A Deep Dive on Ammonia Neutralizer Chemistry - Prime, ClorAm-X, Rongalite and friends.

[taricha]

Are you going to get into the chemistry behind your disks

Nah,
I'll wait go too much into the weeds here except to say
@Dan_P And I spent a lot of time and effort from the point of view that maybe the NH3 films in the seneye and seachem disks are fooling us. But we found no interferences or interesting internal mechanics with the chemicals in question.
So no I didn't find any relevant internal chemistry to discuss.

I wouldn't describe the seachem disks as accurate in anything other than a qualitative sense. Dan figured out how to turn the seneye into a precise NH3 device.

Details on the seneye calibration and measurement work by @Dan_P can be found in this thread.

With less precise methods, we kept observing no evidence of a reaction. Once Dan worked out a way to get a much more precise NH3 value, then the actual ammonia removal trends could be observed, and that's what you see in charts like Fig 3, and Fig 5.
(if you've got your hands on an Ion selective electrode, I'd be interested to see a similar exercise. I expect it will look a lot like what Dan measured via the seneye.)

 

[Formulator]

if you've got your hands on an Ion selective electrode, I'd be interested to see a similar exercise. I expect it will look a lot like what Dan measured via the seneye.

Its in the mail . Along with a hach reagent set and a distillation apparatus. And I have a UV/Vis spectrophotometer.

On a more general note, I’m really excited about the work you and @Dan_P have done on this and can’t wait to read the rest of your write-up. I’m especially interested to see the chemistry and mechanism behind the “detoxification” reaction since you seem to have identified the active compounds. If we know what the product of the reaction is intended to be, we could even make a set of standards to really test out the accuracy of the different test methods.

Thank you for all the work you have put into this to help the hobby!

 

[taricha]

previously, part 1 & 2

Part 3: Why/How does ClorAm-X work? Chemistry of HMS

Hydroxymethanesulfonate (HMS) - as its sodium salt - is the chemical compound that is provided in formulations like ClorAm-X and Amquel, and as we’ll discuss later, is a component of Prime. The chemical formula is CH3NaO4S, and is also known as Sodium Formaldehyde Bisulfite. The name …Formaldehyde Bisulfite is relevant as (partial) dissociation into formaldehyde and sulfite is both a process that occurs under aquarium conditions and is crucial to the intended function of this chemical.

Note from the NaOH in equation 1 that this dissociation is a pH changing process and is reversible, so pH would be lowered by the HMS dissociating, and raised by it forming from sulfite and formaldehyde.

One relevant consequence of the dissociation behavior of HMS is that the compound itself is not very reactive with chlorine, chloramine, ammonia etc. It must dissociate into the parts to become reactive. Sulfite is an immediate dechlorinator, reacting with chlorine or chloramine to become fully oxidized to sulfate [1].

The formaldehyde part of HMS serves no purpose with chlorine or chloramine except in a scheme to bind ammonia, discussed later.
In Fig. 6 below, a demonstration is given of the pH-dependent dissociation that the dechlorination relies on. A 10x dose of ClorAm-X (320 mg/L) was added to tankwater, and variable amounts of the treated tankwater were used to dechlorinate a fixed amount of bleach ( approx. 0.002 mmoles) with starch-iodine as the indicator. What’s plotted is the time that it takes for the different amounts of the treated tankwater (and thus different amounts of HMS) to dechlorinate and decolorize the indicator at different pH levels.

Figure 6. The amount of HMS needed to quickly dechlorinate decreases as the pH rises. At higher pH (yellow and green data) the HMS dissociates more quickly/thoroughly and the sulfite released dechlorinates rapidly.

Note that at pH 9, the dechlorination is very rapid at around the expected 1:1 molar ratio - that is about 0.002mM of HMS dechlorinates 0.002 mM of bleach almost instantly. Also note that the amount of treated tankwater (plotted on x-axis as mmoles of HMS) needed for rapid dechlorination increases with each 0.5 pH unit decrease. At lower pH, a smaller and smaller portion of the HMS rapidly dissociates into available sulfite and formaldehyde, and the dechlorination depends on this release of sulfite. At lower pH, most of the HMS remains as the compound and does not react with chlorine, requiring much larger than the theoretical 1:1 molar ratio to dechlorinate rapidly.

This also indicates that at aquarium-relevant pH, much less than the full amount of HMS is ever released as sulfite and formaldehyde at any one time. Proportionally the large majority remains as the HMS in this equilibrium. This pH-dependent dissociation is also expected from published literature on HMS [2]. The degree of dissociation can be estimated by adding the dose of the solid HMS to aquarium water and measuring the pH drop. The pH drop is small, in the ballpark of ~0.1-0.2 pH units for the recommended amounts of ClorAm-X and measured as an alk consumption, can be calculated to be around the ballpark of ~5% of the HMS at pH 8. In addition to the initial pH lowering dissociation, because this is an equilibrium - any processes in the water that consume some of either the sulfite or formaldehyde (such as reactions with dissolved O2) will lead to further dissociation and a small continued lowering of pH over time. In the Kuhns patent [3] for aquarium ammonia removal by HMS, the data using Ion Selective Electrode to measure removal of NH3 in the first minutes to hours is mostly accounted for by the pH drop from the partial dissociation of HMS.
This pH drop in the first seconds to <1hr should not be confused for the reaction that actually binds ammonia. In Fig. 7 you can see that the early pH drop proceeds nearly the same with or without any ammonia present at all.

Figure 7. This 10mM is ~40x recommended dose of ClorAm-X. The pH drop in saltwater over the first hour is very similar with (open orange data) or without ammonia (solid blue data).


So what about the actual ammonia-binding? While the HMS dechlorinates chlorine and chloramine by the sulfite released, there is equimolar formaldehyde being released as well, which is presumably crucial to the ammonia binding process.
The expected reaction between HMS and ammonia is described in the thesis by Kathryn Brown [4] (see scheme 1.36 in the reference). The HMS dissociates into sulfite and formaldehyde, the formaldehyde reacts with ammonia, and goes through a couple of intermediates, then grabs a sulfite back in the end to become aminomethylsulfonate H2NCH2SO3- . This product is semi-stable, in much the same way as the original hydroxymethanesulfonate (HMS) is, and is expected to also fall apart if the pH is too high. Each step of the reaction should also be thought of as an equilibrium, with the reverse process occurring as well. Each part is necessary - without the sulfite available, the brief association between ammonia and formaldehyde does not last and causes no ammonia decrease.

Fig. 8 below illustrates that you can get the same result in removing total ammonia with HMS if you had dosed the dissociated components - formaldehyde and sulfite - separately.

Figure 8a (left): Ammonia reduction after 48 hours from formaldehyde alone, formaldehyde + sulfite, and Hydroxymethanesulfonate (HMS) - ClorAm-X.
Figure 8b (right): the amounts used in each test.

The HMS behaves the same as if you dosed its dissociated parts separately, supporting the idea of the equilibrium in practice between the compound and its dissociated parts.


Part 4: How are other products different? Chemistry of Rongalite compared to HMS

Thus far we have focused mostly on Hydroxymethanesulfonate (HMS), the ingredient in ClorAm-X as it is the most measurably effective ammonia remover of these types of compounds, but there is another commonly used aquarium chemical for these dechlorinator/ammonia remover applications - Hydroxymethanesulfinate also known as Rongalite. These are confusingly similar, differing by just one oxygen - at least one aquarium ammonia remover product MSDS lists the name of one compound, but the CAS number of the other. Despite many similarities there are a few important differences, a table below attempts to organize some relevant properties. (The sodium salts of each are the most common and are what is in aquarium usage)

Table 1. A list of relevant properties of the very similar compounds HMS and rongalite, used in aquarium ammonia remover products.

One of the relevant differences for aquarium applications is that rongalite is a more effective dechlorinator than HMS. A mole of rongalite can dechlorinate two moles of chlorine/chloramine, while HMS only dechlorinates in a 1:1 ratio. This is because of the different adducts with formaldehyde in each. HMS has sulfite attached to formaldehyde, and the sulfite, SO32- can only be oxidized once before going to sulfate SO42- and becoming essentially non-reactive. Rongalite has a sulfoxylate, SO22- attached to formaldehyde, and it can be oxidized twice before becoming sulfate.
The other way in which Rongalite is a superior dechlorinator is that it reacts with chlorine rapidly regardless of pH, whereas HMS only reacts with chlorine to the extent that it dissociates and releases a sulfite, which is a pH-dependent process as illustrated in fig 6. This means that in an aquarium context, Rongalite will complete the dechlorination reaction in seconds, and HMS will take minutes, perhaps hours if the pH drops to near 7 or below. (Similarly, an iodometric method distinguishes between HMS and rongalite by the fact that HMS doesn’t react with iodine, I2 but rongalite does [5])
In an aquarium, there are a few different pathways that rongalite can actually become HMS.
Some of these can be complicated, but rongalite could get one oxygen and become HMS directly as suggested in the Tetra rongalite patent [6]. Or the sulfoxylate - if freed - can be oxidized to form a number of different intermediates, including dithionite [7], thiosulfate, and sulfite [8] before ending up as sulfate. Interestingly, some of these intermediates (dithionite and sulfite) react with the formaldehyde left behind by the sulfoxylate to form HMS [5]. The purpose for mentioning all of these complications is that in an aquarium context, we can expect that an addition of rongalite over time to form some amount of HMS and its dissociated parts, so the pathway for ammonia removal by rongalite might simply be the same pathway as that for HMS (described in part 3).
How are these two compounds, HMS and rongalite removed from aquarium water over time? The sulfite and sulfoxylate parts of these can react with dissolved oxygen over time to eventually become sulfate - both products are known “oxygen scavengers” that remove oxygen from solutions. The formaldehyde portion can de-gas out from the water (slowly), it can also be broken down by bacteria like many other organics, or it can react with dissolved oxygen first to become formic acid, which is then broken down by bacteria. All of these processes (except de-gassing of formaldehyde) consume oxygen and so the hobbyist should expect some decrease of dissolved O2 from the aquarium water if not aerated.
Below is some data to give a sense of the amount of O2 consumption that might be expected from high doses of these chemicals if tank water is not aerated. The O2 levels were measured in 500mL of tank water covered and un-aerated, with nothing, 10x, and 30x doses of ClorAm-X (HMS).

Figure 9. Dissolved O2 levels measured with a D.O. probe over 1 day for zero (blue) 10x (red) and 30x (yellow) doses of ClorAm-X .

This data illustrates that oxygen is consumed as these compounds break down in the water over time, and aeration should be in place. If we assume that the majority of this O2 consumption is from sulfite reacting with O2 to form sulfate, then the removed O2 represents a reaction involving about 3.5% of the total sulfite in the 10x and 30x ClorAm-X doses. If any of a number of factors were changed: water movement, substrate included, pH kept at 8+, mature biofilms, etc, the breakdown would be expected to be faster. If the compound had been rongalite instead of HMS, the sulfoxylate would have twice the capacity to remove O2 compared to sulfite, so the O2 removal might be expected to be larger.

(In a day or 2 - Part 5: Why do other Products not work? Measurements of product ingredients.)
Part 5: here

References
1. EPA. (2000). Wastewater technology fact sheet dechlorination. Environmental Protection Agency. https://www3.epa.gov/npdes/pubs/dechlorination.pdf

2. Song, S., Ma, T., Zhang, Y., Shen, L., Liu, P., Li, K., ... & McElroy, M. B. (2021). Global modeling of heterogeneous hydroxymethanesulfonate chemistry. Atmospheric Chemistry and Physics, 21(1), 457-481.

3. Kuhns, J. F. (1987). Method and product for removal of chloramines, chlorine and ammonia from aquaculture water. U.S. Patent No. 4,666,610. Washington, DC: U.S. Patent and Trademark Office.

4. Brown, K. H. (1999). Kinetic studies on the reaction of formaldehyde with amines in the presence of sulfite (Doctoral dissertation, Durham University).

5. Danehy, J. P., & Zubritsky, C. W. (1974). Iodometric method for the determination of dithionite, bisulfite, and thiosulfate in the presence of each other and its use in following the decomposition of aqueous solutions of sodium dithionite. Analytical chemistry, 46(3), 391-395.

6. Gunter, R. (1988). Agent for the elimination of active chlorine compounds from water. U.S. Patent No. 4,786,434. Washington, DC: U.S. Patent and Trademark Office.

7. Makarov, S. V., Horváth, A. K., Silaghi-Dumitrescu, R., & Gao, Q. (2016). Sodium dithionite, rongalite and thiourea oxides: chemistry and application.

8. Wikipedia contributors. (2024, May 14). Sulfoxylic acid. In Wikipedia, The Free Encyclopedia. Retrieved June 2024, from https://en.wikipedia.org/w/index.php?title=Sulfoxylic_acid&oldid=1223734514

 

[Formulator]

Amazing work!!!

What are your thoughts about this alternate pathway for ammonia and formaldehyde? I guess the presence of the sulfite interrupts?

1. Ammonia reacts with formaldehyde to form methyleneimine (CH2=NH):

HCHO + NH3 → CH2=NH + H2O

2. Methyleneimine reacts with additional formaldehyde and ammonia to form hexamethylenetetramine (methenamine):

4CH2=NH + 2NH3 → C6H12N4

Overall Rx:

6HCHO + 4NH3 → C6H12N4 + 6H2O

 

[jft]

What about API Tap water conditioner and TeTra Aquasafe used it for YEARS with Good Effect?

 

[Malcontent]

What about API Tap water conditioner and TeTra Aquasafe used it for YEARS with Good Effect?

API TWC doesn't claim to detoxify ammonia and interestingly neither does AquaSafe Plus.

 

[jft]

API TWC doesn't claim to detoxify ammonia and interestingly neither does AquaSafe Plus.

They split Chloramine into Chlorine and Ammonia and if you have a fully functioning biofilter the Bacteria should handle the ammonia molecule, the Chlorine is Neutralized. I have successfully used both on REEFS , Marine and Turtle tanks without spikes or losses for 15 years plus without issues. I used Prime but its more expensive and did not IMO do a better Job. just my personal experience and The results have been favorable.

 

[taricha]

API TWC doesn't claim to detoxify ammonia and interestingly neither does AquaSafe Plus.

API TWC I did look at and it tested as just thiosulfate (consistent with the label.)
Tetra Aquasafe Plus I did not test, I can't find the MSDS or source that I used for including it in this list so I'm removing it. Thanks.

 

[taricha]

Amazing work!!!

What are your thoughts about this alternate pathway for ammonia and formaldehyde? I guess the presence of the sulfite interrupts?

1. Ammonia reacts with formaldehyde to form methyleneimine (CH2=NH):

HCHO + NH3 → CH2=NH + H2O

2. Methyleneimine reacts with additional formaldehyde and ammonia to form hexamethylenetetramine (methenamine):

4CH2=NH + 2NH3 → C6H12N4

Overall Rx:

6HCHO + 4NH3 → C6H12N4 + 6H2O

I don't have enough chemistry kung fu to look at that and say whether it's likely or not or why (all balanced equations look plausible to me!).
But Fig 8, you can see in Dan's clever experiment that even providing 12 mM of formaldehyde did nothing to the measured NH3 value, whereas providing 12 mM of formaldehyde + equimolar sulfite caused a removal of NH3 of the same size as the 12 mM of HMS.

 

[taricha]

Along with a hach reagent set and a distillation apparatus.

Personally, I wouldn't go too crazy on buying test reagents until you get a handle on the interferences that will need to be overcome. (discussion there in part 6)

You can probably figure out a clever distillation procedure to get what you're after, but some of these things decompose under heating (or high pH or low pH) and might complicate things more.

 

[Malcontent]

API TWC I did look at and it tested as just thiosulfate (consistent with the label.)
Tetra Aquasafe Plus I did not test, I can't find the MSDS or source that I used for including it in this list so I'm removing it. Thanks.

Tetra does list the ingredients for AquaSafe Plus. I know they have ammonia binder patents but their only ammonia detoxifying product says it works by lowering pH. Tetra is the only big name that doesn't have an ammonia-binding product. Maybe they found they don't work and are simply being more honest.

 

[Formulator]

I don't have enough chemistry kung fu to look at that and say whether it's likely or not or why (all balanced equations look plausible to me!).
But Fig 8, you can see in Dan's clever experiment that even providing 12 mM of formaldehyde did nothing to the measured NH3 value, whereas providing 12 mM of formaldehyde + equimolar sulfite caused a removal of NH3 of the same size as the 12 mM of HMS.

What is particularly interesting, considering the data, is that the product of my proposed reaction between formaldehyde and NH3 is an uncharged amine base, methenamine. So theoretically it could pass through your gas membrane test disc and induce a pH change thus registering as NH3 with those discs (if I understand the mechanics of the disc tests correctly).

 

[Randy Holmes-Farley]

Amazing work!!!

What are your thoughts about this alternate pathway for ammonia and formaldehyde? I guess the presence of the sulfite interrupts?

1. Ammonia reacts with formaldehyde to form methyleneimine (CH2=NH):

HCHO + NH3 → CH2=NH + H2O

2. Methyleneimine reacts with additional formaldehyde and ammonia to form hexamethylenetetramine (methenamine):

4CH2=NH + 2NH3 → C6H12N4

Overall Rx:

6HCHO + 4NH3 → C6H12N4 + 6H2O

This paper looks at the the kinetics of that in fresh water. At least in fresh water at pH 8, the overall reaction seems much too slow to be useful.

The Kinetics of the Reaction of Formaldehyde with Ammonia

Abstract. The kinetics of the reaction of formaldehyde with ammonia to form hexamethylenetetramine has been investigated in dilute aqueous solutions at 20°

academic.oup.com

"The reaction was first-order with respect to ammonia and second-order with respect to formaldehyde. "

At pH 7.9, k is about 230 M-2sec-1 (from table II; temp = 20 deg C)

let's pick to look at the half life at 2 ppm ammonia (0.12 mM) and 3.5 ppm formaldehyde (0.12 mM) for ease of calculation.

The half life of a third order reaction is
T1/2 = 1.5/(ka^2) where a is the starting concentration
from: https://www.toppr.com/ask/question/halflife-of-a-thirdorder-reaction-is/

Thus,

T1/2 = 1.5/(230 M-2sec-1 x 0.00012 M x 0.00012 M)
T 1/2 = 452,898 seconds = 5.2 days

FWIW, there is lots of detailed discussion of possible intermediates if you want to try to finger those.

"These results may be explained by a mechanism involving a rate-determining attack of methylolamine on free formaldehyde to form dimethylolamine, followed by theformation of hexamine through cyclotrimethylenetriamine and 1, 5-endomethylene1, 3, 5, 7-tetrazacyclobctane."

Thus, one would likely need dimethylolamine (or something before it) to be the thing that reactions with NH3 sensing films to avoid the long delay times.

 

[Randy Holmes-Farley]

The first fast reaction product according tot hat paper is NH2CH2OH, which can react with another formaldehyde to form NH(CH2OH). Those would be what might react with ammonia sensing films (although it does not obviate the fact that Seachem claims these films do show the effect of Prime).

These are not stable compounds one can purchase.

 

[Formulator]

This paper looks at the the kinetics of that in fresh water. At least in fresh water at pH 8, the overall reaction seems much too slow to be useful.

The Kinetics of the Reaction of Formaldehyde with Ammonia

Abstract. The kinetics of the reaction of formaldehyde with ammonia to form hexamethylenetetramine has been investigated in dilute aqueous solutions at 20°

academic.oup.com

"The reaction was first-order with respect to ammonia and second-order with respect to formaldehyde. "

At pH 7.9, k is about 230 M-2sec-1 (from table II; temp = 20 deg C)

let's pick to look at the half life at 2 ppm ammonia (0.12 mM) and 3.5 ppm formaldehyde (0.12 mM) for ease of calculation.

The half life of a third order reaction is
T1/2 = 1.5/(ka^2) where a is the starting concentration
from: https://www.toppr.com/ask/question/halflife-of-a-thirdorder-reaction-is/

Thus,

T1/2 = 1.5/(230 M-2sec-1 x 0.00012 M x 0.00012 M)
T 1/2 = 452,898 seconds = 5.2 days

FWIW, there is lots of detailed discussion of possible intermediates if you want to try to finger those.

"These results may be explained by a mechanism involving a rate-determining attack of methylolamine on free formaldehyde to form dimethylolamine, followed by theformation of hexamine through cyclotrimethylenetriamine and 1, 5-endomethylene1, 3, 5, 7-tetrazacyclobctane."

Thus, one would likely need dimethylolamine (or something before it) to be the thing that reactions with NH3 sensing films to avoid the long delay times.

Ahhh, you just had to go sticking rate constants into my beautifully balanced chemical equations . Agreed though. Not likely the dominant reaction with this information.

I will think about the intermediates and play with arrow pushing later. This is fun!

 

[taricha]

an uncharged amine base, methenamine.

The first fast reaction product according tot hat paper is NH2CH2OH, which can react with another formaldehyde to form NH(CH2OH). Those would be what might react with ammonia sensing films

Can we say anything about the properties of these proposed products...
Are these things gases? Would they diffuse from water into air at the same rate as NH3? Do we expect that they could slip into ion selective electrodes and fool those too?
We might already have experimental data that relates to these

A practical argument against these proposed mechanisms (aside from Randy's point about slow kinetics) is that if low molar ratios of formaldehyde detoxified ammonia, then that would likely have been widely adopted, or at least widely recognized - since people often use formaldehyde in quarantine settings.

 

[Randy Holmes-Farley]

Can we say anything about the properties of these proposed products...
Are these things gases? Would they diffuse from water into air at the same rate as NH3? Do we expect that they could slip into ion selective electrodes and fool those too?
We might already have experimental data that relates to these

A practical argument against these proposed mechanisms (aside from Randy's point about slow kinetics) is that if low molar ratios of formaldehyde detoxified ammonia, then that would likely have been widely adopted, or at least widely recognized - since people often use formaldehyde in quarantine settings.

They would diffuse into air much more slowly than NH3. They also are likely quite toxic, but there likely won’t be studies since they would be unstable and hard to test. I would not assume these are detoxified ammonia.

They could never be isolated but could diffuse into membranes, likely more slowly than NH3.

 

[Randy Holmes-Farley]

Here’s an in vivo study of ammonia and formalin.

https://fish.ku.ac.th/pdf/Bulletin29-38/Fishery%20Bulletin%20no%2032-1-2.pdf

 

[Formulator]

Here’s an in vivo study of ammonia and formalin.

https://fish.ku.ac.th/pdf/Bulletin29-38/Fishery%20Bulletin%20no%2032-1-2.pdf

So the mechanism I proposed is viable, albeit at much higher doses of formaldehyde than those suggested by our products. By my calculations, taking the paper at face value, the neutralization of 0.12 mM NH3 would require about 3 mM formaldehyde.

 

[Randy Holmes-Farley]

Leo Morin (deceased original owner of Seachem) long ago wrote this:

"The classical reaction of ammonia with formaldehyde to form methenamine is the principal of most ammonia removing conditioners. It may be used either directly or as a bisulfite complex. The bisulfite formaldehyde complex has the advantage of odor control, enhanced reaction time, and improved methenamine stability."

https://seachem.com/downloads/articles/Ammonia-Management.pdf