An ion-neutral model to investigate chemical ionization mass spectrometry analysis of atmospheric molecules – application to a mixed reagent ion system for hydroperoxides and organic acids

Abstract

An ion-neutral chemical kinetic model is described and used to simulate the negative ion chemistry occurring within a mixed-reagent ion chemical ionization mass spectrometer (CIMS). The model objective was the establishment of a theoretical basis to understand ambient pressure (variable sample flow and reagent ion carrier gas flow rates), water vapor, ozone and oxides of nitrogen effects on ion cluster sensitivities for hydrogen peroxide (H2O2), methyl peroxide (CH3OOH), formic acid (HFo) and acetic acid (HAc). The model development started with established atmospheric ion chemistry mechanisms, thermodynamic data and reaction rate coefficients. The chemical mechanism was augmented with additional reactions and their reaction rate coefficients specific to the analytes. Some existing reaction rate coefficients were modified to enable the model to match laboratory and field campaign determinations of ion cluster sensitivities as functions of CIMS sample flow rate and ambient humidity. Relative trends in predicted and observed sensitivities are compared as instrument specific factors preclude a direct calculation of instrument sensitivity as a function of sample pressure and humidity. Predicted sensitivity trends and experimental sensitivity trends suggested the model captured the reagent ion and cluster chemistry and reproduced trends in ion cluster sensitivity with sample flow and humidity observed with a CIMS instrument developed for atmospheric peroxide measurements (PCIMSs). The model was further used to investigate the potential for isobaric compounds as interferences in the measurement of the above species. For ambient O3 mixing ratios more than 50 times those of H2O2, O3−(H2O) was predicted to be a significant isobaric interference to the measurement of H2O2 using O2−(H2O2) at m∕z 66. O3 and NO give rise to species and cluster ions, CO3−(H2O) and NO3−(H2O), respectively, which interfere in the measurement of CH3OOH using O2−(CH3OOH) at m∕z 80. The CO3−(H2O) interference assumed one of its O atoms was 18O and present in the cluster in proportion to its natural abundance. The model results indicated monitoring water vapor mixing ratio, m∕z 78 for CO3−(H2O) and m∕z 98 for isotopic CO3−(H2O)2 can be used to determine when CO3−(H2O) interference is significant. Similarly, monitoring water vapor mixing ratio, m∕z 62 for NO3− and m∕z 98 for NO3−(H2O)2 can be used to determine when NO3−(H2O) interference is significant.