CO2 and O2


Junior Poster
Oct 1, 2007
After three months with Mazzei CO2 injection, I've reached a conclusion about CO2 and O2 dynamics.

I've read many times from CO2 mist doubters that the bubbles aren't actually CO2 rich for long, but quickly become filled with other gasses, presumably O2. Without really thinking about it, I figured that was bunk. But is it?

O2 in about 30% of the atmosphere. In freshwater, its about 20 ppm at equilibrium. Clearly the gas partitions into the atmosphere (gas phase) with great preference. So now introduce a CO2 bubble into the water. Initially O2 is zero ppm within the bubble. One would expect O2 in water to partition into the bubble. The question is whether or not this happens quickly enough (fast compared to bubble lifetime). I suppose there's also question about the amount of O2 in commercial CO2 tanks (not zero, but almost certainly very low given the use as a non-reactant in welding and what not).

I suppose one could set up an experiment to determine the rate of exchange. It would be tough given the tiny bubbles involved. Alternatively, one could use fish to answer the question. In my tank, the venturi comes on three hours before the lights. Watching the fish closely, I see that the only time they ever appear stressed is during the last hour of lights-off CO2 and the first hour or so of lights-on CO2. This leads me to suspect (not believe, just suspect) that the CO2 mist is extracting O2 from the water. Until the plants start respiring O2, the ppm of O2 drops and the fish are stressed.

Some extra info that may or may not be relevant. I run an air pump when CO2 is off. So at the start of CO2 injection, O2 should be pretty good (the fish appear to agree). A few hours after lights on, the fish are quite active and aren't gulping. The gulping is limited to a few hours where, if I'm right, O2 would be at a minimum from CO2 mist "scrubbing."

58 gallon tank
192 watts compact florescent (6700 k times 3, colormax times 1)
CO2 mazzei 584 powered by a Eheim 2026
drop checker with 4 dKh that shows yellow-green
second 2026 for extra flow
densely planted using EI with reduced KNO3 given high fish load
12 rummy noses
6 black neon tetras
6 danios
6 large tiger barbs
2 clown plecos
1 yoyo loach (I think. Haven't seen it lately, but it hides a lot)

Tom Barr

Staff member
Jan 23, 2005
O2 is 21% in the air and at equilibrium is it about 6.5-7ppm in our tanks at 82F. CO2 is far more soluble than O2(which is pretty insoluble).
This is why you have differences in the levels at equilibrium.

If you find 20ppm in any FW, it'll be either very cold(you might find 12-15ppm) and there will be a huge massive algae bloom, or both.

There's virtually no reactive gas in CO2.
It's all food grade for the most part. Maybe a little, and I mean very little(I know the CO2 gas guys here pretty well and have asked, and our lab has reagent grade gases- N2, CO2, O2 and He, they dissolve in the same manner)) CO2 is a by product of industrial processing, they collect it and ship it.

I've heard this argument before.
They guy did a lot of speculation, talk, and threatened to test everything and then found out that it takes a lot of work, money and resources:rolleyes:
That sent him running, but he had lots of energy and time to criticize me for some odd reason:p

Always interesting.

I think your idea is interesting(not like the wind bag I mentioned above though with sarcasm, hehe) however...... O2 and CO2 partial pressures do act independently.

So they each act on their own, not dependent on eachother.
You might want to critically measure CO2 and O2 in the system first before hand to get an idea of what is really happening before making a conclusion though.

If your fish are appearing stressed at these 2 points in time, try increasing the current and flow rates, have some rippling, but not break the surface etc.

This will equilibrate O2 much better and then you can look at CO2 independently. If you have high growth and lots of fish in a small tank/glass box, food, waste not clean your filters as much as you should, keep things good and clean, then you run the risk of relying mostly on plant O2 production and that is often when you see the better fish behavior, activity etc.

If CO2 builds up too much, then you might be stressing them.
Seems like this may be the case.

But measuring O2 is generally easier than CO2.
But they make some nice meters for both O2 and CO2 that do not rely on gas consumption and are very accurate. I am getting a CO2 meter soon, (not the 7000$ one mentioned, but something that does the precise same thing, but for 5X less, not cheap, but it's pretty accurate and does not have testing issues like KH/pH methods and it's EPA approved).

I think measuring the gas inside the bubble that rapidly dissolves is a toughy.
How do know which is dissolved and which is in the aqueous phase? Labelling the gas will not tell you that, even if you could track it, it does not help answer that question.

You can sample the microbubbles and take a time series, but that's not easy either. Measuring fish behavior or any sort is tough as well. Lots of experimental bias being added by us. The amount of Gas in side the CO2 bubble will be going out when you add CO2, there might be a little, and very little, gas going in. Now the gas going in to the bubble is not ALL O2, there's a lot of N2, some CO2 will remain as well. As gas bubbles become smaller, the surface tensions go way way up. This seems to make them more resistant to final dissolution. I do not fully understand it myself, but a friend who teaches engineering here discussed it. Mostly in the context of marine systems.

Anyone can add doubt to any discussion, but the clever can test it to fine a way around it. Even if you do this, many will be skeptical.

The doubters claimed it did not help plant growth.

I measured the O2 levels, which is a measure of plant production using a Hach LDO meter, it showed roughly 21-40% higher O2 productions rates vs a control(no mist, same CO2 ppm).

It does not say why the test treatment had higher levels, only that something about the CO2 mist helped.

As far as method that helps grow plants, it's clear to me that it helps.
Is it the CO2 gas phase that helps speed diffusion? Or is it the micro bubble gas breaking up the boundary layers around the leaves?

You could use air as a control or N2 gas even better, and see if you get the same % increase in O2 production.

I know this does not address fish, but a good O2 meter will as well as good accurate CO2 measurement.

I know folks have serious issues measuring CO2.
I do.

I'm not confident even with some of the equipment I have if I can measure to highly accurate levels without a lot of work. So I know any critic who's too lazy to test much themselves will ever bother and be even remotely close to being able to.

I am highly confident with O2 measurements I've got that part and the CO2 meter will address that other part.

Once that is in place, then I can measure the fish response, but then what is it that you measure? Time gasping at the surface? Color changes, reduced activity(if so, what units would you use?), reduced growth rates, feeding?

Maybe they are lazy today?

The critics rarely suggest such things when it comes to fish behavior, again, very interesting that they are so selective and offer little in the way of solutions. I've yet to meet a single person other than George Booth that's ever bothered to measure O2 in planted tanks over a day or more. I have data loggers on a client's tank that tells me the monthly O2 levels(or any other parameter I want really like pH/redox/conductivty)

I know the failings, but I put the idea out there and then go back and pick at it, try and figure out ways around it.

Along the way, I develop many neat ideas, methods, and get a much better understanding.

It's a good exercise to go through this for everyone though.
Even if there is no answer as of yet.

All I do know at this point is there is higher O2 levels, which suggest that something is better as far as increasing growth rates, plant preferences.
I also know I need a very accurate method for CO2 measure that gets around pH/KH.

And I'm on the way there.

Tom Barr


Lifetime Charter Member
Lifetime Member
Nov 21, 2007
Los Angeles, CA
Why are you turning on CO2 3 hours before lights on? The standard recommendation is turning on CO2 one hour before lights on and turning it off one hour before lights off. I know you may already know this, and you may just be doing an experiment.


Junior Poster
Oct 1, 2007
An experiment

Tedr: it was an experiment born of a timer-programming mistake. I'm going back to an hour now that I've watched it for a while and saw no benefit over the standard hour offset.

Tom: Have you checked O2 with lights out when CO2 mist begins? I'm wondering if you could see a drop before plant respiration brings it back up. I suppose the problem here is the plants taking up O2 when not photosynthesizing, presenting a moving baseline. Someday I'll get an ORP meter and will try this myself.

Tom Barr

Staff member
Jan 23, 2005
jsrevenaugh;23444 said:
Tom: Have you checked O2 with lights out when CO2 mist begins? I'm wondering if you could see a drop before plant respiration brings it back up. I suppose the problem here is the plants taking up O2 when not photosynthesizing, presenting a moving baseline. Someday I'll get an ORP meter and will try this myself.

No, I do not add CO2 at night:p

O2 drops slowly during the night, mine peaks about 1 hour prior to lights off(10 hour cycle). It is well above 140% and drops to about 90-95% right before the lights comes on. Take about 2-3 hours to drop to 100% from 140%.

However, aeration does remove high levels of CO2 and O2.
It may occur as you suggest.
I'm not sure, I have not tested for it.
I do not do aeration, nor add CO2 at night.
I see no need based on the data for plants or fish.

There are many variations you can test and measure to answer such questions.
However, you need really good data ACCURACY AND COLLECTION METHODS.

The way I see it, that's the biggest issue for most folks.
Not the method itself.
It works and it works very well.
But if we know why, they we can further hone it and compare it vs other methods.

I do not think an ORP probe will tell you what you want, an O2 meter will.

ORP is good for O3 and for measuring sediment Redox levels.

Tom Barr