ACTA ALIMENTARIA VOL. 10 (A QUARTERLY OF THE COMMITTEE ON FOOD SCIENCE OF THE HUNGARIAN ACADEMY OF SCIENCES, 1981)

1981 / 1. sz. - KURUCZ-LUSZTIG, É.-PRÉPOSTFFY, M.-JERÁNEK, M.: Comparison of analytical methods used for the characterization of carbonyl content of fats

2 KURTJCZ-LUSZTIG et al.: ANALYSIS OF CAKBONYLS IN FATS 2 position products, the so-called secondary products are responsible for the rancid smell and taste of the oxidized fats. The most important oxidized prod­ucts are, as far as flavor and taste go, the carbonyl compounds. A series of alkanals, alkenals and alkyldienals can be derived from the various hydro­peroxides of the fatty acids. It can be assumed that the carbonyl content is correlated with the organoleptic character of the particular fat, and that the correlation is better than that observed for the peroxide number. Apparently, this assumption is straightforward, since the peroxide number indicates only the amount of tasteless and odorless peroxides, indirectly connected only with the changes of the taste. At the analysis of the secondary products, the natural flavor components and the flavor reversion materials are generally also present. Carbonyl analysis is customarily carried out along two main lines (PIEKGIOVANNI & VOLONTERIO, 1976). The first approach calls for the physico­chemical separation of the components followed by their identification by instrumental methods (gas chromatography, mass spectrometry, etc.). The second approach makes use of a suitable reagent to separate all the carbonyl compounds followed by the analysis of the individual derivatives. The methods based on the first approach generally require oil samples of large quantities. DEBRUYNE (1964), e.g. subjected 90 1 oil to steam distilla­tion to remove the volatiles. The volatile fraction thus collected was repeatedly chromatographed and the fractions were identified by infrared spectrometry. Starting with 250 1 soybean oil, CHANG and SMOTJSE (1967) could identify as many as 71 components by infrared spectroscopy and mass spectrom­etry. Since the handling of large samples posed special problems, the need for reduced sample sizes increased. SELKE and co-workers (1970) used only 100 ml oil. The volatiles were separated and analysed by mass spectrometry. In 1971, PBEVOT and co-workers (1971) could decrease the sample size to as little as 100 /Л by injecting it directly onto a gas Chromatograph. A 5-cm long precolumn packed with uncoated support served as a guard column before the analytical column. Apart from the small sample size, the method had another advantage: there were no volatile losses since the original oil sample was directly injected onto the column. DUPTJY and co-workers (1976) described a similar system. 500 mg of the oil sample tested was injected onto a precolumn packed with glass wool. They claimed that the total peak area of the chromato­gram correlated sufficiently well with the taste scores. The method was proposed for the rapid instrumental characterization of the taste of vegetable oils. BLUMENTHAL and co-workers (1976) investigated several deep-fry fats and oils. Volatiles were separated by frceze-drying in high vacuum and subsequently analysed by gas chromatography. The total peak area of the chromatogram and the taste scores correlated reasonably well. 1* Acta Alimentaria 10, 1981

Next