Scientific_Publication

Impact of the nickel interlayer on the electrical resistance of tin–tin interface
submitted to fretting loading
P. Jedrzejczyk a,b
, S. Chad a
, S. Fouvry a,
⁎, P. Chalandon c
a
Laboratoire de Tribologie et dynamique des Systèmes, CNRS UMR 5513, Ecole Centrale de Lyon, 36 av Guy de Collongue, 69134 Ecully Cedex, France
b
Institute of Materials Engineering, Technical University of Lodz, Stefanowskiego 1, 90924 Lodz, Poland
c
PSA Peugeot Citroën, Materials Department, DTI/DPMO/CIMB, Centre Technique de Belchamp / 25208 Montbéliard, France
a b s t r a c ta r t i c l e i n f o
Article history:
Received 16 April 2008
Accepted in revised form 11 December 2008
Available online 24 December 2008
Keywords:
Fretting
Electrical contacts
Tin coatings
Nickel underlayer
Variable displacement amplitude
The behaviour of the tin/tin interface for electrical contact application is addressed in this paper. Two coating
systems were studied: a bronze–tin coating system and a bronze–nickel–tin coating system. The transition
amplitude δt from partial slip (P.S) to gross slip (G.S), defining the domains of infinite and finite lifetime of the
contact, was found using the variable–displacement–amplitude methodology (VDA). Constant–displacement–
amplitude (CDA) tests were performed in order to determine the influence of the nickel interlayer on the
performance of the electrical contact. The conducted analysis concluded that there is no influence of the nickel
interlayer on the electrical endurance during G.S. However, the nickel interlayer, which serves as a diffusion
barrier, eliminates the diffusion of copper through the tin coating; thus preventing the formation of copper oxides
on the top surface. It concludes that the application of the nickel underlayer by maintaining high tin–tin friction
coefficient extends the P.S domain and therefore increases the reliability of the electrical contact.
© 2009 Elsevier B.V. All rights reserved.
1. Introduction
The use of tin as a coating material in electrical contacts is
prevalent. The low-cost application of tin is an alternative to the more
expensive solutions using noble materials such as gold [1] or silver.
Although tin is a cheap solution, it is also very susceptible to the
fretting conditions, which is its major disadvantage.
Small oscillatory movements induced by vibration and thermal
expansion can severely degrade the electrical properties of the
contact. Indeed those movements, referred to fretting, cause wear of
the interface and induce debris formation which can dramatically
decay the electrical resistance.
An extensive effort has been made to define the mechanisms of
damage of electrical contacts. Antler was one of the first researchers to
give a review on this problem [2,3]. Several studies have been devoted
to each of the materials used in connectors. Gold finishes were studied
by Rabinowicz et al. [4,5], Antler and Drozdowicz [6], Togasaki [7].
Silver coatings were subject to thorough investigation by Kassman-
Rudolphi and Jacobson [8–11].
The various aging mechanisms and the range of oxide films formed
on the surface of the tin interface were characterised by Malucci
[12,13]. The formation of intermetallic phases, another phenomenon
very relevant to the performance of the tinned contacts, was
investigated by Braunovic [14], Noel [15], and Haimovich [16]. Park
et al. performed extensive research on the influence of mechanical and
environmental loadings on the electrical resistance. Their thorough
studies showed the influence of temperature [17,18] and fretting
corrosion [19,20] on the contact resistance.
The work of Hannel et al. [21,22], Kassman-Rudolphi and Jacobson
[9,10] provides information on the influence of the sliding regime on
the performance of electric connectors.
The aim of this work is to investigate the influence of the nickel
interlayer, used as a barrier against the copper diffusion, on the
transition from P.S to G.S condition as well as its influence on the
performance of the contact under G.S condition. To study this aspect
the VDA methodology [23] was applied. The second goal was to
confirm the conformity of this methodology with the traditional CDA
test methodology, especially the possible effect of imposed mechan-
ical loadings history on the obtained results. The last aspect of the
research work was to confirm the hypothesis of the low and stable
electrical resistance under P.S and immediate degradation of this
parameter when contact turns to G.S condition.
2. Experimental details
2.1. Experimental setup
An original setup was designed and built for this study. This
machine made it possible to simulate environmental conditions
similar to those due to a car engine. Fig. 1(a) illustrates the schematic
diagram of the machine.
The oscillatory motion of the upper holder, fixed via flexible strips,
wasproducedby theelectromagnetic shaker. Anupper samplewas fixed
Surface & Coatings Technology 203 (2009) 1624–1628
⁎ Corresponding author. Tel.: +33 4 72 18 65 62; fax: +33 4 78 43 33 83.
E-mail address: Siegfried.Fouvry@ec-lyon.fr (S. Fouvry).
0257-8972/$ – see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.surfcoat.2008.12.008
Contents lists available at ScienceDirect
Surface & Coatings Technology
journal homepage: www.elsevier.com/locate/surfcoat

to the upper holder while the second sample was fixed to the lower
table, whose slight horizontal displacement allowed the tangential force
to be measured during the test using a piezoelectric load sensor. The
upper specimen movement was controlled using a laser displacement
sensor to an accuracy of about 0.1 µm. A dead mass placed on the upper
holder was applied to assure the normal force loading. The dedicated
software, based on the Labview platform, was used to provide all
necessary recording and monitoring. The parameters which can be
recorded and controlled during each test are as follows: the relative
displacement δ⁎ [from ±0.5 to ±40 µm], the frequency f [from 1 to
500 Hz], the normal force P [from 0.2 to 5 N], the relative humidity RH
[from 1 to 99%], and the temperature T [from 20 to 160 °C].
3. Contact configuration and electrical resistance
measurement system
The sample system used in this study consisted of two 90° crossed
semi-cylinders (Fig. 1(b)) each with a radius of 2.3 mm. To measure the
electrical resistance of the contact during the test a four wire method
was applied. Two wires, one for each sample, supplied a stabilised
current I=0.005 A±0.2%. Another two wires were used to measure the
contact voltage. This system enabled the electrical resistance from 10−6
to 103
Ω to be measured. The number of sliding cycles NC to reach the
threshold value of the electrical resistance of RC =0.004 Ω was assumed
as a lifetime of the contact.
3.1. Materials studied, roughness and test conditions
3.1.1. Materials
Two different tribocouples were investigated. The first system
consisted of bronze substrate covered with the 2 µm thick pure-tin
coating. The second system was similar, but a 2 µm thick nickel
interlayer was deposited between bronze substrate and tin coating.
3.1.2. Surface roughness
The surface roughness of the studied materials was measured
before the fretting tests. For each studied material ten measurements
along the cylinder's peak (contact areas) were conducted. The
obtained results concerning the Ra parameter are listed in Table 1.
As it can be clearly seen the surface roughness parameters are almost
the same so it can be initially concluded that the roughness of the
samples has no influence on obtained results.
3.1.3. Test conditions
Identical experimental conditions were applied to both tested
couples: Temperature 23 °C, relative humidity 10%, frequency 30 Hz,
normal force 3 N.
3.2. Test methodology
Two different tests were performed on each of the studied coating
systems. The first type of test was a VDA test. The second type was a
conventional CDA test.
The VDA tests allow the determination of the boundary between P.
S and G.S in one single test. The principles of the VDA methodology are
shown in the Fig. 2. For a contact with given geometry and normal
load, first a small displacement amplitude (δ⁎=2 µm) is applied to
ensure that the initial sliding condition is P.S. The displacement
amplitude is being then progressively increased in small increments
(Δδ⁎=1 µm). Each displacement amplitude is maintained for a
sufficient length of time (ΔNC =15,000 cycles), so that dynamic
conditions in the contact have stabilised. In this way the transition
amplitude δt is determined in a single test. The CDA test methodology
consists of the tests carried out in the constant–displacement
amplitudes. The imposed amplitude is maintained constant during
all period of the test; the electrical resistance is being measured as
well. When the value of the electrical resistance becomes higher than
the threshold value (RNRC), the lifetime of the electrical contact (NC)
can be estimated. The lifetime of the electrical contact for different
amplitudes can be distinguished allowing the creation of endurance
chart, so called Wohler-like curve. This curve represents the lifetime
(NC) of the contact as a function of applied displacement amplitude δ⁎.
Fig. 1. (a) Schematic presentation of the experimental setup and (b) crossed-cylinders
configuration used in the study (illustration of the 4 points electrical measurement
method).
Table 1
Surface roughness parameters
Ra [nm] Standard deviation [nm]
Cu–Sn system 851.4 38.8
Cu–Ni–Sn system 848.8 26.2 Fig. 2. Illustration of the principles of the Variable Displacement Amplitude (VDA)
methodology.
1625P. Jedrzejczyk et al. / Surface & Coatings Technology 203 (2009) 1624–1628

4. Results
4.1. Tin/tin interface with nickel interlayer
First the variable–amplitude test on the system with a nickel
interlayer was conducted. The results of this test can be seen in the
Fig. 3. To the left of the graph, the incremental evolution of the force ratio
f=Q⁎/P, can be seen. Every augmentation of the force ratio is caused by
the increase of the displacement amplitude δ⁎. The initial displacement
amplitude was fixed at δ⁎=2 µm. During the test, the amplitude was
increased by increments of Δδ⁎=1 µm which in turn increased the force
ratio. The maximumvalue of the force ratio is f=1.22, which corresponds
to the transition amplitude δt =5 µm. When the amplitude reaches the
value of 5 µm the system turns from P.S to G.S.
The critical impact of the sliding condition on the resistance of the
tin–tin interface is confirmed by the CDA test (Fig. 4). The test performed
at the amplitude δ⁎=4 µm (below the transition amplitude δt =5 µm)
showed very low and stable electrical resistance (NC N106
), whereas the
test performed at the amplitude δ⁎=6 µm (above the transition
amplitude δt =5 µm) revealed a very fast increase of the resistance to a
value over the threshold. It must be clear that this transition is
dependent on the applied conditions and the normal force P in the
contact. The higher the normal force P, the higher the transition
amplitude as expected from Mindlin elastic formalism δtαP2/3
[24].
The results of several constant-amplitude tests are summarised
and graphically presented in Fig. 5. It can be observed that the lifetime
of the electrical contact (number of cycles NC) decreases as the
displacement amplitudes increase. This behaviour is typical for
electrical resistance when contact is operating under G.S condition.
The transition amplitude δt =5 µm marks the border between the
limited and unlimited lifetime of the electrical contact.
Two zones can be observed in the fretting scar (Fig. 6(a)). The
first zone is the central sticking area where good metal to metal
contact is assured. The passage of the electrical current is provided
through this zone. The second zone, represented by the external
annulus, is the sliding zone where the relative displacement
between two contacting surfaces takes place. The difference in
the chemical composition of those two zones can be seen in the
EDX patterns presented in Fig. 6(b). In the inner zone of the contact
only tin is observed, no other concentration of chemical elements
can be seen. On the contrary, in the outer zone the presence of a
few spots of nickel can be observed, as well as a large amount of
oxygen indicating tin oxides. The tin layer was partially removed, in
the form of debris, from the sliding zone as a result of the relative
displacement between two surfaces.
4.2. Tin/tin interface without nickel interlayer
The test schedule used in the investigation of the second coating
system was identical to the one applied to the first system. The VDA
test was initially carried out. The result of this test can be seen in the
Fig. 7. It shows an incremental increase of the force ratio f. It can be
seen, that in this case the system moves from P.S to G.S when the
displacement amplitude reaches the transition value of δt =4 µm. The
corresponding value of the force ratio is f=0.8. After the transition, the
contact operates in the G.S condition. As previously shown just after
the transition the electrical resistance starts to increase slowly for
δt =4 µm and then rapidly for δ⁎N5 µm.
The tests conducted under CDA conditions lead to a similar
evolution with an asymptotic increase of the lifetime of the
electrical contact with the decrease of the applied displacement
amplitude (Fig. 5). As before, the transition amplitude δt =4 µm,
defined thanks to VDA methodology is pertinent to predict the
excellent electrical endurance (NC N106
) of the contact below δt and
limited performance above δt.
The analysis of fretting scar presented in the Fig. 8(a) reveals the
morphology of the contact surface after the test. The zone of contact is
mostly homogenised and it can be noted that some plastic deforma-
tions of the interfaces took place. It is quite difficult to distinguish the
Fig. 3. VDA test of the tin/tin interface with nickel interlayer: plotting of the force ratio
f=Q⁎/P and electrical resistance as a function of the fretting cycles.
Fig. 4. Electrical resistance endurance of tin/tin interface with nickel interlayer,
Constant Displacement Amplitude (CDA) tests, displacement amplitudes: δ=4 µm (P.S),
δ=6 µm (G.S), LT — lifetime of the electrical contact NC =5750 cycles.
Fig. 5. Endurance charts obtained for the both coating systems. ▲ tin–tin coating
without nickel interlayer, □ tin–tin coating with nickel interlayer, , — infinite
lifetime of the contact, sliding transition amplitude δt =5 µm, tin–tin interface
with nickel interlayer, sliding transition amplitude δt =4 µm, tin–tin interface
without nickel interlayer.
1626 P. Jedrzejczyk et al. / Surface & Coatings Technology 203 (2009) 1624–1628

two different zones characteristic of the P.S condition. The EDX
patterns (Fig. 8(b)) show an increased concentration of oxygen in the
outer and central zone of contact. Although the contact was running
under P.S condition, surprisingly high concentration of copper is
clearly noticeable, especially in the central zone of the contact. The
presence of tin coating over the whole interface assures that the
substrate had not been reached.
5. Discussion
5.1. Correlation between the sliding condition and electrical endurance
This work confirmed the direct correlation between the sliding
condition and electrical performance of the contact. Infinite lifetime of
the electrical contact is achieved when contact is running under P.S
condition. Finite endurance is obtained during G.S conditions. The
comparison between the VDA results and endurance chart shows that
the transition amplitude is not exactly the threshold amplitude δe
marking the infinite and finite lifetime. In fact, the contact at the sliding
transition is unstable and can randomly fluctuate between P.S and G.S.
This can explain the finite endurance of the contact in the test conducted
at δt (Fig. 5). The infinite endurance is obtained for stabilised P.S
condition corresponding to displacement amplitude slightly smaller
than δt. This infers the following relationship: δe =δt −δR, with δR —
residual displacement amplitudewhich allows a stabilized P.S condition.
In this study we assumed δR =1 µm. Future investigation is required to
better formalize this parameter.
5.2. Impact of nickel interlayer
The comparison between the two Wohler curves obtained for
studied coating systems concluded that there is no large difference in
the electrical lifetime of the tin–tin interface when the G.S condition is
operating. The tests carried out in the same displacement amplitudes
lead to a very similar electrical lifetime. From this, the conclusion can
be drawn that the nickel interlayer does not influence the lifetime of
the electrical contact under G.S. The only impact that can be noticed is
Fig. 7. VDA test of the tin/tin interface without nickel interlayer: plotting of the force
ratio f=Q⁎/P and electrical resistance as a function of the fretting cycles.
Fig. 8. Observation of the P.S tin/tin (without nickel interlayer) fretting scar (δ⁎=4 µm,
50,000 cycles); (a) SEM micrograph (b) EDX chemical elements distribution maps.
Fig. 6. Observation of the P.S tin/tin (with nickel interlayer) fretting scar (δ⁎=5 µm,
50,000 cycles); (a) SEM micrograph (b) EDX chemical elements distribution maps.
1627P. Jedrzejczyk et al. / Surface & Coatings Technology 203 (2009) 1624–1628

the change of the transition amplitude from P.S to G.S condition. The
system with the nickel interlayer presents wider domain of the P.S
condition than that without nickel interlayer. This extension of the P.S
domain, which is clearly beneficial to electrical performance, can be
only interpreted considering the tribological response of the inter-
faces. It is shown that the interlayer influences the value of the friction
coefficient which is much higher for the system with nickel (µt =1.22)
than for the system without nickel (µt =0.8). Considering the Mindlin
description of the sphere/plane P.S contact it is shown, that the P.S–G.S
transition amplitude is proportional to the friction coefficient. There-
fore, the higher the friction coefficient is, the higher the transition
amplitude δt, and so the wider the P.S domain. The question is how to
explain such tribological difference considering the fact that the tin
layer is still present in the contact area.
The EDX patterns (Fig. 8(b)) show that copper and oxygen are
present in the P.S fretting scar obtained for the system without a nickel
underlayer. It can be concluded that due to the plastic deformation of
the surface, copper atoms diffuse through the tin layer up to the
surface. The subsequent creation of copper oxides, which are known
to provide the low friction coefficient, will favour the lower transition
amplitude from P.S to G.S condition. To confirm presence of the copper
oxides, local investigation involving micro-Raman analysis, which can
not be included in this work are currently planed. However the
combination of copper, tin and oxygen elements suggests that both
copper and tin oxides are generated in the contact zone.
In the case of the system with the nickel interlayer, which serves as
a copper diffusion barrier, the copper oxides can not be created, thus
higher friction coefficient, related to metallic tin–tin and probably tin–
oxides interface, is maintained. Higher friction coefficient will favour
the extension of the domain of P.S. The comparison with the system
without a nickel interlayer suggests that the driving factor reducing
the coefficient of friction and consequently the sliding transition is the
presence or absence of copper oxides on the top of the surface of P.S
contact.
Future investigation will be undertaken to clarify this aspects and
the relative influence of friction and mechanical properties on sliding
response of studied materials.
6. Conclusions
The main results of this research can be resumed in the following
points.
1. The study confirms the stability and reliability of the VDA test for
the determination of the transition amplitude δt from P.S to G.S
condition.
2. The non-noble tin–tin electrical contact performance is fully
controlled by the operating sliding condition (P.S-infinite lifetime,
G.S-finite lifetime).
3. The nickel interlayer has no influence on the lifetime and
performance of the electrical contact under G.S condition but
plays a critical role on tribological behaviour under P.S by extending
this domain.
4. The chemical analysis of the fretted surfaces suggests that nickel, by
stopping the diffusion of the copper atoms through the tin layer,
prevents the formation of the lubricious copper oxides. The high
friction coefficient of the tin–tin interface promotes safe P.S
condition. It demonstrates the fundamental result that the
favourable effect of nickel interlayer is in fact related to its capacity
to modify the friction behaviour under P.S condition (friction and
displacement amplitude at the P.S//G.S sliding transition).
5. This investigation has shown that the nickel interlayer by modifying
the chemical composition of the top surface influences the friction
coefficient in the P.S condition and consequently the P.S//G.S
transition. The diffusion process of the copper into the tin layer can
also modify the mechanical properties of the coating (i.e. hardness
and elastic properties). Indeed the mechanical properties of copper–
tin system can be different from copper–nickel–tin system. In order
to verify the veracity of this assumption the local nano-hardness of
mentioned systems should be performed. Unfortunately this can not
be included in the frame of this work.
References
[1] C.J. Evans, Components Hybrids Manuf. Technol. 3 (2) (1980) 226.
[2] M. Antler, Wear 106 (1985) 5.
[3] M. Antler, IEEE Trans. Components Hybrids Manuf. Technol. 8 (1985) 87.
[4] H. Tian, N. Saka, E. Rabinowicz, Wear 142 (2) (1991).
[5] H. Tian, N. Saka, E. Rabinowicz, Wear 142 (1) (1991) 57.
[6] M. Antler, M.H. Drozdowicz, Wear 74 (1981–1982) 27.
[7] N. Togasaki, Y. Okinaka, T. Homma, T. Osaka, Electrochim. Acta 51 (2005) 882.
[8] A. Kassman-Rudolphi, S. Jacobson, Tribol. Int. 30 (3) (1997) 165.
[9] A. Kassman-Rudolphi, S. Jacobson, Wear 201 (1–2) (1996) 244.
[10] A. Kassman-Rudolphi, S. Jacobson, Wear 201 (1–2) (1996) 255.
[11] A. Kassman-Rudolphi, S. Jacobson, Wear 165 (2) (1993) 227.
[12] R.D. Malucci, IEEE Trans. Compon. Packag. Technol. 22 (1) (1999) 53.
[13] R.D. Malucci, IEEE Trans. Compon. Packag. Technol. 24 (3) (2001) 399.
[14] M. Braunovic, Proc. 49th IEEE Holm Conf on Electrical Contacts, Washington, 2003,
p. 124.
[15] S. Noel, N. Lecaude, S. Correia, P. Gendre, A. Grosjean, Proc. of the 52 IEEE Holm
Conf. Montreal, 2006, p. 1.
[16] J. Haimovich, J. Technol. 3 (1993) 46.
[17] Y.W. Park, T.S.N. Sankara Narayanan, K.Y. Lee, Wear 262 (2007) 320.
[18] Y.W. Park, T.S.N. Sankara Narayanan, K.Y. Lee, Tribol. Int. 41 (2) (2008) 87.
[19] Y.W. Park, T.S.N. Sankara Narayanan, K.Y. Lee, Surf. Coat. Technol. 201 (2006) 2181.
[20] Y.W. Park, T.S.N. Sankara Narayanan, K.Y. Lee, Tribol. Int. 40 (3) (2007) 548.
[21] S. Hannel, S. Fouvry, P.H. Kapsa, L. Vincent, Wear 249 (2001) 761.
[22] S. Hannel, Sur l'importance de la transition de glissment en fretting pour la
connectique, PhD thesis, Ecole Centrale de Lyon, Laboratoire de Tribologie et
Dynamique des Systemes, 2000.
[23] J.M. Voisin, A.B. Vannes, L. Vincent, J. Daviot, B. Giraud, Wear 181–183 (1995) 826.
[24] R.D. Mindlin, J. Appl. Mech. 16 (1949) 259.
1628 P. Jedrzejczyk et al. / Surface & Coatings Technology 203 (2009) 1624–1628

Scientific_Publication

More Related Content

Scientific_Publication