Towards a high quality in-situ observation network for oxygenated volatile organic compounds (OVOCs) in Europe: transferring traceability to the International System of Units (SI) to the field

Abstract. Volatile organic compounds (VOCs) have a large impact on the oxidising capacity of the troposphere and are major precursors of tropospheric ozone and secondary atmospheric aerosols. Accurate measurements and data comparability of VOCs among monitoring networks are essential to assess the trends of these secondary air pollutants. Metrological traceability of the measurements to the international system of units (SI-traceability) contributes to both: measurement consistency and data comparability. Accurate, stable and SI-traceable reference gas mixtures (RGMs) and working standards are needed to achieve SI-traceability through an unbroken chain of calibrations of the analytical instruments used to monitor VOCs. However, for many oxygenated VOCs (OVOCs), such RGMs and working standards are not available at atmospheric amount of substance fraction levels (< 10 nmol mol^-1). Here, we present the protocols developed to transfer SI-traceability to the field by producing two types of SI-traceable working standards for selected OVOCs. These working standards, based on RGMs diluted dynamically with dry nitrogen and on certified spiked whole air samples, were then assessed using Thermal Desorption-Gas Chromatography-Flame Ionization Detector (TD-GC-FID) and Proton Transfer Reaction-Time of Flight-Mass Spectrometry (PTR-ToF-MS) as analytical methods. For that purpose, we calibrated five analytical instruments using in-house calibration standards and treated the new SI-traceable working standards as samples. Due to analytical limitations, the assessment was only possible for acetaldehyde, acetone, methanol and methyl ethyl ketone (MEK).  Relative differences between assigned and measured values were used to assess the working standards based on dilution of RGMs. The relative differences were within the measurement uncertainty for acetone, MEK, methanol and acetaldehyde at amount of substance fractions around 10 nmol mol^-1. For the working standards based on certified spiked whole air samples in pressurized cylinders, results showed a good agreement among the laboratories (i.e., differences within the measurement expanded uncertainties (U) ranging between 0.5 nmol mol^-1 and 3.3 nmol mol^-1) and with the certified amount of substance fraction value for acetaldehyde (15.7 nmol mol^-1 ± 3.6 (U) nmol mol^-1), acetone (17 nmol mol^-1 ± 1.5 (U) nmol mol^-1) and MEK (12.3 nmol mol^-1 ± 2.3 (U) nmol mol^-1). Despite the promising results for the working standards based on the dilution of RGMs and on certified spiked whole air samples filled into pressurized cylinders, the assessment must be considered with care due to the large measurement uncertainty, particularly for methanol. Active collaboration among metrological, meteorological and atmospheric chemistry monitoring communities is needed to tackle the challenges of OVOC monitoring, such as the lack of stable and SI-traceable calibration standards (i.e., RGMs and working standards). Besides from this collaboration, other research applications, such as modelling and remote sensing, may benefit from the transfer of SI-traceability to monitoring stations.