Ice crystals in cold and in mixed phase clouds are assumed to be formed via different freezing mechanisms. While homogeneous freezing and deposition ice nucleation can be expected to be important for cold clouds, it is likely that immersion freezing plays an important role for mixed phase clouds. That is because atmospheric particles usually acquire soluble material during their stay in the atmosphere (through gas phase processes such as e.g. SOA-formation or through wet phase chemistry during cloud cycles). Also, atmospheric ice nuclei (IN) that are needed for heterogeneous freezing processes (such as deposition ice nucleation and immersion freezing) and that generally are insoluble particles, can be expected to be of comparably large sizes (see e.g. Welti et al., 2009), so that slight amounts of soluble material on them will be sufficient to let them activate at supersaturations typically found in the atmosphere. However, this additional material on the particles might hinder the freezing process, due to freezing point depression (for large amounts of soluble material), or due to interactions of the added material with sites on the particle that are responsible for the freezing (active sites).
During the past years, the immersion freezing process has been a focus of studies done at LACIS (Leipzig Aerosol Cloud Interaction Simulator (Stratmann et al., 2004) at the Leibniz Institute for Tropospheric Research, Germany). LACIS is a laminar flow tube of 7m length with a diameter of 15 mm. The temperatures inside LACIS can be varied from 298 down to 223 K under operational pressures from 700 hPa to ambient values. Residence times are in the order of seconds up to 1 minute. Inside LACIS, super-saturation with respect to water and/or ice is achieved by a combined heat and vapor diffusion process.
For the investigations described here, size-segregated, monodisperse Arizona Test Dust (ATD) and SiO2 particles were used as ice nuclei (IN). For some experiments, these particles were coated with sulfuric acid of varying amounts, at temperatures up to 85°C (giving coating thicknesses on the order of up to 4 nm) and/or they were thermally treated in a thermo denuder (at 250°C). Also, for some experiments, ammonia was added.
LACIS measured ice fractions (number of frozen droplets divided by the number of all droplets) as function of temperature, down to -40°C. Simultaneously to the freezing experiments, the ability of the particles to act as CCN (Cloud Condensation Nuclei) was measured with a CCN counter (Droplet Measurement Technologies). The data obtained with the CCN counter could be used to determine the amount of soluble material that was present on the particles. Furthermore, the particle composition was determined with aerosol mass spectrometry (AMS).
For both, pure ATD and pure SiO2 particles, the critical supersaturation for activation was at about 0.4% for a particle size of 300 nm (mobility size). This value is lower than would be expected for a completely insoluble particle, and this shows that even for high grade, clean laboratory generated particles contamination cannot be excluded. For the coated particles, the comparison of the amount of soluble material detected from CCN and AMS measurements gives insights to the nature of the soluble material. It was found that coating in general reduced the particles ability to act as IN (Niedermeier et al., 2010), and that additional thermal treatment could mostly not recover the particles IN ability (Sullivan et al., 2010). While coating and thermal treatment both were done at temperatures much higher than those occurring in the atmosphere, the experiments show how the combined examination of different particle properties can help to get insights on the still not fully understood freezing process.
Niedermeier, D. et al. (2010). Aerosol Chem. Phys., 10, 36013614.
Stratmann, F. et al. (2004). J. Atmos. Oceanic Technol., 21, 876-887.
Sullivan, R. et al. (2010). Aerosol Chem. Phys. Discuss., 10, 16901-16940.
Welti, A. et al. (2009). Aerosol Chem. Phys., 9, 67056715.