5. AOACSPDSMethods-2018AwardsV3

7

K line et al .: J ournal of AOAC I nternational V ol . 100, N o . 3, 2017  5

and placebo of liquid concentrate product. When 20 ppb aloins were spiked into the powder placebo or into the placebo of liquid concentrate product, aloin peaks were clearly observed (matching the RTs of the aloin peaks in the standard injection). For aloe-emodin, the specificity study results clearly showed that there was no significant interference from the placebos of either the powder product or liquid concentrate product (Figure 3). When 20 ppb aloe-emodin were spiked into the powder placebo or into the placebo of the liquid concentrate product, aloe-emodin peaks were clearly observed (matching the RTs of the aloe-emodin peaks in the standard injection). Six replicate samples from each of the four products were prepared according to the previously described test method. Samples were analyzed against a freshly prepared standard solution. The amounts of aloin A and B and aloe-emodin recovered from each sample were then calculated. Tables 1–4 show that the RSDs of six test results for aloin A and B and aloe-emodin for each of the four products were less than 10%. Liquid and powder placebo samples were spiked in triplicate with 10, 20, and 30 ppb spiking solutions of aloins A and B and aloe-emodin at the 50, 100, and 150% level. The spiked samples (three concentrations and three replicates of each concentration) were analyzed according to the internal test method. The amount of aloins A and B and aloe-emodin in the spiked samples were calculated as the percentage recovery. Tables 5–10 show that average recoveries for the spiked samples were 82.9–100.1% for aloin A, 85.5–89.5% for aloin B, and 94.2–109.1% for aloe-emodin, which were all within the acceptable limit range of 80–120%. Accuracy Precision

Linearity/Range

A standard solution containing 1 ppm of aloins A and B and aloe-emodin was prepared. Dilutions from the 1 ppm standard were made to obtain standard solutions containing 10, 20, 40, 80, 160, and 500 ppb of aloinsAand B and aloe-emodin ( see Table 11). Three replicate injections were made for each of the six solutions prepared above. The peak areas for aloins A and B and aloe-emodin that were obtained for each solution were plotted against their corresponding concentrations. Linear regression analyses on the six coordinates were performed. Tables 12–17 and Figures 4–6 show that the linearity of detector response for aloins A and B and aloe-emodin in the range of 10–500 ppb yielded linear correlation coefficients (R) of 0.9999, which were within the acceptable limit of >0.998. The same four products were analyzed (in duplicate) by a second analyst on a different day, using a different HPLC system and a different Phenomenex Synergi Hydro-RP HPLC column. Results were compared with the average results from the precision test for aloins A and B and aloe-emodin. Table 18 shows that there was a difference of <10% in the test results obtained by the two analysts for aloins A and B. The difference in the test results for aloe-emodin in aloe vera 5×, aloe concentrate, and aloe powder was <10% and in the aloe vera gel 200×, the difference was slightly higher, at 15.7%. The obtained 15.7% difference was considered justifiable given the method accuracy requirement was 80– 120%. Therefore, 15.7% was within the method accuracy requirement of 20% variability from 100%. Ruggedness

System Suitability

System suitability parameters for working standards of aloins A and B and aloe-emodin were calculated using Waters

Figure 2. Aloin region overlay HPLC chromatograms of (a) 20 ppb standard mixture, (b) solvent blank, (c) placebo of the powder product, (d) placebo of the liquid concentrate product, (e) 20 ppb standard mixture-spiked placebo of the powder product, and (f) 20 ppb standard-spiked placebo of the liquid concentrate product.