5. AOACSPDSMethods-2018AwardsV3

16

O fitserova & N erkar : J ournal of AOAC I nternational V ol . 99, N o . 6, 2016  1473

A 100 psi back-pressure regulator was installed on the detector outlet line to prevent any liquid from boiling in the postcolumn reactor.

Table 2016.10B. HPLC pump gradient conditions Time, min

Li275, % Li750, % RG003, %

0

100 100

0 0

0 0 0

12 45

G. HPLC Conditions

66

34

45.1

0 0

0 0 0 0

100 100

The equipment was connected in the following order: HPLC pump, autosampler, guard column, analytical column, postcolumn derivatization system, and UV-Vis detector. A Lithium cation- exchange column and lithium-based buffer solutions were used to separate l -theanine. The HPLC pump flow rate was set to 0.35 mL/min and the postcolumn reagent pump flow rate was set to 0.3 mL/min. The column oven temperature was set to 37°C and the postcolumn reactor temperature was set to 130°C. The HPLC pump gradient conditions that were used for the analysis are listed in Table 2016.10B . The UV-Vis detector signal was monitored at 570 nm with the reference wavelength set at 630 nm. An injection volume of 10 μL was used. Before starting the analysis, the system was equilibrated for at least 30 min until all temperatures and pressures were stable. At least one reagent blank was injected to equilibrate the column before injecting the working calibration solutions, control samples, samples extracts, and reagent blank. A mid- range calibration solution was run every 10 injections to confirm the stability of the calibration curve. (a)  Retention times for l -theanine in sample extracts and calibration solutions were within 0.5 min. (b)  Retention times for l -norleucine in sample extracts and calibration solutions were within 0.5 min. (c)  The correlation coefficient R 2 for the weighted linear regression calibration curve was ≥0.9998. (d)  Relative error for the back-calculated concentration for the mid-range calibration standard was within ±4%. H. System Suitability The response ratio for the calibration standards (Area L- theanine /Area IS ) vs its corresponding ratio for the concentrations (Concn L-theanine /Concn IS ) was plotted to obtain a weighted linear regression calibration curve. The concentration of l -theanine (μg/mL) in the sample extracts was calculated by interpolating the calibration curve. The amount of l -theanine in the sample was calculated by using the following formula: I. Calculations

50

50.1

100 100

0 0

62

reagents and different calibration curves. In addition, samples SRM 3254, SRM 3255, and SRM 3256 were analyzed using two different HPLC systems. The Grubbs’ outlier test for a 95% confidence interval was applied to the results with no outliers detected.

K. Accuracy Testing

Method accuracy was evaluated by analyzing SRM 3254, SRM 3255, and SRM 3256, as well as by conducting spike recovery studies for seven matrixes. SRMs were analyzed in triplicate over 4 days by two different analysts using two different HPLC systems and different lots of reagents and columns. For spike recovery studies, each matrix was spiked at two levels and samples analyzed in duplicate over 3 days by two different analysts using different lots of reagents. l -Theanine stock solution and l -theanine intermediate stock solution were used to spike the samples. The overall mean for unspiked samples determined during the course of the precision study was used to calculate the recoveries.

L. Ruggedness Testing

The effect of seven factors ( see Table 2016.10C for a list) was evaluated using the Youden ruggedness trial design (18). In each experiment, the values of four factors were modified as shown in

Table 2016.10C. Ruggedness trial experimental design

Factor

Value 1

Value 2

Formulation of ninhydrin reagent

T100, 1-part ninhydrin reagent (A)

T200, 2-part ninhydrin reagent (a)

HPLC flow rate Extraction volume

0.35 mL/min (B)

0.38 mL/min (b)

25 mL (C)

10 mL (c)

Analyst

Analyst 1 (D)

Analyst 2 (d)

  ×

µ g mL

( )

2 h (E)

1.5 h (e) Lot 2 (f)

Extraction time

Concn extract

Volume extract mL



(

)

=

Concn sample mg g

Extraction solution Li220 Reactor temperature

Lot 1 (F)

( )

× Mass sample g 1000

130 °C (G)

125 °C (g)

Experiment No.

Combination of factors

J. Precision Testing

1 2 3 4 5 6 7 8

ABCDEFG ABcDefg AbCdEfg AbcdeFG aBCdeFG aBcdEfG abCDefG abcDEFg

Each matrix was analyzed in triplicate over 4 days. Working calibration solutions were prepared on each day of the analysis. Repeatability precision was assessed by calculating s r and RSD r (%) for same-day replicates measured under the same conditions. To determine intermediate precision, the conditions of analysis were intentionally varied by performing the analysis on different days by two different analysts using different lots of