X-ray absorption spectroscopy characterization of Zn underpotential deposition on Au(1 1 1) from phosphate supporting electrolyte

Article abstract of DOI:10.1016/j.electacta.2010.07.046

Zn K-edge X-ray absorption spectroscopy (XAS) has been used to investigate the structure of Zn monolayers prepared on Au(1 1 1) electrodes via underpotential deposition (UPD) from phosphate supporting electrolyte. Theoretical modeling of the XAS data indicates that the Zn adatoms adopt a commensurate (√3 × √3)R30° (θ_sc = 0.33) adlayer structure and reside within the 3-fold hollow sites of the Au(1 1 1) surface. Meanwhile, phosphate counter-ions co-adsorb on the UPD adlayer and bridge between the Zn adatoms in a (√3 × √3)R30° (θ_sc = 0.33) configuration, with each phosphorous atom residing above a vacant 3-fold hollow site of the Au(1 1 1). Significantly, this surface structure is invariant between the electrochemical potential for UPD adlayer formation and the onset of bulk Zn electrodeposition. Analysis of the Zn K-edge absorption onset also presents the possibility that the Zn adatoms do not fully discharge during the process of UPD, which had been proposed in prior voltammetric studies of the phosphate/Zn(UPD)/Au(1 1 1) system.

Full text of DOI:10.1016/j.electacta.2010.07.046

Electrochimica Acta 55 (2010) 8532–8538
Contents lists available at ScienceDirect
Electrochimica Acta
journal homepage: www.elsevier.com/locate/electacta
X-ray absorption spectroscopy characterization of Zn underpotential deposition
on Au(1 1 1) from phosphate supporting electrolyte
a,b,∗
, Rachel L. O’Malley , Timothy J. O’Connell , Antje Vollmera,c, Trevor Raymenta,d,∗∗
a
a
Jonathan R.I. Lee
a
Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK
Lawrence Livermore National Laboratory, Livermore, CA 94550, USA
Helmholtz Zentrum Berlin f. Materialien und Energie, BESSY II, Albert-Einstein-Str. 15, 12489 Berlin, Germany
Diamond Light Source Ltd., Diamond House, Harwell Science and Innovation Campus, Didcot, Oxfordshire OX11 0DE, UK
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 6 April 2010
Received in revised form 14 July 2010
Accepted 17 July 2010
Available online 24 July 2010
Zn K-edge X-ray absorption spectroscopy (XAS) has been used to investigate the structure of Zn mono-
layers prepared on Au(1 1 1) electrodes via underpotential deposition (UPD) from phosphate supporting
electrolyte. Theoretical modeling of the XAS data indicates that the Zn adatoms adopt a commensurate
√
√
◦
(
3 × 3)R30 (ꢀsc = 0.33) adlayer structure and reside within the 3-fold hollow sites of the Au(1 1 1)
surface. Meanwhile, phosphate counter-ions co-adsorb on the UPD adlayer and bridge between the Zn
√
√
◦
adatoms in a ( 3 × 3)R30 (ꢀsc = 0.33) configuration, with each phosphorous atom residing above a
vacant 3-fold hollow site of the Au(1 1 1). Significantly, this surface structure is invariant between the
electrochemical potential for UPD adlayer formation and the onset of bulk Zn electrodeposition. Anal-
ysis of the Zn K-edge absorption onset also presents the possibility that the Zn adatoms do not fully
discharge during the process of UPD, which had been proposed in prior voltammetric studies of the
phosphate/Zn(UPD)/Au(1 1 1) system.
Keywords:
Underpotential deposition
XAS
Zinc
Phosphate
Au electrode
©
2010 Elsevier Ltd. All rights reserved.
1
. Introduction
X-ray absorption spectroscopy (XAS) is used to characterize the
structure of Zn monolayers prepared on Au(1 1 1) substrates via
UPD from phosphate supporting electrolyte.
Underpotential deposition (UPD), the phenomenon of metal
monolayer formation on a foreign metal substrate under more
oxidizing potentials than the equilibrium potential for bulk elec-
trodeposition, has been the subject of considerable research in
recent years because it yields model systems for investigation of the
electrode/electrolyte interface [1–4]. Many UPD systems are com-
posed of well-ordered adlayers that are commensurate with the
underlying electrode surface and, as such, offer an excellent oppor-
tunity to study the structure and composition of the interfacial
region. Furthermore, they enable the evaluation and comparison of
different techniques for structural characterization of the electri-
fied interface. The capability for analysis of the interfacial structure
is of fundamental importance in developing an enhanced under-
standing of the modes and mechanisms of electrochemical surface
processes and, therefore, insight into technologically relevant pro-
cedures such as electroplating and electrocatalysis. In this paper,
XAS provides a valuable technique for structural analysis of the
electrified interface. The elemental specificity of XAS enables one
to characterize the local environment (up to a radius of ∼5 A˚ [5,6])
of individual elements within a condensed phase and, as such,
allows the detailed investigation of both ordered and amorphous
systems. Therefore, XAS represents a complementary or alternative
tool to techniques that require longer range order, which include
X-ray diffraction (XRD)/surface X-ray scattering (SXS). One can also
conduct in situ XAS measurements of electrode materials with
the use of synchrotron radiation [7–9], including in situ struc-
tural studies of the electrode surface [10–13]. This is an important
capability because it eliminates the potential for rearrangement
of the interfacial structure upon removal of the working electrode
from the electrolyte and, by extension, electrochemical control.
Despite the apparent strengths of XAS for investigation of the
electrode surface structure, use of the technique to study UPD sys-
tems was largely discontinued after the late 1990s because of the
extended data collection periods required to obtain high quality
spectra. Significantly, a recent study of electrodeposited Cu mono-
layers on Au from sulfate buffer illustrates that improvements in
flux and detection efficiency at synchrotron sources have miti-
gated this problem for UPD adlayers with high surface coverages
(surface coverage, ꢀsc, ≥0.67) [13]. Zn UPD on Au(1 1 1) from phos-
∗ Corresponding author at: Lawrence Livermore National Laboratory, Livermore,
CA 94550, USA.
∗
∗ Principle corresponding author at: Diamond Light Source Ltd., Diamond House,
Harwell Science and Innovation Campus, Didcot, Oxfordshire OX11 0DE, UK.
E-mail addresses: lee204@llnl.gov (J.R.I. Lee), trevor.rayment@diamond.ac.uk
(
T. Rayment).
0
013-4686/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2010.07.046

J.R.I. Lee et al. / Electrochimica Acta 55 (2010) 8532–8538
8533
phate buffer is an excellent system with which to build upon this
preceding Cu(UPD)/Au(1 1 1) study for several reasons: (1) voltam-
metric experiments indicate that the Zn adlayer has a considerably
reduced surface coverage (ꢀsc ∼ 0.33) [14–16] with respect to Cu
UPD and, therefore, provides an opportunity to demonstrate the
suitability of XAS for the investigation of low concentration surface
species; (2) the specific geometrical arrangement of the Zn adlayer
and co-adsorbed anion overlayer remains unresolved, despite prior
research conducted for the phosphate/Zn(UPD)/Au(1 1 1) system
using a range of characterization techniques; (3) the mechanisms
of Zn electrodeposition and structure of the resulting deposit
are of relevance for important industrial applications, particularly
in understanding the formation and properties of anti-corrosion
galvanic coatings. In fact, interest in the processes of Zn electrode-
position has resulted in studies of Zn UPD on a variety of metal
electrodes that include Au [2,14–22], Pd [23,24], Pt [23–33], Ag [34],
Cu [19] and Cd [19].
thick plate to yield an upper surface that is comprised of two
ends chamfered at 45 to a flat and polished plane of dimensions
◦
18 mm × 24 mm. When coated with Au, this plane provides the WE
surface studied with XAS. Meanwhile, Au deposited on the slop-
ing faces allows for direct electrical contact to the WE without
incurring any surface damage. Rigorous cleaning of the glass was
undertaken prior to Au deposition. Initially, the substrates were
immersed in aqua regia (3:1 ratio of concentrated HCl:HNO ) for
3
∼60 s to remove any existing Au film and then rinsed in ultrapure
water (Elga, 18.2 Mꢁ cm). Subsequent agitation in piranha solution
(3:1 ratio of H SO :H O ) for ∼1 h served to remove organic con-
2
4
2
2
taminants from the substrates [35]. After a second phase of rinsing
in ultrapure water, the samples were dried in N2 and immediately
transferred to the evaporation chamber (Balzer UTT440).
Once secured within the evaporation chamber, the substrates
◦
were preheated at ∼300 C for 10 h and then coated with 1500 A˚
−
6
Au at a rate of 1 A˚ /s and a chamber pressure of ∼1 × 10 Torr.
Although the experiments reported to date do not yield a
definitive assignment of the interfacial structure for the phos-
phate/Zn(UPD)/Au(1 1 1) system, they provide an excellent point of
reference and offer invaluable information to assist in an XAS inves-
tigation. In particular, in situ scanning tunneling microscopy (STM)
images recorded by Nakamura and co-workers [14,17] display a
Upon reaching the required thickness of Au deposit, the samples
◦
were annealed at ∼300 C for 6 h to yield domains with a (1 1 1)
surface orientation and to achieve a comparable surface rough-
ness to Au films prepared on mica [36]. Following the annealing
phase, the samples were slowly cooled to room temperature over
a 10 h period, to prevent film fracture under the stresses of rapid
temperature change.
√
√
◦
(
3 × 3)R30 (ꢀsc = 0.33) arrangement of atoms with respect to
the underlying Au(1 1 1) surface that the authors attribute to phos-
phate anions co-adsorbed on the Zn adlayer. The structure of the
UPD monolayer could not be resolved from the STM micrographs,
but a tentative 2D arrangement of the interfacial region was pro-
posed in which the Zn adatoms bridge between at least two Au
atoms at the electrode surface. Meanwhile, individual phosphate
counter-ions co-adsorb at different heights above the plane of the
Zn UPD adlayer within the model.
All electrolytes were prepared in ultrapure water and composed
of 0.1–1 mM Zn(ClO4)2 and 0.1 M KH PO4 (pH 4.4–4.5). Dissolved
2
molecular oxygen was removed from the electrolytes by purging
with Ar for ≥30 min before each experiment. For the purposes of
reference and comparison, three Zn containing standards were pre-
pared for XAS measurement in the transmission mode. The 5 ␮m
Zn metal foil was mounted in a suitable frame without further
modification. Finely ground samples of ZnO were mixed with BN
and pressed into pellets. The concentration of BN in each pellet
was adjusted until an absorption of ꢂt ∼ 1 was achieved. Finally,
a 10 mM aqueous solution of Zn(ClO ) was prepared and trans-
The XAS investigation of the phosphate/Zn(UPD)/Au(1 1 1) sys-
tem presented in this study addresses numerous aspects of the
interfacial structure that were previously unresolved. Modeling
of the experimental XAS spectra yields the structure of the Zn
adlayer buried beneath the co-adsorbed phosphate counter-ions
and the registry between the UPD adlayer and anionic overlayer.
The timescales required to obtain spectra with a sufficient signal
to noise (S/N) ratio for extended analysis are shown to be compet-
itive to alternative X-ray based characterization techniques, such
as SXS/XRD, and, therefore, suitable for routine studies of low con-
centration surface species using synchrotron radiation.
4
2
ferred to a liquid cell with a 2 mm path length for transmission XAS
measurements.
2.3. Instrumentation
XAS and electrochemical measurements were conducted using a
purpose built cell, which was designed for use in the thin layer con-
figuration employed by Samant et al. [37] for SXS studies of UPD. A
thorough description of the composition and operation of an equiv-
alent cell during electrochemical and XAS studies of UPD adlayers
is available in the literature [13]. We note, however, that the cell
allows for exposure of the Au(1 1 1) electrodes to bulk electrolyte
during electrochemical measurements and UPD adlayer deposi-
tion. Moreover, the WE faces a Mylar window, which is deflated
during XAS measurements to trap a thin layer of solution at the
Au(1 1 1) surface. The presence of the thin layer ensures electro-
chemical control over the system throughout the XAS studies, while
ensuring minimal attenuation of the incident X-ray beam by the
electrolyte. Electrochemical control of the system under investi-
gation was provided by an Eco-Chemie PGSTAT12 potentiostat,
which was operated using the General Purpose Electrochemical
System (GPES) software, and a standard three-electrode configu-
ration comprising of the Au(1 1 1) WE, Pt foil CE and SCE RE. The
preparation and XAS measurement of a surface adlayer of Zn was
conducted at a constant electrode potential within the appropri-
ate region for UPD. During adlayer formation, the experimental cell
was held in the fully inflated configuration for 20 min to ensure that
an equilibrium surface coverage was reached. The Mylar film was
then deflated and the cell transferred to the experimental beamline
without loss of the potentiostatic condition.
2
. Experimental
2.1. Reagents and materials
All reagents were obtained from commercial sources and
used without further purification. Boron nitride (99%), potassium
dihydrogen phosphate (99.99%), zinc oxide (99.99%) and zinc per-
chlorate (99.99+%) were purchased from Aldrich. 5 ␮m zinc metal
foil, 3 ␮m thick Mylar film and 5 ␮m Kapton film were purchased
from Goodfellow. Tempax glass was purchased from UQG Optics
Ltd. The Au beads (99.99%) used in preparation of the evaporated
thin film Au(1 1 1) working electrodes (WEs) and the Pt foil (99.99%)
used as the counter electrode (CE) were obtained from Birmingham
Metal. The reference electrode (RE) was a saturated calomel elec-
trode (SCE, REF401) purchased from Radiometer. All potentials in
this manuscript are reported with respect to SCE.
2.2. Preparation of the working electrode, electrolyte and XAS
standards
The WEs were prepared by thermal evaporation of Au onto a
Tempax glass substrate. Each substrate was machined from 5 mm

8
534
J.R.I. Lee et al. / Electrochimica Acta 55 (2010) 8532–8538
XAS measurements were conducted on Station ID26 at the Euro-
ried out according to established protocols that are described in
greater detail in the literature [13,41]: the theoretical structure is
refined to optimize R after the inclusion of each ‘shell’ of symmetry-
related atoms, which are added sequentially in order of proximity
to the absorbing atom (shells at comparable distances were added
simultaneously).
pean Synchrotron Radiation Source (ESRF), Grenoble, France. A
detailed description of the beamline and storage ring properties
during the course of this study are available in the literature [13].
The energy of the incident X-ray photons was calibrated to the max-
imum in the first derivative of the XAS spectrum obtained from the
5
␮m zinc metal foil. An ionization chamber was utilized for the
The position of each backscattering atom is defined relative
to the absorbing element using a spherical polar co-ordinate
scheme. Within this scheme, the colatitudal angle, ꢀ, is measured
with respect to the Au(1 1 1) surface normal and has a range of
measurement of I0 and a photodiode detector positioned behind
the experimental sample provided It. Measurements in the fluo-
rescence mode were conducted using a solid state 13-element Ge
detector array (Canberra). The detector permitted count rates of
up to 100 kHz with deadtime correction and offered a resolution of
◦
0 ≤ ꢀ ≤ 180 , the azimuthal angle, ꢅ, resides in the plane of the WE
◦
surface and has a range of 0 ≤ ꢅ ≤ 360 and the distance, r, repre-
∼
300 eV FWHM for the Zn K˛ fluorescence photons.
All fluorescence XAS measurements were conducted on ID26
sents the interatomic separation. In addition, the atoms contained
in an individual shell are described collectively by their number, n,
element type, t, and relevant Debye–Waller factor (second cumu-
lant), a. Therefore, a total of six degrees of freedom, or relevant
independent points, are required for the characterization of each
model shell.
using a single experimental geometry. In this configuration, the
fluorescence detector was mounted in the horizontal plane at an
angle of 90 with respect to the axis of the incident beam. The
◦
active surface of the detector was covered by a Z-1 filter (Cu, 5 ␮m
thick) to reduce the background signal obtained from Compton and
elastic scattering [38]. In addition to improving the ratio of S/N,
this allowed a closest approach of 5–6 cm between sample and
detector without saturation of the device. The cell was mounted
on the experimental beamline by means of a motorized goniome-
ter (ID26) such that the electric vector of the incident radiation
was perpendicular to the Au(1 1 1) surface plane of the WE. The
angle of incidence, ˛, between the X-ray beam and sample surface
was selected to approach grazing incidence as closely as possible
The number of relevant independent points, NI, available for
modeling was calculated using Eq. (2), which was first reported
by Stern [38]:
2
ꢆr ꢆk
ꢇ
NI =
+ 2
(2)
where ꢆk is the range in k-space over which the spectrum is
analyzed and ꢆr is the range in real space encompassed by
the theoretical model. Structural modeling was conducted almost
◦
exclusively within the limit imposed by N . Additional models that
but was ultimately limited to ˛ = 5 by the relative dimensions of
I
did not meet this criterion were also considered, provided NI was
the WE surface and cross-section of the beam. Individual spec-
tra were collected over a 35–40 min period and a spectral range
of 9500–10,150 eV. Those recorded under identical experimental
conditions, typically 2–3 spectra, were combined to provide an
improvement in S/N.
within six relevant independent points (i.e. a fraction of a single
shell) of encompassing the entire structure, although the conclu-
sions drawn from full refinement were treated with appropriate
caution. The limits of ꢆk were established prior to modeling by
truncation of the background subtracted data. The minimum in k
was defined by the beginning of the EXAFS, while the maximum in
k corresponded to the energy at which the noise and amplitude of
the oscillations reached a comparable magnitude.
2.4. Analysis
Initially, raw XAS data was extracted from the native beamline
All data was phaseshift corrected and is presented accord-
ingly throughout this study. The theoretical atomic phaseshifts and
data format using SPEC and XOP. The energy of the absorption edge,
E0, was identified by the maximum in the first derivative of XAS
spectra collected from the 5 ␮m Zn foil. All spectra were normalized
to I0 and, subsequently, to the magnitude of the edge step, which
was taken to be the difference in absorption intensity between the
pre-edge region and 50 eV beyond the absorption edge. Extraction
of the EXAFS via subtraction of the pre- and post-edge backgrounds
was conducted using PAXAS [39]. The pre-edge region was mod-
eled using a second order polynomial and the post-edge region
was modeled with a cubic spline, which was refined via a least
squares fitting routine. The Cook-Sayers criterion [40] was applied
as an important diagnostic throughout the process of background
subtraction.
‘muffin-tin’ potentials for each structural arrangement were cal-
culated within EXCURV98 (using the parameters outlined in the
Supporting Information). The effects of multiple scattering and, for
structures of defined orientation, polarization dependence were
included in all modeling.
3. Results and discussion
3.1. Electrochemical measurements
Fig. 1 displays a characteristic CV recorded for the UPD of Zn
on Au(1 1 1) from an electrolyte composed of 1 mM Zn(ClO4)2 and
0.1 M KH PO . Obtaining a cyclic voltammogram of the form pre-
For the purposes of structural analysis, theoretical modeling of
the experimental EXAFS was conducted using EXCURV98 [41]. The
fit parameter, R, provides a measure of the match between the
experimental and theoretical spectra and is defined by:
2
4
sented in Fig. 1 was indicative of a well-prepared sample and
provided an important system diagnostic prior to the collection
of any XAS data. The appearance of the CV is consistent with those
reported in the literature[14–18] and, as a consequence, the peaks
denoted A and A* are attributed to the UPD adsorption and desorp-
tion of Zn, respectively. Peaks B and B* are observed even in the
absence of Zn and are assigned respectively to the adsorption and
desorption of phosphate ions on the WE surface [14–16]. In situ
TEM measurements reported by Nakamura and co-workers indi-
cate that neither Zn nor phosphate ions are strongly adsorbed on
the Au(1 1 1) surface at −200 mV vs. SCE and, based on this observa-
tion and the form of the cyclic voltammogram, the authors suggest
that no specifically adsorbed species are present on the electrode
surface in the double layer region between peaks A and B* [14]. This
N
ꢁ
ꢂ
ꢀ
1
ꢃi
e
i
t
i
R =
[|ꢄ (k) − ꢄ (k)|] × 100%
(1)
i
e
t
i
where ꢄ (k) represents the experimental EXAFS, ꢄ (k) represents
i
the theoretical EXAFS derived from the structural model, ꢃ repre-
i
sents a weighting parameter at point i and N is the total number of
data points in the spectrum. An improvement in fitting is accom-
panied by a reduction in R and, as a consequence, structural
refinement using EXCURV98 was conducted with the intent of min-
imizing this parameter, while also ensuring that the model was
chemically and physically viable. The model development was car-

J.R.I. Lee et al. / Electrochimica Acta 55 (2010) 8532–8538
8535
Fig. 1. CV recorded for Zn UPD on Au(1 1 1) from a phosphate supporting electrolyte
using a dedicated electrochemical cell with the ‘hanging meniscus’ configuration.
CVs recorded using the XAS cell were closely comparable despite the fact that it
does not contain an idealized geometrical arrangement of the 3-electrode system.
As such, the CV provides an excellent point of reference for subsequent XAS studies.
Fig. 2. Individual XANES spectra recorded at the Zn K-edge for a Zn UPD monolayer
prepared on Au(1 1 1) from phosphate supporting electrolyte at −612 mV (orange),
−
1
2+
Electrolyte composition: 0.1 mM Zn(ClO4)2, 0.1 M KH2PO4, rate: 20 mV s
.
−625 mV (red) and −650 mV (dark red) vs. SCE region, Zn metal (grey) and Zn in
aqueous solution (blue). All of the spectra are pre-edge background subtracted and
normalized to the edge step. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of the article.)
does not preclude the possibility that Zn adatoms could adsorb on
the Au surface between −200 mV vs. SCE and the onset of peak
A (the shoulder observed at the most oxidizing potential encom-
passed by peak A, ∼−250 mV) without any associated reductive
charge transfer.
layer of Zn accounts for 4–7% of the total signal for each ‘UPD’
spectrum. Subtraction of the electrolyte signal did not have an
appreciable effect on modeling of the UPD EXAFS in EXCURV98.
A detailed analysis of the voltammetric data collected in this
study for the phosphate/Zn(UPD)/Au(1 1 1) system is not presented
in this paper since thorough treatments are available in the litera-
ture [14–16,18]. It is important to note, however, that voltammetric
measurements conducted in preparation for the XAS experiments
yielded data in close qualitative and quantitative agreement with
the published studies. Most significantly, experiments specifically
modeled after the work of Aramata and co-workers [14–16,18] to
identify the shift in electrochemical potential of peak A (EUPD) as
a function of the Zn2+ concentration within the electrolyte yielded
3
.2.1. XANES measurements
Fig. 2 displays the XANES region of XAS spectra recorded for
UPD adlayers prepared via polarization in potential at −612 mV,
−
625 mV and −650 mV vs. SCE for 20 min and two of the exper-
imental standards. The close similarity between the ‘UPD’ XANES
data provides a preliminary, although not conclusive, indication
that the structure of the Zn adlayer is largely constant between
−
612 mV and −650 mV vs. SCE. Equally, the contrasting shape and
energy of absorption onset with respect to the experimental stan-
dards illustrates that the ‘UPD’ XAS signals do not arise from the
formation of bulk Zn deposit or primarily from the thin layer of
2+
equivalent results: the evolution in (EUPD) with [Zn ] was con-
−
sistent with an overall charge transfer of ∼1 e for the combined
process of Zn UPD adsorption and co-adsorption of phosphate.
Such close similarity in the cyclic voltammetry data ensures that
the information obtained from subsequent XAS measurements can
be directly compared with prior research published for the phos-
phate/Zn(UPD)/Au(1 1 1) system.
2+
Zn in solution. The absorption edge for the UPD spectra (9662 eV)
resides between the values for the bulk Zn metal (Zn , 9659 eV) and
the Zn solution (9664 eV), which presents the possibility that the
0
2+
Zn adatoms do not fully reduce during the process of adsorption
and support an intermediate oxidation state in the region of ∼+1,
a feature that has also been proposed for Cu UPD on Au electrodes
from sulfate supporting electrolyte [42–44]. A definitive assign-
ment of the oxidation state cannot be obtained from inspection
of the XANES and comparison with standard spectra alone because
changes in the local environment of Zn can lead to modifications in
the occupied and unoccupied electronic density of states and shift-
ing of the absorption edge, without any variation in oxidation state.
In fact, ex situ X-ray photoelectron spectroscopy studies indicate
small shifts (<1 eV) in the binding energy of core electrons within
the adatoms of Cu(UPD)/Au(1 1 1) [45] and Ag(UPD)/Au(1 1 1) [46]
systems with respect to bulk Cu and Ag, which are attributed to
the effects of adatom–substrate interactions (e.g. orbital expan-
sion/contraction resulting from the size mismatch between the
Cu or Ag and Au). There is also well established discussion in the
community on the effect of charge rearrangement arising from
the difference in workfunction/electronegativity of substrate and
adsorbate as demonstrated by Trasatti [47] and Kolb et al. [1,48,49].
This charge transfer might well cause ionic Zn species at the surface
or result in the formation of a dipole.
3.2. X-ray absorption spectroscopy measurements
Zn adlayers were prepared for XAS measurements at a series
of potentials (−612, −625 and −650 mV vs. SCE) between the
conditions required for UPD and bulk electrodeposition. As a con-
sequence, analysis of the XAS spectra enables an assignment of any
structural evolution in the Zn adlayer and the co-adsorbed phos-
phate overlayer as a function of the applied potential. XAS spectra
were also recorded at 500 mV, where one would not anticipate the
formation or presence of a Zn UPD adlayer. Instead, the Zn XAS sig-
nal should arise solely from the thin layer of solution trapped above
the electrode surface. Since the number of Zn K˛ fluorescence pho-
tons corresponding to the edge step is directly proportional to the
concentration of Zn in the beam, the measurement at 500 mV pro-
vides a quantitative assignment of the contribution from the thin
layer of solution to the UPD spectra. Comparison between the fluo-
rescence signals collected at 500 mV and within the UPD region
under identical experimental geometries indicates that the thin

8
536
J.R.I. Lee et al. / Electrochimica Acta 55 (2010) 8532–8538
Even so, the possibility that the Zn adatoms retain an interme-
diate charge in the region of +1 is intriguing because voltammetric
studies demonstrate that the UPD of Zn on Au(1 1 1) [18] and
polycrystalline Au electrodes [22] from phosphate buffer is a
one electron transfer process. Furthermore, the authors could not
unambiguously attribute an oxidation state of +1 to the Zn adatoms
from voltammetric analysis alone due to the potential for charge
transfer between the co-adsorbed phosphate ions and the UPD
adlayer. For an oxidation state of +1 (i.e. the one electron transfer is
between the Au surface and a Zn adatom), the charge density asso-
ciated with the Zn adsorption (peak A) in the cyclic voltammogram
corresponds to a surface coverage of ꢀ ∼ 0.3, which is within exper-
imental error of the coverage (ꢀ ∼ 0.33) predicted by Nakamura
and co-workers and [14] acts as a guide for subsequent structural
modeling of the XAS.
3.2.2. EXAFS measurements
As an initial diagnostic of the background subtraction and
modeling procedures, EXAFS spectra recorded for the experi-
mental standards were fitted with well-characterized crystallo-
graphic/local structures (e.g. an hcp arrangement for Zn metal, the
wurtzite structure of ZnO and octahedrally co-ordinated Zn2 in
aqueous solution). High quality fits of R ≤ 30% were obtained for
all of the standard materials. Each of the standard structures was
also applied in modeling of the experimental EXAFS recorded for
the Zn UPD adlayers. These models resulted in extremely poor
+
(
R ≥ 70%) fits, which indicates that undesired solids, such as Zn
metal and ZnO, are not responsible for the XAS spectra between
650 mV ≤ E ≤ −612 mV and provides additional confirmation that
−
2+
the contribution from solvated Zn to the EXAFS signal is low.
Subsequent models applied to the EXAFS spectra recorded for
the UPD adlayers were based upon Zn adsorption in the atop, 2-fold
bridging and 3-fold hollow sites of an Au(1 1 1) surface. Deviation
from an idealized Au(1 1 1) surface and from perfect registry of
the Zn adatoms within each adsorption site was accommodated
during model refinement, provided that the structure remained
physically viable. A range of configurations for anions within the co-
adsorbed overlayer were also included as essential components of
the model structures. The best-fitting model was identical for the Zn
adlayers prepared between −650 mV ≤ E ≤ −612 mV, which indi-
cates that the local environment of the adatoms remains unchanged
within this region of electrochemical potential. Hence, modeling in
EXCURV98 is discussed collectively for EXAFS recorded at all three
Fig. 3. (a) A characteristic Zn K-edge EXAFS spectrum (k3 weighted) recorded for
a Zn UPD monolayer prepared in phosphate supporting electrolyte between −612
and −650 mV vs. SCE (red) and the associated best-fitting theoretical model obtained
using EXCURV98 (dark blue). (b) The corresponding FTs for the experimental EXAFS
from the Zn adlayer (red) and the best-fitting theoretical model (dark blue). (For
interpretation of the references to color in this figure legend, the reader is referred
to the web version of the article.)
‘
UPD’ potentials in the following discussion.
Fig. 3(a) displays a comparison between a representative EXAFS
spectrum collected within the UPD region and the best-fitting
model obtained in EXCURV98. Associated Fourier transforms for
the experimental and theoretical EXAFS are presented in Fig. 3(b).
In all cases, the lowest fit parameters (R ∼ 39%) for the ‘UPD’ spec-
tra were obtained using models based upon Zn adsorption within
the 3-fold hollow sites of the Au(1 1 1) electrode surface. Mod-
els based upon adsorption in the atop and 2-fold bridging sites
resulted in lower quality fits to the experimental data (R ∼ 45% and
and anion overlayer (including the (√
3 ×
√3)R30◦
arrangement
observed with in situ TEM [14,17]) and/or with a Zn surface cover-
age of ꢀsc ∼ 33% identified via voltammetric measurements [14–16].
Structural parameters for the best-fitting theoretical model of
the experimental EXAFS are included in Table 1 and corresponding
schematics for the local environment of the Zn adatoms viewed
along multiple and distinct axes are displayed in Fig. 4(a) through
(d). The first shell of backscatterers is composed of three oxygen
atoms, which exhibit 3-fold symmetry about an axis co-incident
with the Au surface normal that passes through the center of the Zn
∼
43%, respectively). Moreover, the corresponding local environ-
ments for Zn adsorbed in the atop and 2-fold bridging positions
were inconsistent with the formation of an ordered UPD adlayer
Table 1
Structural parameters for the best-fitting models of the local environment of Zn atoms within the UPD monolayers prepared on Au(1 1 1) obtained from EXAFS collected
between −612 and −650 mV vs. SCE.
Debye–Waller factor, a (Å2)
ꢀ ( )
◦
ꢅ ( )
◦
Shell
Atom type
Number, n
Radial distance, r (Å)
1
2
3
4
Ef
O
Au
P
Au
3
3
3
3
1.94–1.99
2.86–2.92
3.27–3.35
4.03–4.15
0.018–0.025
0.023–0.031
0.012–0.026
0.022–0.029
42–49
135–143
49–57
0, 120, 240
0, 120, 240
30, 150, 270
60, 180, 300
124–136
−5 to 2

J.R.I. Lee et al. / Electrochimica Acta 55 (2010) 8532–8538
8537
Fig. 4. (a) A schematic of the local environment of Zn adatoms within the UPD adlayer prepared on Au(1 1 1) from phosphate supporting electrolyte viewed along the Au(1 1 1)
surface normal. Note that this structure contains only the atoms included in the best-fitting structural model applied to the experimental EXAFS. (b) A schematic of the same
◦
local environment of Zn viewed along the plane of the Au(1 1 1) surface. (c) Another in-plane schematic in which the view is rotated by 90 with respect to the view in (b).
(
d) A projection of the local environment displayed in (a) assuming that the electrode surface has a well-ordered Au(1 1 1) arrangement and that the oxygen atoms in the
phosphate ions form a tetrahedral arrangement around the central phosphorous atom. (e) The projection of this local environment over the Au(1 1 1) surface (the oxygen
√
√
√
√
◦
◦
atoms have been omitted for clarity in this schematic) to yield the open ( 3 × 3)R30 Zn adlayer with a co-adsorbed open ( 3 × 3)R30 overlayer of phosphate. The green
atoms represent the local environment of the Zn adatoms identified via theoretical modeling of the Zn K-edge EXAFS. (For interpretation of the references to color in this
figure legend, the reader is referred to the web version of the article.)
adatom. The oxygens are located above the Au atoms that directly
co-ordinate to the Zn and, as such, mirror the 3-fold hollow adsorp-
tion site in which the adatom resides. Each oxygen is bonded to a
distinct phosphorous atom, which indicates that three co-adsorbed
counter-ions are associated with the Zn. Importantly, the projection
of the Zn–O–P linkage in the plane of the Au surface is not linear:
an improvement of ꢆR ≥ 1 in the fit parameter was obtained when
each of the three oxygen atoms directly bonded to a Zn adatom sup-
port a degree of negative charge. Given that the phosphorous atom
of each counter-ion bridges between three of these oxygen atoms,
a corresponding charge would be associated with the co-adsorbed
phosphate ions. The interatomic separation for the second shell of
backscatterers (the first Au shell, r = 2.86–2.92 A˚ ) is also indicative
of excess electron density associated with the backscattering Au
atoms, provided the Zn adatoms retain a positive oxidation state.
◦
the P atoms were rotated by ꢅ = 20–30 with respect to the ori-
entation of the Zn–O bond. The combination of this rotation with
An excess of negative charge at the electrode surface would present
√
an interatomic separation of r = 3.28–3.36 A˚ and colatitudal angle
the possibility that the (23 × 3) reconstruction of the Au(1 1 1)
◦
of ꢀ = 54.5 places the P atoms directly above 3-fold hollow sites
surface observed [50–53] in the absence of Zn is not lifted during
Zn UPD.
in the Au(1 1 1) surface that are unoccupied by the Zn adatoms. It
is noteworthy that the Zn–P and Zn–Au (second shell) interatomic
separations are similar (3.28–3.36 A˚ and 2.86–2.92 A˚ , respectively),
Expanding the structural model to include additional shells of
backscattering atoms would, in principle, enable further study of
any surface reconstruction upon formation of the Zn adlayer, since
one could determine the registry between the top two layers of
Au atoms within the electrode. A thorough examination of this
phenomenon is precluded for the data presented herein due to
the limitations imposed by the NI: at a minimum, an additional
shell of Au atoms is required to model adsorption in the fcc or hcp
sites of Au(1 1 1) or the 3-fold hollow sites of a reconstructed elec-
trode surface and the associated increase in degrees of freedom
exceeds the available NI from the truncated EXAFS. Nevertheless,
only a small increase in ꢆk and, therefore, S/N, would be neces-
sary for the requisite improvement in NI to model the fifth shell of
atoms. The gain in S/N obtained following the combination of two
or three spectra indicates that the desired improvement would be
achieved via an additional ∼2 h of data collection (∼4 h total). It is
important to emphasize, however, that in the absence of this addi-
tional data collection, the combination of only three XAS spectra
(∼2 h total) was required for assignment of the local environment
about the Zn adatoms with sufficient extent and detail to deter-
mine the structure of the UPD adlayer and co-adsorbed overlayer
of counter-ions.
The 2–4 h period required to record spectra of sufficient quality
for detailed characterization of the Zn(UPD)/Au(1 1 1) system is in
accordance with the preceding study of Cu UPD on Au(1 1 1) [13],
in which a reduced (≥80 min) timeframe was necessary for struc-
tural analysis of UPD monolayers with 2-fold (ꢀsc = 0.67) or 3-fold
(ꢀsc = 1) higher surface concentrations of the metal adatoms. Mean-
while, even the upper bound of 4 h data collection necessary for
extended analysis is competitive with alternative or complemen-
tary techniques, such as XRD/SXS, and also represents a reasonable
timeframe for the routine use of synchrotron facilities.
which results in the apparent asymmetry observed for the peak
centered at ∼2.9 A˚ in the FT of the experimental EXAFS (Fig. 3(b))
and the presence of a shoulder at ∼3.3 A˚ .
When projected across the surface of the Au(1 1 1) electrode,
the best-fitting local environment of the Zn adatoms is con-
sistent with formation of the interfacial structure displayed in
Fig. 4(e). The Zn adatoms form a commensurate (√
3 ×
√
3)R30
◦
monolayer structure within this arrangement, which supports an
ordered overlayer of co-adsorbed counter-ions. Furthermore, each
phosphate anion bridges between three Zn adatoms such that
they also adopt a ( 3 × 3)R30 configuration with respect to
the underlying Au electrode. The assignment of this XAS-derived
surface structure is particularly interesting because it is consis-
tent with accepted compositional and structural information for
the phosphate/Zn(UPD)/Au(1 1 1) system obtained via alternative
techniques: (1) the adlayer exhibits the surface coverage of Zn
√
√
◦
(
ꢀsc = 0.33) reported for the phosphate/Zn(UPD)/Au(1 1 1) system
√
√
◦
in prior voltammetric studies [14–16] and (2) the ( 3 × 3)R30
arrangement of co-adsorbed anions is consistent with the structure
observed with in situ STM [14,17]. The agreement with prior stud-
ies lends support to the commensurate Zn adlayer arrangement
obtained from the experimental EXAFS and reaffirms the value of
the technique for investigating the composition and structure of
the electrified interface.
Analysis of the interatomic separations for the first two shells
of backscatterers provides additional insight into the experimen-
tal system. In the event that the Zn adatoms do not reduce to Zn0
and retain a positive oxidation state within the UPD adlayer, the
1
.94–1.99 A˚ separation exceeds the anticipated bond length for the
+
first shell of backscatterers (Zn –O ∼1.7 A˚ ) and would suggest that

8
538
J.R.I. Lee et al. / Electrochimica Acta 55 (2010) 8532–8538
4
. Conclusions
[7] J. McBreen, J. Solid State Electrochem. 13 (2009) 1051.
[
8] M. Giorgetti, S. Passerini, W.H. Smyrl, S. Mukerjee, X.Q. Yang, J. McBreen, J.
Electrochem. Soc. 146 (1999) 2387.
In situ XAS measurements conducted at the Zn K-edge enable
[
9] M. Giorgetti, S. Mukerjee, S. Passerini, J. McBreen, W.H. Smyrl, J. Electrochem.
Soc. 148 (2001) A768.
assignment of the local environment of Zn adatoms on Au(1 1 1)
electrodes prepared by UPD from phosphate buffer. The structural
arrangement around the Zn is invariant over a wide potential range
within the UPD region. Projection of this arrangement across the Au
electrode indicates that the UPD adlayer adopts a commensurate
[
[
[
10] J.G. Gordon, O.R. Melroy, M.F. Toney, Electrochim. Acta 40 (1995) 3.
11] S. Wu, J. Lipkowski, T. Tyliszczak, A.P. Hitchcock, Prog. Surf. Sci. 50 (1995) 227.
12] T. Tyliszczak, A. Hitchcock, S. Wu, A. Chen, G. Szymanski, J. Lipkowski, Syn-
chrotron Radiat. News 11 (1998) 31.
[13] J.R.I. Lee, R.L. O’Malley, T.J. O’Connell, A. Vollmer, T. Rayment, J. Phys. Chem. C
113 (2009) 12260.
14] M. Nakamura, A. Aramata, A. Yamagishi, M. Taniguchi, J. Electroanal. Chem. 446
(1998) 227.
[15] A. Aramata, S. Taguchi, T. Fukuda, M. Nakamura, G. Horányi, Electrochim. Acta
444 (1998) 999.
16] S. Takahashi, K. Hasebe, A. Aramata, Electrochem. Commun. 1 (1999) 301.
17] S. Takahashi, A. Aramata, M. Nakamura, K. Hasebe, M. Taniguchi, S. Taguchi, A.
Yamagishi, Surf. Sci. 512 (2002) 37.
18] S. Moniwa, A. Aramata, J. Electroanal. Chem. 376 (1994) 203.
19] M.G. Chu, J. McBreen, G. Adzic, J. Electrochem. Soc. 128 (1981) 2281.
20] A. Manzoli, M.C. Santos, S.A.S. Machado, Thin Solid Films 515 (1997) 6860.
√
√
◦
(
3 × 3)R30 (ꢀsc = 0.33) arrangement on the underlying Au sur-
[
face, with the Zn adatoms located in 3-fold hollow adsorption sites.
Analysis of the XAS-derived local environment also demonstrates
√
√
◦
that the co-adsorbed layer of counter-ions adopt a ( 3 × 3)R30
[
[
(
ꢀsc = 0.33) structure, in which each phosphate ion bridges between
three Zn adatoms and the central phosphorous atom resides above
a vacant 3-fold hollow site on the Au surface. Data collection peri-
ods of only ∼2 h per sample were required to allow the assignment
of the local environment of the Zn (with an estimated 4 h neces-
sary for a more extended assignment), which are comparable to the
timescales of more traditional techniques for studying the struc-
ture at the electrode/electrolyte interface. As such, XAS offers a
competitive and complementary in situ technique for the struc-
tural investigation of low concentration species in the vicinity of
the electrode surface.
[
[
[
[21] A.R. Despic, M.G. Pavlovic, Electrochim. Acta 27 (1982) 1539.
[
[
22] M.H. Fonticelli, D. Posadas, R.I. Tucceri, J. Electroanal. Chem. 565 (2004) 359.
23] A. Aramata, M.A. Quaiyyum, W.A. Balais, T. Atoguchi, M. Enyo, J. Electroanal.
Chem. 338 (1992) 367.
[24] M.A. Quaiyyum, A. Aramata, S. Moniwa, S. Taguchi, M. Enyo, J. Electroanal.
Chem. 373 (1994) 61.
[
25] S. Taguchi, A. Aramata, M.A. Quaiyyum, M. Enyo, J. Electroanal. Chem. 374
1994) 275.
[26] S. Taguchi, T. Fukuda, A. Aramata, J. Electroanal. Chem. 435 (1997) 55.
(
[27] G. Horányi, A. Aramata, J. Electroanal. Chem. 434 (1997) 201.
[28] G. Horányi, A. Aramata, J. Electroanal. Chem. 437 (1997) 259.
[29] S. Taguchi, A. Aramata, J. Electroanal. Chem. 457 (1998) 73.
[30] L.D. Burke, T.G. Ryan, Electrochim. Acta 37 (1992) 1363.
Acknowledgements
The authors acknowledge the ESRF for provision of synchrotron
radiation facilities and would like to thank Thomas Neisius for out-
standing assistance in the operation of beamline ID26. JRIL also
thanks the Engineering and Physical Sciences Research Council for
financial support and Dr. Sven L.M. Schröder for helpful discussions.
Portions of this work were performed under the auspices of the U.S.
Department of Energy by Lawrence Livermore National Laboratory
in part under Contract W-7405-Eng-48 and in part under Contract
DE-AC52-07NA27344.
[31] A.A. El-Shafei, J. Electroanal. Chem. 380 (1995) 269.
[
[
32] K. Igarashi, A. Aramata, S. Taguchi, Electrochim. Acta 46 (2001) 73.
33] T. Boiadjieva, M. Monev, A. Tomandl, H. Kronberger, G. Fafilek, J. Solid State
Electrochem. 13 (2009) 671.
[34] G. Adzic, J. McBreen, M.G. Chu, J. Electroanal. Chem. 128 (1981) 1691.
[
35] Y.S. Lo, N.D. Huefner, W.S. Chan, P. Dryden, B. Hagenhoff, T.P. Beebe, Langmuir
5 (1999) 6522.
36] Y. Golan, L. Margulis, I. Rubenstein, Surf. Sci. 264 (1992) 312.
1
[
[37] M.G. Samant, M.F. Toney, G.L. Borges, L. Blum, O.R. Melroy, Surf. Sci. 1–2 (1988)
L29.
[
[
[
[
[
38] E.A. Stern, Phys. Rev. B 48 (1993) 9825.
39] N. Binsted, PAXAS: Program for the Analysis of X-ray Absorption Spectra, 1988.
40] J.W. Cook, D.E. Sayers, J. Appl. Phys. 52 (1981) 5024.
41] N. Binsted, EXCURV98: CCLRC Daresbury Laboratory Computer Program, 1998.
42] A. Tadjeddine, D. Guay, M. Ladouceur, G. Tourillon, Phys. Rev. Lett. 66 (1991)
Appendix A. Supplementary data
2235.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.electacta.2010.07.046.
[43] A. Tadjeddine, G. Tourillon, D. Guay, Electrochim. Acta 36 (1991) 1859.
[
[
44] A. Tadjeddine, G. Tourillon, Sov. Electrochem. 29 (1993) 82.
45] K.-O. Thiel, A. Vollmer, M. Hintze, E. Avci, C. Donner, Z. Phys. Chem. 221 (2007)
1255.
References
[46] A. Vollmer, E. Avci, K. Thiel, C. Donner, J. Electroanal. Chem. 605 (2007) 15.
[
[
[
[
[
[
[
47] S. Trasatti, J. Electroanal. Chem. 33 (1971) 351.
48] H. Gerischer, D.M. Kolb, M. Przasnyski, Surf. Sci. 43 (1974) 662.
49] D.M. Kolb, H. Gerischer, Surf. Sci. 51 (1975) 323.
50] J. Wang, A.J. Davenport, H.S. Isaacs, B.M. Ocko, Science 255 (1992) 1416.
51] J. Wang, B.M. Ocko, A.J. Davenport, H.S. Isaacs, Phys. Rev. B 46 (1992) 10321.
52] D.M. Kolb, Prog. Surf. Sci. 51 (1996) 109.
[
[
1] D.M. Kolb, J. Electroanal. Chem. 54 (1974) 25.
2] D.M. Kolb, M. Przasnyski, H. Gerischer, Advances in Electrochemistry and Elec-
trochemical Engineering, John Wiley and Sons Ltd., 1978.
3] S. Szabó, Int. Rev. Phys. Chem. 10 (1991) 207.
[
[
[
[
4] E. Herrero, L.J. Buller, H.D. Abru n˜ a, Chem. Rev. 101 (2001) 1897.
5] S.J. Gurman, N. Binsted, I. Ross, J. Phys. C 17 (1984) 143.
6] S.J. Gurman, J. Phys. C: Solid State Phys. 21 (1988) 3699.
53] K.G. Huang, D. Gibbs, D.M. Zehner, A.R. Sandy, S.G.J. Mochrie, Phys. Rev. Lett. 65
(
1990) 3313.