Objective To evaluate the uncertainty of method for determining cadmium and its compounds in the air of workplaces by flame atomic absorption spectrometry.

Methods According to the principle and method of the national occupational health standards, GBZ/T160.5-2004 and JJF 1059.1-2012, the components of uncertainty were calculated.

Results The relative uncertainty of standard sampling volume was 0.008 4 and 0.004 8 for 75 L and 120 L, respectively. The uncertainty for the standard solution and preparation process, standard curve fitting, sample preparation process, and sample parallel determination was 0.006 6, 0.025 8, 0.030 0, and 0.001 5, respectively. The combined uncertainty was 0.041 0. The concentration of cadmium in the test sample was 0.067 mg/m³, and the expanded uncertainty for the measurement was 0.005 mg/m³ (k=2). The preparation process of sample solution, the fitting standard curve and the standard sampling volume were the main sources of the uncertainty of this method, and the other components had little influence on it. The uncertainty introduced by the sample digestion process in the preparation of sample solution made the greatest contribution to the combined uncertainty.

Conclusion During experimental process, attention should be paid to reducing the loss of sample pretreatment and choosing high purity standard liquid and/or standard. Quality control should be improved in the process of sample preparation and calibration curve fitting. The uncertainty of results should be minimized to ensure the accuracy and reliability of the experimental data.