[Stew_822]

Hi guys,

We recently learnt about spectroscopy in Chemistry and it got me thinking. I was wondering if it would be possible to measure the concentration of ammonia in water using an automated procedure - that I could make, DIY style, using a raspberry pi as a controller. The goal would be to monitor the levels constantly so that the thing could warn me if there is any ammonia detected. I would not like to use reagents, because then I would end up with waste materials that I would have to store in the device.

I am interested in finding the concentration of ammonia, nitrite, nitrate and possibly macro and micro nutrients. I have searched for ammonia probes but they are $300+ and seem to require replacement of the membranes (something I would very much like to avoid).

The water I would be testing is from an aquarium and so will obviously contain other (unknown) substances. Would these interfere with the result if I were to use spectroscopy? For example (using the IR spectroscopy), there would be nitrates in the water - am I correct in reasoning that if I chose a specific wavelength to test for ammonia (that it absorbs strongly in) and the nitrates also absorb in that wavelength, then I will have a skewed reading because I will actually be measuring the concentration of ammonia and nitrate? Would the gases in air skew the result too?

So to summarise my main questions:
1) Is it even possible? If so, what technique to measure the concentration would be best? Could I even build this myself?
2) Would the other substances in the aquarium interfere with the result? How can I combat this?
3) Is there an easier way that I have missed?
4) How accurate do you reckon I could get it? Ideally I would like a 0.05ppm detection, but I would settle for 0.1ppm or 0.25ppm. Some fish die at 0.25ppm concentrations though so if I couldn't get it that accurate it would end up being a project mostly for sh!ts and giggles - but all my projects end up like that anyway!

I'm not too good at chemistry, so please forgive me if I sound a little stupid.

Cheers,
Stewart

[Borek]

Possible - yes. Easy and cheap - no.

IMHO your main problems are the sensor and the lamp. No idea if they are commercially available - I guess they can be recovered from some old system. But in general they won't come cheap.

However, compare http://www.consumerphysics.com/myscio/

And yes, you can expect interference, that's quite a normal situation in all spectroscopic methods. You either have to select a wavelength without interference, or mask the other substance, or extract the substance you are interested in. Alternatively, you can do the measurements at several wavelengths and determine concentrations of all substances involved.

[Stew_822]

Thanks for your reply!

That SCiO looks interesting

I have done some research and some googling, and have found this sensor:
http://html.alldatasheet.com/html-pdf/32197/TOSHIBA/TCD1304AP/2525/10/TCD1304AP.html
(the link is to a datasheet with the spectral response)
It can be found on ebay for about $10. Do you think I would be able to detect ammonia with that sensor? What kind of range do you think I should be looking at? Or should I get any sensor that looks alright, and hope ammonia absorbs in the spectrum that my sensor can sense. I'm not sure where I should start - looking at the absorbtion spectrum of ammonia, picking a wavelength that it absorbs strongly in, and finding a sensor to suit (which may be damn near impossible to do it cheaply), OR buying this $10 sensor I've found, spending the hours needed to rig it up, then testing different concentrations of ammonia to see if I can detect a difference.

The other thing I thought of, does the absorbtion spectrum of a substance change if it is dissolved? If I go with buying the sensor first, this doesn't matter, but otherwise I should be looking for dissolved ammonia absorbtion spectrum graphs.

The plan to deal with interference, then, would probably be to select a wavelength without as much interference as possible and try and compensate by determining the concentrations of the other substances.

I really appreciate your advice

Cheers,
Stewart