No abstract

Fortified milks were saponified overnight at room temperature with 1% ethanolic pyrogallol and KOH. The digest was extracted with hexane after adding water and ethanol, and the extract was washed consecutively with 5% KOH, water, and 55% aqueous ethanol to remove polar lipids. After evaporation, the residue was first chromatographed on a column of 5 μm silica. A fraction containing vitamin D was collected, evaporated, and rechromatographed on a reverse phase column for the separation and quantitation of vitamins D2 and D3. Recovery was 96-99% and the coefficient of variation was 3% (8 replicates). Infant formula was diluted and then saponified and extracted as in the analysis of milk. Margarine was saponified by shaking overnight with 1% ethanolic pyrogallol and 80% KOH. Water and ethanol were added to the digest before extraction. Extracts from formula and margarine were chromatographed as milk except, before HPLC, the extract was dissolved in isopropanol-hexane (1 + 99) and passed through 5 cm alumina in a Pasteur pipet, and the concentration of isopropanol in the first high performance liquid chromatographic (HPLC) solvent system was halved to improve the separation of vitamin D from other absorbing lipids. Usually several peaks were obtained during the final HPLC analysis, and the identification of vitamins D2 and D3 was less certain than in the analysis of milk. The coefficients of variation for formula and margarine were 6% (5 replicates) and 9% (6 replicates), respectively.

Retinol was determined in margarine (50 mg), infant formula (1 ml), and fortified milk (1 ml) by saponification in centrifuge tubes, extraction of the unsaponifiable lipid with hexane, and high pressure liquid chromatography (HPLC) in 90% methanol on a 25 cm × 3.2 mm column containing 10 μm LiChrosorb reverse phase. β-Carotene was determined using the same column and 99% methanol as eluant. The vitamins in the eluate were identified and measured from their absorption at 325 (retinol) and 453 nm (β-carotene), using a computing integrator. Retinol and carotene were prominent peaks in the chromatograms from properly fortified samples and were satisfactorily separated from other materials. In a survey of 12 different margarines for retinol, the results from liquid chromatography agreed with those of a variety of Spectrophotometric and fluorometric procedures but they were obtained with greater ease and they could be interpreted more confidently. The recovery of retinol from oil was better than 99%. The coefficients of variation were 6.8% for 12 replicate analyses of a margarine for retinol and 5.4% for 10 replicate analyses of a margarine for β-carotene.

Saponification and subsequent evaporation of extracts can cause losses of vitamin A during analysis and thus cause erratic results. These steps were therefore avoided by measuring retinyl palmitate and β-carotene directly in hexane extracts of milk and margarine. Extracts were purified before high performance liquid chromatography(HPLC) by washing with aqueous alcohol. Retinyl palmitate was measured at 325 nm after chromatography on LiChrosorb Si60, 5 μm, using ethyl ether–hexane (2+98) as the solvent system. Milk (2 mL) was shaken with absolute ethanol (5 mL) and hexane (5 mL). Water (3 mL) was added and, after mixing and centrifugation, 100 μL hexane layer was injected. Margarine (5 g) was dissolved in hexane (100 mL). After 5 mL of the hexane solution was washed with 60% ethanol and centrifuged, a 50 μL aliquot was injected. The results of the retinyl palmitate estimation in milk agreed with those obtained by a fluorometric method; the results in margarine agreed with those obtained by saponification, extraction, and HPLC. The coefficients of variation on analysis of 10 replicate samples of milk and margarine were 3 and 4.5%, respectively. The analysis of margarine for vitamin A was completed by measuring β-carotene with a detector set at 453 nm; the coefficient of variation of this measurement was 3% (10 replicates).

No abstract