CO2 test method development

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113
I was looking and have been for some time, for various responsive methods to measure CO2 in water.

Most methods involve some transition from water to the air. We add gas, then it dissolves(we assume all of it), then it gets to the plants or off gases to the air above.

Test kits, pH meters, colorimeters, drop checkers, and color charts are not that reliable.

We also have O2 and light PAR issues, so even say 30ppm might not be enough or might be too much depending on those 2 parameters.

Still, I'm looking for some reliable method that would yeild +/- 1ppm or within 1ppm. Needs to be cheap and perhaps integrated into a controller type function.

I think I've found a good method that's within 200-600$ range.
Takes a little DIY and is a bit like a large Drop checker in some ways. I'd still need to verify it against a reference to be sure the method is accurate and responsive.

Method 1:
pH reference cell across a membrane is one method(Basically the KH ref of 4 degrees is used to wrap the pH probe tip) and using the selective gas membrane to keep the KH solution in, but allows CO2 and other dissolved gases across.

Method 2:
Using a relative inexpensive air/gas IR analyzer to measure gas CO2 inside a bell that is inverted with a gas bubble at the top to withdraw the degassed tank water. No it will not be 30ppm in the bell:eek: 1500-2500ppm is more the range for us.

This method also allows very low levels of measure with high accuracy. It as well as method 1 are still limited by Fick's 1st law of diffusion across the boundary layer between gas/liquid phases.

Since the gas does need to go back into solution like in the Drop checker, the response is faster, and with a larger bell, we can get faster response times and a more even average of the tank since the area is much(can be). the gas IR can measure fast and very accurately, as well be linked to a controller function. It typically uses the air as a reference to self calibrate.

These run 200-600$ or so on line.
The bell collection needs DIY'ed to the tank and then a sample hole that's sealed needs added to the gas bubble area.

I'm not sure it would work, or not.

Seems like it would be worthwhile to do/try.
Using degassed CO2 seems like a good method and has much higher % accuracy and the ppm's are much higher for gas. You could also use O2 for this method as well if you used an O2 gas analyzer.

They also make human health CO2/O2 analyzers that might be able to do the trick for both CO2 and O2 pretty accurately in liquid/gas phases perhaps.

Regards,
Tom Barr
 

Philosophos

Lifetime Charter Member
Lifetime Member
Mar 12, 2009
1,346
0
36
Method 1 is something I've seen you and VaughnH (Hoppy for anyone from TPT) talk about. I'd love to give one of these a try eventually. If they could be manufactured to fit with a few common brands of pH meter, I'm sure they'd change the hobby for the better. Even +/- 2.5ppm would be significantly more meaningful to the average hobbyist than playing with a drop checker, especially given the increased responsiveness.

Your second design seems like it'd be a bit more expensive, but if it handles low range better then I can see the appeal. What sort of response times would it give compared to the membrane? Could it be moved around within the tank as easily?

-Philosophos
 

ordloh

Prolific Poster
Jul 30, 2009
39
0
6
37
Singapore
Method 2 is a great idea. the IR analyizer might be DIYable with an IR source and a detector. Not sure how to control the IR source wavelength though without a prism.
 

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113
The IR part is not much of an issue, the chamber that collects the gas and uses partial pressure relation is the only real issue. I'd also like the bubble some reference CO2 at a known ppm to test it also.

Regards,
Tom Barr
 

jonny_ftm

Guru Class Expert
Mar 5, 2009
821
2
16
First, great project

I also find method 1 technically easier and very interesting. I see some major issues with it: it will need more time to react than method 2 which will lead also to fluctuations due to CO2 going on/off. We know it can get very low in 30mn in well aerated aquariums while it could take 2h to show in the KH4 solution. It will cause 1h30 gap with low CO2 before probe reacts. Once the probe reacts, CO2 will be injected. It will take 30mn to 2h to diffuse well in the aquarium, thus leading to +2h low CO2 fluctuation periods. Also, the most sensitive PH regulators have a +/-0.05 PH trigger before switching the control on or off. With a KH4 solution this leads to an adjustment of +/-6ppm CO2 at the best. Finally, the biggest problem will be deviation and the need for a regulat calibration of those PH electrodes

For method 2, we will need a calibration to adjust readings between the CO2 concentration in the bell and the CO2 concentration in the aquarium. Again, using a mist effect can affect drastically this correlation, also the flow near the bell and also the concentration of other gases, mainly oxygen. I'm really not sure that a correlation can be found with method 2 to the 1ppm precision
 

Biollante

Lifetime Charter Member
Lifetime Member
Jun 21, 2009
3,210
3
36
Surprise, AZ
Co2 & N2

Hi,

If I understand method 2, great precision should be possible by calibrating against a known mix of nitrogen and carbon dioxide.

Not being very smart, :eek: I know I tend to irritate, that is not my intention.

Biollante
 

ordloh

Prolific Poster
Jul 30, 2009
39
0
6
37
Singapore
It might be possible to just pass the IR through a column of water itself, provided that the detector is sensitive enough and you can tune the source down to a very specific wavelength. I'm not sure if 30ppm is a high enough concentration to detect though, maybe a long path length can increase the sensitivity. I think a detecting CO2 concentratinos in a gas bubble above water will be pretty tricky since CO2 solubility and the vap pressure of water both vary with temperature.
 

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113
Gas ppms[g] and water [aq] concentrations are vastly different.

370ppm for air for CO2, 210,000 ppm for O2.

Water? about 0.5ppm CO2 and only 7-8ppm of O2(or close to these ranges).

CO2 is very soluble, O2 is not in water.
Less space too.

Oxygen Solubility in Fresh and Sea Water

Salinity also affects things.

Air has something like 20 times more oxygen in it than the same volume of water.
I'm trying to recall where I found that info:confused: Oh well......

Regards,
Tom Barr
 

ordloh

Prolific Poster
Jul 30, 2009
39
0
6
37
Singapore
Yeah that's why i don't like thinking in ppm, I prefer molarity, more useful for calculations too. 30ppm of co2 in water works out to be about 0.00068M? Sounds pretty detectable. I couldn't find any databases for absorption coeff. of CO2 but i found this Determination of the molar absorption coefficient for the infrared absorption band of CO2 in rhyolitic glasses -- Behrens et al. 89 (23): 301 -- American Mineralogist
CO2 at 2349 cm-1 is 1214, so at 30ppm with a path length of 1cm, you'd get absorbace of 0.82. nice number within the detectable range:). I think 2349 is pretty unique to CO2.
 

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113
Well one system has a solvent, the other is just a gas mixture.
Lots of differences there.

Nothing that cannot be out done.

As far as Molar......0.7-0.8 mM is about 25-30ppm of CO2[aq].
So a range of 0.01 mM to 2 mM would be the detection range, with a +/- of say 0.01 to 0.03mM.

Yes, the IR band spectra is pretty unique to CO2.

Regards,
Tom Barr
 

ordloh

Prolific Poster
Jul 30, 2009
39
0
6
37
Singapore
Ah, then that eliminates the need for the air bubble :)
wish i had the money to try these things out. Good luck with the project, subscribing and keenly awaiting the results.
 

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113
ordloh;43214 said:
Ah, then that eliminates the need for the air bubble :)
wish i had the money to try these things out. Good luck with the project, subscribing and keenly awaiting the results.

Kind of an "if", not sure how well it will work out.

Seems like this is certainly the way to go for CO2 specific measurement detection for both precision and accuracy, a sore spot in the field, as well as for aquariums.

Regards,
Tom Barr
 

billb

Junior Poster
Jun 8, 2009
7
0
1
Hi Tom,

I may not have fully grabbed onto the concept of method 1, but I thought that is basically what most of the cheap CO2 electrodes do. I understood them to basically be pH electrodes within a buffer sollution separated from the sample by a selective membrane. Electrodes direct has some designed to be built in line
pH Electrodes Products. I am not sure about the accuracy though.

Bill
 

Biollante

Lifetime Charter Member
Lifetime Member
Jun 21, 2009
3,210
3
36
Surprise, AZ
DIY Sub US$ 350 ? Maybe

billb;43251 said:
Hi Tom,

I may not have fully grabbed onto the concept of method 1, but I thought that is basically what most of the cheap CO2 electrodes do. I understood them to basically be pH electrodes within a buffer sollution separated from the sample by a selective membrane. Electrodes direct has some designed to be built in line
pH Electrodes Products. I am not sure about the accuracy though.

Bill

Hi Bill,

I think the object is to do what the electrodes and so forth that cost US$ 1,000 plus for US$ 200-400 instead.

I do not have the brand you reference, but they seem accurate, as a result of questions (objections?) Tom Barr raised, I had a fellow with appropriate letters behind his name verify my equipment, storage, processes and calibration procedures and tells me (certifies) accuracy for my purposes as a minimum of 0.08 ppm and technically I could reach 0.02 ppm which is significantly better than I had anticipated frankly +- 0.1 ppm makes me happy, for hobbyist purposes I figure +- 1 ppm is good.

If I understand it is Method 2 that really seems to have the greatest hope for the best accuracy at a low cost, I think around US$ 300-350.

It is possible we could pick up the Original Equipment Manufactures (OEM) NDIR (non-dispersive infrared) CO2 sensor. We might be able to knock the price down under US$ 200.

I am seriously unsure of the calibration or reference samples, but if I understand the process it is not terribly difficult to get a good N2 CO2 reference sample.

I am giving it a shot with a Extech Desktop Indoor Air Quality Carbon Dioxide (CO2) Monitor - Mfg# CO200. It is going to take some serious messing with, but I figure I have equipment to tell me how accurate I can be with this thing. :)

Biollante
 

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113
billb;43251 said:
Hi Tom,

I may not have fully grabbed onto the concept of method 1, but I thought that is basically what most of the cheap CO2 electrodes do. I understood them to basically be pH electrodes within a buffer sollution separated from the sample by a selective membrane. Electrodes direct has some designed to be built in line
pH Electrodes Products. I am not sure about the accuracy though.

Bill

pH Electrodes Products

Well, it's okay, but uses the same idea, response time is good, but the range is a bit weird. Most ion probes have some interferences, you can do secondary treatment to remove those, but then it's not a direct measure in the tank, you have to take a subsample...........treat that with some chemical to remove the interference, then measure.

Not clear on how accurate the thing is.


Regards,
Tom Barr
 

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113

Tom Barr

Founder
Staff member
Administrator
Jan 23, 2005
18,699
786
113
This product is extremely accurate, likely extremely $$$:

http://www.martekinstruments.com/pdfs/co2.pdf

I like the accuracy ranges though.
But it's only over a 10ppm range, or 10000ppb range
If it went to say 200 ppm, that would be cooler.

Ideally, a slight change in CO2 to say 10ppb or so for lower resolution accuracy, 0.00 to 200.00 ppm would be oaky

Response times of 30-60 seconds

Seems possible to me.


Regards,
Tom Barr