Heterogeneous ice nucleation of α‐pinene SOA particles before and after ice cloud processing

Abstract

The ice nucleation ability of α‐pinene secondary organic aerosol (SOA) particles was investigated at temperatures between 253 and 205 K in the Aerosol Interaction and Dynamics in the Atmosphere cloud simulation chamber. Pristine SOA particles were nucleated and grown from pure gas precursors and then subjected to repeated expansion cooling cycles to compare their intrinsic ice nucleation ability during the first nucleation event with that observed after ice cloud processing. The unprocessed α‐pinene SOA particles were found to be inefficient ice‐nucleating particles at cirrus temperatures, with nucleation onsets (for an activated fraction of 0.1%) as high as for the homogeneous freezing of aqueous solution droplets. Ice cloud processing at temperatures below 235 K only marginally improved the particles' ice nucleation ability and did not significantly alter their morphology. In contrast, the particles' morphology and ice nucleation ability was substantially modified upon ice cloud processing in a simulated convective cloud system, where the α‐pinene SOA particles were first activated to supercooled cloud droplets and then froze homogeneously at about 235 K. As evidenced by electron microscopy, the α‐pinene SOA particles adopted a highly porous morphology during such a freeze‐drying cycle. When probing the freeze‐dried particles in succeeding expansion cooling runs in the mixed‐phase cloud regime up to 253 K, the increase in relative humidity led to a collapse of the porous structure. Heterogeneous ice formation was observed after the droplet activation of the collapsed, freeze‐dried SOA particles, presumably caused by ice remnants in the highly viscous material or the larger surface area of the particles.

The efficiency of secondary organic aerosol particles acting as ice-nucleating particles under mixed-phase cloud conditions

“…Further studies found that SOA from α-pinene is initially an inefficient INP under cirrus cloud conditions and shows an increased ice nucleation ability when precooled or preactivated, i.e. after cloud processing, where the aerosol first activated to supercooled cloud droplets and then froze homogeneously (Ladino et al, 2014;Wagner et al, 2017). These and other studies on the ice-nucleating ability of SOA particles are summarized in Hoose and Möhler (2012) and Knopf et al (2018).…”

“…In addition to the different particle generation procedures, the resulting particle sizes vary, too, which certainly alters the ice nucleation capability (larger particles provide a larger surface area and are more likely to form ice). The findings of the above-mentioned studies can be summarized as follows: Wang et al (2012) and Ignatius et al (2016) found that atmospheric SOA particles are potentially important for ice nucleation due to their semi-solid or solid phase states by investigating SOA from naphthalene and α-pinene, respectively, whereas Ladino et al (2014) and Wagner et al (2017) found that α-pinene SOA at first is an inefficient INP at cirrus temperatures, but after precooling of the SOA particles, ice nucleation ability is increased. Schill et al (2014) found that semi-solid or glassy SOA from aqueous processing of methylglyoxal with methylamine is a poor depositional INP; however, Wilson et al (2012) found other aqueous glassy aerosol to nucleate ice heterogeneously at temperatures relevant for cirrus formation in the tropical tropopause layer.…”

“…Ice nucleation has been tested e.g. in expansion chambers (Murray et al, 2010;Wilson et al, 2012;Wagner et al, 2017), in continuous-flow diffusion chambers and flow tubes (Ladino et al, 2014;Ignatius et al, 2016), and in microscope systems (Wang et al, 2012;Baustian et al, 2013). Whether these formation pathways and ice nucleation initiation methods have an impact on the ice nucleation ability of the SOA particles is not clear.…”

“…Several further studies have investigated the ice-forming capability of different SOA particles (e.g. Prenni et al, 2009;Wang et al, 2012;Ladino et al, 2014;Schill et al, 2014;Ignatius et al, 2016;Wagner et al, 2017) in the laboratory. The procedures for generation of the SOA particles as well as the methods of initiating ice nucleation do vary substantially among these experiments.…”

“…The procedures for generation of the SOA particles as well as the methods of initiating ice nucleation do vary substantially among these experiments. For example, SOA formation has been initiated by dark ozonolysis (Prenni et al, 2009;Wagner et al, 2017) or photochemical reactions (Ladino et al, 2014;Ignatius et al, 2016), using gas-phase reactions (Wang et al, 2012;Ignatius et al, 2016) or aqueous processing (Wilson et al, 2012;Baustian et al, 2013;Schill et al, 2014). Ice nucleation has been tested e.g.…”