Mixed-phase clouds, which contain both supercooled liquid water droplets and ice crystals, cover >34% of the Earth's surface (Zhang et al., 2018) and significantly impact the atmospheric radiation budget: they cool the Earth by reflecting incoming solar radiation and warm the Earth by prohibiting the escape of outgoing longwave radiation (IPCC, 2013; Rosenfeld et al., 2014; Zhao et al., 2018). Mixed-phase clouds are also essential to cloud electrification and the production of precipitation (Liu et al., 2019; Rosenfeld et al., 2008). Ice formation in the atmosphere is primarily the result of competition between the activation energy and interface energy for water molecules (DeMott, 2002; Seifert, 2011). To generate ice particles, the energy barrier needs to be broken by increasing the interface energy, which depends on temperature and the logarithm of supersaturation. This breakage can be realized either by homogeneous nucleation occurring at temperatures below −38°C (Pruppacher, 1995) or by heterogeneous nucleation at warmer temperatures ranging from −38°C to 0°C with some insoluble aerosols acting as ice-nucleating particles (INPs). For mixed-phase clouds, ice crystals are produced by heterogeneous nucleation process via four basic ice nucleation modes, e.g., deposition nucleation, immersion nucleation, contact nucleation, and condensation nucleation (Pruppacher & Klett, 1997). However, the heterogeneous nucleation process is still poorly understood due to the complexity of the INP nucleation efficiency corresponding to its chemical and microphysical properties (Cantrell & Heymsfield, 2005; Kanji et al., 2017). Moreover, current assessments suggest that the net radiation forcing related to the interaction between aerosols and mixed-phase clouds shows a large uncertainty and even exhibits a contradictory variation from −0.67 to +0.70 W m −2 (Fan et al., 2016; IPCC, 2013). Dust, soot, and biological particles are considered the three most effective INPs in the atmosphere. The yearly averaged emission of dust aerosols is estimated to be 1,000-3,000 Tg, which is far more than that of either soot (∼50 Tg) or organic material (∼60 Tg) (Seifert, 2011), and thus provides a promising opportunity to study heterogeneous nucleation in an actual atmospheric environment. Numerous studies concerning dust-related heterogeneous ice formation have involved laboratory experiments (