Atmospheric Measurement Techniques (Feb 2024)

A traceable and continuous flow calibration method for gaseous elemental mercury at low ambient concentrations

  • T. D. Andron,
  • T. D. Andron,
  • W. T. Corns,
  • I. Živković,
  • I. Živković,
  • S. W. Ali,
  • S. W. Ali,
  • S. Vijayakumaran Nair,
  • S. Vijayakumaran Nair,
  • M. Horvat,
  • M. Horvat

DOI
https://doi.org/10.5194/amt-17-1217-2024
Journal volume & issue
Vol. 17
pp. 1217 – 1228

Abstract

Read online

The monitoring of low gaseous elemental mercury (GEM) concentrations in the atmosphere requires continuous high-resolution measurements and corresponding calibration capabilities. Currently, continuous calibration for GEM is still an issue at ambient concentrations (1–2 ng m−3). This paper presents a continuous flow calibration for GEM, traceable to NIST 3133 Standard Reference Material (SRM). This calibration approach was tested using a direct mercury analyser based on atomic absorption spectrometry with Zeeman background correction (Zeeman AAS). The produced continuous flow of GEM standard was obtained via the reduction of Hg2+ from liquid NIST 3133 SRM and used for the traceable calibration of the Zeeman AAS device. Measurements of atmospheric GEM using the calibrated Zeeman AAS were compared with two methods: (1) manual gold amalgamation atomic fluorescence spectrometry (AFS) calibrated with the chemical reduction of NIST 3133 and (2) automated gold amalgamation AFS calibrated using the mercury bell-jar syringe technique. The comparisons showed that a factory-calibrated Zeeman AAS device underestimates concentrations under 10 ng m−3 by up to 35 % relative to the two other methods of determination. However, when a calibration based on NIST 3133 SRM was used to perform a traceable calibration of the Zeeman AAS, the results were more comparable with other methods. The expanded relative combined uncertainty for the Zeeman AAS ranged from 8 % for measurements at the 40 ng m−3 level to 91.6 % for concentrations under 5 ng m−3 using the newly developed calibration system. High uncertainty for measurements performed under 5 ng m−3 was mainly due to instrument noise and concentration variation in the samples.