Characterisation of the Dekati Oxidation Flow Reactor: A New Platform for Secondary Organic Aerosol Formation Research

Since emerging with the concept of potential aerosol mass (Kang et al, 2007), oxidation flow reactors (OFRs) have shown great promise for atmospheric research. Combining high oxidant concentrations with low residence times enables simulating days of atmospheric photochemical aging within seconds. The Dekati oxidation flow reactor (DOFR) is a new OFR aiming to provide a platform for secondary organic aerosol (SOA) formation for both atmospheric emissions regulation and research. A compact design and short residence times enable characterisation of SOA formed from changing emission profiles such as a vehicle exhaust (Kuittinen et al, 2021). Developed from the TUT Secondary Aerosol Reactor (TSAR) (Simonen et al, 2017), the DOFR retains much of the same operational principles within a commercialised design. For the DOFR to succeed as a commercial instrument for regulatory and research use, consistent performance and atmospheric relevance of formed SOA are essential. To evaluate this a characterisation project has been undertaken testing fundamental DOFR parameters including VOC losses and particle penetration alongside application to a range of case studies representative of future use cases.

VOCUS PTR-MS has been used to measure VOC and IVOC losses for 15 different compounds including α-pinene, isoprene, benzene, and σ-xylene. Losses throughout the DOFR and sample conditioning unit (SCU) were not uniform across the tested compound range, however the SCU and post reactor dilutor were frequently identified as causing the most significant losses.

The DOFR uses 254 nm UV lamps to initiate oxidation and OH formation, requiring ozone supply from the SCU. Implications of pre-reactor ozone exposure was tested by repeating VOC loss measurements in the presence of 5 ppm of ozone. Losses for ozone reactive species such as α-pinene and β-caryophyllene showed more significant losses than in the absence of ozone compared to non-ozone reactive species. Figure 1 shows remaining fraction of injected VOC after the SCU both with and without ozone.

Figure 1- Fraction of remaining injected VOC after the SCU with and without ozone vs. species ozonolysis rate constant.

Particle penetration data has been collected for particles between 10-100 nm for two separate DOFRs and conditioning units. Full analysis is ongoing but initial results for the entire system have shown high transmission (≥ 90%) in the 70–100 nm range with losses becoming more significant (penetration < 80%) below 30 nm. Particle penetration data probing dependence on illuminated reactor lamps and system flow rates has also been collected, showing notable impacts on particles less than 30 nm. Characterisation of these losses is critical for the DOFR to succeed as a platform for aerosol research ensuring losses of primary and secondary particles are understood.

Final characterisation of the DOFR will combine both the VOC and particle loss data above as well as stability and repeatability of parameters such as OH concentration, humidity and temperature, and reactor flows and dilution factors. Parameter data from case studies including single VOC systems and complex emission sources such as a combustion engine has been collected for this characterisation and will aid development of standard operating procedures.