AOAC RI ERP EBOOK FOR FERTILIZERS

AOAC RESEARCH INSTITUTE AOAC OFFICIAL METHODS OF ANALYSIS (OMA)

OMAMAN-28/AOAC 2006.03 Study Director: Sharon Webb, University of Kentucky, Division of Regulatory Services, 103 Regulatory Services Bldg, Lexington , Kentucky 40546-0275 different results if compared with AOAC 965.09 c) AOAC 2006.03 does not adequately recover Fe if compared with other methods, including AOAC 965.09. As a consequence, it appears that the addition of HCl to the digestion system is useful. With relation to Group A- metals, comparisons have been made with AOAC 2006.03 and the hot-block system. In general, there are not statistical differences (P = 0.05) except for Co and Mo; in addition, Cr shows high statistical differences (P = 0.01): this behaviour can be interpreted as a possible bias of AOAC 2006.03. Calibration and linearity The most suitable type of calibration curve has been determined (linear or quadratic) using the software provided by the instrument. For this reason, the calibration should be easy enough. Observed correlation coefficient values are between 0.99909 and 1.000000. Linearity has been determined by comparing responses for the working standard solutions over the range of expected concentrations. All results reported have undergone corrective calculations to include the test portion weight and total volume to yield the corrected values. Trueness or Accuracy Accuracy has been determined by results of analysis for all elements of interest, Group A and Group B, in NIST SRM 695 Trace Elements in Multi-Nutrient Fertilizer. Additionally, the results by the proposed method have been compared to consensus values for Magruder 2009-06, and expressed as percentage recovery of the certified or consensus value. Generally, percent recovery results are in the range: 74.6 – 102.5 (for NIST SRM 695) and 94.9- 149 (for Magruder 2009-06). Precision Method precision has been determined by independent analysis of the validation materials in triplicate, and variation has been expressed as relative standard deviation (RSD). Many results show HorRat values below 0.3 for a single validation study. In detail, HorRat values (total samples: 15) are below 0.3: a) seven times for calcium b) five times for cuprum c) eight times for iron (in addition, two HorRat values are reported are reported to be > 1.3) d) 12 times for magnesium e) three times for manganese (in addition, five HorRat values are reported to be > 1.3) f) three times for zinc (in addition, two HorRat values are reported to be > 1.3). However, RSD values for the proposed method are low if compared with collaborative study Page 6 of 14 The following strengths of the proposed method have to be also mentioned: