Researchers at Duke University tested the effectiveness of 14 commonly available masks in a proof-of-concept study. A low-cost laser device was used to demonstrate that a simple optical measurement method could effectively determine the efficacy of masks to reduce the transmission of respiratory droplets during regular speech.  Due to the recent COVID-19 pandemic, worsened by global shortage of commercial supplies, widespread use of homemade masks and mask alternatives have been used by the public.

Unsurprisingly, the study found that surgical masks and N95 respirators without valves worked the best at blocking respiratory droplets from projecting into the air when a person talks.  However, a variety of double-layer polypropylene and cotton masks also reduced a significant amount of spray from normal speech. Least effective, were bandannas, knitted masks and neck fleeces (“gaiter masks”). Neck fleeces actually dispersed more spray into the air because the material broke down larger respiratory droplets into smaller particles.  View Full Study (PDF)



Masks Ranked From Most Effective to Least Effective

1. Fitted N95, no valve (14 in photo)
2. 3-layer surgical mask (1)
3. Cotton-polypropylene-cotton mask (5)
4. 2-layer polypropylene apron mask (4)
5. 2-layer cotton, pleated style mask (13)
6. 2-layer cotton, pleated style mask (7)
7. Valved N95 mask (2)
8. 2-layer cotton, Olson style mask (8)
9. 1-layer Maxima AT mask (6)
10. 1-layer cotton, pleated style mask (10)
11. 2-layer cotton, pleated style mask (9)
12. Knitted mask (3)
13. Double-layer bandanna (12)
14. Gaiter type neck fleece ( 1)

In the experimental setup, a lens turns an optical laser into a sheet of light. This light sheet, shone through a dark enclosure (made up of cardboard sheets and duct tape), reveals when droplets pass through it, with the results being filmed by a mobile phone camera.


Source: Low-cost measurement of facemask efficacy for filtering expelled droplets during speech, Science Advances 07 Aug 2020

Mask Image Courtesy Emma Fischer, Duke University