SPSFAM Heavy Metals ERP Book

AOAC Expert Review Panel for Heavy Metals 

ER 1 Mean recovery for NIST SRM 1643e was 96% determined as iAs. Although not mentioned in the  SMPR, mass balanced is an important parameter to consider when performing speciation  analysis. Good mass balance indicates that the majority of the total arsenic in the samples is  accounted for during the speciation analysis. The mass balances obtained for all 13 samples  were within the method's acceptable range of 65‐ 115%. ER 2 It had been included for the review for First Action Method.  ER 3 I don't know. I can't find SLV data. I only have the MLV data.  MLT was performed with 8 laboratories using 13 juice samples including apple, grape and pear  juices and apple juice concentrate. One major challenge of an MLT is finding foods that have  levels of analytes >LOQ.  In order to have food samples with meaningful levels of analytes,  selected samples were fortified with various combinations of inorganic arsenic, DMA, and/ or  MMA. These fortification values were not disclosed to participating laboratories Levels were  selected to mimic those reported in the literature and, thus, provide a meaningful challenge for  the MLT. The LOQs for the various species were either 10 or 11ppb compared to the SMPR requirement  of 10ppb. However, the method uses a very conservative approach to LOQ (30*sigma). If the  usual 10*sigma equation was used then all LOQ values would be well below the SMPR  requirement. The RSDR for inorganic arsenic (iAs) ranged from 3.9% to 12.0% for the 11 samples with iAs >  LOQ. As would be expected wit fruit juices, the only sample with DMA > LOQ was the sample  spiked by the central lab before the samples were sent out to the participating labs. The RSDR  for this sample was 6.7%. Four samples had MMA > LOQ. The RSDR for these samples ranged  from 6.0% to 9.3%.  Each laboratory prepared fortified samples induplicate at three spike concentration levels and  then determined spike recoveries. The nominal spike concentrations in the RTD juices were 5,  25, and 50 ppb for As(III), As(V), DMA, and MMA. The nominal spike concentrations in the juice  concentrates were 35, 100 and 200 ppb. Average spike recoveries ranged from 97% to 105%.  These fortification recovery results demonstrate acceptable accuracy and precision of the  method over the range of concentrations of interest and do not indicate any significant matrix  interferences. including signal enhancement. Yes. Multilibrary validation document demonstrates the reproducibility and probability of  detection.  The LOQ published in first action method is based on multilibrary validation report.  Uncertainty of measurement is in section N of first action method with a reference to EAM  section 3.3 for detailed discussion for combined and expanded uncertainty published in this  method.

ER 4

5 of 9