A vacuum-purge gas nitriding furnace was modified to develop a process and a furnace enhancement to produce a controlled in-situ oxide layer on the surfaces of steel parts using various oxidation techniques. The process is an effective alternative to conventional grit blasting of materials as a means of surface preparation for uniform and consistent nitriding results.


Pre-oxidation is known to enhance receptivity of steel part surfaces to the effects of nitriding, and in-situ oxidation is inherently efficient and economical. We will discuss the type of oxidizing carrier used in the furnace, practical methods used to control the oxidation, and a gas delivery system developed to inject gases with an elevated dew point for the purpose of providing a controlled oxidizing atmosphere. Comparative tests with other activation techniques and results with no activation are discussed along with approaches to technical process difficulties encountered.


Introduction and Background

Nitriding is a thermochemical process where nascent nitrogen (N) is diffused into the surface of steel for the purpose of increasing wear resistance. The process is performed by heating steel parts in a controlled atmosphere to a temperature range of approximately 750-1150°F (400-621°C). Because of the absence of a quenching requirement, the attendant volume changes and comparatively low temperatures employed in this process, nitriding of steels produces less distortion and deformation than either carburizing or conventional hardening.[1] Ammonia gas (NH3) is introduced into the furnace atmosphere and reacts catalytically on the hot steel parts. Some of the nascent nitrogen produced from the dissociated ammonia (Eq. 1) diffuses into the steel.

NH3 → N + 3/2 H2               (1)


    The nascent nitrogen that is adsorbed by the steel dissolves into alpha iron or forms nitrides with alloying elements in the steel. The atomic nitrogen and hydrogen components shown in Equation 1 are unstable and unite with other like atoms to form molecules.[2] Some methods of controlling the nitriding potential in the furnace include the use of an external ammonia dissociator and/or dilution of ammonia with other gases such as hydrogen and/or nitrogen.

    Traditionally, a hot-walled metal retort furnace with substantial mass is used for the nitriding process. Prior to heating the parts in the metal retort, nitrogen gas is used to purge air from the furnace to remove residual oxygen for safety. The amount of nitrogen gas required to purge the vessel is substantial, and the purging process is time consuming. After the retort is purged of air, the furnace is heated to the nitriding temperature. The metal retort is inefficient in terms of the large amount of thermal energy required to heat to the nitriding temperature. At the end of the nitriding process, the metal retort is again at a disadvantage because of the time to cool down to ambient temperature in order to unload the parts. The furnace also has to be purged of ammonia with large quantities of nitrogen gas during cooling and prior to unloading.

    A vacuum furnace was designed to have the capability of performing the nitriding process.[3] The furnace incorporates a lightweight, graphite-lined hot zone receiving the parts to be processed (Fig. 1). The graphite-lined interior of the furnace is non-reactive with ammonia gas. For this reason, lower ammonia flow rates are required compared to metal retort designs. The exterior of the vacuum furnace walls are cooled by means of a water jacket.

    In lieu of purging the furnace with nitrogen, a vacuum pumping system is utilized to quickly, safely and effectively purge the vessel of air followed by backfilling with nitrogen gas. The furnace is heated convectively by means of graphite heating elements coupled with recirculation fans and a baffling system.

    The nitriding process is initiated by pumping out a portion of the nitrogen gas and backfilling with ammonia gas to the desired nitrogen/ammonia ratio. The process is controlled by analyzing the hydrogen gas in the effluent, which is generated from the reaction in the furnace. The amount of hydrogen generated combined with ammonia and nitrogen mass flow controllers are used as inputs to an algorithm and feedback loop to automatically maintain a nitriding potential in the furnace atmosphere. At the end of the nitriding process, the furnace atmosphere is purged by the vacuum pumping system, then backfilled with nitrogen gas. A water-cooled stainless steel fin-and-tube heat exchanger and cooling fan are then used to rapidly cool the furnace and parts for unloading.

    The importance of a clean, active surface for consistent and reproducible nitrogen diffusion is essential. Parts that are free from contamination will generally nitride without issues. However, some pre-nitride machining processes that result in burnished or polished surfaces will impede the nitriding process.[4] One method of overcoming machining effects is to lightly grit blast the surface of the steel and thoroughly clean off the grit media prior to nitriding.

    Grit blasting has certain drawbacks, especially in terms of time and labor costs. The type of grit used, the pressure utilized, nozzle distance from the part and operator technique must be consistent for reproducibility.

    As an alternative surface activation method, workpieces are preheated in a controlled oxidizing atmosphere at temperatures varying from 650°F (343°C) to just below the nitriding temperature.[5] The oxidizing atmospheres are typically air, nitrous oxide or water-saturated gas. Oxygen pretreatment is not new. It is typically done in a separate furnace or in a dual-chamber furnace, however, in order to avoid possible mixing of the oxidizing atmosphere and the ammonia. The use of a vacuum furnace allows for a controlled oxidation step followed by nitriding in a single chamber, thereby eliminating the need for two furnaces or a dual-chamber vacuum furnace (Fig. 2).

    The addition of an oxide film on steel parts may be perceived as a source of contamination that would inhibit nitriding. It has been proposed that the surface oxide actually increases the surface area of the part, however, which in turn facilitates nitrogen uptake, nucleation and growth of the nitride compound layer.[6] As an added benefit, undesirable surface residues on the part surfaces are oxidized and vaporized prior to nitriding, resulting in cleaner parts and increased nitrogen diffusion.[7]



The current study looked at the use of a water-saturated nitrogen (N2) gas and dry nitrous oxide (N20) pretreatment to determine the effectiveness of pre-oxidation on the resultant nitriding properties. The material used for the experiments was ground and polished, decarb-free AISI 4140, pre-heat treated and tempered to a hardness of 28-32 Rockwell C (HRC). The resulting case hardness, case depth and white-layer thickness were analyzed after the various pre-oxidation treatments and compared to the process without pre-oxidation.

    The nitrogen gas was passed into a stainless steel vessel filled with de-ionized water controlled at a temperature of 90-100°F. The water-vapor content of the nitrogen increases as the nitrogen gas enters the water-filled vessel through a tube extending to the bottom of the vessel and exits through the top of the vessel (Fig. 3). The water-saturated gas exits the container directly into the preheated furnace. The oxidation temperatures used for these tests were 650°F, 750°F and 980°F. For each trial, the exposure time of oxidation was constant at 20 minutes each (Fig. 4).

    The nitrous oxide pretreatment test used a nitrous oxide/nitrogen blend. The three different nitrous oxide blends used included 5% N2O/95% N2, 25% N2O/75% N2 and 50% N2O/50% N2. Only one oxidation temperature was studied, 980°F (527°C), and the oxidation exposure was 20 minutes for each test. A detailed description of the test procedure for each trial is shown in Table 1 and Table 2.



In the nitrous oxide pre-oxidation testing, the hardness and case depth increased with an increase in the nitrous oxide concentration, which was expected (Fig. 5). Increased oxidizing concentration resulted in increased surface area of the part and enhanced nitriding. Compared with the baseline hardness of 51.3 HRC, the hardness value for 5% nitrous was 54.2 HRC, 54.6 HRC for 25% nitrous and 55.4 HRC for 50% nitrous. As can be seen, increasing the nitrous concentration by a factor of 10 gave improved hardness by only 1.2 HRC points. Such a small increase in hardness compared to the cost of the nitrous oxide may not be justified. The wet nitrogen value of 55 HRC at 980°F is essentially comparable to that of the 50% nitrous test, and the use of wet nitrogen is more economical (Table 3). The effect of pre-oxidation temperature and resultant hardness and case depth with the wet nitrogen runs was more significant compared to the nitrous oxide concentration testing.

    An additional test was conducted to determine resultant case hardness and case depth utilizing the wet nitrogen pre-oxidation technique and comparing a grit-blasted activated test part with a virgin test part with no activation. Figure 6 demonstrates the similarity of activation by grit blasting with alumina oxide (220 grit size) and pre-oxidation utilizing wet nitrogen at 925°F (496°C) for 30 minutes. The case-depth and case-hardness values achieved with the two activation methods are comparable and significantly higher in surface hardness and deeper in case depth compared to the virgin test part.



The water-saturated nitrogen trials indicate that an increase in the oxidation temperature increases surface hardness and case depth when compared to the baseline test with no pre-oxidation. Increasing the oxidation temperature from 650°F to 980°F increased the surface hardness by approximately 2 HRC points. Compared to no pre-oxidation, however, the surface hardness for the 980?F pre-oxidation increased by approximately 4 HRC points. Additionally, the water-saturated nitrogen pre-oxidation test showed a case-depth increase of approximately 0.0015 inch compared to the baseline test. The thickness of the white layer also increased with wet nitrogen pre-oxidation by 5.0 x 10-5 inches compared to the baseline without pre-oxidation. The resultant increase in hardness and slight increase in white-layer formation with increasing pre-oxidation temperature is not unexpected. A thicker oxide case forms as the temperature is increased. The oxide film increases the surface area of the part and provides more opportunity for nitrogen activity at the surface.

    The nitrous oxide pre-oxidation tests also exhibited an increase in surface hardness and case depth when compared to the baseline test with no pre-oxidation. Only a minimal increase in hardness was observed when comparing the various compositions of nitrous oxide from 5-50% N2O, but approximately 3.5 HRC points were gained when comparing pre-oxidizing with N2O to the baseline test. The case depth of the nitrous oxide pre-oxidation also increased by approximately 0.0013 inch compared to the baseline tests. The thickness of the white layer did increase with nitrous oxide trials compared to the baseline by 6.0 x 10-5 inch, similar to the wet nitrogen gas trials.

    The pre-oxidation methods of water-saturated nitrogen with nitrous oxide are comparable. Only a slight difference in case hardness was observed with increasing composition in nitrous oxide matching the increasing pre-oxidation temperatures with water-saturated nitrogen gas (Table 4). An exception may be the 650°F (343°C) wet nitrogen pre-oxidation test, which indicates a slightly lower surface hardness of approximately 1 HRC point when compared to the average surface hardness of all other pre-oxidation trials. Regardless of the pre-oxidation method, the white-layer thicknesses were all comparable. Additionally, results obtained with pre-oxidation methods are equivalent to results obtained with activation using grit blasting. IH


The author acknowledges the following individuals for their contributions: Donald Jordan, vice president, corporate metallurgist; Virginia Osterman, Ph.D., technical director; Aaron Moyer, research technician, all of Solar Atmospheres Inc.; and Harry Antes, Ph.D., consultant.


For more information: Contact Trevor M. Jones, principal engineer, Solar Atmospheres Inc., 1969 Clearview Rd., Souderton, Pa.; tel: 215-721-1502 x1220; fax: 215-723-5039; e-mail: trevor@solaratm.com; web: www.solaratm.com



1. Knerr, C. H. et al., ASM Handbook, ASM International, Vol. 4 (1991), p. 387

2. Davis, J. R., Surface Hardening of Steels: Understanding the Basics, ASM International, (Ohio, 2002), p. 141

3. Jordan, D., “Vacuum Gas-Nitriding Furnace Produces Precision Nitrided Parts,” Heat Treating Progress, (2009), p. 9

4. AMS 2759/6B section 8.6 (2009)

5. Friehling, P. B. et al., “On the Effects of Pre-oxidation on Nitriding Kinetics,” Technical University of Denmark, (2000), p. 103

6. Stiles M. et al., Heat Treat Materials, Vol. 53 (1998), p. 211

7. Linde Group, “Furnace Atmospheres,” Gas Nitriding and Nitrocarburizing, No.3 p. 10