Sciencemadness Discussion Board

UV/VIS Spectra

Ninja - 9-12-2014 at 09:06

Hi!

Where can i get good UV/VIS spectra on the internet?
I want to build a spectrophotometer myself with a simple 254nm UV source.
What do you think? Might it be a useful tool?
Thanks for any suggestions.

forgottenpassword - 9-12-2014 at 09:21

UV maxima are reported in scholarly journals as a matter of course. That might be the best place to look for information. There are books that give print-outs for pharmaceuticals and pesticides etc., but they are perhaps not the data that you are looking for?

[Edited on 9-12-2014 by forgottenpassword]

unionised - 9-12-2014 at 12:26

If it's just a 254 nm source you won't get much ofa spectrum.
You won't, for example be able to measure absorption at 253 nm or 255nmm or any other wavelength.

you would need broad bans source.
Deuterium lamps are commonly used. I think a xenon ark would also work.
(some so-called "xenon" car lamps are not really xenon lamps)

SupaVillain - 18-12-2014 at 11:09

Look at my DIY FTIR thread here and the GC/MS and raman spectrometer threads on other sites online, you will find it is very possible to make these spectrometer machines but also you may find good info on tips to make it less expensive. You can also see in my thread how I came across info in research for tips on how to research your own project, i found looking at teardowns of the machines and finding out the industry terms for everything helped the most

also if you look at the UV/vis wikipedia page i think it says some have a beamsplitter to compare the beams, and some you will just compare the data with the sample in the UV light path and the data you get without having the sample there at all. So if you can do without the beamsplitter, i highly suggest you do without it because they are insanely expensive (at least for IR, maybe not for UV though...)

diddi - 18-12-2014 at 14:09

sorry ninja, but a spectrometer uses a variable wavelength source to produce the output you see. each frequency is directed thru the sample and the transmittance (absorbance) is measured. this is what produces the graph. with a single wavelength of light, your graph will only have 1 point on it rather than a spectrum of points. you need a rethink I am afraid :(

phlogiston - 19-12-2014 at 13:28

Another solution to make a dual beam spectrophotometer that is used in many good instruments and maybe easier/chepaer to build than a beamsplitter is to use a chopper and rotating mirror to alternately measure the sample and the reference cuvet.
The light source must not significantly vary in intensity on the time scale of a measurement, but that is typically the case.

Ninja - 22-12-2014 at 11:40

Quote: Originally posted by diddi  
sorry ninja, but a spectrometer uses a variable wavelength source to produce the output you see. each frequency is directed thru the sample and the transmittance (absorbance) is measured. this is what produces the graph. with a single wavelength of light, your graph will only have 1 point on it rather than a spectrum of points. you need a rethink I am afraid :(

No i want to analyse the spectrum of the fluorescence.
By using 254nm i suppose many aromatic compounds and not so many organic compounds will show fluorescence.
But how many fluorescent compounds are there?
How can i know there will be a useful fluorescence spectrum before trying it out?
i might it combine it with the analysis of the transmittance of the light emitted by my fluorencent lamp. Or i might use a phosphor of multiple ones which can optionally give a wide spectrum of >254nm light.
That's basically my idea, because xenon arc lamps and alike are extremely expensive.
Please provide me with further ideas to build a cheap fluorescence spectrometer.:)

IrC - 22-12-2014 at 21:05

Quote: Originally posted by Ninja  
Please provide me with further ideas to build a cheap fluorescence spectrometer.:)


I realize you are new here but asking to be spoon fed is not well liked. People have their own jobs, lives, families, school, etc., and expect someone to show they have put forth real effort before asking others to do their work for them. Mind you I am not trying to scold, rather to inform so you will have a better experience on SCM. As example, entering the words fluorescence spectrometer into a Google search came up with some interesting links. While you may have already done this search, based upon your posts it did not sound to me like you have tried searching all that hard. For example:

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

http://en.wikipedia.org/wiki/X-ray_fluorescence

Attachment: Fluorescence Spectroscopy.pdf (315kB)
This file has been downloaded 618 times

Attachment: Fluoromax-4.pdf (1.8MB)
This file has been downloaded 411 times

Attachment: Fluorescence brochure 012309.pdf (8.8MB)
This file has been downloaded 664 times

These are in the first few hits. Often it helps to study designs of working models for ideas and greater understanding.

These links should give you things to study and many ideas, such as using several LED's of varying frequencies and high output, relatively cheap I might add. No doubt you will think of other ideas besides LED's which are also within a reasonable cost by doing further searching.

smaerd - 23-12-2014 at 09:11

The word's cheap and spectrometer just don't go hand in hand unfortunately. So you're interested in fluoresence? Do you have a specific compound which fluoresces that you are interested in?

For example if you know the emission of the fluorophore and that it is excited at 254nm, you could build a simple fluorimeter. Excite the sample with your light source in a quartz or UV transparent sample cell. Collect the scattered light after filtering it through damn near anything to soak up the UV source (glass, plastic, whatever). Use a detector that is somewhat specific to the sample. For example say your compound emits at 450nm. Find a cheapy detector (photodiode) that is pretty sensitive in that range. Maybe amplify the signal if it is weak.

Irc I like the idea of using LED's. Do you know if they have the same band-gap or spectral range if they are reverse-biased (IE: as used in detection rather then excitation)? I don't see why they wouldn't, same semiconductor, different direction. Could provide an idea for a ~15nm range cheap detector that could be swapped out for very crude analysis. Then again a diffraction grating, and CCD would be infinitely better.

IrC - 23-12-2014 at 11:47

Quote: Originally posted by smaerd  
Irc I like the idea of using LED's. Do you know if they have the same band-gap or spectral range if they are reverse-biased (IE: as used in detection rather then excitation)?


I do not know. Not something I have ever had a reason to investigate but it sounds like an interesting enough question I think I will look into it. Not even sure where to look for I cannot recall reading about such a study although of course much mention has been made about the ability of a LED to detect photons.

smaerd - 23-12-2014 at 15:19

Yea I've never heard of it being done either. Not to say it hasn't but I just haven't. Could provide a cheap means for a 'narrow' band detector.

Could do a quick test of say using an LED in reverse bias to detect. Then read voltages from the same type of LED incident on it, one/two of higher wavelength, and one/two of lower. Hmm I do have a few LED's sitting around, maybe a megaohm resistor somewhere?

This could be interesting :). I'll see if I can do some really crude testing tonight, why not. Although the LED may detect incident light from another LED source, if it were to be used for a fluorimeter type detector, an amplification stage and/or visible light collimating lens would likely be required. A collimating lens could also serve as a UV filter.

Couldn't do all the fancy analysis with such a device either, but if you made a compound which excites at 254nm, and emits at say 500nm in it's protonated form, and 700nm in it's deprotonated form. It could end up being a useful little tool.

[Edited on 23-12-2014 by smaerd]

smaerd - 23-12-2014 at 16:17

Alright here was the result of my basic test.

Experimental design: An LED was used as a detector by wrapping it's legs around multimeter probes. A series of LED's as emitters were wired to two AA batteries with a small resistor(270ohm) in series. 'Maximum' voltages were collected when the emitter was in close proximity to the detector.

505nm LED as detector
Emitter - 440nm, 0.0915V 300mcd
Emitter - 505nm, 0.0205V 9000mcd
Emitter - 590nm, 0.000V 3500mcd

590nm LED as detector
Emitter - 440nm, 0.0027V 300mcd
Emitter - 505nm, 0.1057V 9000mcd
Emitter - 590nm, 0.945V 3500mcd (note different amount of digits due to different multimeter setting)

I'm not really sure what to make of the results to be honest. Maybe I'll buy some more LED's of the same viewing angle, and power ratings so it's more controlled. Then again I am kinda lazy.