12640 Phys. Chem. Chem. Phys., 2011, 13, 12640–12645 This journal is c the Owner Societies 2011

Cite this: Phys. Chem. Chem. Phys., 2011, 13, 12640–12645

XPS revelation of tungsten edges as a potential donor-type catalyst

Yanguang Nie,aXi Zhang,

aShouzhi Ma,

aYan Wang,

bJisheng Pan

cand

Chang Q. Sun*ade

Received 27th January 2011, Accepted 13th May 2011

DOI: 10.1039/c1cp21421g

We report an efficient yet simple technology of photoelectron spectroscopic purification for

identifying the capability of, and direction of charge flow in, a catalyst in a reaction, which has

enabled the finding, for the first time, of the similarity of the valence band of tungsten edges to

that of Rh adatoms and Ag/Pd alloy and hence suggested that W undercoordinated atoms could

be a suitable candidate for replacing the costly Rh adatoms and Ag/Pd alloy as a cheaper, richer,

and efficient donor-type catalyst for CO and NO oxidation applications. The new technology and

new findings will be stimulating to the community for new catalyst design and identification and

provide a better understanding of the electronic process of a catalytic reaction associated with

undercoordinated atoms.

1. Introduction

Atomic undercoordination associated with vacancies, defects,

terrace edges, and nanostructures of various shapes demon-

strate excellent properties, such as the extremely high catalytic

ability that cannot be seen even from a flat surface of the same

specimen, such as Au, Rh, and Pt.1 The altered local structural

and electronic environment modifies the bond length, bond

energy, potential trap depth, and hence the Hamiltonian, work

function, electroaffinity, and the atomic cohesive energy that

locally determine the performance of a material, such as the

catalytic, electronic, dielectric, optic, magnetic, mechanical,

and thermal properties.2 Thus, understanding the bonding and

the energetic behavior of electrons localized in atomic-scaled

zones surrounding undercoordinated atoms is the key for one

to harness the process of catalytic reaction.

X-Ray photoelectron spectroscopy (XPS) is a powerful tool

for detecting the energetic behavior of electrons in the valance

band and below, showing the fingerprints of the crystal

potential change with the local chemical and coordination

environment and its consequences on the electronic energy and

structure in the deeper core bands.3 Generally, XPS data can

be decomposed into several components corresponding to

contributions from bulk (B) and surfaces (Si, i = 1,2. . .) of

different atomic layers in sequence. Three kinds of surface core

level shift (SCLS), i.e., positive, negative, and mixed shift are

generally assumed for the component assignment. Fig. 1a

illustrates the positive SCLS, which means that the B and Si(i = 1,2,. . .) are arranged in the sequence of S1, S2,. . ., and B

from lower (larger absolute value of energy) to higher binding

energy in the XPS profile.4 A represents the entrapped

components due to the undercoordinated adatoms or edge

atoms and P represents the screened polarization states.

The SCLS is often attributed to the ‘‘initial–final’’ states5 or

the ‘‘surface bond contraction’’6 effects. The latter has been

confirmed experimentally from Ta,7 Nb,8 and Mo9 surfaces

and Au10 and Cu11 atomic clusters using XPS and low-energy

electron diffraction. The first interlayer spacing of a W(320)

surface has been found to contract by up to 25%.12 However,

decomposing the XPS spectra with derivatives of quantitative

information of bonds and electronic energy has long been

problematic because of the lack of constraints for the XPS

profile deconvolution: (i) the number of components in one

XPS spectrum; (ii) the energy separation between the components

and their correlations; (iii) the reference point from which the

core level shifts; and (iv) the direction of the energy shift

upon surface and edge formation. Establishment of these

constraints is highly needed, in addition to a consistent under-

standing of the nature and the physical origin of the SCLS and

their indications in catalyst design.

The SCLS ofW(100), (110), and (111) surfaces13 and their (320)

and (540) vicinals14 have been well measured for more than 30

years using different techniques, such as the synchrotron and

XPS, unfortunately with discrepancies in the assignment of the

direction of the energy shifts and the energy of the bulk

component.13b,d,14b,15 A negative shift was assigned with the bulk

component at 31.4 eV and the surface component at 31.1 eV.

a School of Electrical and Electronic Engineering,Nanyang Technological University, Singapore 639798.E-mail: ECQSun@ntu.edu.sg

b School of Information and Electronic Engineering,Hunan University of Science and Technology, Xiangtan 411201,China

c Institute of Materials Research and Engineering, Agency for Science,Technology and Research (A*STAR), Singapore 117602

d School of Materials Science, Jilin University, Changchun 130012,China

e Faculty of Materials, Photoelectronics and Physics,Xiangtan University, Changsha 400073, China

PCCP Dynamic Article Links

www.rsc.org/pccp PAPER

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 24

Jul

y 20

11Pu

blis

hed

on 1

4 Ju

ne 2

011

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

P214

21G

View Online

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 12640–12645 12641

Much more information regarding the interatomic binding

energy and the effect of undercoordination is supposed to be

given by the sophisticated measurements. For the past 30 years,

progress towards such information has been limited because of

the lack of suitable theories and decoding technologies.

Recently, we incorporated the bond order–length–strength

(BOLS) correlation mechanism into the tight-binding theory16

(BOLS-TB) to examine the CLS of undercoordinated systems

from the perspective of Hamiltonian perturbation. The BOLS-TB

algorithm has enabled quantitative information about the

bonds and electronic energy from the SCLS of Ru,17 Be,18

Rh and Pd surfaces,19 carbon allotropes,16,17 Pt and Rh

adatoms,19 and Cu/Pd and Ag/Pd nanoalloys.20 It has been

consistently confirmed that the shorter and stronger bonds

between undercoordinated atoms derive globally positive

CLS because of the undercoordination-induced quantum

entrapment. The occupancy of the polarized states at the

upper edge of the valence band dictates the direction of charge

flow between the catalyst and the reactant in the process of

catalytic reaction.19,20

The objective of this work is to show that an incorporation

of the BOLS-TB algorithm and the recently-developed atomistic

photoelectron distillation spectroscopy (APDS) purification

method into the well-measured XPS data 13d,14b,21 has enabled

the discrimination of the edge states from those of the surface

and the bulk, which suggests that the W edge could perform

the same as Rh adatoms and AgPd alloy as a donor-type

catalyst with quantitative information about the 4f level

energy of an isolated W atom and its bulk shift, as well as

the local bond length, binding energy density, and the atomic

cohesive energy.

2. Principles

According to the BOLS-TB algorithm,22 a specific nth energy

level of a specimen and its shift upon interatomic interaction

being involved follow the relations,18

DEvðzÞ ¼ EvðzÞ �Evð0Þ ¼ az þ zbz ðCore level shiftÞEvð0Þ ¼ �hn; ijVatomðrÞjn; ii ðAtomic core levelÞ

az ¼ �hn; ijVcrystðrÞð1þDHÞjn; ii / Ez ðExchange integralÞbz ¼ �hn; ijVcrystðrÞð1þDHÞjn; ji / Ez ðOverlap integralÞ

8>><>>:

ð1Þ

where |n,ii is the specifically localized Bloch wave function at a

specific ith atomic site. DH is the perturbation to the Hamiltonian

due to the effect of bond-order loss. The intraatomic potential,

Vatom(r), intrinsically defines the core level of an isolated atom,

Ev(0); the interatomic potential, Vcryst(r), determines the shift

of the core level, DEv (z = 12), when the bulk is formed.

The z is the effective coordination number (CN or z) of the

considered atom. For a bcc bulk, the effective z value is

12 rather than 8 for normalization purposes. The Ev(0) and

DEv (z = 12) are intrinsic constants that do not change with

the environment. The Ev(12) moves deeper or positively with

respect to the Ev(0) because of the additional Vcryst(r). Most

importantly, the energy shift of DEv(z) is proportional to the

cohesive energy per bond at equilibrium, Ez. Any perturbation

to the interatomic potential or the bond energy will lead to the

energy shift.

According to the BOLS scheme, bonds between under-

coordinated atoms are shorter and stronger. Local quantum

entrapment and densification of bonding charge and binding

energy will happen, which perturbs the interatomic potential.

The core level shift of the undercoordinated atoms is therefore

deeper than that of the bulk. On the other hand, polarization

of the electrons at the upper-edge of the valence band by the

locally, densely, and entrapped core electrons also provide

perturbation to the Hamiltonian through screening and splitting

the crystal potential, and thus,

Fig. 1 (a) Illustration of the XPS spectral components with positive

CLS caused by atomic undercoordination. The P, B, Si, and A

components represent, respectively, contributions from the polarization,

bulk, surfaces, and adatoms. (b) The surface atomic structures

of the (110) vicinal (540) and (320) surfaces with edge atoms

density. The edge density of (320) is 0.28 monolayer (ML) and the

(540) is 0.16 ML.

1þ DH ¼C�mz ¼ Ez=E0 ðTrap depressionÞ

p ¼ ðEvðpÞ � Evð0ÞÞ=ðEvð12Þ � Evð0ÞÞ ðPolarizationÞ

(

Cz ¼ dz=d0 ¼ 2=f1þ exp½ð12� zÞ=ð8zÞ�g ðBond contraction coefficientÞ

ð2Þ

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 24

Jul

y 20

11Pu

blis

hed

on 1

4 Ju

ne 2

011

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

P214

21G

View Online

12642 Phys. Chem. Chem. Phys., 2011, 13, 12640–12645 This journal is c the Owner Societies 2011

En(p) represents the peak energy of the polarization component in

the XPS. Cz is the Goldschmidt–Pauling coefficient of bond

contraction. E0 and d0 represent, respectively, the bond energy

and bond length in the ideal bulk. The bond nature indicator

m has been optimized to be unity for metals. The p is the

coefficient of polarization to be determined from the XPS

measurement. Thus, correlation between the XPS components

follows the criterion,

EvðxÞ � Evð0ÞEvð12Þ � Evð0Þ

¼ C�1z ðEntrapmentÞp ðPolarizationÞ

�ð3Þ

The x represents z or p. If the polarization-entrapment

coupling effect is apparent, the term C�1z is then replaced by

pC�1z , the trapped states will be moved up from the otherwise

low-z position to energy close to the bulk component. For

situations without apparent polarization, the relation evolves,

EvðzÞ � Evð0ÞEvðz0Þ � Evð0Þ

¼ Cz0

Cz; or; Evð0Þ ¼

Cz0Evðz0Þ � CzEvðzÞCz0 � Cz

ð4Þ

This relation allows us to determine the Ev(0) for an isolated

atom and the bulk shift, which have been a long pursuit of the

community. The accuracy of the determined Ev(0) depends on

the database size collected from the same materials. If there

are a total of n components of B and Si components for

various surfaces of the specimen, there will be a combination

of C2n = n!/(2!(n � 2)!) possible Ev(0) values. In the present

case, n = 7 and C2n = 21, as given in Table 1. By taking the

average of the Ev(0) and the standard deviation, s, we have theexpression for the z-resolved CLS for the XPS components:

E4f(z) = hE4f(0)i � s + [E4f(12) � E4f(0)]/Cz (5)

Besides, with the derived z values for various surface and

subsurface layers of differ crystal orientations, we are able to

elucidate the z-resolved local strain, binding energy density,

and the cohesive energy per discrete atom in the surface

skins.18 Such information is particularly fundamentally important

for one to understand and control processes at sites surrounding

undercoordinated atoms.

One can imagine what will happen if we subtract the XPS

spectrum collected from the flat surface from that collected

from the edged surface under the same measurement

conditions, upon background correction and spectral area

normalization of both. The residual spectrum keeps only the

features due to the edge atoms within zones of only a one or

two atomic layer size. Such an APDS process enables the

purification of the edge states as the APDS filters out the

artifact background and bulk information, such as the back-

ground uncertainty and the ‘‘initial–final states’’ effect that

exists throughout the course of measurements.23

3. Results and discussion

3.1 The APDS of edge atoms

Fig. 1b shows the atomic arrangement of the (320) and (540)

surfaces with a considerable fraction of undercoordinated

atoms compared with the flat (110) surface. From the original

W 4f spectra shown in Fig. 2a for the (110), (320) and (540)

Table 1 BOLS elucidated information regarding the atomic-layer (S1, S2) and crystal-orientation resolved effective CN (z), local strain (Cz � 1),the relative binding energy density (C�4z ) and atomic cohesive energy (zibC

�1z ) from the measured XPS W 4f SCLS. The zib = zi/zb is the relative

coordination number. The spectral deconvolution using the BOLS-TB algorithm derives the energy level of an isolated W atom as E4f(0) = 28.910 �0.006 eV and its bulk shift DE4f(12) = 2.173 eV with the z-resolved CLS: E4f(z) = 28.910 � 0.006 + 2.173C�1z for surface and edge atoms

i E4f(i) (eV) z Cz � 1 (%) E-density [C�4z ] DEC(i)/EC(B) [zibC�1z � 1](%)

W B 31.083 12 0 1 0W(100)21 S2 31.283 5.161 �8.26 1.41 �53.12

S1 31.398 3.970 �12.57 1.71 �62.16W(110)14b S2 31.240 5.829 �6.61 1.31 �47.99

S1 31.402 3.942 �12.71 1.72 �62.36W(111)13d S2 31.275 5.270 �7.96 1.39 �52.28

S1 31.370 4.195 �11.58 1.64 �60.46

Fig. 2 From the normalized the (110), (540), and (320) XPS 4f7/2

spectra24 (a) one can hardly discriminate the contribution of the edge

atoms from those of the un-edged (110) surface; but (b) the APDS, or

subtraction of the un-edged from the edged W(540) and (320) surfaces,

can resolve the edge states unambiguously with the P and P + T extra

states and the B and the additional valley at the bottom edge.

The resultant APDS almost satisfies the criterion of spectral area

conservation.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 24

Jul

y 20

11Pu

blis

hed

on 1

4 Ju

ne 2

011

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

P214

21G

View Online

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 12640–12645 12643

surfaces collected under the same experimental conditions,24

one can hardly discriminate the spectral features from one

another; the APDS process, however, makes a great difference,

as shown in Fig. 2b, proceeded by subtracting the spectrum

collected from the un-edged (110) surface from the edged ones

of the (540) and (320) surface after the standard process of

spectral normalization and background correction using the

Shirley method. The correctness of the APDS outcome is

justified by the criterion of spectral area conservation, which

means that the spectral loss and gain should be identical. The

normalization of the spectra is to minimize the artifacts due to

scattering of the X-rays by the edge atoms. Unexpectedly, two

extra components centered at 30.945 and 31.310 eV are present

and two valleys centered at 31.083 and 31.454 eV are generated.

The emergence of these spectral features indicate that the

electronic structure for the edge atoms is indeed different from

those at the bulk interior or at the flat surface, but at this

moment one can hardly tell why.

3.2 SCLS analysis of the W(110), (100) and (111) surfaces

In order to calibrate and understand the APDS features

in Fig. 2b, we need to decompose the XPS spectra from the

well-faceted surfaces with respect to the reported best fits19

using three components, the bulk, B, and the second and the

first surface layers, S2 and S1. Experimental conditions, such as

the incident beam energy and the emission angle, may change

the spectral appearance because of the X-rays’ penetration

depth. A spectrum collected at larger emission angles or with

lower incident beam energy collects information dominated by

the shallow surface, otherwise more bulk information is

collected. Varying experimental conditions can never change

the intrinsic properties of the surfaces, such as the atomic

coordination numbers of the surface and sub-surfaces, that are

the key factors used herewith. The order of the B, S2, and S1components and the separation between them follows the

constraint given in eqn (4). The decomposed z-resolved

components of the W(100), W(110), and W(111) 4f7/2 spectra

in Fig. 3 show that the atomic CN reduction leads to the

positive CLS. The CNs of the S1 and the S2 components across

the three W surfaces vary slightly because of the anisotropy of

crystal structure and atomic density.13,14 Deconvolution of the

three surfaces aims to enhance the accuracy in determining

the E4f7/2(0) and the DE4f7/2

(12) by minimizing the standard

deviation value, s. From the BOLS-TB enabled deconvolution,

we derived the following information: (i) the W 4f energy level of

E4f7/2(0) = 28.910 � 0.006 eV for an isolated W atom and its bulk

shift of DE4f7/2(12) = 2.173 eV; (ii) an analytical expression for

the z-resolved CLS: E4f7/2(z) = 28.910 + 2.173C�1z for the under-

coordinated edge and adatoms; (iii) the effective atomic CN of the

sublayers of different orientations and their derivatives on the local

bond strain, local bond energy, the ratio of binding energy density

and the atomic cohesive energy to the respective bulk values, as

summarized in Table 1.

3.3 Edge CLS purification

Based on the derived CN-dependent SCLS, E4f7/2(z) = 28.910 �

0.006 + 2.173C�1z , we can incorporate the z values into the

APDS spectra in Fig. 2b and hence clarify the origin of the

extra spectral features:

(i) The B valley at 31.083 eV = 28.910 + 2.173 eV is

unambiguously the bulk component that was defined in the

APDS without needing intuitive assignment. This finding

clarifies the long confusion13c regarding bulk component that

was always assumed at the bottom edge of the 4f band.

(ii) The additional T states below the bulk component are

the entrapped states because of the undercoordination-induced

bond strain and bond strength gain. The locally entrapped T

states polarize the valence electrons that cannot be directly

Fig. 3 Deconvolution of the XPS 4f7/2 spectra of (a) W(100),21

(b) (110),14b and (c) (111)13d surfaces using the B, S2 and S1 components

with the derived 4f level of an isolated W atom as E4f(0) = 28.910 �0.006 eV and its shift upon bulk formation, DE4f(12) = 2.173 eV, both

of which are intrinsic constants changing with neither the experimental

conditions nor the crystal orientations. The component energies

follows: DE4f(z) :DE4f(12) = C�1z with Cz being the bond contraction

coefficient. Deconvolution also confirms the positive SCLS and leads

to the quantitative information, as listed in Table 1. The baselines are

the standard spectral background correction.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 24

Jul

y 20

11Pu

blis

hed

on 1

4 Ju

ne 2

011

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

P214

21G

View Online

12644 Phys. Chem. Chem. Phys., 2011, 13, 12640–12645 This journal is c the Owner Societies 2011

detected using XPS, but it happens to the Rh and Ag/Pd alloy,

as confirmed using ultra-violet photoelectron spectroscopy,

showing the consistency of charge polarization direction in all

the bands of the same specimen.20,22

(iii) The P component at 30.945 eV results from the screening

and splitting of the crystal potential by the polarized valence

electrons. We can estimate the polarization coefficient with the

known energies of the P and the B components, p = [E4f7/2(p) �

E4f7/2(0)]/[E4f7/2

(12) � E4f7/2(0)] = (30.945 � 28.910)/2.173 =

93.6%, which means that the crystal potential has been

partially screened and elevated by 6.4% of the bulk value.

The otherwise T component turns out to be T + P with an

additional valley at the bottom edge of the core band because

of the coupling effect of entrapment and polarization. The T

component is supposed to add a component at energy corres-

ponding to z o 4, if the polarization is absent or it is

sufficiently weak. The state loss (second valley) at 31.454 eV

with an effective CN of 3.57 is supposed to be absent; the

strong interaction between edge atoms should enhance the

intensity of the states at lower-z positions instead, if no

polarization happens. However, as we discussed, the screening

effect also applies to the trapped states, and therefore, this

valley becomes present, and the T component becomes P + T.

The C�13 is replaced with pC�13 = C�13.75, which means that the

original edge states located at z = 3 shift up to energy being

equivalent to z = 3.75. The edge bond is strengthened by

[E4f7/2(T + P) � E4f7/2

(0)]/[E4f7/2(12) � E4f7/2

(0)] = (31.310 �28.910)/2.173 = 1.104, or 10.4%, because of the joint effect of

entrapment and polarization. It should be 1.104/p = 1.104/

0.936 = 1.18 instead, if no polarization occurs.

3.4 Potential catalytic behavior

The extra P and the P + T states in the APDS are due to the

edge atoms only, as the APDS has filtered out the background

and bulk information. It has been confirmed that the valence

and the core electrons of a specimen shift simultaneously in the

same direction because of the screening effect to the core

charge, such as the cases of AgPd and CuPd bimetallic alloy

catalysts,20 and the Pt and Rh adatoms.22 The APDS of W

edges share the same attribute to those of Rh adatoms and

AgPd alloy, as shown in Fig. 4. The latter have been identified

as donor-type catalysts compared to the Pt adatoms and CuPd

alloy that are the opposite, because of the respective

dominance of polarization and entrapment effects. From the

electronic structure, we can suggest that W edges should

perform the same as Rh and Ag/Pd as n-type catalysts, though

experimental confirmation is needed. Nevertheless, the APDS

can help us to search for new catalysts and identify the

catalytic nature of existing catalysts.

4. Conclusion

The BOLS-TB enabled XPS deconvolution of the W(100),

(110) ,and (111) surfaces has led to quantitative information

about the 4f energy level of an isolated W atom as 28.910 eV

and its bulk shift of 2.173 eV. The positive core-level shift

originates from the stronger and shorter bonds between

undercoordinated atoms, which follow the prediction of the

Goldschmidt–Pauling rule of bond contraction and the theory

of BOLS correlation. The deconvolution provides profound

information about the effective atomic CN, local bond strain,

bond energy, binding energy density and the atomic cohesive

energy in the surface skin of up to two atomic layers of

different orientations. Further APDS processing revealed

extra features of the bulk valley, polarization and the joint

effect of entrapment and polarization and their physical

indications, which confirm the positive core-level shift and

clarify the long confusion in the bulk component assignment.

Results show the BOLS expectation of the edge states as

resulting from the undercoordination-induced local bond

contraction and the associated quantum entrapment and

densification of core electrons, and the polarization of

the otherwise conducting electrons of W edge atoms by the

entrapped core charge. Most strikingly, being similar to the

spectral features of Rh adatoms and Ag/Pd alloy, the W edge

is suggested to serve as a donor-type catalyst. If it works, the

impact to catalytic industrial processes would be enormous.

As demonstrated, the APDS should provide a powerful tool

for one to purify information from atomic-scaled zones

surrounding undercoordinated atoms regarding the local bond

and energetic behavior of electrons, which is helpful for us to

search for new catalysts.

References

1 (a) S. W. Lee, S. Chen, J. Suntivich, K. Sasaki, R. R. Adzic andY. Shao-Horn, Role of Surface Steps of Pt Nanoparticles on theElectrochemical Activity for Oxygen Reduction, J. Phys. Chem.Lett., 2010, 1(9), 1316–1320; (b) G. Jones, M. J. Gladys, J. Ottal,S. J. Jenkins and G. Held, Surface geometry of Cu{531}, Phys.Rev. B: Condens. Matter Mater. Phys., 2009, 79(16), 165420.

2 (a) A. l. Dıez, J. Fernandez, E. Lalinde, M. T. Moreno andS. Sanchez, High-Nuclearity Pt–Tl–Fe Complexes: Structural,Electrochemistry, and Spectroelectrochemistry Studies, Inorg.Chem., 2010, 49(24), 11606–11618; (b) Y. Niimi, H. Kambaraand H. Fukuyama, Localized Distributions of Quasi-Two-Dimensional Electronic States near Defects Artificially Createdat Graphite Surfaces in Magnetic Fields, Phys. Rev. Lett., 2009,102(2), 026803; (c) X. L. Ji, K. T. Lee, R. Holden, L. Zhang,J. J. Zhang, G. A. Botton, M. Couillard and L. F. Nazar,Nanocrystalline intermetallics on mesoporous carbon for directformic acid fuel cell anodes, Nat. Chem., 2010, 2(4), 286–293;(d) T. Capron, Y. Niimi, F. Mallet, Y. Baines, D. Mailly, F. Y. Lo,

Fig. 4 The APDS for Rh adatoms with the inset of Pt adatoms22 for

comparison with Fig. 2b. The spectral similarity of the Rh adatoms

and W edge atoms suggests that the W edge may perform the same to

Rh adatoms as a donor-type catalyst.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 24

Jul

y 20

11Pu

blis

hed

on 1

4 Ju

ne 2

011

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

P214

21G

View Online

This journal is c the Owner Societies 2011 Phys. Chem. Chem. Phys., 2011, 13, 12640–12645 12645

A. Melnikov, A. D. Wieck, L. Saminadayar and C. Bauerle,Low-temperature dephasing in irradiated metallic wires, Phys.Rev. B: Condens. Matter Mater. Phys., 2008, 77(3), 033102.

3 E. Sacher, Asymmetries in Transition Metal XPS Spectra: MetalNanoparticle Structure, and Interaction with the Graphene-Structured Substrate Surface, Langmuir, 2010, 26(6), 3807–3814.

4 F. Matsui, T. Matsushita, Y. Kato, M. Hashimoto, K. Inaji,F. Z. Guo and H. Daimon, Atomic-layer resolved magnetic andelectronic structure analysis of Ni thin film on a Cu(001) surface bydiffraction spectroscopy, Phys. Rev. Lett., 2008, 100(20), 207201.

5 A. Howard, D. N. S. Clark, C. E. J. Mitchell, R. G. Egdell andV. R. Dhanak, Initial and final state effects in photoemission fromAu nanoclusters on TiO2(110), Surf. Sci., 2002, 518(3), 210–224.

6 R. A. Bartynski, D. Heskett, K. Garrison, G. Watson,D. M. Zehner, W. N. Mei, S. Y. Tong and X. Pan, The 1stinterlayer spacing of Ta(100) determined by photoelectrondiffraction, J. Vac. Sci. Technol., A, 1989, 7(3), 1931–1936.

7 D. M. Riffe and G. K. Wertheim, TA(110) Surface and subsurfacecore-level shifts and 4F7/2 line-shapes, Phys. Rev. B: Condens.Matter, 1993, 47(11), 6672–6679.

8 B. S. Fang, W. S. Lo, T. S. Chien, T. C. Leung, C. Y. Lue,C. T. Chan and K. M. Ho, Surface band structures on Nb(001),Phys. Rev. B: Condens. Matter, 1994, 50(15), 11093–11101.

9 N. Wu, Y. B. Losovyj, Z. Yu, R. F. Sabirianov, W. N. Mei,N. Lozova, J. A. C. Santana and P. A. Dowben, The surfacerelaxation and band structure of Mo(112), J. Phys.: Condens.Matter, 2009, 21(47), 474222.

10 (a) W. J. Huang, R. Sun, J. Tao, L. D. Menard, R. G. Nuzzo andJ. M. Zuo, Coordination-dependent surface atomic contraction innanocrystals revealed by coherent diffraction, Nat. Mater., 2008,7(4), 308–313; (b) J. T. Miller, A. J. Kropf, Y. Zha,J. R. Regalbuto, L. Delannoy, C. Louis, E. Bus and J. A. vanBokhoven, The effect ofgold particle size on Au–Au bond lengthand reactivity toward oxygen in supported catalysts, J. Catal.,2006, 240(2), 222–234; (c) V. Petkov, Y. Peng, G. Williams,B. H. Huang, D. Tomalia and Y. Ren, Structure of goldnanoparticles suspended in water studied by X-ray diffractionand computer simulations, Phys. Rev. B: Condens. Matter Mater.Phys., 2005, 72(19), 195402.

11 D. Q. Yang and E. Sacher, Initial- and final-state effects on metalcluster/substrate interactions, as determined by XPS: copperclusters on Dow Cyclotene and highly oriented pyrolytic graphite,Appl. Surf. Sci., 2002, 195(1–4), 187–195.

12 J. H. Cho, D. H. Oh and L. Kleinman, Core-level shifts of lowcoordination atoms at the W(320) stepped surface, Phys. Rev. B:Condens. Matter, 2001, 64(11), 115404.

13 (a) G. K. Wertheim and P. H. Citrin, Surface-atom core-level shiftsof W(111), Phys. Rev. B, 1988, 38, 7820; (b) A. M. Shikin,A. Varykhalov, G. V. Prudnikova, V. K. Adamchuk, W. Gudatand O. Rader, Photoemission from Stepped W(110): Initial or

Final State Effect?, Phys. Rev. Lett., 2004, 93, 146802;(c) D. M. Riffe, B. Kim, J. L. Erskine and N. D. Shinn, Surfacecore-level shifts and atomic coordination at a stepped W(110)surface, Phys. Rev. B: Condens. Matter, 1994, 50, 14481;(d) K. G. Purcell, J. Jupille, G. P. Derby and D. A. King,Identification of underlayer components in the surface core-levelspectra of W(111), Phys. Rev. B, 1987, 36, 1288.

14 (a) D. Chauveau, P. Roubin, C. Guillot, J. Lecante, G. Treglia,M. C. Desjonqueres and D. Spanjaard, Surface core level spectro-scopy of the stepped surface, Solid State Commun., 1984, 52(7),635–639; (b) X. Zhou and J. L. Erskine, Surface core-level shifts atvicinal tungsten surfaces, Phys. Rev. B: Condens. Matter Mater.Phys., 2009, 79, 155422.

15 K. G. Purcell, J. Jupille and D. A. King, Coordination number andsurface core-level shift spectroscopy: Stepped tungsten surfaces,Surf. Sci., 1989, 208(3), 245–266.

16 M. A. Omar, Elementary Solid State Physics: Principles andApplications, Addison-Wesley, New York, 1993.

17 Y. Wang, Y. G. Nie, L. L. Wang and C. Q. Sun, Atomic-Layer- and Crystal-Orientation-Resolved 3d(5/2) Binding EnergyShift of Ru(0001) and Ru(1010) Surfaces, J. Phys. Chem. C, 2010,114(2), 1226–1230.

18 Y. Wang, Y. G. Nie, J. S. Pan, L. Pan, Z. Sun and C. Q. Sun, Layerand orientation resolved bond relaxation and quantum entrapmentof charge and energy at Be surfaces, Phys. Chem. Chem. Phys.,2010, 12, 12753–12759.

19 Y. Wang, Y. G. Nie, J. S. Pan, L. K. Pan, Z. Sun, L. L. Wang andC. Q. Sun, Orientation-resolved 3d(5/2) binding energy shift of Rhand Pd surfaces: anisotropy of the skin-depth lattice strain andquantum trapping, Phys. Chem. Chem. Phys., 2010, 12(9),2177–2182.

20 C. Q. Sun, Y. Wang, Y. G. Nie, B. R. Mehta, M. Khanuja,S. M. Shivaprasad, Y. Sun, J. S. Pan, L. K. Pan and Z. Sun,Interface quantum trap depression and charge polarization in theCuPd and AgPd bimetallic alloy catalysts, Phys. Chem. Chem.Phys., 2010, 12(13), 3131–3135.

21 J. Jupille, K. G. Purcell and D. A. King, W{100} clean surfacephase transition studied by core-level-shift spectroscopy:Order-order or order-disorder transition, Phys. Rev. B, 1989,39(10), 6871.

22 C. Q. Sun, Y. Wang, Y. G. Nie, Y. Sun, J. S. Pan, L. K. Pan andZ. Sun, Adatoms-Induced Local Bond Contraction, QuantumTrap Depression, and ChargePolarization at Pt and Rh Surfaces,J. Phys. Chem. C, 2009, 113(52), 21889–21894.

23 N. Pauly and S. Tougaard, Surface and core hole effects in X-rayphotoelectron spectroscopy, Surf. Sci., 2010, 604(13–14),1193–1196.

24 X. B. Zhou and J. L. Erskine, Surface core-level shifts at vicinaltungsten surfaces, Phys. Rev. B: Condens. Matter Mater. Phys.,2009, 79(15), 155422.

Dow

nloa

ded

by N

anya

ng T

echn

olog

ical

Uni

vers

ity o

n 24

Jul

y 20

11Pu

blis

hed

on 1

4 Ju

ne 2

011

on h

ttp://

pubs

.rsc

.org

| do

i:10.

1039

/C1C

P214

21G

View Online