Determination of 27 Elements in Animal Feed by Inductively Coupled Plasma-Mass Spectrometry

A method was developed to determine 27 elements (aluminum, arsenic, boron, cadmium, calcium, chromium, cobalt, copper, iron, lead, magnesium, manganese, mercury, molybdenum, nickel, phosphorus, potassium, selenium, sodium, sulfur, vanadium, zinc, antimony, uranium, beryllium, thallium, and thorium) in animal feeds and pet foods using closed vessel microwave digestion and inductively coupled plasma-mass spectrometry. These elements can adversely affect animal health when amounts become excessive or deficient in food. For potentially toxic elements, the target LOQ was determined to be 1/10 the lowest maximum tolerable level (MTL) for the most sensitive animal species. For nutritionally essential elements, the target LOQ was determined to be 1/10 the concentration that would be considered deficient in feed if that level is lower than 1/10 the lowest MTL. The targeted high end of the quantitation range was selected as twice the highest MTL. A single-laboratory validation (SLV) was performed to determine the accuracy and precision of the method, and the results were evaluated with respect to predetermined performance characteristics. The test materials used in the SLV included two National Institute of Standards and Technology certified reference materials, two Association of American Feed Control Officials (AAFCO) check samples, and one in-house previously analyzed feed sample. The concentrations of each element determined were the result of spiked analyte, incurred analyte, or a combination of spiked and incurred analyte. All samples were analyzed seven times on different days at 2 x LOQ and at the midrange concentration. For most data, the results of the SLV met or exceeded the criteria for accuracy and repeatability. For accuracy, K at the midrange level had a mean recovery of 95%, which is just below the low preferred accuracy threshold of 97%. For repeatability, all the 2 x LOQ CVr values were below the preferred values. Be, Cr, Ni, Na, Mn, and S all had midrange CVr values exceeding the preferred values. Be had the highest midrange CVr value of 9.93. Eight AAFCO check samples were also analyzed to determine the method's accuracy and repeatability for elements at the manufacturer's claimed levels. For accuracy, all results had z-scores <1.5. For repeatability, three CVr values from two AAFCO check samples were greater than the preferred limits.