> Surely, if it can filter heavy isotopes, coronavirus won’t fit through.
Surprisingly, that reasoning doesn't necessarily actually work.
It turns out that there are actually several different mechanisms by which a filter can stop particles.
Big particles, for example, might not fit between the gaps in the filter--think fish in a net. This is called sieving.
Particles that are too small for sieving but are heavier than the surrounding flow keep moving in a straight line when the flow goes around the filter fibers. They collide with the fibers and get stuck. This is called inertial impaction.
The smallest particles that the filter can handle are not held in place by the fluid they are flowing in and so move around a lot by diffusion. This diffusion can lead them to hitting the fibers and getting stuck.
Particles too big for diffusion but too small for inertial impaction can follow the flow around fibers, but in doing so they can still hit the fiber and get stuck. This is called interception.
There are also electrostatic effects with some filter materials that can ensnare some kinds of particles.
When you put this all together, the result is that filters do not work the way we would intuitively expect, where there is some particular size and everything above that is stopped and everything below that makes it through. That would only be true if sieving was the only mechanism in play.
The curves of efficiency vs. particle size for all of the non-electrostatic mechanisms are S curves. As size goes up, sieving, inertial impact, and interception all go up, but at different rates.
Sieving's curve rising section is almost vertical. Inertial impact's is fairly rapid but nowhere near as rapid as sieving's. Interception's is much more relaxed.
Diffusion is also an S curve, but it goes the other way, being high for small particles and dropping for large particles.
When you add them all up you end up with a curve that is high and flat for small particles, then dips down around some particular size, and then rises back up to high efficiency.
There's some nice illustrations and graphs here [1].
This is why 0.3 microns is used when rating HEPA filters. It's around the size that is hardest for them to handle.
[1] http://donaldsonaerospace-defense.com/library/files/document...