TECHNICAL PAPER
Gaseous Nitriding:
In Theory and In Real Life
PAPER: Gaseous Nitriding: In Theory and In Real Life
Gaseous Nitriding: In Theory and In Real Life
Karl-Michael Winter
PROCESS-ELECTRONIC GmbH, A member of United Process Controls, Heiningen, Germany
[email protected], phone +49 7161 94 888 0
Abstract
Expert systems for gaseous nitriding, be it simulators or
controllers, are largely based on the Lehrer Diagram, which
shows the correlation between nitrogen-iron phases,
temperature, and the partial pressure ratio of ammonia and
hydrogen. While this theory is widely held, in real-world
applications, nitriding cycles do not always work in this way,
and thus materials, parts and results do not match Lehrer
Diagram’s calculations. In this presentation we will identify
additional parameters needed to complement the Lehrer
Diagram by using a variety of processes and samples.
Introduction
Dear readers, let me begin by saying that this article is based
on my field experience with nitriding process control, from a
novice to the veteran years. When I started working with
nitriding furnaces about 15 years ago, nitriding was a wellknown process in heat treating with many factors affecting the
process results. When the results are not as expected, like the
computer techie whose first troubleshooting question is “is
your computer plugged in?”, metallurgists and other experts
ask if the pre-nitriding process was performed properly. While
there is truth to this statement, to a large extent it does not
solve most problems.
My First Experience with Nitriding
The first nitriding control system we installed start-up was
relatively easy. All we had to achieve was a white layer
thickness of 0.0005-0.0006” and an oxidation layer on top of
the white layer within 5.5 hours. Having set up the process
recipe and running production in four furnaces successfully,
the parts were unexpectedly coming out faulty - the oxidation
layer could be easily wiped off. Examining the samples and
their micrographs we observed a negligible white layer, which
was insufficient for the oxide layer to adhere to properly.
As the parts came directly from the in-house machining shop
next door, we had the possibility to check if there were any
changes to the manufacturing process prior to nitriding. It was
concluded that a change in the cutting oil concentrations had
an adverse affect on the nitriding process. Because of the high
oil concentration, cleaning could not adequately remove all
residuals, and nitriding was therefore inhibited. Neither the
introduction of a pre-oxidation treatment with air at 660°F nor
nitrocarburizing with CO2 at 1080°F were sufficient at
removing the oil contaminants. Consequently, a return to the
original cutting-oil concentration and the nitriding process was
working fine again.
As it happened, the first setback I encountered was directly
related to a pre-nitriding process. But nitriding is complex, and
in the following sections, we’ll discuss a few factors that can
affect results.
As a General Rule of Thumb
It is common knowledge that the key control parameters used
to control the nitriding process and determine its outcome are
temperature as well as residual ammonia or dissociation, or
the more modern nitriding potential.
As a general rule of thumb:
• increasing the temperature will increase the case
depth and increase the white layer;
• increasing the residual ammonia and decreasing the
dissociation measured will increase the case depth
and white layer; and
• increasing the nitriding potential will increase the
case depth and the white layer.
This is valid as long as the same material is treated.
Temperature is without a doubt an important influence in
nitriding and all other case hardening treatments. At higher
temperatures the iron lattice provides more space for nitrogen
atoms to diffuse further into the metal part. Typical nitriding
processes are carried out at temperatures between 840-1100°F.
Beyond this temperature limit, for example at 1830°F a case of
interstitially-placed nitrogen is produced, which cannot be
considered nitriding. This process is more similar to
carburizing where the carbon is replaced by nitrogen.
To control the atmosphere conditions we have to know what
measuring equipment is used. An ammonia analyzer calculates
the ammonia percentage in the furnace, but readings are
actually taken of the residual ammonia in the exhaust. With a
burette or a hydrogen analyzer, the dissociation rate is based
on the percentage of all gases except ammonia in the exhaust.
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 2 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
Does point 2 of the rule of thumb hold true (ie., the more
ammonia the higher the nitriding effect)? To answer this
question we must look at what happens on the steel surface.
We begin by producing an iron nitride Fe4N, known as gamma
prime. If we use iron and nitrogen gas and apply massive
pressure to crack the N2 and then have the two Ns react with 8
Fes to build two Fe4Ns, the thermodynamic properties will
give us an equilibrium equation, depending on temperature
N2 + 8 Fe → 2 Fe4N
K1 = a²Fe4N / pN2 / a8Fe = f1(T)
pN2 = exp(2 * δG0Fe4N / R / T)
powder in it, heated up to a specific temperature and had
varying ammonia-hydrogen mixtures flowing though it to
determine the iron-nitrogen-phase of the powder. To avoid
changes in the phase due to cooling effects, the powder
samples were quenched.
In order to have proper controlled process gas, he set the
mixture on the inlet and measured the percentages on the
outlet. Up to a temperature of 1170°F, no differences were
noted. With increased temperatures, he detected a different
ratio of ammonia to hydrogen on the exhaust due to the
presence of nitrogen out of the thermal dissociation of the
ammonia in the vessel. From these results he designed a set of
diagrams, giving the phase boundaries between the ironnitrides as a function of temperature and %NH3,
log(p²NH3/p³H2) and pN2.
with R = 8.290 J/mol and T in Kelvin
δG0Fe4N = -17250 + 58.6*T [J/mol]
1, 2
Solving the equation using the standard enthalpies of the
participants, the nitrogen partial pressure needed at 930°F
comes to roughly 5000 atmospheres. This represents a water
level of more than 30 miles! Since there is no industrial
furnace where you can create such a big pressure, it must
come from somewhere else.
Ammonia Production
Ammonia is created by putting hydrogen together with
nitrogen into a vessel at high temperatures and high pressures.
3 H2 + N2 -> 2 NH3
In order to speed things up slightly, the whole process is done
using a catalyst. Obviously there is also a thermodynamic
equilibrium function for this reaction
K2 = pN2 * p³H2 / p²NH3 = f2(T)
which allows us to predict what temperature and pressure
conditions are expected at a certain ratio of N2, H2 and NH3.
As ammonia production takes place at temperatures between
850-1100°F and at pressures between 200-1000 atmospheres,
the efficiency is comparably low; at 930°F and 250
atmospheres, efficiency is about 15% 3.
Nevertheless, a dissociating ammonia molecule can easily
provide the nitrogen partial pressure needed to create iron
nitrides.
The Lehrer Diagram
Early in the last century Lehrer 4 did a lot of tests to evaluate
the iron-hydrogen-ammonia equilibrium. He used a noncatalytic and very small reactor, placed decarburized iron
Fig. 1: Iron-Nitrogen-Phase-Diagram according to Lehrer 4
depending on temperature and nitrogen partial pressure.
The Equilibrium
As %NH3 was always meant to be 100% of the gas minus the
percentage of hydrogen and the nitrogen gas pressure is
practically negligible the diagram we use today shows the
phase boundaries as a function of temperature and
log(pNH3/p1.5H2) known as the nitriding potential
KN = pNH3 / p1.5H2
out of the nitriding reaction
NH3 → N[ad] + 1.5 H2
Giving the equilibrium constant:
K3 = pNH3 / aN / p1.5H2
When setting the nitrogen activity as the partial pressure of
nitrogen, KN is related to the nitrogen pressure as
KN = K3 * aN = pNH3 / p1.5H2
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 3 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
And the effective nitrogen partial pressure out of the
dissociation can be calculated as
will not follow the NH3 content in the atmosphere or the
dissociation.
Kinetics
pN2 = KN² * exp(2 * δG0NH3 / R / T)
with R = 8.290 J/mol and T in Kelvin
δG0NH3 = -52090 + 113.9*T [J/mol] 1
Today we know that there is a temperature-dependent
equilibrium between the percentages of nitrogen within or
bound to iron that will balance in a specific concentration ratio
of ammonia and hydrogen. This can be measured and
therefore also controlled.
With an industrial nitriding process using ammonia, we will
undoubtedly encounter thermal dissociation. In that case, we
cannot merely control the measured value of NH3 or
dissociation (100%- NH3) in the exhaust. We must also
calculate the amount of nitrogen (Fig. 2), otherwise we may
end-up with different results.
KN=f(NH3,H2) vs KN=f(NH3,dNH3)
1
1000
0.9
0.8
100
0.7
KN
0.5
1
pH2
0.6
10
0.4
0.3
0.1
0.2
0.1
0.01
0%
20%
40%
60%
80%
100%
0
120%
NH3
KN[H2]
KN[d]
pH2[H2]
pH2[d]
pN2[d]
Fig. 2: Nitriding potentials and partial pressures of hydrogen
and nitrogen over ammonia percentage measured at the
exhaust for ammonia-hydrogen mixtures and ammoniadissociated ammonia mixtures.
But could there be a difference in the nitriding results between
an ammonia/hydrogen mixture and an ammonia/dissociated
ammonia mixture? Tests done in a small bell-type furnace at
the Institute for Material Science (IWT) in Bremen where the
nitriding potential was controlled both ways showed similar
results. As in Lehrer’s tests, at higher KNs the thermal
dissociation became more eminent and the hydrogen
controlled KN[H2]-curve moved towards the KN[d]-curve.
Therefore, if the partial pressure ratio of NH3 and H2 in the
form of the nitriding potential is giving the equilibrium
content of nitrogen in an iron surface, the nitriding process
To prove equilibrium in industrial processes the nitriding
committee of the German Association for Materials and Heat
Treatment (AWT) conducted a set of tests at commercial and
in-house heat treaters. AWT provided the iron powder, small
iron cubes, and a small container to hold the powder while in
an agitated furnace atmosphere. The samples were then given
to companies with in-house heat treat departments equipped
with nitriding potential-controlled furnaces. In a few words,
the test results were not very promising. The powder was
blown away by the furnace fan, and ultimately the test method
proved to be invalid – in fact what was measured was the
quality of the measuring and control systems.
However, one noteworthy exception was an in-house shop that
treated identical parts regularly using two different recipes and
yielded similar case results. The specification called for
processing small bars to have a defined case depth and surface
hardness with no more than 0.0002” white layer. In the first
process, KN-controlled uses dissociated ammonia for dilution,
in the second, the process uses nitrogen for dilution. When
measuring the residual ammonia in the exhaust, both processes
operated at about 12% NH3. When calculating the nitriding
potential, the first process ran at 0.25 whereas the second at 4,
and yet the parts come out identical in terms of surface
hardness, case depth and white layer thickness.
If we abide by point 3 of our Rule of Thumb increasing the
nitriding potential from 0.25 to 4, with temperatures held
constant, should have increased the white layer thickness
dramatically. So there must be a second effect influencing the
outcome, besides the equilibrium given by the nitriding
potential and the temperature.
Grabke 5 stated that after measuring the denitriding speeds of
iron in argon-hydrogen mixtures, the speed of either the
second stage (II) or the third stage (III) of ammonia
dissociation gives the speed for the nitriding reaction.
NH[ad] + H[ad] ↔ NH2[ad]
(II)
NH2[ad] + H[ad] ↔ NH3[ad]
(III)
In both cases the speed is proportional to the hydrogen partial
pressure in the gas, which gives nitrogen flow into the alphastructure based on the following equation(s)
j = Kv * pv/2 * (cαNequ – cαN,s)
with v representing the dissociation stage 2 or 3.
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 4 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
Moreover, the higher the hydrogen partial pressure, the faster
the nitrogen content in the surface will reach the equilibrium
content given by the nitriding potential.
If that’s the case, then the second process in the example
above did not have enough time to build a decent white layer
as should be expected with the higher nitriding potential.
We now have a higher hydrogen percentage but the hydrogen
partial pressure in the vessel is at 0.5 atmospheres with 0.19
atmospheres about 2/3 of the 0.33 atmospheres under normal
pressure. So, we should see an effect. If we decrease the
pressure even further at 0.1 atmospheres, we will see
NH3 [KN=3, 0.1 atm]
H2 [KN=3, 0.1 atm]
Diss [KN=3, 0.1 atm]
Nitriding at High Pressures
Jung 6 studied the influence of overpressure on the kinetics of
the nitriding process. He assumed that the epsilon growth will
follow the same principles as the kinetics in the alphastructure given by Grabke. His results proved that when
keeping the nitriding potential constant, increasing the
pressure will increase the growth rate of the white layer due to
the higher hydrogen partial pressure.
= 33 %
= 50 %
= 67 %
The graph shows the varying hydrogen partial pressure for a
constant nitriding potential dependent on furnace pressure
(Fig. 3)
pH2 = f(KN=3, furnace pressure)
2.00
KN = pNH3 / p1.5H2 [atm-0.5]
NH3 [KN=3, 1 atm]
H2 [KN=3, 1 atm]
= 56 %
= 33 %
and the measured dissociation comes to
Diss [KN=3, 1 atm]
= 44 %
Increasing the pressure in the furnace to 10 atmospheres and
controlling the same nitriding potential shifts these readings of
the exhaust to
NH3 [KN=3, 10 atm]
H2 [KN=3, 10 atm]
Diss [KN=3, 10 atm]
= 75 %
= 18 %
= 25 %
But even with a lower hydrogen percentage, the hydrogen
partial pressure in the vessel is at 10 atmospheres with 1.8
atmospheres 5.6 times higher compared to the 0.33
atmospheres under normal pressure. Consequently the white
layer growth increased dramatically on Armco-Iron and
carbon steels.
Nitriding at Lower Pressures
What if this time we kept the nitriding potential constant but
lowered process pressure? Using the same example as above
but lowering the pressure to 0.5 atmospheres, our readings will
change to
NH3 [KN=3, 0.5 atm]
H2 [KN=3, 0.5 atm]
Diss [KN=3, 0.5 atm]
= 49 %
= 38 %
= 51 %
pH2 [atm]
For a nitriding potential of 3 at normal pressure, the
percentages of ammonia and hydrogen in dissociated ammonia
measured on the exhaust would come to
1.50
1.00
0.50
0
2
4
6
8
10
12
furnace pressure [atm]
Fig. 3: Hydrogen partial pressure for a controlled nitriding
potential of 3 as a function of furnace pressure.
Nitriding in Diluted Atmospheres
Considering that the molecular nitrogen (pressure) is not or
only marginally present in the nitriding process, we can
assume that the dilution will act like a pressure decrease.
Zimdars 7 did a study on the influence of diluting oxinitriding
atmospheres with nitrogen while keeping the nitriding
potential and the oxidation potential constant. He found that
the layer growth rate depended on the ammonia percentage of
the process gas. At a nitriding potential of 3 and process
temperatures between 1020-1100°F, the dilution with a 15%
ammonia content had no practical influence on the formation
of the white layer or case. When diluted further to 5% NH3 in
the exhaust, the reaction is delayed as the nitrogen intake
through the material surface slows down the diffusion towards
the center of the part.
The slower reactions can be explained with the lower transfer
coefficient due to the lower hydrogen pressure (Fig. 4).
Obviously, in the presence of oxygen, the surface reaction or
the dissociation reaction will be affected, but the general
tendency is identical.
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 5 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
Fig.5: White layer thicknesses in microns on samples of
stamped 1006 nitrided at different temperatures and aiming
for the same nitrogen weight percentage.
transfer coefficient in Nitrogen diluted
atmospheres [570°C, KN=1]
5.00E-06
60%
The Influence of Carbon
4.50E-06
4.00E-06
3.50E-06
40%
3.00E-06
2.50E-06
30%
2.00E-06
20%
1.50E-06
1.00E-06
% NH3, H2
50%
10%
5.00E-07
0.00E+00
0%
20%
40%
60%
80%
0%
100%
nitrogen dilution
k3[d]
pNH3
pH2
Fig.4: Influence of Nitrogen dilution on the nitrogen transfer
coefficient in an nitriding atmosphere controlled to a nitriding
potential of 1 at 570°C / 1060°F.
General Rule of Thumb - Modification # 1
Given the aforementioned considerations, we can modify our
General Rule of Thumb as follows:
•
•
•
•
increasing the temperature will increase the case
depth and increase the white layer, provided the
atmosphere allows for formation of a white layer;
set the nitriding potential to match the desired
nitrogen content in the part surface at the given
temperature;
the reaction speed will be given by the partial
pressure of hydrogen. A dilution with nitrogen will
not dramatically slow down the process, as long as
the nitrogen intake does not come below the nitrogen
flow needed for diffusion and white layer growth.
increasing the furnace pressure will increase the
growth of the white layer.
This provides a basic recipe on how to change the nitriding
parameters in order to reach certain specifications. Figure 5
shows samples with the same white layer thickness and same
composition processed at different temperatures.
Additional factors that greatly influence nitriding results
include:
•
•
Parts composition and condition
Process cycle
Let’s begin with the material first. As with carburizing, in
nitriding there is also an alloying factor. And just as in
carburizing, this factor will decrease or increase the
equilibrium nitrogen content at the surface 8, 9. This is due to
the influence of alloying elements on the nitrogen activity
within the steel.
Carbon increases nitrogen activity 9, which lowers the
solubility, but at the same time shifts the phase boundaries to
lower nitrogen pressures, that is to say to lower nitriding
potentials.
When treating samples with different carbon contents, such as
AISI 1010, 1020 or 1045 in the same process, the white layer
thickness will become bigger with the increasing carbon
weight percentage (Fig. 6).
Fig.6: white layer thickness increases with increasing carbon
content in the material. Left AISI 1020, right AISI 1045, both
treated at 570°C / 1060F, 2 hrs, NH3 + 10% CO2.
Might this be due to the higher local pressure or nitrogen
activity, which simultaneously decreases the case depth as the
solubility decreases and in turn decreases the slope of
interstitially dissolved nitrogen towards the core?
Let’s just say partially. First of all, in untreated carbon steels
with higher carbon contents, we treat more or less a statistical
material, as ferritic grains mix up with pearlitic grains,
averaging a 1045 for example. But on a smaller scale we
either treat a grain of 1010 or a grain of 1075. Second, at
temperatures between 930-1110°F, the interstitially dissolved
carbon in iron maximizes to less than 0.02 wt%. We actually
deal with ferrite and cementite (Fig. 7).
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 6 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
The cementite structure can transform immediately into a Fe23[NC] epsilon phase, whereas the ferrite will start with the
slower growing gamma prime phase during the nitriding or
nitrocarburizing cycle aiming for a white layer.
Figure 9 shows the nitrogen and carbon profile of the nitrided
white layer of low carbon steel. The white layer ends at about
12 microns / 0.0005” where the carbon has peaked. The steel
was heavily decarburized, which is seen from the low carbon
content at 20 microns / 0.0008” where it dropped from the
original 0.15 wt% down to about 0.05 wt% (unfortunately the
GDOES does not show the original substrate C content which
goes up to 0.15 wt% at approximately 40 microns / 0.0016”
depth).
Fig.7: Nitrocarburized Ck 45 with grains of ferritic and
pearlitic structure. The pearlitic grain shows inner nitriding of
the cementite lamellae to epsilon.
Carbon content in the steel, as well as the grain condition and
the inner stresses from the manufacturing process will have an
influence on the white layer growth. Figure 8 compares the
white layers of a stamped 1006 and a not-stamped 1045
nitrocarburized in the same process.
Fig.9: Typical carbon peek at the interface of the white layer
to the diffusion layer on a nitrided carbon steel.
As long as there is no white layer acting as a diffusion barrier,
due to the lower diffusion coefficient for carbon as compared
to iron, carbon will effuse from the part’s surface leading to a
drop in the original carbon content of the substrate.
Fig.8: Nitrided samples of stamped 1006 (left) and not
stamped 1045 (right) show the same white layer thickness
when treated in the same load for4 hrs at 1020 F, KN =2.64.
Generally speaking, a stamped low carbon steel part can be
treated like a non-stamped part of a higher grade carbon.
Nitriding vs. Nitrocarburizing
Nitriding of a high-carbon part is similar to nitrocarburizing a
low-carbon part. When nitriding carbon steel, we typically see
a decarburization of the material surface.
The moment a closed white layer is built, the carbon effusion
will slow down and a peak of carbon builds up at the interface
to the diffusion layer. When treating the material in such a
way as to create a white layer as fast as possible by providing
a high nitriding potential at the beginning of the process, the
carbon effusion is minimized. Having high carbon content at
the interface, which is actually mostly kept within cementite,
will enforce the formation of carbonitrides. It need few
nitrogen atoms to transform several Fe3C into Fe2-3[NC].
Naumann and Langenscheid 10 developed phase diagrams for
the combined iron-nitrogen-carbon phases, where the weight
percentages of carbon and nitrogen in an epsilon phase vary
with temperature. Figure 10 shows the phase boundaries of
epsilon in wt% C and wt% N depending on temperature.
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 7 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
Figure 11 shows that the nitriding potentials needed to form
typical nitrides in alloyed steels are much lower than the
potentials needed to build a white layer (red lines).
Nitrides
600
Temperature
580
560
540
520
500
Fig.10: Phase boundaries of epsilon Fe2-3[NC] for
temperatures between 500°C / 930°F and 650°C / 1200°F
according to Naumann-Langenscheid 10. The inner red section
marks a region that will give epsilon independent of the
process temperature.
Using the equations given by Kunze 11 we can calculate the
nitriding potential KN and the carburizing potential out of the
Boudouard reaction KCB needed to form an epsilon phase at a
temperature of 1080°F, and staying within the red section
given by figure 8, and gradually exchanging the nitrogen
content with carbon (Table 1).
Table 1: Table shows the different nitriding and carburizing
potentials needed to form epsilon at 1080°F. Calculations
done according to Kunze 11, Epsilon phase determined
according to Naumann and Langenscheid 10.
wt % N
8
7.5
7
6.5
wt% C
0.1
0.75
1.25
1.75
KN
1.99
1.75
1.44
1.21
KCB
0.02
0.19
0.36
0.55
1.E-24
1.E-20
Where carbon increases the nitrogen activity in steel and
therefore lowers the solubility, most nitride building elements
like titanium, vanadium or chromium etc., have the reverse
effect. In theory titanium will increase the solubility of
interstitially kept nitrogen atoms, but the interaction force of
titanium and nitrogen is so high that they will immediately
form titanium nitrides.
1.E-12
1.E-08
1.E-04
1.E+00
Nitriding Potential
gamma prime
e
AlN
epsilon
CrN
VN
xi
Cr2N
Mo2N
g'
TiN
Si3N4
Fig.11: Nitriding potentials needed to form MexNy in a
nitriding atmosphere for temperatures between 500°C/930°F
and 600°C/1110°F, calculated using thermodynamic data
from Barin and Knacke 12.
Contrary to carburizing where the higher soluble amount of
carbon, due to the alloying elements, is kept interstitially and
diffuses into the steel, in nitriding, the nitrogen atoms will
diffuse more or less immediately forming nitrides until all
nitride builders are bound. This causes a completely different
profile inside the part similar to what we are used to in
carburizing processes (Fig. 12).
C
Precipitation Profile vs.
If we consider the carbon content in the steel as the driving
force for the epsilon formation, we can actually see that the
original phase boundary of gamma prime to epsilon given by
the Lehrer diagram shifts to lower nitriding potentials with
increasing carbon contents.
Nitride Building Elements
1.E-16
Interstitial Diffusion Profile
Depth
Fig.12: Different types of case profiles, with and without
precipitation of Me-nitrides.
As the amount of nitrogen able to enter the steel surface is still
dependent on the space within the lattice and therefore nearly
constant, the precipitation front moves at a much slower speed
compared to the diffusion front in pure iron. On the other
hand, the more nitrides that form due to an increasing
percentage of alloying elements, the total amount of nitrogen
kept in the surface increases. This has a dramatic effect on the
nitrogen flux through the surface.
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 8 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
Using the data given by Fromm and Gebhardt 9, the
equilibrium nitrogen content in iron at 1080°F is
approximately 0.1 wt%, whereas 1 wt% of chromium in the
alloy can hold 0.27 wt% in chromium nitrides.
To have repeatable and somewhat predictable results the use
of dissociated ammonia instead of nitrogen is a much better
way to control the nitriding potential.
Acknowledgments
General Rule of Thumb - Modification # 2
We can now modify the “General Rule of Thumb” even
further to include:
• Increasing the temperature will increase the case
depth and increase the white layer, provided an
atmosphere allowing for formation of a white layer
• Set the nitriding potential to match the desired phase
on the parts surface. Carbon will shift the boundary
to the epsilon phase to lower nitriding potentials,
increasing amounts of nitride building elements will
shift the boundary to higher nitriding potentials
respectively
• Nitride building elements have a high impact on the
nitrogen flux needed to saturate the structure and
therefore
• Diluting the nitriding atmosphere with nitrogen or
treating the part at low pressures will stop a proper
nitriding of high alloyed steels earlier compared to
low alloyed or carbon steels.
• Increasing the furnace pressure will increase the
growth of the white layer, but this effect will slow
down by increasing the nitride building alloying
elements.
• Stamped parts will behave like non-stamped parts
with higher carbon content.
Conclusions
Obviously there are many more factors influencing the
nitriding process than covered in this article. While I am still
on the learning curve, I must point out that nitriding is far
more complicated than carburizing and that a proper
prediction of the process outcome is dependent not only on
temperature and nitriding potential or even dissociation.
At lower temperatures the steel structure does not change, but
machining the part prior to heat treating has a big influence.
The carbon content and the alloying elements will either shift
the phase boundaries or lead to a high nitrogen intake which is
actually driven by the hydrogen partial pressure in the furnace
atmosphere.
Ammonia and hydrogen together define the nitriding
probability, with KN setting the equilibrium nitrogen content,
and hydrogen giving the reaction speed. Yet in spite of this a
nitriding potential control should be favored over dissociation
or residual ammonia.
I wish to thank Paulo Abrantes (Nitrex Metal Inc. Canada) for
having performed numerous test cycles and for supplying
most of the micrographs included in the article. I also want to
thank Professor Dr. Hoffmann and Dr. Klümper-Westkamp at
IWT in Bremen for the GDOES evaluations and the endless
discussions on nitriding phenomena.
References
[1] Jentzsch, W.-D., Mathematische Modellierung der
Ausbildung von Stickstoffkonzentrationsprofilen und des
Wachstums der gamma’ –Phase (Fe4N) während der
Nitrierung von Eisen in Ammoniak-WasserstoffGemischen, Dissertation Bergakademie Freiberg (1976).
[2] Jentzsch, W.-D. and Böhmer, S., „Thermodynamische
Betrachtungen zur Nitridbildung auf Eisen beim
Gasnitrieren“, Kristall und Technik 12 (1977), p 1275
[3] Schröter, W., Lautenschläger, K.-H., Bibrack, H.,
Taschenbuch der Chemie, 17. durchgeseh. Aufl. – Thun
(Frankfurt am Main 1995), p 476.
[4] Lehrer, E., „Über das Eisen-Wasserstoff-AmmoniakGleichgewicht“, Zeitschrift für Elektrochemie 36
(1930),pp. 383-392.
[5] Grabke, H. J., „Reaktionen von NH3, N2, H2 an der
Oberfläche von Fe, I. Zur Kinetik der Nitrierung von
Eisen mit NH3/H2-Gemischen und der Denitrierung. II.
Zur Kinetik der Nitrierung von Fe mit N2 und der
Desorption von N2“, Berichte Bunsenges. phys. Chemie
72 (1968) Nr. 4, pp. 533-548.
[6] Jung, M., Entwicklung eines geregelten Drucknitrierprozesses, Dissertation Universität Bremen (1999), pp.
126-131.
[7] Zimdars, H., Technologische Grundlagen für die
Erzeugung nitridhaltiger Schichten in stickstoffangereicherten Nitrieratmosphären, Dissertation Bergakademie
Freiberg (1987), pp. 37-39
[8] Zheng, X., Nitrogen Solubility in Iron-Based Alloys and
Powder Metallurgy of High Nitrogen Stainless Steels,
Dissertation Swiss Federal Institute of Technology
(1991), pp. 9-38.
[9] Fromm, E. and Gebhardt, E., Gase und Kohlenstoff in
Metallen, Springer Verlag Berlin / Heidelberg (1976), pp.
578-613.
[10] Naumann, F.K., Langenscheid, G., „Beitrag zum System
Eisen-Stickstoff-Kohlenstoff“, Archiv Eisenhüttenwesen
36 (1965) 9, pp. 677-682.
[11] Kunze, J., „Thermodynamische Gleichgewichte im
System
Eisen-Stickstoff-Kohlenstoff“,
HärtereiTechnische Mitteilungen HTM 51 (1996), pp. 348-354.
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 9 of 10
PAPER: Gaseous Nitriding: In Theory and In Real Life
[12] Barin, I. and Knacke, O., Thermochemical Properties of
Inorganic Substances, Springer Verlag Berlin/Heidelberg
(19
Copyright © 2009, United Process Controls. All rights to copy, reproduce and transmit are reserved.
Page 10 of 10