Search

A method was developed for the analysis of vitamins D2 and D3 in a variety of nutritional products. To extract vitamins D2 and D3 from products containing substantial amounts of fat, a saponification with alcoholic potassium hydroxide is required to release the vitamin D. Trideuterium-labeled vitamin D is added to the sample prior to saponification, and quantitation is achieved using linear regression of the ratio of peak response for 2H3-D and vitamin D. Acceptable linearity was achieved between 0.6 and 27 microg/100 g with a correlation requirement of >0.999. The method detection limit of 0.02 microg/100 g was verified by spiking placebo products carried through the saponification and extraction steps of the method. At the quantitation limit (0.12 microg/100 g), the signal was easily distinguished from the background. Vitamin D3 spike recoveries ranged from 107 to 119% at the low level and 104 to 116% at the high-level spike. Vitamin D2 recoveries were 105 to 116% and 91 to 110% for the low- and high-level spikes, respectively. SRM 1849a has a certified concentration of 11.1 +/- 1.7 microg/100 g; using this standard reference material, the range of 9.4 to 12.8 microg/100 g was met on each of the 6 days. Method repeatability, determined in 12 vitamin D3 product matrixes over 6 days, ranged from 3.9 to 48%. The adult nutrition-milk protein sample was the most notable; it failed within-day, as well as day-to-day, precision requirements. There was no attempt to optimize the sample preparation to accommodate any problem matrix.