What Is UV Disinfection and How Does It Actually Work?

UV disinfection uses ultraviolet light to kill or inactivate bacteria, viruses, fungi, and other microorganisms by damaging the molecular machinery they need to reproduce. The core of the process is photochemistry: UV photons in the germicidal range (roughly 200 to 300 nm) are absorbed by the nucleic acids and proteins inside a pathogen, creating structural damage that the organism either cannot repair or cannot repair fast enough to survive. It is used today to treat drinking water, sterilize hospital surfaces, clean indoor air, and process food, all without adding chemicals. But the details of how it works, what it works well against, and where it falls short are more interesting than the one-line explanation suggests.

How UV Light Destroys Microbes at the Molecular Level

The primary target of germicidal UV is the genetic material inside a microorganism. When UV photons in the 250-to-270 nm range strike DNA or RNA, they cause adjacent pyrimidine bases (the “letters” of the genetic code) to fuse together into abnormal structures called dimers. These dimers act like roadblocks: they prevent the cell’s replication machinery from reading the genetic instructions correctly. If enough dimers accumulate, the organism loses its ability to reproduce, and it effectively dies even if the cell membrane is still physically intact.¹

Genetic damage is the main event, but it is not the only one. UV light also damages proteins, particularly at wavelengths below 240 nm and around 280 nm where aromatic amino acids absorb strongly. For viruses, this matters because the protein coat (capsid) is what a virus uses to latch onto and enter a host cell. Damage the capsid proteins enough, and the virus can no longer infect even though its genome might be partially intact.² Research on herpes simplex virus, for instance, showed that different UV wavelengths attack different viral components: 260 and 280 nm UV caused both DNA photoproducts and protein degradation, while longer wavelengths like 310 nm mainly damaged proteins without strongly affecting the genome.³ Studies on adenovirus similarly confirmed that UV-induced protein damage occurs across the entire germicidal spectrum, from 200 to 300 nm.⁴

Why Some Pathogens Need Far More UV Than Others

One of the most practical things to understand about UV disinfection is that different organisms vary enormously in how much UV dose they require. The dose, usually expressed as fluence in millijoules per square centimeter, combines the intensity of the light with the exposure time. A low dose might eliminate most common bacteria, but the same dose could barely dent a tough virus or a protozoan cyst.

Viruses as a group tend to require higher UV doses than vegetative bacteria. Among viruses, double-stranded DNA viruses like adenovirus are exceptionally resistant, needing around 125 mJ/cm² to achieve a 99.9% (3-log) reduction in drinking water treatment. That is many times the dose needed for the same level of kill against common bacteria like E. coli.⁵ Bacterial spores and protozoa (like Cryptosporidium) also sit on the tougher end of the spectrum, with protozoa generally requiring higher UV doses than bacterial spores.⁶

This variation is why UV disinfection systems are engineered for specific target organisms. A municipal water plant designed to handle Cryptosporidium delivers a different dose than a hospital surface-disinfection robot designed to kill MRSA. If someone tells you a UV gadget “kills 99.9% of germs,” the meaningful follow-up question is: which germs, and at what dose?

UV Disinfection in Water Treatment

Water treatment is the oldest and most widespread application of UV disinfection. The principle is straightforward: water flows past UV lamps in a sealed chamber, and every organism that absorbs enough photons during its time in the chamber is inactivated. Unlike chlorine, UV leaves no chemical residual in the water and produces no disinfection by-products like trihalomethanes, which are a long-standing concern with chlorine-based treatment.⁷

The flip side of leaving no residual is that UV cannot protect water after it leaves the treatment chamber. Chlorine, for all its drawbacks, keeps working as water travels through miles of pipe. UV-treated water can become recontaminated if the distribution system has a leak or biofilm. That is why many municipal systems use UV as one step alongside a low dose of chemical disinfectant rather than as a standalone treatment.

Water clarity is a major limiting factor. Turbidity, the cloudiness caused by suspended particles, shields microbes from UV photons. Research on agricultural water found that UV-C treatment significantly reduced E. coli at turbidity levels up to about 23 NTU, but as turbidity dropped from roughly 23 to 11 NTU, the additional microbial reduction was dramatic, improving by more than 2 log units.⁸ Even low concentrations of suspended solids can begin to interfere with UV disinfection performance.⁹ This is why pre-filtration is standard practice before UV treatment: the cleaner the water, the more effectively the light can reach every organism in it.

How UV Cleans Indoor Air

Air disinfection using UV takes a fundamentally different approach than water treatment because you cannot force all the air in a room through a narrow chamber the way you can with water flowing through a pipe. Two main strategies exist: in-duct systems that irradiate air inside HVAC ductwork, and upper-room systems that bathe the air near the ceiling of an occupied space while keeping the lower, occupied zone safe.

Upper-room UV germicidal irradiation is a well-studied approach. Fixtures are mounted high on walls or ceilings, aimed horizontally or upward so that the germicidal zone stays above head height. As natural air currents and mechanical ventilation push contaminated air upward, it passes through the UV zone and gets treated. Research has found that this method can be more effective for air disinfection than mechanical ventilation alone, at lower installation and operating costs.¹⁰

Mixing fans make a real difference. Testing with Serratia marcescens (a common test bacterium for air studies) showed that adding a ceiling fan to an upper-room UV setup boosted kill rates from roughly 46% to 62% at low ventilation rates, and from about 53% to 86% at higher ventilation rates.¹¹ The logic is simple: faster air mixing brings more contaminated air into the UV zone per unit time. Research modeling classroom environments confirmed that when room air speed exceeds about 0.1 m/s, upper-room UV systems can achieve roughly 90% disinfection of aerosol particles in the 1 to 10 micrometer range.¹² The takeaway for building managers is that installing UV lamps without thinking about airflow patterns wastes most of the system’s potential.

Surface Disinfection and Its Stubborn Limitations

UV surface disinfection is the application most people encounter through consumer products and hospital “UV robots.” The appeal is obvious: wheel a device into a room, turn it on, and the light kills pathogens on every visible surface. The reality is more complicated than the marketing suggests.

UV light travels in straight lines. It cannot bend around obstacles, reach inside crevices, or treat the underside of an object facing away from the lamp. This shadowing problem is the single biggest practical limitation of surface UV. Anything blocked from direct line of sight to the UV source receives little or no germicidal dose. Surface roughness compounds the problem: irregular textures create tiny shadows and increase the effective surface area that needs to be irradiated, diluting the dose across a larger area.¹³

Distance matters too. UV intensity follows the inverse-square law: double the distance from the lamp, and the intensity drops to a quarter. A UV device effective at one meter may deliver an inadequate dose at three meters. Organic soil or dirt on surfaces can also absorb UV before it reaches the microbes underneath, which is why hospital protocols typically call for manual cleaning before UV treatment rather than using UV as a substitute for wiping down surfaces.

Reflected UV introduces another consideration. While reflection can sometimes help by bouncing light into partially shadowed areas, it also creates unintended human exposure risks and can generate small amounts of ozone depending on the wavelength and materials involved.¹⁴

Humidity and Other Environmental Factors That Weaken UV

UV disinfection performance in air is surprisingly sensitive to humidity. Classic experiments found a sharp decline in the fraction of airborne organisms killed at relative humidity above 60 to 70%, with evidence of UV-induced microbial reactivation above 80% relative humidity.¹⁵ More recent research has confirmed this general pattern while adding nuance: UV rate constants tend to peak around 40% relative humidity, with performance declining at both lower and higher humidity levels. Larger bioaerosol particles in very humid environments need higher UV doses to achieve the same kill.¹⁶

The mechanism behind this humidity effect is not fully settled, but the leading explanation is that water molecules surrounding airborne bacteria at high humidity can absorb some UV energy before it reaches the DNA, and may also facilitate enzymatic repair processes. For anyone designing or evaluating an air-disinfection system, this means that UV performance can vary significantly between a dry winter and a muggy summer in the same building, unless the system is sized with high-humidity conditions in mind.

Far-UVC and the Promise of Occupied-Space Disinfection

Conventional germicidal UV at 254 nm is effective but dangerous to human skin and eyes, which is why traditional systems are designed to keep people out of the UV zone. Far-UVC light, centered around 222 nm from filtered krypton chloride excimer lamps, has emerged as a potential game-changer because it appears to kill pathogens just as efficiently while posing far less risk to people.

The biophysical reasoning is straightforward. At 222 nm, photons are absorbed so intensely by the proteins in the outermost dead layer of skin (the stratum corneum) that very little energy penetrates to the living cells beneath. Research using 3D human skin models found that 222 nm light killed methicillin-resistant Staphylococcus aureus (MRSA) efficiently but produced almost no premutagenic DNA lesions in the skin, unlike conventional 254 nm lamps.¹⁷ A study using human skin biopsies confirmed that 222 nm induces only minor DNA damage at high doses, particularly in lighter skin types with less melanin.¹⁸

The picture is not perfectly clean, though. Compared to 254 nm light, 222 nm is absorbed more intensely by proteins (especially aromatic side chains), causing photooxidation and crosslinking in biological materials it hits. While inflammatory and reactive oxygen species responses seem less pronounced with 222 nm, some researchers have flagged concerns about alterations in skin regeneration pathways under chronic exposure.¹⁹ Far-UVC technology is moving toward deployment in public spaces like airports and hospitals, but long-term safety data from real-world chronic exposure is still accumulating. The optimism is warranted; the certainty is not quite there yet.

Mercury Lamps Versus UV LEDs

For decades, the workhorse of UV disinfection has been the low-pressure mercury vapor lamp, which emits almost all its UV output at 253.7 nm, close to the peak germicidal wavelength. These lamps are cheap, well understood, and efficient at converting electrical energy to UV. But mercury is toxic, the lamps are fragile, they take minutes to warm up, and their disposal is an environmental concern.

UV-C LEDs are the rising alternative. They can be tuned to specific wavelengths, they turn on and off instantly, they are compact, and they contain no mercury. In disinfection tests using real wastewater, UV-C LEDs achieved pathogen removal rate constants about 1.4 times higher than conventional mercury lamps.²⁰ The trade-off, for now, is economics and environmental footprint at scale. For systems running continuously, mercury lamps still come out ahead on cost and overall environmental impact because LED manufacturing is energy-intensive and LED lifespans, while improving, have not yet matched the total photon output per dollar of mercury technology.

Where LEDs shine (literally) is in intermittent use. Because mercury lamps degrade faster with frequent on-off cycling while LEDs are unaffected by it, installations that do not run around the clock, like point-of-use water purifiers, portable disinfection devices, and on-demand surface treatment, favor LEDs.²⁰ Life cycle assessments have also shown that under certain operating conditions, LED systems can reduce environmental impact by roughly 25% across most impact categories compared to mercury systems, with even larger reductions in categories like mineral resource use.²¹

Can Bacteria Become Resistant to UV?

With antibiotic resistance dominating public health conversations, a reasonable question is whether microbes could eventually evolve resistance to UV disinfection the way they have to drugs. The short answer from the current evidence is: not in any practical sense, at least not through the kind of repeated low-level exposure that occurs in hospital and water treatment settings.

A study exposed three epidemiologically important multidrug-resistant organisms, including MRSA and two strains of carbapenem-resistant Klebsiella pneumoniae, to 25 serial rounds of UV disinfection using both xenon and mercury UV sources. Colony counts stayed low throughout, and whole-genome sequencing afterward showed no significant genetic changes that would indicate developing resistance.²² The researchers concluded that UV disinfection is unlikely to generate UV-resistant hospital flora.

That said, the picture is different under sustained evolutionary pressure over hundreds of generations. Laboratory experiments propagating E. coli for 600 generations under UV selection did find that all UV-selected lines became more UV-resistant, with correlated changes in other traits.²³ This is a very different scenario from what happens in a hospital room, where organisms receive a single high dose rather than 600 generations of sub-lethal selection. The practical lesson: UV resistance through evolution is biologically possible but requires conditions that proper disinfection protocols are designed to prevent. Giving an inadequate dose repeatedly is the one scenario where resistance could theoretically emerge.

Microbial Repair After UV Exposure

Microbes are not passive victims of UV damage. Many bacteria carry built-in DNA repair systems, and the most relevant one for UV disinfection is photoreactivation. When bacteria that have been UV-damaged are subsequently exposed to visible light (particularly blue light), an enzyme called photolyase can directly reverse the pyrimidine dimers that UV created, essentially undoing the damage.

This is a real engineering concern, not a theoretical curiosity. In one study on E. coli, photoreactivation after a low UV dose of 5 mJ/cm² allowed up to 46% of one bacterial strain to recover after four hours of exposure to visible light. At higher UV doses (15 mJ/cm² and above), photoreactivation was essentially negligible.²⁴ The revived bacteria retained their original characteristics, including antibiotic resistance coded in their DNA, confirming that photoreactivation is genuine recovery rather than regrowth from survivors.

Experiments under strict dark conditions have confirmed that the core UV damage response (a transient loss of culturability followed by partial recovery, with dose-dependent mutation rates) occurs with or without photolyase-mediated repair, though photoreactivation can amplify recovery when visible light is available.²⁵ The practical implication for water treatment is that UV-treated water exposed to sunlight shortly after treatment (in open reservoirs, for instance) could see some microbial recovery unless the initial dose was high enough to overwhelm repair capacity. This is one reason UV systems are designed with safety margins well above the minimum lethal dose.

How UV Dose Gets Measured and Validated

Claiming that a UV system delivers a specific germicidal dose is meaningless unless you can verify it, and verification turns out to be trickier than you might expect. UV lamps degrade over time, quartz sleeves accumulate fouling, and the geometry of a treatment chamber creates uneven dose distribution. You cannot just trust the manufacturer’s sticker.

The standard approach uses chemical actinometry: a solution that reacts to UV in a predictable, measurable way. One widely used method relies on uridine, a molecule containing the same UV-absorbing base (uracil) found in RNA. Because uridine’s absorption spectrum in the 240 to 290 nm range closely matches the spectral sensitivity of most microorganisms, it serves as an ideal stand-in for measuring germicidal photon delivery.²⁶ You expose the uridine solution to the UV system, measure how much it has degraded using a standard spectrophotometer, and back-calculate the dose. It is simple, environmentally friendly, and does not require expensive equipment.

For air disinfection systems where you need to map dose distribution across a three-dimensional space, a different technique uses iodide/iodate solutions in small spherical quartz chambers placed at various points in the room. The spherical shape allows UV from all directions to be captured, giving a true fluence measurement at each location.²⁷ Arranging a battery of these spheres at different positions creates a dose map that tells you where the UV zone is strong, where it is weak, and where pathogens might slip through undertreated.

UV Combined with Chemistry

UV disinfection on its own is a physical process: photons hit targets, damage accumulates, organisms die. But UV can also be combined with chemical oxidants to create what are known as advanced oxidation processes, which are used less for disinfection and more for destroying chemical contaminants that UV alone cannot break down, like pharmaceuticals, pesticides, and industrial pollutants in water.

The principle is that UV photons split hydrogen peroxide or persulfate molecules into highly reactive radicals (hydroxyl radicals and sulfate radicals, respectively) that attack organic compounds indiscriminately.²⁸ These radicals are among the strongest oxidizers available in water treatment. The UV component provides the energy to generate them, and the radicals do the heavy chemical lifting. This is a distinct application from germicidal UV, and the two should not be confused, but they sometimes coexist in the same treatment plant, with germicidal UV handling pathogens and UV-driven oxidation handling chemical pollutants in separate or combined stages.