This is an interesting phenomenon that I haven't seen before. To start off with a discussion of coronas in general, they are fundamentally caused by diffraction, a form of optical scattering where the path of light from the sun is "bent" as it passes by the edge of a small particle. A simple illustration is case D in this example figure showing light scattering about a small cloud droplet:
The other interactions shown include reflection and refraction. For diffraction, when multiple interactions take place, the diffracted light produces interference patterns. A simple example is shown here for light being diffracted from two sides of a particle:
Ultimately, a corona is the result of a great many of these interactions taking place in the presence of a very large number of particles. However, as to pollen coronas in particular, it turns out that the type of particle largely doesn't matter - diffraction is such a major component compared to other optical scattering processes (e.g., reflection or refraction) that a corona may be formed whether the particles are liquid or solid:
Corona formation, to a good approximation, needs no knowledge of the droplet interior because the surface scattered waves predominate. It could be water, ink or coal - the pattern is almost the same. It depends primarily on the droplet size, shape and the wavelength of the light.
Pollen, it turns out, sometimes happens to be in a similar size range to cloud droplets. With diameters ranging between 15 and 200 μm, smaller pollen particles can be of similar size as cloud droplets (~20 μm in diameter), allowing them to have similar diffraction properties. In the end, the main difference is that pollen are less uniformly shaped than cloud droplets; preferential orientation as they are suspended in the air can cause the resultant corona to become elongated vertically or have particular bright patches in some cases.