The Corrosion Behaviour of Objects

Electroplated with Platinurn

IMPORTANT INFLUENCE OF INTERMEDIATE LAYERS
By M. Baumgartner and Ch. J. Raub
Forschungsinstitut fur Edelmetalle und Metalkhemie, SchwSbisrh Gmund, West Germany
If base metals are thinly coated with platinum from an aqueous
electrolyte the deposit is likely to be porous, thus enhancing the
susceptibility 0.f the substrate metal to corrosion and the surface to
discolouration in certain environments. This paper considers the
processes involved and shows that the application of a suitable inter-
mediate layer can effectively prevent this, so enabling the inherent
properties of platinum to be utilised for a greater number of functional
and decorative applications.
Thin layers of metallic platinum can be
produced in a variety of ways: by electrolytic
deposition from aqueous electrolytes or salt
melts, by thermal evaporation or cathodic
sputtering (the physical vapour deposition or
PVD processes) or by firing printed pastes.
Depending on the deposition process and the
parameters involved, the properties of the
coating can vary considerably. Those deposited
from aqueous electrolytes at thicknesses of a
few microns or less are highly stressed and
porous, although bright, hard and wear
resistant. Deposits from molten salts are
generally more. than xpmthick; they are low
stressed and have low porosity, being semi-
bright in appearance, soft and highly ductile.
Layers produced by the PVD processes are
generally in the thickness range of o.~pm or
less, and therefore rather porous. Fired ceramic
pastes often form thick layers with internal
porosity. These processes may compete for
technical applications, for example coatings
produced from aqueous electrolytes or by the
PVD processes are used for decorative and
functional applications by the electronics
industry, on razor blade edges and on spectacle
frames. Deposits from salt melts are found in
the electronic, chemical and glass industries;
fired platinum-containing pastes form decora-
tive and functional coatings in similar areas.
With the exception of coatings deposited
from salt melts, platinum layers a few microns
thick are generally quite porous, which makes
them difficult to use for the protection of less
noble substrate materials. Where porosity in the
platinum coating exposes a base metal surface,
and in the presence of an electrolyte, the
electrochemical nobility of platinum makes
these regions very susceptible to dissolution of
the base metal, discolouration of the platinum
surface by corrosion products and changes in
the chemical or mechanical behaviour which
may result in premature failure of the coated
object.
During the course of a research programme
undertaken at our institute on the properties of
platinum and platinum alloy layers deposited
from aqueous electrolytes we found that even a
thin intermediate layer could have a profound
influence on the corrosion behaviour of nickel
or copper alloy substrates coated with thin
porous platinum coatings.
Electrochemical Considerations
If two metals of different electrochemical
potentials are in direct contact in an electrically
conductive aqueous solution then contact
corrosion may occur. In this case an inter-
mediate potential forms with a value between
those of the two individual metals. Thus the
Platinum Metals Rev., 1985,29, (4), 155-162 155
more noble metal is cathodically and the less
noble metal anodically polarised, with the result
that the corrosion rate of the latter is increased
while that of the nobler metal is decreased.
Such combinations frequently occur, for
example cracked or porous layers of platinum,
rhodium, gold or silver may overlay brass,
bronze, copper or nickel. I n many instances also
a bright sulphur-containing nickel deposit is
applied as an intermediate coating beneath
rhodium or gold layers (I ).
As a first approximation, the normal poten-
tial of metals and alloys can be used to predict
the corrosion behaviour of certain metal com-
binations ( 2) . However caution is essential as
many factors not taken into account when
establishing these normal potentials have a
significant effect on corrosion taking place
under practical conditions; such factors include
for example solution composition (3). Table I
shows the potentials of selected metals in a
sodium chloride-containing solution and in
artificial perspiration, as measured on suitably
prepared samples with a Knick millivoltmeter.
I n order for contact corrosion to occur, a
potential difference must exist between the two
metals and they must be in electrical contact. I t
is also necessary to consider their interaction
Platinum 400
Gold I 240
Palladium 190
Miralloyldl 140
Si l ver 140
Nickel 99.6 50
Bright nickel 40
Copper 10
Table I
Potential of Various Metals as Observed
in Sodi um (:hloride Sol uti on and
Artificial Perspiration
420
2 50
230
150
220
65
45
15
I Potenti al la1
la1 As measured against a silverlsilver chloride electrode
lbl According to DIN 50905. pH 7.5. 20aC
IcI According to DIN 642411
Id1 50 copper-50 tin
Fig. 1 The principle of a contact corro-
sion localised couple is shown, as i t
occurs when a crack through the
platinum plating brings electrolyte into
contact with the platinum and the
copper substrate
with water and its constituents, that is with the
H+ ions, the OH- hydroxyl ions and the
dissolved oxygen. The principles of contact
corrosion as it may occur at a crack in a
platinum layer on a copper substrate are shown
in Figure I .
The dissolution of the anode is determined
by Faradays law, and from it the weight loss,
AG, can be established:
where G =weight loss (grams), t =time
(seconds), F = Faraday constant (96,500
coulombs), M =atomic weight, n =valence of
the ions formed, and I =galvanic current
(amperes). The current I depends on the
diffcrence between the cathodic and anodic
areas, and obeys Ohms law, the electrical
resistance being made up of the resistance
between the corrosion partners (internal
resistance) and that within the electrolyte
(external resistance). Reaction schemes I , 2, 5
and 6 apply. If the anode is insoluble Reactions
3 and 4 take place. The cathodic reactions may
Platinum Metals Rev., 1985,29, (4)
156
Anodic and Cathodic Reaction Schemes
1
Anodic
I M e j Me X+ + xe-
z MeX+ + e
Cathodic
5 Me X+ + xe- j M e
6 Me( X+' ) ++ e- -) Me X+
7 H 3 0 + + e - + H20+f Hz
8 + O 2 + H 2 0 + 2 e - j z(0H)
be just the anodic reactions reversed. If no
metal ions are present-especially in acid
solutions-protons or Ht ions may be
discharged, if hydrogen overvoltage permits.
In the considered copper/platinum system,
copper is dissolved anodically below defects in
the platinum layer. At the cathodic platinum
coating, either H+ ions are discharged as
hydrogen or dissolved oxygen is reduced to
OH- ions, depending on the conditions in the
electrolyte (Reactions 7 and 8 apply). The
dissolved copper ions migrate over the platinum
surface discolouring it around the defects.
Test Procedures
Using copper substrates the corrosion protec-
tion provided by platinum layers deposited
from a commercial electrolyte onto the inter-
mediate layers listed in Table I1 was studied.
The test procedures used were immersion in
artificial perspiration (ISO/DIS 6424/I), the
salts spray test DIN 50021 and the sulphur
dioxide-containing Kesternich test (DIN
50018). All tests were terminated after 48
hours. During the Kesternich test the samples
were examined after I , z and 5 cycles. The
criterion considered was the visual appearance
of a typical part of the surface after testing
compared with that of an as-prepared sample;
exceptional areas such as those showing edge
corrosion were disregarded. The samples
measured zcm x z.gcm and the thickness of the
intermediate layers varied between I and IZ
pm. The thickness of the platinum coating,
which was deposited from a sulphuric acid-
based commercial bath (Platinbad K, DEGUSSA
AG), was determined by X-ray fluorescence.
Corrosion in Perspiration
A summary of the corrosion behaviour
observed in artificial perspiration is shown in
Figure 2. All copper samples p!ated directly
Fig. 2 The summary of
corrosion data, obtained
after 48 hours in artificial
perspiration ( I S0 DIS
6424/I) shows the benefit
of imposing a thin layer of
gold, silver or copper-tin
between the platinum
coating and the copper sub-
strate
I-
z
W
V
.cI
%
t t
--
Platinum Metals Rev., 1985, 29, (4)
157
Substrate
Copper
Copper
Copper
Copper
Copper
Copper
Copper
Copper
Copper
Intermediate layer* Outer plating
None
Bright nickel
Semi-bright nickel
Silver
Gold
Palladium
Palladium on gold
50 copper-50 tin
Gold on silver
Platinum
Platinum
Platinum
Platinum
Platinum
Platinum
Platinum
Platinum
Platinum
In the electroplated form
with platinum showed very poor corrosion
resistance, with corrosion covering the whole
area. Sometimes the coating became separated
from the base. The thickness of the platinum
layer had no significant influence on this
behaviour, since the attack was identical for 0.5
and I .opm thick layers. Microscopic investiga-
tions showed that it was not possible to produce
crack-free deposits at these thicknesses from
this electrolyte, see Figure 3. Attempts were
therefore made to identify suitable materials for
intermediate layers which would prevent cracks
and/or porosity penetrating from the platinum
surface through to the copper substrate, thus
avoiding contact corrosion.
Bright, or semi-bright intermediate nickel
layers proved to have no positive effect. Bright
nickel coatings between 5 and Iopm were
studied. The attack resembled that observed on
samples where platinum had been deposited
directly onto copper, Figure 4. Bright nickel
layers, especially very bright ones with high
sulphur contents, are much less noble than
semi-bright coatings with considerably less
sulphur. The electrochemical potential of semi-
bright nickel is closer to that of platinum than
is the potential of very bright nickel. I t was
therefore interesting to see if the use of the
former would reduce corrosion. However the
corrosion resistance of samples with a 5 to
12pm thick intermediate layer was no better
thdn that of samples with a very bright nickel
layer, as can be seen in Figure 5.
Fig. 3 The platinum layer 1 micron thick
produced from the specified sulphuric acid
based commercial electrolyte on a copper sub-
x 200 strate was severely cracked
Fig. 4 A crack through the 0.2 micron
platinum coating has resulted in localised
corrosion of the bright nickel intermediate
layer (2pm) and the copper substrate x 1000
Fig. 5 Defects penetrate the 0.5 micron
platinum coating enabling corrosion of the 12
micron thick semi-bright nickel intermediate
layer to commence x 500
Platinum Metals Rev., 1985,29, (4) 158
Table II
Layer Combinations
Fig. 6 Despite an artificial crack in the
platinum layer (arrowed) the presence of
the gold intermediate layer (3pm) prevents
the occurrence of contact corrosion x 1750
If the copper substrate is plated with silver
before the platinum coating is applied a
pronounced improvement in the corrosion
resistance is observed. Depending on the quality
of the substrate surface, for example freedom
from defects and low roughness, a 0.5 to r.opm
thickness of silver is sufficient to completely
prevent corrosion, in the conditions employed.
There are two possible explanations for this: [i]
If there is no connection between the copper
substrate and the platinum, through cracks or
pores in the silver, then there can only be
contact between silver and platinum. However,
the contact potential between these two metals
is too small (about 200mV) to cause appreciable
damage. [ii] Silver chloride formed during any
initial corrosion may block cracks or pores thus
preventing any further corrosion.
An intermediate layer of gold with
electrochemical potential of 25omV in artificial
perspiration gives superb corrosion protection if
it is fairly free of pores; a section through such
a coating is shown as Figure 6. The porosity of
plated layers depends on features such as the
quality of the substrate and the electrolyte used,
and in the conditions employed here a
practically pore free coating was obtained at
thicknesses in the range 2 to ym. Indeed some
samples were free of corrosion when the inter-
mediate gold layer was only one r p n thick.
Despite the fact that palladium has a slightly
I'ig.7 The corrosion behaviour of platinum plated copper substrates showing the marked c*Ft'c.c.i
of the composition of the intermediate layer. Top row, left to right: copper/platinum (0. 5~111).
copper/semi-bright nickel ( 12pm)/platinum (0.Spm), copper/silver ( I,um)/platinum (0.5p11):
bottom row, left to right: copper/gold (3pm)/platinum ( OSp m) , copper/palladium ( 2pm) /
platinum (0.5pm). copper/copper-tin ( lpm)/platinum ( 0. 5pm)
Platinum Metals Rew., 1985,29, (4) 159
e
a
e - \
Fig. 8 Corrosion data for
? f various metal combinations
after 48 hours salt spray
test (SS DIN 5002 1 ) show
marked similarity with the
piration tests
- c
c
results of the artificial pers-
higher potential than silver a rather strong
contact corrosion is observed when a palladium
intermediate layer is tested. Apparently a zpm
thick palladium layer was too porous to prevent
contact corrosion effectively. Furthermore the
possibility has to be considered that a localised
contact couple may have been formed between
the palladium intermediate layer and the copper
substrate. A gold strike o.r,um thick between
the copper and the palladium had little
influence on the corrosion behaviour, as com-
pared with a gold-free sandwich structure.
If a white bronze or speculum type of
electroplated deposit, such as Miralloy is used
for the intermediate layer, excellent corrosion
resistance can be achieved, Figure 7. A layer
about Ipm thick is sufficient to give a behaviour
comparable with that of a gold or silver inter-
mediate coating, even if the copper-tin deposit
is less ductile. However, if cracks penetrate
from the platinum outer surface through to the
copper strong corrosive attack is observed.
Corrosion in Artificial Salt Spray
Even when exposure in artificial perspiration
enabled us to make detailed statements on the
behaviour of the various systems, the
standardised salt-spray test (DIN 5002 I , 48
Fig. 9 The behaviour of
plated samples in sulphur
di oxi de-contai ni ng at-
mospheres ( Kesternich
test: DIN 50018 - SFW
2.0s) after 5 cycles showed
some similarities with pre-
vious tests, although a
silver intermediate layer
was not as effective unless
overlaid with a gold strike
Plarinum MeraIs Rev., 1985,29, (4)
160
Fig. 10 The appearance of various platinum plated samples after f i ve rycles of the Kes1c.rnivh
test, showing that the use of an appropriate intermediate layer, or layers t-nablvs ;I vw!
satisfactory corrosion-resistant finish to be produced on a copper substrate
hours) was also applied. The results were quite
similar, as can be seen by comparing Figures 2
and 8. No corrosion was observed for the coni-
binations using gold, silver or copper-tin inter-
mediate coatings. However, with nickel layers a
much stronger attack sometimes occurred, as in
artificial perspiration, possibly due to the loss of
passivity of the nickel in solutions with a high
chlorine concentration.
Sulphur Dioxide Test Results
I n order to observe the behaviour of the
plated objects during attack by gaseous
corrosives we studied their reaction in the
Kesternich test (DIN 5001 8). For combinations
without intermediate coatings, or using nickel
as an intermediate, the attack resembled that
observed in the salt-spray test. A 2pm thick
palladium intermediate layer showed no
improvement either.
Of the precious metal intermediate layer?
used a 2pm or more thick gold coating proved
very effective in obtaining good corrosion
resistance, as can be seen in Figures 9 and 10.
Silver layers were not as effective here as in the
other tests, which can be explained by the
attack of sulphur dioxide on silver and by the
formation of fairly soluble silver corrosion
products, compared with the rather insoluble
silver chloride present in the other tests.
However if the silver intermediate layer is given
a gold strike below the final platinum deposit,
and o.2pm is an effective thickness, the com-
posite is resistant to the sulphur dioxide
atmosphere, see Figure 10. The speculum
copper-tin alloy again proved to be very effec-
tive for corrosion protection in this test, and it
appears that it might be a suitable alternative to
precious metals for the intermediate layer.
Summary
The inherent properties of platinum make its
use as thin electroplated layers interesting for a
variety of applications. However as a result of
its high chemical nobility there is a danger that
localised couples may form between the subs-
trate material and the platinum near any
defects in the platinum coating, thus causing
high corrosion at these spots. I n standard corrc-
sion tests made on samples plated from a
Platinum Metals Rev., 1985,29, (4) 161
specified sulphuric acid based commercial
electrolyte intermediate layers of nickel proved
to be ineffective for corrosion protection, as did
palladium. In chloride-containing solutions
silver proved to be very effective, not only due
to its rather high nobility but also because of
the formation of insoluble silver chloride corro-
sion products. Very good corrosion resistance
under the test conditions was shown by the
layered combination: platinum (outer)/Miralloy
copper-so tin (intermediateycopper substrate,
making this a very satisfactory alternative to
totally precious metal systems.
Acknowledgements
We wish t o thank Impala Platinum Ltd. for
supporting our work on the electroplating of
platinum and DEGUSSA A.G. for providing us with
electrolytes and chemicals. Platinbad K and
Miralloy are registered tradenames of DEGUSSA
products.
References
I K. E. Becker, H. Schlegel and A. Kubat,
2 G. Kortum, Lehrbuch der Electrochemie,
3 J . Elze and G. Oelsner, Metallob@he, 1958, 12,
4 E.Raub and W.Pfaer,Gold BulI.,1980,13,(3),1 I Z
MetallobJache, 1968, ZZ, ( 5) , 145
Verlag Chemie, Weinheim, 1958
(9, 132
Symposium on Fine Chemical Manufacture
Platinum metal catalysts have an increasingly
important role to play in producing materials
for the healthcare, plant protection and
electronics markets. This is a conclusion drawn
from the recent symposium entitled Catalysis
in Fine Chemical Manufacture organised by
the Royal Society of Chemistry from 1~12t h
J uly, at the University of Birmingham. Of the
fourteen papers given, which covered a broad
spectrum of catalytic subjects, eight contained
references to platinum catalysed reactions, and
three drew extensively on platinum metal
catalysis to illustrate the importance of
matching structure and function in catalysis.
These are reviewed here.
Professor P. B. Wells of the University of
Hull considered the objectives of catalyst design
to be the preparation of catalytically active
material in which the chemical and physical
properties of the active sites are not degraded
during use, and the minimisation of side reac-
tions so that the catalysed reaction results in
marketable products. These can be achieved
through modification of the active site environ-
ment, by surface treatment to give car-
bonaceous or sulphur deposits, by occlusion of
hydrogen during reaction, or by the deliberate
choice of a metal-support combination which
interacts beneficially under the reaction condi-
tions. With the platinum and titania combina-
tion, reduction of the TiO, to Ti,O, occurs
locally, facilitating the passage of electrons
from the metal to the support. With certain
supported osmium clusters activity is dependent
on the extent of electron transfer. Sintering is
linked to metal-support interactions, and the
anchoring of small platinum crystallites on
silica by Pt-O-Si bonds helps to retard sintering.
Selectivity has been improved by the use of
multifunctional systems, for example
bifunctional platinum reforming catalysts and
polyfunctional zeolite-based catalysts.
In his lecture Composition and Structure of
Catalysts for Vapour Phase Reactions,
Professor G. C. Bond of Brunel University
described fixed and fluidised bed catalysts.
Fixed bed reactors utilise large particles such as
pellets, while fluidised bed reactors employ
small particles of uniform size. Otherwise the
catalysts are very similar, being composed of an
active phase (a platinum metal), a support (such
as alumina or silica), and a promoter which may
alter the texture, electronic structure or poison
resistance of the combination - as with the
addition of rhenium to a platinum reforming
catalyst. In phenol hydrogenations the catalyst
is bifunctional, the platinum metal reducing the
aromatic ring and the support acid function
rearranging the intermediate to give a cyclic
ketone. Metal location is important as it affects
the mass transport within the pores. This factor
is to a large extent determined by the method of
preparation; fast absorption from solution gives
uniform deposition within the pores while slow
absorption allows metal to deposit at pore
mouths.
Metal dispersion and location and the effect
of these parameters on the activity-selectivity of
reactions was the main theme of Mr I. L.
Dodgsons lecture. He described the types of
palladium on charcoal catalysts manufactured
by J ohnson Matthey Chemicals for liquid phase
reactions, discussed how their physical
structure controls their behaviour in selected
industrial reactions and recommended
appropriate catalysts. A.J.B.
Platinum Metals Rev., 1985,29, (4)
162