Peroxy radical measurements by ethane – nitric oxide chemical amplification and laser-induced fluorescence during the IRRONIC field campaign in a forest in Indiana

Abstract

Peroxy radicals were measured in a mixed deciduous forest atmosphere in Bloomington, Indiana, USA, during the Indiana Radical, Reactivity and Ozone Production Intercomparison (IRRONIC) during the summer of 2015. Total peroxy radicals ([XO2]≡[HO2]+Σ[RO2]) were measured by a newly developed technique involving chemical amplification using nitric oxide (NO) and ethane (C2H6) followed by NO2 detection by cavity-attenuated phase-shift spectroscopy (hereinafter referred to as ECHAMP – Ethane CHemical AMPlifier). The sum of hydroperoxy radicals (HO2) and a portion of organic peroxy radicals ([HO2*]=[HO2]+Σαi[RiO2], 0<α<1) was measured by the Indiana University (IU) laser-induced fluorescence–fluorescence assay by gas expansion instrument (LIF-FAGE). Additional collocated measurements include concentrations of NO, NO2, O3, and a wide range of volatile organic compounds (VOCs) and meteorological parameters. XO2 concentrations measured by ECHAMP peaked between 13:00 and 16:00 local time (LT), with campaign average concentrations of 41±15 ppt (1σ) at 14:00 LT. Daytime concentrations of isoprene averaged 3.6±1.9 ppb (1σ), whereas average concentrations of NOx ([NO] + [NO2]) and toluene were 1.2 and 0.1 ppb, respectively, indicating a low impact from anthropogenic emissions at this site. We compared ambient measurements from both instruments and conducted a calibration source comparison. For the calibration comparison, the ECHAMP instrument, which is primarily calibrated with an acetone photolysis method, sampled the output of the LIF-FAGE calibration source which is based on the water vapor photolysis method and, for these comparisons, generated a 50 %–50 % mixture of HO2 and either butane or isoprene-derived RO2. A bivariate fit of the data yields the relation [XO2]ECHAMP=(0.88±0.02;[HO2]+[RO2])IU_cal+(6.6±4.5) ppt. This level of agreement is within the combined analytical uncertainties for the two instruments' calibration methods. A linear fit of the daytime (09:00–22:00 LT) 30 min averaged [XO2] ambient data with the 1 min averaged [HO2*] data (one point per 30 min) yields the relation [XO2]=(1.08±0.05)[HO2*]-(1.4±0.3). Day-to-day variability in the [XO2]/[HO2*] ratio was observed. The lowest [XO2]/[HO2*] ratios between 13:00 and 16:00 LT were 0.8 on 13 and 18 July, whereas the highest ratios of 1.1 to 1.3 were observed on 24 and 25 July – the same 2 d on which the highest concentrations of isoprene and ozone were observed. Although the exact composition of the peroxy radicals during IRRONIC is not known, zero-dimensional photochemical modeling of the IRRONIC dataset using two versions of the Regional Atmospheric Chemistry Mechanism (RACM2 and RACM2-LIM1) and the Master Chemical Mechanism (MCM 3.2 and MCM 3.3.1) all predict afternoon [XO2]/[HO2*] ratios of between 1.2 and 1.5. Differences between the observed ambient [XO2]/[HO2*] ratio and that predicted with the 0-D modeling can be attributed to deficiencies in the model, errors in one of the two measurement techniques, or both. Time periods in which the ambient ratio was less than 1 are definitely caused by measurement errors (including calibration differences), as such ratios are not physically meaningful. Although these comparison results are encouraging and demonstrate the viability in using the new ECHAMP technique for field measurements of peroxy radicals, further research investigating the overall accuracy of the measurements and possible interferences from both methods is warranted.