Expanding volatility calibration range of FIGAERO-ToF-CIMS

Filter Inlet for Gases and Aerosols (FIGAERO) inlet coupled with Time-of-Flight Chemical Ionization Mass Spectrometers (ToF-CIMS) have been successfully employed in numerous studies during past years (Thornton et. al., 2020). The instrument can be used to study both gas-phase and particle-phase chemical compounds over a broad range of chemical functionalities. The inbuilt controlled thermal desorption mechanism also allows for the investigation of volatility of the compounds present in the particle phase. Such a particle volatility measurement using FIGAERO-ToF-CIMS relies on accurate identification of the T_max for individual compounds, i.e. the temperatures at which highest respective signals are observed. These T_max values can be converted to saturation vapor pressure P_sat and saturation mass concentration C* values with a reliable calibration (Ylisirniö et. al., 2021). 

However, due to a lack of calibration compounds with known P_sat at the low end of the volatility range (P_sat <10^-9 (Pa) / C* < -10^-4 (μg/m³)), the current calibration procedure can only cover the desorption temperature range up to ~80-100 ˚C, while desorption temperatures of FIGAERO-ToF-CIMS can reach 200 ˚C and T_max values are routinely identified up to 160 ˚C. In this study we aim to extend the FIGAERO-ToF-CIMS calibration values to cover lower P_satvalues by utilizing a range of different approaches to increase the accuracy of the volatility measurement.

Methods

T_max values of Polyethylene Glycols (PEG’s) from PEG 5 to 15 were measured with FIGAERO-CIMS and corresponding C* values were either measured (for PEG 5-9) with isothermal evaporation experiments or estimated (PEG 5-15) with several different methods. Used estimation methods were quantum chemical modelling (COSMOtherm), desorption modelling (Schobesberger et. al., 2018), and parametrizations including Modified Grain Model, EVAPORATION, SIMPOL, and by Li et. al., 2016, Stolzenburg et. al., 2018 and Mohr et. al., 2019. Additionally, a parametrization equation according to Li et. al. 2016 was fitted to PEG 4-9 literature data and C* values of PEG 10-15 were estimated with this fit.                   

Results

Figure 1 shows the measured and estimated C*(298 K) values as function of the measured T_maxvalues. The spread of results with different models is increasingly larger especially at the higher order PEG’s, with over 10 orders of magnitude difference between models. Two independent measurement methods agree with each other within one order of magnitude. Best agreement with measurements is with desorption model, Modified Grain Model and COSMOTherm.

Figure 1. T_max vs. C*. Results from different measurements or estimates are shown in the figure legend. Note that some markers are very close to each other.

 

This work was supported by Academy of Finland Flagship funding (grant no. 337550)

 

Li et. al., 2016, Atmos. Chem. Phys., 16, 3327-3344.

Krieger et. al., 2018, Atmos. Meas. Tech., 11(1), 49–63.

Mohr et. al., 2019, Nat. Comm. 10, Article number: 4442.

Schobesberger et. al., 2018, Atmos. Chem. Phys., 18(20), 14757–14785.

Stolzenburg et. al., 2018, PNAS, 115 (37) 9122-9127.

Thornton et. al., 2020, Acc. Chem. Res., 53(8), 1415–1426.

Ylisirniö et. al., 2021, Atmos. Meas. Tech., 14(1), 355–367.