Nitriding for duplex coating

In 1992, I presented my doctor dissertation focussed on combination of Titanium nitride and plasma nitriding. So, it’s been more than 30 years since I tried to clarify the problem of adhesion of the PVD layer on the nitrided substrate. But time passed and nothing fundamental changed. Heat treatment facilities offer nitriding technology, and users continue to complain that the layer of TiN, DLC, CrN, etc. does not stick to the surface. What is the problem?

Well, mainly because commercial heat treatment shops have no experience with coating, and conversely, commercial coating suppliers have little idea about nitriding.

The main problem of good adhesion of the PVD layer on the nitrided surface is the phenomena occurring at the interface between the layer and the substrate.

Fig. 1 – Interface between TiN and nitrided surface

The idea that the epitaxial growth of the TiN crystal lattice on the α´- Fe(N_x) lattice can occur was not confirmed, except for cases where epitaxial growth of a layer on MC or M₆C type carbides was recorded. However, these carbides are represented as a minor component in steel, and therefore cannot fundamentally affect the overall adhesion.

Obr.č. 2 – Interface strength MC/TiN > M₆C/TiN > HSS/TiN [1]

[1] : Strength ranking for interfaces between a TiN hard coating and microstructural constituents of high speed steel determined by micromechanical testing, Elsevier, Materials & Design, vol 204, June 2021, Materials Center Leoben Forschung GmbH)

Thus, the bonding between the PVD layer and the substrate is almost certainly based primarily on covalent and metallic bonding at the atomic level. Therefore, in order to maximize this bond, we need to maintain a metallic clean surface of the substrate after nitriding, without any impurities and oxides.

Fig. 3 –Experiment result of stress annealing Ti6Al4V at b) high vacuum  < 5*10-5 mbar, c) in retort furnace with Ar 99,9995, c) on air [2]

In order to achieve this, we have to deal not only with processes, but also with technical gases and their purity. Even if, for example, we use argon with a purity of 99.9995 at a pressure of 1 bar abs in a retort furnace, we work even under these conditions with argon containing 5 ppm H₂O, 3 ppm O₂. Converted to a total of 8*10^-1 Pa (H₂O+O₂). This amount of oxygen and water vapor is sufficient to form a layer of TiO₂, Al₂O₃, V₂O₅ oxides on the surface of the titanium alloy with thickness of 0.3 µm.

 [2] Thermal Treatment of 3D printed Titanium Alloy (Ti6Al4V), Manufacturing Technology, Michaela Fousová, Dalibor Vojtěch, April 2018, ISSN 1213-2489

Even if we work with pure gases, we still have oxygen and water vapor in the nitriding chamber, which can create oxides. But most nitriding processes are based on NH₃ ammonia. Do we know how it is with his purity? The highest purity 4.0 is supplied by GHC (Gerling, Holz & Co). It can be seen from the technical sheet that even this pure ammonia contains 100 ppm of water. If we work with a retort nitriding furnace with 1 bar abs. of NH₃ pressure, there will be 10 Pa of water partial pressure (H₂O) in the retort.

Fig. 4 – Technical parameters of ammonia from GHC

So even if we try our best for nitriding in gas, ammonia itself brings so much water into the process that we will not perform nitriding but oxynitriding to form α´-Fe(N_xO_y).

The relationship between the adhesion of the TiN layer on the HSS (1.3343) substrate and the nitrogen content in the layer is shown in Fig. 5.

Fig. 5 – The values of adhesion Lc, measured by the scracth test method with acoustic emission, on a nitrided surface with different nitrogen contents in wt. %

It follows from the above that the optimal nitriding for duplex layers will be in the area with a partial pressure of nitrogen in the range of p_N2 = 2 to 4 Pa. Even if we achieve a comparable Kn in gas nitriding furnace, we still have the problem of 10 Pa of water partial pressure, which is higher than our proposed regulated value for nitrogen.

If we want to control the nitrogen partial pressure in the above range, we have to suppress the residual atmosphere at least one order lower. We are already in the area of pressures that can be achieved only with the primary and Roots vacuum pumps. Therefore, we must ensure that before starting the nitriding process, the starting pressure is at least 2 – 4 * 10^-1 Pa, i.e. 2 – 4 * 10^-3 mbar, and at the same time the leakage of the device is not greater than 4 Pa*l/s.

And what about the hardness of the nitrided layer on the HSS (1.3343) substrate under these conditions? It grows steeply even with minimal nitrogen content. In the area corresponding to the partial nitrogen pressure of 2 – 4 Pa, the values are according to Figure 6. Here, the lattice parameter approximately α’- Fe (N_x) = 0.2870 nm and the hardness 1600 to 1700 HV0,1 was measured.

Fig. 6 – Dependence of the microhardness of the nitrided surface on the lattice parameter

What are the conclusions? In practice, 3 typical nitriding processes are available

• Nitriding in gas
• Nitriding under reduced pressure (LPN)
• Nitriding in plasma

Since the first two work exclusively with ammonia NH3, based on the above, we can completely exclude these two nitriding technologies from processes suitable for duplex coating.

So, what are we left with? Only and exclusively plasma nitriding. If the manufacturers of PVD/PACVD coating equipment manage to implement this nitriding technology in the deposition equipment, we win, we don’t have to worry about anything. The entire process, nitriding and deposition, takes place in one device and under almost ideal conditions.

But if we don’t have such a device, the processes must be split into two different cycles,  nitriding and coating. Nitriding in plasma then have to fulfill all conditions stated above:

• we have to work with a Roots vacuum ensuring a starting pressure of 1* 10^-3 mbar, i.e. with acceptable cleanliness of the atmosphere in vacuum chamber
• the plasma nitriding process itself requires a tight device, even so we must guarantee a leakage rate of less than 4 Pa*l/s
• as we are working  with N₂ and H₂ gases at a low working pressure of 1-5 mbar, so the influence of impurities in the gas in the form of oxygen and water vapor is significantly suppressed
• the reducing effect on surface impurities and oxides is ensured in the plasma by high-energy particles bombarding the surface, when the voltage on the cathode/substrate can be up to 2 kV
• the dissociation of hydrogen in the glow discharge provides us with a sufficient amount of atomic hydrogen so that the chemical reduction of oxides by reaction with hydrogen takes place simultaneously
• it is not a technical problem to regulate the nitrogen partial pressure at the level of 2 to 4 Pa in the plasma nitriding device

Jiří Stanislav

May 25, 2023