Electrochemical pitting behaviour of type 321 stainless steel in sulde-containing

chloride solutions
Y.M. LIOU, S.Y. CHIU, C.L. LEE and H.C. SHIH*
Department of Material Science and Engineering, National Tsing Hua University, Hsinchu, Taiwan 300,
Republic of China
(*author for correspondence, e-mail: [email protected])
Received 16 December 1998; accepted in revised form 27 April 1999
Key words: cyclic potentiodynamic polarization, pitting, sodium chloride, sodium thiosulfate, type 321
stainless steel
Abstract
The pitting behaviour of type 321 stainless steel in sulde-containing chloride aqueous environments was studied
using cyclic potentiodynamic polarization. A well-established correlation between H
2
S and Na
2
S
2
O
3
in the study of
corrosion was applied, that is, H
2
S was simulated by Na
2
S
2
O
3
. The major factors aecting the pitting corrosion of
type 321 stainless steel are the Cl

concentration, solution pH and temperature. The results clearly indicate that

both i
pit
and i
pp
decrease with increasing Cl

concentration and temperature, while s

pass
is more sensitive to
temperature variation. i
pit
decreased with decreasing pH in the range 2 ` pH ` 7X5. The surface morphology and
chemistry of the corroded type 321 stainless steel resulting from anodic polarization in 0.01 M S
2
O
2
3
-containing Cl

solution were analysed by XRD, SEM and EPMA. A higher concentration of sulfur was found in the pits, and the
dark surface lm was mainly composed of FeS and c-Fe
2
O
3
. The results describe the pitting behaviour of type 321
stainless steel in sulde-containing Cl

aqueous environments.
1. Introduction
Corrosion has always been a problem in petroleum
rening and petrochemical operations. Uniform cor-
rosion of equipment can be readily detected by
various inspection techniques. In contrast, isolated
pitting is potentially much more serious because
leakage can occur at highly localized areas that are
dicult to detect [13]. Stainless steels (SS) are
extensively used in petrochemical reneries, because
of the highly corrosive nature of the catalysts and
solvents that are often used [3]. Most stainless steels will
pit in the presence of chlorides. Hydrogen sulde is
the main constituent of renery sour waters and is also
formed by the decomposition of organic sulfur com-
pounds that are present at elevated temperatures [46].
The type 321 SS used in this study was a titanium-
stabilized austenitic stainless steel. Sensitization result-
ing from the precipitation of chromium carbide along
the grain boundaries is avoided by alloying titanium
with the native carbon in the form of TiC precipitates
at much higher temperatures. The use of H
2
S in this
study was replaced by Na
2
S
2
O
3
, the latter being more
convenient and less hazardous to the environment. A
series of simple immersion tests in deaerated 20 wt %
NaCl solution at 80 C indicated a good correlation
between the critical pitting concentration of S
2
O
2
3
and the critical H
2
S pressure for pitting [7], leading to
a standard simulated Cl

+ H
2
S environment being
proposed [8, 9] consisting of deaerated 20 wt % NaCl +
10
3
10
2
M S
2
O
2
3
aqueous solution of pH 4 at
80 C. However, in the present work a more acidied
environment of pH 2 was also adopted in the polar-
ization tests of type 321 SS.
By means of the cyclic potentiodynamic polarization
[10], the eects of Cl

concentration, temperature, and

solution pH on the pitting corrosion behaviour of type
321 SS in sulde-containing Cl

environments were
investigated.
2. Experimental details
2.1. Materials
Coupons of type 321 SS were cut from commercial
stock, fully annealed, abraded with a series of SiC
papers to 600 grit, followed by ultrasonic cleaning
in acetone and nally rinsing in deionized water prior
to the immersion tests. The specimen area exposed
to the test solution was 1 cm
2
and the remaining area
was shielded by epoxy resin. The chemical compo-
sitions (wt %) analysed by atomic emission spec-
troscopy were: 0.069 C, 1.30 Mn, 0.023 P, 0.006 S,
Journal of Applied Electrochemistry 29: 13771381, 1999.
1377
1999 Kluwer Academic Publishers. Printed in the Netherlands.
0.60 Si, 10.96 Ni, 17.89 Cr, 0.50 Ti, 0.86 Mo and
balanced Fe.
2.2. Cyclic potentiodynamic polarization
The polarization apparatus was similar to the stan-
dard assembly as described elsewhere [1113]. A
standard calomel electrode (SCE) and a graphite
electrode were used as reference and counter electrode,
respectively and an EG&G model 273 potentiostat/
galvanostat for the polarization of the specimen.
Figure 1 shows a schematic of the cyclic polarization
curve for a metal that pits. A scan rate of 1 mV s
1
was found convenient and suciently slow to avoid
any loss of conguration of the potentiodynamic
polarization curves. All specimens were initially po-
larized from their natural corrosion potential, desig-
nated as i
corr
and subsequently controlled cathodically
at 1X2 V vs SCE, a potential which is signicantly
below i
corr
, for about 10 min to reduce the possible
existing surface lms, then reversed to where the
polarization started. A new potential i
c
was found to
be more active than i
corr
. The second reverse occurred
at potentials signicantly more noble than i
corr
or at
the current density b10
3
A cm
2
. The potential was
then swept to more active values and the cyclic
polarization was thus completed.
2.3. Test solution
The test solutions were prepared with deionized water
and reagent-grade NaCl, and Na
2
S
2
O
3
X5H
2
O. The
solution pH was adjusted by either adding HCl or
NaOH. Deaeration of the test solution was achieved by
purging the solution with deoxygenated nitrogen for 1 h
prior to the test and continuing the purging during
the test.
3. Results and discussion
3.1. Eect of Cl

concentration
The eects of Cl

concentration on the potentiodynamic

polarization curve of type 321 SS in deaerated 0.01 M
S
2
O
2
3
-containing Cl

aqueous solution (pH 7) at 25 C

are shown in Figure 2(a) and (b). The curves indicate
that: (i) increasing Cl

concentration shifts both the

pitting potential (i
pit
) and the protection potential (i
pp
)
in the active direction; Cl

exerts a noticeable inuence

on the pitting susceptibility; (ii) linear relationships
between i
pit
and i
pp
as functions of Cl

concentration
were observed [1416], these are: i
pit
(v)=0X1100X153
log [Cl

wt %] and i
pp
(v) = 0X2520X136 log [Cl

wt %], respectively, indicating that i

pit
and i
pp
are
highly susceptible to Cl

concentration, while the

difference between i
pit
and i
pp
, that is, Di remains
Fig. 1. Schematic of cyclic polarization curve for a metal that pits.
Fig. 2. Eect of Cl

concentration on the polarization behaviour of

type 321 SS in deaerated 0.01 M S
2
O
2
3
-containing Cl

aqueous
solutions at pH 7 and 25 C, (a) anodic polarization curves for
Cl

(wt %) of 0.1, 1, 10, and 20, and (b) variation of E

pit
, E
pp
, and DE
with the Cl

concentration.
1378
constant. The magnitude of Di was often considered as
an indicator for pitting resistance; that is, the smaller the
Di, the higher the pitting resistance [14, 17]. However,
the difference between i
corr
and i
pp
has been considered
as a critical indication of pitting resistance in recent
years. i
corr
changes only slightly with Cl

concentration
within 0.3 V (Figure 2(a)) and lies much below the i
pp
by 0.73 to 0.45 V as Cl

concentration increases from

0.1 wt % to 20 wt % where type 321 SS is entirely free
from pitting corrosion.
3.2. Eect of temperature
Temperature increases the rate of almost all chemical
reactions. It is vital in this particular environment
containing S
2
O
2
3
. Figure 3 shows the potentiodynamic
polarization results of type 321 SS in a deaerated 0.01 M
S
2
O
2
3
-containing Cl

solution (pH 7) at temperatures

3092 C. It is clear that both i
pit
and i
pp
decrease with
increasing temperature. It is worth noting that i
corr
shifts slightly in the noble direction with increasing
temperature, which suggests that depolarization of the
cathodic reaction takes place, resulting in a higher
corrosion rate while the polarization of the anodic
reaction remains constant. The passive current density
s
pass
increases from 2 10
5
to 10 10
5
A cm
2
as the
temperature is varied from 30 to 92 C (Figure 3(a) and
(b)). Increasing temperature apparently promotes the
ionic activities of Cl

and S
2
O
2
3
and a possible
synergistic effect causes breakdown of the passive lm
and reduces pitting resistance.
3.3. Eect of solution pH-corrosion behaviour diagram
Various characteristic potentials, such as i
corr
, i
c
, i
pp
and i
pit
, derived from the cyclic potentiodynamic
polarizations for type 321 SS in deaerated 0.01 M
S
2
O
2
3
-containing Cl

aqueous solutions expressed as a

function of solution pH, also known as the corrosion
behaviour diagram, was constructed, as shown in
Figure 4. It is apparent that i
pit
decreases with decreas-
ing solution pH, thereby reducing the resistance for
pitting, while Di (i
pit
i
pp
) which covers the region of
imperfect passivation increases linearly with pH in the
range pH 2 to 7.5. The extent of the perfect passivation
region starts to reduce more signicantly in solution of
pH ` 4. The pH 1.9 line divides the diagram into areas
of general corrosion and passivation. This boundary
was carefully determined by polarizing the specimen
anodically using small pH increments, and noting a
transition from the active state of general corrosion to
the passive state [10, 18, 19]. The type 321 SS used in
this study remains passive to pH values as low as 1.9,
compared with 8.2 for mild steel in a similar environ-
ment [10]. The pH-varying i
corr
lies along the boundary
of i
pp
and intersect it at pH 4, or more quantitatively,
i
corr
b i
pp
at pHO4 where type 321 SS suffers from
pitting corrosion; on the other hand, i
corr
` i
pp
at
pHP4 where type 321 SS is entirely immune from
any type of corrosion (Figure 4). i
corr
species the
potential of a metal that exhibits passivity in most air-
saturated solutions. For instance, chromium, although
Fig. 3. Eect of temperature on the anodic polarizations of type 321
SS in 0.01 M S
2
O
2
3
-containing 20 wt % Cl

aqueous solution at pH 7
and 25 C, (a) polarization curves taken at different temperatures, and
(b) variation of E
pit
, E
pp
, E
corr
, DE and I
pass
with the temperature.
Fig. 4. Corrosion behaviour diagram for type 321 SS in deaerated
0.01 M S
2
O
2
3
-containing 20 wt 7 Cl

aqueous environment with

varying solution pH values at 25 C.
1379
near zinc (i

= 0X763 V) in the electromotive series,

behaves more like silver (i

= 0X799 V) in many air-

saturated aqueous solutions because of its passive
surface lm. Chromium and stainless steels can even
be passivated in air. Pitting was observed at
1X9 ` pHO4, while no pitting occurred for pH b 4.
This observation is summarized in the corrosion behav-
iour diagram of Figure 4 as derived from anodic
polarization curves. For pH ~ 2 where curves of i
corr
and i
pit
meet, pits developed almost instantaneously,
whereas for pH ` 1X9, pitting corrosion could be con-
sidered as totally masked by general corrosion. The
region below i
pass
also suffers from general corrosion
and is not a major concern, since i
corr
is more noble
than these potentials, especially for pH b 4 where
passivity remains intact. The passive current density,
s
pass
increases slightly with decreasing pH in the pH
range 7.5 to 2, showing that the protective ability of the
passive oxide lm becomes weaker and less corrosion-
resistant in acid solutions.
3.4. Pit morphology and corrosion product
Either colloid formation or lower metal overvoltage
resulting from the chemisorption of Cl

ions under
anodic polarization may cause local breakdown of the
passive lm. The appearance of pits, in which sulfur is
deposited in a solution pH 2, is shown in Figure 5(a)
and (b), while the alloying element Ti (0.5%) for the
stabilizing does not show any localized enrichment, as
shown in Figure 5(c). However, when uniform corrosion
takes place for pH ` 1X9, only a dark layer of FeS,
along with c-Fe
2
O
3
, was detected by XRD as shown in
Figure 6. Both S and H
2
S originate from S
2
O
2
3
as a
result of the following chemical and electrochemical
decompositions [1, 7, 20, 21].
Formation of S:
S
2
O
2
3
H

S HSO

3
(1)
S
2
O
2
3
6 H

4 e

2 S 3 H
2
O (2)
Formation of H
2
S:
S 2 H

2 e

H
2
S (3)
4 S 4 H
2
O 3 H
2
S H
2
SO
4
(4)
These reactions can be further enhanced by increasing
temperature and/or acidity of the environments.
The resulting FeS lm is also conrmed by the
potentialpH diagram for FeSH
2
O [22] where a
constant pH line occurs at pH ~ 2 where FeS and
Fe
2
+ H
2
S are in electrochemical equilibrium. For
pH b 2 FeS is a stable phase from 0X15 to 0X65 V vs
SCE, which corresponds to the areas of pitting, imper-
fect passivation, perfect passivation, and the complete
distribution of i
corr
, as observed in the corrosion
behaviour diagram (Figure 4).
4. Conclusions
The following conclusions can be drawn:
(i) Increasing Cl

concentration in the presence of

0.01 M Na
2
S
2
O
3
decreases i
pit
and i
pp
with the
following relations: i
pit
(v) = 0X110 0X153
log [Cl

wt %] and i
pp
(v) = 0X252 0X136
log [Cl

wt %]. Both i
pit
and i
pp
are highly sus-
ceptible to the Cl

concentration, while the Di or

i
pit
i
pp
remains constant at about 180 mV and is
almost independent of Cl

concentration.
Fig. 5. Photomicrograph and element distribution of type 321 SS in
deaerated 0.01 M S
2
O
2
3
-containing 20 wt % Cl

aqueous solution at
pH 2 and 25 C, (a) SEM micrograph, (b) X-ray mapping of sulfur
and (c) X-ray mapping of titanium.
1380
(ii) Both i
pit
and i
pp
decrease with increase in tem-
perature, accounting for the increase in ionic activity
of Cl

and S
2
O
2
3
or, more precisely, their synergism
eects, which promote passivity breakdown. s
pass
increases with increasing temperature.
(iii) i
pit
decreases with increasing acidity of S
2
O
2
3
-con-
taining Cl

solutions and therefore increases the

tendency for pitting. The perfect passivation region
shrinks rapidly as i
pass
rises cathodically (more
noble potentials) in solutions of pH ` 4. The
superimposed i
corr
distribution intersects i
pp
, which
is the boundary between imperfect passivation and
perfect passivation at pH 4. Type 321 SS is ex-
tremely susceptible to pitting corrosion in the range
2 ` pH ` 4. It is apparently safe to use type 321 SS
in similar environments but of higher pH, that is
pH b 5.
(iv) Sulfur was the only chemical residue observed in the
pits (2 ` pH ` 4), while FeS, along with c-Fe
2
O
3
,
was detected as the major chemical components of
the surface lm resulting from general corrosion
(pH ` 1X9).
Acknowledgements
The authors gratefully acknowledge nancial support
from the Chinese Petroleum Corporation and also from
the National Science Council of the Republic of China
under contract NSC 85-2216-E-007-012.
References
1. R.D. Kane, Int. Metals Rev. 30 (1985) 291.
2. H.I. McHenry, P.T. Purtscher and T.R. Shives, Corros. Sci.
27 (1987) 1041.
3. R.D. Merrick, Mater. Perform. 27 (1988) 30.
4. Z.A. Foroulis, Corros. Prevent. Control 40 (1993) 84.
5. G.E. Moller, I.A. Franson and T.J. Nichol, Mater. Perform., Vol
20 (No. 4), 1981, p 4150.
6. J. Gutzeit, R.D. Merrick and L.R. Scharfstein, `Corrosion in
Petroleum Rening and Petrochemical Operations, Metals Hand-
book', Vol. 13, Corrosion, 9th edn. (ASM International, Metals
Park, OH, 1987), p.1262.
7. NACE Standard TM 01-77, `Laboratory Testing of Metals for
Resistance to Sulde Stress Cracking in H
2
S Environment'
(NACE, Houston, TX, 1990).
8. S. Tsujikawa, A. Miyasaka, M. Ueda, S. Ando, T. Shibata,
T. Haruna, M. Katahira, Y. Yamane, T. Aoki and T. Yamada,
Corrosion 49 (1993) 409.
9. T. Haruna and T. Shibata, Corrosion 50 (1994) 785.
10. H.C. Shih, J.C. Oung, J.T. Hsu, J.Y. Wu and F.I. Wei, Mater.
Chem. Phys. 37 (1994) 230.
11. N.D. Greene, `Experimental Electrode Kinetics' (Rensselaer Poly-
technic Institute, Troy, NY, 1965).
12. G.A. Cragnolino and N. Dridhar, Corrosion 25 (1969) 233.
13. M. Hubbell, C. Price and R. Heidersbach, `Crevice and pitting
corrosion test for stainless steels', in `Laboratory Corrosion Tests
and Standards' (edited by G.S. Haynes and R. Baboian) (ASTM
STP 886, 1985), p. 324.
14. B.E. Wilde, Corrosion 28 (1972) 283.
15. T. Suzuki and Y. Kitamura, Corrosion 28 (1972) 1.
16. D.A. Jones and B.E. Wilde, Corros. Sci. 18 (1978) 631.
17. H.P. Leckie and H.H. Uhlig, J. Electrochem. Soc. 113 (1966) 1262.
18. S.D. Chyou and H.C. Shih, Bull. Electrochem. 3 (1987) 1.
19. S.D. Chyou and H.C. Shih, Corrosion 47 (1991) 31.
20. A.M. Pedraza, I. Villegas, B. Chornik and P.L. Freund,
J. Electroanal. Chem. 250 (1988) 443.
21. G. Schmitt, Corrosion 471 (1991) 288.
22. R.C. Newman, K. Sieradzki and H. S. Isascs, Metall. Trans. A 13
(1982) 2015.
Fig. 6. X-ray diraction pattern of type 321 SS after anodic polariza-
tion in deaerated 0.01 M S
2
O
2
3
-containing 20 wt % Cl

aqueous
solution at pH 2 and 25 C, showing the major corrosion product of
FeS along with c-Fe
2
O
3
.
1381