my nootloss - 8-6-2005 at 12:17
Hey guys, I think it would be helpful for us to find a way to make a spectroscope, so we could find out what stuff is in the chemicals we buy
(example, in drain cleaner, we could boil all the h2so4 and h2o off and use the spectroscope to find out what left in it…)
I was thinking of using a blank CD as the prism?
Sandmeyer - 28-6-2005 at 18:11
hehe, you must be kidding, right?
12AX7 - 28-6-2005 at 18:46
I've always wanted to make a mass spectrometer, but ya know, I haven't the slightest idea how the hell you ionize a solid substance. Shoot
a thick e-beam at it and sputter??
Tim
neutrino - 28-6-2005 at 19:20
I think he meant to put things in a flame and watch spectral lines. Clever, although somewhat out of the scope of amateur science.
12AX7 - 28-6-2005 at 21:38
Howso? A prism and scale (or better yet, color photographic film) aren't very diffcult at all.
Tim
my nootloss - 28-6-2005 at 22:32
Yeah that’s the amazing thing about it, it’s very EZ and a very good thing to have in the lab, the reason I say this is just think about all the
crappie chemicals we are some times forced in to using i.e. sulfuric acid. The drain cleaner I got had some very weird things in it, so what I could
do is boil it don’t and use the spectroscope to find out what I have left (sorry I used the same example 2 times, I forgot.) I am sure any one of
you guys has some chemical in your lab that you would love to find out what is in it. I really don’t know a whole lot about spectroscopes, I have
looked some up “spectroscope” in my books (old books, most use a prisms or so it says in the book. So I guess that would be fine).
Neutrino- I don’t really so how this is out of the range of any person who is unwavering in there quest for chemical understanding. Thank you by the
way; in my book “Clever” is one of the most highly recognized words.
12AX7- please accept my apogee for my lack of knowledge, but what is a “color photographic film” and how accurate and how many times can you use
one/it? All so how would we get are hands on these?
To all - I am really open to any and all suggestions on how to make one of these great things
12AX7 - 29-6-2005 at 00:12
You could use B&W film too, but it'd be more interesting with color. 
Just a strip of photographic film covering the viewing screen, you project the spectral lines on it exposing those areas, which are then developed at
the 1 hour photo into visible color. Well, if they do large strips.
Tim
my nootloss - 29-6-2005 at 06:40
i work at one so i will ask
chromium - 29-6-2005 at 13:58
I once made simple CD spectroscope. It is very easy - just piece of CD disk and some lightproof box with two holes. I did this just for fun so i did
not bother do draw scale. I used gas flame and sometimes voltaic arc as light source. Some spectral lines are clearly visible but there is still lot
problems to solve.
There is lot of chemical elements everywhere and its not at all easy to guess which combinations can make that particular pattern. Especially if you
have no proper scale. Another problem is that some common chemical elements have no intense spectral lines in visible region. For proper determination
of those one needs infrared or UV spectroscopy.
I probably will continue with these experiments as its indeed usefull to have spectroscope.
12AX7 - 29-6-2005 at 15:37
Dunno about IR (maybe one of those IR test cards from RadioShack, for testing your remote?), but UV you could add a fluorescent screen to the scale.
The only problem with diffraction gratings is the same wavelength is reflected at a number of angles, so it repeats every number of degrees. At least
as I recall. So you see red orange yellow green cyan blue purple red orange...
Sodium in a gas flame would certainly give you a bright reference. I'd like to see someone clearly photograph the doublet bands on their
homemade apparatus.
Tim
jimwig - 29-6-2005 at 17:34
go to journal of chemical education site
do title search on spectroscope
there are several for the building in the lab - if memory serves
also more amateur constructed scope are to be found in the amateur scientist column in scientific american
there is an online index for that also
google at close range
jim wiggins
chromium - 30-6-2005 at 12:30
Journal of chem education is quite usefull reading. I did not know it before. Thanks jim!
Twospoons - 30-6-2005 at 20:51
If you remove the IR filter in the lens then most webcams will detect infra-red down to 1000nm. So thats one way to view at least part of the IR
spectrum.
The image intensifiers in night vision scopes also work well into the IR, though I'm not sure how far the response extends.
Archimede - 1-7-2005 at 12:59
You can take a screenshot from a webcam and save the picture in a standard size, bmp 16 colors. With a visual basic program it shouldnt be too hard to
find out how much % of each of the basic colors you have in the pictures.
If looking at a printed picture someone is able to determine what element that is, I am sure that it can be done with a simple program on a pc.
unionised - 4-7-2005 at 13:27
The typical sensors in webcams only work in the near IR about 900nm. The spectra that get used for identification by IR cover the region around 10000
nm.
Measuring UV might be easier but most condensed phase UV spectra are pretty dull, a splodge at the high energy end.
"If looking at a printed picture someone is able to determine what element that is, I am sure that it can be done with a simple program on a
pc."
Interesting idea, picture recognition, like voice recognition, is actually quite difficult for a computer. If you find a simple program to do it I
sugest that you sell it to the millitary for lots of $$$. The odd thing is that people are very good at it.
Anyway, there are spectrum matching programs available, the tricky bit is getting the data into the PC.
Twospoons - 4-7-2005 at 14:55
The pyroelectric sensors used in motion detectors cover the 5-14um IR range. I guess some kind of scanning aperture would be needed to pick out
individual IR lines.
That still leaves a hole from 1um to 5um - maybe the old military image intensifiers could be used ? I've got one that uses IR down to 3um IIRC.
[Edited on 4-7-2005 by Twospoons]
unionised - 5-7-2005 at 08:57
Or you could cover a coil of fine wire with soot from a candle flame.
jimwig - 5-7-2005 at 11:49
would you be so kind as to elaborate
aka duh!!!
on the soot idea.
thanks
unionised - 9-7-2005 at 02:56
I seem to be talking bolometers.
that should be enough for you to google.
Oxydro - 9-7-2005 at 21:04
Um... I have to make a quick guess about what such a thing is, before googling. Is it a coil of light-absorbing wire that changes resistance with
temperature, therefore acting as a primitive light meter? I will edit to say, I was wrong, when I find what it is.
edit: crazy - I guess I was somewhat right. However, they are not primative at all - cept in the variety we might make, of course - some are used in
submillimeter astronomy and involve cooling to millikelvins.
edit2: to add a question, what kind of light sensor is used in a IR spectroscope unit?
[Edited on 10-7-2005 by Oxydro]
[Edited on 10-7-2005 by Oxydro]
unionised - 10-7-2005 at 07:07
This might help
http://hiq.linde-gas.com/international/web/lg/spg/likelgspg....
Archimede - 16-7-2005 at 22:47
Voice recognition is quite a bit more complicated than reading a bitmap and count how many pixels of each color you have. And we are not talking about
a complex image. Just a 16 or 256 colors bitmap with a 'rainbow' in it.
[Edited on 17-7-2005 by Archimede]
[Edited on 17-7-2005 by Archimede]
LED's as a light sorce for a Spectrophotometer
Lambda - 17-7-2005 at 00:38
Posted on 12-7-2005 by Lambda:
"The narrow bandwidth of LED's"
https://sciencemadness.org/talk/viewthread.php?tid=4114&...
Addition:
In the UV-spectrum quarz cuvets will be required due to the high absorbtion of normall glass.
In the visible spectrum glass cuvets may be used.
In the IR-spectrum kitchen salt (NaCl) works out just fine. You will have to press the mixed sample into tablets, or.....good luck !
kyanite - 17-7-2005 at 12:54
Oh, I am so confused.
From what I understand from this thread, we can/want to(?) create a primitive IR spectrophotometer, which could be made up of:
an IR detector(bolometer or what not)
a prism/diffraction grating (a cd/dvd)
a sample holder made of NaCl for the plates
and some monochromatic IR LEDs for the light?
What confuses me is why use a picture and software to determine the number of colours. Aren't we after the IR portion of the spectrum spread? How
do you take a rgb picture of that?
Anyways,
My org-chem textbook's IR spectrum graphs go from 2.5 to 15 microns. I suppose thats the range we could be aiming for. Is there any other info to
verify this?
and also pyroelectric detectors(from IR motion detectors) only have such a narrow range of detection because of a filter window. Maybe if we could
remove the window or replace it, we could use it for all the other wavelengths?
Info taken from here
12AX7 - 17-7-2005 at 13:09
AFAIK, you don't have to (and can't! )worry about "color"
(i.e., wavelength expressed as an appearance, which loses meaning outside of the narrow visible spectrum of course!), so you use the spectrometer
which alters wavelength to position on a scale. Then you just read off position of the intensity bands. So yeah, you'd probably get IR and UV
as a monochrome scale photograph, unless you want to use a more sophisticated spectrum such as intensity varying color (heat-sensitive video and radar
images for instance use this), or try to transpose visual colors down a few orders of magnitude - of course, this needs a camera that senses IR bands
individually.
Tim
[Edited on 7-17-2005 by 12AX7]
kyanite - 17-7-2005 at 13:26
Wow, I'm dumber that I thought!
So we can skip all the bolometers and detector crap by using an IR camera...
Brilliant. That makes things so much more simpler!
cd spectroscope
Damn, this looks fun. Easy too.
I think I'm going to take a whack at it.
Archimede - 18-7-2005 at 13:52
I was told that most B/W cameras cover a portion of IR as well, thats why when its very dark you still see some shadows on the screen. I didnt check
any specs to see how good it would be for this.
Also I think that some home video cameras have the night vision button that all it does (I think) converts IR to green so you can record and see in
the dark.
unionised - 21-7-2005 at 10:46
"Wow, I'm dumber that I thought! "
Does anyone remember a few posts back where I pointed out that the bit of the IR spectrum that normal cameras will see isn't the bit usually used
for IR spectroscopy.
What about the bit where someone cited the IR spectra in their textbooks as a good place to start when looking for the range of IR you need (well
done, BTW, and quite correct)?
[Edited on 21-7-2005 by unionised]
Interferometer
Quibbler - 25-7-2005 at 06:19
UV and visible spectra provide very little information. The best method to identify chemicals is nuclear magnetic resonance (NMR) - not really
feasible at home. IR is fairly useful, but It may be better to use another approach to recording IR spectra.
Most IR spectrometers use a Michelson interferometer. It does away with any need for frequency separation (no prism/grating).
So source of IR - glowing filament (light bulb)
Detector - some kind of solid state device should work nicely.
Beamsplitter - a plastic sheet can work
moving mirror - tricky part needs to be precise, its position is usually monitored by a laser.
Have fun!
my nootloss - 25-7-2005 at 09:28
I am sorry but what propose would the moving mirror serve? If a mirror is indeed needed then I think it would have to be a frontal mirror.
And why would we not need the UV and visible spectra? Why not just use frequency separation, and pick up as many wavelengths as we can? And I had
thought that UV was all most the only thing some substances put out. And if IR works better with out frequency separation then we could just let it
pick up the light before it gets separated
12AX7 - 25-7-2005 at 13:53
Said inferometer uses two crossed beams (actually more than two, Google it) and an adjustable mirror to cause adjustable interference. Unless
I'm thinking of the speed-of-light measuring apparatus. I seem to recall both with the same device though...
Tim
Quibbler - 27-7-2005 at 07:07
Michelson did dabble in speed of light measurements. The interferometer was however used to disprove the existence of luminiferous aether.
The method has been sucessfully adapted to measure IR spectra. What happens is the IR beam is split into two. These two beams travel different
distances, then are recombined. If one of the path lengths is varied (moving mirror) an interference pattern is generated at the detector. Now if a
sample is placed in the beam some frequencies will be removed and a different interference pattern will be obtained. The missing frequencies can be
found by fourier tranformation. Sounds improbable, but it works.
Quibbler - 29-7-2005 at 08:35
Ok maybe that sounds a bit difficult. What about NQR (nuclear quadrupole resonance) I made a spectrometer with just one active component (a FET) and
it worked. The computer interface was a little more difficult. Next I am going to try it through the sound card to bypass the interface which is a bit
temperamental.