Is looking at a bare mercury lamp light dangerous?

[Roger Breton]

I am thinking about setting up some observation experiments with my students
and the help of a spectroscope I bought from Edward Scientific to
demonstrate the dicontinuous spectral nature of some light sources. I was
thinking of observing a 175W bare mercucy lamp and a 150W sodium lamp
because those, I think, would be most revealing as compared to observing the
spectrum of a tungsten lamp. But someone told me that directly observing a
mecury lamp would be dangerous? I don't want to make myself or any of my
students blind because of my ignorance.

Any suggestions? Comments?

Roger Breton

 

[Victor Roberts]

I am thinking about setting up some observation experiments with my students
and the help of a spectroscope I bought from Edward Scientific to
demonstrate the dicontinuous spectral nature of some light sources. I was
thinking of observing a 175W bare mercucy lamp and a 150W sodium lamp
because those, I think, would be most revealing as compared to observing the
spectrum of a tungsten lamp. But someone told me that directly observing a
mecury lamp would be dangerous? I don't want to make myself or any of my
students blind because of my ignorance.

Any suggestions? Comments?

Don't do it! High pressure mercury arc tubes emit UV radiation that
can be harmful to the eyes and skin. When the arc tube is enclosed in
the glass outer jacket they are safe to look at from a reasonable
distance, but the bare arc tube is not safe.

You can view bare HPS arc tubes, but they will self destruct in a few
hours if operated without their protective outer jackets.

Why do you want to view the bare arc tubes instead of the lamps
themselves?

 

[JB]

Don't do it! High pressure mercury arc tubes emit UV radiation that
can be harmful to the eyes and skin. When the arc tube is enclosed in
the glass outer jacket they are safe to look at from a reasonable
distance, but the bare arc tube is not safe.

You can view bare HPS arc tubes, but they will self destruct in a few
hours if operated without their protective outer jackets.

Why do you want to view the bare arc tubes instead of the lamps
themselves?

--

I reckon he really means 'bare lamp' as in 'lamp not in a luminaire'?

JB

 

[Victor Roberts]

I reckon he really means 'bare lamp' as in 'lamp not in a luminaire'?

You may be right, but until we know what he means I think the warnings
are in order.

 

[Roger Breton]

OK, I won't do it. But what is that 'glass outer jacket' made of: is it
clear glass as in transparent or is it layered with some kind of UV filter?
As I would want to be able to observer the spectral lines of mercury with my
spectrograph I would not be interested in a lamp that is covered with white
like a fluorescent tube.

Is there anything that you know of that would not be harmfull that I could
look at from a distance and that would let the spetral line through?

Thank's.

Roger Breton

 

[Danny Rich]

Roger,

Victor Roberts' advice is spot on. Even with the protective shell, a 175W
lamp produces a lot of UV energy at 264nm, 313nm, 365nm (see the first
chapter of Wyszecky & Stiles). Those lines are all ionizing radiations for
the components of the eye and skin. Why kind of spectroscope do you have?
Perhaps it might be better to put the lamp in a box and open a small hole
and then capture the image from the spectroscope with a digital camera. You
could even do a "wavelength calibration" experiment like the famous Canadian
spectroscopist, Herzberg. Using the grid function in Photoshop. And no
ones skin or eyes need to be irradiated. By the way, using a linear
variable interference filter (also available from Edmund Scientific) I can
see both the mercury lines and the Yittrium Oxide red lines through the
white phosphor of a 3-band fluorescent lamp. You can also see that such red
lines are missing in the standard coolwhite lamp light.

Danny

 

[Victor Roberts]

OK, I won't do it.

What is "it"? As you can see, we have a certain amount of disagreement
here about what you want to do.

But what is that 'glass outer jacket' made of: is it
clear glass as in transparent or is it layered with some kind of UV filter?

Most high pressure mercury lamps have a clear glass outer jacket. Some
"color corrected" high pressure mercury lamps have a phosphor coating
on the inside of the glass bulb to convert the UV radiation from the
arc tube into visible light at wavelengths that are missing from the
discharge. You can easily get lamps without the phosphor coating, but
the coating will not prevent you from observing the mercury visible
lines.

High pressure mercury arc tubes generate far less 245nm radiation per
watt of input power than a low pressure mercury discharge such as a
fluorescent lamp. However, because high pressure mercury lamps operate
at such a high power density, the quartz arc tube does emit enough
254nm radiation to be harmful to human eyes and skin.

The glass does not have a UV filter, but it is naturally less
transparent to 254nm UV than the quartz arc tube. The bulb is normally
made from a borosilicate glass, such as Corning 7740 that will trap
almost all the 254nm radiation generated by the arc. 7740 glass does
not absorb the 365nm near UV generated by the arc so you will be able
to measure this, with appropriate instruments. You will obviously be
able to see all the mercury visible lines.

As I would want to be able to observer the spectral lines of mercury with my
spectrograph I would not be interested in a lamp that is covered with white
like a fluorescent tube.

The phosphor coating does not prevent observation of the mercury
visible lines, even in a fluorescent lamp. You do need to use a narrow
bandwidth spectrometer in order to allow the mercury visible lines to
be observed "above" the phosphor background.

Is there anything that you know of that would not be harmfull that I could
look at from a distance and that would let the spetral line through?

Which line? If you mean the 254nm line, you can mount the arc tube in
a box that prevents the light from shining on human eyes and skin, and
then observe the spectra with a spectrometer and appropriate detector.
Remember, even if the 254nm radiation was not dangerous, you can't see
it with your eyes.

There are lots of interesting lines in the visible region, so I am a
bit confused about your desire to "see" the 254nm line, if I
understand you correctly.

 

[zxcvbob]

OK, I won't do it. But what is that 'glass outer jacket' made of: is it
clear glass as in transparent or is it layered with some kind of UV filter?
As I would want to be able to observer the spectral lines of mercury with my
spectrograph I would not be interested in a lamp that is covered with white
like a fluorescent tube.

Is there anything that you know of that would not be harmfull that I could
look at from a distance and that would let the spetral line through?

Thank's.

Roger Breton

How about using a 15W fluorescent germicidal lamp? That should give you a
similar spectrum as an uncoated MV lamp, but at a lower and safer power level.

Do they make R50 "sun lamps" anymore? Those were self-ballasted MV lamps
with little-if-any coating. Because of the reflector you could control the
light a little better than just a bare MV bulb, and I don't think I've ever
seen a clear normal MV lamp. A sun lamp will still be dangerous if someone
is stupid enough to stare into it.

Best regards,
Bob

 

[Ben Newsam]

You may be right, but until we know what he means I think the warnings
are in order.

http://www.nichol.co.uk

Look under Products... Optics... Spectral Sources

I took a peek at a mercury tube yesterday with my digital spectrograph,
and the visible mercury lines show well. Unfortunately, there is a
component in the infrared that looks suspiciously like argon to me.

 

[Jeff Waymouth]

The mercury arc tube is dosed with elemental mercury and a "fill
pressure" of about 20-30torr of argon. (or it was when I was production
engineer for that line lo these many moons ago)

Jeff Waymouth

 

[Jeff Waymouth]

How about using a 15W fluorescent germicidal lamp? That should give
you a similar spectrum as an uncoated MV lamp, but at a lower and
safer power level.

Do they make R50 "sun lamps" anymore? Those were self-ballasted MV
lamps with little-if-any coating. Because of the reflector you could
control the light a little better than just a bare MV bulb, and I
don't think I've ever seen a clear normal MV lamp. A sun lamp will
still be dangerous if someone is stupid enough to stare into it.

Best regards,
Bob

Looking at a gemicidal lamp ain't safe either. It desn'ttake much 253.7
nm energy to start frying things!

The R type sunlamps were discontinued long ago (in the USA, anyways).

Jeff Waymouth

 

[Victor Roberts]

How about using a 15W fluorescent germicidal lamp? That should give you a
similar spectrum as an uncoated MV lamp, but at a lower and safer power level.

Compared to a high pressure mercury discharge, the low pressure
mercury discharge used in the germicidal lamp produces far fewer lines
in the visible portion of the spectrum. And, as Jeff has stated, there
is that pesky problem of the 254 nm radiation.

 

[Martin Brown]

OK, I won't do it. But what is that 'glass outer jacket' made of: is it
clear glass as in transparent or is it layered with some kind of UV filter?
As I would want to be able to observer the spectral lines of mercury with my
spectrograph I would not be interested in a lamp that is covered with white
like a fluorescent tube.

You will still see the main mercury lines clearly even with a common or
garden fluorescent light (and certainly well enough for teaching
purposes).

Other interesting light sources are sunlight, low pressure sodium, neon
lamps, and the little thin argon mix lamps that have become popular for
putting in "go faster" PC's. LED spectra are also worth showing students
- they are not as monochromatic as they might first appear.

Another demonstration that you can use to show your students spectra
requires much less investment than buying a spectroscope.

Hold one of the aluminium coated shovelware CDs off the front of a PC
mag so that sunlight glances off it at a very shallow angle and then
look down at right angles onto the face of the disk to see the solar
spectrum. The dispersion is easily good enough to see Fraunhofer
absorbtion lines in the first and second order spectra. Not useful for
quantitative work - but very cheap, quick and educational. No slit
needed it is self collimating.

(Check posts in sci.astro.amateur for precise details of how to do it)

Is there anything that you know of that would not be harmfull that I could
look at from a distance and that would let the spetral line through?

Boxed so that no stray light escapes except through a UV filter glass
window. I once had UV induced conjunctivitis (welders flash?) from a
badly set up school biology experiment that irradiated the whole lab
with a bunch of high power mercury gro-lights. It isn't nice at all -
like having sand in the eyes for a few days.

Regards,

 

[Ben Newsam]

Another demonstration that you can use to show your students spectra
requires much less investment than buying a spectroscope.

Hold one of the aluminium coated shovelware CDs off the front of a PC
mag so that sunlight glances off it at a very shallow angle and then
look down at right angles onto the face of the disk to see the solar
spectrum. The dispersion is easily good enough to see Fraunhofer
absorbtion lines in the first and second order spectra. Not useful for
quantitative work - but very cheap, quick and educational. No slit
needed it is self collimating.

And if the same trick is done with an "energy bulb" instead of a full
fluorescent strip light, you can see quite distinct images of the bulb
in the various colours of the mercury spikes. In the blue and turquoise
anyway, the orange and red are a little more confused of course.

 

[Roger Breton]

What is "it"? As you can see, we have a certain amount of disagreement
here about what you want to do.

I meant to say that I won't go ahead with this experiment without the proper
safety for me and my students. The kind of spectrograph I have was made by
Wabash Instrument Corp, in the US, Model SP-125. It has a variable
adujstable slit at one end to point it at a source to observe. It must have
some dispersive element inside but since the instrument is sealed I can't
really tell. Probably some kind of grating. So the light comes in at one
end, hits the grating and is then projected on some graduated screen where I
can read the wavelength in Angstroms.

Last year, when I gave my color course, I brought the spectroscope in class
along with a plain 18" GE warm white and cool 18W white domestic fluorescent
lamp and some tungsten 60W bulbs. We started our series of observations by
pointing the spectroscope at the 60W bulb. In the spectroscope, students
were able to come in contact with a full continuous spectrum in the form of
a 'rainbow', that's how the dispersed light of the 60W bulb appears when
seen through the spectroscope. Then, we turned the spectrocope to a
fluorescent lamp. I wish it would have been a 'bare', clear transparent
fluorescent tube because, while some of the mercury lines shined stronger at
certain wavelengths in the spectroscope, and it was clear to see them, the
phosphor coating inside the fluorescent tube also created a continuous
spectrum, thereby lessening the impact of the observation. I guess even if I
would have had a bare fluorescent tube to look at (one not coated inside
with any material) that would still have been too harmful for direct
observation? After looking at the fluorescent tube, since it was night time
in january, we turned the spectroscope at some sodium street lamps hoping to
catch some real line spectrums? It was not as sharp as I imagined it would
be. Finally, we pointed the spectroscope at a CRT, on a white patch. There
we saw two clearly distinguishable lines, one at 6500 Angstroms and one at
about 6400 A (I know it's nanometres but that's how the instrument scale is
graduated...). Tha was quite eye opening for the students. To top it all, we
alternated some Wratten photographic filters like 25A, 47B and 58 in front
of the telescope while looking at a 60W bulb to demonstrate that part of the
spectrum was absorbed by the filter.

This time, I am preparing for a new iteration of that class in january and
I'd really like to have some real 'line sources' to observe safely for the
students. I strongly believe that direct observation of the visible spectrum
is a good pedagogical approach to introduce students to color science. One
that I believe allow me to explain illuminants, spectrums of objects and,
later, the Standard Observer and the CIE colorimetry system.

Roger Breton

 

[Don Klipstein]

I meant to say that I won't go ahead with this experiment without the proper
safety for me and my students. The kind of spectrograph I have was made by
Wabash Instrument Corp, in the US, Model SP-125. It has a variable
adujstable slit at one end to point it at a source to observe. It must have
some dispersive element inside but since the instrument is sealed I can't
really tell. Probably some kind of grating. So the light comes in at one
end, hits the grating and is then projected on some graduated screen where I
can read the wavelength in Angstroms.

Last year, when I gave my color course, I brought the spectroscope in class
along with a plain 18" GE warm white and cool 18W white domestic fluorescent
lamp and some tungsten 60W bulbs. We started our series of observations by
pointing the spectroscope at the 60W bulb. In the spectroscope, students
were able to come in contact with a full continuous spectrum in the form of
a 'rainbow', that's how the dispersed light of the 60W bulb appears when
seen through the spectroscope. Then, we turned the spectrocope to a
fluorescent lamp. I wish it would have been a 'bare', clear transparent
fluorescent tube because, while some of the mercury lines shined stronger at
certain wavelengths in the spectroscope, and it was clear to see them, the
phosphor coating inside the fluorescent tube also created a continuous
spectrum, thereby lessening the impact of the observation. I guess even if I
would have had a bare fluorescent tube to look at (one not coated inside
with any material) that would still have been too harmful for direct
observation? After looking at the fluorescent tube, since it was night time
in january, we turned the spectroscope at some sodium street lamps hoping to
catch some real line spectrums? It was not as sharp as I imagined it would
be.

For a cheap "bare" fluorescent tube, see if you can find an "unfiltered
blacklight" such as F20T12/BL or F20T12/350BL (or is it F20/BL or
F20/350BL?). These are 20 watt 2-footers that will work in any fixture
that takes 20 watt T12 lamps (bulbs).
The UV from these should have no significant skin hazard, except maybe
to anyone taking certain photosensitizing medications (I don't know what
these are). Eye hazards from prolonged staring into these at close range
may be significant. That can be negated by GAM 1510 filter gel, available
from some theater supply shops. Or else make some sort of box to surround
the lamp with so that the only way out is through the spectroscope - and I
imagine it's unlikely any eye-endangering amount gets through the
spectroscope. Just in case, maybe put a piece of GAM 1510 filter gel or
polycarbonate plastic sheet in front of the slit.

A clear 175 watt mercury lamp would need similar precautions - its UV
output is mainly 365-366 nm. The glass outer bulb largely blocks the
major shorter wavelengths (probably not a minor one in the shortwave end
of the UVA range). But if you want to seriously play around with a high
pressure mercury lamp, I would use that GAM 1510 gel or polycarbonate
(such as "Lexan") to block the 365-366 nm UV - severe eye exposure to UVA
is not good for the lens of the eye.
And be ready for the heat produced by these things! If you box one into
a largely light-tight box, you get most of that 175 watts being converted
into heat (close to 600 BTU/hour). If the light can get out, you still
get a lot of heat, since a mercury arc produces some light, some UV, at
least a fair amount of heat and not so much infrared. Tungsten filaments
produce mostly infrared, most of which passes through the glass bulb and
escapes and develops heat in the room but not in the bulb - but in a 175
watt mercury, you have about 50 watts heat conducted from the arc, about
18 watts of heat from electrode losses, at least 20 watts from shorter
wavelength UV that is absorbed in the glass bulb if it makes it out of the
arc tube, and a few watts of 365-366 nm UV, visible, and IR will probably
be absorbed at the arc tube surface and become heat. Probably at least
100 watts or around 350 BTU/hour of non-radiant heat, which I guesstimate
to be like that from a 300 watt incandescent, maybe almost as bad as a 500
watt halogen for non-radiant heat...

- Don Klipstein ([email protected])

 

[Victor Roberts]

Last year, when I gave my color course, I brought the spectroscope in class
along with a plain 18" GE warm white and cool 18W white domestic fluorescent
lamp and some tungsten 60W bulbs. We started our series of observations by
pointing the spectroscope at the 60W bulb. In the spectroscope, students
were able to come in contact with a full continuous spectrum in the form of
a 'rainbow', that's how the dispersed light of the 60W bulb appears when
seen through the spectroscope. Then, we turned the spectrocope to a
fluorescent lamp. I wish it would have been a 'bare', clear transparent
fluorescent tube because, while some of the mercury lines shined stronger at
certain wavelengths in the spectroscope, and it was clear to see them, the
phosphor coating inside the fluorescent tube also created a continuous
spectrum, thereby lessening the impact of the observation.

If you can reduce the width of the input slit on your spectrometer you
can increase the apparent intensity of the mercury lines relative to
the light generated by the phosphor.

I guess even if I
would have had a bare fluorescent tube to look at (one not coated inside
with any material) that would still have been too harmful for direct
observation?

Correct. Actually, no more harmful than looking at a coated lamp. The
glass used to make a normal fluorescent lamp does not pass the
relatively strong 254 nm or 185 nm UV lines generated by the mercury
discharge.

After looking at the fluorescent tube, since it was night time
in january, we turned the spectroscope at some sodium street lamps hoping to
catch some real line spectrums? It was not as sharp as I imagined it would
be.

The spectra of metal halide lamps is much more interesting than that
of high pressure sodium lamps - but you may need more resolution than
your spectrometer has to resolve the many lines in a metal halide
lamp.

[snip]

This time, I am preparing for a new iteration of that class in january and
I'd really like to have some real 'line sources' to observe safely for the
students.

One of the most interesting line sources is a low pressure neon
discharge. The old neon indicator lamps, which have been almost
completely replaced by LEDs, have a rich line spectra and operate at
low power. If you can find neon lamps, perhaps at a Radio Shack or
equivalent store, they are both interesting and safe to observe.

 

[TKM]

Victor wrote:

This time, I am preparing for a new iteration of that class in january and
I'd really like to have some real 'line sources' to observe safely for the
students. I strongly believe that direct observation of the visible spectrum
is a good pedagogical approach to introduce students to color science. One
that I believe allow me to explain illuminants, spectrums of objects and,
later, the Standard Observer and the CIE colorimetry system.

Roger Breton

<SNIP>

An interesting and educational spectral demonstration is to look at the
spectrum of an HPS lamp as it warms up. As the light output increses, the
original low pressure sodium lines are absorbed and replaced by a gap with a
broader spectral discharge on either side. From the safety standpoint,
there is no UV emission from the lamp and a 50 or 70 watt HPs lamp provides
plenty of light.

By the way, it is not difficult to project such spectra on a large screen so
a whole classroom can see the demonstration at once. You need a light
source, an enclosure with a hole to let some of the light out, an adjustable
slit and a diffraction grating arranged so that you can adjust the distance
between the grating and the slit (for focusing). Make a slide with a
wavelength scale and project that on the screen below the spectrum and you
have a way to identify the lines.

Image quality is a direct function of the quality of the diffraction grating
and the adjustment of the slit.

Terry McGowan



Terry McGowan

 

[Roger Breton]

One of the most interesting line sources is a low pressure neon
discharge. The old neon indicator lamps, which have been almost
completely replaced by LEDs, have a rich line spectra and operate at
low power. If you can find neon lamps, perhaps at a Radio Shack or
equivalent store, they are both interesting and safe to observe.

Is it because they contain different gases that different neon lamps produce
different color of lights? I am looking at some neon lamps on eBay, right
now. I think I'll order. It's not very expensive and is bound to be fun.

Regards,

Roger Breton

 

[Victor Roberts]

Is it because they contain different gases that different neon lamps produce
different color of lights? I am looking at some neon lamps on eBay, right
now. I think I'll order. It's not very expensive and is bound to be fun.

Well, my answer may very well start a flame war

As far as I am concerned, a neon lamp contains ----neon as the active
gas. However, there are some people who believe that any lamp with
cold electrodes, and certainly any lamp with cold electrodes that is
bent into interesting shapes, is a "neon" lamp, even if the lamp does
not contain a single atom of neon. These lamps produce colors other
than the characteristic neon red color because they use gasses other
than neon.

On the other hand, there are "real" neon lamps that use a phosphor to
convert the small amount of UV produced by the neon discharge into
other colors. These are still neon lamps, IMHO. However, there is very
little useful UV produced by a neon discharge, so the phosphor output
is rather weak. I would assume that a "neon" lamp that does not have
the neon color is using other gasses.