Determination of Soluble Vitamins in Beverages

Fat-Soluble Vitamin Standards: The fat-soluble vitamins were prepared in acetonitrile to yield a one mg/mL solution of DL-α-tocopherol acetate and a 0.2 mg/mL solution of retinol palmitate. Retinol palmitate requires several minutes of vortexing to dissolve. These solutions were stored in the dark at 4°C . Calibration standards of retinol palmitate and DL-α-tocopherol acetate were not prepared in a matrix of 0.015% formic acid. Instead, due to solubility limitations, working standards of these vitamins were prepared in mobile phase B from stocks prepared in acetonitrile, and a separate standard curve was prepared for the fat-soluble vitamins.

click for large version

Figure 2. Separation of Brand C, a Fruit-Flavored, Sugar-Sweetened, Vitamin-Enhanced Water, on the Acclaim PolarAdvantage II Column.

Nine Vitamins Separated

Figure 1 (see p. 34) shows the separation of nine vitamins with the proposed method. Sodium citrate and citric acid, common ingredients that enhance tartness, were added to the mixture to demonstrate the separation between nicotinamide and the citrate peak. As shown in Figure 1, citrate and the vitamins are well separated from each other.

Vitamins are a structurally diverse group of compounds with different absorbance maxima. For example, retinol palmitate has a strong absorption at 350 nm, while D-pantothenic acid absorbs in the UV between 200 nm and 225 nm. Given the wide UV absorbance range for the vitamins commonly added to enhanced waters, wavelengths of 210 nm, 280 nm, and 350 nm were used for detection. In Figure 1, the chromatogram collected at 210 nm captures each of the vitamins except for vitamin A.

The linearity, limit of detection, limit of quantitation, and precision data for this gradient method were determined for nine vitamins: pyridoxine HCl (B₆), ascorbic acid (vitamin C), nicotinic acid (B₃), nicotinamide (B₃), pantothenic acid (B₅), cyanocobalamine (B₁₂), folic acid, tocopherol acetate (vitamin E), and retinol palmitate (vitamin A).

Correlation coefficients ranged from 0.9985 to 0.9997. Retention time precisions (relative standard deviations or RSDs) ranged from 0.07% to 0.23% with peak area precisions (RSDs) ranging from 0.28% to 0.97%, with the exception of ascorbic acid, which had a peak area precision of 3.47%. Ascorbic acid exists in solution in equilibrium with the oxidation product dehydroascorbic acid (DHAA). While both compounds are biologically active as vitamin C, DHAA does not strongly absorb in the UV and is therefore difficult to quantify. The oxidation reaction is reversible and is minimized at low pH or by adding a reducing agent, such as dithiotheitol, to prevent oxidation of ascorbic acid by dissolved oxygen.⁶

click for large version

Table 1. Analysis of Three Brands of Vitamin Enhanced Waters

Functional Beverage Samples

Three brands of vitamin-enhanced water were analyzed over three days to evaluate the precision of the method. Representative data are presented in Figure 2 (see above) with a summary in Table 1 (see left). Intra-day retention time precision ranged between 0.04% and 0.23%, which is equivalent to the precision of the standards. Inter-day peak area precision ranged from 0.37% to 9.5%. The increased imprecision observed in the folic acid results is due not only to the presence of a closely eluting peak in the brand B sample but also to the low concentration of folic acid in the sample. Even though this is a challenging matrix and there is a low concentration of folic acid, it is still easily quantified.

Recoveries for the water-soluble vitamins ranged from 93% to 119%. As with precision, the extremes in recoveries were for ascorbic acid and folic acid. Despite the challenges in quantifying these two vitamins, the recoveries were good, showing that the method is accurate.

Direct spiking of the acetonitrile stocks of vitamins A and E into vitamin-enhanced waters led to the formation of an unstable suspension; therefore, a control sample was used to indirectly measure recovery. Recoveries for fat-soluble vitamins were determined by comparison of spiked samples to a control sample of retinol palmitate and DL-α-tocopherol acetate in 0.015% formic acid, leading to recovery values ranging from 101% to 110%.