6. AOACSPIFANMethods-2018Awards

37

148 L in et al .: J ournal of AOAC I nternational V ol . 100, N o . 1, 2017

(d)  Calculation of working standards concentration level 1–6 (WS1–6): WS1–6 = (volume of WS7 used in mL) × (concentration of WS7 in ng/mL) ÷ (WS1–6 volume in mL) (e)  Calculation of the biotin concentration in the injected product samples (C i ) is from its biotin peak area and the standard curve generated from the standards. (f)  Calculation of original product concentration (C p ) is, therefore, based on the dilution scheme used for the sample: = the injected sample’s biotin concentration, from standard curve (ng/mL); D 0 = the dilution of original product before filtration: D 0 = 50 mL; ss = the sample size (g); D 1 = the unit conversion (from per g to per 100 g sample and from ng to μg of biotin): D 1  = (100 g/100 g)(1 μg/1000 ng) = 1/10 g·μg/ng/100 g; and D 2 = the conversion from sample to reconstituted finished product: for powder, D 2 = 25 g powder ÷ (25 g powder + 200 g water) = 25 g ÷ 225 g = 1/9; for liquid (2× dilution by weight), D 2 = 2; and for ready-to-feed (RTF; as is), D 2 = 1. C p = C i × D 0 ÷ ss × D 1 × D 2 where C p = the original product concentration (μg/100g); C i This method has undergone a thorough single-laboratory validation (SLV) usingAOAC INTERNATIONAL guidelines to probe its linearity, LOQ, specificity, accuracy, and ruggedness. The analytical range for SPIFAN biotin-fortified matrixes was found to be between 1.7 and 142 μg/g reconstituted final product or RTF. Calibration fit . — During each analytical run, seven standards with biotin concentrations ranging from 5 to 100 ng/mL were injected before and after each sample set. Calibration curves were constructed from these standards using a polynomial regression curve (cubic-fit) and used to back-calculate the concentration of each working standard in order to calculate calibration error at each level. The method demonstrated good polynomial regression (cubic) fit, over a standard range of 5–100 ng/mL biotin, with r 2  >0.9990. The calibration errors for the lowest two levels (near the LOQ level) are around 25% and 10%; the calibration errors for the remaining levels were <8% (Table 1; Figure 2). Suitable calibration curve range . — Due to the characteristics of the postcolumn protein binding reaction, saturated Results and Discussion Method Validation

switch valve position starts at 1 → 2. At 5 min, after biotin is eluted and detected by the FD, the valve position is switched to 1 → 6. The flow rate then ramps to 1.5 mL/min over 1 min and keeps until 12 min, when riboflavin is eluted and detected by the FD. Lastly, the flow rate decreases to 0.4 mL/min over 1 min before the valve position is switched back to 1 → 2. The flow rate keeps at 0.4 mL/min until 15 min. ( 6 )  Postcolumn pump flow rate .—0.2 mL/min. (b)  Instrument configuration with an optional column switch .—The system should be configured as shown in Figure  2016.11A . (c)  System pressure .—Column pump head pressure maximum at 600 bar or per column manufacturer’s instructions; postcolumn reaction coil head pressure maximum at 40 bar or per manufacturer’s instructions. (d)  System equilibration .—( 1 ) Turn on the FD at least 1 h before start of analysis. ( 2 ) Inject the most concentrated standard (approximately 100 ng/mL) onto the column and observe the response on the FD. If necessary, adjust the detector gain and sensitivity settings so that the standard curve is within the range of the detector. After the detector settings have been determined, inject the most concentrated standard three to four times and note the peak areas. If the system is equilibrated, the RSD of the standard peak areas should be <2%, and the peak areas should not steadily increase or decrease by more than 4% from the first injection to the third or fourth injection. If the RSD is >2%, then locate the source of the imprecision and correct it before beginning the sample analysis. If peak areas steadily increase or decrease by more than 4%, the system is not equilibrated and must be allowed to equilibrate longer. After the system has reached equilibrium and the RSD is ≤2%, inject a set of standards, unknown samples, and another set of standards. Every set of unknown samples must be bracketed by standards. (e)  Column and system maintenance .—The column, postcolumn reaction coil, and system may be cleaned by using 50% methanol at an appropriate pressure, referring to F(c) . Quantification is obtained by using a seven-level external standard consisting of the following concentrations: 5, 10, 20, 40, 60, 80, and 100 ng/mL. The calibration of the standards is determined by using a polynomial regression curve (cubic-fit). (a)  Calculation of stock standard concentration: SS = S w × P ÷ 500 × D 1 where SS = the stock standard concentration (μg/mL); S w = the standard weight (g); P = the purity of the standard (g/g); 500 = the volume of the stock solution (mL); and D 1 = the unit conversion factor: D 1 = 1000000 μg/g. (b)  Calculation of intermediate standard (IS) concentration: IS = (volume of SS used) × (SS concentration) ÷ dilution volume (c)  Calculation of working standards concentration level 7 (WS7): WS7 = (volume of IS used in mL) × (concentration of IS in µ g/mL) ÷ (WS7 volume in mL) × 1000 ng/ µ g G. Calculations

Table 1. Representative calibration standard data

Concentration, ng/mL

HPLC peak area

Level

Amount, ng/mL Error, %

1 2 3 4 5 6 7

5

0.2019 0.6600 2.2932 6.1743 9.2305 11.353 12.863

6.21 9.00

24 10

10 20 40 60 80

18.61 41.03 60.45 77.66 103.9

7 3 1 3 4

100