Ice can form homogenously in the atmosphere at temperatures below −35 ∘C (Koop and Murray, 2016) or heterogeneously at warmer temperatures when an ice-nucleating substance (INS) is present to initiate freezing (Murray et al., 2012; Kanji et al., 2017; Hoose and Möhler, 2012). Heterogeneous ice nucleation can take place via several different modes: immersion freezing, deposition nucleation, pore-condensation freezing, and contact freezing (Vali et al., 2015; David et al., 2019). Here we study immersion freezing, which involves the initiation of ice formation by an INS immersed in an aqueous droplet (Vali et al., 2015). This mechanism is thought to dominate ice formation in mixed-phase clouds (Ansmann et al., 2009; Westbrook and Illingworth, 2011).
Atmospheric INSs include mineral dust, soil dust, and bioaerosols (Murray et al., 2012; Kanji et al., 2017; Hoose and Möhler, 2012; Tang et al., 2016). While in the atmosphere, INSs can be transported over long distances and coated with organic and inorganic solutes (Burrows et al., 2009; Fröhlich-Nowoisky et al., 2016; Hinz et al., 2005; Tinsley et al., 2000; McNaughton et al., 2009; Usher et al., 2003; Falkovich et al., 2004). Therefore, to effectively predict ice nucleation in the atmosphere, the effects of solutes on the freezing properties of INSs in the immersion mode need to be determined. A better understanding of the effects of solutes on freezing properties may also lead to a better understanding of the mechanism of heterogeneous ice nucleation in general, which remains highly uncertain (Coluzza et al., 2017). Additionally, if different INS-solute combinations produce known and unique changes in freezing properties, it may be possible to use freezing responses to solute additions as “fingerprints” for different INSs in atmospheric samples, as suggested by Reischel and Vali (1975).
Solutes can decrease the ice-nucleating ability of INSs in the immersion mode by lowering the water activity in the solution (i.e., freezing point depression) (Rigg et al., 2013; Koop et al., 2000; Koop and Zobrist, 2009; Zobrist et al., 2008). Solutes can also modify the ice-nucleating ability of an INS by interacting with and/or modifying its surface, even at low solute concentrations (< 0.1 M). Several studies have investigated the effects of solutes at low concentrations on the freezing properties of mineral dusts in the immersion mode. Aqueous NH3 and NH4+ salts at low concentrations improve the ice nucleation ability of feldspars, micas, gibbsite, quartz, and kaolinite and have little to no effect on the ice nucleation ability of amorphous silica particles (Kumar et al., 2019a, b, 2018; Reischel and Vali, 1975; Whale et al., 2018). In some cases, K+ salts improve the ice nucleation ability of feldspars depending on the concentration of the salts and the freezing temperature (Yun et al., 2020; Perkins et al., 2020). LiI was found to increase the freezing temperature of kaolinite particles in one study (Reischel and Vali, 1975) but not in a more recent study (Ren et al., 2020). Other inorganic salts, including NaOH and NaCl, decrease the freezing temperatures of some types of mineral dust (Kumar et al., 2019a, b, 2018; Reischel and Vali, 1975; Whale et al., 2015). Inorganic acids either decrease the ice nucleation ability of mineral dust particles or have little effect, depending on the type of acid, exposure time, concentration of the acid, and type of mineral dust (Kumar et al., 2018; Burkert-Kohn et al., 2017; Sullivan et al., 2010b; Tobo et al., 2012; Augustin-Bauditz et al., 2014; Wex et al., 2014; Sullivan et al., 2010a; Link et al., 2020). On the other hand, organic solutes have often been found to have no effect on the ice-nucleating ability of mineral dust particles (Zobrist et al., 2008; Koop and Zobrist, 2009; Tobo et al., 2012; Wex et al., 2014; Kanji et al., 2019).
In comparison to mineral dust, there have only been a small number of studies that have investigated the effect of solutes at low concentrations on non-mineral dust INSs. Reischel and Vali (1975) studied the effects of a range of inorganic salts on the freezing properties of leaf-derived nuclei and found only small changes (less than 1.5 ∘C) in the freezing temperatures of this INS in the presence of each of the tested solutes. Whale et al. (2018) studied the effects of (NH4)2SO4 and NaCl on the ice nucleation ability of humic acid and found no significant change in freezing temperature in the presence of either solute. Attard et al. (2012) studied the effect of pH on the freezing properties of several Pseudomonas strains and found that acidic solutions decreased the ice nucleation activity of the Pseudomonas strains studied. Koop and Zobrist (2009) studied the effects of the solutes (NH4)2SO4, glucose, H2SO4, and PEG400 on the freezing properties of Snomax (a commercial product for artificial snow production made from components of Pseudomonas syringae) and found no effect of the solutes on the freezing temperature other than freezing point depression. Chernoff and Bertram (2010) studied the effects of H2SO4 coatings on the freezing properties of Snomax and similarly found that the coating caused no significant change in the ice-nucleating properties other than freezing point depression. Amato et al. (2015) injected Pseudomonas syringae suspensions in (NH4)2SO4 into a cloud simulation chamber and observed a slight decrease in the ice-nucleating activity compared to the cells in water, although these data were not corrected for freezing point depression by the solute. Weng et al. (2016) studied the effects of the cryoprotectants ethylene glycol, propylene glycol, and trehalose on Pseudomonas syringae and found no effect of the solutes on freezing temperature other than freezing point depression. Desnos et al. (2020) studied the effects of the cryoprotectant Me2SO on the freezing properties of Snomax and observed a decrease in the ice nucleation activity that was greater than that produced by freezing point depression. Schwidetzky et al. (2021) studied the effect of the inorganic salts NaCl, NH4Cl, NaSCN, and MgSO4 on Snomax and found that NaSCN and NH4Cl decreased the freezing temperature of Snomax, NaCl had no effect on the freezing temperature, and MgSO4 increased the freezing temperature of Snomax.
To expand on the limited studies mentioned above, we investigated the effect of (NH4)2SO4 at a low concentration (0.05 M) on the freezing properties of several types of non-mineral dust INSs of atmospheric relevance. (NH4)2SO4 was chosen because it is a common inorganic solute in the atmosphere. A concentration of 0.05 M was chosen because it is relevant for mixed-phase clouds in the atmosphere. Because (NH4)2SO4 causes an increase in the ice nucleation ability of most mineral dust particles even at low concentrations, we investigated whether it would have a similar effect on non-mineral dust INSs in the immersion mode. If (NH4)2SO4 has little to no effect on the freezing properties of non-mineral dust INSs, then a change in freezing temperatures of atmospheric samples in response to the addition of low concentrations of (NH4)2SO4 could potentially be used to identify the presence of mineral dust INSs in atmospheric samples.
