Evaluation of Pd electroless films ability to be used as H-permeable anodes

Article abstract of DOI:10.1016/j.electacta.2008.06.022

The prospective ability of using thin Pd electroless films made of pure Pd and Pd-P alloys as hydrogen permeable anodes was evaluated by a comparative characterization of these materials. The morphology, structure and composition effect on their ability to adsorb and oxidize hydrogen in 0.1 M NaOH solution was investigated. The results revealed that pure Pd (prepared at low T) followed by Pd-P alloy (low P content) exhibit the highest activity to absorb and oxidize hydrogen. It was concluded that the Pd alloy amorphicity (brought by a low content of P) and the pure Pd core porosity (brought by a low deposition rate) are the main factors responsible for such behaviour. In acid medium, it was found that Pd alloy films hinder the hydrogen absorption.

Full text of DOI:10.1016/j.electacta.2008.06.022

Electrochimica Acta 53 (2008) 8138–8143
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Evaluation of Pd electroless films ability to be used as H-permeable anodes
M. Cristina F. Oliveira∗
Centro de Química de Vila Real, Departamento de Química, Universidade de Trás-os-Montes e Alto Douro, Quinta dos Prados, Apartado 1013, Vila Real, Portugal
a r t i c l e i n f o
a b s t r a c t
Article history:
The prospective ability of using thin Pd electroless films made of pure Pd and Pd–P alloys as hydrogen
permeable anodes was evaluated by a comparative characterization of these materials. The morphology,
structure and composition effect on their ability to adsorb and oxidize hydrogen in 0.1 M NaOH solution
was investigated. The results revealed that pure Pd (prepared at low T) followed by Pd–P alloy (low P
content) exhibit the highest activity to absorb and oxidize hydrogen. It was concluded that the Pd alloy
amorphicity (brought by a low content of P) and the pure Pd core porosity (brought by a low deposition
rate) are the main factors responsible for such behaviour. In acid medium, it was found that Pd alloy films
hinder the hydrogen absorption.
Received 26 March 2008
Received in revised form 4 June 2008
Accepted 7 June 2008
Available online 22 June 2008
Keywords:
Pd electroless
Pd alloys
H-permeable anodes
Pd membranes
Pd–P
© 2008 Elsevier Ltd. All rights reserved.
1. Introduction
hypophosphite ion is less toxic and gives rise to the formation of a Pd
alloy (phosphorous is inevitably incorporated in the metal bulk dur-
The use of palladium-based membranes in hydrogen separa-
tion/purification devices is broadly known but in contrast, the
use of this material as a hydrogen permeable anode has not been
much explored. Its application is confined to some hydrogenation
reactions [1] and quite few fuel cells. The applicability of hydrogen-
permeable anodes in hydrogen-air fuel cells have been demon-
strated in alkaline [2,3] and proton conducting fuel cells (PCFCs)
[4,5], on which a thick commercial Pd foil (25 ␮m) has been used
as the Pd membrane. The employment of this material, even when a
Pd-black film is deposited on it, presents several drawbacks: mate-
rial cost, low hydrogen flux and mechanical instability due to hydro-
gen embrittlement. Reduction of the Pd thickness is an attractive
approach to solve some of the above problems as it would increase
the H2 flux and reduce the cost effectiveness of the membrane.
According to the literature, Pd membranes can be made of pure
palladium or can be made of a Pd alloy (e.g. Pd–Ag) [6]. Pd alloys
present better mechanical properties than pure Pd because phase
transitions from ␣ to ␤-Pd hydrides rest below room temperature
which means that hydrogen embrittlement can be avoided. Among
the most used deposition methods to prepare Pd membranes, elec-
troless plating has been the most popular one because of its high
simplicity. Several reducing agents are available for electroless
deposition, however hydrazine has been the most used one on the
preparation of Pd electroless membranes [7–11]. Nonetheless, the
ing the metal deposition process giving rise to a Pd–P alloy). Despite
these apparent advantages, literature reports on the preparation of
Pd-electroless membranes from a hypophosphite ion solution, are
very scarce [12,13].
In this work, the prospective ability of thin Pd-electroless films
made of pure Pd and Pd–P alloys, to be used as hydrogen-permeable
anodes (exhibiting no pinholes and displaying a high ability to
absorb and to oxidize hydrogen) will be evaluated by a com-
parative characterization of these materials. This characterization
will be made in terms of analysing their structure, morphology,
composition and electrochemical behaviour in alkaline medium
within a potential domain suitable for hydrogen absorption and
oxidation. The permeability of the prepared films will not be
appraised because experiments will be performed under imperme-
able boundary conditions (smooth metal substrates will be used as
the film support). A porous support was not employed at this stage
in order to allow the roughness factor comparison of the different
prepared materials.
2. Experimental
2.1. Preparation of Pd-electroless films
Four different Pd-electroless membranes were prepared: two of
pure Pd and two of a Pd–P alloy. Pure-Pd membranes were assigned
as Pd (low T) and Pd (high T) because the plating conditions for
their preparation differed mainly on the deposition temperature.
Despite preparation of Pd (low T) membrane required a reducing
∗
Tel.: +351 259350286; fax: +351 259350480.
E-mail address: mcris@utad.pt.
0013-4686/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2008.06.022

M.C.F. Oliveira / Electrochimica Acta 53 (2008) 8138–8143
8139
Table 1
Electroless baths composition and deposition conditions of Pd pure and Pd alloy films on Ni and Cu substrates, respectively
Pd film
Reducing agent (mM)
Pd2+ (mM)
NH3 (in H2O) at 25% (v/v) (cm3)
EDTA (M)
Tdep (◦C)
tdep (min)
Reference
Pd (low T)
Pd (high T)
Pd alloy (low P)
Pd alloy (high P)
27.0
28.0
20.0
10.0
10.0
30
23
11.2
8.8
0.11
0.11
0.01
0.29 M NH4Cl
25
50
50
50
300
50
90
[10]
[11]
[12]
[13]
N2H4
12.0
1.0
10.0
−
H2PO2
90
agent concentration that was approximately double than for depo-
sition of Pd (high T), it was found that on using a lower hydrazine
concentration solution (10 mM), a homogeneous Pd film was not
regularly obtained. Concerning the Pd–P alloys, these membranes
were assigned as Pd alloy (low P) and Pd alloy (high P) because
the main difference on the electroless plating composition was the
hypophosphite ion concentration, and therefore the P content of
the alloy.
The films were prepared using electroless plating bath compo-
sitions and plating conditions described in the literature (Table 1).
Excepting Pd alloy (high P), all of these films have been reported
for H2 purification/separation devices. Time deposition was moni-
tored upon the metal substrate immersion into the plating solution.
Pure Pd and Pd alloy films were deposited on smooth metal discs
of nickel (ꢀ = 5 mm) and copper (ꢀ = 9.1 mm), respectively, which
were previously inserted, under pressure, into a Teflon holder and
polished to mirror finishing with successively finer grades of alu-
mina, down to 0.3 ␮m. This electrode assembly allowed deposition
of Pd film just on one face of the disc and its removal by attack of
acid (for electrothermal analysis) without damage of the electrode
support.
3. Results and discussion
3.1. Characterization of the morphology, composition and
structure of Pd-electroless films
Scanning electron microscopy was carried out to determine the
Pd film thickness and assess the film morphology and uniformity.
Fig. 1 shows that the thickness of the Pd-black film is extremely
uniform (which is a characteristic of electroless deposition) and
approximately the same on the different prepared materials (it
varies from 2.2 to 2.8 ␮m). Regarding pure-Pd films, spherical par-
ticles trapped on the surface were found on Pd (high T) film. Some
of these were not well fixed to the surface, leaving holes behind
it. The formation of these spherical particles is probably related to
an insufficient stability of the deposition bath. It may be generated
within the solution due to its self-decomposition, falling down on
the growing film. Looking at the cross-section of the films it was
observed that both Pd films were dense and compact at the sur-
face, but a porous and spongy structure was found on Pd (low T)
core. This remarkable structure seems to be the end result of a very
low deposition rate.
Concerning Pd alloys films, these also revealed to be dense and
compact, either at the surface, either in the core. However, Pd alloy
(low P) displays some small pinholes on the surface, leaving the
copper substrate uncovered on some surfaces sites. EDS analysis
showed that the high P content film contains about 9.5 at.% of P
and the lower P content film contains about 4.0 at.% of P.
2.2. Characterization of Pd-electroless films
Structural analysis of the films was carried out in a Philips X’Pert
diffractometer by X-ray diffraction using Cu K␣ radiation. The iden-
tification of crystalline phases was done using the JCPDS database
cards. On the SEM/EDS analysis a Philips-FEI Quanta 400 micro-
scope was employed. Atomic force microscopy was used for the
topographic characterization of the Pd-black surface. The mea-
surements were performed in a Nanoscope IIIa Multimode AFM
Microscope (Digital Instruments, Veeco) in tapping mode using
etched silicon probes (RTESP7 NanoprobeTM, Digital Instruments)
with a resonance frequency of about 300 kHz.
Atomic force microscopy allowed looking in more detail the sur-
face morphology of these films, particularly, in a region free of the
high-dimension particles trapped on the surface (Fig. 2). It is clear
that Pd (low T) presents the most irregular surface, displaying a
wide range of particles dimension and the highest rms roughness
(Rq). In contrast, Pd alloy (high P) shows Pd particles with the most
uniform grain size and the smoothest surface. Curiously, the same
trend has not been found on evaluating the roughness factor by
electrochemical means [14] (Table 2). This discrepancy seems to
reveal that the surface area is not only dependent on the average
particles height (which determines the rms roughness), but is may
also be ruled by the particles density on the surface. The roughness
factor (rf) was not calculated for Pd alloy (high P) because it was
found that PdO reduction peak overlapped with a non-identified
cathodic peak at rather positive anodic limit potentials (whereas
the anodic dissolution of P probably occurs).
The electrochemical instrumentation consisted of an Autolab
(model 100) potentiostat/galvanostat. Experiments were per-
formed in one compartment cell with a Pt flag and a saturated
calomel electrode (to which the potential is referred) as the counter
and reference electrode, respectively. In the cyclic voltammetric
experiments the potential was initially scanned from the anodic
limit potential in the cathodic direction, at a scan rate of 10 mV s−1
.
The solutions were deaerated with N2 and an oxygen-free nitrogen
atmosphere was kept in the cell during the measurements.
The typical hydrogen loading procedure consisted in applying
a rather negative electrode potential (typically −1.3 or −1.4 V) for
variable time in the 0.1 M NaOH solution. A fresh NaOH solution
was prepared every day from high purity water (Millipore system)
and 99.998% pure NaOH (Aldrich). Following the loading portion
of the experiment, it was applied −0.10 V to discharge hydrogen
incorporated in the Pd film. From the total amount of anodic charge
collected, the H content in the metal was calculated assuming the
discharge of absorbed hydrogen is the only electrolytic reaction at
the working electrode and that hydrogen is quantitatively removed
from the electrode. The amount of Pd in the film was determined
by electrothermal atomisation after dissolving it in an HNO3 + HCl
solution (1:1). Reproducible results were obtained from new pre-
pared electrodes.
It is important to remark that despite electroless deposition
yields a quite low surface roughness in contrast to others deposition
Table 2
Surface characterization parameters of Pd-electroless films (roughness factor, Rq,
particles dimension) and palladium loading
Roughness
factor
Rq (5 × 5 ␮m)
∅ particles
Pd loading
(mg cm−2
(nm)
)
Pd (low T)
Pd (high T)
Pd alloy (low P)
Pd alloy (high P)
5.3
6.1
4.2
–
77
24
21
18
22–106
40–50
8–18
2.8 (0.52)a
2.6 (0.43)a
1.2 (0.28)a
2.2
18–22
a
These values rely on the real surface area of Pd films.

8140
M.C.F. Oliveira / Electrochimica Acta 53 (2008) 8138–8143
Fig. 1. SEM images of as-deposited Pd-electroless films.
∼
methodologies, like electrodeposition (rf 10) [15] or chemical dis-
As expected, the obtained amount of Pd on the prepared films
(per geometric surface area) is higher than on anode materials pre-
pared with Pd nanoparticles dispersed on a supporting material
(0.08–1.0 mg cm−2) [18–20]. However, it is important to remark that
an effective comparison should only rely on Pd loading determined
=
placement reaction (rf = 30) [16], it may be enhanced by applying
a thermal treatment to the electroless alloys [17]. Notwithstanding
such procedure was not applied in this work as it is out of the main
objectives of the present study.

M.C.F. Oliveira / Electrochimica Acta 53 (2008) 8138–8143
8141
Fig. 2. AFM surface topographies of Pd-electroless films at a magnification of 5␮m × 5 ␮m.
per real surface area and unfortunately such data was not available
in the literature data.
Fig. 3 shows the XRD pattern of as-deposited Pd-electroless
films. The sharpest peaks are ascribed to the substrate due to the
thin thickness of the deposit. The XRD pattern exhibit clearly five
diffraction peaks at about 2Â = 40◦, 47◦, 68◦, 82◦ and 86◦ which are
indexed to (1 1 1), (2 0 0), (2 2 0), (3 1 1) and (2 2 2) facets of palla-
dium. Contrasting to pure Pd films, the broadness of Pd peaks at the
Pd alloy films reveals that these films are amorphous. In agreement
with others electroless alloys [21], the degree of amorphicity of the
alloy seems to increase as the phosphorus content of the Pd–P alloy
increases as well.
3.2. Electrochemical characterization of the Pd-electroless films in
alkaline medium
Fig. 4 shows typical steady-state voltammograms obtained in
0.1 M NaOH for pure-Pd and Pd-alloys electroless films at 10 mV s−1
,
which is a scan rate high enough to avoid non-electrochemical
recombination reaction and low enough to allow diffusion of hydro-
gen from the bulk of the electrode to the surface electrode [22].
Comparison of charge density of peak I on the several prepared Pd
films will allow evaluating, in a whole, the films ability to absorb
and oxidize hydrogen. Concerning pure Pd films, it can be concluded
that Pd (low T) film exhibits a higher ability to absorb and to oxi-
dize hydrogen than Pd (high T) film. It was found that peak I charge
is 2.7 times higher on Pd (low T) than on Pd (high T) (if charge
density comparison relied on the real surface area, the difference
ability would be even higher). On account of the different structure
of Pd core on both films, it is suggested that the spongy struc-
ture observed on Pd (low T) film promotes hydrogen absorption
(and consequently its oxidation) being responsible for the observed
behaviour.
Fig. 3. XRD pattern of as-deposited Pd-electroless films.

8142
M.C.F. Oliveira / Electrochimica Acta 53 (2008) 8138–8143
Fig. 5. Two typical current–time transient for the charge (at −1.4 V) and discharge
(at −0.1 V) of Pd (high T) film in 0.1 M NaOH solution.
that Pd (low T) presents a better performance than Pd (high T) film.
Comparing pure Pd and Pd alloys it can be concluded that pure Pd
(low T) exhibits the highest activity to absorb and oxidize hydrogen,
followed by Pd alloy (low P).
The plateau observed in Fig. 6 corresponds, in terms of H/Pd
ratio, to 1.10, 0.44 and 0.88 in Pd (low T), Pd (high T) or Pd alloy (high
P) and Pd alloy (low P) films, respectively. These data allows con-
cluding that, Pd (low T) and Pd alloy (low P) films present a better
activity towards hydrogen absorption/oxidation than Pd films pre-
pared by others methodologies (electrodeposition or laser pulsed
deposition) [22,25,26]. The rather high H/Pd ratio for Pd (low T)
Fig. 4. Cyclic voltammograms of Pd-electroless films in 0.1 M NaOH; 10 mV s−1
.
Compared to pure Pd films, the different ability to absorb and
oxidize hydrogen on the two prepared Pd-alloy membranes is not so
distinct. Pd alloy (low P) shows apparently a higher ability to absorb
and oxidize hydrogen than the high P content alloy (the anodic
charge is 1.9 times higher), but a reliable comparison is not much
feasible when the real surface area of the surface film is unknown.
Differences on the peak potentials of peak I on the two films are
not fully understood but may be attributed to different locations
of absorbed hydrogen, i.e. whether it is preferentially accumulated
close to the electrode surface or in the Pd bulk.
Additional experiments in the alkaline medium were also per-
formed to evaluate quantitatively the films ability to absorb and to
oxidize hydrogen. In such experiments the Pd films were loaded and
unloaded by constant potential electrolysis. Fig. 5 presents typical
charge and discharge curves for pure Pd-membranes in 0.1 M NaOH.
Integration of these curves allowed plotting the anodic charge ver-
suscathodiccharge, both dividedbythe Pdload foreachfilm(Fig. 6).
The observed plateau reflects the saturation of the electrode with
hydrogen which is consistent with the concomitant observation
of evolved bubbles. The obtained results allow concluding that Pd
alloy (low P) displays a higher ability to absorb and oxidize hydro-
gen than the high content P alloy. This result is in agreement with
Paseka [23] and Lu et al. [24] data for Ni–P alloys. These authors have
also found a higher amount of absorbed hydrogen on low P con-
tent alloys (3–5 at.%) than on electrodes with a P content of about
13–14 at.%. They both claimed that despite no electrocatalytic syn-
ergetic reaction occurs by the presence of the non-metallic element,
the formation of an amorphous structure, to a certain low limit
content of phosphorus, increases Pd ability to absorb hydrogen.
Concerning pure Pd films, the obtained results (Fig. 6) reveal
to be in good agreement with the voltammetric data, confirming
Fig. 6. Plot of Qa vs. Qc for Pd-electroless films in 0.1 M NaOH solution.

M.C.F. Oliveira / Electrochimica Acta 53 (2008) 8138–8143
8143
does not depend on the cathodic limit potential and that its peak
potential is exactly the same in a Ni pure substrate. So, it maybe
suggested that this peak is due to the oxidation of the metal sub-
strate through the small holes that were found on the film surface.
Comparison of the anodic charge assigned to hydrogen oxidation
(at 0.0 V) on both pure Pd films reinforces the conclusion that Pd
(low T) displays a higher activity to absorb and oxidize hydrogen
than Pd (high T).
4. Conclusions
The overall results seem to reveal that the Pd amorphicity
(brought by a low content of P) and the Pd core porosity (brought
by a low deposition rate) are the main factors responsible for the
high activity of the Pd-based electroless films to absorb and oxidize
hydrogen in the alkaline medium. It is foreseen that these materials
are suitable to be used in the future as hydrogen-permeable anodes
in fuel cells.
Acknowledgements
I am grateful to Dr. Ana Viana (FCUL) for the AFM measurements
and Dr. Nuno Martins (UTAD-UME) for the SEM/EDS analysis.
References
[1] H. Inoue, T. Abe, C. Iwakura, Chem. Commun. (1996) 55.
[2] P. Cabot, E. Guezala, J. Calpe, M. García, J. Casado, J. Electrochem. Soc. 147 (2000)
43.
[3] P. Cabot, E. Guezala, J. Casado, J. New Mater. Electrochem. Syst. 2 (1999) 253.
[4] S. Yamaguchi, T. Shishido, H. Yugami, S. Yamamoto, S. Hara, Solid State Ionics
162/163 (2003) 291.
[5] S. Yamaguchi, S. Yamamoto, T. Shishido, M. Omori, A. Okubo, J. Power Sources
129 (2004) 4.
[6] S. Adhikari, S. Fernando, Ind. Eng. Chem. Res. 45 (2006) 875.
[7] H.-B. Zhao, K. Pflanz, J.-H. Gu, A.-W. Li, N. Stroh, H. Brunner, G.-X. Xiong, J.
Membr. Sci. 142 (1998) 147.
Fig. 7. Cyclic voltammograms of Pd-electroless films in 0.1 M H2SO4; 10 mV s−1
.
[8] G.B. Sun, K. Hidajat, S. Kawi, J. Membr. Sci. 284 (2006) 110.
[9] A. Li, W. Liang, R. Hughes, Catal. Today 56 (2000) 45.
[10] B. Nair, J. Choi, M. Harold, J. Membr. Sci. 288 (2007) 67.
[11] D. Tanaka, M. Tanco, S. Niwa, Y. Wakui, F. Mizukami, T. Namba, T. Suzuki, J.
Membr. Sci. 247 (2005) 21.
[12] D. Tanaka, M. Tanco, T. Negase, J. Okasaki, Y. Wakui, F. Mizukami, T. Suzuki, Adv.
Mater. 18 (2006) 630.
[13] E. Robertis, A. Fundo, A. Motheo, L. Abrantes, J. Braz. Chem. Soc. 16 (2005)
103.
may be endorsed to H2 trapping within the porous structure of the
film.
3.3. Electrochemical characterization of the Pd-electroless films in
acid medium
In order to compare Pd-electroless films activity in alkaline and
acid medium, cyclic voltammograms were also recorded in sul-
phuric acid medium. Typical cyclic voltammograms for pure Pd and
Pd alloy membranes are shown in Fig. 7. It is clearly revealed that Pd
alloys do not absorb hydrogen in this medium, which is consistent
with the formation of gas bubbles, observed from about −0.70 V. In
order to evaluate whether this behaviour was exclusively related
[14] A.N. Correia, L. Mascaro, S. Machado, L. Avaca, Electrochim. Acta 42 (1997)
493.
[15] A. Czerwinski, R. Marassi, S. Zamponi, J. Electroanal. Chem. 316 (1991) 211.
[16] M.C. Oliveira, A. Viana, Int. J. Hydrogen Energy 32 (2007) 3100.
[17] A.M. Fundo, L.M. Abrantes, J. Electroanal. Chem. 600 (2007) 63.
[18] D. Yuan, C. Xu, Y. Liu, S. Tan, X. Wang, Z. Wei, P. Shen, Electrochem. Commun. 9
(2007) 2473.
[19] J. Pascual, S. Cigarroa, O. Feria, J. Power Sources 172 (2007) 229.
[20] J. Liu, J. Ye, S. Jiang, Y. Tong, Electrochem. Commun. 9 (2007) 2334.
[21] M.C. Oliveira, J. Appl. Catal. 329 (2007) 7.
[22] A. Czerwinski, I. Kiersztyn, M. Grden, J. Electroanal. Chem. 492 (2000) 128.
[23] I. Paseka, Electrochim. Acta 40 (1995) 1633.
[24] G. Lu, P. Evan, G. Zangari, J. Electrochem. Soc. 150 (2003) A551.
[25] J. Paillier, L. Roué, J. Electrochem. Soc. 152 (2005) E1.
[26] M. Baldauf, D. Kolb, Electrochim. Acta 38 (1993) 2145.
−
to the presence of HSO4 ions, others electrolytes (HClO4 and HCl)
were used, but the hydrogen absorption hindrance retained. The
understanding of this peculiar behaviour is currently under inves-
tigation. Concerning Pd pure films, a small peak at about −0.30 V
appears on Pd (high T). It was found that the charge of this peak