To retrieve the physical properties of aerosols from multi-channel ground-based/satellite measurements, we developed a shape model of coated soot particles and created a dataset of their optical properties. Bare soot particles were assumed to have an aggregate shape, and two types of aggregates with different size–shape dependences were modeled using a polyhedral Voronoi structure. To simulate the detailed shape properties of mixtures of soot aggregates and adhered water-soluble substances, we propose a simple model of surface tension derived from the artificial surface potential. The light-scattering properties of the modeled particles with different volume fractions of water-soluble material were calculated using the finite-difference time-domain method and discrete-dipole approximation. The results of the single-scattering albedo and asymmetry factors were compared to those of conventional internally mixed spheres (i.e., effective medium spheres based on the Maxwell–Garnett approximation and simple core–shell spheres). In addition, the lidar backscattering properties (i.e., lidar ratios and linear depolarization ratios) of the modeled soot particles were investigated. For internally mixed soot particles, the lidar backscattering properties were sensitive to the shape of the soot particles and the volume mixing ratio of the assumed water-soluble components. However, the average optical properties of biomass burning smoke, which have been reported from in situ field and laboratory measurements, were difficult to explain based on the individually modeled particle. Nonetheless, our shape model and its calculated optical properties are expected to be useful as an alternative model for biomass burning smoke particles in advanced remote sensing via multi-channel radiometer and lidar measurements.