Roko Mijic, Alexey Turchin

Epistemic status: Many different uncertainties here, but the idea has some good evidence in favor of it and a high potential payoff.

Tl;dr: We should urgently investigate putting special human-safe Far-UVC lamps all over our built environment to ‘kill’ virus particles whilst they are in the air, thereby vastly reducing covid-19 spread.

Inspired by: https://www.nature.com/articles/s41598-018-21058-w


One of the most promising and neglected ideas for combating the spread of covid-19 is the use of ubiquitous ultraviolet light in our built environment (trains, offices, hospitals, etc). Ultraviolet light is already being used as a disinfecting agent across the world; it goes by the acronym UVGI - “Ultraviolet germicidal irradiation”. The energetic photons of UVC light break chemical bonds in DNA and kill/inactivate both viruses and bacteria.

Ultraviolet light on earth exists on a spectrum between 200nm and 400nm. Light above 400nm is blue visible light. Light below 200nm is called “vacuum UV” because it is strongly absorbed by the oxygen in ordinary air and therefore cannot exist except in a vacuum or some other non-air medium. Within the 200-400nm range we have UVA, UVB and UVC, and at the short-wave edge of the UVC band we have “Far-UVC”, from roughly 200nm–220 nm.

Safety considerations

Human beings are also vulnerable to UV radiation. It causes skin cancer and serious eye damage.

However, recent research suggests that the Far-UVC band is actually safe for human skin because it cannot penetrate through the thin layer of dead skin cells on the surface of our skin.

This means that it might be possible to mount a long-term response to covid and other pathogens by constantly illuminating our built environment with light from specifically the Far-UVC band. If the Far-UVC light is indeed safe for humans, the Far-UVC could be on at all times and could destroy or deactivate viral particles before they can spread from person to person.

Why hasn’t this already been considered by relevant authorities? Far-UVC appears in a literature review by WHO, but it is not currently being acted upon as the amount of evidence in favor of safety and efficacy is small.

There is some uncertainty about whether Ozone generation by this band (200nm-220nm) would be problematic. Ozone is not great for your health. However, it seems to be the case that the 200-220nm band is not a strong producer of Ozone. In addition, UV degradation of surfaces might result from chronic UV exposure.

Balancing harms of action and inaction

Even if Far-UVC is somewhat harmful it might still be a good idea to implement. Small harms from Far-UVC light might be much less bad than large harms from covid-19, or from the economic damage caused by the lockdown which one author estimates to be roughly $10 million per minute, plus much personal hardship which will be caused by the forthcoming recession.

Furthermore, UV light is easier to defend a person against than a virus. Sun-creams, clothing and eyewear that defend against UVC may be less bad than a semi-permanent lockdown or an exponentially growing covid-19 outbreak that results in millions or tens of millions of deaths. UV in the built environment could even be managed intelligently - computer vision could identify where the people were and turn on UV lights only in unoccupied areas, though such a project would at best be ready by the start of 2021 (and then only with wartime levels of effort and purpose).

If the safety claims of Far-UVC are partially true rather than fully true, a combination of using Far-UVC with physical protection like eyewear may still cause only acceptable losses to cancer and eye damage. In the longer term, such “almost safe” Far-UVC could be combined with intelligent management at various levels of granularity; imagine a lift that is bathed in Far-UVC every time people leave it, or “walls” of Far-UVC separating people that automatically turn off momentarily when a person walks through them. The ultimate system might even adjust the power of the Far-UVC using AI.

Epidemiological considerations

Even an ideal Far-UVC solution that was harmless to humans, 100% lethal to covid-19 particles and easy to deploy at scale might not be sufficient to reduce R to exactly 0. But the key question is whether it could reduce R below 1 whilst also allowing most economic activity. An easy preliminary experiment to run would be to put virus samples in mouse cages - perhaps in aerosolized form - treat some cages with Far-UVC, leave other cages alone, see if infection rates go down in the treated cages.

This is an important source of uncertainty and further research is needed.

Scaleup considerations

Even a perfect system is useless if it cannot be scaled up and implemented across the globe. Far-UVC can be produced from Krypton Chloride (Kr-Cl) Excimer Lamps, but modern Aluminium Nitride (AlN) Far-UVC LEDs are a better solution for the long term. In the even longer term, collimated Far-UVC could be produced by lasers. This is an important source of uncertainty and further research and expert input is needed.

Power considerations:

The amount of Far-UVC energy required to kill 99% of the viral particles is estimated to be around 20J/m^2. With a power of, say 5W/m^2, a system would need 4 seconds to mostly sterilize a viral aerosol that could travel from person to person. However a lower power system would still have some benefits - we know that people can be infected by air that was contaminated 30 minutes earlier. Higher power in these wavelengths could be difficult to achieve with Kr-Cl Excimer Lamps as the overall efficiency from electricity to Far-UVC is ~10%. AlN Far-UVC LEDs would likely have a much higher conversion efficiency.

Generality

One of the greatest benefits of Far-UVC is that it would be a very general weapon against pathogens. Far-UVC kills/deactivates bacteria, viruses and other pathogens. MRSA, C-DIFF, influenza, etc are all killed by UVC, as is the next problematic pathogen, whatever it is.

Summary

There are many different reasons that Ubiquitous Far-UVC might not work out, but if it did work out it could have huge benefits. For this reason the authors believe that it should get more attention at this critical time. Scaleup and safety and efficacy trials must all be carried out as quickly as possible, preferably in parallel. More importantly, the idea needs more attention from experts in the relevant fields - UV physics, epidemiology, and people who study the etiology of skin cancers. As of writing there are reports that the US government estimates the epidemic could last for 18 months, so a plan like Far-UVC that will take months to implement may still be a critical component of a response later this year.

Appendix. Other ways to use UV light to fight coronavirus

One of the explanations of the flu and other infections seasonality is that the Sun's UV kills viruses. However, people spend a lot of time indoors even during summer, and especially during self-isolation. Most of our infections are happening indoors: at home, in transport and in working places. UV from Sun could be part of the explanation of the lower instances of coronavirus in southern countries.

If we replace light bulbs everywhere with light sources which also emit UV light of some special wavelength, we will kill most of the airborne viruses and will clean fomites. Thus, we will create artificial summer everywhere and will lower R0 of coronavirus below 1.

The main obstacles are the duration of exposition and possible harm to people. Recently in Moscow 20 children had burns in their eyes after a school teacher forgot to turn off the UV cleaner in the classroom.

There are several other ideas, besides Far-UVC light, to prevent human eye and skin damage:

1) Intelligently controlled UV lighting. UV light source turns on the maximum level when there are no people in the room. We already have motion detectors for lighting, but here they will work in reverse. Light with motion detection could also direct light in directions, where there is no motion, so no people. On the video, one can see UV light sources on sale with motion detectors:

The power of light could be temporarily increased after the sound of sneezing. But it will make all the system more complex and its large-scale implementation will take longer time. If Krypton Chloride Excimer bulbs are used, their lifetime is not great, so they can’t run constantly. But if we can get the Aluminium Nitride LEDS then lifetime and efficiency will be better.

2) Not “too strong” sources of UV, which are producing Sun’s intensity of UV and which act mostly on fomites. As we know, humans can survive at least 1 hour of sunlight UV exposure without strong damage (on beaches). We could use it as a reference point to calibrating UV sources.

3) Strong UV lighting + gloves. Everyone will wear gloves, masks and glasses outside. In that case, no skin will be exposed to the UV lighting (and to viruses). Wearing PPE will be effective anyway. Women in the East are wearing full cover clothes, and they are ok.

4) Wearable headlight UV will direct UV light in the opposite direction to the person’s eyes but will cover everything he inhales or touches, as well as his hands. The light will be strongest near the human face (but not affecting the face), and will attack droplets which the person is about to inhale. However, the light will dissipate in the distance of 1- 2 meters to safer levels. UV headlamps already exist and on sale, but may be not strong enough for disinfection. It will be especially effective if wearables Far UVC light sources will be used.

5) UV flashlight - Torch that emits UV radiation in a wide beam. Runs off main power. Could be used by cleaners as an additional step when cleaning surfaces.

Pros:

  1. Simpler, easier, cheaper and faster to build than other solutions
  2. Less harm to people, as UV light can be directed, and is not always on
  3. Proving ground (an MVP, in startup terms) for more advanced implementations
  4. Mobile; could be used in multiple locations

Cons:

  1. Less effective than always on UV lights and lamps
  2. Requires additional time/effort on top of normal cleaning routines

Artificial light exists currently almost everywhere, where contemporary humans live: in homes, in any shop, in cars and even on the streets. All we need is to replace electric lamps. Large amounts of lamps could be manufactured in 0.5-1 year, and smaller amounts for critical places like elevators in the even shorter notice.

However, there is a problem of actual testing the technology until it will be approved as safe and effective by the FDA. It is technically difficult to make deep UV (220nm) light-emitting diodes.

A good start will be to put UV lights in the places of short use: elevators, shops, restrooms.

It is much more convenient to wear protection against light than protection against viruses, and after a few months of lockdown, the idea of returning to almost normal life but with sun cream and/or gloves will be quite nice.

References

Welch, D., Buonanno, M., Grilj, V. et al. Far-UVC light: A new tool to control the spread of airborne-mediated microbial diseases. Sci Rep 8, 2752 (2018). https://doi.org/10.1038/s41598-018-21058-w

Narita K, Asano K, Morimoto Y, Igarashi T, Nakane A (2018) Chronic irradiation with 222- nm UVC light induces neither DNA damage nor epidermal lesions in mouse skin, even at high doses. PLoS ONE 13(7): e0201259. https://doi.org/ 10.1371/journal.pone.0201259

Willie Taylor, Emily Camilleri, D. Levi Craft, George Korza, Maria Rocha Granados, Jaliyah Peterson, Renata Szczpaniak, Sandra K. Weller, Ralf Moeller, Thierry Douki, Wendy W.K. Mok, Peter Setlow DNA damage Kills Bacterial Spores and Cells Exposed to 222 nm UV Radiation Applied and Environmental Microbiology Feb 2020, AEM.03039-19; DOI: 10.1128/AEM.03039-19

Colorado company uses UV lighting technology to kill 99.9 percent of bacteria and viruses. Fox Denver, 7 Macrh 2020

New to LessWrong?

New Comment


39 comments, sorted by Click to highlight new comments since:

I think understanding all three of the following papers (which are predecessors to the 2018 Nature paper) is important to guessing the efficacy and safety of the intervention. I'll add to this comment as I gain more understanding.

EDIT: I found thinking about this very difficult, so I made progress very slowly and have mostly stopped. Here's my current story.

The biophysical reason to believe that 207nm light preferentially harms viruses and bacteria over mammalian cells:

Mammalian cells are big, like >50 wavelengths long, while bacteria and viruses are <5 wavelengths long. 207nm light activates certain peptide bonds that are found in ~ all proteins. So energy transfers at a large constant rate, let's call it lambda (I feel pretty confident about the argument if lambda > 10% per wavelength. I think it can be estimated from the papers but I just haven't tried hard enough) from light to organic material. When you shine light at a virus, you get something like lambda * (1 - lambda)^3 fraction of the energy deposited in the squishy RNA bits. But when you shine it at a mammal cell, you get like lambda * (1 - lambda)^30 fraction of the energy in the squishy DNA bits.

Thing that's empirically shown in these papers:

There is a dosage at which 207nm is clearly quite toxic (here toxicity means "causes cell death / virus inactivation") to bacteria/viruses but is clearly not toxic to mammalian cells.

Thing that's not empirically shown in any paper that I know of:

There is a power level of 207nm such that constant exposure keeps a space disinfected but which does not cause cancer in humans.

An incomplete argument for the above claim:

  1. Cell-killing and cancer-causing from radiation are both caused by excess energy induced into the nucleus of mammal cells, in such a way that total risk scales linearly with the integral over time of power of radiation exposure.
  2. The toxicity study yields an upper bound on how much radiation power makes it to the nucleus given a certain power of exposure to the cell.
  3. The cancer literature yields "safe exposure" levels that can be cast in terms of "radiation power that reaches the nucleus".
  4. Combining 2 and 3 gives a safe exposure power level for 207nm lamps, and I conjecture (or maybe just hope?) that this level is greater than the level used in the toxicity study. That level of power is already shown effective at killing viruses and bacteria, and does so in a way that only depends on them being small and made of replicating nucleic acid.

It would be also in interesting to look at visual light as possible virus killer (as visual light is known to be safe and there are many its sources). There was a claim that it kills viruses in blood (but in pulses). Visual light also kills bacteria. But what about surfaces?

Sun also helped cure patients in 1918 - maybe by killing viruses?

thanks!

I have a handheld blue laser ..111A 450 nm .I have tried to find out if will kill the covid 19 virus on things like paper money,door handles and food etc instead of using sanitizers .I do not know where to turn.Of course I have to leave the house for food pickup .getting the mail.or even taking the dog out.I live in an appt so its easy to get contaminated and i am 73 years old .Please steer me in the right direction

450nm blue light will not kill viruses in any reasonable time & power level.

Also, worry much less about fomites and much more about breathing the virus in.

I didn't realize at first that this was cross-posted from EAF, so since this seems to be getting more attention here I'll repost my comment from there here:

You don't mention this, and maybe there is no research on it, but do we expect there to be much opportunity for resistance effects, similar to what we see with antibiotics and the evolution of resistant strains?

For example, would the deployment of large amounts of far-ultraviolet lamps result in selection pressures on microbes to become resistant to them? I think it's not clear, since for example we don't seem to see lots of heat resistant microbes evolving (outside of places like thermal vents) even though we regularly use high heat to kill them.

And even if it did would it be worth the tradeoff? For example, I think even if we knew about the possibility of antibiotic resistance bacteria when penicillin was created, we would still have used penicillin extensively because it was able to cure so many diseases and increase human welfare, although we might have done it with greater care about protocols and their enforcement, so with hindsight maybe we would do something similar here with far-ultraviolet light if we used it.

Some bacterial species have impressive level of radiation resistance, so it's a quality that can certainly be evolved much like antibiotic resistance. The most extreme radiation-resistance strategies involve the presence of multiple genomes within a cell as "backups", this is coupled with various metabolic activities which have the goal to replace/repair the "main one" when it is damaged by radiation. Viruses, on the other hand, completely lack a metabolism and therefore will have serious trouble developing this kind of resistance.

I agree with this, plus note that UVC is actually more dangerous than radiation for bacteria, because the number of photons per watt is significantly greater and the energy per photon is already enough to do damage.

Is number of photons greater or energy of each photon?

the number of photons per second per watt of power is a lot greater for UVC than for gamma radiation.

If the number of photons per Joule* is higher for UVC, that means that each photon carries less energy.

*A Watt is a Joule per second; sustaining a Watt of power for one second requires a Joule of energy.

I am not an expert on viral and bacterial resistance to UV radiation. Some googling reveals that bacteria can evolve partial resistance to some forms of UV. In th following paper, the bacteria survived roughly a minute of UVB exposure from a 15W source: https://onlinelibrary.wiley.com/doi/full/10.1111/j.1558-5646.2011.01438.x

But I don't think they can evolve to be totally resistant, i.e. just sit under a UV light at 250nm or 220nm indefinitely. UV is fundamentally harmful to bacteria because their germ line cannot be protected from it, kind of like how humans are fundamentally vulnerable to gamma rays.

The same goes for viruses, except they also cannot repair themselves and they're even smaller which makes them more vulnerable.

I think that for viruses it will be difficult to become completely radiation resistant, as it would require complete overhaul of their makeup: thicker walls, stronger self-repair.

Three points:

1) Any risk-benefit calculation should consider that COVID-19 appears to be only minimally harmful to the young and healthy. Deep UV can cause skin cancer at any age, which seems like a good reason to be careful here. Human DNA is not fundamentally different than virus DNA where UV light is concerned.

2) It might be safer in Africa or India where melanin protects the local populations from UV. Yes, that's what melanin is for.

3) Can we do the same thing, but safer, with a detergent mist?

The idea of Far-UVC is that it is supposed to be safe due to the dead skin layer. Until we know for sure it's difficult to make the tradeoffs.


re (2), I suspect that melanin is not effective against UVC.

re (3), it's an interesting line of thought, but my suspicion is no.

1) 254 nm is the same part of the spectrum that the ozone layer protects us from, and also the absorption peak for DNA. IIRC non-aquatic life didn't take off until the ozone layer formed, probably for that reason. UV does bad things to DNA, and I wouldn't bet on dead skin providing adequate protection from large exposures. The DNA spectrum has another peak around 212 nm (although the dropoff at low energy may be from the limitations of the optics rather than from lower absorption by the molecule).

2) The absorption spectrum for melanin shows its strongest peak between 270 and 220 nm. I'd say it depends on whether the dropoff below 220 nm was due to the absorption characteristics of the molecule or was an artifact of the measurement (spectrometer, cuvette, solvent, etc.). I'm inclined to guess artifact; such a sharp dropoff toward higher energy doesn't make physical sense (unless I'm missing something).

As I said, Far-UVC from 200-220nm is supposed to be safe to humans because of our layer of dead skin. This is kind of the whole point of the post and is in the references.

254nm has less energy per photon but it penetrates further through skin, meaning that 254nm is definitely dangerous.

I was curious about how much we could rely on that safety, and it turns out there are threshold limit values (see the sixth slide) for UV-C. Between 200 and 220 nm the TLVs are .02 to .08 J/cm^2 (200 to 800 J/m^2), according to the American Conference of Governmental Industrial Hygienists. At 5W/m^2 (your suggested irradiation) that gives you 40 to 160 seconds of reasonably safe human exposure.

Apparently those slides contradict the studies cited in my article.

I don't know which to believe. The rational action is to urgently run more experiments to assess the risk from 200nm-220nm band.

It would be really great if those slides had a references section.

I spent a while looking into safety for this idea, and far-UVC has a ways to go before it's established safe enough for public settings---it's only been shown to be non-carcinogenic for one of two relevant DNA damage mechanisms. I've posted a more in depth discussion here.

(I added Roko to the metadata as a coauthor. Tagged coauthors are a beta-feature which currently can only be edited by moderators, since we haven't implemented the mechanics of having authors approve each other.)

Could you give Roko editing rights to this post?

Unfortunately the LW codebase doesn't support multiple authors with edit access to the same post yet; we're working on this (as part of a broader overhaul of the post editor, which will allow Google Docs-style simultaneous editing), but it isn't ready yet. In the mean time, the easiest way to handle this is to make edits in a shared Google Doc, and have the primary author paste them in and save. (Copy-paste between the post editor and Google Docs should just work.)

We did it, but there is a formating problem with content menu of the left - all items should be equal, no sub-stections.

The indentation of the table of contents is determined by the heading levels. It looks like you may have set some of the headings to "Heading 2" style and others to "Heading 3" style in Docs, then adjusted the font size to make them look the same. If you use Docs' heading format presets and use the same one for all the sections, they should be at the same indent level.

I am merely a layman that wishes to pose a question. Could LED driven UV-C fiber-optics be used internally in the throat and lungs to kill the virus in the infected? Is it possible to manage it so that harm to human tissue and risk of cancer would be minimized?

no, definitely not. UVC is highly carcinogenic, and furthermore your lungs are a fractal which makes it impossible to shine light into all of the surface.

Typo thread: the sentence "In addition, UV degradation of surfaces might result from chronic UV exposure." has a hyperlink with an extra character at the end.

It doesn't sound like this would require much in the way of coordination. That makes me a bit more hopeful about it than most options. Less room for a tragedy of the commons. Once demonstrated safe and effective, individuals and businesses could deploy UV lighting and derive benefit from it, without worrying about whether their investment will be wasted by the inaction of others.

Yes, it is something that one nation could develop and then others could pick it up once it was shown to work.

How about something with the form-factor of a mask with UV-LEDs inside? Oh wait, but then you might as well just wear a normal mask. :-P And it wouldn't protect eyes...unless it also emitted a sheet of light upward.

OK try again: Take your "wearable headlight" proposal, but with the form-factor of a baseball cap, emitting a narrow sheet of light downward from the rim, a couple inches in front of the face. So it wouldn't really hit other people, although it might get your hands and legs a bit sometimes. (And yes there are lenses to collimate UVC—e.g. UVFS—although they obviously won't be as cheap as molded plastic lenses.) This is only helpful if aerosol transmission is significant, which I don't know either way. :)

Actually, it could be combined with full face mask, which also covers eyes. Such mask also could include wearable display, so no need in smartphones. If Apple will design it, it will look nice.

This topic got extensive coverage in the latter half of Making Sense (Sam Harris) open-access podcast titled 'Engineering the Apocalypse'.

My main takeaways were:

  • It's (too) early (as this post also highlights regarding safety)
  • The technology is being tested in LEDs (cheaper) but is quite far away from commercial use
  • It could be a dud, but the ROI would be amazing (millions for billions)

Starting at 3:11

Roko, kudos.

I'm not familiar with your background but your post is one of the most inclusive among many professional ones. Being an expert of CIE in this field and having numerous patents on optics let me humbly add a number of short comments.

  1. Harmful effect of UVC is a common place, but it is greatly exaggerated as a result of misinterpreted raw data which generally assume much larger irradiated areas of our body. And in any case they look as laughing stock compared to other radiation treatments.
  2. UVC is the only method to kill 99.9% of any Virus, given a sufficiently long irradiation time.
  3. Quoted by you figure of merit is more typical for FOMITES (surfaces) rather then droplets.
  4. Just imagine that we have a compact wearable Headmouted system concentrating all emitted light only on areas of your mouth-nose of say 10cm^2 and is absorbed there. Thus principally it is possible to completely "sterilize both you and a user (two-way action) with say 100 mW lamp with less than 0.1s of through-flow time.
  5. I'm close to form a consortium and start production shortly.

I am also a layman wishing to ask a question. Could/would far ultra violet light exposed to the blood of a patient with Covid -19 possibly weaken or kill the virus without damaging blood cells ... etc ?

Bob D

No, this Far-UVC stuff is very nasty. It is only ok for us on the outside because our skin stops it.

Exposing your blood to large amounts of Far-UVC is almost certainly going to lead to cancer IMO.

It looks like it can't go through water, so it can't reach virus in blood. However, it seems that visual light is beneficial for immunity, and during 1918 flu sun was used as one of the therapies