[Allan Rogerson]

Allan Rogerson  asked the Naked Scientists:
   
My son is wanting to test amount of UV light emitted from the cfl lights
for his science project. What instrument (uv meter) would I need to
measure the results?

Thanks for your help.


Allan Rogerson

What do you think?

[Chemistry4me]

What instrument (uv meter) would I need to measure the results?

Didn't you just answer your own question?

[RD]

It is possible to photograph the UV from a CFL using a digital or film camera and a CD as a diffraction grating to split the light into a spectrum ...

http://www.thenakedscientists.com/forum/index.php?topic=19676.msg220177#msg220177

The brightness of the violet (false colour) line at the end of this spectrum would be a measure of the UV emitted.

A photoshop-type programme would allow you to quantify the brightness of this violet line in a photo of the spectrum, and thus enable you to compare the amount of UV emitted by one brand of CFL with another.


These UV sensitive beads (which change colour when exposed to UV) could provide a cheap and simple method of measuring UV ...

These ultra-low cost beads are extremely sensitive to natural light and change colour dramatically from white to various colours when taken outside - even on an overcast day. Containing trace amounts of a photochromic pigment that responds the the ultraviolet component of daylight, the beads have many different uses ranging from scientific testing of (UV) sun block materials to the creation of 'smart' jewellery.

http://www.mutr.co.uk/product_info.php?products_id=3505
 
However I don't know if these UV beads would be sensitive enough to change colour in the small amount of UV emitted by a CFL, perhaps if they were positioned very close to the bulb.   

[techmind]

Allan Rogerson  asked the Naked Scientists:
   
My son is wanting to test amount of UV light emitted from the cfl lights
for his science project. What instrument (uv meter) would I need to
measure the results?

Thanks for your help.


Allan Rogerson

What do you think?

Working with people who measure light for a living, I think it's going to be very difficult to get a quantitive (or even quantitative-comparative between bulbs) measure.

Complexities are (I believe) that the UV is mostly emitted from the ends of the tube where the phosphor is typically thinner. You won't be able to average over all directions without sophisticated equipment. Digital cameras will be relatively insensitive to UV light, and if you try to use a CD as a diffraction grating you will likely still struggle to distinguish the UV line from the two or three blue lines.


However... what you could do would be to obtain a "blacklight" filter which absorbs (most) visible light yet passes UV, and put that next to the bulb. Then put a fluorescent sample eg fluorescent red sticker on the opposite side of the filter to the bulb and see how bright it glows.
You might be able to measure the brightness of the sticker-glow semi-quantitively (enough to compare relative strength of several bulbs) if you use a digital camera with fixed, known manual exposure. Make sure you keep the sticker at exactly the same distance from all the bulbs as otherwise you'll mess up the measurements.


* Unfortunately I can't really suggest where you'd get a blacklight filter from on a school-project budget. You might be able to ask a theatrical supplier eg Lee Filters in the UK. For this application you might get away with a deep blue filter anyway. You could try Lee's "181 Congo Blue" ( http://www.leefilters.com/lighting/products/colours/ref:C4630710C6C881/ ) for starters - or ask for their suggestions. I can't image their filter "gels" (plastic sheet) will cost more than a few UK-pounds per square foot.
The wavelength you're trying to pass is 365nm, you want to block as much of the rest as you can.

[RD]

To put the amount of UV emitted from fluorescent lighting in context a 1993 study found that UV exposure from sitting under typical office fluorescent lights for eight continuous hours is equivalent to just over one minute of sun exposure (Lytle et al, 1993).

http://www.environment.gov.au/settlements/energyefficiency/lighting/publications/fs1.html


If this is true then a there should be a UV level equal to daylight a few centimeters from the fluorescent bulb.
If the UV level of the fluorescent light is 1/480th (1minute/8hours) of sunlight level at say 1 meter from the bulb,
then the UV level from the bulb will be equal to sunlight if measured 4.56 centimeters from the bulb (due to the inverse square law). 

[RD]

Some additional thoughts on this subject …

The UV emitted by a CFL is a few narrow lines in an emission spectrum, whereas the UV in sunlight is a band of frequencies (UVA UVB UVC etc), so making comparisons between UV from fluorescent light and UV from daylight is going to be difficult.

The main UV line in the CFL spectrum is 404nm (+/-1). Sunlight whose UV emission at that wavelength was the same as the CFL would be much more biologically damaging than the CFL because sunlight contains a much broader range of UV emissions, (most of which are of higher frequency so more energetic and more damaging to biology).

Just to make matters more complicated...
some photochemical reactions require light of a very specific wavelength (frequency) to occur.
The specificity of this phenomenon could allow a sparse emission spectrum (like the CFL's) to be as effective as sunlight at initiating a chemical reaction if the emission spectrum contained a line at the exact frequency required for the photochemical reaction to occur.   
E.g people with the immune disorder SLE, which is exacerbated by sunlight, claim fluorescent lighting also causes their condition to flare up, but that incandescent lighting does not*. This would be possible if the photobiochemical reaction in their immune system was specific to UV light of 404nm.

 (* incandescent lighting does not contain UV)

[Bored chemist]

The 404nm line is not UV, it's visible violet light.

CDs are made from polycarbonate which absorbs UV very well indeed. Some of the UV from a CFL will be the 254nm line and I think that is pretty close to the absorbtion maximum.

The 366 nm radiation is usually a relatively minor part of the emision from a low pressure mercury arc but it might be a major part of the UV that gets through the glass envelope..

Incandescent lamps can emit UV. There were potential problems with this when tungsten/halogen lamps first became popular. They now include UV blocking filters

[RD]

No UV emission from a 200 Watt incandescent bulb (see below)...

Spectra from another website which seems to confirm the violet line is actually UV (~400nm)...

 [ You are not allowed to view attachments ]

http://ioannis.virtualcomposer2000.com/spectroscope/amici.html#1fluop


According to this site 436nm is "the approximate limit of human vision at shorter wavelengths,
 i.e. a 400nm line would not be visible to the human eye.

http://www.thenakedscientists.com/forum/index.php?topic=19676.msg220189#msg220189

This supports my claim that the 404nm line is not visible violet light but it is a false violet colour recorded by the camera caused by invisible UVA.

[BTW I have photographed spectra from fluorescent sources using a SLR film (not digital) camera using a CD as a diffraction grating, and can confirm that lines appear on the photographs of spectra which were not visible through the camera's (optical) viewfinder, i.e. invisible ultraviolet and infrared emission lines are recorded on film so must not be absorbed by the polycarbonate of the CD. 

[Bored chemist]

I can see the 405nm line. It is visible and therefore it is not UV.

[RD]

I can see the 405nm line. It is visible and therefore it is not UV.

If you are referring to the image above the 405nm line is recorded on film as a false colour violet line in the photographic image but this line would not be visible to the naked eye. Film and digital cameras are sensitive to UV and IR light which are invisible to the normal unaided human eye*.

Only near UV is of interest for UV photography, for several reasons. Ordinary air is opaque to wavelengths below about 200 nm, and lens glass is opaque below about 180nm. UV photographers subdivide the near UV into:

Long wave UV that extends from 320 to 400 nm, also called UV-A,
Medium wave UV that extends from 280 to 320 nm, also called UV-B,
Short wave UV that extends from 200 to 280 nm, also called UV-C.
(These terms should not be confused with the parts of the radio spectrum with similar names.)

http://en.wikipedia.org/wiki/Ultraviolet_photography


[* Aphakic persons may be able to see UV]


Re: original question, cheapish UV meters are available to measure sun exposure, e.g.,
 but not I'm not sure they would be sensitive enough to measure the UV from a CFL.

[techmind]

Here's the spectrum of a modern fluorescent tube I measured a few years ago. It is fairly typical.

 [ Invalid Attachment ]

The instrument I used only measures from 780-380nm.


As to when visible becomes UV or "near UV" (or infra-rad) is not precisely defined. The eye becomes progressively less sensitive at the edges of the "visible" spectrum. If you make the light source strong (bright) enough it may still evoke a physical response. Whether or not an individual can see a given wavelength does not really define it as "visible" or not in the general sense. That said, I think 405nm would generally be regarded as visible.

See also the first plot on my webpage: http://www.techmind.org/colour/
This shows the internationally-agreed standardised response-functions (strictly colour-matching functions) of the human eye, and as colourimetrically-accurate representation of those colours as you can really get on a computer-screen.


Mercury natively puts out several strong UV lines, principally 254nm (UV-C) but this is absorbed by almost everything (except quartz-glass) and won't get through the glass of the tube. I think there's a 306nm line, which is quite weak, and will probably still be significantly attenuated by the glass. Mercury also emits a modest 365nm wavelength. 365nm is "near UV" (UV-A), the dominant wavelength used for theatrical and disco/club effects, and penetrates thin glass reasonably well. This wavelength provokes strong fluorescent effects in clothes whashed with "whitening" detergents, and in common "fluorescent-coloured" paints, inks, and plastics.

365nm "blacklight" tubes are normally perceived as a hazy deep purple, but it's not completely clear whether this sensation comes primarily from the 365nm emission itself, or from some longer wavelengths which also penetrate the blacklight (visible-absorb) filter.
A completely unfiltered mercury tube still emits a certain amount of visible light and looks a sky-bluey-white, but is much less efficient at making visible light than a fluorescent phosphor which converts the high-power 254nm UV inside the tube into visible light.

[RD]

Daveshorts took the pic below, perhaps he can tell us if the violet line was visible to the naked-scientist's naked eye,
 (seeing it on a digital camera's LCD screen doesn't count).
 

[ You are not allowed to view attachments ]

http://www.thenakedscientists.com/HTML/content/kitchenscience/exp/colours-in-cds/

I suspect the violet line in this spectrum is caused by UV and was not visible to the eye but has been recorded
 by the camera as a false violet colour. Cameras, film & digital, can "see" UV, human eyes cannot

http://www.thenakedscientists.com/forum/index.php?topic=19676.msg220177#msg220177

[techmind]

Re DaveShorts' picture above, I'd be quite certain you're not seeing IR at the right.
For one, fluorescent lamps are not known for producing IR (it'd be a waste of energy anyway) and secondly digital cameras typically render near-IR (eg the 750nm-ish of TV remote-controls) as purply anyway - just to confuse you!

I've just tested my digital camera and it seems to see a 365nm "blacklight" tube at a broadly comparable brightness to what I perceive (compared to other visibly-illuminated things in the scene at the same time), although the camera sees it as almost pure blue whereas I see a more purply colour. Of course your mileage may vary. You wouldn't expect cameras to be very sensitive to UV or IR otherwise it'd cause them to record obviously-wrong colours.


Dave's CD-photo (quoted by RD) looks like an extremely good match to my spectrum, with the one he's labelled "UV?" matching my 405nm peak, and his third line from the right being noticably broader that the two leftmost lines etc. Most excellent!

One day I'll get around to modifying my simulated-spectrum computer-program to take measured spectra and generate correctly-simulated colour-accurate pictures.


Actually Dave's CD-diffraction grating works even better with a DVD than a CD, probably because the pitch of the tracks is finer, so you get better spectral resolution (even in the first-order diffraction pattern).

I can confirm that the second purple line which Dave has marked "UV?" and must be the 405nm is barely visible (but just discernable if I look for it and catch a bright reflection of the light) using a DVD and simple setup. With a better optical setup (some black card to cut the background light) it would be much more clearly visible.

I strongly suspect that the setup Dave used to get his photo is a bit more sophisticated than he's letting on (although nothing you couldn't do at home).

[RD]

I can confirm that the second purple line which Dave has marked "UV?" and must be the 405nm is barely visible (but just discernable if I look for it and catch a bright reflection of the light) using a DVD and simple setup.

I added the "UV? and "IR?" text to Dave's photo of the spectrum from a CFL: my fault if these labels are incorrect.

[Bored chemist]

RD, when I said that the 405nm radiation is visible I meant neither more nor less than that I could see it.
As it happens I saw it from a CFL and through a calibrated monochromator.
The 405 line is, at least from my point of view, quite clearly visible. It is violet in colour and, therefore not ultaviolet. Other people, particularly the elderly, may not be able to see it because it may be absorbed by lens of the eye if this has become yellowed.

What cameras, film, video or bumble bees make of it isn't the issue.
If you want to measure UV from one of these lamps then looking at the 405 nm light isn't any more helpful than looking at the yellow doublet.

[lyner]

hey BC, are you an X Man?

[rosy]

The webcam Dave (daveshorts) uses as a near-IR camera renders near-IR as a sort of pinky-grey colour (at least on his computer) I think.
You can see the images in the kitchen science article about making the IR camera:
http://www.thenakedscientists.com/HTML/content/kitchenscience/exp/make-an-infra-red-camera/

[rosy]

Oh, and on the fluorescent lights producing IR thing... they really don't produce anything much to speak of... at least in the part of the spectrum not cut out by crossed polarizing filters. If you turn a flurorescent light on and off it makes very nearly no difference at all to the amount of light picked up by the camera (in stark contrast to incandescent bulbs).

[techmind]

Well here's what my digital camera makes of remote-control IR wavelengths (without any physical hacks like removing filters):

 [ Invalid Attachment ]


This purply colour seems fairly typical for unmodified colour cameras exposed to IR.

As a very general rule, I've found that the cheapest cameras (eg. mobile phone cameras) will be most sensitive to IR, then digital compacts, and digital SLR will be least sensitive to IR because they have a better IR-cut filter.

[Bored chemist]

Fascinating stuff about IR.
What does it have to do with measuring UV?