AOAC Final Action Methods in 2017

P ACQUETTE & T HOMPSON : J OURNAL OF AOAC I NTERNATIONAL V OL . 101, N O . 2, 2018 537

Table 1. SPIFAN SMPR parameters for Methods 2011.19 and 2015.06

Element

Analytical range

LOQ

RSD r , %

Recovery, %

RSD R , %

Na

10 – 850 mg/100 g

10 mg/100 g

≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5 ≤ 5

90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110 90 – 110

≤ 10 ≤ 10 ≤ 10 ≤ 10 ≤ 10 ≤ 10 ≤ 10 ≤ 16 ≤ 10 ≤ 16 ≤ 10 ≤ 15 ≤ 15 ≤ 15

Mg

3 – 110 mg/100 g

3 mg/100 g

P K

15 – 800 mg/100 g

15 mg/100 g 10 mg/100 g 20 mg/100 g 0.01 mg/100 g 0.001 mg/100 g 0.001 mg/100 g 0.001 mg/100 g 0.001 mg/100 g

10 – 2000 mg/100 g 20 – 1280 mg/100 g 0.01 – 20 mg/100 g 0.005 – 1.0 mg/100 g 0.001 – 0.005 mg/100 g

Ca

Fe

Mn Mn Cu Cu

0.005 – 1.2 mg/100 g

Downloaded from https://academic.oup.com/jaoac/article-abstract/101/2/536/5653941 by guest on 05 May 2020

0.001 – 0.005 mg/100 g

Zn

0.1 – 18 mg/100 g

0.1 mg/100 g

Cr

20 – 1600 µg/kg

20 µg/kg 10 µg/kg 20 µg/kg

Se

10 – 500 µg/kg

Mo

20 – 1000 µg/kg

2011.14 . Thus, the accuracy of the ICP-AES and ICP-MS data can be ascertained by comparison of the mean results of the two independent studies. Other accuracy and validation data for Method 2015.06 can be found in the publication of the single- laboratory validation (SLV) data (4) and in a separate publication specifically focused on the performance of this method at very low levels (5). Note that one other comparison is also available: the reproducibility for Cr, Mo, and Se reported for the present MLT compared to that reported for the same elements by the same method in the initial MLT (3). Different samples were used in the two independent MLTs, but it is interesting to compare the reproducibility of the original MLT that required known duplicate determinations to be averaged with the present MLT that required single determinations of each sample. Table 2 lists the laboratories and equipment involved in the study. Although 17 laboratories originally signed up for the study, several dropped out for various reasons and did not submit any data; the 10 laboratories shown in the table all completed the study. There is representation from seven countries, five different model ICP-MS units, and six different kinds of microwave units. All ICP-MS instruments were equipped with modern collision/reaction cell (CRCs) that Multilaboratory Collaborative Study

are thought to be necessary for this method to avoid low-mass molecular interferences. In the first phase of the study, participating laboratories set up the method and checked the linearity, LOQ, and accuracy of their respective instruments. These same qualification tests were used in the first MLT (3), as these techniques have consistently been shown to lead to robust method transfers within Abbott Nutrition ’ s network of laboratories. Specifically, laboratories were asked to transfer the Abbott method provided (which is more detailed than the published First Action Method 2015.06 ) and translate the given operating conditions for Agilent ICP-MS units into those suitable for their system. It was mandatory to keep the prescribed analyte and internal standard (ISTD) masses, as well as the nominal standard concentrations (made from a reputable vendor ’ s master multielement standard solution), but other settings, such as ion lens voltages, collision cell gas flows, argon gas flow rates in plasma and the nebulizer, type of nebulizer, microwave settings, etc., could be varied to suit the laboratory ’ s need and to obtain the necessary qualification data. The goal was to develop robust pulse/analog detector responses that would yield highly linear calibration curves with correlation coefficients all greater than 0.9995 and all calibration residuals below 4%. Next, the laboratories were asked to make up independent standard solutions and to run them as samples over 3 days. These “ samples ” were spaced at 10%, 40%, 800%, and 2400% of the

Table 2. Participating laboratories and equipment for Method 2015.06 MLT

Lab No.

Lab name

ICP-MS model

Microwave model

Country

4 5 6 9

Laboratory Aquanal

Agilent 7800

CEM MARS 6 CEM MARS 5 CEM MARS 6 CEM MARS 5

France

Syngene International, Ltd

Agilent 7700x

India

Covance Laboratories

Thermo iCAP Q

United States

Abbott Nutrition Qlip, Netherland

Agilent 7700x

Singapore

12 13 14 15 16 17

Thermo iCap Q Agilent 7500ce PE NexION 300D

Milestone Ultrawave

The Netherlands The Netherlands

Eurofins Fonterra

CEM MARS 6

NovaWAVE FA

New Zealand

Meiji Co., Ltd

Agilent 7700x

PE Multiwave 3000 TOP Analytikjena TOP Analytikjena

Japan Japan Japan

Morinaga Milk Industry Co., Ltd Megmilk Snow Brand Co., Ltd

Agilent 7800

PE NexION 300