Abstract. The volatility of organic aerosols plays a key role in determining
their gas–particle partitioning, which subsequently alters the
physicochemical properties and atmospheric fates of aerosol particles.
Nevertheless, an accurate estimation of the volatility of organic aerosols
(OAs) remains challenging because most standards for particulate organic
compounds are not available, and even for those with standards, their vapor
pressures are too low to be measured by most traditional methods. Here, we
deployed an iodide-adduct long time-of-flight chemical ionization mass
spectrometer (LToF-CIMS) coupled with a Filter Inlet for Gases and AEROsols
(FIGAERO) to probe the relationship between the molecular formulae of
atmospheric organic aerosols' components and their volatilities. Tmax
(i.e., the temperature corresponding to the first signal peak of thermogram)
for calibrants was abstracted and validated from the desorption thermograms
of mixed organic and inorganic calibrants that were atomized and then
collected on a PTFE filter, leading to a linear correlation between
Tmax and volatility. In addition, 30 ambient filter samples were
collected in winter 2019 at Wangdu station in the Beijing–Tianjin–Hebei region
and analyzed by FIGAERO-LToF-CIMS, leading to the identification of 1448
compounds dominated by the CHO (containing carbon, hydrogen, and oxygen
atoms) and CHON (containing carbon, hydrogen, oxygen, and nitrogen atoms)
species. Among them, 181 organic formulae including 91 CHO and 90 CHON
compounds were then selected since their thermograms can be characterized
with clear Tmax values in more than 20 out of 30 filter samples and
subsequently divided into two groups according to their O / C ratios and
different thermal desorption behavior. The mean O / C of these two groups is
0.56±0.35 (average ± 1 standard deviation) and 0.18±0.08, respectively. Then the parameterizations between volatility and
elemental composition for the two group compounds were obtained. Compared
with previous volatility parameterizations, our functions provide a better
estimation for the volatility of low-volatility organic compounds (LVOCs) in
ambient organic aerosols. Furthermore, our results suggest that volatility
parameterizations should be specialized for organic compounds with different
O / C ratios.