Abstract. Heterogeneous ice nucleation is an important mechanism for the glaciation of mixed phase clouds and may also be relevant for cloud formation and dehydration at the cirrus cloud level. It is thought to proceed through different mechanisms, namely contact, condensation, immersion and deposition nucleation. Supposedly, deposition nucleation is the only pathway which does not involve liquid water but occurs by direct water vapor deposition on a surface. This study challenges this classical view by putting forward the hypothesis that what is called deposition nucleation is in fact homogeneous or immersion nucleation occurring in pores and cavities that may form between aggregated primary particles and fill with water at relative humidity RHw < 100% because of the inverse Kelvin effect. Evidence for this hypothesis of pore condensation and freezing (PCF) originates from a number of only loosely connected scientific areas. The prime example for PCF is ice nucleation in clay minerals and mineral dusts, for which the data base is best. Studies on freezing in confinement carried out on mesoporous silica materials such as SBA-15, SBA-16, MCM-41, zeolites and KIT have shown that homogeneous ice nucleation occurs abruptly at T=230–235 K in pores with diameters (D) of 3.5–4 nm or larger but only gradually at T=210–230 K in pores with D=2.5–3.5 nm. Melting temperatures in pores are depressed by an amount that can be described by the Gibbs–Thomson equation. Water adsorption isotherms of MCM-41 show that pores with D=3.5–4 nm fill with water at RHw = 56–60% in accordance with an inverse Kelvin effect. Water in such pores should freeze homogeneously for T < 235 K even before relative humidity with respect to ice (RHi) reaches ice saturation. Ice crystal growth by water vapor deposition from the gas phase is therefore expected to set in as soon as RHw > 100%. Pores with D > 7.5 nm fill with water at RHi > 100% for T < 235 K and are likely to freeze homogeneously as soon as they are filled with water. Water in pores can freeze in immersion mode at T > 235 K if the pore walls contain an active site. Pore analysis of clay minerals shows that kaolinites exhibit pore structures with pore diameters of 20–50 nm. The mesoporosity of illites and montmorillonites is characterized by pores with T = 2–5 nm. The number and size of pores is distinctly increased in acid treated montmorillonites like K10. Many clay minerals and mineral dusts show a strong increase in ice nucleation efficiency when temperature is decreased below 235 K. Such an increase is difficult to explain when ice nucleation is supposed to occur by a deposition mechanism, but evident when assuming freezing in pores, because for homogeneous ice nucleation only small pore volumes are needed, while heterogeneous ice nucleation requires larger pore structures to contain at least one active site for immersion nucleation. Together, these pieces of evidence strongly suggest that ice nucleation within pores should be the prevailing freezing mechanism of clay minerals for RHw below water saturation. Extending the analysis to other types of ice nuclei shows that freezing in pores and cracks is probably the prevailing ice nucleation mechanism for glassy and volcanic ash aerosols at RHw below water saturation. Freezing of water in carbon nanotubes might be of significance for ice nucleation by soot aerosols. No case could be identified that gives clear evidence of ice nucleation by water vapor deposition on a solid surface. Inspection of ice nuclei with a close lattice match to ice, such as silver iodide or SnomaxTM, show that for high ice nucleation efficiency below water saturation the presence of impurities or cracks on the surface may be essential. Soluble impurities promote the formation of a liquid phase below water saturation in patches on the surface or as a complete surface layer that offers an environment for immersion freezing. If porous aerosol particles come in contact with semivolatile vapors, these will condense preferentially in pores before a coating on the surface of the particles is formed. A pore partially filled with condensed species attracts water at lower RHw than an empty pore, but the aqueous solution that forms in the pore will freeze at a higher RHi than pure water. The ice nucleation ability of pores completely filled with condensed organic species might be totally impeded. Pores might also be important for preactivation, the capability of a particle to nucleate ice at lower RHi in subsequent experiments when compared to the first initial ice nucleation event. Preactivation has often been explained by persistence of ice embryos at specific sites like dislocations, steps, kinks or pores. However, it is not clear how such features can preserve an ice embryo at RHi < 100%. Rather, ice embryos could be preserved when embedded in water. To keep liquid water at RHw well below 100%, narrow pores are needed but to avoid a strong melting point depression large pores are favorable. A narrow pore opening and a large inner volume are combined in "ink bottle" pores. Such "ink bottle" pores would be suited to preserve ice at RHi < 100% and can arise e.g. in spaces between aggregated particles.
Influence of particle size on the ice nucleating ability of mineral dusts
