1036 J.T. CARLIN ET AL. 20. Ohloff, G., and I. Flament, Forschr. Chem. Org. Naturst. 36:231 (1979). 21. Maga, J.A., in Fenaroli's Handbook of Flavor Ingredients, 2nd ed., Vol. 1, edited by T.E. Furia and N. Bellanca, CRC Press, Inc., Cleveland, Ohio, 1975, p. 228. 22. van der Linde, L.M., J.M. van Dort, P. de Valois, H. Boelens and D. de Rijke, in Progress in Flavor Research, edited by D.G. Land and H.E. Nursten, Applied Science Publishing, Ltd., London, England, 1979, p. 219. 23. Kazeniac, S.J., and R.M. Hall, J. Food Sci. 35:519 (1970). [Received October 28, 1985] 9 %Heterogeneous Catalytic Hydrogenation of Canola Oil Using Palladium N. Hsu, L.L. Diosady*, W.F. Graydon and L.J. Rubin Department of Chemical Engineering, University of Toronto, Toronto, Ontario, Canada M5S 1A4 The hydrogenation of canola oil was studied using palladium black as a potential catalyst for producing partially hydrogenated fats with low trans-isomer content. Pressure (150-750 psig) appeared to have the largest effect on trans-isomer formation. At 750 psig, 90 C and 560 ppm metal concentration, a maximum of 18.7% trans isomers was obtained at IV 53. A nickel catalyst produces about 50% trans isomers at the same IV. For palladium black, the linolenate and linoleate selectivities were 1.2 and 2.7, respectively. The max- imum level of trans isomers observed ranged from 18.7 % to 42.8% (150 psig). Temperature (30-90 C) and catalyst concentration (80-560 ppm) affected the reaction rate with little effect on trans-isomer formation and selectivities. At 250 psig and 50 C, supported palladium (5% Pd/C) appeared to be twice as active as palladium black. At 560 ppm Pd, 5% Pd/C produced 30.2% trans (IV 67.5), versus 19.0% trans for palladium black (IV 68.9). Respective linoleate selectivities were 15 and 6.6, while linolenate selectivities were approximately unity. Analysis of the oil samples by neutron activation showed approximately a 1 ppm, Pd residue after filtration. The search is still on for an active, heterogeneous catalyst for hydrogenation of edible oils with production of a low level of trans isomers. The need for such a catalyst arose following a Canadian study which reviewed the health effects of these isomers (1). This subject was reviewed thoroughly by Applewhite (2). In his review it was shown that the studies of the health effects of trans isomers were inconclusive. The FDA recently commis- sioned the Federation of American Societies for Experimental Biology to undertake a study of trans fatty acids (3) in the hope of settling this controversial issue. Thus, we feel it would be prudent to develop alternative catalysts which would minimize the forma- tion of trans fatty acids. Such a development would, in any case, offer the processor an alternative, should one be needed, for this or any other reason. In an earlier review of catalysts (4), it was reported that heterogeneous palladium catalysts were unsuitable for the hydrogenation of triglycerides because they were nonselective and produced large quantities of trans *To whom correspondence should be addressed. acids. More recent work indicates that, in general, palladium forms more trans isomers than nickel (5}, especially under conditions normally employed with the latter (6). Thus, we initially investigated the homoge- neous catalytic hydrogenation of canola oil using a palladium complex. It had been reported by Itatani and Bailar (7) that the mixture of dichlorobistriphenylphos- phine palladium (I I) and stannous (I I) chloride dihydrate was a very active homogeneous catalyst for the hydrogenation of soybean oil methyl esters. Further- more, at 575 psi and 60 C, a total trans content of less than 20% was observed. We used this catalyst mixture to hydrogenate canola oil at 500 psig and 110 C. After 5 hr of reaction time the IV dropped 12 units. A black precipitate, most likely palladium, was present in the partially hydro- genated oil. Itatani and Bailar (7) reported that some of their runs produced aprecipitate. In another report (8), the same authors observed a black precipitate following the hydrogenation of soybean oil methyl ester using the platinum analog of the palladium complex. The precipitate, assumed to be platinum black, was then used in a run, but it proved to be relatively inactive. We observed that our precipitate, when used alone in the hydrogenation of canola oil, was very active at 750 psig and 70 C. The homogeneous palladium catalyst subsequently was abandoned, and palladium black was made the catalyst of choice. The objective of our study was, therefore, to evaluate the performance of palladium with respect to pressure, temperature, concentration and catalyst support during the hydrogenation of canola oil. The earliest published work on the hydrogenation of any edible oil in which palladium was used was in 1953; in it, the promoter effect of platinum and palladium on nickel was examined (9). It was found that the addition of platinum increased the bonding strength of hydrogen to the catalyst surface, whereas addition of palladium decreased the bonding strength. Zajcew (10) hydro- genated castor oil to castor wax using 5% and 1% palladium on carbon. Although palladium on carbon was used by Zajcew in subsequent reports (11,12), it was for the hydrogenation of soybean and cottonseed oil. In almost all cases, hydrogenation was accompanied by 20-50% trans-isomer formation. The lowest trans content reported was 15.0% at IV 67.1 (11). The catalyst used was palladium (200 ppm) in the form of 1% JAOCS, Vol. 63, no. 8 (August 1986)