Evidence

Proof that it works

It works, and the proof has moved out of the laboratory and into spaces. 222nm light inactivates airborne coronavirus, flu and norovirus, bacteria in the air and on surfaces, and even bacterial spores, at doses inside the human exposure limits. A few ceiling units in a working room cut airborne norovirus by 99.8%. In a busy dental clinic, 222nm light cut airborne bacteria by about 80%, the same as adding six and a half air changes an hour. Below are the main studies, what they tested and what they found.

In the laboratory

Columbia University, 2018. Airborne H1N1 flu in a chamber built to mimic coughing and breathing. 2 mJ/cm² of 222nm inactivated more than 95% of it. Read the paper.

Columbia University, 2020. Two airborne human coronaviruses. 1.2 and 1.7 mJ/cm² inactivated 99.9%. The authors worked out that at the exposure limit for an occupied room, roughly 3 mJ/cm² an hour, you would see about 90% inactivation in 8 minutes, 99% in 16 and 99.9% in 25. Read the paper.

Hiroshima University, 2020. The COVID virus itself, dried on plastic. 1 mJ/cm² (ten seconds under the lamp) removed 88.5% of it; 3 mJ/cm² (thirty seconds) removed 99.7%, to below the level the test could detect. Read the paper.

St Andrews, Leeds and Dundee, 2022. Airborne Staphylococcus aureus in a room sized chamber, with the lamps set inside the American exposure limit. 98% gone in under five minutes. They chose S. aureus because it is harder to kill than flu or coronavirus, so the result is a cautious one. Read the paper.

Columbia University, 2017. MRSA killed efficiently, with almost no DNA damage in a model of human skin exposed to the same light.

Korea, 2023. A broad range of bacteria, gram positive and gram negative, plus SARS-CoV-2, all inactivated by 222nm.

Norovirus on surfaces, 2025. Four strains of human norovirus on steel and skin. 222nm worked well on wet virus; it worked less well when the virus had dried or was mixed with vomit, which shields it. The practical lesson is that Far-UVC cuts the airborne and surface load all day, and you still clean up after an incident.

In spaces

Columbia University, 2024. Four 222nm ceiling fixtures in a room where staff cleaned animal cages all day, with the lamps inside the regulatory limits. Airborne infectious norovirus down by 99.8%. The first study to show Far-UVC working against a pathogen in an occupied room. Read the paper.

Columbia University, 2025, Journal of Hospital Infection. A working dental clinic that already had ten air changes an hour of filtered ventilation. 222nm cut airborne bacteria by about 80% on top of that, equivalent to another six and a half air changes. In a room with a typical three air changes, the same lamps would cut the load by around 92%.

Denmark, 2025. Ceiling mounted 222nm fixtures in occupied clinical areas significantly reduced the bacteria on surfaces.

Denmark, 2026. Twelve care homes fitted 222nm units in common areas and residents' rooms. Over six months, antibiotic prescribing fell compared with the homes that did not. Not a randomised trial, but a real result in a real setting.

Taiwan, 2025. In a randomised trial, a 222nm device used on pressure sores in intensive care reduced bacteria in the wound and improved healing.

Japan, 2023 and 2024. Lower bacterial contamination of endoscopists during procedures; germicidal effect on dental spittoons.

Systematic review, 2025, Journal of Hospital Infection. Pooling the published studies, 222nm disinfected a wider range of pathogens than 254nm and with a better safety record.

What this means in your area

At the dose we use with people present, a unit is doing the work of several air changes an hour, continuously, without moving any air. It carries on when the doors are propped open, when the air handling is recirculating and when the ventilation is being serviced. The dental clinic figure of 6.5 air changes was measured in a room with excellent ventilation already; in a restaurant, bar or crew mess running at two or three air changes, the difference it makes is bigger.