Influence of pulse plating parameters on the electrocodeposition of matrix metal nanocomposites

Article abstract of DOI:10.1016/j.electacta.2007.06.009

Electrocodeposition of alumina particles with copper and nickel from acidic electrolytes has been investigated using different deposition techniques. Compared to direct current (DC) deposition, both pulse plating (PP) and pulse-reverse plating (PRP) facilitated higher amounts of particle incorporation. With conventional DC plating the maximum alumina incorporation is ～1.5 wt% in a nickel and ～3.5 wt% in a copper matrix. However, the implementation of rectangular current pulses can give considerably higher particle contents in the metal layer. A maximum incorporation of 5.6 wt% Al₂O₃ in a copper matrix was obtained by PP at a peak current of 10 A dm^-2, a duty cycle of 10% and a pulse frequency of 8 Hz. In general, low duty cycles and high pulse frequencies lead to an enhanced particle codeposition. The microstructure and the hardness of both pure metal films and nanocomposite coatings showed only a weak dependence on the PP and PRP conditions.

Full text of DOI:10.1016/j.electacta.2007.06.009

Electrochimica Acta 52 (2007) 7362–7371
Influence of pulse plating parameters on the electrocodeposition
of matrix metal nanocomposites
Denny Thiemig, Ronny Lange, Andreas Bund∗
Department of Physical Chemistry, Dresden University of Technology, D-01062 Dresden, Germany
Received 5 March 2007; received in revised form 31 May 2007; accepted 2 June 2007
Available online 12 June 2007
Abstract
Electrocodeposition of alumina particles with copper and nickel from acidic electrolytes has been investigated using different deposition
techniques. Compared to direct current (DC) deposition, both pulse plating (PP) and pulse-reverse plating (PRP) facilitated higher amounts of
particle incorporation. With conventional DC plating the maximum alumina incorporation is ∼1.5 wt% in a nickel and ∼3.5 wt% in a copper
matrix. However, the implementation of rectangular current pulses can give considerably higher particle contents in the metal layer. A maximum
incorporation of 5.6 wt% Al2O3 in a copper matrix was obtained by PP at a peak current of 10 A dm−2, a duty cycle of 10% and a pulse frequency
of 8 Hz. In general, low duty cycles and high pulse frequencies lead to an enhanced particle codeposition. The microstructure and the hardness of
both pure metal films and nanocomposite coatings showed only a weak dependence on the PP and PRP conditions.
© 2007 Elsevier Ltd. All rights reserved.
Keywords: Electrocodeposition; Copper; Nickel; Alumina; Nanocomposite; Pulse-plating
1. Introduction
the metal matrix were found to be crucial to improve the
coating properties [10,11]. In many cases the enhanced per-
formance of the coatings is mainly caused by a change in the
growth mode or grain size of the metal matrix [11,12] and
not so much by the presence of the particles themselves. The
particle incorporation is affected by several interrelated pro-
cess variables, e.g. current density, pH and bath composition,
hydrodynamics and particle size, type and shape [2,13–17]. Fur-
thermore, thetimedependenceoftheappliedcurrentisoneofthe
most important parameters in the electrocodeposition process
[17].
The structure of electroplated metal and alloy coatings
is specified by the electrocrystallization process, particularly
by the interplay between nucleation and crystal growth [18].
An attractive way of controlling these two processes is the
application of a periodically changing current [19]. Thus the
microstructure and the properties of the films can be easily
controlled by varying, e.g. cathodic pulse length (ton), pulse
pause (toff), cathodic peak current density (ip), anodic pulse
length (trev), anodic peak current density (ian) and average cur-
rent density (iave) [19]. Compared to DC plating, PP can yield
nanocrystalline coatings with improved surface appearance and
properties, such as smoothness, refined grains and enhanced
corrosion resistance [12,20–24].
The incorporation of inert particles during metal electrode-
position has stimulated scientific and technological interests
for decades [1–3]. Recently, electrocodeposition was industri-
ally used to produce abrasion and corrosion resistant coatings,
e.g. for automotive applications [4]. The effect of the operat-
ing conditions influencing the quantity of incorporated particles
was studied for a variety of interesting metal–particle combi-
nations [5]. Several theoretical models have been proposed to
describe the electrocodeposition phenomena [1,3,5–8]. How-
ever, up to date the mechanism of particle incorporation is not
fully understood. This lack of fundamental understanding of
the electrocodeposition process has resulted in a trial and error
approach being the only method of developing parameters for
industrially applicable codeposition of composite films with
fairly reproducible particle content.
The properties of metal matrix composites (MMCs) mainly
depend on their composition and structure [9]. Uniform dis-
tribution and high amounts of incorporated particles within
∗
Corresponding author. Tel.: +49 351 463 34351; fax: +49 351 463 39820.
E-mail address: andreas.bund@chemie.tu-dresden.de (A. Bund).
0013-4686/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2007.06.009

D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
7363
Table 1
Electrolyte composition and deposition parameters
Copper sulfate
Nickel sulfamate
Metal
Other
0.8 M CuSO4 × 5 H2O
1.08 M Ni(NH2SO3)2; 0.04 M NiCl2 × 6H2O
0.55 M H2SO4; 3.26 mM HCl
0.65 M H3BO3
pH
1.7
45
4.3
40
Temperature (◦C)
Particles in suspension (g L−1
Total electrical charge (C)
)
0, 1, 5, 10
140
0, 1, 5, 10
140
Besides the electrocrystallization the particle inclusion dur-
ing MMC plating is affected by pulse plating. It has been
observed that PP [15,17,25–27] as well as PRP [28,29] could be
used to enhance the particle incorporation. In the case of PRP
of Cu–Al2O3 composites, the maximum particle incorporation
was found when the deposit thickness per cycle approaches the
particle diameter [28]. Furthermore, it has been shown in ref.
[30] that PRP with pulses in the order of seconds can be used for
a size-selective incorporation of alumina nanoparticles into cop-
per matrices. However, the distribution of incorporated particles
can be also improved by PP. In the case of Ni–Al2O3 compos-
ites, pulse plating leads to an incorporation of smaller and less
agglomerated nanoparticles [31,32].
There are some reports on the effects of PP and PRP on
the electrocodeposition of alumina with copper and nickel from
acidic electrolytes [25,27,28,30–33]. Promising improvements
in terms of the amount and the distribution of incorporated alu-
mina particles have been reported with a pulse length in the order
ofmilliseconds[27,31,32]. However, noextensiveinvestigations
on the effect of the several interrelated PP and PRP parameters,
such as ton, ip, toff, trev and ian, can be found in the literature.
The present paper aims at shedding some light on this issue.
The codeposition behavior of ␥-Al2O3 with copper and nickel
was studied using three different electrodeposition techniques.
The results of the DC plating experiments were published in
detail in previous papers [11,13]. Here we discuss the effect of
pulse plating (PP) and pulse reverse plating (PRP) conditions,
e.g. pulse frequency, duty cycle and value of the peak current
density, and particle loading on the composition and properties
of the MMCs. The duration of the employed rectangular current
pulses was in the order of milliseconds. The particle content,
structure and hardness of the films have been characterized and
their dependence on the deposition parameters will be discussed.
D1, accuracy 1 ◦C). During the electrocodeposition the elec-
trolyte was magnetically stirred at a speed of about 300 rpm in
order to keep the nanoparticles in suspension. The anode was
a bent plate of pure nickel or copper cylindrically surround-
ing the working electrode. Copper plates of 2.3 cm2 area were
used as working electrodes. Prior to the electrocodeposition the
substrates were ground using 4000 grade silicon carbide paper,
electrochemically degreased and activated with uniclean 675
solution (Atotech Germany GmbH). A potentiostat/galvanostat
model 273 (EG&G Princeton Applied Research) was used for
the deposition of coatings by dc currents. PP and PRP experi-
ments were carried out using a CAPP pulse-plating system (Axel
Akerman S.A., Denmark). After electrodeposition the samples
were sonicated in distilled water for 10 min in order to remove
loosely adhering particles.
In the pulse deposition experiments the pulse frequency [i.e.
f = 1 /(ton + toff)] and the duty cycle [i.e. θ = ton/(ton + toff)] of the
imposed rectangular pulses were varied. The pulse on-time (ton)
was varied between 25 and 200 ms and the pulse pause (toff
)
was kept constant at 100 and 200 ms, respectively (Table 2). For
the PP experiments the cathodic peak current density (ip) was
maintained at 5 and 10 A dm−2, respectively. The overall elec-
trocodeposition times were chosen so that the total charge was
always 140 C. This corresponds to a layer thickness of about
21 ␮m for the pure metals. In the case of PP depositions the
particle content in the electrolyte was varied from 1 to 10 g L−1
Al2O3, whereas in the PRP experiments the particle content was
maintained constant at 10 g L−1. The cathodic peak current den-
sity (ip) was 5 A dm−2 whereas the anodic peak current density
(ian) varied between 1 and 5 A dm−2 (Table 3).
Polarization curves in electrolytes with and without particles
were recorded using a rotating disk electrode (RDE) at rota-
tion rates of 500 and 1000 rpm and a scan rate of 10 mV s−1
.
2. Experimental
Table 2
Applied pulse plating current programs
Copper/alumina and nickel/alumina nanocomposites were
deposited from an acidic copper sulfate bath and acidic nickel
sulfamate bath, respectively. The basic composition of the elec-
trolytes and the deposition parameters are given in Table 1. The
Al2O3 particles (Aeroxide Alu C, Degussa) with an average size
of 13 nm and a specific BET-surface of ∼100 m2 g−1 were used
as received without any further treatment.
Deposition experiments were performed in a three-electrode
system with a cell volume of 300 mL. The bath temperature
was maintained at 45 ◦C using a Haake thermostate (model G
toff (ms)
ton (ms)
Pulse frequency (Hz)
Duty cycle (%)
25
50
100
25
50
50
100
100
100
200
200
100
200
200
8.0
6.7
5.0
4.4
4.0
3.3
2.5
1.7
20.0
33.3
50.0
11.1
20.0
66.7
50.0
66.7
200
400
The cathodic peak current density was 5 and 10 A dm−2, respectively.

7364
D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
Table 3
Pulse-reverse plating conditions and resulting particle incorporation
ton (ms)
ic (A dm−2
)
trev (ms)
ian (A dm−2
)
Cu–Al2O3, wt% Al2O3
Ni–Al2O3, wt% Al2O3
50
100
200
400
50
100
200
400
5
5
5
5
5
5
5
5
20
20
20
20
20
20
20
20
1
1
1
1
5
5
5
5
3.7
3.5
3.7
4.1
2.7
2.8
3.5
4.0
3.1
3.0
3.2
4.8
1.2
1.4
2.4
4.2
All samples were plated at a concentration of 10 g L−1 Al2O3 in the electrolyte.
The working electrode material was a Pt disk (0.2 cm2), and the
cylindrical counter electrode was of pure copper or nickel. A
saturated calomel electrode was used as the reference electrode.
All polarization scans were performed by ramping the poten-
tial from 0 V down to −1 V (yielding a composite coating of
approximately 250 nm thickness) and then anodically to +1 V
and back to the starting potential of 0 V.
The surface morphology and microstructure of the coat-
ings were investigated using a Zeiss DSM 982 (Oberkochen,
Germany) scanning electron microscope. The samples were
embedded in epoxy resin, cut with a diamond saw, mechani-
cally grinded with 800–4000 grade silicon carbide paper and
polished with diamond solution down to 1 ␮m. Finally, the cop-
per cross-sections were etched with 0.5 M nitric acid and the
nickel ones with a mixture of 50 mL concentrated acetic acid and
50 mL concentrated nitric acid. The particle content in the films
was determined by energy dispersive X-ray analysis (EDX). The
EDX analyses were performed on the surface as well as in the
cross-section. The reported values of incorporated Al2O3 parti-
cles in the copper or nickel matrix are average values of at least
three measurements for each sample.
films ranged from ∼1.9 to ∼5.4 wt% alumina (Fig. 1a and b)
whereas the particle incorporation in the nickel coatings was
limited to amounts between ∼0.9 and ∼2.2 wt% alumina. The
maximum particle incorporation of about 5.4 wt% ␥-Al2O3 was
observed in a copper nanocomposite deposited at a duty cycle of
11% and a cathodic peak current density of 10 A dm−2 (Fig. 1b).
Compared to direct current (DC) plating of Cu–Al2O3 compos-
ites with a maximum particle content of about 3.5 wt% [13],
an obvious enhancement of particle concentration occurred due
to the application of rectangular current pulses. However, for
nickel the improvement of the codeposition is less pronounced
when going from DC plating [11] to PP (Fig. 2).
The Vickers microhardness of the films was determined using
a FischerScope HM2000 S as described in DIN EN ISO 14577.
The reported data are averaged values of 10 measurements. The
thickness of all coatings was more than ten times the maximum
indention depth of 1 ␮m ensuring that effects from the copper
substrates are negligible.
3. Results and discussion
3.1. Effect of pulse plating (PP) conditions on the particle
incorporation
Electrolytic PP codeposition of alumina particles with copper
and nickel was carried out using the current programs shown in
Table 2, where the cathodic peak current density was maintained
between 5 and 10 A dm−2, respectively.
The variation of the amount of embedded alumina nanopar-
ticles in the metal matrix with the pulse duty cycle is given in
Figs. 1 and 2 for various particle contents of the electrolyte.
In general, the particle content of the coatings increases with
decreasing duty cycle. We found that shorter on times are benefi-
cial for the particle inclusion at the same duty cycle. Depending
on the plating conditions, the particle incorporation in copper
Fig. 1. Effect of the duty cycle on the alumina content of copper-alumina
composites electrodeposited at various Al2O3 contents in the electrolyte and
ton = 200 ms. (ꢀ) 1 g L−1; (᭹) 5 g L−1; ( ) 10 g L−1. (a) iP = 5 A dm−2 and (b)
iP = 10 A dm−2
.

D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
7365
Fig. 3. Influence of the pulse frequency on the incorporation of alumina particles
Fig. 2. Effect of the duty cycle on the incorporation of alumina particles in nickel
at a pulse current density of 5 A dm−2. (ꢀ) 1 g L−1; (᭹) 5 g L−1; ( ) 10 g L−1
.
films deposited at various Al2O3 contents in the electrolyte and ton = 200 ms. (ꢀ)
(a) Cu–Al2O3 composites and (b) Ni–Al2O3 composites.
1 g L−1; (᭹) 5 g L−1; ( ) 10 g L−1. (a) iP = 5 A dm−2 and (b) iP = 10 A dm−2
.
diffusion. Finally, the particle incorporation is mainly governed
by the reduction of adsorbed metal ions on the particle surface.
In PP deposition, the reduction of free metal ions as well as
ions adsorbed on the particles takes place during ton. The rate of
metal reduction increases with the absolute value of the cathodic
peak current density. Compared to ions adsorbed on the parti-
cles free metal ions in the solution have a higher mobility. The
increasing peak current density leads to a preferential reduction
of free ions at the electrode [26]. Hence, the particle content of
the electrocodeposited composite films decreases with increas-
ing peak current density. Independent proof for the adsorption of
the metal cations comes from measurements of the zeta potential
of alumina particles in sulfate containing solutions [11,13].
The effect of the individual pulse plating parameters (particu-
larly pulse length, ton, pulse pause, toff, and peak current density,
ip) can be summarized by introducing the average plating current
density, Eq. (1).
One of the most influential parameters during electrocode-
position is the particle content of the electrolyte [2]. Regardless
to the other working conditions, an increase of the alumina con-
centration in the plating bath leads to an enhanced alumina
codeposition (Figs. 1–3). For example, the amount of particle
codeposition with nickel was enhanced by about 50% due to
an increase of the particle loading of the electrolyte from 1 to
10 g L−1 (Figs. 2 and 3). Furthermore, the particle incorporation
showed a tendency to increase with the pulse frequency (Fig. 3).
In general, the effects of duty cycle and pulse frequency can be
explained by the relationship of ton and toff. Decreasing the duty
cycle as well as increasing the pulse frequency corresponds to a
longer relaxation time, i.e. the time at which no current passes
throughthesystem. Hence, duetothelongerpausemorealumina
particles arrive at the electrode without a “concurrent” metal
deposition taking place in parallel, and as a result the particle
incorporation increases.
The particle content of the nickel nanocomposites increased
when the peak current density was decreased from 10 and
5 A dm−2 (Fig. 2). In the case of the Cu–Al2O3 composites this
tendency was less pronounced (Fig. 1). The increase of particle
content with decreasing peak current can be explained by the
model of Celis et al. which describes the particle codeposition
as a multi-step process [7]. According to this model, the parti-
cles are transferred to the electrode by means of convection and
tonip
iave
=
= θiP
(1)
ton + toff
Taking into account that the average current density increases
with the duty cycle, Θ, and peak current density, iP, one can con-
clude that the incorporation of alumina nanoparticles is favored
by low average current densities. The average current density in
PPiscomparableto theappliedcurrentdensity in DCdeposition.
Similar to the currently observed dependence, the maximum

7366
D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
duration of the cathodic and the anodic pulse is kept constant
the composites plated with a lower anodic peak current density
incorporate more alumina particles (Table 3). The amount of
dissolved metal increases with the reverse pulse current density
and the amount of dispersed particles (Fig. 4). Hence, the particle
incorporation seems to be favored by anodic cycles with little
metal dissolution. Another interesting observation is that the
alumina content of the composites is greatly enhanced at longer
cathodic pulse times. Assuming 100% current efficiency for the
anodic dissolution, 0.07 and 0.35 nm of nickel are dissolved
during one cycle at an anodic current density of 1 and 5 A dm−2
,
respectively. In the case of copper the thickness of the dissolved
metal layer during stripping cycle is comparable, due to the
similar values of atomic mass and density of both metals.
The results shown in Table 3 indicate that the amount of
codepostied particles increases with the duration of the cathodic
pulse. Furthermore, a lower anodic peak current density and
thus a smaller amount of dissolved nickel leads to an increased
particleincorporation. Thehighestincorporationachievedinthis
Fig. 4. Anodic polarization curve of as deposited nickel nanocomposites in the
acidic nickel sulfamate bath containing various amounts of alumina particles at
a rotation rate of 500 rpm. (
) 0 g L−1; (
) 1 g L−1; (
) 5 g L−1
.
amount of codeposited particles in the case of DC deposition
was often found in the range of low current densities at which the
metal ion reduction changes from charge transfer to mass trans-
port control [2,3,33–35]. A maximum incorporation of 5.4 wt%
alumina in a copper matrix was achieved at an average current
density of about 1 A dm−2 and a particle content of 10 g L−1
alumina in the electrolyte. In comparison a copper composite
film prepared under the same conditions with DC plating at a
current density of 1 A dm−2 contained only ∼0.3 wt% alumina.
study was 4.8 wt% Al2O3 in nickel deposited at iP = 5 A dm−2
,
ton = 400 ms and ian = 1 A dm−2, trev = 20 ms.
This corresponds to a deposition of about 7 nm (equivalent to
the radius of one alumina nanoparticle) followed by the dissolu-
tion of 0.07 nm nickel. One can assume that most particles that
were captured during the cathodic cycle stay in contact with the
electrode during the stripping cycle [27]. This sticking process
would be facilitated by induced polarization forces between the
particle surface and the surface of the growing composite [27].
Thus, PRP with properly chosen parameters leads to a selective
metal dissolution and an effective increase of the particle content
of the layer.
3.2. Pulse-reverse plating (PRP) of nanocomposites
As a preliminary test polarization curves in the copper and
nickel electrolytes were measured using an inert Pt rotating
disk electrode (see Section 2). Fig. 4 shows the anodic polar-
ization behavior obtained in the acidic nickel sulfamate bath.
An important result was that the anodic polarization is signifi-
cantly affected by the addition of alumina nanoparticles to the
electrolyte. In accordance with the literature [25,34], the pres-
ence of alumina nanoparticles in the electrolyte caused a shift
of the anodic dissolution peak to more cathodic potentials and
a significant increase of the anodic dissolution current density
(Fig. 4). The magnitude of both, the shift of the peak as well as
the increased anodic current density strongly depends on the alu-
mina concentration in the electrolyte and on the rotation rate of
the electrode. Thus, due to an increasing particle loading from 1
to 5 g L−1 Al2O3 the maximum stripping current increased from
∼0.8 to ∼4.9 A dm−2 (Fig. 4). This finding can be rationalized
with an reduced tendency for passivation of the nickel films and
an abrasive effect of the alumina particles in the solution dur-
ing the anodic nickel dissolution. The following discussion is
based on the hypothesis that short phases of anodic dissolution
are beneficial for the codeposition process.
It has been reported [27,28] that the maximum particle incor-
poration during PRP was found when the metal deposit thickness
approaches the particle diameter size. In the case discussed
above only half of the particle volume is incorporated during the
cathodic pulse. Hence, it might be possible to further increase
the amount of alumina incorporation by increasing the metal
layer thickness per cathodic cycle. However, increasing both iP
as well as ton resulted in deposits with a powdery appearance
which did not adhere well to the substrate.
3.3. Surface morphology and microstructural
characterization
The surface morphology of pure copper films plated with dif-
ferent deposition techniques is shown in Fig. 5. For DC plating
the surface morphology of the copper film is made up of rel-
atively large grains (Fig. 5a). In the case of PP (Fig. 5b) and
PRP (Fig. 5c) the surface morphology of the copper coating
was only slightly affected by the pulse parameters (ton, toff, ip,
trev and ian). This finding is at first glance hard to rationalize,
because one might assume that due to the application of PP and
PRP the crystal growth mode and the surface morphology of the
copper matrix change significantly [13,24]. According to the
electrocrystallization as discussed in detail by Budevski et al.,
the grain size and growth mode is defined by the interplay of
nucleation and growth [18]. In the case of PP and PRP experi-
The PRP conditions were optimized to obtain deposits with
a smooth surface morphology and are given in Table 3. The
achieved particle contents are significantly higher than those
of the composites plated with DC [11,13] and in the case of
nickel also higher than the PP plated films (Figs. 2 and 3). If the

D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
7367
Fig. 5. Surface morphology of pure copper films prepared by different deposition techniques. (a) DC, 10 A dm−2; (b) PP, θ = 67%, f = 1.67 Hz, iP = 10 A dm−2; (c)
PRP, iP = 5 A dm−2, ton = 50 ms and ian = 1 A dm−2, trev = 20 ms.
ments the number of atoms deposited during one cathodic cycle
is defined by ton and ip. Furthermore, sorption processes occur-
ring during pulse pause may influence nucleation and growth
processes in the following cathodic pulse cycle [18]. In our PP
and PRP experiments (Tables 2 and 3), the cathodic pulse cur-
rent density is comparable to the average current density in DC
plating. Finally, the cathodic overpotential during PP and PRP
is close to that obtained during DC plating. Hence, the surface
morphology of the copper film is expected to be similar for the
different deposition techniques, i.e. DC, PP and PRP (Fig. 5).
The surface morphology of the electrodeposited nickel films,
i.e. the size and the shape of the grains were significantly reduced
and altered due to both, PP conditions and incorporation of
alumina nanoparticles (Fig. 6). In the case of Ni–Al2O3 com-
posites (Fig. 6b and d) the agglomerates of alumina particles
appear as bright spots in the dark nickel matrix, as validated
by energy dispersive X-ray spectroscopy. The nickel growth
asmoothsurfacewithfinegrains(Figs. 6aandc). Thepurenickel
films plated at a higher pulse current density (Fig. 6a) and longer
pulse on-time exhibited a coarser grain structure. Higher values
of the peak current density and cathodic pulse length correlate
with an increased average current density (Eq. (1)). Hence, the
coarsening of the nickel structure with increasing current den-
sity is in good agreement with the results for nickel plated from
an acidic sulfamate bath under DC conditions [11,41].
Figs. 7–9 show cross-sectional SEM micrographs of pure
metal films and metal matrix nanocomposites plated by means
of PP and PRP. Regardless of the deposition technique, the pure
copper and nickel films exhibited a columnar structure belong-
ing to the field-oriented texture (FT) type [11,13] (Figs. 7a, 8a
and 9a). The codeposition of Al2O3 particles strongly affects the
microstructure of the metal matrix. As a result of the Al2O3 addi-
tion to the plating electrolyte the microstructure of the copper
and the nickel matrix changed to the UD (unoriented dispersion)
type (Figs. 7b, 8b and 9b) [11,13,42].
is affected by several interfacial inhibitors such as H2, Hads
,
Ni(OH)2 [20,36,37]. It is assumed that PP conditions as well as
particle incorporation perturb the nickel growth and induce an
increase of the number of nucleation sites resulting in a refined
grain structure [5,31,38,39]. Thereby the perturbation of the
nickelgrowthresultsfromachangeintheadsorption–desorption
phenomena at the nickel/electrolyte interface which is probably
caused by an local pH increase induced by an adsorption of H+
on the dispersed particles [17,36].
In the case of nickel electrodeposition from an acidic sulfa-
mate bath, a bimodal grain structure of “truncated pyramidal
type” could be observed [11,40] (Figs. 6a and c). A cluster of fine
grains surrounds relatively large grains. The PP coatings showed
An increase of the average plating current density affects cop-
per and nickel in a different way. A general observation was that
with increasing cathodic pulse current density and amount of
incorporated alumina nanoparticles a refinement of the copper
grains occurred. However, the differences in the microstructure
were more pronounced in the case of the PP plated films. The
changes of the columnar microstructure were directly related to
the refinement of the surface morphology due to an increase in
the plating current density [13,43]. In contrast to this, an increase
of the average plating current density induced an obvious coars-
ening of the field-oriented nickel structure (micrographs not
shown), which is directly related with the coarsening of the sur-

7368
D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
Fig. 6. Surface morphology of pure Ni and Ni–Al2O3 composites plated at a duty cycle of 67% and a pulse frequency of 1.67. 0 g L−1 Al2O3: (a) 10 A dm−2, (c)
5 A dm−2. 10 g L−1 Al2O3: (b) 10 A dm−2, (d) 5 A dm−2
.
face morphology (Fig. 6). Of course, SEM does not provide the
best microstructural analysis. For an extensive characterization
of the grain size distribution and the crystallographic texture
TEM and XRD investigations are needed which are currently
under way in the authors’ lab.
250 and 380 HV. Compared to direct current conditions pure
nickel films from an acidic sulfamate bath exhibited a hardness
between 330 and 360 HV [11]. Thus, in accordance with the
literature [42,44] it appears that both PP and PRP conditions
have only little impact on the hardness of nickel coatings from
thesulfamatebath. Thedependenceof themicrohardness ofpure
copper and copper alumina composites on the pulse duty cycle
is shown in Fig. 10a. The microhardness values for the copper
films range from 45 for pure copper to 85 HV for composite
films (Fig. 10a). These hardness values are in good agreement
with literature [45] where the hardness of copper films plated
3.4. Microhardness of electrodeposited nanocomposites
The microhardness of the nickel coatings plated under PP
and PRP conditions varied between 250 and 520 HV (Fig. 10b),
whereas the hardness values of pure nickel films were between
Fig. 7. Cross-sectional scanning electron micrographs of pure copper and composite films deposited by PP deposition techniques at θ = 67%, f = 1.67 Hz,
iP = 10 A dm−2. (a) Pure copper and (b) composite film, 10 g L−1 Al2O3 in the electrolyte.

D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
7369
Fig. 8. Cross-sectional scanning electron micrographs of pure copper and composite films deposited by PRP deposition techniques at iP = 5 A dm−2, ton = 50 ms and
ian = 1 A dm−2, trev = 20 ms. (a) Pure copper and (b) composite film, 10 g L−1 Al2O3 in the electrolyte.
from an acidic sulfate bath is reported to be in the range from 50
to 105 HV. Furthermore, the hardness of the PP and PRP films is
comparable to our previous results of copper films plated under
DC conditions [13].
In accordance with the literature [26] the microhardness of
pure nickel films and nanocomposites tends to increase with
decreasing duty cycle (Fig. 10b). These results are in good agree-
ment with the effects of duty cycle on the surface morphology of
the coatings (Fig. 6). It is well known that the hardness of nickel
films plated from the acidic sulfamate bath under DC condi-
tions increases with decreasing plating current density [45]. In
PP experiments the average plating current density (Eq. (1))
increases with decreasing duty cycle and increasing pulse fre-
quency. Hence, the improved hardness of the nickel films due to
a variation of the plating parameters (Fig. 10b) can be explained
by the change of the average current density, which is in good
agreement with the results of DC plating [45]. A lower duty
cycle means a relatively longer pulse pause that leads to an
enhanced incorporation of alumina particles within the metal
matrix (Figs. 1 and 2).
Finally, thehardnessofthecopperandnickelnanocomposites
is only slightly affected by the absolute amount of incorporated
alumina particles. The main parameter governing the hardness of
the coatings is the structure of the metal matrix that is principally
influenced by the average plating current density. As a result
of the particle incorporation, the microstructure of metal films
plated from acidic electrolytes changes from the FT-type (Figs.
The hardness of metal films as well as MMCs is known to
be related to the structure of the matrix and the amount and
distribution of the reinforcing metal oxides particles [9,11,13].
The results shown in Fig. 10a indicate that the shorter the pulse
off-time (i.e. the higher the duty cycle) the higher the hardness
of the copper coatings. At the first glance, it might be difficult
to rationalize this finding. Note that the amount of codeposited
particles in the copper layer decreases with increasing duty cycle
(Fig. 1). Hence, the enhanced hardness of the composites is not
directly related to the absolute amount of particle incorporation.
However, in accordance with our results from DC deposition
[13] the effect of nanoparticle codeposition on the structure
of the copper films has to be considered. The structure of the
copper film changes due to a variation of the deposition param-
eters (Figs. 7 and 8). At low peak current densities the average
grain size of pure copper films is known to decrease with a rela-
tive increase of the pulse pause [24]. Hence, the dependence of
the hardness of copper coatings on the duty cycle (Fig. 10a) is
probably related to microstructure of the copper films.
Fig. 9. Cross-sectional scanning electron micrographs of coatings prepared by PP at θ = 67%, f = 1.67 Hz, ip = 10 A dm−2. (a) Pure nickel and (b) nickel alumina
composite deposited from an electrolyte containing 10 g L−1 Al2O3.

7370
D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
The microhardness of the pure metal and the nanocompos-
ite films was characterized as a function of the interdependent
PP working conditions such as duty cycle and pulse frequency.
Similar to the results found in the case of DC deposition of metal
matrix nanocomposites, it has been shown that the hardness of
the composite films prepared by PP and PRP mainly depends on
the microstructure of the metal matrix.
In general, both PP and PRP are proved to be advantageous
for controlling and improving both the particle content and the
properties of metal matrix nanocomposites.
Acknowledgement
This work was financially supported by the “Deutsche
Forschungsgemeinschaft” (Grant BU 1200/10-1).
References
[1] P.M. Vereecken, I. Shao, P.C. Searson, J. Electrochem. Soc. 147 (2000)
2572.
[2] J.L. Stojak, J. Fransaer, J.B. Talbot, in: R.C. Alkire, D.M. Kolb (Eds.),
Advances in Electrochemical Science and Engineering, vol. 7, Wiley-VCH,
Weinheim, 2002.
[3] A. Hovestad, L.J.J. Janssen, J. Appl. Electrochem. 25 (1995) 519.
[4] J.P. Celis, J. Fransaer, Galvanotechnik 88 (1997) 2229.
[5] C.T.J. Low, R.G.A. Wills, F.C. Walsh, Surf. Coat. Technol. 201 (2006) 371.
[6] N.J. Guglielmi, Electrochem. Soc. 119 (1972) 1009.
[7] J.P. Celis, J.R. Roos, C. Buelens, J. Electrochem. Soc. 134 (1987) 1402.
[8] J.L. Fransaer, J.P. Celis, J.R. Roos, J. Electrochem. Soc. 139 (1992) 413.
[9] A.F. Zimmerman, G. Palumbo, K.T. Aust, U. Erb, Mater. Sci. Eng. A328
(2002) 137.
[10] I. Garcia, A. Conde, G. Langelaan, J. Fransaer, J.P. Celis, Corros. Sci. 45
(2003) 1173.
[11] A. Bund, D. Thiemig, Surf. Coat. Technol. 201 (2007) 7092.
[12] A.F. Zimmerman, D.G. Clark, K.T. Aust, U. Erb, Mater. Lett. 52 (2002)
85.
Fig. 10. Correlation between the vickers microhardness of pure metal and
nanocomposite films and the pulse duty cycle at a plating current density of
10 A dm−2; (ꢀ) 0 g L−1; (᭹) 1 g L−1; ( ) 5 g L−1; ( ) 10 g L−1. (a) Copper
and (b) nickel.
7a, 8a and 9a) to the UD-type (Figs. 7b, 8b and 9b), which
is known to significantly improve the hardness of the coating
[11,13].
[13] A. Bund, D. Thiemig, J. Appl. Electrochem. 37 (2007) 345.
[14] I. Garcia, J. Fransaer, J.P. Celis, Surf. Coat. Technol. 148 (2001) 171.
[15] S.C. Wang, W.C. Wei, J. Mater. Chem. Phys. 78 (2003) 574.
[16] M. Kaisheva, J. Fransaer, J. Electrochem. Soc. 151 (2004) C89.
[17] E.A. Pavlatou, M. Stroumbouli, P. Gyftou, N. Spyrellis, J. Appl. Elec-
trochem. 36 (2006) 385.
4. Summary
[18] E. Budevski, G. Staikov, W.J. Lorenz, Electrochemical Phase Formation
and Growth, VCH, Weinheim, 1996.
Copper and nickel alumina nanocomposites have been
obtained by means of DC plating, pulse plating (PP) and pulse
reverse plating (PRP). The percentage of alumina particles in the
metal coating varied from 0 to 5.6 wt% depending on the metal
matrix and the deposition technique. A maximum incorpora-
tion of about 5.6 wt% alumina was obtained in the case of PP
[19] J.C. Puippe, Schriftenreihe Galvanotechnik: Pulse-Plating, Leuze Verlag,
Bad Saulgau, 1986.
[20] C. Kollia, Z. Loizos, N. Spyrellis, Surf. Coat. Technol. 45 (1991) 155.
[21] P.T. Tang, P. Watanabe, J.E.T. Andersen, G. Bech-Nielsen, J. Appl. Elec-
trochem. 25 (1995) 347.
[22] R. Mishra, R. Balasubramaniam, Corros. Sci. 46 (2004) 3019.
[23] C.Y. Dai, Y. Pan, S. Jiang, Y.C. Zhou, Surf. Rev. Lett. 11 (2004) 433.
[24] S. Tao, D.Y. Li, Nanotechnology 17 (2006) 65.
Cu–Al2O3 composite at a peak current density of 10 A dm−1
,
[25] A.B. Vidrine, E.J. Podlaha, J. Appl. Electrochem. 31 (2001) 461.
[26] M.E. Bahrololoom, R. Sani, Surf. Coat. Technol. 192 (2005) 154.
[27] P. Xiong-Skiba, D. Engelhaupt, R. Hulguin, B. Ramsey, J. Electrochem.
Soc. 152 (2005) C571.
[28] E.J. Podlaha, D. Landoldt, J. Electrochem. Soc. 144 (1997) L200.
[29] W. Wang, F.Y. Hou, H. Wang, H.T. Gou, Scripta Mater. 53 (2005) 613.
[30] E.J. Podlaha, Nano Lett. 1 (2001) 413.
[31] B. Mu¨ller, H. Ferkel, Nanostruct. Mater. 10 (1998) 1285.
[32] J. Steinbach, H. Ferkel, Scripta Mater. 44 (2001) 1813.
[33] A.M.J. Kariapper, J. Foster, Trans. Inst. Met. Finish. 52 (1974) 87.
[34] P.R. Webb, N.L. Robertson, J. Electrochem. Soc. 141 (1994) 669.
[35] K. Helle, F. Walsh, Trans. Inst. Met. Finish. 75 (1997) 53.
a duty cyle of 20%, a pulse frequency of 8 Hz and a particle
content of 10 g L−1 alumina in the electrolyte. The alumina
content in the metal matrix increased with the particle loading
of the electrolyte, pulse frequency and decreasing duty cycle.
Although the particle content of the deposits prepared by PRP
washighercomparedtothoseofDCplating, theabsoluteamount
of particle incorporation in a copper matrix was only slightly
affected by a change in the PRP working conditions. In the case
of Ni–Al2O3 PRP seems to be a useful method to increase the
particle inclusion almost two times in comparison to DC plating.

D. Thiemig et al. / Electrochimica Acta 52 (2007) 7362–7371
7371
[36] C. Kollia, N. Spyrellis, J. Amblard, M. Froment, G. Maurin, J. Appl.
Electrochem. 20 (1990) 1025.
[41] W.-H. Lee, S.-C. Tang, K.-C. Chung, Surf. Coat. Technol. 120–121 (1999)
607.
[37] M. Stroumbouli, P. Gyftou, E.A. Pavlatou, N. Spyrellis, Surf. Coat. Technol.
195 (2005) 325.
[38] V. Medeliene, Surf. Coat. Technol. 154 (2002) 104.
[39] L. Du, B. Xu, S. Dong, H. Yang, Y. Wu, Surf. Coat. Technol. 192 (2005)
311.
[42] W. Kleinekatho¨fer, Ch.J. Raub, E. Raub, Metalloberfla¨che 36 (1982) 411.
[43] Y. Gan, D. Lee, X. Chen, J.W. Kysar, J. Eng. Mater. Technol. 127 (2005)
451.
[44] W. Kleinekatho¨fer, Ch. Raub, J. Surf. Technol. 7 (1978) 23.
[45] M. Schlesinger, M. Paunovic, Modern Electroplating, fourth ed., John
Wiley and Sons, New York, 2000.
[40] S.W. Banovic, K. Barmak, A.R. Marder, J. Mater. Sci. 33 (1998) 639.