Ammonia and its use in heat treatment

The effort to improve the surface properties of low-alloy or pure carbon steels is quite evident. They are cheap, and nitriding or nitrocarburizing processes can improve their usability at affordable prices, within one cycle. Therefore, the need for ammonia for nitriding or nitrocarburizing in gas has an increasing tendency. So, it is worth considering what ammonia to use for these technologies, when we start this technology. What should be our basic criterion?

I have already written about this once here: https://www.jstconsultancy.cz/karbonitridace-a-kvalita-cpavku/. The condition for the success and reproducibility of the process is the purity and stability of our ammonia.

Fig. 1 – Amount of water in different grades of ammonia

The graph shows the difference between the individual grades of ammonia. If we look at their purity, even though all grades contain a certain amount of oil up to 10 ppm, they differ fundamentally in the water content. Even though ammonia is called “anhydrous”, the difference is precisely in the water content.

From my own experience, I can say that grades 2.5 and 2.8 can be excluded from nitriding and nitrocarburizing processes. And not only for processes at atmospheric pressure, but also for low-pressure processes such as NITRAL from Fours BMI. The reason may be the instability of the process, manifested in the formation of a surface layer. I have personally experienced this, and it is very unpleasant to tell the customer that it did not work.

Fig. 2 – Layer after low-pressure nitrocarburizing on steel 11 373
Fig. 3 – Layer after low-pressure nitrocarburizing on steel 12 050

The presence of water can lead to various reactions both in the volume of the retort and on the surface of the steel components (Fig. 4), however, their precise description is almost impossible, if only because the thermal decomposition of ammonia and water has its own laws.

Fig. 4 – Gas nitriding model according to https://pubs.acs.org/doi/10.1021/jp410947d
Fig. 5 – Mass spectra of ammonia without and with water content, measured on an Inficon mass spectrometer, PGA 100, NH3 flow rate 150 ml/min, total pressure ≈1Pa.

From the mass spectrum in Fig. 5 the spectrum of pure ammonia without water content is not only made up of NH3⁺ ions, but also NH2⁺, NH⁺ and nitrogen atoms N^+. Their proportion will vary depending on the degree of ammonia dissociation, i.e. depending on the process temperature (orange bars).

However, if water is also present in ammonia, its peak appears at m/e = 18, and radicals O⁺, OH⁺, CO⁺, N2⁺ and NO⁺ increase. These ions are the result of reactions of ammonia ions with water. They will not have a significant effect on the dissociation of NH3 itself, but their effect will be significant on the processes on the steel surface. The decomposition of water NH3.H2O will increase hydrogen and oxygen in the working atmosphere. Hydrogen will affect the nitriding potential represented by the nitriding number Kn, oxygen will then react with N⁺ to form, for example NO⁺, but it can also be a source of oxidation or decarburization of the steel surface by reaction with carbon.

From Fig. 1 the water content of 100 to 400 ppm represents huge partial pressures, in the conversion of 10 to 40 Pa, or 0.1 to 0.4 mbar. Why huge? Because it is verified that the activity of the atmosphere a(N) ≈ a(pN2).

From the processes of nitriding in plasma it is known that ≈7 Pa of partial pressure of nitrogen is sufficient to form a nitriding layer ƴ´- (Fe4N). In other words, if any reactive element C, N, B, S, or its carrier gas were in the atmosphere in an amount of 10 to 40 Pa of partial pressure, we have enough amount of this element for diffusion and formation of various types of nitrides, carbides, borides, sulphides according to the relevant phase diagram.

The nitriding number itself is based on the following formula. It follows that the more hydrogen there is in the atmosphere, the lower the nitriding number will be.

Therefore, if we nitride or nitrocarburize, the nitriding number Kn will significantly depend on the process temperature. The temperature therefore determines the degree of ammonia dissociation.

Fig. 6 – Dependence of the degree of dissociation of ammonia on temperature

Fig. 7 –  Lehrer diagram of dependence of Kn and temperature

At a classic nitrocarburizing temperature of 550 °C, the degree of dissociation will be approximately 13% according to the graph in Fig. 6. This means that the atmosphere will contain approximately 32.5 mbar of N2, 97.5 mbar of H2 and 870 mbar of NH3. This results in a nitriding number of approximately 1. This is high enough to allow nitriding to ƴ´- Fe4N according to the Lehrer diagram in Fig. 7.

Fig. 8 – Kn and atmospheric composition in the case of using pure ammonia NH3
Fig. 9 – Kn and atmospheric composition in the case of nitrogen addition of 800 mbar

But if I return to our problem with water in ammonia, what would the water content have to be to reduce the nitriding number from 1 to 0.9 and thus stop the formation of the nitride layer? The calculations show that it would have to be 8,600 ppm H2O.

But we only have 400 ppm even in ammonia with a purity of 2.5, why should that be a problem? A partial answer is in Fig. 9. Adding nitrogen to the working atmosphere significantly narrows the range for the partial pressure of hydrogen, and in that case our nitriding or nitrocarburizing will be much more sensitive to the amount of hydrogen from the water content. The same applies to low-pressure processes.

The second, and even more fundamental, influence is the cleanliness of the bottles or barrels. This is probably exactly what happened to us. Water accumulated in the bottles that were not cleaned regularly, and this caused the collapse of the nitriding cycle. From the spectrum in Fig. 5 shows that in this case the water content in ammonia was measured in the ratio of peaks at m/e = 17 (NH3⁺) and 18 (H2O⁺), i.e. in our case in the volume of about 8.3%. This is 83,000 ppm at atmospheric pressure. This number is 10 times higher than the amount of water to reduce Kn below 1, and really sufficient for the nitriding process to stop working. How can this happen?

Ammonia in a bottle or barrel is liquid, but there is gas above the liquid level. The pressure of this gas is tied to the temperature of the container and is, for example, at 0°C: about 4 bars, at 20°C: about 8.5 bars, at 30°C: about 11 bars. It therefore corresponds to the pressure of the saturated vapor of the gas in the container. If the gas is removed too quickly, the liquid cools down because evaporation takes away heat, and the pressure therefore drops. In extreme cases, the outlet valve can even freeze.

On the other hand, water has a low vapor pressure, only 23 mbar at 20 °C. Therefore, it evaporates significantly less than ammonia and accumulates in the bottle or barrel. Its concentration in the liquid gradually increases. If we do not clean the ammonia container at all, and we take ammonia with a purity of 400 ppm H2O, after five fillings without cleaning we will have 8 000 ppm of water in the container. This is already a critical number for our nitriding or nitrocarburizing process.

Periodic cleaning of transport packaging is carried out according to the following scheme:

Fig. 10 – – Workflow for periodic cleaning of barrels or bottles

What to say in conclusion?

• With higher ammonia purity, higher process reliability can be expected
• For controlled nitriding and nitrocarburizing processes, water content below 200 ppm can be recommended
• Hydrogen released by H2O decomposition reduces the nitriding number
• Since Kn is evaluated indirectly via the H2 sensor, this excess hydrogen can also distort the Kn evaluation
• Due to the different evaporation characteristics of NH3 and H2O, for the actual process we collect saturated NH3 vapors with a certain proportion of water, but a large part of it will remain in the container. Repeated filling of bottles or barrels will therefore cause an increase in water. Based on the input ammonia purity, the number of permitted fillings of the container can be calculated so that the water content does not exceed the limit
• Just as important as the water content in ammonia is the maintenance plan for barrels and bottles. The supplier must be able to control the number of fillings and guarantee that complete regeneration of ammonia transport containers will be included in the correct time interval

In the Czech Republic, there are basically only two major suppliers of ammonia for quenching plants, Linde and GHC Invest (https://www.ghcinvest.cz/3-Kontakty). GHC supplies NH3 from the German company GHC Gerling, HOLZ & Co., in the best quality on the market, with a purity of 4.0. However, it supplies ammonia exclusively in bottles or barrels.

Linde then, if I am not mistaken, ammonia from Belgium with a purity of 3.8. In addition to bottles and barrels, it can also fill large-capacity tanks.

If you find it interesting and want to try it, contact me or directly at GHC in Prague:

GHC Invest, s.r.o.
Korunovační 103/6
170 00 Praha 7
tel: +420 233 374 806
email: info@ghcinvest.cz
www.ghcinvest.cz

Jiří Stanislav

April 27, 2025