude oil with water and sell water-degummed oil. The gums could then be mixed with the meal or treated in such a way that the oil yield improves or the fatty acid moieties are recouped. It could also treat its crude oil with an aqueous solution of PLC at a temperature at which this enzyme does not yet lose its activity, raise the temperature to reduce the oil viscosity, and isolate the degummed oil in a centrifugal separator. Because these processes are carried out at the same stage, they have been put in the same box in Figure 2.

Figure 2 also shows a box that refers to a membrane degumming process as applied to the miscella resulting from the extraction process. Such processes have been disclosed in several patents and have been the subject of industrial trials, but since these suffered from serious membrane fouling, the membrane degumming process is not being used industrially.

The dry degumming process has been represented on the left side of Figure 2. The use of this process permits a two-step refining process consisting of the dry degumming step, which also serves to bleach the oil, and the physical refining step. The dry degumming process can be profitably applied to low-phosphatide oils such as palm oil, lauric oils such as coconut oil, palm kernel oil, and animal fats such as edible tallow and lard. For high-phosphatide oils such as most vegetable oils, the amount of bleaching earth required would be so high as to make the process uneconomic. As I mentioned during my presentation, I think that the dry degumming process could profit from further investigation and development. One possibility might involve the partial neutralization of the degumming acid prior to the bleaching earth addition such as disclosed by Nock (1995) for oils containing more phosphatides than the oils now being dry-degummed.

Figure 2 also includes the acid degumming process, but as mentioned before (Dijkstra, 1993), this process does not ensure sufficiently complete removal of phosphatides so it has to be followed by an alkali refining process or a dry degumming process. Accordingly, the acid degumming process can be considered as obsolete.

The Complete (Deffense, 1999) and SOFT^® (Choukri et al., 2001) degumming processes use EDTA as complexing agent for the calcium and magnesium ions that form part of the NHP. Figure 2 shows it as a further treatment of water-degummed oil, but in principle, it is also highly effective for other oil grades. Since EDTA is quite expensive, its use may be most appropriate as a kind of polishing operation that requires only little reagent. One such polishing operation would be the removal of pro-oxidant iron from acid oils that have become contaminated during transport and/or storage.

The largest box in Figure 2 refers to the acid refining process, either on its own or preceding an enzymatic treatment of the gums originating from this acid refining process. In the acid refining process, the NHP are decomposed by a degumming acid that is finely dispersed through the oil, and the resulting phosphatidic acid is then made hydratable by converting it to its alkali salt by the addition of caustic. There are various acid refining process variants on the market. TOP is now offered by Westfalia, and several other suppliers offer their own type of acid refining process. It has been proven in practice on a large variety of oils over a number of years, and I continue to regard it as the process of the future.

Whether to treat the acid refined oil with a PLA1 or LAT enzyme or treat the gums with these enzymes depends on local circumstances. If a plant with large holding vessels is already present, it does not cost much to treat the oil with enzymes and thereby increase the oil yield. Installing such large vessels may well be unduly costly, so in that case, treating the gums may well be preferable. This treatment can be enzymatic or chemical depending on what the refiner wants to achieve. If the processor wants to recuperate the fatty acids present in the gums, a chemical treatment is called for. If the plant manager has an outlet for oily material (a mixture of partial glycerides and triglycerides) and can dispose of residual organics, the enzymatic treatment offers opportunities, but then enzyme cost and patents have also to be taken into account.

Albert J. Dijkstra, consultant, France, can be reached via email at albert@dijkstra-tucker.be .

information

For further reading:

Choukri A., M.A. Kinany, V. Gibon, A. Tirtiaux, and S. Jamil, Improved oil treatment conditions for soft degumming. J. Am. Oil Chem. Soc. 78:1157-1160 (2001).

Clausen, K., Enzymatic oil-degumming by a novel microbial phospholipase, Eur. J. Lipid Sci. Technol. 103:333-340 (2001).

Deffense, E.M.J., Method for eliminating metals from fatty substances and gums associated wth said metals, PCT Patent Application Publication WO 99/02630 (1999).

Dijkstra, A.J., Degumming, refining, washing and drying fats and oils, in Proceedings of the World Conference on Oilseed Technology and Utilization, edited by T.H. Applewhite, American Oil Chemists' Society, Champaign, Illinois, 1993, pp. 138-151.

Dijkstra, A.J., and M. Van Opstal, The total degumming process, J. Am. Oil Chem. Soc. 66:1002-1009 (1989).

Kellens, M.J., and W.F.J. De Greyt, Oil recuperation process, US Patent 7,696,369 (2010).

Naudet, M., E. Sambuc, and P. Desnuelle, Recuperation des acides gras contenus dans les mucilages. Bull. Mens. ITERG:112-115 (1954).

Nock, A., Process for refining glyceride oil, PCT Patent Application Publication WO 95/30727 (1995).