Tin-manganese alloy electrodeposits II. Corrosion performacne studies

Article abstract of DOI:10.1149/1.2806773

The corrosion performance of tin-manganese alloy electrodeposits either immersed in a quiescent sodium chloride solution or exposed to neutral salt spray (NSS) was studied. The results obtained were then compared with those for conventional pure zinc and zinc-nickel alloy electrodeposits with or without chromate passivation treatment. It was found that tin-manganese alloy electrodeposits provided much longer sacrificial protection against the corrosion of mild steel substrates than zinc-nickel alloy coatings, although they were inferior to pure zinc coatings. The prolonged sacrificial property was attributed to the high polarization resistance of tin-manganese alloy electrodeposits as revealed by linear polarization resistance techniques, which would only allow the corrosion and the consequent ennoblement of these coatings to proceed at a slow rate. The result of NSS tests indicated that tin-manganese alloy electrodeposits, with a time to red rust of about 1200 h, were as good as chromated zinc-nickel alloy coatings but far outperformed pure zinc coatings and nonchromated zinc-nickel coatings. The high corrosion resistance of tin-manganese alloy electrodeposits without any passivation treatment suggests that they may be an environmentally friendly replacement to zinc alloy coatings, which need toxic chromate conversion coatings for corrosion-resistance enhancement.

Full text of DOI:10.1149/1.2806773

Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
C53
0013-4651/2007/155͑2͒/C53/9/$23.00 © The Electrochemical Society
Tin–Manganese Alloy Electrodeposits
II. Corrosion Performance Studies
Keming Chen and Geoffrey D. Wilcoxz
Institute of Polymer Technology and Materials Engineering, Loughborough University, Loughborough,
Leicestershire LE11 3TU, United Kingdom
The corrosion performance of tin–manganese alloy electrodeposits either immersed in a quiescent sodium chloride solution or
exposed to neutral salt spray ͑NSS͒ was studied. The results obtained were then compared with those for conventional pure zinc
and zinc–nickel alloy electrodeposits with or without chromate passivation treatment. It was found that tin–manganese alloy
electrodeposits provided much longer sacrificial protection against the corrosion of mild steel substrates than zinc–nickel alloy
coatings, although they were inferior to pure zinc coatings. The prolonged sacrificial property was attributed to the high polar-
ization resistance of tin–manganese alloy electrodeposits as revealed by linear polarization resistance techniques, which would
only allow the corrosion and the consequent ennoblement of these coatings to proceed at a slow rate. The result of NSS tests
indicated that tin–manganese alloy electrodeposits, with a time to red rust of about 1200 h, were as good as chromated zinc–nickel
alloy coatings but far outperformed pure zinc coatings and nonchromated zinc–nickel coatings. The high corrosion resistance of
tin–manganese alloy electrodeposits without any passivation treatment suggests that they may be an environmentally friendly
replacement to zinc alloy coatings, which need toxic chromate conversion coatings for corrosion-resistance enhancement.
© 2007 The Electrochemical Society. ͓DOI: 10.1149/1.2806773͔ All rights reserved.
Manuscript submitted June 28, 2007; revised manuscript received September 26, 2007. Available electronically December 3, 2007.
Electrodeposited zinc and zinc-based alloys such as zinc–nickel,
zinc–cobalt, and zinc–iron have been widely used for the protection
of steels from corrosion, particularly in aggressive chloride-bearing
environments.1,2 They provide protection to the underlying steel by
means of physical exclusion and sacrificial action. However, their
improved protective ability relies on a postplating chemical passiva-
tion treatment using chromium͑VI͒ compounds to convert the coat-
ing surface from the initial active state to an inert passive oxide
film.2-4 This complex oxide film on the coating surface serves as a
barrier to the environment, while the chromium͑VI͒ contained in the
film leaches out when in contact with water, forming a local chro-
mate solution which repassivates the exposed zinc surface.3 Unfor-
tunately, chromium͑VI͒ is highly toxic and is also categorized as a
possible carcinogen.5 The great concern over the allergic effects for
workers handling chromated components and the toxic effects from
their discharge into the environment has led to the initiation of the
European Union’s “End-of-Life Vehicle Directive,” which prohibits
the use of chromium͑VI͒ in materials and components of vehicles
from July 1, 2007. As a result of the pending legislation, there is an
urgent requirement for the development of new chromium͑VI͒-free
passivation treatments,6-8 while it is also important to develop new
alternative corrosion-resistant electrodeposits which need not rely
on toxic chromate to improve their corrosion resistance.
Tin–manganese alloy electrodeposits have considerable potential
as environmentally friendly sacrificial coatings.9 However, there are
few reports on the corrosion properties of these coatings due to the
great difficulty encountered in their electrodeposition. Recently we
developed a process for the successful electrodeposition of tin–
manganese alloy coatings from an acidic sulfate/gluconate
solution.10 The advantages of this process include adherent and
compact coatings with manganese content up to 40 wt %, relatively
high current efficiency ͑above 50%͒, low deposition current density
͑1–4 A/dm2͒, and adequate bath stabilities. In this paper, the corro-
sion performance of the coatings obtained, in terms of their sacrifi-
cial properties and barrier properties on mild steel substrates, was
characterized using electrochemical techniques including measure-
ments of the open-circuit or corrosion potential ͑Ecorr͒, the galvanic
corrosion current, and the linear polarization resistance. The coat-
ings were also subjected to accelerated corrosion tests in a neutral
salt spray ͑NSS͒ cabinet. For the purpose of comparison, pure zinc
and zinc–nickel alloy electrodeposits with and without chromating
passivation treatments were used as control samples. Finally, the
composition and the types of corrosion products of tin–manganese
electrodeposits after NSS tests were analyzed in order to gain insight
into the corrosion mechanism.
Experimental
For corrosion-performance studies, all the electrodeposits, in-
cluding the control ones, were electroplated to approximately 8 ␮m
on mild steel panels. The electrolyte formulation and electroplating
conditions for the electrodeposition of tin–manganese alloy coatings
have been described in Ref. 10. For the electrodeposition of other
coatings used in this study, the electrolytes and the corresponding
electroplating conditions are given in Table I. The zinc–nickel alloy
electrodeposits obtained were measured to contain 11–14 wt %
nickel. During all electroplating, a platinized titanium mesh was
used as the anode.
After electrodeposition, pure zinc and zinc–nickel alloy coatings
were thoroughly rinsed and immediately passivated in chromate so-
lutions. Pure zinc coatings were immersed in a room-temperature
solution containing 200 g/L sodium dichromate for 30 s to obtain
an iridescent yellow conversion coating.4 The pH of the solution
was adjusted to 1.5 with 20 vol % solution of sulfuric acid ͑SG͒
1.84. The passivation treatment of zinc–nickel alloy coatings was
found to be much more difficult than that of pure zinc coatings.
Only black conversion coatings were obtained from a solution con-
taining 30 g/L CrO3, 10 mL/L H3PO4 ͑SG 1.69͒, 5 mL/L HCl ͑SG
1.18͒, HNO3 ͑SG 1.42͒, and H2SO4 ͑SG 1.84͒ each.4 The coatings
were immersed for 30 s at room temperature. After the chromating
process, all the samples were stored in a desiccator for two days
before any corrosion testing.
The corrosion potential of samples were measured against a satu-
rated calomel reference electrode ͑SCE͒ using a digital voltammeter.
Note that only the coated side of the samples was exposed to the
corrosion medium, which was quiescent 5 wt % sodium chloride
solution with a volume of 400 mL. The surface area of the samples
exposed to the corrosion medium was approximately 20 cm2 formed
by an O-ring ͑⌽5 cm, where ⌽ indicates internal diameter͒.
The sacrificial properties of tin–manganese alloy electrodeposits
were further investigated by measuring the galvanic corrosion cur-
rent ͑Igalv͒ generated when coupled to bare mild steel panels im-
mersed in 500 mL of 5 wt % sodium chloride solution. A purpose-
built corrosion cell was used11 containing two electrodes, one
prepared from a coated steel panel and one from a bare mild steel
panel. The surface area of the two electrodes in contact with the
corrosion medium was equally 20 cm2 formed by a ⌽5 cm O-ring.
The distance between the two electrodes was maintained at 7 cm.
The galvanic current generated was recorded using an ACM Instru-
z E-mail: G.D.Wilcox@lboro.ac.uk
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C54
Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
Table I. Electrolytes and electroplating conditions for the deposition of coatings used in this study.
Sn–Mn10
Zn
Zn–Ni
Sn
Mn
20 g/L SnSO4
70 g/L MnSO4·4H2O
40 g/L Sodium gluconate
70 g/L ͑NH4͒2SO4
30 g/L H3BO3
4–6 mL/L Tween 20
pH 3.5
350 g/L ZnSO4·7H2O
60 g/L ZnSO4·7H2O
130 g/L NiSO4·6H2O
160 g/L Na2SO4·10H2O
pH 1.5–2.0
50 g/L SnSO4
4 mL/L Tween 20
pH 1.0
105 g/L MnSO4·4H2O
140 g/L ͑NH4͒2SO4,
pH 8.0
60 g/L Al2͑SO4͒
3
30 g/L H3BO3
pH 1.5–2.0
20–25°C
50°C
20–25°C
55–60°C
Agitation
No agitation
Slight agitation
No agitation
1 A/dm2
4 A/dm2
5 A/dm2
4 A/dm2
20–25°C
No agitation
1–4 A/dm2
ments GalvoGill 12 zero-resistance ammeter over a period of
2000 h. In addition, the coupled potentials ͑Ecouple͒ were also moni-
tored against a SCE reference electrode. Potential measurements
showed that the coupled potential measured was virtually the same
͑within 1 mV variation͒ and independent of the distance of the ref-
erence electrode to the electrodes, indicating that the resistance of
the external circuit and that of the solution were low and therefore
their effect on the galvanic corrosion current could be neglected.
The barrier properties of coatings were determined using the lin-
ear polarization resistance technique. One purpose-built polarization
cell was employed, with a working electrode ͑i.e., the coating͒ of
approximately 1 cm2 exposed surface area formed by a ⌽11 mm
O-ring, a platinum auxiliary electrode, and a SCE reference elec-
trode. The polarization was performed using an ACM Instruments
computer-controlled potentiostat ͑Auto Tafel HP͒. Prior to each lin-
ear polarization resistance ͑LPR͒ sweep, the test sample was ex-
posed to 5 wt % sodium chloride solution ͑300 mL͒ for 2 h in order
for the working electrode to reach a relatively stable Ecorr. Other-
wise, there would be a large corrosion potential drift on the polar-
ization curve obtained for the tin–manganese alloy coatings. The
potential sweep was carried out 20 mV on either side of Ecorr at a
scan rate of 0.1 mV/s. The same experimental arrangement was also
used to obtain potentiodynamic polarization curves for tin–
manganese alloy coatings over a wide potential range at a sweep
rate of 0.5 mV/s, from which the values of Tafel constants, i.e., ␤a
and ␤c, were derived. The corrosion rate ͑icorr͒ of coatings was then
calculated using the Stern–Geary equation12
Results and Discussion
Sacrificial properties.— To define the sacrificial nature of tin–
manganese alloy electrodeposits, four samples prepared at
1–4 A/dm2 were immersed in 5 wt % NaCl solution and the change
of corrosion potential with time was monitored. The coating depos-
ited at 1 A/dm2 had approximately 20 wt % manganese, while the
coatings deposited at 2–4 A/dm2 had a manganese content of
around 40 wt %.10 As can be seen from Fig. 1, after an immersion
time of 2 h, tin–manganese alloy electrodeposits have corrosion po-
tentials within the range of approximately −950 to − 1250 mV vs
SCE. With comparison to the corrosion potential of the steel sub-
strate, which was measured as −700 mV vs SCE, there is no doubt
that tin–manganese alloy electrodeposits are anodic to the steel sub-
strate and therefore should provide sacrificial protection to the un-
derlying steel. It was also expected from the composition depen-
dence of the corrosion potential of alloys13 that tin–manganese alloy
coatings with higher manganese content should have more negative
corrosion potentials, as pure manganese was measured to have a
corrosion potential of around −1350 mV vs SCE after being im-
mersed in 5 wt % NaCl for 2 h, while the corrosion potential of
pure tin electrodeposits under similar conditions was measured to be
around −500 mV vs SCE. However, it can be seen from Fig. 1 that
such a simple relationship between the corrosion potential and the
manganese content did not exist for the first few hours of immer-
sion, although it did hold for longer periods of immersion. The
coating deposited at 1 A/dm2 had approximately 20 wt % manga-
nese but it exhibited a corrosion potential more negative than that of
␤a␤c
icorr
=
͓1͔
2.3͑␤a + ␤c͒Rp
where Rp is the value of the LPR.
NSS testing was carried out using a humidified 5 wt % NaCl
spray at 35°C in accordance with ASTM B117. Triplicate samples
were electroplated over an area of 77 ϫ 72 mm. After electrodepo-
sition, some samples were scratched diagonally to expose the steel
substrate, simulating the situation when the coating is damaged. The
corrosion resistance of the coatings was presented as the time to the
first appearance of red rust, i.e., when the unscratched coating had
been penetrated by corrosion and the mild steel substrate had started
to corrode or when sacrificial protection had been lost on the
scratched specimens. After the NSS tests, the composition of the
corrosion products was analyzed using an energy-dispersive X-ray
spectrometer ͑EDX͒ coupled to a scanning electron microscope
͑SEM͒. The phases composing the corrosion products were identi-
fied by X-ray diffractometry ͑XRD Bruker D8 X-ray diffracto-
meter͒, with Cu K␣ radiation ͑␭ = 1.5406 Å͒, 2␪ in the range of
10–75°, at a scan rate of 0.02°/s, a voltage of 40 kV, and a current
of 40 mA. The patterns recorded were then indexed using the Pow-
der Diffraction Files ͑PDFs͒ published by the Joint Committee on
Powder Diffraction Standards ͑JCPDS͒ International Center for Dif-
fraction Data.
Figure 1. Variation of corrosion potential with the immersion time of
Sn–Mn alloy coatings deposited at different current densities.
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
C55
Figure 2. Comparison of the corrosion potential of Sn–40 wt % Mn ͑depos-
ited at 3 A/dm2͒ with different passivated and unpassivated electrodeposits
immersed in 5% NaCl solution for an extended period of time.
Figure 3. Micrograph showing the formation of tin whiskers during the
immersion test of Sn–Mn alloy coatings in quiescent 5% NaCl solution.
coatings deposited at 2 and 3 A/dm2, which contained approxi-
mately 40 wt % manganese.10 This result may be related to our
earlier observation10,14 that among the Sn–Mn alloy coatings studied
within the current density range of 0.5–4 A/dm2, the coatings de-
posited at approximately 1 A/dm2 had the most active surface. They
suffered from a rapid surface oxidation even during ambient storage.
As a result, their surface color turned from bright silver-gray
͑shortly after electrodeposition͒ to dark gray in just a few hours.
Sn–Mn coatings deposited at other current densities also showed the
same characteristic, but it took them much longer ͑a few days͒ to be
oxidized.
For longer-term immersion, the corrosion potential of a tin–
manganese alloy coating electrodeposited at 3 A/dm2 was recorded
over a period of 2000 h and compared with those of pure zinc and
zinc–nickel alloy electrodeposits. As shown in Fig. 2, the corrosion
potential of the tin–manganese alloy electrodeposit remained below
−1000 mV vs SCE for approximately 240 h before it started to shift
in the positive direction. Similar ennoblement was also observed for
zinc–nickel alloy coatings but in a much shorter period of time,
which is in agreement with the results obtained by Wharton et al.6
The ennoblement behavior of zinc–nickel alloy coatings has been
attributed to a dealloying, or more specifically a dezincification
process,15-17 during which the active zinc component from the coat-
ing surface is selectively dissolved. As a result the surface concen-
tration of the more noble nickel component of the alloy is increased,
thereby causing the corrosion potential to shift with time in the
positive direction. For tin–manganese alloy electrodeposits, it is
worth noting that, as shown in Fig. 3, on the surface of tin–
manganese coatings after the long-term immersion tests, many tin
whiskers were found. The rapid formation of tin whiskers on tin–
manganese alloys has been observed previously during ambient
storage.10,14 Such growth of tin whiskers may contribute to the en-
nobling behavior of tin–manganese alloys observed in Fig. 2. A
more detailed discussion regarding this ennoblement is given in sub-
sequent sections of this paper.
when coupled to bare mild steel panels. The galvanic coupling was
to simulate the situation that may arise following coating damage,
which would lead to the exposure of a small area of underlying steel
to the aggressive environment. For coatings which are sacrificial in
nature, they would become anodic and suffer an accelerated rate of
attack, whereas the steel would be cathodically polarized, thereby
reducing its corrosion rate.18 The variation with time in the galvanic
corrosion current generated by the Sn–Mn/steel couple and the cor-
responding coupled potential are shown in Fig. 4. A positive sign
was assigned to the galvanic current flowing from the coating to the
steel ͑through the solution͒, a situation in which the steel was sac-
rificially protected by the dissolution of the coating. For the Sn–Mn/
steel couple ͑Fig. 4͒, after an initial rapid decay, the galvanic corro-
sion current then fluctuated around a value of approximately
4 ␮A/cm2, although the coupled potential was observed to shift
gradually in the positive direction. As the coupled potential in-
creased to about −700 mV vs SCE, ͑i.e., the corrosion potential of
mild steel͒, after 350 h of test, the galvanic corrosion current then
decreased rapidly to small values. Eventually the galvanic corrosion
For pure zinc coatings, especially those chromated, no such early
ennoblement occurred ͑Fig. 2͒. The rapid increase in the corrosion
potential of as-plated pure zinc coating after approximately 1000 h
of immersion was due to its advanced consumption and the conse-
quent loss of sacrificial protection. This was indicated by the occur-
rence of red rust from the corrosion of the underlying steel soon
after the corrosion potential of zinc coating reached that of steel. For
the tin–manganese coating, red rust occurred at approximately
1200 h, while for zinc-nickel alloy coatings, red rust was observed
after their corrosion potentials dropped from the peak values at
around 1000 h.
Figure 4. Galvanic corrosion current ͑Igalv͒ and potential measured for the
Sn–Mn/mild steel couple ͑Ecouple͒. The Sn–Mn coating was deposited at
3 A/dm2 with a manganese content of around 40 wt %.
The sacrificial properties of tin–manganese alloy coatings were
further studied by monitoring their galvanic corrosion behavior
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C56
Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
diffusion-controlled. As a consequence, the galvanic corrosion rate
for all the coatings was in fact determined by the limiting diffusion
current density for oxygen reduction at the mild steel surface, which
is not related to the nature of the anodic coatings but governed by
the concentration of dissolved oxygen in NaCl solution and the cor-
responding diffusion layer thickness in the vicinity of the mild steel
surface. Baldwin et al.15 also found a similar phenomenon in their
study on the galvanic corrosion behavior of electrodeposited zinc
and zinc–nickel coatings coupled with steel under total immersion
conditions. As they reported, the galvanic corrosion current for pure
zinc coatings stabilized at a value of approximately 4.4 ␮A/cm2,
which was believed to be the limiting current density of oxygen
reduction on steel immersed in their quiescent NaCl solutions.
Because the galvanic corrosion was assumed to be governed by
the oxygen reduction on the steel cathode and all the coatings had
shown a similar galvanic corrosion rate, it could be expected that for
the same type of coatings, no matter whether they were chromated
or not, they should have similar time to current reversal and also
similar ennoblement behavior. However, as can be seen in Fig. 5, the
time to current reversal and the ennobling behavior for pure zinc
coatings and zinc–nickel alloy coatings were in fact quite different
and showed apparent dependence on the postplating chromating
treatment. The difference in the galvanic corrosion behavior for the
same type of coatings suggested that during the galvanic corrosion
tests, apart from the galvanic corrosion which was revealed by the
galvanic current flowing in the couple, there was also self-corrosion
of the coatings. That is, the total corrosion rate of the coatings
coupled to mild steel panels was the sum of the self-corrosion rate
that would occur on an isolated coating in the same environment and
the galvanic corrosion rate due to coupling. So, Itotal = Iself + Igalv
.
Such a phenomenon could normally be observed when the coupled
potential is close to the open-circuit corrosion potential of the coat-
ing. For example, Baldwin et al.19 found that when coupled to alu-
minium alloys, the self-corrosion rate of “as-plated” pure zinc and
zinc–nickel alloy coatings accounted for more than 50% of their
total corrosion rates. In this circumstance, the difference in the time
to the reversal of galvanic current observed in this study for pure
zinc or zinc–nickel alloy coatings with or without passivation treat-
ment may be due to the difference in their self-corrosion rates, as it
is known that the self-corrosion rate of coatings is dependent on the
postplating chemical passivation treatment, with the chromated coat-
ings having a much lower self-corrosion rate.
Figure 5. Comparison of the variation of galvanic corrosion current ͑Igalv
͒
and coupled potential measured for different types of coatings coupled to
mild steel.
current reversed its direction, with the mild steel panel becoming the
anode and the coating losing its sacrificial protection ability. As a
consequence, red rust was observed soon after the reversal of the
galvanic corrosion current.
Figure 5 compares the galvanic corrosion currents and the corre-
sponding coupled potentials for five different coating/steel couples.
The time to the current reversal is important for sacrificial coatings
as it shows how long the coating under investigation can protect the
steel coupling electrode from corrosion through sacrificial action. It
can be seen that the Sn–Mn alloy coating, with a time to current
reversal of approximately 430 h, was superior to zinc–nickel alloy
coatings but inferior to zinc coatings. The best sacrificial protection
was achieved with the chromated pure zinc coating. Its corrosion
current did not change direction even after more than 2000 h of
immersion, although the coupled potential shifted to the corrosion
potential of bare mild steel after approximately 650 h of immersion
͑Fig. 5b͒.
In spite of the different nature of the coatings studied ͑Fig. 5͒,
they exhibited similar galvanic corrosion behavior when coupled to
mild steel, namely, all the galvanic corrosion currents generated
fluctuated around a similar value of about 4 ␮A/cm2 before the
eventual current reversal. This independence of the nature of the
coupling coatings is presumably because the galvanic corrosion cur-
rents were limited by the rate of the cathodic reaction taking place
on the mild steel surface. In neutral NaCl solutions, the predominant
cathodic corrosion reaction is oxygen reduction. Under the present
experimental conditions, the samples were in contact with 500 mL
quiescent NaCl solutions confined in a small corrosion cell with a
solution surface area of about 40 cm2, so the solutions were not
air-saturated and the oxygen reduction reaction may have become
Barrier properties.— Another important aspect of concern for
sacrificial coatings is their service life in an electrolyte or when
subjected to outdoor exposure. Such barrier properties are decided
by the self-corrosion rate of the coatings, which can be characterized
by the LPR technique. Figure 6 shows the variation of the LPR
value with immersion time of the tin–manganese alloy coating de-
posited at 3 A/dm2. The LPR value obtained exhibited a consider-
able increase with increasing immersion time, although large fluc-
tuations were also observed, especially within the first 500 h of
immersion. After about 1130 h of immersion, a sharp drop in the
LPR value was observed. Following this, red rust soon occurred on
the surface of the sample. Again, the increase in the LPR may be
either due to the formation of a layer of manganese oxide/hydroxide
with poor conductivity ͑see the section on Corrosion product analy-
sis͒, as found in the corrosion of zinc–manganese alloys,20 or due to
the enrichment of tin at the surface as tin is less active than manga-
nese. It was observed that within the first few hours of immersion, a
black film was formed on the surface of the tin–manganese alloy
coating. This black film then turned “chocolate” brown with in-
crease in immersion time. These films may not be sufficiently dense
and uniform, therefore leading to variation with time of the number
and size of pores in the films and consequently the variation of the
LPR. The protective property of the manganese oxide/hydroxide
film was further confirmed by a similar increase in the LPR of a
pure manganese coating during its long-term immersion in the
5 wt % NaCl solution as shown in Fig. 7. The LPR of pure manga-
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
C57
Figure 8. Anodic polarization curves for the Sn–Mn alloy coating deposited
at 3 A/dm2 and the mild steel substrate.
Note that the linear region of the polarization curve is limited to
within less than 1 decade of current ͑Fig. 8͒. For the measurement of
Tafel constants, the linear region would be ideally over at least 1
decade of current. For the cathodic polarization, the Tafel constant,
␤c, was measured to be approximately 195 mV/decade. However, it
is most likely that this linear behavior arose from the evolution of
hydrogen rather than the reduction of oxygen, which is the predomi-
nant cathodic reaction during corrosion in the NSS. The corrosion
current density for the tin–manganese alloy coating after 2 h immer-
sion ͑Rp = 2700 ⍀/cm2͒ was thus calculated as 17 ␮A/cm2 by ac-
cepting ␤c = 195 mV/decade, or 38 ␮A/cm2 by assuming the ca-
thodic reaction is under the control of oxygen diffusion and
consequently ␤c → ϱ. Such corrosion rates fall between the values
reported for as-plated pure zinc ͑approximately 40 ␮A/cm2 and not
time dependent21͒ and zinc–nickel alloy electrodeposits ͑approxi-
mately 6–10 ␮A/cm2 as reported in Ref. 21͒. This confirmed that,
initially, tin–manganese alloy coatings were active through a disso-
lution mechanism.
Figure 6. Change of LPR value and corrosion potential ͑Ecorr͒ with immer-
sion time of Sn–Mn alloy coatings deposited at 3 A/dm2.
nese increased from approximately 500 ⍀ cm2 after 1 h of immer-
sion to approximately 5000 ⍀ cm2 after 30 h of immersion, a value
of the same order of magnitude to that of the tin–manganese alloy
coatings under study. It can also be seen from Fig. 7 that the corro-
sion potential of pure manganese reached that of mild steel substrate
after approximately 300 h of immersion, suggesting that pure man-
ganese is too active as a sacrificial coating for the protection of steel
against corrosion.
In order to obtain the corrosion rate ͑icorr͒ of the tin–manganese
alloy coating from the LPR value ͑Rp͒ using the Stern–Geary
equation,12 separate potentiodynamic polarization tests were carried
out to obtain the Tafel constants needed, i.e., ␤a and ␤c. The anodic
and cathodic polarization curves for the coating ͑and also the mild
steel substrate͒ recorded after 2 h immersion are shown in Fig. 8 and
9, respectively. The anodic Tafel constant, ␤a, of the tin–manganese
alloy coating was measured to be approximately 233 mV/decade.
In the long-term corrosion rate measurements, however, it may
not be appropriate to use the initial values of Tafel constants for the
calculation of icorr because of their time-dependent nature.21 There-
fore, the experimental results are simply reported as Rp values ͑Fig.
6͒.
During the anodic potentiodynamic polarization tests of tin–
manganese alloy coatings, it was observed that vigorous evolution
of gases occurred as the potential was swept to approximately
Figure 7. LPR value and corrosion potential ͑Ecorr͒ of a pure Mn coating
measured at different immersion times in 5% NaCl solution.
Figure 9. Cathodic polarization curves for the Sn–Mn alloy coating depos-
ited at 3 A/dm2 and the mild steel substrate.
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C58
Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
Figure 11. Effect of applied current density on the time to red rust of Sn–Mn
alloy coatings with a thickness of approximately 8 ␮m.
Figure 10. Variation of corrosion potential of Sn–Mn alloy coatings during
the NSS testing. The Sn–Mn alloy coatings were deposited at different cur-
rent densities shown in the figure.
The effects of current density on the time to red rust of tin–
manganese alloy coatings is shown in Fig. 11. In general, no con-
siderable difference in the time to red rust, which was approximately
1100 h, was observed for the tin–manganese coatings deposited
within 2–4 A/dm2. This may be attributed to their similar chemical
composition, i.e., around 40 wt % manganese in the coatings.10 At
1 A/dm2, the time to red rust showed a much wider variation, i.e.,
720–1600 h. The variation was found to be due to the difficulty in
controlling the deposit composition, as 1 A/dm2 was the current
density at which marked composition differences in the Sn–Mn al-
loy occurred.10 The coatings deposited at 1 A/dm2 with a manga-
nese content lower than 20 wt % exhibited a shorter time ͑around
720 h͒ to red rust, while the coatings with a higher manganese con-
tent ͑approximately 28 wt %͒ exhibited a much longer time ͑around
1600 h͒ to red rust. For the effect of coating thickness on the cor-
rosion performance of tin–manganese alloys ͑deposited at
3 A/dm2͒, Fig. 12 shows that the time to red rust increased with
increasing deposit thickness, a result which may be expected when
the coating is not cracked or defected as its thickness increases.
Figure 13 compares the time to red rust for different types of
coatings in both unscratched and scratched states. With a time of
−400 mV vs SCE. As the potential was further swept in the positive
direction, the current peaked at around −200 mV vs SCE ͑as indi-
cated by the arrow in Fig. 8͒. After the peak, the evolution of gases
stopped and then the corrosion of steel substrate started. A similar
phenomenon was also observed with pure manganese coatings but at
much more negative potentials. The gas evolved was probably oxy-
gen. However, the reaction leading to the evolution of oxygen may
not be due to the electrolysis of water. One possible explanation is
that with increasing potential and consequently an increasing oxidiz-
ing action of the anode, manganese was able to be oxidized to per-
manganate ions ͑MnO−4͒. However, MnO−4 ions are not stable in the
presence of water and they tend to oxidize water with the evolution
of oxygen and the separation of stable MnO2 according to the fol-
lowing reaction22
4MnO−4 + 4H+ → 4MnO2 + 3O2 + 2H2O
͓2͔
Accordingly, once the coating was completely consumed and no
more MnO−4 ions were formed, the anodic current went down and
the evolution of oxygen stopped.
NSS.— Shortly after electrodeposition ͑within 2 h͒, tin–
manganese alloy coatings obtained were subjected to NSS tests. It
was found that once exposed to the salt fog, tin–manganese alloy
coatings were soon corroded with the formation of brown corrosion
products on some areas in less than 1 h. These corrosion products
could be construed as the normal red-brown rust from the corrosion
of the steel substrate; however, they can still be differentiated from
their difference in color as a more chocolate brown. Also, the cor-
rosion products of tin–manganese alloys were not as voluminous as
those of steel. On continued exposure, the coating surface became
darker and more homogenously covered with corrosion products
until the eventual occurrence of red ͑ferrous͒ rust on certain local-
ized areas.
To further confirm that the occurrence of the chocolate brown
corrosion products was not due to the corrosion of steel substrate,
the variation of corrosion potential of tin–manganese coatings dur-
ing the NSS was also monitored ͑Fig. 10͒. It was found that the
corrosion potentials of tin–manganese alloys remained much more
negative than those of steel for a long period of time, irrespective of
the formation of chocolate brown corrosion products on the surface
of the coatings. This suggests that the chocolate brown corrosion
products formed during the tests were not due to the corrosion of the
steel substrate but due to the corrosion of the coating itself.
Figure 12. Effect of deposit thickness on the corrosion resistance of Sn–Mn
alloys. The Sn–Mn alloy coatings were deposited at 3 A/dm2.
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
C59
Figure 13. Comparison of NSS results for different types of coatings with
and without scratches.
approximately 1150 h to red rust, the as-plated tin–manganese alloy
coatings, even without any passivation treatment, were as corrosion
resistant as chromated zinc–nickel alloy coatings. For scratched
samples, no significant decrease in the time to red rust was found
with all the coatings studied, and red rusts were not always first
observed along the scratches. In fact, most of the scratched samples
failed due to the appearance of red rust in other areas rather than
along the scratches, indicating their good sacrificial properties.
Effect of tin whisker growth.— Tin–manganese alloy coatings
were not stable during ambient storage. They were susceptible to
accelerated oxidation and also growth of tin whiskers, a unique phe-
nomenon which has been reported elsewhere.17 The effect of the
changes in the surface state of tin–manganese alloy coatings during
storage on their corrosion performance is therefore of concern and
was studied through the NSS tests. Four tin–manganese alloy coat-
ings deposited at 1–4 A/dm2 were stored at room temperature for
2 months to allow the oxidation of the coatings and the growth of tin
whiskers to occur. Figure 14 shows the surface morphology of such
coatings. After 2 months of ambient storage, a large number of long
tin whiskers were formed and tangled together on the surface of the
coating deposited at 1 A/dm2, whereas on the surface of coatings
deposited at 2–4 A/dm2, less and shorter tin whiskers were ob-
served. These samples after the ambient storage were then subjected
to the NSS tests. Despite the surface oxidation ͑as indicated by the
formation of dark gray oxide͒ and the growth of tin whiskers ͑as
shown in Fig. 14͒ after ambient storage, the NSS corrosion perfor-
mance of these tin–manganese alloy coatings was not appreciably
affected. The time to red rust for all these coatings was similar to
those without ambient storage. Also, there was not a considerable
difference between the corrosion potentials of the coatings with am-
bient storage and those subjected to NSS tests shortly after elec-
trodeposition. This phenomenon may be explained by the fact that
the formation of oxide and the growth of tin whiskers during the
storage probably only took place in a thin surface layer ͑less than
1 ␮m from the surface͒ as shown in Fig. 15, with the remaining
coating underneath not affected. In addition, only sporadic tin whis-
kers were observed on the surface of samples after the NSS tests,
whether they were subjected to a prior ambient storage or not. This
means that tin whiskers, formed either before or during the NSS
tests, may be washed away by the condensed NaCl fog.
Figure 14. Micrographs showing the surface morphology of Sn–Mn coat-
ings after 2 months of ambient storage. The coating was deposited at ͑a͒
1 A/dm2 and ͑b͒ 3 A/dm2, respectively.
was much lower than that in the circled areas. The protrusions in
Fig. 16 ͑marked 3͒ corresponded to the red rust and so were Fe rich.
Figure 17 shows the result of XRD analysis of the corrosion
products of tin–manganese alloy coatings and pure manganese coat-
ings. The corrosion products of both types of coatings produced
similar XRD patterns. They mainly consisted of manganese oxide
͑Mn3O4͒ or manganese iron oxide ͑Mn,Fe͒3O4, and manganese ox-
ide hydroxide ͑MnOOH͒. For tin–manganese alloy coatings, espe-
cially those deposited at 1 A/dm2 with lower manganese content
Corrosion products.— The surface morphology of corrosion
products formed on tin–manganese alloy coatings ͑40 wt % Mn,
deposited at 3 A/dm2͒ after the NSS tests is shown in Fig. 16. From
the compositional analysis, the corrosion products were found to
mainly consist of O, Mn, Fe, and Sn ͑Table II͒. However, no Sn was
detected in areas as indicated by the circle ͑marked 1͒ in Fig. 16,
while the Fe content in the areas indicated by the square ͑marked 2͒
Figure 15. Fractography showing that the growth of tin whiskers on Sn–Mn
alloy coatings only affected the topmost surface layer.
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

C60
Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
Table II. Chemical composition of corrosion products of Sn–Mn
alloy coating deposited at 3 AÕdm2, atom %.
O
Mn
Fe
Sn
Area 1
Area 2
Area 3
Overall
57.69
68.09
63.65
67.06
29.35
23.14
10.96
20.89
12.96
3.69
25.40
9.69
5.08
2.36
may be reduced back to metallic tin, as Sn–Mn alloy coatings had
negative corrosion potentials ͑around −1000 mV vs SCE͒, while the
standard reduction potential for tin is just −378 mV vs SCE.22 As
the best corrosion performance of tin–manganese alloy coatings was
achieved by coatings with a manganese content around 30 wt %, it
can be suggested that there was a combined protection from both the
manganese oxide/hydroxide and tin and tin oxide/hydroxide. The
metallic tin may be deposited around the pores in the manganese
oxide/hydroxide, thus sealing up the active corrosion sites and con-
sequently suppressing further corrosion of the coating. This may
explain the fact that in areas where tin is present, the corrosion
product is more compact ͑Fig. 16͒ and the areas are more corrosion-
resistant ͑as less iron detected in the corrosion products, see Table
II͒. For the ennoblement of the tin–manganese alloy coatings during
the corrosion tests, possible contributing factors may include ͑i͒ the
preferential dissolution of manganese and consequent enrichment of
tin at the surface, as found in the corrosion of Zn–Ni alloys,14-16 ͑ii͒
the growth of tin whiskers, and ͑iii͒ the formation of manganese
oxide/hydroxide, as observed with the corrosion of Zn–Mn alloy
coatings.20 However, it is difficult to determine whether only one of
these factors or a combined effect of these factors is responsible for
the ennoblement.
Figure 17. XRD patterns of corrosion products of pure Mn coating and
Sn–Mn alloy coatings deposited at 1 A/dm2 ͑approximately 20 wt %͒ and
3 A/dm2 ͑approximately 40 wt % Mn͒.
͑around 20 wt %͒, phases identified also included tin oxide ͑SnO͒,
free tin ͑Sn͒, and some manganese tin hydroxide ͓Mn͑Sn͑OH͒ ͔͒.
6
As we reported earlier,10 Sn–Mn alloy coatings with lower man-
ganese content ͑around 20 wt %͒ prepared at 1 A/dm2 mainly con-
sisted of fine-grained MnSn2 intermetallic compounds, while
Sn–Mn alloy coatings with higher manganese content ͑around
40 wt %͒ prepared at 2–4 A/dm2 were amorphous-like. Based on
these microstructural characteristics and also the results of XRD and
compositional analysis of corrosion products described above, it is
proposed that Sn–Mn alloy coatings corroded through the decompo-
sition of manganese-containing intermetallic compounds like
MnSn2. The manganese ions from the decomposition reaction were
Conclusion
The result of this study has confirmed that tin–manganese alloy
coatings, with higher manganese contents ͑more than 20 wt %͒, in
the as-plated state, were as good sacrificial coatings as chromate
passivated zinc–nickel alloy coatings, while they far outperformed
unpassivated pure zinc or zinc–nickel alloy coatings. The major rea-
son for their good corrosion resistance was believed to come from
the physical barrier effect provided by the manganese oxide/
hydroxide formed from the decomposition of intermetallic com-
pounds in the tin–manganese coatings, which was further strength-
ened by the concomitant deposition of metallic tin or tin oxide/
hydroxide. However, tin–manganese alloy coatings were not stable
during ambient storage and suffered from the spontaneous growth of
tin whiskers. Fortunately, the formation of tin whiskers only affected
a thin top surface layer ͑less than 1 ␮m͒ of the coating. Tin whiskers
formed may be washed away during the corrosion; as a result, this
instability did not show any appreciable effect on the corrosion re-
sistance of the coatings, although it did lead to the deterioration of
their appearance. It can therefore be expected that tin–manganese
alloy coatings may find applications as a substitute for chromated
zinc–nickel alloys, where the high corrosion resistance is of critical
importance but the visual effect is not important. For wider indus-
trial acceptance, however, further work is necessary to inhibit the
growth of tin whiskers. Potential solutions to whisker growth would
include postplating annealing heat-treatment,24 the addition of a
third alloying element,24,25 and also possibly the modification of the
coating electrolyte chemistry.
then precipitated as manganese hydroxide ͓Mn͑OH͒ ͔, as the pH
2
value of the corrosion medium was near neutral, which is too high
for manganese ions to exist as free ions.22 Once Mn͑OH͒ was
2
formed, it may be further oxidized, thus leading to the formation of
the more-stable oxide/hydroxides, such as Mn3O4 and MnOOH,23 as
detected in the corrosion products ͑Fig. 17͒. The formation of
Mn3O4 and MnOOH could provide the uncorroded coating under-
neath a physical barrier from the corrosion medium. However, the
tin ions formed during the above-mentioned decomposition reaction
Acknowledgments
This work was supported by a European Community ͑EC͒ grant,
project no. CRAF-1999-70714.
Figure 16. SEM micrograph showing the surface morphology of corrosion
products of Sn–Mn alloy coatings deposited at 3 A/dm2.
Loughborough University assisted in meeting the publication costs of this
article.
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).

Journal of The Electrochemical Society, 155 ͑2͒ C53-C61 ͑2008͒
C61
13. N. D. Tomashov, Theory of Corrosion and Protection of Metals: The Science of
Corrosion, Macmillan, New York ͑1966͒.
14. K. Chen and G. D. Wilcox, Phys. Rev. Lett., 94, 066104 ͑2005͒.
15. K. R. Baldwin, M. J. Robinson, and C. J. E. Smith, Br. Corros. J., London, 29, 299
͑1994͒.
16. N. R. Short, A. Abibsi, and I. K. Dennis, Trans. Inst. Met. Finish., 67, 73 ͑1991͒.
17. M. Pushpavanam, S. R. Natarajan, K. Balakrishnan, and L. R. Sharma, J. Appl.
Electrochem., 21, 642 ͑1991͒.
18. H. P. Hack, Galvanic Corrosion, ASTM, Philadelphia, PA ͑1988͒.
19. K. R. Baldwin, M. J. Robinson, and C. J. E. Smith, Br. Corros. J., London, 29, 293
͑1994͒.
20. M. Sagiyama, T. Urakawa, T. Adaniya, and Y. Fukuda, Plat. Surf. Finish., 74, 77
͑1987͒.
21. K. R. Baldwin, M. J. Robinson, and C. J. E. Smith, Corros. Sci., 36, 1115 ͑1994͒.
22. M. Pourbaix, Atlas of Electrochemical Equilibria in Aqueous Solutions, p. 290,
Pergamon Press, Oxford ͑1966͒.
References
1. G. D. Wilcox and D. R. Gabe, Corros. Sci., 35, 1251 ͑1993͒.
2. E. Budman and R. R. Sizelove, Met. Finish., 102, 291 ͑2004͒.
3. C. W. Ostrander, Electroplating Engineering Handbook, 3rd ed., A. K. Graham,
Editor, pp. 437–447, Van Nostrand Reinhold, Toronto ͑1971͒.
4. T. Biestek and J. Weber, Electrolytic and Chemical Conversion Coatings, Portcullis
Press, Redhill, Surrey, U.K. ͑1976͒.
5. S. Langard and T. Norseth, Handbook on the Toxicity of Metals, Elsevier, New
York ͑1979͒.
6. J. A. Wharton, G. D. Wilcox, and K. R. Baldwin, Trans. Inst. Met. Finish., 77, 152
͑1999͒.
7. C. Müller, M. Sarret, E. García, and J. A. Ortega, J. Electrochem. Soc., 151, C149
͑2004͒.
8. E. Almeida, T. C. Diamantino, M. O. Figueiredo, and C. Sá, Surf. Coat. Technol.,
106, 8 ͑1998͒.
9. J. Gong and G. Zangari, Mater. Sci. Eng., A, A344, 268 ͑2003͒.
10. K. Chen and G. D. Wilcox, J. Electrochem. Soc., 153, C634 ͑2006͒.
11. W. R. A. Meuleman and S. Roy, Trans. Inst. Met. Finish., 81, 55 ͑2003͒.
12. A. T. Kuhn, Techniques in Electrochemistry, Corrosion and Metal Finishing: A
Handbook, Wiley, Chichester ͑1987͒.
23. B. Messaoudi, S. Joiret, M. Keddam, and H. Takenouti, Electrochim. Acta, 46,
2487 ͑2001͒.
24. M. Jordan, The Electrodeposition of Tin and Its Alloys, p. 183, Saulgau/Wurtt.,
Germany ͑1995͒.
25. S. M. Arnold, Plating, 53, 96 ͑1966͒.
Downloaded on 2015-03-03 to IP 128.59.222.12 address. Redistribution subject to ECS terms of use (see ecsdl.org/site/terms_use) unless CC License in place (see abstract).