As part of a chemistry course, I recently observed a demonstration in which the instructor electrified a mercury gas discharge tube (the tube looked something like this image). The glow from the tube appeared to be a light blue, but when I performed an internet search afterwards, I found that mercury gas discharge tubes apparently emit ultraviolet light, which can be harmful to the eyes and body. I was standing within one foot of the discharge tube (which looked something like this while lit) for several minutes. The tube was completely clear, which I am assuming means that it did not contain a phosphor to block the UV light (please correct me if I am wrong or if I am misunderstanding something).

The same demonstration was performed with a hydrogen gas discharge tube, which also was completely clear (see above).

Since I was looking at the tubes, could the higher frequencies of electromagnetic radiation (e.g. the ultraviolet light from the mercury) have damaged my eyes, skin, etc.?

After the demonstrations, my eyes seemed irritated, but I'm not sure if this was just a result of the placebo effect.

  • 3
    $\begingroup$ Most glasses absorb uv quite well. Ask your instructor. $\endgroup$ – Jon Custer Nov 11 '20 at 2:41
  • 5
    $\begingroup$ "Placebo" not really. The correct term is Nocebo. $\endgroup$ – Nilay Ghosh Nov 11 '20 at 3:15
  • 1
    $\begingroup$ If you did not feel like sand in your eyes it was likely nocebo. $\endgroup$ – Alchimista Nov 11 '20 at 10:52
  • $\begingroup$ "Is it safe" is a question without a binary answer. It isn't safe if the intensity of the light is high and duration of exposure is long. But that is also true of bright sunlight. $\endgroup$ – matt_black Nov 11 '20 at 12:56

Could it have caused irritation or injury? Possibly, but not likely, depending on a few factors:

  • The intensity of the emitted ultraviolet.
  • Proximity to the light source.
  • Length of exposure.

You'd need to calculate the dosage integrated over time.

That said, it is very unlikely any harm was done, because of the following:

  • The power supplies are limited to to only a few mA current. The total power is the arc voltage drop, ~30 to 100 VAC, times the current. For the supply abvove, that would be 100 VAC * 0.01 A = 1 W. One watt is low power for a mercury lamp.
  • Demonstration spectral lamps, such as this capillary mercury vapor lamp, usually have thick borosilicate glass capillaries. Borosilicate glass does not transmit shorter wavelengths well. Most of the UV would have been absorbed by the thick capillary wall.
  • The length of time was given as "several minutes", so the integrated intensity over time (i.e., total energy absorbed) was low.
  • That you state you were, "within one foot of the discharge tube," ~30 cm, is of concern, because some older discharge tube power supplies had exposed high voltage terminals. Though limited to 10 mA, considered below a dangerous current for electric shock, in most cases, one could still get burnt by the ~10,000 VAC arc, should one come in contact with exposed terminals, or could be startled and hurt by a muscular reflex action. Likely, though, the ends of the tube and associated wiring were safely enclosed.

So, all things considered, there might have been fumes in the lab or other cause of irritation, but I believe the UV light was not likely an issue.

  • $\begingroup$ Borosilicate glass does not transmit shorter wavelengths well. Most of the UV would have been absorbed by the thick capillary wall. This is not quite right. This type of tube emits a 365 nm line that is easily observable. That wavelength is not strongly absorbed in glass, and one can for example put it through a glass diffraction grating. However, that's a fairly long wavelength in the UVA band, and so not harmful at this intensity. $\endgroup$ – Ben Crowell Nov 12 '20 at 14:31

Without even knowing much about the tube itself, I can tell you the amount of UV it emits is a small fraction of what you're exposed to when you walk outside on a sunny day.

What matters for eye safety is actually power per unit area --- If you were to concentrate that light with a magnifying glass, and get it to focus on the back of your retina, then "maybe" you did some damage.

Note: I'm not a chemist. I just happened to spot this question on the list of 'open questions' in Stack Exchange. I'm an electronics engineer, and one of my professional responsibilities is the application and understanding of the European laser/LED eye safety standards (EN60825). I actually use a mercury vapor lamp to calibrate our spectrometer.

If it really concerns you, next time wear a pair of sunglasses. Even cheap ones do a very good job of blocking all types of UV. You're not wrong to be concerned a little.... UV light from an arc welder can seriously damage one's eyes if you stare at it (this is why welders wear nearly black eye protection). But the emissions from your little gas discharge tube is probably about a million times weaker than those from an arc welder.

It's like anything else -- Exposure time & exposure levels matter. You can smoke a single cigarette and have no ill effects. But smoke 20 a day for 20 years, much different story....

  • 3
    $\begingroup$ Even cheap [sunglasses] do a very good job of blocking all types of UV. One should be careful with cheap sunglasses that may not block UV but instead open your pupils up, allowing even more UV to enter the eye (see for instance healthcare.utah.edu/healthfeed/postings/2018/07/…) $\endgroup$ – WoJ Nov 12 '20 at 8:41
  • $\begingroup$ @WoJ Yeah, I should have said that. Funny thing, a few years ago I read in Consumer Reports this same topic, and they'd found even cheap sunglasses (nearly always) work fine. I was curious, so I went and bought a dozen pairs from various dollar stores. I found the same thing, every one blocked UV very well. So there surely could be some out there that are just dark plastic, but you'd have to (per my limited experience) try really hard to find them. $\endgroup$ – Kyle B Nov 13 '20 at 0:24
  • $\begingroup$ as an ex-physicist I approve of your approach :) What you found out is good news. $\endgroup$ – WoJ Nov 13 '20 at 11:14

Mercury discharge tubes such as germicidal lamps that are intended to emit short wave (UVC) radiation have to be made with an expensive quartz envelope rather than borosilicate glass.

Depending on the gas pressure they can be made to emit primarily UVC which is invisible and can be dangerous (at ~200nm) or UVA at ~400nm. Since you could see the discharge it was probably the latter, and probably any UVC was heavily absorbed by the envelope.

The 400nm emission line is actually longer wavelength than that used as "black light" in night clubs (~350 to 370nm) so less harmful.

The danger with strong UVC sources is that the light is completely invisible and does not trigger the blink reflex so you can get damage to the cornea (like snow blindness, usually temporary). Symptoms are described as follows:

Symptoms may not start for 6 to 12 hours after exposure to the ultraviolet light. Symptoms include:

Eye pain.

A feeling that something is in the eyes (foreign body sensation).

Sensitivity to light (photophobia).


  • $\begingroup$ The 400nm emission line is actually longer wavelength than that used as "black light" in night clubs (~350 to 370nm) so less harmful. The 405 nm line is visible, not UV. Mercury tubes actually do emit a 365 nm line, which is UVA and not strongly absorbed by the glass. $\endgroup$ – Ben Crowell Nov 12 '20 at 15:06

The type of mercury discharge tube used for these classroom demonstrations has five strong lines in its emission spectrum. Four are visible, while one is UV with a wavelength of 365 nm, which is in the UVA range. The wavelengths that are harmful to the human body are mainly the ones in the UVB and UVC bands, which have wavelengths shorter than 315 nm. This fact alone makes it very unlikely that you were harmed.

We can also think about the intensity. UVA is about 6% of the energy content of natural sunlight.

Ultraviolet is absorbed somewhat by glass, so some of the UV would have not have escaped from the tube. However, the absorption of the 365 nm UV line in a typical glass discharge tube of the types I've worked with is not very efficient. The UV line is clearly observable through fluorescence, and the intensity of the UV line is about the same as the intensities of the four visible wavelengths that it emits as well. There are shorter wavelengths emitted by some mercury discharge tubes, but in the classroom-demo tubes I've used, those wavelengths are not observable. This is probably due to the intentional design of the tube (electrodes, voltage, pressure, etc.), and the glass would probably also absorb these shorter wavelengths much more strongly.

So roughly speaking, there are five lines of equal intensity, and the UV emitted by the tube is probably on the order of 20% of the light it emits. (The exact proportion depends on the details of the tube, and can even vary between different parts of the same tube.) What this tells us is that the proportion of UVA to visible may be somewhat higher for the light from the mercury discharge tube than for sunlight (perhaps by a factor of 3 or 4), but not massively higher. Therefore the intensity of the light as sensed by your eye is not too bad a gauge of the intensity of the UVA from this source. From my experience doing this sort of demonstration in a classroom, I assume the light seemed much less bright to you than bright sunlight. The eye-brain system senses intensity sort of logarithmically, so I'm guessing that the brightness of the light was maybe $10^{-4}$ of the brightness of bright sunlight. The UVA exposure you experienced would therefore be many orders of magnitude less than what you would have experienced by being outside in bright sunlight for a similar amount of time.

Note that ultraviolet also has positive health effects. You need UVB so that your body can synthesize vitamin D.


Not the answer you're looking for? Browse other questions tagged or ask your own question.