Electrodeposition of amorphous/nanocrystalline and polycrystalline Ni-Mo alloys from pyrophosphate baths

Article abstract of DOI:10.1016/j.electacta.2004.08.028

Pyrophosphate plating bath was found to be a good alternative to citrate bath for deposition of Ni-Mo amorphous alloys. The addition of wetting agents such as 2-butyne-1,4-diol and rokafenol N-10 to the pyrophosphate bath resulted in the removal of bumps, spheres and cracks from the Ni-Mo alloy surface. The plated alloy layers adhered well to Cu-Zn brass and steel, were of thickness from a fraction to tens of micrometers and the molybdenum content was independent of the distance from the support. An increase in the concentration of the molybdate ion in the bath leads to an increase in the amount of Mo in the alloys up to 33-35 at.% and to a decrease in the deposition rate. These changes and the influence of pH are discussed in the paper. The atom arrangement in the alloys changes from (2 2 0) preferred for pure-nickel deposition to (1 1 1) for content of Mo higher than 15 at.%. For 20 and more at.% of Mo the structure of the alloy is amorphous like. An analysis of SEM and STM micrographs obtained indicates that contrary to the Ni-W alloy the amorphous phase is made of circa 10-50nm in diameter objects and not by long needles perpendicular to the substrate.

Full text of DOI:10.1016/j.electacta.2004.08.028

Electrochimica Acta 50 (2005) 1405–1412
Electrodeposition of amorphous/nanocrystalline
and polycrystalline Ni–Mo alloys from pyrophosphate baths
Mikołaj Dontena, Henrikas Cesiulisb, Zbigniew Stojeka,∗,1
a
Department of Chemistry, Warsaw University, ul. Pasteura 1, PL-02-093 Warsaw, Poland
b
Department of Physical Chemistry, Vilnius University, Naugarduko 24, LT-2006, Vilnius, Lithuania
Received 20 July 2004; received in revised form 16 August 2004; accepted 21 August 2004
Available online 3 October 2004
Abstract
Pyrophosphate plating bath was found to be a good alternative to citrate bath for deposition of Ni–Mo amorphous alloys. The ad-
dition of wetting agents such as 2-butyne-1,4-diol and rokafenol N-10 to the pyrophosphate bath resulted in the removal of bumps,
spheres and cracks from the Ni–Mo alloy surface. The plated alloy layers adhered well to Cu–Zn brass and steel, were of thickness
from a fraction to tens of micrometers and the molybdenum content was independent of the distance from the support. An increase in
the concentration of the molybdate ion in the bath leads to an increase in the amount of Mo in the alloys up to 33–35 at.% and to
a decrease in the deposition rate. These changes and the influence of pH are discussed in the paper. The atom arrangement in the al-
loys changes from (2 2 0) preferred for pure-nickel deposition to (1 1 1) for content of Mo higher than 15 at.%. For 20 and more at.%
of Mo the structure of the alloy is amorphous like. An analysis of SEM and STM micrographs obtained indicates that contrary to
the Ni–W alloy the “amorphous” phase is made of circa 10–50 nm in diameter objects and not by long needles perpendicular to the
substrate.
© 2004 Elsevier Ltd. All rights reserved.
Keywords: Electrodeposition; Ni–W alloys; Amorphous alloys; Pyrophosphate bath
1. Introduction
denum in these materials apparently increases their catalytic
activity and improves their mechanical properties. The high
electrochemical activity and relatively low cost of the Ni–Mo
alloys make them widely used in modern electrochemistry-
related industry [3–5]. Applications of the Ni–W alloys
include their use as electrodes to promote the hydrogen
evolution [3–5] and as catalysts in the hydro processing of
aromatic oils [6] and hydrogenation of benzene in the gas
phase [7].
Ni–Moalloyscanbepreparedusingseveralmethods. Met-
allurgical methods are not very convenient because of easy
oxidation at the crystallisation step and high melting temper-
ature of molybdenum. To prepare the molybdenum alloys,
other advanced processing methods such as powder metal-
lurgy and mechanical alloying [8,9], spark plasma sintering
[10], laser cladding [11] and electrodeposition are used.
Mechanical alloying offers an opportunity to prepare alloys
Nickel–molybdenum alloys are of interest due to their
useful properties. They are resistant to corrosion (also in the
corrosion media containing Cl− ions) and wear, exhibit low
thermal-expansion coefficients, are known of premium hard-
nessandpremiumcatalyticproperties, andsomeofthemhave
unusually soft magnetic characteristics [1–7]. Regarding the
resistance to corrosion, the alloy films probably cannot be
too thin to be effectively resistant. The large catalytic activity
of Ni–Mo alloys comes not only from the significant content
of nickel, a main component, but also from the unique amor-
phous internal structure of the alloy. The presence of molyb-
∗
Corresponding author. Tel.: +48 60 228 9705; fax: +48 22 822 4889.
E-mail address: stojek@chem.uw.edu.pl (Z. Stojek).
ISE member.
1
0013-4686/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2004.08.028

1406
M. Donten et al. / Electrochimica Acta 50 (2005) 1405–1412
of various structural and chemical factors that are important
for their final catalytic properties [8]. The electrodeposition
approach to the production of the Ni–W alloys has also
some apparent advantages, i.e. nanocrystalline/amorphous
materials can be obtained and there is a possibility to control
the grain size of the alloys from several to hundreds of
nanometers. In addition, thin layers of the alloys can be plated
on a surface of any shape and there is no post-processing
requirement.
2. Experimental procedure
All of the alloys where deposited using an EG&G PARC
galvanostat/potentiostat model 173A. The electrodepositions
were done in a special plating cell, with two symmetrically
placed anodic compartments, separated from the cathodic
chamber with glass frits. Two inert, passivated titanium elec-
trodes (TiO2/RuO2) were used as the anodes. The area of
each anode was circa 10 cm2. Polished Cu–Zn brass and steel
plates and silicon wafers with vapour deposited Cu (Kocour,
Chicago, and Plating Test Cell Supply Company, Cleveland,
OH) had the working area of 2 cm2 and were used as the sub-
strates for the deposits. However, all the results shown in the
figures, except for Fig. 5, were obtained with brass substrates.
Just before plating the substrate was degreased and slightly
etched and activated with dilute sulfuric acid.
For the electrodeposition of the amorphous molybdenum
alloys with the iron-group metals the citrate–ammonia baths
are used in most of the cases [12–16]. However, if the content
of Mo does not exceed 12–13 at.%, the deposits obtained in
the citrate ammonia bath consist of just one crystallographic
phase, i.e. the solid solution of Mo in nickel [13,14]. Then
the structure of the alloy is polycrystalline. When the con-
tent of Mo increases up to 18 at.%, the initial indications of
the formation of the amorphous/nanostructured phase arise.
The alloys become really amorphous/nanostructured when
the content of Mo exceeds 22–25 at.%. The electrodeposi-
tion of such a high-molybdenum-content alloys from the
citrate–ammoniasolutionispossible, however, thevery-high-
molybdenum-content alloys (63 or 38 at.% of Mo), indepen-
dently of the alloy deposition rate, exhibit cracks and small
pits which diminishes their practical importance [15].
The nickel alloys can be also deposited from other baths.
The pyrophosphate bath has been employed for the electrode-
position of nickel [17], and nickel alloys with Cu [18], Fe [19]
and Sn [20]. A possibility of electrodeposition of the Ni–Mo
alloys from pyrophosphate bath was mentioned some time
ago by Stasov and Pasechnik [21] and recently by Jovic et
al. [22]. The composition proposed in Ref. [21] enables the
deposition of the alloys under reasonable current efficiency
at room temperature with Mo content up to 40 wt.%. The
deposits are well adhered to the copper and steel substrates.
Also, the content of Mo in the alloys does not depend sub-
stantially on temperature in the range 20–40 ◦C. The Ni–Mo
alloys electrodeposited by pulse current [22] might contain
up to 41 at.% of Mo. A current-efficiency drop below 10%
accompanied that high content. Moreover, the morphology
of the deposits was particularly sensitive to the co-evolution
of hydrogen, and most of the deposits exhibited a dense net
of bumps and cracks.
Polarization experiments were done with the three-
electrode system using
a type PI-50-1.1 potentio-
stat/galvanostat, Russia, connected with a PR-8 function
generator. A platinum sheet served as the counter electrode
and a SCE was the reference electrode. The top of a copper
rod of the working area equal to 0.125 cm2 was used as the
working electrode. Before each experiment the surface of
the electrode was coated with either Ni or Ni–Mo in the
examined solution for 30 min. The ohmic drop was evaluated
in separate experiments by switching the galvanostatic pulse
off after 30 s of electrolysis and measuring the resulting
potential drop with the use of an oscilloscope.
The solution with the main components at the
concentrations such as those reported for the elec-
trodeposition of molybdenum alloys by Stasov and
Pasechnik [21] was the starting point in this work.
Stasov’s solution is called the initial solution through-
out this paper and its composition can be writ-
ten as: Na4P2O7·10H2O—160 g dm−3, NH4Cl—20 g dm−3
,
NiSO4·7H2O—40g dm−3 and Na2MoO4·2H2O. The sug-
gested pH value for the initial solution is 8.5. pH of the final
solutionswasadjustedbyaddingappropriateamountsof30%
ammonium solution. In the course of the work the initial solu-
tion was changed by adding 2-butyne-1,4-diol—50 mg dm−3
and rokafenol N-10 (nonionic detergent)—100 ␮l dm−3, and
by varying the concentration of Na2MoO4 in the bath. All so-
lutions were prepared using deionized water produced by a
NanoPure Milli-Q purification system, Millipore. The alloys
were deposited at 20 ◦C.
The aim of this work was to modify the pyrophosphate
bath so that thin layers of Ni–Mo alloys of good appearance,
integrity and smoothness could be obtained. Such smooth
layers are advantageous from e.g. trybological point of view.
Theaboveaimhasbeenachievedbyadding2-butyne-1,4-diol
and rokafenol N-10 (polyoxyethylene–phenol class nonionic
detergent) to the bath, which compounds were used success-
fully in the electrodeposition of tungsten alloys [23–25]. The
pyrophosphate bath should be more resistive to degradation
(oxidation)comparedtocitrates. Theinfluenceofvariousfac-
tors on the current efficiency, the deposition rate, the molyb-
denum content and the structure of the alloys have been ex-
amined.
Compositions of the deposits were examined using a
Roentec, model M1, EDX analyzer (Germany) integrated
with a LEO, model 435 VP, scanning electron microscope
(SEM). The mean from the multiple EDX composition data,
the weight of the alloy layer and the assumption that six and
two electrons are used in the reduction of the Mo and Ni ions,
respectively, allowed the calculation of the current efficiency.
The deposition rates have been calculated from the difference
in the weight of the substrates after and before deposition.
A NanoScope scanning tunneling microscope (STM) was
used to observe the alloy surfaces and to estimate their

M. Donten et al. / Electrochimica Acta 50 (2005) 1405–1412
1407
smoothness. In the STM study, the tunneling current was set
in the range up to 1 nA and the bias voltage up to 300 mV.
The scanning rate was varied from 0.9 to 2.0 Hz. The STM
data obtained were further computer processed to create 3D
images and to perform a roughness analysis.
The structure of the deposits was determined by the X-
ray diffraction method using a DRON-1 instrument equipped
˚
with a Fe X-ray tube (K␣: λ = 1.9373 A).
3. Results and discussions
3.1. Influence of electrodeposition conditions on
composition anddeposition rates of Ni–Mo alloys
The Ni–Mo alloys electrodeposited from the bath corre-
sponding to the initial solution are rather rough due to the
presence of many micro-sphere fragments on the surface.
The introduction of the brightening and wetting agents into
the baths makes the surface much smoother and eliminates
the cracks in the alloy layer. The bumps and spheres virtu-
ally disappear from the surface. This is illustrated in Fig. 1A
and B. The edges of the covered support plates (not shown in
Fig. 1) are also reasonably smooth and well defined. The ad-
ditives used together lower the surface tension, adsorb on the
alloy surface, decrease the hydrogen evolution at a given Mo
content, keep the seed size small and apparently help to keep
the deposition rate at similar level throughout the surface.
The dependencies of the molybdenum content and the de-
position rate versus Na2MoO4 concentration for the solutions
with and without 2-butyne-1,4-diol and rokafenol are shown
in Fig. 2. Apparently, the amount of Mo in the alloys in-
creases up to 33–35 at.% with increasing the concentration
of the molybdate ion in the bath. This increase is accompa-
nied by a decrease in the deposition rate. The decrease in the
deposition rate (expressed in mg h−1 cm−2) is a result of the
decrease in the current efficiency from 50–57% to 2–10% (pH
8.5; j = 30 mA cm−2). The primary cause of these changes is
an enhancement in the catalytic properties of nickel towards
the evolution of hydrogen, which takes place after adding
molybdenum to nickel. Furthermore, for a fixed concentra-
tion of the molybdate ions in the solution, the content of Mo
in the alloys is smaller by up to 7% when the deposition is
carried out in the presence of the brightening and wetting
agents.
Fig. 1. SEM images of Ni–Mo alloy (84:16 at.%) deposits on iron covered
with Zn–Cu brass layers obtained with j = 30 mA cm−2 from solutions. (A)
Initial solution with 4 g dm−3 of Na2MoO4·2H2O, as proposed in [21]. (B)
Initial solution (as is A) containing also 2-butyne-1,4-diol (50 mg dm−3) and
rokafenol (100 ␮l dm−3).
The deposition rate is very low at low current densities
applied while the Mo content in the formed Ni–Mo alloy
exceeds 30 at.%. This is illustrated in Fig. 3. In this case,
i.e. at the potentials corresponding to low current densities,
molybdenum is deposited preferentially [21]. Since the ex-
change current density for the hydrogen evolution for Ni–Mo
alloys in the alkaline solutions is relatively high and can
reach the values up to 11 mA cm−2 [26], the hydrogen evo-
lution becomes the major electrode process. Therefore, the
alloy deposition rate decreases. With an increase in current
density, the cathode potential becomes more negative, the
Fig. 2. Influence of Na2MoO4 concentration in the bath on Mo content in
Ni–Mo alloys (1, 2; left axis) and on deposition rate of alloys (1ꢀ, 2ꢀ; right
axis). Bathscomposition:1ꢀ, 2ꢀ—thatofinitialsolution;1, 2—initialsolution
with 2-butyne-1,4-diol (50 mg dm−3) and rokafenol (100 ␮l dm−3). Current
density: 30 mA cm−2. The arrows indicate corresponding ordinate axis.

1408
M. Donten et al. / Electrochimica Acta 50 (2005) 1405–1412
Fig. 3. Effect of current density on Mo content in Ni–Mo alloys (1, left axis,
filled circles) and on current efficiency (2, right axis, open circles). Bath
composition: initial solution containing also 6 g dm−3 of Na2MoO4·2H2O,
2-butyne-1,4-diol (50 mg dm−3) and rokafenol (100 ␮l dm−3).
Fig. 4. Calculated percentage fractions of all ammonia (A) and pyrophos-
phate (B) nickel complexes in total Ni(II) plotted as functions of pH. Solution
composition: 0.36 M Na4P2O7 + 0.14 M NiSO4 + 0.374 M NH4Cl.
electroreduction of Ni(II) accelerates and the contents of
molybdenum in the alloys decreases. This causes an increase
in the overvoltage for the hydrogen evolution and correspond-
ingly an increase in the deposition rate.
phate is the key intermediate species. For the deposition of
the Ni–Mo alloys in the citrate bath it has also been sug-
gested that the electroreduction of MoO4 to Mo proceeds
2−
pH influences the Mo content in the alloys and the depo-
sition rate. Some examples are presented in Table 1. As it
follows from these data, the content of Mo increases with de-
creasing pH in the range 8.5–10, while the deposition rate and
the current efficiency decrease. The increase of pH above 10
neither affects substantially the deposition rate nor the current
efficiency.
in several steps, one of which is the formation of an intermed−i-
ate species, probably MoO2 [29,30] or (NiCitMoO2H3)ads
[15].
The pH effect on the deposition rate and the Mo content in
the alloys is not easy to explain either. This is just because the
electrodeposition is accomplished by several conjugated and
pH-dependent reactions. After taking into account the forma-
tion of various Ni(II) complexes in the pyrophosphate baths,
the Nernst potentials for both Ni and Mo couples can be esti-
mated as circa −0.6 V. So, as it can be seen from the polariza-
tion curves presented in Fig. 5, the electrodeposition of both
Ni and Ni–Mo alloy occurs under kinetic control (with some
overpotential). A similar effect was also observed during the
The relation between the contents of Mo and pH is strik-
ingly similar to the dependence of the partial percentage
fractions of Ni pyrophosphate species and the Ni–ammonia
complexes on pH. These percentage fractions have been cal-
culated using the available formation constants for all py-
rophosphate and ammonia complexes, and are plotted versus
pH in Fig. 4. In the pH range from 8.5 to 10 a substantial
drop in the fraction of the Ni–pyrophosphate complexes and
an increase in the fraction of the Ni–ammonia complexes
take place. The absolutely dominating species in the bath is
NiP2O72− and, within the ammonia complexes, Ni(NH3)4
2+
and Ni(NH3)32+. It is very probable that by analogy to the de-
position of Ni–W alloys in the citrate bath [27,28], for which
theimportanceofabimetalliccomplex[(Ni(WO4)Cit)(H)]2−
in the deposition process has been shown, in the pyrophospate
bath, a bimetallic complex of Ni, molybdate and pyrophos-
Table 1
Mo contents, deposition rate and current efficiency for Ni–Mo al-
loys deposited at various pH from initial solution containing addition-
ally 12 g dm−3 of Na2MoO4·2H2O, 2-butyne-1,4-diol (50 mg dm−3) and
rokafenol (100 ␮l dm−3), j = 30 mA cm−2
pH
Contents of
Mo (at.%)
Deposition rate
Current
efficiency (%)
(mg h−1 cm−2
)
Fig. 5. Galvanostatic polarization curves for Ni (1) and Ni–Mo alloy on
copper (2) electrodes in the initial solution containing also 2-butyne-1,4-diol
(50 mg dm−3), rokafenol (100 ␮l dm−3), and Na2MoO4·2H2O at concentra-
tion 0 g dm−3 (1) and 10 g dm−3 (2) (pH 8.5).
8.5
9.0
9.75
29.7
22.1
19.7
9.1
15.6
14.4
16.5
32.7
31.5

M. Donten et al. / Electrochimica Acta 50 (2005) 1405–1412
1409
Table 2
The deposited Ni–Mo alloys adhere well to copper-, brass-
and iron substrates. The deposits can be folded without ev-
ident peeling. Only some cracks may appear in the folding
area. This is illustrated in Fig. 6.
Mo content at various distances from the substrate
Distance from substrate
Mo content (at.%)
Sample 1
Sample 2
15–25 ␮m, top of layer
1 ␮m
Close to substrate
12.9 0.5
13.2 0.5
10.7 0.7
23.3 1.1
–
24.0 0.7
3.2. X-ray diffraction and SEM analysis
Alloy deposited at current density of 30 mA cm−2 from initial solu-
A brief analysis of the X-ray diffractograms of the elec-
trodeposits obtained from the plating pyrophosphate bath
leads to a conclusion that an addition of molybdenum to the
deposit changes its texture. The atom arrangement changes
from (2 2 0) preferred for pure-nickel deposits to (1 1 1) for
the alloys containing 15 and more at.% of Mo. Fig. 7 presents
selected diffraction spectra obtained for the deposits plated
from the pyrophosphate bath. The structure of the elec-
trodeposited Ni–Mo alloy becomes amorphous-like (the ap-
propriate widening of the diffraction peaks appears) when
the amount of molybdenum exceeds 20 at.%. Characteristic
changes in the patterns of the fracture cross-sections obtained
for various deposits are shown in Fig. 8. Apparently, the struc-
ture of the deposits transforms from that typically columnar
tion containing 2-butyne-1,4-diol (50 mg dm−3), rokafenol (100 ␮l dm−3
)
and Na2MoO4·2H2O at concentration 4 g dm−3 (sample 1, pH 9.23) and
9 g dm−3 (sample 2, pH 8.75).
Ni–Mo electrodeposition from the citrate–ammonium baths
[13]. Apparently, the hydrogen evolution process is faster at
the Ni–Mo alloy compared to that at the Ni electrode. An
increase in pH results in decreasing the current related to the
hydrogen evolution and, correspondingly, in increasing the
alloy deposition rate.
The Mo content in the alloy was determined at several
points on the surface of the fracture cross-section. Three lo-
cations are selected for Table 2: (a) the position as close as
possible to the copper substrate; (b) that at the distance of
circa 1 ␮m from the substrate; (c) on the top of the deposit.
As it follows from the data presented in Table 2, the content
of Mo in the alloys is almost uniform and independent of the
film thickness. A small decrease (not shown in the table) in
the Mo content from 23.3 to 21.8 at.% is obtained at the sub-
strate edges (the thickness of the films examined was circa
40 ␮m). This decrease is rather caused by the increase in the
current density at the edges (diffusional edge effect).
Fig. 7. X-ray diffractograms obtained for Ni and Ni–Mo alloys plated from
pyrophosphate baths. Mo content: 6.9 at.% (A), 15.6 at.% (B), 21.7 at.% (C)
and 23.9 at.% (D). Current density: 30 mA cm−2
.
Fig. 6. SEM micrograph of a folded sheet coated by Ni–Mo alloy (13 at.%
of Mo). Thickness of deposit: circa 15 ␮m. Current density: 30 mA cm−2
.

1410
M. Donten et al. / Electrochimica Acta 50 (2005) 1405–1412
Fig. 8. SEM images of fracture cross-sections of thin (several ␮m) layers of Ni and various Ni–Mo alloys deposited from pyrophosphate bath. Current density:
30 mA cm−2: (A) Ni; (B) Ni–Mo alloy with 6.23 at.% of Mo; (C) Ni–Mo alloy with 23.89 at.% of Mo.
observed for pure Ni to a uniform-like for the deposits with
a high Mo content. The most pronounced irregularities in the
fracture cross-section structure, seen as multilayer fracture,
start at the Mo content of 6–8 at.%. This is probably related
to the beginning of the transformation of the deposit texture
from (2 0 0) to (1 1 1). The deposits with Mo content higher
than 15 at.% have an almost uniform appearance of their frac-
ture cross-sections. However, it is possible to observe slight
irregularities in their structure along the direction perpen-
dicular to the substrate. Namely, the first circa 5 ␮m of the
alloy layer (those located next to the substrate) give cracks
slightly irregular, while the rest of the layer brakes very uni-
formly. This difference in the morphology of the deposit can
be caused bythe changesinthe orientation of the nanocrystals
in the internal and the external layers or by the transformation
of the deposits from nanocrystalline to glass-alike structure.
Interestingly, the above differences in the morphology of the
cross-brakes do not result in significant changes in the X-ray
diffraction patterns.
tion. The mean crystal sizes determined from the peak width
using the Scherrer equation modified by Warren and Biscoe
are 93 and 113 A for the 7 and 30 ␮m deposits, respectively.
It suggests that the changes in the fracture cross-section pat-
terns seen in Fig. 8 are caused by the reasons independent of
the nanocrystal size.
˚
Some changes in the crystal grain size can be observed for
the deposits with various Mo content: the lower the molyb-
denum content the larger are the crystals. The calculated
grain sizes for the alloys of various Mo content are shown
in Table 3. It is clear that with increase in molybdenum con-
tent the Ni–Mo alloy changes its structure from micro- to
nanocrystalline. The change in the 2θ value for the signal
placed in the position corresponding to the Ni (1 1 1) plane
(2θ = 56.88◦, referred to Fe K␣) gives additional information
regarding the deposited alloys. The changes in the d(1 1 1)
distance plotted versus molybdenum content in the alloy are
shown in Fig. 9. This distance increases linearly in the range
of Mo content up to 15 at.%. Such a phenomenon is typical
for the formation of solid solution of two mixed metals. At
the Mo content of 16–20 at.% d reaches a maximum and then
starts to decrease. This may be related to the formation of
The diffractograms obtained for both 7 and 30 ␮m thick
deposits exhibit the shapes typical for the nanostructured de-
posits and contain one broad peak placed in the same 2θ posi-

M. Donten et al. / Electrochimica Acta 50 (2005) 1405–1412
1411
Table 3
3.3. STM examination
Estimated grain size and d(1 1 1) distance obtained for Ni and Ni–Mo de-
posits plated from the pyrophosphate bath
The surface roughness of the Ni–Mo alloys was examined
by STM. The growing Ni–Mo alloy forms a fine structure
already at the early stages. When the thickness of the film
increases, the surface becomes more uniform and smoother,
and the characteristic pattern of the substrate is eliminated.
This can be seen in Fig. 10A and B. When the Mo content
in the alloys does not exceed 25–30 at.% and the thickness
of the film is in the range from 0.25 to 12 ␮m, the smallest
roughness of electrodeposited Ni–Mo alloys is 1–3 nm (on
a 5 ␮m × 5 ␮m area scanned). The thin, perpendicular to
the surface, crystallites have the thickness of circa 10–50 nm,
however, their length is limited. This conclusion can be drawn
after comparing the STM image in Fig. 10B with the SEM
micrograph of the fracture cross section given in Fig. 8C.
Apparently, the long, nanometer thick needles seen in the
amorphous/nanocrystalline Ni–W alloys [32] are not formed
in the Ni–Mo alloy.
˚
Mo
content
(at.%)
Estimated grain
(1 1 1) maximal
intensity 2θ
d(1 1 1) (A)
˚
size (A)
0
6.9
8.5
15.6
19
21.4
21.7
23.9
24.8
415
313
280
304
168
113
94
56.84
56.44
56.24
55.88
55.92
56.04
56.04
55.90
56.20
2.0353
2.0485
2.0552
2.0673
2.0660
2.0619
2.0619
2.0667
2.0565
77
61
either nanocrystals of pure Mo that cannot be dissolved in Ni
or of disturbed structure of known intermetallic Ni4Mo com-
pound. Another evidence for the transformation phenomenon
is the parallel decrease in the estimated crystal size observed
in the same region of Mo content.
When the Mo content exceeds 20 at.% the alloy looks
smoother in its fracture cross-section images and gives wide
X-ray diffractograms, which proves that the deposit is more
uniform, probably nanostructured. The results obtained in
this work agree well with the theoretical calculations done by
Zhang et al. [31], who have determined, by using the molec-
ular dynamics simulation, that the glass forming range of the
Ni–Mo binary system starts at 21 at.% of Mo in Ni.
As we have found, similarly to the increase in the molyb-
denum content, the codeposition of phosphorus (reduction of
hypophosphates) with nickel and molybdenum results in a
nanostructured material. For example, an addition of 8 at.%
of phosphorus to the Ni–Mo alloy of 6:1 ratio decreases the
˚
grain size from 300 to 45 A. The addition of phosphorus leads
also to significant smoothing of the fracture cross-section
structure of the layers.
Fig. 10. STM surface images of: brass substrate (A), and Ni–Mo alloy
(16 at.% Mo) electrodeposited from initial solution containing 4 g dm−3 of
Fig. 9. Changes in mean d(1 1 1) distance in Ni–Mo alloy deposits contain-
ing increasing amount of Mo in nickel. Points are experimental, line is a
polynomial cubic function fitted to the points.
Na2MoO4·2H2O, 1,4-butyndiol (50 mg dm−3) and rokafenol (100 ␮l dm−3
)
at current density of 30 mA cm−2 (B). Thickness of coating: circa
12 ␮m.

1412
M. Donten et al. / Electrochimica Acta 50 (2005) 1405–1412
[9] D. Oleszak, V.K. Portnoy, H. Matyja, Mater. Sci. Forum 312 (4)
4. Conclusions
(1999) 345.
[10] S.D. De la Torre, D. Oleszak, A. Kakitsuji, K. Miyamoto, H.
Miyamoto, R. Martinezm-S, F. Almeraya-C, A. Martinez-V, D. Rios-
J, Mater. Sci. Eng. A 276 (2000) 226.
[11] G.L. Goswami, S. Kumar, R. Galun, B.L. Mordike, Lasers Eng. 13
(2003) 1.
Similarly to the deposition of Ni–W alloys from the citrate
bath, the addition of wetting agents such as 2-butyne-1,4-diol
and rokafenol N-10 to the pyrophosphate bath resulted in the
removal of bumps, spheres and cracks from the Ni–Mo alloy
surface.
[12] L.S. Sanches, S.H. Domingues, C.E.B. Marino, L.H. Mascaro, Elec-
trochem. Commun. 6 (2004) 543.
On the other hand, there is no similarity between the fi-
brous, nanocrystalline structure of the Ni–W alloys, where
the needles grow perpendicular to the support surface, and
the structure of the Ni–Mo amorphous alloys. For the Ni–Mo
system, the nanocrystallites are rather oval. The amor-
phous/nanocrystalline structure appears for the deposits of
Mo content higher than circa 20 at.%. In addition, at that con-
tent of Mo, the transformation of the solid solution of Mo in
Ni into either the deformed structure of Ni4Mo or a mixture
of the saturated Ni–Mo solid solution with nanostructured Ni
takes place.
[13] L.I. Stepanova, O.G. Purovskaja, V.V. Sviridov, Russ. J. Appl. Chem.
73 (2000) 66.
[14] L.I. Stepanova, O.G. Purovskaja, V.N. Azarko, V.V. Sviridov, Pro-
ceedings of the Academy of Sciences of Belarus, Ser. Chem. Sci.
(1997) 38 (in Russian).
[15] E.J. Podlaha, D. Landolt, J. Electrochem. Soc. 143 (1996)
885.
[16] E.J. Podlaha, D. Landolt, J. Electrochem. Soc. 143 (1996)
899.
[17] V.A. Purin, Metal electrodeposition from pyrophosphate baths, Riga
(1975) 194 (in Russian).
[18] M. Ishikawa, H. Enomoto, M. Matsuoka, Ch. Iwakura, Electrochim.
Acta 40 (1995) 1663.
[19] A. Ramachandran, C.N. Tharamani, S.M. Mayanna, Trans. Inst. Met.
Finish. 79 (2001) 195.
Acknowledgement
[20] V.V. Orekhova, I.D. Roi, Sov. Electrochem. 27 (1991) 638.
[21] A.A. Stasov, S.A. Pasechnik, Izv. Vysh. Uch. Zav., Ser. Chem. and
Chem. Technol. 16 (1973) 600 (in Russian).
[22] V.D. Jovic, B.M. Jovic, G.R. Stafford, AESF Proceedings, 2003.
[23] M. Donten, Z. Stojek, Polish J. Chem. 68 (1994) 1193.
[24] M. Donten, H. Cesiulis, Z. Stojek, Electrochim. Acta 45 (2000)
3389.
HC thanks the Polish Foundation Kasa Mianowskiego for
a fellowship.
References
[25] H. Cesiulis, A. Baltutiene, M. Donten, M.L. Donten, Z. Stojek, J.
Solid State Electrochem. 6 (2002) 237.
[1] D. Landolt, Plating Surf. Finish. 88 (2001) 70.
[2] T.E. Tsenta, V.M. Knyazheva, T.V. Svistunova, Y.M. Kolotyrkin, D.S.
Zakharin, Prot. Met. 25 (1989) 28.
[3] C.-C. Hu, C.-Y. Weng, J. Appl. Electrochem. 30 (2000) 499.
[4] W.K. Hu, X.J. Cao, F.P. Wang, Y.S. Zhang, Int. J. Hydrogen Energy
22 (1997) 621.
[5] C. Fan, D.L. Piron, A. Steb, P. Paradis, J. Electrochem. Soc. 141
(1994) 382.
[6] J.F. Kriz, H. Shimada, Y. Yoshimura, N. Matsubayashi, A. Nishijima,
Fuel 74 (1995) 1852.
[26] S.D. De la Torre, D. Oleszak, F. Almeraya, A. Martinez, R. Martinez,
D. Rios, H. Miyamoto, Mater. Sci. Forum 343 (3) (2000) 855.
[27] O. Younes-Metzler, L. Zhu, E. Gileadi, Electrochim. Acta 48 (2003)
2551.
[28] O. Younes-Metzler, E. Gileadi, J. Electrochem. Soc. 149 (2002) 100.
[29] S. Rengakuji, Y. Nakamura, M. Inoue, K. Nishibe, H. Imanaga,
Denki Kagaku 59 (1991) 885.
[30] Y. Zeng, M. Ma, X.M. Xiao, Z.L. Li, S.X. Lian, S.M. Zhou, Chin.
J. Chem. 18 (2000) 29.
[31] Q. Zhang, Z.C. Li, C. Lin, B.X. Liu, E. Ma, J. Appl. Phys. 87 (2000)
4147.
[7] M.P. Astier, G. Dji, S.J. Teichner, Appl. Catal. 72 (1991) 321.
[8] P. Kedzierzawski, D. Oleszak, M. Janik-Czachor, Mater. Sci. Eng.
A 300 (2001) 105.
[32] M. Donten, Z. Stojek, H. Cesiulis, J. Electrochem. Soc. 150 (2003)
95.