Don Klipstein wrote:
[snip]
The 365 nm triplet is actually a weak feature of the low pressure
mercury vapor discharge. It is related to the 577-579 nm yellow
triplet, which is also a strong feature of a high pressure mercery
vapor discharge but a weak feature of the low pressure one.
Most fluorescent lamp phosphors do not utilize 365 nm, though the
usual mid-blue component of triphosphors 3500K and higher does
utilize it.
Also, soda lime glass is nearly enough transparent at least down to
340 nm.
My Optics reference (K.D. Alexopoulos, General Physics, Optics, Athens, 1966),
seems to agree with Don's 340nm for regular glass.
However, as much as I would hate to disagree with Don and with the above
reference, in my experience the above is a tad low. I have in front of me the
OSRAM catalog which gives the peak for UVA sun-tanning indium-amalgam
fluorescents, L40W/79 K In, as 350nm.
That sounds long to me for suntanning. Wavelengths longer than about
330-335 nm or so don't do that much.
If the 340nm lower bound for regular glass was correct, suntanning would be
possible behind regular glass windows. But it's not. Hence the 340nm
figure for soda lime glass cannot be correct.
I thik the effective cutoff wavelength is a little longer for windows
and a little shorter for fluorescent lamps, in part from window glass
being thicker, and in part from window glass having a slight tinting by
iron.
As for fluorescents made specifically for suntanning - I think they
would use a different glass to pass even shorter wavelengths.
Note however that the direction of radiation with respect to the glass envelope
and phosphor may be crucial to the transmissivity of UV. Whereas *outcoming*
365nm radiation may pass through the glass envelope and hit the phosphor and
excite it, this *DOES NOT* mean that there is *incoming* 365nm radiation which
manages to survive both the phosphor AND the CFL glass envelope and radiate
outwards.
As I said, spectroscopic analysis does not show any appreciable amounts of
radiation below 390nm, except traces in regular 2700K CFLs.
I will gladly stand corrected if anybody shows a specific spectrogram from a
calibrated spectroscope with non-trace amounts of UV below 390nm.
I don't expect there to be a whole lot because the 365-366 nm triplet is
a weak feature of the low pressure mercury vapor discharge.
Some spectral power distribution curves:
http://ledmuseum.home.att.net/spectra7.htm
Look for "Trisonic" 6500K (halophosphor) ones, all others noted as
"dollar store" ones, and compact fluorescents noted as 2700 K, Sylvania
"white" (halophosphor). The 365-366 nm feature is about half as strong as
the 404.7 nm one in all of these.
Two non-phosphor lamps that emit UVC are shown also: A "Water Purifier"
one and a "UVC" one (though with weak UVC output). The latter also has
365-366 about half as strong as 404.7, and the former has more 365 (I have
seen a similar lamp get hotter and have higher mercury vapor pressure -
that may be the explanation).
I tried getting the spectral power distribution of a Sylvania F40/350BL.
(Copying and pasting links is messy with Sylvania's website - I would try
searching their USA "business" lamp catalog for F40350BLECO 30/CS 1/SKU
or 24922.)
The spectral power distribution has the 365-366 nm spike smaller than
the 404.7 nm one, and there is a little one shown at 310 nm.
- Don Klipstein (
[email protected])