Laserfiche WebLink
<br />, I <br /> <br />'.1 <br /> <br /><.:) <br />C'J <br />1''' <br />,,::, <br /> <br />,-,-, <br />v <br /> <br />w <br /> <br />., <br /> <br />Muscle plugs were analyzed for selenium by neutron activation analysis at the University <br /> <br />of Missouri Research Reactor (I\1URR), Columbia, MO, using a gamma-ray spectrophotometer <br /> <br />with a Tennelec 244 amplifier coupled to a Nuclear Data 59910ss-free counting module and a <br /> <br />Nuclear Data 581 ADC. Samples were prepared for analysis at the Colwnbia Environmental <br /> <br />Researcf1 Center (CERC), Columbia, MO, according to the methods described in Waddell and <br /> <br />May (1995) Determination of radionuclide Se77m followed the methods of McKown and Morris <br /> <br />(1978), <br /> <br />Quality Control <br /> <br />Quality control measures for HGAAS analysis of selenium included the preparation and <br /> <br />analysis of triplicate samples (to measure precision), digested sample spikes (to measure loss of <br /> <br />seleniwn), analysis spikes (to measure matrix effects), procedural blanks (to measure <br /> <br />contamination and determine detection limits), and standard reference material (to measure <br /> <br />accuracy). The standard reference materials (SR1v!) used were water from the National Institute <br /> <br />of Standards and Technology (SRM 1643d, Gaithersburg, MD) and dogfish tissue from the <br /> <br />National Research Council of Canada (DORM-2, Ottawa, Canada), <br /> <br />Consistent sample handling during preparation, digestion, and analysis was indicated by <br /> <br />the low mean percent relative standard deviations, which ranged from 3.7 to 6. I % (Table 2). <br /> <br />Recoveries of selenium in digested spiked samples and analysis spikes were excellent <br /> <br />(97-103%), which indicated no significant loss of seleniwn from the original sample during the <br /> <br />digestion procedure and no significant interference from the other matrix components, Seleniwn <br /> <br />concentrations in all procedural blanks (n=16) were lower than the calculated limit of detection. <br /> <br />The analysis of reference materials were within acceptable ranges established by the CERC (8. <br /> <br />Brumbaugh, CERC, personal communication) and were within 10% of certified values, which <br /> <br />21 <br />