Pd deposition onto Au(111) from nitrate solution

Article abstract of DOI:10.1016/j.electacta.2009.02.005

The initial stages of palladium deposition onto Au(111) from 0.1 M HNO₃ + 0.2 mM Pd(NO₃)₂ have been studied by cyclic voltammetry and in situ scanning tunnelling microscopy. It is demonstrated that nucleation starts exclus

Full text of DOI:10.1016/j.electacta.2009.02.005

Electrochimica Acta 54 (2009) 3830–3834
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
Pd deposition onto Au(111) from nitrate solution
1
∗,1
C. Köntje, L.A. Kibler , D.M. Kolb
Institute of Electrochemistry, University of Ulm, Ulm D-89069, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 5 August 2008
Received in revised form 28 January 2009
Accepted 1 February 2009
Available online 10 February 2009
3 3 2
The initial stages of palladium deposition onto Au(111) from 0.1 M HNO + 0.2 mM Pd(NO ) have been
studied by cyclic voltammetry and in situ scanning tunnelling microscopy. It is demonstrated that nucle-
ation starts exclusively at surface defects such as monoatomic high steps, which is at variance with
recently published work. From this and our previous work it thus appears that surface defects are the
preferred nucleation sites indeed for nitrate, sulphate and chloride containing solutions.
©
2009 Elsevier Ltd. All rights reserved.
Keywords:
Au(111)
Electrodeposition
Nucleation
Palladium
STM
1
. Introduction
tials [5–7]. This was demonstrated in great detail for Pd on Au(111),
deposited from either chloride or sulphate containing electrolytes
[8,9]. Only in the case of high overpotentials deposition starts also
on flat terraces [10]. In a recent paper by Pandelov and Stim-
ming [11], describing Pd deposition onto Au on glass from 0.5 mM
Pd(NO3)2 in perchloric acid solution, it was stated that nucleation
starts on the flat terrace in such a solution. The conspicuous absence
of defects as nucleation sites was explained by a weakly adsorbing
anion, namely nitrate.
Metaldepositionfromaqueoussolutions is stillaprocessofgreat
industrial relevance. The production of metal coatings for either
protection or decorative purposes requires a detailed knowledge of
the influence of all the deposition parameters, including substrate
structure and electrolyte composition. The initial stages of metal
deposition are of particular interest as they play a decisive role
in the growth behaviour and the emerging morphology of over-
layers [1–3]. Furthermore, since bimetallic surfaces are often used
in electrocatalysis, deposition of small amounts of metal A onto
a metal B is frequently employed for creating model catalysts, for
which structure-reactivity relations can be studied in detail. In this
respect, Pd on Au(111) may serve as an example [4].
Initiation of the deposition process requires nucleation cen-
tres, and it is generally agreed that surface imperfections are the
preferred sites for the new phase to grow. Based on simple ther-
modynamic considerations monoatomic high steps in general, and
kink sites in particular, are considered to be the most effective
nucleation centres for well-prepared single crystal surfaces like
Au(111). For less-ordered surfaces like textured Au(111) films on
glass, grain boundaries and regions of step-bunching add to the
available nucleation sites.
Since such an observation would severely question the valid-
ity of our current understanding of electrocrystallization [12,13], it
seems justified to revisit Pd deposition onto Au(111) from nitrate
solution. Our results from cyclic voltammetry and in situ STM for
the initial stages of Pd deposition onto a freshly prepared massive
Au(111) crystal in 0.2 mM Pd(NO ) + 0.1 M HNO are described in
3
2
3
the following. Special consideration has been given to the fact that
Au(111) surfaces reconstruct and that lifting of the reconstruction
at positive potentials is an inherent source of surface imperfections.
Attention is also drawn to nitrate reduction on Au(111) at negative
potentials, which is shown to occur preferentially at step edges.
2. Experimental
In a number of publications on the initial stages of metal depo-
sition we have shown by in situ STM, that indeed monoatomic high
steps are the preferred and only nucleation sites for low overpoten-
The STM images were recorded with a Topometrix TMX Dis-
coverer 2010. For the preparation of the STM tips, a Pt/Ir wire
(80/20) was etched in 4.5 M NaCN and coated with an elec-
trophoretic paint to reduce Faraday currents at the tip/electrolyte
interface. All images were recorded in the constant-current mode,
typically at IT = 2 nA. Pt wires were used as counter and quasi-
reference electrodes for the STM. The potential difference between
∗ Corresponding author. Tel.: +49 731 50 25400; fax: +49 731 50 25409.
E-mail address: dieter.kolb@uni-ulm.de (D.M. Kolb).
1
ISE member.
0
013-4686/$ – see front matter © 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2009.02.005

C. Köntje et al. / Electrochimica Acta 54 (2009) 3830–3834
3831
a saturated calomel electrode (SCE) and the Pt wire is in this
case EPt = +0.67 ± 0.05 V. An SCE served as reference for the cyclic
voltammograms obtained in a standard electrochemical cell. In the
following, all potentials are quoted with respect to SCE.
The Au single crystal electrodes were discs (MaTecK Jülich, Ger-
◦
many) which had surfaces oriented to better than 1 and were
polished down to 0.03 ␮m. Discs with a diameter of 4 mm were used
for cyclic voltammetry, while the crystals used for STM were 12 mm
in diameter. Before each measurement the electrodes were flame-
annealed with a Bunsen burner or in a hydrogen flame at orange
heat for typically 5 min and cooled down in air (CV) or in a stream
of nitrogen (STM). In order to show the importance of monoatomic
high steps for nitrate reduction, a stepped Au(111) crystal was also
used for cyclic voltammetry, the surface of which had a miscut
◦
of 6 . After each deposition study, the electrode was polarized in
0
.1 M H SO for around 10 s at 10 V to remove Pd residues [14]. The
2 4
brown Au oxide overlayer was then dissolved in 1 M HCl, before the
electrode was again flame-annealed.
Fig. 2. Current–potential curve for the deposition of palladium onto Au(111) from
All electrolytes were made from HNO3 (Merck, suprapur), pal-
0
.1 M HNO3 + 0.2 mM Pd(NO3)2 showing formation of the first Pd monolayer in a
ladium powder (Aldrich, 99.999%) and high-quality Milli-Q water
−
1
peak at +0.55 V. Scan rate: 1 mV s
.
◦
(
18.2 Mꢀ cm at 25 C and <2 ppb total organic carbon). Commer-
cially available Pd(NO3)2 turned out to be not sufficiently pure; it
caused speckles in the STM images of the gold surface, which pre-
vented high-resolution viewing of Pd deposition. We mention in
defects of the otherwise catalytically rather inactive Au(111) sur-
face. To demonstrate the importance of monoatomic high steps
on the substrate for nitrate reduction, a cyclic voltammogram has
passing that Pd deposition was also studied in 0.1 M HClO + 0.2 mM
4
◦
been recorded with a stepped Au(111) crystal (miscut 6 ). It is obvi-
Pd(NO3)2 with essentially identical results.
ous that the activity for nitrate reduction is significantly enhanced
for the Au(111) surface with larger density of monoatomic high
steps. The negative end of the CV is shown in the inset of Fig. 1
together with the corresponding curve in 0.1 M H SO , for compar-
3. Results and discussion
2
4
3.1. Cyclic voltammetry
ison.
The current–potential curve for palladium deposition onto
Au(111) from 0.1 M HNO + 0.2 mM Pd(NO ) , starting at +0.72 V
Fig. 1 shows the cyclic current–potential curve for Au(111) in
3
3 2
−
.1 M HNO , recorded at 10 mV s . The Au(111) surface is ther-
3
1
0
in negative direction, is shown in Fig. 2. The sharp peak around
mally reconstructed at the immersion potential of 0 V, the scan
being started in negative direction. The peak pair between 0.15 V
and 0.4 V represents the lifting of the reconstruction for positive
scan direction, while the potential-induced reconstruction starts
in negative direction. At the negative end of the cyclic voltammo-
gram for nitric acid a slight down-bending is noticed, indicative
of a reduction current superimposed on the double-layer charg-
ing current in a potential region, where hydrogen evolution does
not yet occur. Hence, this current can be assigned to the elec-
troreduction of nitrate [15], which occurs preferentially at surface
0
.55 V indicates underpotential deposition of the first monolayer.
−2
The charge for this peak is about 460 ␮C cm
its set to 0.40 and 0.64 V), which is in good agreement with the
calculated charge of 445 ␮C cm
pseudomorphic Pd monolayer on Au(111). On the basis of our STM
(integration lim-
−2
for the electrodeposition of a
◦
Fig. 1. Cyclic current–potential curve (first cycle) for Au(111) (miscut < 1 ) in 0.1 M
−1
HNO3. Scan rate: 10 mV s . The curve for Au(111) in 0.1 M H2SO4 is included for
comparison. The inset shows the negative part of the CV for stepped Au(111) (miscut
Fig. 3. STM image (300 nm × 300 nm) of Au(111) in 0.1 M HNO3 after lifting of the
reconstruction. ESCE = 0.77 V.
◦
6
) in 0.1 M HNO3 and 0.1 M H2SO4.

3
832
C. Köntje et al. / Electrochimica Acta 54 (2009) 3830–3834
measurements, we can tentatively assign the small shoulder at
.47 V to Pd deposition on top of the monoatomic high gold islands,
adsorption of anions is avoided. However, under the conditions of
Pd deposition, which starts at rather positive potentials (E> pzc),
reconstruction inevitably is lifted. As a consequence, formerly per-
fectly flat terraces of the reconstructed surface are littered with
monoatomic high gold islands, which cover about 5% of the total
surface and which constitute nucleation centers for the metal depo-
sition reaction. Hence it is desirable to characterize the structure
of a Au(111) electrode after lifting of the surface reconstruction in
nitrate solution and before Pd deposition.
0
which completes monolayer formation (see below).
The Nernst potential of Pd2+/Pd for our solution has been deter-
mined to 0.53 V. However, bulk deposition is seen to start only
around 0.35 V, suggesting a very slow deposition kinetics in nitrate
solution.
3.2. In situ STM measurements
Fig. 3 shows an Au(111) surface after lifting of the reconstruction
at a potential of 0.77 V. The flame-annealed reconstructed surface
A freshly prepared, flame-annealed Au(111) electrode is recon-
√
structed, the surface structure being identical to the ( 3 ×
was immersed into 0.1 M HNO at −0.05 V, where reconstruction is
3
2
2) structure observed for samples prepared in UHV [16]. The
stable, and the potential was then stepped to 0.77 V. The terraces
are covered with monoatomic high gold islands with diameters of
around 4–8 nm.
reconstructed surface is stable under electrochemical conditions,
provided negative surface charges are applied (E < pzc) and specific
Fig. 4. Sequence of 6 STM images (350 nm × 350 nm) showing the initial stages of Pd deposition onto Au(111) from 0.1 M HNO3 + 0.2 mM Pd(NO3)2. Deposition potentials and
times as indicated in the figures. Tunneling current: 2 nA. Tip potential: 0.8 V vs SCE.

C. Köntje et al. / Electrochimica Acta 54 (2009) 3830–3834
3833
Fig. 5. STM images showing nucleation and growth of Pd on Au(111) in chloride (a,b), sulphate (c,d) and nitrate (e,f) containing acidic solutions. Images a, c and e: Au(111)
before Pd deposition. Images b, d and f: initial stages of Pd monolayer formation at underpotentials.
The same type of surface is seen again in Fig. 4a, which shows
bare, unreconstructed Au(111) in 0.1 M HNO + 0.2 mM Pd(NO )
ity. As has been shown by various groups, there is a significantly
increased surface mobility for gold in chloride containing solution,
which leads to the so-called electrochemical annealing [17,18]. As
a direct consequence, lifting of the Au(111) reconstruction leads to
monoatomic high gold islands that are larger in size, but fewer than
those in sulphuric acid solutions (compare Fig. 5a and c). In that
case nucleation and growth of Pd can be studied by STM excep-
tionally clearly. Note the chemical contrast in the STM images for
Au and Pd (Fig. 5b). Gold adatom diffusion is much slower in sul-
phate or nitrate (and perchlorate, not shown) containing solutions,
resulting in a high number of small gold islands (Fig. 5c and e).
The triangular shape of the Pd islands that grow in sulphuric acid
solution is noteworthy [9]. In all three cases the growth is two-
dimensional up to almost two Pd layers, regardless of the anion in
solution. The statement in ref. [11] that scan rates of 1 mV/s should
be applied in order to achieve 2-D growth is in clear contrast to our
findings. However, it is correct to say that such low scan rates are
required to detect different nucleation sites in cyclic voltammetry
[9]. It is also incorrect to state that Pd monolayer formation is much
faster in nitrate than in sulphate [11]. Our observations published in
ref. [9] did not show any inhibiting effect of the adsorbed sulphate
layer.
3
3 2
at 0.62 V as a starting point for Pd deposition. Lowering the elec-
trode potential to 0.57 V initiates Pd deposition, which is seen to
occur exclusively at the monoatomic high steps. The rims of the
gold islands seem to be most effective nucleation centers, quickly
leading to a Pd overlayer which uniformly covers the lower ter-
races (Fig. 4c). As noted on previous occasions [8], a clearly higher
overpotential is required for Pd deposition on the tops of the gold
islands, which obviously lack the classic nucleation sites like kinks.
At 0.49 V an STM image emerges (Fig. 4d) which is practically iden-
tical to that of the bare surface (see Fig. 4a), indicating completion
of the first Pd monolayer, i.e., terraces and islands are now covered
with Pd. Nucleation and growth of the second Pd layer is demon-
strated in Fig. 4e and 4f. Although at such high overpotentials the
Pd-covered gold islands still provide nucleation sites, Pd deposition
is also noted to occur on the first Pd monolayer in regions where
the Au(111) substrate apparently had no surface defects (see region
on the upper terrace near the step in the centre of Image 4a). This
may be a consequence of the rather negative deposition potential.
The third layer starts to grow before the second layer is completed.
This is seen by the bright spots in Fig. 4f, indicating the beginning
of 3-dimensional growth.
Acknowledgements
4
. Conclusions
This work was supported by the Fonds der Chemischen Indus-
trie.
In summary we can safely state that Pd deposition onto Au(111)
starts always at the monoatomic high rims of islands or the
monoatomic high steps of terraces in nitrate, sulphate and chlo-
ride containing solutions. This is demonstrated again in Fig. 5. The
nature of the anion has no effect on the nucleation behaviour.
It does, however, influence the growth and hence the shape of
the emerging Pd islands simply by affecting surface atom mobil-
References
[
1] K. Jüttner, W.J. Lorenz, Z. Phys. Chem. N.F. 122 (1980) 163.
2] R.J. Nichols, in: J. Lipkowski, P.N. Ross (Eds.), Imaging of Surfaces and Interfaces,
Wiley-VCH, Weinheim, 1999, p. 99.
[

3834
C. Köntje et al. / Electrochimica Acta 54 (2009) 3830–3834
[
3] D.M. Kolb, in: R.C. Alkire, D.M. Kolb (Eds.), Advances in Electrochemical Science
and Engineering, vol.7, Wiley-VCH, Weinheim, 2002, p. 107.
[12] E. Budevski, in: B.E. Conway, J.O‘M. Bockris, E. Yeager, S.U.M. Khan, R.E. White
(Eds.), Comprehensive Treatise of Electrochemistry, vol.7, Plenum Press, New
York, 1983, p. 399.
[13] E. Budevski, G. Staikov, W.J. Lorenz, Electrochemical Phase Formation and
Growth, Wiley-VCH, Weinheim, 1996.
[
[
4] L.A. Kibler, Chem. Phys. Chem. 7 (2006) 985.
5] T. Will, M. Dietterle, D.M. Kolb, in: A.A. Gewirth, H. Siegenthaler (Eds.),
Nanoscale Probes of the Solid-liquid Interface, Nato-ASI Series E, vol.288,
Kluwer, Dordrecht, 1995, p. 137.
[14] M. Baldauf, Ph.D. thesis, University of Ulm, 1996.
[
6] H.-F. Waibel, M. Kleinert, L.A. Kibler, D.M. Kolb, Electrochim. Acta 47 (2002)
[15] G.E. Dima, A.C.A. de Vooys, M.T.M. Koper, J. Electroanal. Chem. 554–555 (2003)
15.
1461.
[
[
[
7] R. Hoyer, L.A. Kibler, D.M. Kolb, Surf. Sci. 562 (2004) 275.
8] L.A. Kibler, M. Kleinert, R. Randler, D.M. Kolb, Surf. Sci. 443 (1999) 19.
9] J. Tang, M. Petri, L.A. Kibler, D.M. Kolb, Electrochim. Acta 51 (2005) 125.
[16] A.S. Dakkouri, D.M. Kolb, in: A. Wieckowski (Ed.), Interfacial Electrochemistry,
Marcel Dekker, New York–Basel, 1999, p. 151.
[17] J.L. Stickney, I. Villegas, C.B. Ehlers, J. Am. Chem. Soc. 111 (1989) 6473.
[18] H. Striegler, P. Skoluda, D.M. Kolb, J. Electroanal. Chem. 471 (1999) 9.
[
10] H. Naohara, S. Ye, K. Uosaki, J. Chem. Phys. B 102 (1998) 4366.
[11] S. Pandelov, U. Stimming, Electrochim. Acta 52 (2007) 5548.