Many big purifier companies claim only their HEPA filters capture nanoparticles, like this Molekule ad.

Although this claim is false, even journalists repeat this myth. Why is this belief so pervasive?

What size micron is a HEPA filter?

One reasonable explanation is that the definition of “HEPA” leads people down that route. For example, this definition on Wikipedia says HEPA filters must remove: “99.97% of particles that have a size greater than or equal to 0.3 µm.”

It only mentions particles 0.3 microns and above. So what about nanoparticles? Nanoparticles range from 0.001 microns to 0.1 microns. Can HEPA filters capture these?

This exact question has led to years of confusion and false advertising claims about HEPA filters, so let’s get to the bottom of it.

Why Our Intuitions About Filters Are Wrong

The answer is surprising. It’s surprising because our intuition about filters is wrong. Like most people, I had the intuition that HEPA filters work like a net.

If a particle is smaller than the holes in the net, it gets through. Makes sense!

That intuition is true for big particles. When large particles fly into a HEPA filter, they’re too big to get through, so they get stuck.

But if we zoom down to the really small particles—like nanoparticles—things start getting weird. Nanoparticles are so small that they bounce around like a pinball when they hit gas molecules. (Scientists call this Brownian Motion.) That means they fly in random zigzag patterns.

Nanoparticles are so small that they can fit through the fibers in filters, but they get stuck anyway. Flying in zigzag patterns means they end up hitting the fibers and getting stuck.

When a filter captures a particle in this way, scientists call it “diffusion.”

How Effective Is Diffusion at Filtering Nanoparticles?

OK, so diffusion captures nanoparticles. But how many nanoparticles? Maybe 50% of nanoparticles? It turns out that diffusion is far more effective than that. According to NASA, HEPA filters capture “virtually 100% of particulates.”

But where’s the data? Researchers at the University of Minnesota tested this question with weaker fiberglass furnace filters and higher-grade HEPA filters. In their test, they shot particles of silver from 3 to 20 nanometers at the filters.

The results showed that filters captured 99.99% of particles smaller than 5 nanometers.

Their results demonstrate that the miracle of Brownian motion isn’t just a HEPA thing. This principle works for any fiber filter, including furnace filters (also called “MERV filters”). In the graph above, the MERV filters captured about 99% of particles at 10 nanometers and 99.9% at 4 nanometers. That’s impressive for a cheap fiber filter.

Brownian motion works for masks too! For example, scientists tested 3M masks and found they were 96% effective down to 0.007 microns (7 nanometers). Researchers at the University of Massachusetts found similar results in their mask tests.

Whence The Focus on 0.3 Microns?

So why does Wikipedia mention the 0.3-micron particle size? Rather than being the lower limit, these larger 0.3-micron particles are right in the middle. It turns out, particles in this middle region are the hardest to capture.

That’s because 0.3-micron particles don’t fly in zigzag patterns much. They’re too big. But at the same time, they’re small enough that they don’t easily get caught in the fibers (fancy names when particles get caught in the fibers: “impaction” and “interception”).

Why Should I Care?

Purifier companies can prey on our faulty intuition and use it to overcharge us. For example, the Molekule purifier makes claims like this on their website:

And here’s IQAir explaining why their HyperHEPA filter is better than an “ordinary filter.”

The National Head of Sharp claimed on Quora that ordinary HEPA filters can’t capture anything below 0.3 microns.

I suspect that some people really do get so far into designing purifiers (and ads) without discovering this fact about filtration. Even a journalist at Wired repeated this myth. So some of these tricks may be ignorance rather than deception. But now you know the science!