Determination of lead and cadmium in ceramicware leach solutions by graphite furnace atomic absorption spectroscopy: method development and interlaboratory trial.

This method was developed to improve sensitivity and eliminate time consuming, evaporative pre-concentration in AOAC Method 973.82 and American Society for Testing and Materials method C738 for testing foodware. The method was developed using leach solutions obtained by leaching 9 differently decorated ceramic vessels with 4% acetic acid for 24 h at room temperature. Lead and cadmium concentrations in leach solutions were 0.005-17,600 and 0.0004-0.500 microg/mL, respectively. Concentrations were determined using peak area, phosphate chemical modifier (8.3 microg PO4(-3)), and a standard curve for quantitation. Optimized pre-atomization and atomization temperatures were 1,300 and 1,800 degrees C, respectively, for Pb and 1,100 and 1,700 degrees C, respectively, for Cd. Characteristic masses (mo) were 10 and 0.4 pg for Pb and Cd, respectively. Precision of repeated analyses of calibration solutions was < or =3% relative standard deviation. Precision of duplicate leach solution analyses on different days was 0-9% relative difference. Recovery from fortified leach solutions was 96-106%. Results obtained by this method agreed 92-110% with those of confirmatory analyses. Results of certified reference material solutions agreed 94-100% with certificate values. Pb and Cd limits of quantitation (LOQ) were 0.005 and 0.0005 microg/mL, respectively. Results from 3 trial laboratories for 4 leach solutions containing Pb and Cd concentrations of 0.017-1.47 and <0.0005-0.0864 microg/mL, respectively, agreed 89-102% with results of the author. Two attributes of this method were noteworthy: (1) Background absorbance due to organic matter was entirely absent from atomization profiles, making the use of short pre-atomization hold times (2 s) possible. (2) Instrument precision was excellent and only one determination per solution was needed. Acceptance criteria for quality control measurements and a practical procedure for estimating the method LOQ during routine regulatory analyses are described.