Nuclear Instruments and Methods in Physics Research B68 (1992) 398-401
North-Holland
Beam Interactions
withMaterials$Atoms
Oxidation and corrosion studies of Al-implanted stainless steel
AISI 321 using nuclear reaction and electrochemical techniques
F. Noii ', P. Misoelides ', P. Spathis ', M. Piiakouta z and H. Baumann
3
1 Department of Chemistry, Aristotle University,
GR-54006 Thessaloniki, Greece
z
Tandem Accelerator Laboratory, NCSR "Demokritos", GR-15310 Aghia Parasketd - Attiki, Greece
s Nuclear
Physics Institute, J. W. Goethe University, D-6000 Frankfurt a.M. 90, Germany
The oxidation of Al-implanted (implantation energy 40 keV, dose 10 6-10" AI ions/cmZ) AISI 321 stainless steel samples in
air has been studied at temperatures between 450 and 650°C using the "O(d,p) 170 nuclear reaction. The determination of the
distribution of! the implanted Al atoms has been performed using the resonance at 992 keV of the Z7AI(p,1')28 Si nuclear reaction .
The determined oxygen profiles indicate that the implantation of 5 x 10 6 and 10" AI ions/cmZ leads to an improvement of the
oxidation resistance of the studied steel samples. The passivation/corrosion behaviour of the AI-implanted steel samples in 0.5M
aqueous sulphuric
The acid solution has also been investigated electrochemically using potentiodynamic and cyclovoltammetric
techniques .
passivation potential values and the repassivation moving to more positive values indicate an improvement of the
corrosion resistance of the AI-implanted steel samples.
1. Introduction
Ion implantation has been proven to be successful
in modifying the properties of near-surface layers of
materials without affecting the bulk . It is well known
that ion implantation improves the thermal oxidation
resistance of metals and alloys . This is due to both
chemical and radiation damage effects [1]. The selection of the suitable thermal oxidation inhibitor is primarily based on its affinity to oxygen in comparison
with that of the matrix atoms and its solubility in the
metallic matrix.
During this work the thermal oxidation of Al-implanted (implantation energy 40 keV, dose between
10'6 and 11)" Al ions/cm') austenitic sta:nless steel
AISI 321 (he/Cr18/Ni8/Ti) samples has been studied in the temperature region between 450 and 650°C.
Aluminum was chosen because of its high affinity to
oxygen, as well as its ability to form double oxides and
its solubility in iron.
The behaviour of the Al-implanted steel samples
was also electrochemically investigated in acid solution
(0 .5M H2SO4) although the implantation of ions does
not always improve the corrosion resistance of metallic
surfaces (see e.g. refs . [2,3]).
2. Experimental
Stainless steel AISI 321 (Fe/Cr18/Ni8/Ti) samples (thickness 0.25 mm, supplied by Goodfellow) :;ere
0168-583X/92/$0.'..00 © 1992 - Elsevier
Science Publishers
B.V.
implanted at room temperature with 40 keV Al ions
(implantation dose between 10'6 and 10"AI ions/cmZ)
using the 50 kV implanter of the Nuclear Physics
Institute of the University of Frankfurt, Germany. The
vacuum in the implantation chamber was better than
10 -6 mbar . A liquid nitrogen trap was surrounding the
target arrangement during the implantation. The AI +
ion beam intensity was about 10 RA/cmZ.
The distribution of the implanted Al atoms on the
sample surfaces has been determined by means of the
resonance at 992 keV of the 27AI(p,y)28 Si nuclear reaction at the 2.5 MV van de Graaf accelerator of the
same institute .
Implantations with 2 x 10" Al ions/cmZ (EA, = 40
keV) at 10 -5 mbar oxygen partial pressure lead to the
formation of gold-brown coloured layers with up to 50
at.% Al concentrations.
The implanted samples were oxidized in air at temperatures between 450 and 650°C. The duration of the
oxidation varied from one to six days . The oxygen
depth distribution on the oxidized samples was determined using the "O(d,p)"O nuclear reaction at the 5
MV Tandem Accelerator Laboratory of the National
Centre for Scientific Research "Demokritos", Athens,
Greece.
Electrochemical investigation of the implanted steel
samples has been performed in aqueous solution of
0.5M H 2 SO4 by means of potentiodynamic and cyclovoltammetric techniques using rapid (50 V/h) and slow
(0.6 V/h) scan rates. The slow scan rates allow measurements under nonconstant conditions of the metalAll rights reserved
F. Noli et al. / Oxidation and corrosion ofAl-implanted stainless steel
lic surface (coating formation) and the corroding environment. In this way predictions of the general corrosion behaviour of the material can be made. The rapid
scan rates allow measurements under constant conditions (without or with very thin coating formation) of
the metallic surface and the corroding medium.
The technique of studying the corrosion behaviour
of a metal-environment system by potentiodynamic
polarization curves can be used for a direct calculation
of the corrosion rate as well as for indicating the
conditions of activity, inactivity, passivity and tendency
of the metal to suffer local, pitting and crevice attack.
An indication of the susceptibility to initiation of localized corrosion is given by the potential at which the
anodic current increases rapidly (pitting potential).
The determination of the corrosion rate and potential regions with intense anodic activation is possible .
The corrosion rate is found by intersecting the extrapolation of the linear portion of the second cathodic
curve with the equilibrium stable corrosion potential .
The intersection corrosion current is converted to a
corrosion rate (penetration : mil per year - mpy) by the
use of a conversion factor (based upon Faraday's law,
the electrochemical equivalent of the metal, its valence
and gram-atomic weight) . A conversion factor of 0.5 is
used in most cases for the determination. The accuracy
of the corrosion rate calculation is dependent upon the
degree of linearity of the second cathodic curve: the
longer the better. This technique of calculating the
corrosion rate does not indicate local attack, only an
average uniform corrosion rate [14,15] .
3. Results and discussion
The theoretical distribution of the implanted Al
ions in the AISI-321 steel matrix, calculated using the
computer code TRIM90 [4] in comparison with the
experimental results obtained by means of the
2`Al(p,y) 2sSi nuclear reaction is given in fig . 1 . The
experimental aluminum profile, before the thermal
treatment of the samples, is characterized by a lower
mean ion range (ca 160 A in comparison to the calculated value of 281 Â) and shows a diffusion tail form .
The differences between the experimental and the
calculated distributions are most probably due to sputtering effects . The aluminum profile of the thermally
treated samples was also determined using the same
technique. In both cases a 60° angle of the target
surface to the proton beam direction facilitated the
measurement of the rather narrow Al distribution,
improving the depth resolution of the technique . The
Al distribution in the implanted AISI-321 steel samples
(dose ; 10 17 ions/cm 2 ) after thermal oxidation for one,
two and four days at 650°C is given in fig . 2 .
399
70 2.50
E 2.00
U
N
1 .50
a
81
w
AI - IONS RANGE
IN AISI-321 STEEL
É 1 00
U
ô 0.50
â
0.00
0.00
40000
1300.00
Target Depth
1200.00
1600 .00
Fig. 1 . Experimental (*) and calculated (o) range distribution
of 40 keV Al ions in AISI-321 stainless steel .
The oxidation of the studied Al-implanted steel
samples in the temperature range between 450 and
650°C shows a parabolic behaviour (Am 2/S 2 =k p t,
where Om/S is the mass increase per surface unit, kp
the parabolic rate constant and t the time). Fig. 3 gives
the concentration of oxygen on Al-implanted steel
samples at 450°C. The corresponding parabolic rate
constants are 4.4 x 10 -6, 9 .0 x 10 - ' and 2 .1 x 10-5
((mg/cm2 ) 2 per day), for the samples implanted by
10 17, 5 x 1016 and 10 16 Al ions/cm2 , respectively. The
value for the nonimplanted steel samples at 450°C is
2.3 x 10 -5 ((mg/cm2 )2 per day). This is an indication
that the oxidation resistance improvement by means of
implantation of 10 16 Al ions/cm2 at 450°C is rather
limited, whereas the implantation with higher dose
leads to a more significant reduction of the oxidation
rates . Fig . 4 shows the oxygen concentration on 10 17 AI
ions/cm2 implanted stainless steel samples at 450, 550
and 650°C. There is a considerable difference in the
oxidation behaviour at the 450 and 550°C on the one
hand and the 650°C on the other hand, indicating the
presence of different oxidation mechanisms. A preliminary evaluation of the Arrhenius plot of the parabolic
rate constant k p in the case of the implantation with a
dose of 10 17 Al ions/cm2 gave an activation energy
value of Ea = 26 kJ mol -1 for the temperature region
below 550°C. The corresponding activation energy value
for higher temperatures is about 7 times higher, also
indicating the different oxidation mechanisms concerned. These results are in good agreement with data
obtained during oxidation studies using AI-implanted
pure iron [5] . It seems more likely that the low temperature (below 600°C) oxidation mainly proceeds, as in
the case of pure iron, by fast diffusion paths transport,
Vlll. ION BEAM MODIFICATION
400
F. Noli et al. / Oxidation and corrosion of Al-implanted stainless steel
30 .00
25 .00
0.00 11. ...
-15.00
. . ... . . . ...
35 .00
85 .00
135.00
185.00
Depth ( nm )
235.00
Fig. 2. Experimental depth distribution of 40 keV AI ions (dose: 10 17 cm -2) on AISI-321 steel samples after oxidation in air at a
temperature of 650°Cfor one (o), two(+) andfour ( ")days.
which is known to be less activated than the bulk
diffusion taking place at higher temperatures. Aluminum being soluble in a-Fe is equally distributed
between bulk and grain boundaries of iron oxides,
slightly affecting both fast diffusion paths transport
and bulk diffusion. The improvement of the oxidation
resistance of the AISI-321 by implantation with a dose
higher than 5 x 10'6 Al ions/cm 2 is of the order of a
factor of 2, whereas the implantation of less soluble
ions segregating to dislocations and grain boundaries
could lead to better blocking effects and more drastic
reduction of the low temperature oxidation rates .
Oxidation time (days)
Fig. 3. Oxygen concentration on nonimplanted (x) and AI
implanted AISI-321 steel samples treated at 450°C. (+ : 1017,
o: 5 x 10 16 , and + : 10 16 Al ions/cm2).
There is an increasing number of publications dealing with the improvement of the corrosion resistance of
ion-implanted metallic surfaces, especially iron and
steel surfaces, in acid solutions (see e.g . refs . [2,3,6,12]) .
In most of the cases an improvement of the corrosion
resistance of the implanted surfaces is observed, although it is well known that ion implantation is unsuitable for long-term corrosion protection, because of the
rather limited width of the modified surface layer
produced by low energy ions (implantation energies up
to a few hundred keV). The partial or total amorphiza-
Fig. 4. Oxygen concentration of AI-implanted (dose: 10 17
ions/cm2) steel samples treated at 450
550 (o) and 650°C
(+).
F. Noli et al. / Oxidation and corrosion of Al-implanted stainless steel
tion of the surface of the implanted substrates should,
on the other hand, be thermodynamically more reactive in corrosive environments . However, the lack of
defects, grain boundaries, etc., usually found in crystalline materials, leads to an improvement of the resistance of the amorphous state to pitting corrosion [13] .
Most of the corrosion studies in solutions are using
electrochemical techniques (potentiodynamic, potentiostatic and cyclovoltammetric methods), although
these measurements cannot substitute long-term
open-circuit immersion tests and can be, in most of the
cases, interpreted only qualitatively .
Potentiodynamic measurements indicated a slight
increase of the pitting potential (Ep ) of the Al-implanted samples compared to the nonimplanted ones.
The EP values varied from 1300 for the nonimplanted
to 1400 mV for the 10 1 ' Al ions/cm Z implanted samples. The EP values are characteristic of the potential
region, where an intense anodic activation and pitting
corrosion begins. The corresponding repassivation (E,)
and corrosion (E,o,) potentials varied from +900 to
+950 mV and from -370 to +130 mV, for the above
mentioned samples, respectively. The corrosion rate
was found to be 27.8 and 20 .2 x 10 -5 mg s -1 cm -2 for
the nonimplanted and the 1017 Al ions/cm Z implanted
AISI 321 steel samples in 0 .5M H2SO4, respectively .
The samples with lower Al dose showed similar behaviour with values between the above mentioned. All
curves show a hysteresis loop during the reverse scan,
indicating sensitivity to pitting corrosion . There are
differences among the subsequent scans showing the
gradual dissolution of the implanted layer.
4. Conclusions
There is an improvement of the oxidation resistance
of almost a factor of 2 of the AISI 321 steel samples
implanted with 5 x 10" and 10 17 Al ions/cmZ in comparison with the nonimplanted material . The oxidation
of the above mentioned Al-implanted samples in air at
temperatures below 550°C seems to proceed by a different mechanism than at 650°C. In the former temperature region, the oxidation of the samples, as in the
case of AI-implanted pure iron, most probably proceeds by fast diffusion paths transport, whereas at
higher temperatures it proceeds by bulk diffusion .
401
The corrosion resistance of the AISI 321 steel in
0.5M H,SO4 solutions seems to be improved by Al
implantation, as indicated by potentiodynatnic m.,.:
surements of the pitting, repassivation and corrosion
potentials . On the other hand, the 40 keV AI implantation does not seem to provide, because of the limited
width of the layer, any modified material surface suitable for any long-term practical application.
5. Acknowledgements
The partial support of this work by the General
Secretariate of Research and Technology, Athens,
Greece, and the Federal Ministry of Research and
Technology, Bonn, Germany, is thankfully acknowledged.
References
[1] M . Pons, A . Galerie and M . Caillet, Defect and Diffusion
Forum 57-58 (1988) 189.
[2] G. Wolf and H . Ferber, Nucl. Instr. and Meth. 209/210
(1983) 197 .
[3] H.J. Kim, W .B . Carter, R .F . Hochman and E.I . Meletis,
Mater . Sci. Eng. 69 (1985) 297.
[4] J .F. Ziegler and J .P. Biersack, Computer code TRIM90,
based on : The Stopping and Range of Ions in Solids
(Pergamon, New York, 1985) .
[5] M . Pons, M. Caillet and A . Galerie, Nucl . Instr. and
Meth. 209/210 (1983) 1011 .
[6] S .T . Picraux, Ann. Rev. Mater. Sci. 14 (1984) 335.
[7] G.K. Hubler, I.L. Singer and C.R . Clayton, Mater. Sci.
Eng. 69 (1985) 203 .
[8] M . Iwaki, T. Fujihana and K . Okitaka, Mater. Sci . Eng .
69 (1985) 211 .
[9] P.'_. Bonora, Mater. Sci . Eng. 69 (1985) 253 .
[10] H. Ferber, G.K. Wolf, H. Schmiedel and G. Dearnaley,
Mater . Sci . Eng. 69 (1985) 261 .
[11] H. Ferber and G.K. Wolf, Mater. Sci. Eng. 90 (1987) 213 .
[12] Y. Okabe, M. Iwaki and K. Takahashi, Mater. Sci . Eng.
90 (1987) 223.
[13] M. Naka, K . Hashimoto and T. Masumoto, Corrosion 32
(1976) 146.
[14] R . Baboian and G . Hayner, Electrochemical Corrosion
Testing, ASTM STP 727 (1981) pp. 274-282.
[15] O . Siebert, Laboratory Corrosion Tests and Standards,
ASTM STP 866 (1985) pp. 65-90.
VIII. ION BEAM MODIFICATION