Galvanizing flux: Ammonium chloride requirements and analysis

A. Hi Abreham,
I am not quite sure what your actual question is, whether it's how to make galvanizing flux, how to analyze and control it, or how it is used, etc.

But I think Dr. Cook's article "LOWER ZINC IN HOT DIP GALVANIZING" might be what you are looking for :-) Luck & Regards,


Ted Mooney, P.E. RET
Striving to live Aloha
finishing.com - Pine Beach, New Jersey

A. So called "Zinc ammonium chloride" is a mixture of zinc chloride and ammonium chloride. It is mixed in various ratios of the two components, mainly double salt (2:1 NH4Cl to ZNCl2), and triple salt. (3:1).
It is used as an aqueous solution of typically double salt 450 g/L and triple salt 300 g/L (but this varies according to local practice).
The purpose of the ZnCL2 is to prevent rusting between the flux tank and the zinc tank. The NH4Cl is involved in a series of reactions at the interface between the steel and the zinc.
Buy the pre-mixed powder, or liquid, and dilute to required strength. The steel to be galvanized, after cleaning in a pretreatment regime gets dipped in that aqueous solution before immersing in zinc.

A. Measuring the pH of flux is difficult, and often misleading. Most of the usual techniques do not work.
ICP [on eBay affil links] is one way to measure the ammonium chloride concentration.

A. Sirs:
Until about 1985 I did ammonium chloride in flux using Kjeldahl distillation and published the method in the magazine, METAL FINISHING. Now I use a self-developed heat of solution method which is extremely simple. This method takes advantage of the fact that when zinc chloride is dissolved in water the solution gets hot, whereas when ammonium chloride is dissolved in water the solution gets cold. Under the conditions that I use there is about a 35 °C differential. I have taught this method to many galvanizers who use it once or twice a month.
In another of my publications in METAL FINISHING, I sent out 13 samples all from the same batch of flux made up here in my lab with an ACN of 1.00 (ACN = ammonium chloride/zinc chloride; e.g., equal amounts of ammonium chloride and zinc chloride, this is between the double and triple salts). All 13 labs failed to provide a usable result even though most of these labs are "certified" by the US government and charge a few hundred dollars/test.
It is true that a newly dissolved flux solution has a pH near 3. The best operating pH for flux is 4.2. "pH" is pH not "ph." This is the negative log base ten of the hydrogen ion concentration (e.g., when the pH changes by one unit the hydrogen ion concentration changes ten-fold, "pH" is actually a mathematical term applied to chemistry.) Ammonium chloride is NOT what to add to change the pH! pH is NOT measured in degrees.
Geoff, I am not familiar with using ICP (inductively coupled plasma) [on eBay affil links] for measuring ammonium chloride in flux solution. I am under the impression that ICP is used like AA or AE except that it is more sensitive because the temperature is hotter. AA or AE is used for 30 or 40 metals that have electrons much further from the nucleus and more easily promoted to outer shells to emit energy upon return. Actually it doesn't really matter because the folks with ICP are very reluctant to run samples with zinc in them because of the memory effect.

Regards,

A. Sir:

It is possible that your local fertilizer company may have a way to test for NH4Cl. The method would likely be Kjeldahl distillation or total nitrogen. I personally would not trust the result however. If the total (ZnC2 plus NH4Cl is 800 g/liter and the ZnCl2 alone is 545 g/liter then the NH4Cl alone is 255 grams/liter (800-545). This would be an "ACN" of only 0.47 (255/545). This ACN corresponds to the mono salt which is commonly used at high concentration for continuous sheet galvanizing. It would NOT be a good flux for general galvanizing.

EDTA [on eBay or Amazon] is the proper titrant for zinc +2 ion. There would be no use for permanganate in this titration (except to oxidize iron+2 to iron+3). This is a very messy way to determine zinc because the orange sludge formed from iron +3 would contain liquid that would also still contain zinc.
For approximate zinc chloride the density in baumé does quite well if the NH4Cl is fairly well known. I have full tables for this.

Could you explain a bit more about your operation, especially the type of galvanizing that you do.

Regards,

A. I would determine Zn via an EDTA titration, and the ammonium chloride via an ion selective electrode.

If I/my employer was too cheap to buy the probe, I'd distill off the NH3, with the Kjeldahl method, then titrate the distillate with 0.02 N H2SO4.

You might have luck using Nessler's reagent [on Amazon] for the ammonia determination, if you prepared the proper matrix matched standards. Lots of set up time, but possibly quicker than the above two methods, after the initial work of making the standard curve is done.

I also don't see how ICP could be used for NH3.

A. Sir:

The "bake and shake" method of testing flux does NOT yield ACN. It gives ACNV which is the correct way to control the volatile (ammonium chloride) and the non-volatile components (zinc chloride, ferrous chloride, and sulphate) of hot dip galvanizing flux. As I previously mentioned on this thread this method is based on the fact that ammonium chloride is endothermic (gets cold) when it is being dissolved in water and zinc chloride and iron (+2) are both exothermic (gets hot) when dissolved in water. Also ammonium chloride contributes little to baumé, whereas zinc chloride and iron chloride give high baumé solutions when dissolved. The method depends on a series of known ACNV's and equipment that is calibrated. It is not possible for me to give the methods or calibrations on a web site.

Regards,

Barlow:

My "bake and shake method" is not quick but it is simple once it is properly set up. It gives ACNV (ACNV = ammonium chloride/(zinc chloride + iron (+2)chloride) NOT ACN. The ACNV value is quite accurate and extremely useful in controlling flux. ACN is NOT useful in the presence of high iron(+2) chloride. pH and baumé do NOT control ACN and in fact are not even related to ACN.

Regards,

A. Barlow:

I have read, with great concern, your last posting several times. In fact I have given your last posting much thought today. You and I are obviously NOT communicating on this website. I have numbered your sentences and I can start to respond as follows:
(1) pH, baumé, HCl, or original formulation are in no way related to ACNV of a process flux solution.
(2)ZnCl2 does not dissociate into HCl. Zinc oxide is insoluble in water. Zinc chloride provides a layer that inhibits rust.
(3)Kindly define your "zinc chloride number." What does "zinc chloride number pulls down the pH," Mean? What is your proof that--"a small portion is believed to be dissociated to ammonium hydroxide?"

I cannot further respond to postings that are contrary to my education and experiences of nearly 40 years working in hot dip galvanizing.

I think I will buy a fishing license and some worms tomorrow and go fishing for walleye and bass :-)

Regards,

A. Barlow:

It occurred to me that you may have tried to get my attention. It is unlikely to happen again. The fishing today was a bit slow but we saw a mother goose and four babies.

The "V" part means using both zinc chloride and iron (II) chloride in the denominator as I previously reported.

I am not able to follow your reasoning in your next few sentences.

In the USA there are some flux companies that use "ratio" which is the inverse of ACN (1/ACN = ratio). This ratio is NOT the OPPOSITE of ACN. Amphoteric means that a material can act as an acid or a base and thus soluble in both.

The "bottom line" on the bake and shake method is that any solid flux when dissolved in water gets colder when the ACNV is higher. Or with sufficiently low ACNV gets hotter (as it is dissolved in water) as the ACNV gets lower.

The method(s) are easy and quite accurate once it is fully calibrated and tables are developed, as I have done.

Now I have other "fish to fry."

A. Sir:

There are three forms of iron that can be in a galvanizing flux tank:

(1) Metallic iron with oxidation number 0. This would come from workpieces falling into the flux tank or steel shot from "sandblasting". This metallic iron dissolves slowly or moderately fast depending on the pH of the flux. Metallic iron can be removed using a strong magnet.

(2) Ferrous iron with an oxidation number +2 is formed when metallic iron dissolves in a flux solution with a pH lower than 5. This type of iron is titratable using hexavalent chromium. In excess in the flux solution it causes excess ash and in the case of a "wet" kettle causes dross. It also causes thicker zinc on the product. If the ferrous iron is more than about 3% there is no reason to try to purify (remove iron) because the entire tank will turn to solid ferric hydroxide.

(3) Ferric iron with oxidation number +3 is caused by oxidation of ferrous iron by a pH higher than 5 by air or peroxide. Ferric iron precipitates with three hydroxides to form orange, insoluble ferric hydroxide that goes to the bottom of the flux tank.

Nearly forty years ago I went around the USA purifying flux solution (removing iron and sulphate, if sulfuric acid was used as the pickle). When the sulphate was about 1.5% and the iron (+2) was about 0.6%, upon purification the %GZU decreased from about 9% to about 6%. The money yearly savings were usually between $0.5 to $1 million/year from zinc savings and much, much less zinc clean-up. Ahh--The Old Days.

To remove the iron (+3) sludge you get scoop shovels and put the orange sludge into barrels. It's a sweaty job.

Regards,

A. William,
I'm not sure if what you really need is treatment or just better prevention. I found an MSDS for zinc ammonium chloride that lists first aid for skin contact, of course all it says is the standard rinse the area and seek medical attention. The page

http://www.johnsonmfg.com/temp/HWhtml/ZCFS3p~1.htm ⇩

goes into more detail about it: "ZnC12 absorption through your skin isn't likely, but absorption onto your skin is highly likely. This is due to the desire of ZnC12 to be wet. It will cling to the water molecules in the cracks of your skin, and once there, it holds on tight."

I am not a doctor, but due to the self-renewing nature of skin I rather suspect that any trapped zinc chloride will be cast off on its own eventually as long as you are not continuing to add more zinc chloride to the area. It sounds like you need better protective clothing, and maybe in the meantime some good skin moisturizer. You may want to discuss the matter further with a dermatologist and your plant's safety manager.

[Ed. update: The above link has broken. Thankfully, The Internet Archive preserved a copy
These days information often appears on the internet, then disappears forever. If that bothers or frightens you please consider a donation to The Internet Archive.

? Sir: Do you use nickel in the kettle?

Regards,

A. In the 1980's I did many KJ tests on flux and it works just fine, but does take skill. Fertilizer companies normally have good skills in this test and likely is less expensive than getting the glassware to do it yourself. The answer is normally g/L of ammonium chloride.

You have made a mistake thinking you need to add NaOH to take out iron. You need both hydrogen peroxide and ammonium hydroxide and you must add the ammonium hydroxide first. NaOH will destroy your flux.

What is your iron (Fe+2) content? and what is flux tank size?

Regards,

A. The Kjeldahl analysis is not specific for ammonia. It measures total nitrogen content which, of course, includes ammonia.
It can be highly accurate in skilled hands but is slow and tedious.
A far simpler method is to boil a known sample with a known excess of sodium hydroxide until all ammonia is driven off and then titrate the residual hydroxide.
However, you cannot just buy the kit and assume that the knowledge of how to use it comes free. Even the simple method requires some basic knowledge of technique and without it you will get unreliable and misleading results.
You probably have a local college with a chemist who will give you the basics but my best advice would be to speak to your flux supplier.

A. Geoff, KJ works perfectly for ammonium chloride in galvanizing flux. I have used it for many years and have made standard solutions containing zinc chloride, iron chloride, and other chlorides and there are no interferences. I do not share your confidence in flux suppliers testing fluxes properly. 22 years ago I made up a flux solution with a specific composition and sent portions of it to 13 companies for testing. All results were useless. World- wide there are many contaminated galvanizing fluxes. I do not expect your alteration to the KJ to work. What about the OH-1 tied up with iron or zinc at the bottom? Also what about the colors interfering with the end-point detection?

Sir, Iron (III) hydroxide at the bottom of the boil flask does not interfere. If you can find a fertilizer company nearby, that does KJ every day, they may do it for a small fee. Here the cost is $25/sample and I have one client that does testing this way. The specific ion electrode for ammonia is high priced and the flux needs to be well diluted with distilled water. From which country do you buy flux? Is your kettle very deep as you indicate?

I now use my bake and shake method and it is easy with excellent results. I have a number of clients that also use the method to advantage. It is based on "heat of solution." ZnCl2 gets hot when dissolved and ammonium chloride gets cold when dissolved. Also there is a second method using density to get ACN directly.

We are working on a Dynamic test (four steps) for testing flux. The first takes a 2 inch x 2 inch x 1/4 inch angle 5 inches long and pickling it and fluxing it and seeing if it will galvanize in 20 seconds. All of my clients have passed this test. There are three more steps in the dynamic testing. I am considering posting it on this web site.

Regards,

Dr Cook: I agree completely that KJ is a perfectly good and accurate method in capable hands.
My point was to question if it is suitable for someone with limited analytical experience; otherwise the question would not have been asked.
The indirect method I suggested is simple, quick and in inexperienced hands probably as accurate as KJ would be.
However, your advice to farm out the work is by far the best option.

A. Pham, By pipes galvanizing 1 year you are certain to have much Fe+2 carry-over from your HCl pickle into the flux and there may be as much FeCl2 in the flux as there is ZnCl2. The Fe+2 in your flux can be easily titrated with dichromate.

What does "inflated" mean? Does it mean "peeled off"

I prefer P at 0.01% or lower.

Your methods for testing ZnCl2 and NH4Cl, I understand perfectly, but they do not work. Iron interferes with the zinc test and if you are using "hard" well water so does calcium. For NH4Cl you are getting total Cl-1 and doing a subtraction. You cannot get correct NH4Cl from wrong ZnCl2 and total Cl-1.

I recommend that you do the first and fourth tests in my new dynamic flux tests as shown in the on-line library on this website. Then post the results on this website. photo (of 20 second test) and minutes (for cook-off time).

FLUX: What country do you buy your flux? What is your flux density (g/ml)? What is temperature of flux? What is color with 50 ml flux + bromophenol blue [on eBay or Amazon] ? Do you have an effective wetting agent in your flux solution? Are your problems (post photos) mostly on the outside or inside of the pipes? What diameters of pipes do you galvanize?f What is TDS (total dissolved solids) of your tap water? Do you have water softener?

Regards,