christina j williamson

Cloth Masks to reduce COVID19 transmission

My colleagues and friends have been finding short explanations of the basic science behind using masks to reduce transmission of COVID19 and the current knowledge on what makes effective masks useful. I’m therefore sharing this information here for anyone who wishes to access it. Before we dive in, a couple of caveats:

  1. I am an aerosol scientist specializing in how particles affect climate. While there is a lot of cross-over to understanding how aerosols transmit diseases, there are also many aspects of this that are far outside my expertise. The information that follows is not medical advice, and I am not qualified to give medical advice.
  2. Scientific knowledge of COVID19 transmission is advancing at a rapid pace. I will endeavor to keep this page current, but there may well be advances I am unaware of.
  3. The contents of this post are mine personally and do not necessarily reflect any position of CU Boulder, CIRES, the U.S. Government or the National Oceanic and Atmospheric Administration.

I would also like to note that similar information to what I cover in this post, has been covered by Prof. John Volckens, Mechanical Engineering Professor at Colorado State University, in an excellent online video about masks for COVID19. I highly recommend watching this.

People to follow on twitter with expertise relevant to airborne transmission of viruses:

Prof. Linsey Marr, Engineering Professor at Virginia Tech with expertise in airborne transmission of viruses @linseymarr

Prof. Shelley Miller, Environmental Engineering Professor at the University of Colorado, Boulder with expertise in indoor air pollution @ShellyMBoulder

Prof. Jose Luis Jimenez, Chemistry Professor at the University of Colorado, Boulder with expertise in aerosols. Tweets in English and Spanish @jljcolorado.

Dr. Richard Corsi, Dean of the Maseeh College of Engineering and Computer Science at Portland State University, expertise in indoor air quality and climate, @CorsIAQ.

Prof. John Volckens, Mechanical Engineering Professor at Colorado State University, expertise in masks for COVID19, @Smogdr

We emit aerosols and droplets when we cough, speak and just breathe

Particles are naturally present in our breath, and are emitted when we breath, speak and cough.

Fig 1. This is part of a plot from a paper “Modality of human expired aerosol size distributions” published in 2011 in the Journal of Aerosol Science by Johnson et al. It shows the number of particles (left axis) of different sizes (bottom axis) measured when a person speaks (top) and coughs (bottom). Particles are also present just when we breath.

The particles are present in two groups according to size (we call these modes), smaller ones, about 0.2 to 10 μm in size, and larger ones, about 10 to 1000 μm in size. The larger mode, as I’ll explain below, is easily stopped by most masks, so we focus on what makes masks good for the smaller mode.

μm may be an unfamiliar unit to you, a mm is 10-3 m, and a μm is 10-6 m. 0.1 μm is roughly the size of the particles in smoke, 10 μm is roughly a grain of pollen, 100 μm is roughly the width of a human hair.

For discussion of virus transmission, we tend to differentiate between aerosols and droplets. Aerosols are particles that stay in the air a long time and are smaller than about 100 μm. Droplets drop to the ground in a short amount of time, and a small distance, and are larger than about 100 μm. I will explain why they have this different behavior below. A lot of scientists are using this terminology to differentiate between aerosols and droplets now, but you may also hear the terms used differently from time to time.

Viruses themselves are only about 0.1 μm in size, smaller than any of the particles in your breath. They are not present in your breath on their own, they will always be attached to an aerosol or a droplet (shown in this paper by Vejerano and Marr published in 2018 in Journal of the Royal Society Interface, and this one by Blachere et al. published in Clinical Infectious Diseases in 2009).

COVID19 is transmitted on the aerosols and droplets in our breath

Airborne transmission of COVID19 via aerosols and droplets in our breath has been established by a number of scientific papers, effectively summarized in a concise and easy to read commentary in Clinical Infectious Diseases, “It Is Time to Address Airborne Transmission of Coronavirus Disease 2019 (COVID-19)“, published in July 2020 by Morawska and Milton, supported by 237 clinicians, infectious-disease physicians, epidemiologists, engineers and aerosol scientists. Prof Linsey Marr (Engineering Professor at Virginia Tech with expertise in airborne transmission of viruses) also summarize the science behind this in an op-ed in the New York Times, and Prof Jose Jimenez (Chemistry Professor at the University of Colorado, Boulder with expertise in aerosols) did likewise in Time Magazine.

Droplets drop to the ground quickly, aerosols stay in the air and move around

Aerosols (the smaller particles in our breath) and droplets (the larger particles in our breath) behave differently in the air. All particles in the air are affected by gravity, which makes them fall to the ground. Smaller particles are also affected by Brownian Motion, which is when they are bumped around by hitting into gas molecules in the air. The combination of gravity and brownian motion means that smaller particles take a lot longer to fall to the ground than big particles. An aerosol of 0.2 μm diameter would take about 10 days to fall from of a person’s mouth above the ground. A droplet of 200 μm would do it in just 3 seconds. The speed of exhaled aerosols and droplets can vary a lot, but we can take around 1 m/s as a rough guide for breathing, and around 5 m/s for sneezing, using measurements in this paper from Tang et al.

Therefore, you can avoid a lot of the large droplets (bigger than about 100 μm) coming out of someone’s nose or mouth simply by standing a few meters away from them, but this DOES NOT WORK for aerosols. Outside (or in a well-ventilated, uncrowded indoor space) aerosols will become more dilute the further away from the person emitting them you are. This makes distancing helpful, but not fully protective.

The size of a particle also affects the direction they travel in. Large droplets have a lot of inertia, and so tend to travel in straight lines. Small particles have very little inertia and so tend to follow the direction of the air flow, even as it bends around corners. This has important implications of mask fit, which I will talk about later.

This picture, adapted from the paper “Airborne spread of infectious agents in the indoor environment” by Wei and Li in the American Journal of Infection Control in 2016, illustrates how larger droplets in our breath quickly fall to the ground (or other surfaces), but smaller aerosol stay floating in the air.

Masks filter aerosols and droplets when we exhale and inhale – they protect the wearer and the community

Filtration of aerosols and droplets is more complex than thinking of fabric like a sieve/strainer that catches particles that are bigger than the holes (this is good news, since aerosols are so small!). Super small aerosols move around a lot in the air thanks to the Brownian motion we discussed above. This means that when air passes through a filter, these smallest aerosols are likely to bump into a fiber in the filter. This process is known as diffusion. As air flows through the filter, it will follow tight curves to get through all the microscopic holes in the filter. The inertia we talked about above means that particles can’t always follow the tighter bends the air flows around, and eventually will go straight and bump into a fiber instead of flowing around it. This happens more frequently the larger a particle is, because it has more inertia, and so this filters out larger aerosols and droplet better. This process is known as impaction. For it to work, we need the holes in our filter to be small enough that the air flow has to twist and bend to get through. If the holes are too large, air goes in straight lines and the particles will go with it – turning it into the sieve, which is not very effective at all!

The graph below shows these two filtration mechanisms, plus a few others that we don’t need to go into here. The main point is that diffusion makes it quite easy to filter out the smallest aerosols, and impaction makes it quite easy to filter out droplets and the largest aerosols, but aerosols between about 0.1 and 0.5 µm are the hardest to filter out.

Good filter materials for preventing COVID19 transmission are the ones that not only get the easy-to-catch smallest aerosols and large droplet, but also are pretty efficient in this hard-to-catch range.

Graph from Lindsley, William. (2016). “Filter Pore Size and Aerosol Sample Collection“, showing the collection efficiency of different filtration methods (left axis) for different particles sizes (bottom axis). Diffusion and impaction are the two methods I’ve described above.

It’s helpful to note that, once aerosols and droplets bump into a fiber in a filter, they are generally stuck there and won’t be blown or sucked off. Be aware though, aerosols and droplets on the mask can get transferred to your fingers when you touch your mask, which is why we avoid touching the outside of our masks while wearing them, use the ear-loops/ties to remove the mask, and wash hands before and after putting on and taking off masks. It is the same reason behind why we cannot share masks with other people.

Masks protects the wearer by filtering aerosols and droplets in the air, stopping them from reaching the wearer’s nose or mouth.

Masks protect the community in two ways. By stopping aerosols and droplets from getting out into the general air they stop other people from breathing them in, and they stop them from ending up on surfaces that other people might touch. We also expect that aerosols in our breath are largest immediately upon being exhaled. This is because our breath is very humid, and the moisture causing that humidity also swells the aerosols making them bigger. Once particles have been in the open air a little bit, that extra moisture can evaporate off, shrinking them slightly. Since bigger particles are generally easier to filter out than smaller particles, masks are likely slightly more effective in protecting the community from the wearer than the wearer from the community. I’m not aware of any studies that quantify this effect though, so we don’t know how important this is yet.

Non-medical grade masks can still be effective protection against airborne viruses

You may have heard of N95 masks. These are used in medical and laboratory settings, for protection from wildfire smoke, and often in construction and woodwork to protect from dust. The name refers to a standard the masks have to reach, in a very specific lab test, they have to filter out 95% of particles that are 0.3 μm in diameter (this is the hardest size to filter our – see above). These have been in short supply since the pandemic hit, and are generally being reserved for medical workers. Thankfully, most of the particles in our breath are larger and easier to filter out than this tricky 0.3 µm size, so, in every-day settings, simple cloth masks can still provide large benefits in terms of protection for the wearer and for the community.

Three factors for a good cloth mask

When thinking about what makes a good cloth mask we need to consider 3 things (this framework was first brought to my attention by John Volckens of Colorado State University):

  1. Filtration
  2. Fit
  3. Breathability

Filtration

Filtration is how effectively the materials that make up your mask remove particles in air that flows through the mask.

John Volckens, Professor of Mechanical Engineering at Colorado State University, set up a laboratory to test cloth masks and medical grade masks at the beginning of the pandemic. He’s made his results available in a helpful, interactive format online. I’m showing a graph of all is cloth mask results here below. He sucks air with a measured number of aerosols in it through different materials that can be used to make masks, and measures how the number of aerosols and droplets decreases after passing through the mask. The graphs show the collection efficiency of each material on the left axis. This is the percentage of aerosols that get stopped by the mask. High collection efficiencies are good mask materials, low collection efficiencies are less useful. These collection efficiencies are shown as a function of the particle size, as each material filters differently for different sized particles. What you can see in this first graph is that not all materials perform equally well as filters. Some are very good, staying above 80% at all sizes, some are quite poor, with less than 10% for the smallest aerosols. This is why it is helpful for all of us not only to wear a mask, but also to put some consideration into what mask we are wearing.

The aerosol particles between 0.5 μm and 10 μm are where we see the biggest difference is mask efficiencies. You can see in the graph below that for particles bigger than 10μm the different materials all perform quite well and similarly to each other.

This graph shows the performance of different cloth masks tested by John Volckens’ lab at Colorado State University. These results are available in interactive plots online. The left axis shows the collection efficiency of the mask i.e. the percentage of particles caught by the mask. A collection efficiency of 100% would be a perfect mask, no particles would get through it. A collection efficiency of 0% would be a useless masks, all particles would go straight through it. Collection efficiencies are shown for different sized particles (bottom axis).

Now we know that different materials can make better or worse masks, we of course want to know how to pick materials the best masks. Using John’s helpful interactive website, I’ve pulled out his results for a number of different types of cotton materials. He’s used two layers of each type of cotton in his tests. You can see that the worst performing cotton is quite a light weave. The best performing is a tight weave with a high thread count.

Efficiencies of different kinds of double layer cotton masks, from John Volckens’ lab at Colorado State University. Generally tighter weaves, higher thread counts, smaller fibers make for better performing masks. Flannel is quite good, likely because it is sort of fuzzy, so there are fibers going in all kinds of directions, making it harder for an aerosol particle to pass through without interacting with a fiber.

There are ways you can test the masks you have, or the fabric you plan to make masks with at home, to see whether it is performing well like the tight weave, high thread count cotton above, or less well like those loose weave at the bottom of that same graph. Researchers at Georgia Tech have put together a rapid response team helping with different aspects of combatting COVID19. The have come up with a quick at-home test using a spray bottle and a mirror. Essentially, you spray a mist of clean water through your fabric, not letting it touch the mirror, and then see how much of the mist ends up on the mirror. Less mist on the mirror = better filter! Check out their full instructions for doing this test yourself.

A group of researchers from Georgia Tech have set up a COVID Rapid Response Team. They provide advice on making good masks, and recommend this easy test to see if your mask/fabric is good for filtering out aerosols. Check out more instructions and other useful information on their website.

To make even more effective masks than high thread-count, tight weave cotton, we can add a layer of filter material between the two layers of cotton (it’s really important for a lot of filter materials like the one’s John Volckens tested below, that they go between two layers of fabric, as it’s not safe to breathe directly through them). Filter materials are specially made to filter aerosols well. In addition to the standard ways fabrics filter aerosols, they often have a static electric charge added to them, which actually attracts the aerosols to the filter fibers, making it harder for them to get through. Often these filter materials are matted, rather than woven or knitted, and this leaves fewer tiny holes between fibers for the odd aerosol particle to get though. Many of these filter materials, often used in air purifiers and ventilation systems, are available commercially.

Adding filter materials between 2 layers of cotton can massively improve the performance of a mask, as shown by these efficiencies measured by John Volckens’ lab at Colorado State University.

I carried out some simple tests at home, sandwiching different paper materials between two layers of cotton as a very simple filter. Because I did these tests at home, they are not as accurate as proper lab tests like John’s and the results have high levels of uncertainty. Nevertheless, they are useful to help us make broad, qualitative conclusions. What I found was, when I added two layers of paper products, like toilet paper, between two layers of cotton, substantially fewer aerosol particles could pass through, making it a better filter. You can see my measurements in the graph below.

My at-home tests also showed that by putting different paper materials between two layers of cotton, aerosol particles were filtered out better. Note these results are from rough, at-home tests and so have large uncertainties. They are useful in a qualitative sense, so for example, here we can say that adding paper materials between two layers of cotton filters more aerosols than adding more cotton layers.

When the pandemic first started, and even now on hiking trails, I notice lots of people using neck-gaiters (also known as buffs), and bandanas as masks. These items are often to hand, and don’t require buying or making something specific, so it’s easy to understand why we’d want to use them. John Volkens’ results show that both a neck gaiter and a doubled over bandana performed quite poorly as mask materials. As you can see in the plot below. They do catch a good percentage of the droplets and largest aerosols, so they’re better than nothing, but they’re just not as helpful as using a more thought-through mask.

Neck Gaiters and Bandana perform worse that most 2 layer cotton masks, as shown by these efficiencies measured by John Volckens’ lab at Colorado State University. Even the worst performing masks here filter out some droplets, so they are better than nothing at all, but they are not nearly as protective as a couple of layers of tight-weave, high thread count cotton.

I also took a look at a neck-gaiter with my simple at-home system. The neck-gaiter I happened to have was merino wool, and I tested it as a single layer, because that is how it’s most commonly used. I’m showing the results below. You can see that the neck-gaiter, in yellow, lets a lot more aerosols through than 2 layers of a cotton sheet, so it’s much less protective as a mask.

In my rough at-home tests I also saw that my buff (or neck-gaiter) performed much worse than 2 layers of an old cotton sheet. My buff is merino wool, it is quite stretchy, and, in normal wearing, only 1 layer can go over my nose and mouth. All these factors likely combine to make it less effective at filtering out aerosols.

There are more studies on the efficacy of different materials for cloth masks out there. For example, Smart Air has a website showing measurements of filter efficiency and breathability of many different materials for making homemade masks.

A note on seams

Sewing puts small holes in fabric, and so seams are likely to be weak-points in your mask that may let more aerosols through. Masks that avoid a seam down the middle (or at least have that seam fully covered by an extra filter layer), and masks where the edge seams are outside of where you mask seals to your face are likely to perform better.

The conclusion of the different studies on filtering materials are:

  1. Tight weave, high thread count materials are the best fabrics to use in masks.
  2. Adding a second cloth layer makes masks better. Don’t just keep adding more layers, though. For top performance, using special materials or adding some kind of paper or other filter is better than going to 4 or 8 layers of cloth.
  3. Neck-Gaiters/Buffs and bandanas do not make great masks. They do stop some of the droplets, but will let a lot of the aerosols through. And we need to be stopping both aerosols and droplets.

Fit

Fit is how well the mask seals to your face. For droplets, the fit isn’t super important. Even if air escapes around the sides of your mask, droplets have enough inertia that they travel in fairly straight lines, and many of them will hit the mask in a straight line from your nose and mouth. For aerosols, fit it critical. If air is leaking around the sides of your mask, or up past your nose, the particles will go with it. Air will follow the path of least resistance, so if you give it a choice of passing through a mask, or going around the side, it will go around the side.

This figure from a paper entitled “Effectiveness of facemasks to reduce exposure hazards for airborne infections among general populations” published by Lai, Poon and Cheung, Journal of the Royal Society Interface in 2012. Shows how the mask they tested could protect from over 90% of particles when fully sealed, but on 40% of particles in the standard way it was worn because of leaks.

Luckily, it is quite easy to test how well your mask fits you.

Signs of a leak:

  1. Can you see a gap?
  2. Can you feel air moving past your face?
  3. If you exhale quickly, do you blink?
  4. Do your glasses/shades fog up over time?

If you answer yes to any of the questions above, your mask is leaking. Find a way to seal better to your face, and your mask will protect you and your community better. If you answered yes to #3 or #4, the air is leaking out the top of your mask, a likely place to look is around the sides of your nose. Nose strips (a strip of moldable plastic or metal, perhaps a twisty-tie) allow you to shape the mask to the contours of your nose and cheeks and can significantly reduce leaks in this area.

To form a good seal, your mask needs to big big enough. It must go over your chin and nose, and span from cheek to cheek. Mask fit can also be specific to face shape. Whether you’re making your masks, or buying commercially produced masks, it can help to try some different designs and sizes to get something that really works for your own face.

Many sources say that facial hair can make it very difficult to form a decent seal with your mask. I’ve looked unsuccessfully for resources to deal with this. If I find anything, I’ll post it here.

Signs of a good seal:

  1. Breathe in deeply and then exhale fast. Your mask should suck in and then puff out slightly.

A note on valved masks

Some N95 masks (and perhaps others) have valves that let unfiltered air out, and only filter air when you breathe in. This feature is designed when the only purpose of the mask is to protect the wearer e.g. from wildfire smoke or woodworking dust. In the case of a pandemic, it is equivalent to having a mask that leaks badly when you exhale. Aerosols will get out. Some droplets will be stopped, so it is still better than no mask at all, but it’s not doing the full job.

Fit and Filters – something to be aware of

Because some of the filter materials we use either aren’t washable (paper products), or become less effective with multiple washes (e.g. those that use static electric charge to help filter), lots of home-made and commercially available masks have filter pockets. This allows the mask to be washed, and the filter to be either left to decontaminate over a few days, or disposed of.

As I mentioned above, air will follow the path of least resistance as you breath in and out. The difficulty with filter pockets is that it can be hard to get a filter in them to reach right to the place where the mask seals to your face. This small gap is the path of least resistance, and the air will prefer to pass through here. The bigger the gap, and the bigger the difference in breathability with and without the filter, the more air will pass through the gap and not through the filter.

The ideal for a filter therefore is to get it to seal to you face just like the rest of the mask. This is much easier to achieve with a filter that is sewn into the mask, but of course that can cause issues with washing the mask. There is evidence that the COVID19 virus decays on different materials over time, and many people suggest leaving a masks in a separate area to decontaminate for a few days. Some researchers are also looking into methods to decontaminate N95 masks safely, to help medical workers deal with the shortage of masks.

I don’t have a perfect solution to the complications of fit and filter pockets at the moment. If I find something useful I’ll add it here.

Breathability

A good mask is a used mask! This means that you need to be able to breathe through your mask at a rate appropriate to your level of activity. It may even mean that you want a different mask for highly aerobic activities outdoors, to the one you use to sit in an office and shop for groceries. If you cannot breathe well enough through your mask, you will find yourself removing it, or creating gaps for the air to go around the mask (which is problematic, see the fit section above).

The less breathable your mask is, the more pressure will build up inside it when you exhale (or outside it when you inhale). This makes it more likely that a leak will occur. Finding/making masks you can breathe well enough in for the different activities where you should be wearing one is important.

Smart Air have produced some very helpful graphics on both the filtration and breathability of different mask materials, and include some helpful discussion of where extra layers are and are not helpful because of a breathability-filtration trade-off.

Summary

I hope that this article has been helpful. The main points I have covered are:

COVID19 is spread by both aerosols (small particles that stay in the air for hours and move around) and droplets (larger particles that fall to the ground within a few meters).

We all emit both aerosols and droplets when we cough, speak and even just breath.

Cloth masks can reduce COVID19 transmission by removing aerosols and droplets from the air we breathe out (protecting the community) and the air we breathe in (protecting ourselves).

A good mask needs to work well in three ways:

  1. filtration – double layered tight weave, high thread count fabrics are best, and additional filter materials sandwiched between fabric can make it even better
  2. fit – it needs to properly cover both nose and mouth, and to fully seal around the face)
  3. breathability – if you can’t breathe through it at the rate you need to for your level of activity, it is more likely to leak and you are more likely to take it off or create a by-pass

What you can do:

  1. Wear a mask whenever you are around other people not in your immediate circle. This is especially true indoors, and in crowded situations. Outdoors the risk of transmission is lower, but wearing a mask still makes things safer.
  2. Make sure your mask fits well. This is one of the most important things under your control.
  3. Use a better mask than just a bandana or neck-gaiter. Unless you are highly exposed (e.g. a medical worker), consistently wearing a decent, well-fitting mask is more important than worrying about every detail of mask construction.