Gas-exchange experiments

[Andy Pierce]

You could add HCl to tap water and pour it into the tank, it will not lose the CO2 that quickly. But then you could just make fresh soda water and use that.

I used to use HCl this way as a temporary CO2 booster, but have abandoned that approach.

If the goal is primarily to temporarily add dissolved CO2 to the tank then making fresh soda water and adding that makes sense to me (although just directly injecting CO2 gas is my preferred approach). Presumably you could even pull some water out of your tank and carbonate that. For me the main reason for adding HCl is to permanently decrease the alkalinity. In fact, when I add HCl to tap water before a water change now I subsequently degas the treated water before adding to the aquarium to remove the excess dissolved CO2 so the pH won't be too different - as noted previously an acid-caused decrease of 1 dKH only changes the pH by less than 0.1 units after gas equilibration but by considerably more than that before equilibration. Unlike adding carbonated water where your potential to go dramatically wrong seems pretty limited, I expect you need to be a lot more careful about HCl addition to be sure you don't run all way out of (bi)carbonate and crash your pH.

 

[PeerUnk]

My well aerated tank follows the ambient room CO2 levels pretty accurate.

Well, I need to edit this statement I made. After successfully installing my own pH sensor, the CO2 could not be following the ambient room CO2 reading that close with heavy aeration; The pH (and therefore CO2 readings as it is fixed to the KH) was much more stable.

I found out I had a air leak in the gas chamber next to the sensor. So the CO2 reading was actually following the ambient room CO2 level, only not via the water of the tank, but via the air leak.

So I fixed it and the CO2 reading is much more stable as the pH reading is. Below the last 36 hours. The yellow trend is the CO2 in the gas chamber and therefore of the water. And it is reacting slower on the increase of the ambient air CO2 measured at the air pumps (green line) The pH in orange is faster in it's reading, following the ambient CO2 reading in (green) and increase in pH affected by the uptake of CO2 by the plants during lighting (in blue):

converting those reading all to CO2 in the water, the trends are even more obvious. I deleted the actual readings, as re-calibration still needs to be done:

After calibration I will conduct the soda bottle experiment.

 

[hax47]

converting those reading all to CO2 in the water, the trends are even more obvious. I deleted the actual readings, as re-calibration still needs to be

Interesting data with the parallel pH readings! What formula and constants have you used to calculate CO2 from pH?

 

[PeerUnk]

Interesting data with the parallel pH readings! What formula and constants have you used to calculate CO2 from pH?

Well, I use the formula of Henderson Hasselbalch, currently with a constant of 6.3 (so at 20 *C), see screen dump.

I found a empirical formula which takes in account the temperature. I still have to integrate that in my C++ code for the ESP32.
I've read amounts of other ions which can influence the constant as well, but that's far beyond my chemistry knowledge...

 

[PeerUnk]

Last weeks I'm going slower on the subject, but I've built a user friendly calibration process on the micro computer which measures pH, which took a while. And I made the pH calibration and reading taking in account the temperature.
On top of that I've lowered the noise in the pH readings by:

• eliminating noise caused by the T5 tubes and other electrical wiring.
• using median value out of a set of 200 readings in 1 minute.
• using galvanic isolated current (temporarily from a power bank)

Not there yet entirely as noise could be lower quite a bit. So I still want to shield the pH probe wire and part of the micro computer for EMI/RFI (electromagnetic and radio frequency interference) by creating a faraday cage, using copper pipe and fabric.

I want to finish this as before conducting new experiments on gas exchange.

 

[hax47]

Well, I use the formula of Henderson Hasselbalch, currently with a constant of 6.3 (so at 20 *C), see screen dump.

I am unsure what the right value is. In the Aquatic Chemistry Concepts book (James F. Pankow), these are the temp/pK values:
temperature (°C), pK1, pK2
0 6.58 10.63
5 6.53 10.56
10 6.46 10.49
15 6.42 10.43
20 6.38 10.38
25 6.35 10.33
30 6.33 10.29
35 6.31 10.25
40 6.30 10.22
50 6.29 10.17
(the references: Harned and Scholes (1941), Harned and Davies (1943), Ellis (1959), and Buch (1960).)

Calculating from this recent paper that @Andy Pierce brought to our attention, the pK seems to be 6.53 at 25°C. I wonder which of these values would better fit the measured data. Or, if you fit the pK to fit the pH measurements, what value would you get?

 

[hax47]

On top of that I've lowered the noise in the pH readings by:

• eliminating noise caused by the T5 tubes and other electrical wiring.
• using median value out of a set of 200 readings in 1 minute.
• using galvanic isolated current (temporarily from a power bank)

How much noise are we talking about here?

 

[PeerUnk]

I am unsure what the right value is. In the Aquatic Chemistry Concepts book (James F. Pankow), these are the temp/pK values:
temperature (°C), pK1, pK2
0 6.58 10.63
5 6.53 10.56
10 6.46 10.49
15 6.42 10.43
20 6.38 10.38
25 6.35 10.33
30 6.33 10.29
35 6.31 10.25
40 6.30 10.22
50 6.29 10.17
(the references: Harned and Scholes (1941), Harned and Davies (1943), Ellis (1959), and Buch (1960).)

Calculating from this recent paper that @Andy Pierce brought to our attention, the pK seems to be 6.53 at 25°C. I wonder which of these values would better fit the measured data. Or, if you fit the pK to fit the pH measurements, what value would you get?

Interesting! Well, considering the Henderson Hasselbalch formula:

CO2(mg/L) = KH(dH) * 1000 * 0.000357 * 44.009 / 10 ^ ( pH - pKz )

there are 4 factors, of which 2 of them are measured by sensors: CO2 and pH. (Or we could fully degass the water outside, measure the pH and put CO2 as 419 mg/L in the equation)

But still we have KH and pKz as factors... I have used pKz = 6.3 to determine the KH. Assuming this would be the end of that story for me. Now you come in with this question...

But my chemistry knowledge falls short on this matter. I always understood KH is the alkalinity as a big part in acid buffer capacity, which is really the driver here. But I don't know whether other molecules which play their part in acid buffer capacity, the ionic load and the formula. For instance: When I have RO water of 5 uS/cm, and I would remineralise leanly with CaSO4 and MgSO4, I might/could measure CO2 and pH (and know HCO3 in the alkalinity/KH is low). I might calculate pKz. And when I would add a known simple macro mix of 10 parts KNO3 and 1 part KH2PO4, so N becomes N=4 mg/L. Still HCO3 might or should be very low and CO2 should be the same, but I wouldn't be surprised the pH changed as these added compounds play their part... Again, my knowledge falls short on chemistry. Someone even might comment this formula is useless with low alkalinity... I don't have the knowledge or arguments to defalsify that.

I could used my drop test to determine KH, but it presumably would not be accurate enough to respond to you what pKz would be in my case.
Or in other words, When I fit the measurements of CO2 and pH (which I did), do we know enough to really say something specific about pKz and/or the KH?
It became clear to me aware this is a hard topic and even the scientific community is still busy with it.

Thanks you for your initial comment!

 

[PeerUnk]

How much noise are talking about here?

I've determined noise is pretty stable at 0.14 pH. But the pH shifts in time as a trend is clear. Please see screen dump.


I'm not try to hijack this topic on gas experiments...
At 12:36 (blue line below) I've take the probe out of the water and put it in a jar at the same spot (a few centimeters above it to be more precise), and instantly the noise it a lot less:

So an educated guess a lot of noise is created via the water. In the water I only have 2 flow pumps (and a grounding wire), I will see what happens when I disconnect them one at a time.

 

[hax47]

But my chemistry knowledge falls short on this matter. I always understood KH is the alkalinity as a big part in acid buffer capacity, which is really the driver here. But I don't know whether other molecules which play their part in acid buffer capacity, the ionic load and the formula. For instance: When I have RO water of 5 uS/cm, and I would remineralise leanly with CaSO4 and MgSO4, I might/could measure CO2 and pH (and know HCO3 in the alkalinity/KH is low). I might calculate pKz. And when I would add a known simple macro mix of 10 parts KNO3 and 1 part KH2PO4, so N becomes N=4 mg/L. Still HCO3 might or should be very low and CO2 should be the same, but I wouldn't be surprised the pH changed as these added compounds play their part... Again, my knowledge falls short on chemistry. Someone even might comment this formula is useless with low alkalinity... I don't have the knowledge or arguments to defalsify that.

I could used my drop test to determine KH, but it presumably would not be accurate enough to respond to you what pKz would be in my case.
Or in other words, When I fit the measurements of CO2 and pH (which I did), do we know enough to really say something specific about pKz and/or the KH?
It became clear to me aware this is a hard topic and even the scientific community is still busy with it.

You are probably right. Things get complicated when the ionic composition/strength changes, and with that, the pK might change as well. The alkalinity should not change because of these ions though.

Even if RO water is used, bicarbonate will be produced, which can be calculated from CO2/pH, provided that we know the pK 🙂. However, probably a better approach would be to add a known amount of bicarbonate to RO water and measure pH and KH.

 

[hax47]

I've determined noise is pretty stable at 0.14 pH. But the pH shifts in time as a trend is clear. Please see screen dump.

Thanks for the detailed information. I knew that electrical noise was an issue when monitoring pH, but I had no idea to what extent.