You are on page 1of 13

Corrosion Science, Vol. 37, No. 3, pp.

467-479, 1995
Pergamon Copyright 0 1995 Elsevier Science Ltd
Printed in Great Britain. All rights reserved
OOlG938)3/95 $9.50+0.00

0010-938x(94)00148-0

THE INFLUENCE OF NITROGEN AND MOLYBDENUM


ON PASSIVE FILMS FORMED ON THE AUSTENO-
FERRITIC STAINLESS STEEL 2205 STUDIED BY AES
AND XPS

CLAES-OLOF A. OLSSON

CITU, University College of Falun Borlkge, P.O. Box 764, S-781 27 Borkinge, Sweden

Abstract-Passivation of the austeno-ferritic (duplex) stainless steel 2205, and the corresponding
austenitic and ferritic single phase alloys, has been studied with Auger electron spectroscopy (AES) and
X-ray photoelectron spectroscopy (XPS) with a focus on the lateral and depth distributions of nitrogen
and molybdenum in and near the passive layer. The results can bc explained by the bipolar model of the
passive film; the film is stabiliscd by the interaction of molybdates and ammonium ions on the surface. The
synergism between nitrogen and molybdenum is of particular importance for duplex stainless steels. since
they owe part of their corrosion resistance to interaction between the nitrogen-rich austenite and the
molybdenum-rich ferrite.

INTRODUCTION

Several attempts have been made to investigate the beneficial effects of nitrogen on
the passivity of austenitic stainless steels using electron beam and ion spectroscopy
techniques. For austeno-ferritic stainless steels, henceforth referred to as duplex
steels, nitrogen has also been found to be, through its relatively high diffusion
coefficient, the element determining the sites for the austenite and ferrite regions.
The steels used in this study were the commercial duplex stainless steel 2205, and two
single phase alloys with compositions corresponding to those of the austenite and
ferrite phases in 2205. The intention was to investigate the influence of nitrogen and
molybdenum on the passivity of stainless steels and the effects of these elements on
the interaction between the austenite and ferrite phase. Interaction between the
phases is an important factor for the corrosion resistance of duplex stainless steels.
It is not until recently that surface analytical methods have reached the sensitivity
needed to detect nitrogen concentrations of about 0.5 at%, with a lateral resolution
sufficient for distinguishing between the different phases on duplex steels. Even so,
nitrogen analysis is rather tedious and an average acquisition time for a phase-
resolved, Auger sputter depth profile from a duplex stainless steel may amount to
approximately 10 h. By combining the depth information from Auger sputter depth
profiles and the oxidation state and depth information from angular-resolved X-ray
photoelectron spectroscopy, or ARXPS, it is possible to draw conclusions about the
interaction between nitrogen and molybdenum.
Molybdenum as an alloy element in stainless steels has attracted a great deal of
interest and has been studied extensively in the literature.‘.2 Its relatively low bulk

Manuscript received 5 September 1994.

467
468 Claes-Olof A. Olsson

concentration combined with a complex chemistry showing a large variety of


oxidation states has opened the field for a number of different theories on the
passivating mechanisms involving molybdenum. The theories concerning the ben-
eficial effects follow two main paths: either that molybdenum enhances the stability
of the passive film from the solution side, or that the presence of molybdenum in the
metal phase near the oxide/metal interface limits the dissolution rate of iron and
chromium.3
Electron spectroscopy studies of nitrogen in stainless steel have been carried out
previously, e.g. by Lu et al.” who found a strong enrichment of nitrogen at the
oxide/metal interface by acquiring Auger sputter depth profiles. Traces of am-
monium irons on the surface led to the conclusion that ammonium ion formation
plays a decisive role in the prevention of pit initiation, as was previously indicated by
findings of ammonia in the solution by Osozawa and Okato.’ The formation of
ammonium ions near the oxide/metal interface is believed to cause a local increase in
pH, thus mitigating the environment and making it more stable for molybdenum
compounds beneficial to passivity.” However, AES studies of the interaction
between molybdenum and nitrogen are further complicated by the difficulty of
detecting molybdenum on the surface using AES.’ Mechanisms involving chemical
reactions in the liquid part of the Helmholtz double layer are also, for obvious
reasons, difficult to trace with analysis techniques demanding UHV.

EXPERIMENTAL METHOD
Corrosion experiments
To investigate passive layer formation on duplex stainless steels, three types of alloys were used: single
phase austenitc and ferrite, and samples from the commercial duplex stainless steel 2205 (DIN I .4462.
UNS S3180.3). The single phase samples were dcsigncd to have compositions close to those of the
respective phases in the duplex stainless steel, with the cxccption of the nickel content in the single phase
austenitc that was increased by about 2 wt’% units to cnsurc phase stability. For the 220.5 samples. the
samples were cut from the outer thirds (parallel to the plate surface but avoiding the outermost parts) of a
40 mm thick plate. The composition of the single phase alloys and the Q and y phases of the duplex steel
wcrc checked using electron-probe X-ray micro-analysis, with special emphasis on the nitrogen
concentration,x the concentrations can bc found in Table I.
Samples. 17 X I7 X I mm, from the three different steels were wet ground and then potishcd with
diamond paste; the finest size being I ,um, to achieve a clean and smooth surface. After polishing, the
samples were ultrasonically cleaned in trichloro-cthylcne, acetone and methanol, and then stored in
laboratory atmosphere until they were subject to corrosion experiments. lmmcdiatcly before corrosion
expcrimcnts the samples were once again ultrasonically cleaned in ethanol and dcionised water. A heat of
2205 with a nitrogen content close to the upper limit of the specifications was sclcctcd to make nitrogen
detection easier.
The samples were polar&d using a Santron EMS potcntiostat in a corrosion cell dcsigncd to avoid
crevice corrosion."After activation at -700 mV(SCE). the samples were polarised in a 1 M NaCl solution
kept at 300K to potentials between -200 and +WO mV(SCE) to invcstigatc passive layers formed under
different conditions. The test arca was 0.5 cm’, which is smaller than the arca of I cm’ normally used for
this type of corrosion cell, to make introduction of the samples into the spectromctcr possible. After the
corrosion cxpcriment, the sample was rinsed in dcioniscd water, dried in an argon jet and transfer& into
UHV. The transfer time from corrosion ccl1 to UHV was. for the Auger measurements, in no case longer
than S min.

Surfuce
unu1ysi.s
The Auger analyses wcrc pcrformcd using a Pcrkin-Elmer PHI 660 Scanning Auger Microprohc.
operated with a primary beam voltage of IO kV and an electron current of 250 nA. The Auger electron
spectra wcrc numerically differentiated and a \emi-quantitative analysis was performed using peak-to-
m
=:
8
Table 1. Concentrations in wt% and at% for 2205 and the respective single phase alloys. The bulk analyses were obtained using B
optical emission spectrometry and the phase resolved analyses were obtained using EPXMA

Fe Cr Ni MO Mn N
Sample/phase wt% at% wt% at% wt% at% wt% at% wt% at% wt% at%

Single phase ferrite bal. 65.6 24.4 26.9 3.9 3.8 3.6 2.2 1.4 1.4 - -
Single phase austenite bal. 65.4 20.7 22.2 8.8 8.4 2.1 1.2 2.0 2.0 0.2 0.8
Ferrite in 2205, 22.3% 64.4 66.1 23.9 26.3 3.9 3.8 3.7 2.2 1.4 1.5 0.05 0.2
Austenite, 2205 66.6 67.5 20.4 22.2 6.6 6.4 2.2 1.3 1.6 1.6 0.2 1.0
Bulk, 2205 bal. 67.1 21.8 23.5 5.7 5.5 2.9 1.7 1.4 1.5 0.2 0.7
470 Claes-Olof A. Olsson

peak heights of the Auger transitions of the respective clemcnts, using specific sensitivity measured for the
sputter-cleaned duplex stainless steel 2205. For Cr, the negative excursion of the 529 cV peak was used
because of the overlap with the oxygen peak. Depth profiling was performed using a rastcred 1 keV argon-
ion beam. The sputtering rate varied between 1.5 and 3.2 nm mini’ on TaZOS and was calibrated on a
weekly basis.
The XPS measurements were performed using a Scienta ESCA 300. Monochromatised Al K,,
radiation was used for excitation. The total energy resolution of the X-ray radiation and the electron
analyser was 0.4 eV. No sputtering was performed prior to analysis. The recorded binding energies of the
photoelectrons from the metal substrates arc in good agrcemcnt with published values.“’ The spectra were
fit using Shirley background subtraction and Gaussian line shapes with addition of an asymmetry factor for
the metal peaks. The compositions of the passive layer and the metal wcrc calculated for different take-off
angles assuming homogeneous distribution of the elements, and by using the electron energy analyser
transmission function, Scoficld cross-sections” and the equation:”
x
In = uo,(hv)T(E/J N,,(Z)exp~;l”u(E,)\‘“/~dz.
(1)
1 ;=_(I

where 1, is the photoelectron intensity for element A, a is a spectrometer constant, a, is the cross-section
for emission of a photoelectron, hv is the energy of the X-rays. r(E,) is the analyser transmission
function, E, is the kinetic energy of the photoelectron. R, is the inelastic mean free path for
photoelectrons in a given medium, N, is the atomic density of the A atoms at a depth z in the sample.
where the z-axis is parallel to the sample normal and 0 is the exit angle of the photoelectron. measured
with respect to the sample surface; the more grazing the analyser angle, the higher the surface sensitivity of
the analysis.

EXPERIMENTAL RESULTS

To establish the stability regions of the three different alloys, potentiodynamic


scans were carried out in a 1 M NaCl solution at 300 K. The breakdown potentials,
defined as where the sample current exceeded 0.2 mA cm-‘, were 1350, 1350 and
1150 mV(SCE) for the duplex steel and the single phase austenite and ferrite alloys,
respectively. The difference in breakdown potentials could, at least in part, be
explained by non-metallic inclusions in the single phase alloys, produced as labora-
tory heats.
By comparing the passive layers formed during exposure to a corrosive medium
of the three different alloys, it is possible to investigate the interaction between the
alloy elements and the respective phases in the duplex stainless steel. Fig. 1(a) shows

b) Ausknile
70

0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14

Depth. ““I Depth. nm

Fig. 1. Auger sputter depth profiles acquired from: (a) a single phase ferrite and a ferrite
region on 2205; and (b) the corresponding plot for austenite. The samples were all polarised
to +600 mV(SCE) in a IM NaCl solution at 300 K.
Effects of N and MO on the passivity of 2205

Iron and Oxygen b) Chromium

?
0
0 2 4 6 8 10 12 14

c) Nickel d) Molybdenum

OO-_4 00-I 14
Depth, nm Depth. nm

Fig. 2. Comparisons between the austenite and the ferrite regions on 2205 samples
polarised to different potentials for 60 min at 300 K for: (a) oxygen and iron; (b) chromium;
(c) nickel; and (d) molybdenum. The thickness increases with increasing potential (a) and
the chromium content drops as the potential approaches the critical pitting potential. The
nickel enrichment at the oxide/metal interface is not correlated to the polarization. The
apparently constant increase in molybdenum content with depth is caused by preferential
sputtering effects. What appears to be surface molybdenum is actually small amounts of
chlorine adsorbed on the surface.

a comparison of the major constituents of the passive film between Auger sputter
depth profiles acquired from a ferrite region in 2205 and the single phase ferritic
alloy. Figure l(b) shows the corresponding plot for austenite. The passive layers
were formed after exposure to a 1 M NaCl solution for 60 min at 600 mV(SCE). The
single phase oxide films are thicker than the corresponding duplex oxide films. Apart
from oxide layer thicknesses, the compositions of the passive films show good
correlation between the single phase alloys and the respective phase regions on the
duplex steel. Figure 2 shows further comparisons between duplex samples kept at
different potentials in the 1 M NaCl solution of: -700, -200, +600 and +900
mV(SCE).
Figure 2(a) displays Auger sputter depth profiles of oxygen and iron, showing
that the passive layer thickness increases with a higher polarisation, i.e. when the
sample is closer to the transpassive potential; this has previously been established for
single phase alloys by, for example, Olefjord et ~1.‘~ A slightly lower bulk concen-
tration of iron in the ferrite is also clearly distinguishable. The variations in passive
layer thicknesses caused by the different polarisations are also reflected in Fig. 2(b)
and (c), where the chromium and nickel peaks are shifted to larger depths in the
material with increasingpolarisation. Enrichment of nickel in the metal closest to the
oxide/metal interface, as shown in Fig. 2(c), shows no clear dependence on the
472 Claes-Olof A. Olsson

Fig. 3. Typical Auger nitrogen sputter depth profiles from adjoining sites on 2205.
Nitrogen is present on top of the passive layer and at the oxide/metal interface. Nitrogen is
cnrichcd also in the ferrite phase.

polarisation potential, neither for the austenite nor for the ferrite phase. The
absolute level of the nickel enrichment in the ferrite phase is lower due to a lower
bulk content. The chromium contents of the passive films, shown in Fig. 2(b), are
lowered as the polarising potential approaches the breakdown potential. Higher
fractions of chromium in the passive film on the ferrite phase are also what would be
expected by simply comparing the bulk contents. For molybdenum, the apparent
concentration gradients in the molybdenum profiles, that can be seen in Fig. 2(d).
were caused by preferential sputtering effects;14 the sensitivity factor for molybde-
num was chosen to match the bulk content at equilibrium sputtering conditions.
What, at a first glance, may be interpreted as surface molybdenum, is actually small
amounts of chlorine adsorbed on the surface; the effect is caused by the overlap
between the CILMM and MO,,, Auger transitions.
Figure 3 shows Auger sputter depth profiles for nitrogen on the respective phases
in 2205 kept for 60 min at +600 mV(SCE) in a 1 M NaCl solution. The profiles clearly
indicate an enrichment of nitrogen to the oxide/metal interface on top of the
austenite as well as on the ferrite phases. The noise level for nitrogen was suppressed
by utihsing very long acquisition times, in this case 10 min, for each nitrogen
spectrum, resulting in a noise level of the order of 0.2 at%.
The XPS measurements, albeit suffering from limited lateral resolution, give
more information about the oxidation states of the species present on the surface. By
comparing ARXPS data from single-phase alloys with those acquired from duplex
samples, it is possible to, within certain limits, deduce which features belong to which
phase on the duplex material. Figure 4 shows the curve fits for the XPS spectra from
the elements having the highest concentrations in the passive films. Of particular
interest for duplex steels are the nitrogen and molybdenum spectra. The presence of
Cr(V1) is also unexpected, given the electrochemical conditions under which the
passive film was formed. One explanation was given by Clayton and Lu,‘” who
argued that CrO:- could be formed in the highly hydrated state where the CrJ’
species are surrounded with Hz0 and OH- ions, i.e. an environment having a higher
pH more suitable for formation of CrOi-. For the solid state Cr03 species, they
argued that due to the fact that Cr03 and Cr?O, have similar standard free
Fe Zp, TOA=PO CT Zp, TO/\=90 0 Is, TOA.40’ N 1s nnd MO 3p, TOA=IO’
I I 1 I, 1 I, I I-
,-“‘I.“,
..
Fm-lle ,,‘.
.
. *.- .

-* l,...
.. ..*. _.*_ * * .
-..‘.*’
.*. *:’ .‘...
*. i..;. ’
. :*\:*. _.
....
::.* . :. * . ...- .** ‘_ ._*
. .* .*....

1...1..~1.~.1...1...,_
..
I I , 1
2% 234 232 23 2uI 226
Binding Ewgy, cv

Fig. 4. Curve fits used for quantification of XPS spectra. The use of commercial alloys with fairly low contents of molybdenum and nitrogen makes the
quantification of these elements uncertain.
474 Claes-Olof A. Olsson

Molybdenum
1

IL I I I I /
30 40 50 60 70 SO 90 loo 0 20 40 60 80 100
Take-Off An& Take-Off Angle

Fig. 5. Calculations of apparent concentrations of the most important oxidation states for:
(a) molybdenum; and (b) nitrogen for different analyser take-off angles. The lower the
analysing angle, the higher the surface sensitivity of the measurement. Of particular interest
is the relatively high abundance of MO(W) on the austenite having the lowest average bulk
concentration of the samples analyscd.

enthalpies, one would expect them to co-exist in the glassy inner oxide of the passive
film. The same reasoning can also be carried out for molybdenum.
The curve fits of Fig. 4 form the base for the calculations of apparent compo-
sitions for molybdenum and nitrogen; their variation with the respective take-off
angles is shown in Fig. 5. The take-off angle is defined with respect to the sample
surface; the surface sensitivity being highest at analyser angles close to grazing. The
angular dependence of the major alloy elements-iron, chromium and nickel- is
close to the dependence usually found on austenitic stainless steels. This means that
virtually no oxidised nickel was found on the surface, whereas for chromium and
iron, hydroxide bonds and higher oxidation-number compounds dominated towards
the outer surface monolayers, as compared to the more stable oxide formed deeper
down in the film.
The apparent molybdenum concentrations, as seen in Fig. 5(a), were calculated
for take-off angles 90” and 40” with respect to the sample surface. At a nearer-to-
grazing analyser angle, the Mo(V1) contribution dominates even more and there is
only a small contribution from metallic molybdenum, which makes a comparison
between the two oxidised states less reliable. Of particular interest is the steeper
slope of the molybdenum content in the single phase austenite, which indicates a
higher surface content of MO(W) on the single phase austenite than on the other two
samples, although its bulk content of molybdenum is lower than for the other two
alloys. The corresponding calculations for nitrogen are shown in Fig. 5(b) and show a
strong surface enrichment of ammonium on the duplex steel. The angular depen-
dence of the line designated as NH3 is not as pronounced, since the peak attributed to
NH, has contributions from atomic nitrogen in the bulk as well. This fact, combined
with the relatively low abundance of nitrogen in the steel, explains the comparatively
high noise level in these calculations.
Effects of N and MO on the passivity of 2205 475

The thicknesses of the passive films were calculated using a model developed for
XPS data by Brox eta1.;16 the calculated passive film thicknesses were 3.4 nm for the
single phase austenite and the duplex samples and 3.1 nm for the single phase ferrite.
The difference, albeit very small, gives the same ranking of oxide layer thicknesses as
the sputter depth profiles shown in Fig. 1.

DISCUSSION

Nitrogen and molybdenum are two elements with a strong beneficial influence on
the passivity of stainless steels. This influence is, for example, expressed by their high
coefficients, compared to chromium, in the pitting resistance equivalence (PRE)
formula:

PRE = %Cr + 3.3 x %Mo + 16 x %N. (2)


The factor given for nitrogen in the literature varies between 10 and 30, but 16 is a
frequently used value. l7 There are several proposed mechanisms for how molybde-
num may interact with other elements in the passive film to improve the stability, and
enhance the ability, for pit repassivation. One mechanism, supported by AES
measurements within corrosion pits, is that molybdenum assists in the repassivation
process through adsorption in pits as Mo(VI).‘* Another mechanism ascribed to
molybdenum is that it could limit the diffusion rate by replacing iron and chromium
at the oxide/metal interface. ly
The fact that the corrosion resistance of duplex stainless steels is higher for stress
corrosion cracking and crevice corrosion than for austenitic stainless steels having
the same PRE values has been indicated previously, e.g. by Combrad et al.20 As for
pitting, the picture is less clear, but the remarkably high corrosion resistance of some
duplex stainless steels in certain chloride media, suggests the existence of synergistic
effects.
Some of the properties of passive layers on stainless steels can be explained by a
bipolar model proposed by Sakashita and Sate.” The passive film is, in this model,
divided into a negatively charged, cation-selective outer layer and a positively
charged, anion-selective inner layer. The passive film will then operate as a diode
under reverse bias and resist further dissolution. The negatively charged outer layer
will repel the ingress of hydroxide and chloride ions; the positively charged inner
layer will resist the dissolution of metal in the oxide/metal interface. The influence of
molybdenum as an alloy element and as a corrosion inhibitor in the solution has been
investigated by Lu et al. 22who found molybdenum to be a critical element for making
the passive film rectifying. If there was no molybdenum present, neither as a
corrosion inhibitor in the solution nor in the metal, the bipolarity of the passive film
was lost. The XPS measurements in this study also gave results that could be
interpreted as signs for bipolarity; the indications, however, should be stronger in
solutions having a lower pH.
Molybdenum is ascribed multiple properties in the bipolar model. It is believed to
take part in the deprotonation of hydroxide by acting as an electron acceptor,
creating the oxygen abundance in the inner regions of the passive layer and helping in
the formation of a stable CrO&r,O,-protecting oxide. The other effect of this
reaction is the production of protons; they will be drawn towards the solution by the
cation selective properties of the outer layer. l5 This will increase the proton activity
476 Claes-Olof A. Olsson

on the surface and assist in the formation of ammonium ions, according to reaction
(3). This is one of the conceivable mechanisms for synergism between molybdenum
and nitrogen.
The idea of nitrogen dissolving from the stainless steel in the form of ammonium
ions were originally suggested by Osozawa and Okato,’ probably via the reaction:

N + 4Hf + 3e- $ NH:. (3)

An important effect for nitrogen is the enrichment to the metal side of the
oxide/metal interface that will expose a comparatively inert surface should the film
fracture or undergo electrochemical breakdown.2’ Willenbruch et al. have pointed
out that, whereas nickel and molybdenum nitrides are relatively unstable, combined
nitrides such as NizMosN show very strong lattice bonding.‘” Such a nitride could, if
present at the oxide/metal interface, bring down the dissolution rate considerably.
Likely sites for nitride formation are kinks, which also may serve as sites for metal
dissolution. Measurements of nitrided layers show that this is the case in neutral
chloride media, whereas in acidic solutions, nitrided iron is dissolved faster than pure
iron.25 If nitrogen is present on the surface as some form of nitride, Willenbruch et al.
suggest the reaction:2”

2 CrN + 3 H20ad, -+ Cr203 + 2 NH,,

which will result in the formation of NH, ligands, or perhaps NH: via further
protonation. Clayton and Martin emphasised that nitrogen reactions, particularly at
higher potentials, are extremely sluggish,2h and that there is a possibility for the
formation of soluble oxidized species, for example, via a reaction suggested by
Pourbaix.*’

NH,+ + 2H,O $ HNOz + 7Hi 6ee, (5)

which may be dispersed in the solution and not observed in surface analyses. That
nitrogen reactions are largely controlled by kinetics rather than by reaching the
thermodynamic equilibrium has also been confirmed by Jargelius,28 who demon-
strated the dissolution of nitrogen as NH: in a potential range considerably wider
than that normally referred to as the region of stability for NH:. The sluggish nature
of the nitrogen reactions makes them susceptible to different types of stimulation,
e.g. an increased production of protons in the liquid/oxide interface assisted by the
presence of molybdates.
The marked enrichment of nitrogen to the oxide/metal interface in the Auger
sputter depth profiles, shown in Fig. 3, is in good agreement with previous
results.22,23 Even if there is no formation of nitrides, nitride enrichment in the
interface could block kink sites and, in that way, bring down the dissolution
velocity.2Y If, on the other hand, the steel is tested in a solution with additions of
NaNO,, to simulate near-passive-layer electrolytes, as investigated by Newman and
Shahrabi in low-pH environments. Their results suggest that nitrogen plays an
important role in the electrolyte.2y They also found the passivating characteristics of
the steel to be almost independent of the nitrogen bulk content.2y For nitrided pure
iron, the dissolution rate is higher in acidic media;30 the solution in and near pits has
Effects of N and MO on the passivity of 2205 477

been found to be highly acidic due to metal hydrolysis. 31 If the dissolution rate of iron
is indeed increased, the result would, on a stainless steel, be an increase in the
enrichment of chromium and molybdenum in the surface layers. One synergistic
effect between nitrogen and molybdenum could be that nitrogen buffers the pH of
the near passive layer solution and thus alters the passive film environment to
become more stable for compounds containing molybdates, as suggested by, e.g. Lu
et ~1.~Since the presence of molybdate in the passive film is an important parameter
for the deprotonation rate in the film,‘5 molybdates would assist in the formation of
ammonium. This is supported by Fig. 5(b), where the ammonium content of the
duplex alloy proves to be more superficial than for the austenite. This effect is caused
by the virtually nitrogen-free ferrite bulk regions that decrease the average content
of nitrogen deeper into the film, but not on the outermost surface where a fast
redistribution of mobile ions is possible.
The Auger sputter depth profiles and the XPS measurements show similar
amounts of nitrogen on top of the passive layer, on the single phase alloys as well as
the respective phases in the duplex steel. From Auger measurements it is impossible
to determine whether this signal originates from nitrogen adsorbed during the
sample transfer process, or from ammonium, as suggested by, e.g. Clayton et ~1.~~
This is possible, however, if XPS is used. In the ARXPS plots, shown in Fig. 5(b), the
ammonium ion fraction is higher in the outer passive layer. The Auger sputter depth
profiles in Fig. 3 show marked enrichments of nitrogen and nickel to the oxide/metal
interface, but with different magnitudes for the different phases. In corresponding
measurements of the single phase ferrite, the nitrogen signal dropped below the
noise level within 1 nm. This could be explained either by a higher nitrogen content in
the duplex ferrite phase, or by diffusion in the oxide/metal interface. XPS data in Fig.
5(b) give an angular dependence for the atomic nitrogen, indicating that it should be
present at the oxide/metal interface. For the single phase ferrite, it was not possible
to identify any atomic nitrogen in the XPS spectra. The presence of ammonium ions
on the surface of the single phase austenite and the duplex steel thus indicates that
nitrogen, in the form of ammonium ions, acts according to the mechanisms suggested
in reactions (4) and (5), i.e. as a mobile repassivator, influencing the passivity on the
austenite as well as the ferrite phase sites. The correlation between a high fraction of
Mo(V1) and the higher nitrogen bulk content of the single phase austenite, seen in
Fig. 5(a), is also striking. This could be taken as an indication that nitrogen, in the
form of ammonium ions, buffers the solution where the stability of molybdates is
higher, a mechanism previously suggested by Lu et a1.3’
When the passive film on the austenite and ferrite phases of the duplex steel is
compared to the passive layer formed on single phase alloys under similar experi-
mental conditions, it is seen that the single phase passive film is clearly thicker. The
increase in thickness of the passive film at polarisations closer to the breakdown
potential, as seen for example in Fig. 2(a), suggests that the single phase alloys are
less resistant to oxidation than the duplex material. The composition of the oxide
film on the respective phases reflects the composition of the underlying alloy to a
certain degree; the chromium content in the duplex stainless steel is slightly higher on
top of the ferrite phase whereas the nickel content is higher in the austenite, but also
enriched in the ferrite, as indicated in Figs 2(b)-(d).
The differences in the passive layers on the duplex steel are small with respect to
the major alloy elements. Larger differences are found for molybdenum and
478 Claes-Olof A. Olsson

nitrogen in the XPS measurements. Even though there is an enrichment of nitrogen


to the oxide/metal interface on the ferrite phase as well, the ferrite phase level of
nitrogen is so low that nitride formation could not possibly play a significant role in
the stabilisation of the passive film. However, the marked enrichment of austenite
stabilisers to the oxide/metal interface, cf. Figs 1 and 3, suggest that the metal nearest
the oxide/metal film could have a laterally continuous austenite-like arrangement;
the concentration levels are well above the bulk level for austenite stability. A
laterally continuous austenite-like phase could be one factor for the formation of a
stable passive film on a duplex stainless stee1.32
There are several independent observations indicating that the surface reactions
of nitrogen and molybdenum are of the utmost importance for the corrosion
resistance of the duplex steel. This is further supported by the electrochemical
characteristics of the single phase alloys that indicate a lower corrosion resistance
when the possibility of nitrogen/molybdenum interaction was eliminated. This type
of interaction is also in good agreement with general observations, such as ‘Duplex
Stainless Steels . . . , but with respect to their resistance to pitting, crevice corrosion
and stress corrosion cracking that, in part, will exceed the resistance of austenitic
stainless steels in chloride containing media. ’ ,X3where the corrosion resistance is said
to depend on the electrolyte, as would be expected if the passivity is strengthened by
reactions at the oxide/liquid interface.

CONCLUSIONS

(1) On duplex stainless steels, there is an enrichment of nitrogen and nickel to the
oxide/metal interface on top of the austenite as well as the ferrite phase regions.
(2) The molybdenum and nitrogen interaction in and near the passive film is
important for the passivity of austenitic stainless steels in general and is particularly
relevant for duplex stainless steels.
(3) Indications are given that the molybdenum and nitrogen interaction is taking
place in the near passive layer region. The highly mobile ions move around and form
an outer region of the passive film that is independent of the underlying phase
(having either an austenitic or a ferritic structure) and enhances the protective
properties of the passive film according to mechanisms described previously for
austenitic stainless steels.

Acknowledgemenfs-Gunilla Runnsjo, Avcsta Sheffield R & D, and Rachel Pettersson. the Swedish
Institute for Metals Research, are gratefully acknowledged for supplying the duplex and the single phase
steels, respectively, and for valuable remarks. Also, Sven Erik Hornstrom, Materialcentrum. is acknowl-
edged for valuable remarks on the script. 1 am indebted to Bertil Eriksson, Avesta Sheffield for phase
characterisation with EPXMA, and to Ulrik Gclius, Department of Physics, Uppsala University. for
assistance with the XPS measurements. The financial support from the Avesta Sheffield Research
foundation is gratefully acknowledged.

REFERENCES
1. Y. C. Lu and C. R. Clayton, Corros. Sci. 29,927 (1989).
2. 1. Olefjord, B. Brox and U. Jelvestam, .l. electrochem. Sot. 132.2854 (1985).
3. S. Jin and A. Atrens, Appl. Phys. A 46, 51 (1988).
4. Y. C. Lu, R. Bandy, C. R. Clayton and R. C. Newman, J. electrochem. Sot. 130, 1774 (1983)
Effects of N and MO on the passivity of SS 2205 479

5. K. Osozawa and N. Okato, Passivity and its Breakdown on Iron and Iron Based Alloys, U.S.A.-
Japan Seminar. Honolulu), p. 135. NACE, Houston, TX (1976).
6. Y. C. Lu, M. B. Ives and C. R. Clayton, Corros. Sci. 35,89 (1993).
7. R. Goetz and D. Landolt, Electrochim. Acta 29,667 (1984).
8. G. Runnsjo, Ph.D. Thesis, Almqvist & Wiksell International, Stockholm (1981).
9. R. Qvarfort, Corros. Sci. 28, 135 (1988).
10. J. F. Moulder, William F. Stickle, P. E. Sobol, K. D. Bomden and J. Chastain (ed.), Handbook of
X-ray Photoelectron Spectroscopy, 2nd Ed. Physical Electronics Industries Inc., Minnesota (1992).
11. J. H. Scofield, J. Electron Spectrosc. Relat. Phenom. 8, 129 (1976).
12. M. P. Seah, in Practical Surface Analysis (eds D. Briggs and M. P. Seah), 2nd Ed., Vol. 1, Wiley.
Chichester, U.K. (1990).
13. I. Olefjord and L. Wegrelius, Corros. Sci. 31, 89 (1990).
14. H. J. Mathieu and D. Landolt, Appl. Surf. Sci. 3, 348 (1979).
15. C. R. Clayton and Y. C. Lu, J. electrochem. Sot. 133.2465 (1986).
16. B. Brox and I. Olefjord in Proc. on Stainless Steel-84, p. 134. The Institute of Metals, London (1985).
17. J.-O. Nilsson, Mater. Sci. Technol. 8,685 (1992).
18. A. Schneider, S. Hofmann and R. Kirchheim, Werkst. Korros. 42, 169 (1991).
19. S. Jin and A. Atrens, Appt. Phys. A 46, 51 (1988).
20. P. Combrad and J.-P. Audouard, in Proc. Conf. Duplex Stainless Steels 1991, p. 257, les Editions de
Physique, Les Ulis France (1991).
21. M. Sakashita and N. Sato, in Passivity of Metals (eds R. P. Frankenthal and J. Kruger), p. 479. The
Electrochemical Society, Corrosion Monograph Series, Princeton, NJ, U.S.A. (1978).
22. Y. C. Lu, C. R. Clayton and A. R. Brooks, Corros. Sci. 29, 863 (1989).
23. R. C. Newman, Y. C. Lu, R. Bandy and C. R. Clayton, in Proc. of the 9th Int. Congress on Metallic
Corrosion, Vol. 4, p. 394. National Research Council, Ottawa, Canada (1984).
24. R. D. Willenbruch, C. R. Clayton, M. Oversluizen, D. Kim and Y. Lu, Corros. Sci. 31,179 (1990).
25. Y. C. Lu, J. L. Luo and M. B. Ives, Corrosion 47,835 (1991).
26. C. R. Clayton and K. G. Martin, in Proc. High Nitrogen Steels-HNS ‘88, p. 256. The Institute of
Metals. London (1988).
27. M. Pourbaix and N. De Zoubov, in Atlas of Electrochemical Equilibria in Aqueous Solutions (ed. M.
Pourbaix), p. 493. Pergamon Press, Oxford, U.K.
28. R. F. A. Jargelius and T. Wallin, in Proc. 10th Scandinavian Corrosion Congress, pp. 161-164.
Stockholm, (1986).
29. R. C. Newman and T. Shahrabi, Corros. Sci. 27,827 (1987).
30. Y. C. Lu, J. L. Luo and M. B. Ives, Corrosion 47,835 (1991).
31. J. L. Luo, Y. C. Lu and M. B. Ives, Proc. Conf. Corrosion ‘92, Paper No. 233. NACE, Houston,
Texas, U.S.A. (1992).
32. C.-O. A. Olsson and S. E. Hornstrom, Proc. Conf. DuplexStainlessSteels ‘94. The Welding Institute,
Cambridge, in press.
33. H. Kaesche. in Die Korrosion der Metalle, p. 225,3rd Ed. Springer-Verlag, Berlin (1990).

You might also like