1
$\begingroup$

What pathogens or harmful stuff IN WATER can you detect with these homemade video spectrometers (webcam) + UV light?

https://www.youtube.com/watch?v=IA5BTD-aelo

I'm trying to do a cheap spectral analysis to analyze water quality, something electronic, so I can do continues monitoring remotely.

$\endgroup$
  • $\begingroup$ Unfortunately the answer is probably no. Water absorbs very little in the visible spectrum which is good for your application, but absorbs a huge amount in the infra-red and UV, which is not good as it absorbs all the light so none reaches detector. Detection in visible will require visibly coloured pathogens and to get sensitivity, a very long cell of water to pass the light through; by 'long' this could be metres. Detection in near UV (250-350 nm) requires UV transmitting windows and detector sensitive here. $\endgroup$ – porphyrin Mar 13 '17 at 16:52
  • $\begingroup$ @porphyrin so UV light is the problem, what if I use a different kind of light? Will it solve the problem? If so what kind of light should I choose? $\endgroup$ – Leo Zurick Mar 13 '17 at 19:17
  • $\begingroup$ no; its not the light that is the problem, in general you have to have absorbance from substances that you want top study in a wavelength range that you can detect. It is common to detect in the uv with commercial or special lab built instruments but hard at home unless you buy special uv optics and uv detectors. And especially hard for trace amounts of substances if you don't know before hand at what wavelengths to look. Also the change in transmitted light intensity you are looking for may be only 1 part in 1000 or less. $\endgroup$ – porphyrin Mar 14 '17 at 13:19
1
$\begingroup$

In general, with this setup, nothing. You would need a UV detector instead of a webcam. And, assuming you had a UV light source and the right detector, it would depend on the UV source spectrum. You could use a blacklight for mid 300 nm or a Steripen for 254. A broad spectrum lamp would be best. Water doesn't start to absorb UV strongly until you get below approx. 200 nm, so compounds that absorb above that will show up. This would be on a commercial unit like a Hach DR6000. On a DIY kit, I doubt you'll be able to detect much of anything unless it is either at high concentrations (very high) or your path length is long (ie Beers law). With all that being said, UV spec isn't a very good "fingerprint" analysis like ftir or mass spec. For example DNA and many organics all absorb at 254 nm, so you won't be able to differentiate them. Probably the easiest water quality parameter you could measure is called UV254, but that's really only useful for determining organic content and disinfection byproduct formation potential.

Okay, will all that being said, you might have a better option using visible light. Hach sells chemical kits for measuring certain constituents - they are coloromertric so they convert the constituent to a colored compounds that is detectable at a certain visible light wavelength. For example, you could test for nitrate or chloride. A used portable Hach meter will run you around 500, these are basically a fancy visible light spec. They work decent, but not great and even that would be worlds better than any of these DIY kits if that tells you anything.

$\endgroup$
  • $\begingroup$ what's the cheapest way to do it under $100 ? What about this? renaud.schleck.free.fr/spectrometre.php It won't work for water analysis? $\endgroup$ – Leo Zurick Mar 14 '17 at 15:27
  • $\begingroup$ could you tell me more about the UV254 ? is this the one? -> ebay.com/itm/… $\endgroup$ – Leo Zurick Mar 14 '17 at 16:08
  • $\begingroup$ Example of nitrate powder pillow, absorption wavelength is in the documents: hach.com/nitraver-5-nitrate-reagent-powder-pillows-10-ml-pk-100/… $\endgroup$ – prof.kvothe Mar 15 '17 at 13:32
  • $\begingroup$ Yes, that lamp would give you the wavelength you want. It will be monochromatic though, so only 254 nm (well, actually with a bit of 185 nm too). $\endgroup$ – prof.kvothe Mar 15 '17 at 13:33
  • $\begingroup$ Again though, I think you need to be a little realistic here - what type of water are you trying to measure? Tap water? Surface water? Ground water? If tap water, you likely won't have concentrations high enough and if surface water there will likely be too much "stuff" to differentiate anything. So...you might be overestimating the ability of a homemade spectrometer. $\endgroup$ – prof.kvothe Mar 15 '17 at 13:36

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.