| Abstract Title: | Demonstrating Traceability for Mercury Measurements in Water Samples |
| Presenter Name: | Teodor Andron |
| Company/Institution: | Jožef Stefan Institute |
| Session: | Special Session - Metrological Traceability for mercury analysis and speciation |
| Co-Authors: | Teodor Andron,Warren T. Corns,Milena Horvat,Jože Kotnik |
Abstract Information :
Traceable and reliable calibration in combination with complete uncertainty evaluation is the key part of ensuring comparable mercury measurements. Two calibration approaches are commonly used in mercury analyses: (i) saturated gaseous elemental mercury inside the bell-jar unit and (ii) aqueous solution of the certified reference material (CRM) NIST 3133. In the case of the gaseous standard, the temperature-dependent concentration of GEM in the bell vessel is calculated using the Dumarey equation, which has been questioned in recent years. Therefore, the purpose of this work was to practically demonstrate the possible differences between the two calibration methods in the determination of dissolved gaseous Hg (DGM) in the river waters. Using NIST 3133-certified reference material (CRM), we first tested the Tekran? 2505 gas calibration unit based on the Dumarey equation. The first step was to check the linearity in the working range 27.76 ? 277.6 pg for the gas standard at 10 §C and NIST 3133 and to compare the signals corresponding to the same Hg masses from each standard. After a two-tailed paired T-test, the p value was 0.638, which indicated that there is no significant difference between the two standards. To assess the validity of the gas standard, the net Hg mass of this standard was determined, using NIST 3133 as the calibration standard. Quantities corresponding to 250 pg Hg were analyzed in the temperature range of 10 to 24 §C of the gas calibrator unit. The relative combined uncertainty of the measurement results varied from 2.6 to 8.2% for a coverage factor of 2, with a major contribution coming from repeatability of the standards. In all cases, the 250 pg target was well within the uncertainty range, which again demonstrate that both calibration approaches are comparable. DGM from river waters in the concertation range between 0.5 ng/L to 34.51 ng/L was then analysed using both calibration methods. Taking into account the measurement uncertainties, the resultswere comparable and no significant difference were observed. This work demonstrates that the gas standard is a suitable calibration method for DGM.
