Available online at www.sciencedirect.

com

surface science
reports
Surface Science Reports 68 (2013) 273–304
www.elsevier.com/locate/surfrep

Review

The interpretation of XPS spectra: Insights into materials properties

Paul S. Bagusa,n, Eugene S. Iltonb, Connie J. Nelinc
a
Department of Chemistry, University of North Texas, Denton, TX 76203-5017, USA
b
Pacific Northwest National Laboratory, Richland, WA 99352, USA
c
C. J. Nelin Consulting, Austin, TX, USA

Received 14 January 2013; received in revised form 14 March 2013; accepted 14 March 2013

editor: H. Freund

Abstract

We review basic and advanced concepts needed for the correct analysis of XPS features. We place these concepts on rigorous foundations and
explore their physical and chemical meanings without stressing the derivation of the mathematical formulations, which can be found in the cited
literature. The significance and value of combining theory and experiment is demonstrated by discussions of the physical and chemical origins of
the main and satellite XPS features for a variety of molecular and condensed phase materials.
& 2013 Elsevier B.V. All rights reserved.

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 273
2. Theoretical models and methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 276
3. One-electron features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
3.1. Initial and final state effects and Koopmans' theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
3.2. Core-hole screening and relaxation effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 281
3.3. Initial state mechanisms leading to BE shifts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
3.4. BE shifts in molecules, clusters, and solids. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 283
4. XPS satellites: covalent interactions and screening . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
4.1. XPS satellites of chemisorbed molecules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
4.2. XPS satellites for the NO dimer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
5. Metal oxides and covalency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
6. XPS satellite features and many-body effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
6.1. Multiplet splitting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
6.2. Near-degeneracy effects: angular momentum recoupling and FACs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
6.3. Satellites in closed shell oxides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 300
Acknowledgments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
Appendix A. List of acronyms and abbreviations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301

1. Introduction

n
Corresponding author. The fundamental relationship in Photoemission Spectro-
E-mail address: [email protected] (P.S. Bagus). scopy, PES, follows from the photoelectric effect established

0167-5729/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.surfrep.2013.03.001
274 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

by Einstein. In its simplest form, this relationship is relatively low intensity, which are not shown [19]. Much more
BE ¼ hν–KE; ð1Þ complex XPS spectra are found for the core-level spectra of
CeO2 [20–22]. In Fig. 2, the Ce 4s XPS of CeO2 is shown
where hν is the energy of the incident photon, KE is the kinetic [21]. Since CeO2 is a closed shell oxide, one would have
energy of the photoemitted electron, and BE is the binding expected a single peak; instead, a distinct triplet is observed.
energy, or ionization potential, IP, of the photoemitted There is agreement that the complex CeO2 XPS spectra cannot
electron. It is important to stress that the BE is not a one be explained by a one-electron model, as was the case for the
electron quantity; rather it is the difference in the energies of Ne XPS (Fig. 1). In order to explain the CeO2 XPS, it is
the N electron initial, un-ionized, state and the N−1 electron necessary to use a theoretical framework where several
final, ionized, state: electronic configurations contribute to the XPS spectra.
BE ¼ EN−1 ðfinalÞ−EN ðinitialÞ; ð2Þ Because they rely on several configurations, these theoretical
methods are described as many-body to distinguish them from
where, by convention, BE 40 for a bound state. In order to
one-body methods where a single electronic configuration for
obtain absolute BEs, corrections to Eq. (1) may include a work
the photoionized system is associated with each XPS peak.
function and an energy reference for the BE. For an explicit
However, as we discuss in Section 5, there is disagreement
discussion of these corrections, the reader is referred, for
about the character of the many-body effects, especially
example, to Refs. [1,2]. In the present paper, we are mainly
depending on whether rigorous or semi-empirical theory is
interested in the relative BEs, BErel, of different PES features
used to analyze the XPS spectra.
rather than with the absolute BEs.
The different spectra in Figs. 1 and 2 illustrate two important
We restrict the discussion to the analysis and interpretation
features of XPS spectra that a proper analysis must explain
of core-level spectra. We use the acronym XPS, X-ray
and, ideally, relate to materials properties. The first is the
photoelectron spectroscopy, in place of PES when we empha-
number of features, attributed to ionization of a given shell,
size that X-rays are used, whether they are obtained with
that receive significant intensity; in Fig. 1, there is only one
conventional laboratory or synchrotron sources; we also use
peak while in Fig. 2, there are three. A second, closely related
the phrase “XPS Spectra” to refer to the core-level spectra,
feature is the relative intensity, Irel, of peaks when the XPS
themselves. Our principle objective is to relate features of XPS
spectra of a particular shell has multiple peaks. Why only one
BEs both to physical processes, such as multiplet splittings
peak in one case and multiple peaks in another and why the
[3,4], and to chemical interactions. Such correlations go
differences in intensity among multiple peaks are questions
beyond the use of XPS to identify the elemental composition
that rigorous theoretical analysis can be called upon to answer
of the sample studied [1]. In fact, relating XPS to chemical
and to relate to materials properties. Another feature of XPS
bonding has been an objective essentially since the inception
spectra that theory can help to understand concerns the shifts
of modern XPS studies [5,6]. However, in order to correctly
of XPS BEs, denoted ΔBE, which have been extensively
connect XPS with materials properties, one needs definitive
studied to determine whether and how the ΔBEs reflect the
understandings of a number of concepts which, unfortunately,
chemistry of the materials [5,6,23]. The ΔBE may be between
have sometimes been elusive. Three important concepts to be
inequivalent atoms in a single sample, as for example surface
considered in detail are (1) screening of a core-hole and related
or bulk atoms in a solid [11,24,25], or of atoms in different
matters of relaxation and Koopmans' theorem [7–14]; (2) the
compounds [26]. As we shall show, theoretical analyses of
extent to which the shift of a core-level BE of an atom in
these features permit an understanding of the interplay of
different chemical and physical environments can be attributed
several effects including: physical, dominantly intra-atomic;
to charge transfer, CT, between the atom and its environment
chemical bonding; environmental; and initial as opposed to
[11,15,16]; and (3) for open shell systems, including high spin
final state, or screening, effects. In other words, an important
oxides, multiplet splitting arising from the angular momentum
goal is to relate the number, intensity, and energy of these
coupling of the ionized core-level with the valence open shell
features to the materials chemistry and we will indicate the
[3,17,18].
directions that need to be taken to achieve this goal. Another
An idealized XPS spectra for the Ne atom is illustrated
schematically in Fig. 1. The figure shows peaks corresponding
to the 2p, 2s, and 1s shells at BE≈10, 20, and 750 eV,
respectively. The Ne XPS also contains satellite features with

Fig. 1. Schematic XPS of the Ne atom showing three peaks corresponding to

the Ne 2p, 2s, and 1s shells. Fig. 2. Ce 4s XPS of CeO2 measured with 875 eV photons; see Ref. [21].
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 275

feature of interest is peak broadening. Theoretical studies yield The overall organization of this review is as follows: In
individual lines at particular energies with particular intensities Section 2, we describe the models that are used. The
while the experimental spectral features, or peaks, are broadened descriptions are brief, but necessary for the reader to under-
by both extrinsic and intrinsic mechanisms. Extrinsic broadenings stand the physical and chemical significance of the work;
arise from the resolution of the electron energy analyzer and from technical details for the models and methods are given in the
the energy distribution of the photon flux. For a laboratory cited references. This section includes a discussion of the
apparatus that uses unmonochromated Mg or Al Kα X-rays, the sudden approximation (SA) [38,39], method that we use to
photon flux will have a broadening of ∼1 eV full width at half determine the relative intensities of the XPS peaks. We also
maximum (FWHM) [27], although this width can be reduced compare our methods and models to those of others. In
considerably if a monochromater is used. With high resolution particular, we contrast the analyses that we present, based on
XPS at synchrotrons, the broadening due to the photon flux and wavefunction theory, to analyses based on density functional
the energy analyzer can be reduced to ∼0.25 eV or even less; see, theory (DFT). In Section 3, we discuss the basic concepts
for example, Refs. [2,28]. Another extrinsic source of broadening necessary to understand the origins of BE shifts in molecular
for insulating samples arises from charging of the material. There and condensed phase XPS spectra and we review several
are several origins for intrinsic broadening. One is the lifetime of examples of these BE shifts. In particular, we present rigorous
the core-hole state created in the XPS photo-ionization [29]; definitions of initial and final state effects and show how
another is the broadening due to inhomogeneities in the material Koopmans' theorem is a way of separating these effects [40].
studied. There is also the possibility of broadening from This separation is necessary to make it possible to distinguish
unresolved final states for different multiplets in, for example, initial and final state contributions to BE shifts. The driving
open shell transition metal compounds, as will be discussed in force for final state effects is the response of the “passive”, un-
detail later [10,30]. Another mechanism is vibrational broadening, ionized, electrons to the hole created by core-level ionization,
often described as the Franck–Condon (FC) broadening [31], which is described as relaxation and as screening of the core-
which has been known since the mid-1970s for the C(1s) XPS hole. We relate the relaxation and core-hole screening to the
spectra of gas phase CH4 [32] and for broad XPS peaks of bulk concept of an equivalent core model introduced by Jolly
oxides [33,34]. Recently, it has been shown that chemical [41,42]. It is common to describe screening, especially in high
interactions with the substrate are responsible for variations in spin oxides [7,8,12–14,20,43], in terms of charge transfer from
the FC vibrational broadening of thin oxide films [35]. These ligands to the unoccupied levels of the core-ionized atom. This
broadenings may have the Gaussian or Lorentzian character [29] usage of screening by charge transfer invokes a change of
and are best described with a Voigt convolution [36]. The details configuration and it neglects orbital hybridization and the
of the broadening functional form are beyond the scope of this closed shell screening which arises because of this hybridiza-
paper and when we discuss XPS broadening we will use a tion. The concept of closed shell screening is introduced in this
Gaussian broadening. section, although major applications and use of this concept
As discussed above, several physical and chemical mechan- are deferred to later sections after the discussion of covalency
isms contribute to the complexity of XPS spectra. Since the in oxides, which is presented in Section 5. In order to relate BE
BEs in Eqs. (1) and (2) are the difference of the energies of N shifts to chemistry and chemical bonding, we introduce and
and N−1 electron wavefunctions, all peaks are rigorously use methods to decompose the properties of the wavefunctions
many-electron rather than one-electron quantities. However, it into the contributions from different chemical and physical
can be useful to distinguish the mechanisms according to one- mechanisms. This is accomplished using constrained variations
body or many-body physics and chemistry. When the wave- [44–46] where the orbitals to be varied and the orbital space of
function, Ψ, for a state is well described by a single config- the variation are constrained so that certain types of chemical
uration, perhaps even a single anti-symmetric orbital product interactions are allowed and others are explicitly excluded. In
or determinant, we describe this as a state where one-body Section 4, the interpretation of satellites is introduced for
effects dominate. This case is distinguished from states where examples where these satellites can be explained in terms of
a single configuration is not sufficient and Ψ must be described screening and covalency as discussed in Section 2. The
as a linear combination of several configurations. Such a analysis of the complex satellite structure in oxides is
distinction is reflected in observed XPS spectra. Spectra where presented in a later section. In the following section, Section
one-body effects dominate have, associated with each shell, a 5, rigorous criteria for determining the extent of the covalent
single “main” peak and weak satellites, as in Fig. 1, while character in nominally ionic systems, e.g., metal oxides and
spectra, as in Fig. 2, which have several intense peaks for a halides are discussed; this analysis is necessary to understand
given shell are cases where many-body effects dominate. In the satellite XPS spectra of these materials. We also describe
cases where an effect could be described either as one-body or how uncertainties may be estimated for the assignments of
as many-body, we explain our choice. For example, we chose effective charges on the various atoms. We examine the
to include the multiplets arising from angular momentum covalency in several oxides both with and without core-
coupling as a many-body effect because angular momentum holes. This is necessary to lay the foundation for the discussion
coupling may also involve recoupling of the valence open shell of the many-electron contributions to the XPS of oxides. In
[30,37], which is inherently a many-body effect; see the Section 6, several topics are considered. We introduce, the
discussion in Section 6.1. contributions of angular momentum coupling and multiplets to
276 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

XPS. This starts with a discussion of the N(1s) and O(1s) XPS bonds in the molecules. For bulk, surface, and chemisorption
of the NO molecule and shows that the singlet and triplet studies, we use clusters that contain the critical chemistry that
multiplets may or may not be resolved. We also discuss determines the XPS features of interest. For the interactions of
angular momentum coupling and recoupling as well as near- atoms and molecules with surfaces, we have examined how the
degeneracy effects for transition metal [3,30,37,47] and properties of the interaction are affected by the size of the
actinide cations [48–50] and, surprisingly, even for rare-gas cluster used to simulate the substrate [54–56]. We find that the
atoms [51]. We describe selection rules for XPS allowed and chemical character of the interaction converges very rapidly
XPS forbidden configurations and show how “forbidden” with the number of atoms used in the cluster model of the solid
states can steal intensity from XPS “allowed” states through substrate. The rapid convergence of the character of the bond
configuration mixing of the allowed and forbidden configura- formed has allowed us to tailor the clusters to correctly model
tions. In addition, effects that lead to satellites in closed shell the XPS features that we are studying and to relate these
oxides where angular momentum coupling and recoupling will features to the chemical interactions and bonding. Sometimes
not play dominant roles are considered. For these systems, a dominant aspects of the XPS of cation core levels in high spin
particular concern is to contrast interpretations in terms of CT oxides are adequately represented by isolated cations
satellites or shake satellites [52,53]. Finally, in Section 7, we [30,48,50,57], due to the importance of the multiplet splittings
summarize the various contributions that we have identified for caused by angular momentum coupling of the core-hole with
the interpretation of the complex XPS spectra of metal oxides. the unpaired valence electrons. In Fig. 3, we show three
We point out that, in order to correctly interpret the physical representative clusters. Fig. 3(a) shows an 18 atom cluster to
and chemical origins of XPS spectra it is necessary to take the model the XPS of bulk and surface atoms of Cu(100); the bond
interplay of several mechanisms into account. Our major distances and the geometry of the atoms are taken from those
objective with this review is to show that valuable information of a bulk Cu crystal. The surface and bulk atoms whose core
can be gained by a proper, non-empirical analysis of this levels are studied, labeled S and B, have coordinations of 8 and
interplay. In Appendix A, we give an alphabetical list of 12, respectively. This cluster was used to analyze the core-
abbreviations and acronyms used in the paper. level BE shifts of surface atoms relative to the bulk [24,25].
The clusters in Fig. 3(a) and (b) as well as a smaller, 13 atom,
2. Theoretical models and methods cluster have also been used to study BE shifts due to particle
size and lattice strain in small particles [58,59].
In this section, we review the underlying quantum mechan- In order to study lattice strain, the bond distances were allowed
ical foundations of the theoretical approaches used to model to breathe while keeping the fcc crystal geometry. Fig. 3(c)
the XPS spectra as described in the following sections. In this shows a cluster used to model the XPS of CeO2 [21,22], which
discussion, we stress the physical and chemical content of the contains a central, nominally 4+, Ce cation, although the
approximations that are used rather than the details of the actual effective charge on the cation is determined from
mathematical formulations, which, in any case, can be found in variational calculations of the cluster wavefunctions [10].
the referenced literature. As a result, this section contains only The metal cation is surrounded by the eight nearest neighbor
three equations; this is the minimum number needed to oxygen anions and by a large set of point charges to provide
understand the significance of the theoretical considerations the Madelung potential of the extended crystal; all atoms and
presented. Three topics are discussed. First, we discuss the point charges are placed at lattice positions of the fluorite
materials models used to describe the physical systems that crystal. Other metal, M, dioxides, MO2, with a fluorite
will be described in the rest of the paper. As we point out, it is structure can be modeled with analogous embedded MO8
necessary to formulate substantive technical reasons that these clusters; for example, UO2 [50]. In a similar fashion,
models contain the necessary physical and chemical features to embedded MO6 clusters, not shown in Fig. 3, can be used to
properly describe the XPS properties of interest. Second, we model octahedral metal oxides. These clusters contain a central
discuss the general approaches taken for the determination of metal cation, six nearest neighbor O anions, and embedding
the electronic structure underlying the XPS processes; this point charges. These clusters have been used to model MnO
includes placing the methods that are used in the later sections [10] and UO3 [60]; these oxides will be discussed later in this
of this paper in the context of other methods that have been paper. Embedded clusters with even larger numbers of atoms
used for the theoretical analysis of XPS. Finally, there is a brief explicitly included were used to model MgO [35], also
discussion of the determination of the intensity of the transi- discussed later.
tions to the XPS photoionized states. It is important to stress For the molecules or clusters, wavefunctions are determined
that a theoretical analysis is incomplete without theoretical using either the non-relativistic program system CLIPS [61], or
predictions of the XPS intensity. the relativistic program system DIRAC [62]. The choice of
The key considerations underlying the choices of materials using relativistic or non-relativistic wavefunctions depends on
models, which are used for the analysis of XPS presented in whether the atoms in system are heavy or light and on whether
the rest of this review are the following. For isolated we are modeling core holes in s shells or in shells with
molecules, we use the experimentally determined geometries non-zero orbital angular momentum where it is normally
since our concern is not to establish the geometries of these necessary to take spin–orbit splitting into account. Thus, for
molecules but to relate properties of the XPS to the chemical example, the study that we describe below of the XPS of the
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 277

Fig. 3. Representative clusters used to model XPS BEs for different condensed systems: (a) A Cu18 model of the Cu(100) surface where the representative bulk and
surface atoms are marked with B and S, respectively. (b) A Cu115 model of a Cu nanoparticle. (c) An embedded cluster model of CeO2 with a central Ce cation, 8
nearest nearest O anions, and point charge embedding.

NO molecule is based on non-relativistic wavefunctions while configurations [68] and the angular momentum coupling is
the study of CeO2 used relativistic wavefunctions. One of the introduced at a later step. HF and, by direct extension, DHF
special features of theoretical calculations of wavefunctions is wavefunctions provide reasonably accurate absolute core-level
that it is possible to decompose a property into contributions BEs with uncertainties of ∼1 eV [70–72] but, even more
from different chemical and physical mechanisms. This is done important for our purposes, accurate relative BEs for BE shifts
by imposing constraints on the orbitals that are varied and on [25] and for multiplet splittings [71]. Of course, when the
the space in which these orbitals can be varied. The name that angular momentum coupling becomes more complex [30,37]
we have given for these constraints is CSOV for constrained and when other many-body effects need to be treated
space orbital variations [44–46], although similar methods of [12–14,47,52,53], it is necessary to go beyond one configura-
decomposition have been developed by others [63]. tion HF wavefunctions. This is done through the method of
The orbitals or spinors are expanded in terms of sufficiently configuration mixing or configuration interaction (CI) where
large sets of Gaussian basis functions [64], so that the results the wavefunction is expanded as a sum over determinants or
are reliable. Pseudo-potentials, described as effective core over angular momentum coupled configurations called config-
potentials (ECPs) [65,66] represent the core electrons of the uration state functions (CSFs) [73]. The general form of a CI
“environmental” atoms [67] but are not used for atoms where a wavefunction, Ψ, is
core electron may be ionized. In every case, variationally
Ψ i ¼ ∑k C i;k Φk ; ð3Þ
optimized orbitals or spinors are determined through either
Hartree–Fock (HF) [64] or Dirac–Hartree–Fock (DHF) [68] where the index i denotes the ith state and the coefficients Ci,k
calculations. In the case of open shell systems, the non- are determined by solving the exact Hamiltonian in the space
relativistic orbitals are normally determined taking into of the Φk determinants or CSFs [64]. In principle, for a
account the angular momentum coupling and solving the HF sufficiently large expansion, the Ψi can be exact solutions.
equations for a specific multiplet [69] while for the relativistic However, the accuracy and reliability of the CI wavefunctions
spinors, the DHF equations are solved for the average of of Eq. (3) depend on the selection of orbitals to construct the
278 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

determinants Φk and of the Φk to be used in the sum. Often, it The overlap integral for Irel(k,α) provides selection rules;
is possible to base these choices on chemical and physical when the integral is zero by symmetry for the final state ΨN−1 α ,
considerations of how particular many-body effects contribute then Irel will be zero. The selection rules are stronger if we are
to the XPS spectra. These considerations will be discussed in able to neglect closed shell screening and assume that the same
the appropriate sections below. orbital set is used for both initial and final states. This
The standard approximation for the calculation of XPS Irel is assumption is most useful for studying the XPS multiplet
the sudden approximation (SA) [38], which is exact in the limit features that arise from the angular momentum coupling
of infinite photon energy. However, the SA provides accurate between the core-hole and open valence shells. Here, as
Irel for photon energies that are ∼100–200 eV above the discussed in Section 6, these stronger selection rules can be
ionization threshold for a core or valence level [74]. The SA used to qualitatively understand the XPS multiplet features.
starts with the assumption that the final state N electron However, we stress that closed shell screening is neglected
wavefunction after photo-ionization can be written as an only to allow qualitative analyses to made; the exact SA Irel
anti-symmetrized product of an N−1 electron bound state, that are used to predict the XPS spectra take full account of the
ΨN−1
k , and a one electron continuum orbital, ε. Further, it is closed-shell screening.
assumed that at time t¼ 0 the N−1 electron bound state is Implicit in the discussion above is that Hartree–Fock or
given by suddenly removing an electron from the kth shell and Dirac–Hartree–Fock methods will be used to determine the
leaving the other electrons unchanged. This is written formally orbitals for configurations and, in some cases, for multiplets.
by using the annihilation operator, ak, such that Then, CI wavefunctions will be used to determine additional
many-body effects and, if it has not already been done with the
Ψ k N−1 ¼ ak Ψ N : ð4Þ Hartree–Fock calculation, to determine the multiplets. As well
as wavefunction theory, based on HF and CI wavefunctions,
This wavefunction is often called a frozen orbital (FO)
DFT has also been used to study core-level spectroscopies and
wavefunction [40] since the “passive” orbitals are frozen, or
has been very successful determining the main XPS peaks of a
fixed, as they are in the initial state where the core-shell is
very large set of organic molecules [79–81]. The comparison
filled. However, ΨN−1
k is not an eigenfunction of the N−1
of HF and CI with DFT methods is not central to the main
electron Hamiltonian. Since energies are measured in XPS, it is
object of this review, which concerns the relationships between
necessary to expand the FO wavefunction, ΨN−1
k , in terms of a
features of an XPS spectra with the electronic structure and
complete set of N−1 electron wavefunctions that are solutions
properties of the material. However, since DFT is a widely
of the N−1 electron Hamiltonian:
used approach to determine the electronic structure of a wide
H N−1 Ψ αN−1 ¼ E αN−1 Ψ α N−1 and Ψ kN−1 ¼ ∑α Ck;α Ψ N−1
α : ð5Þ range of materials, it is appropriate to contrast and compare HF
and CI methods, for the states involved in XPS, with DFT
The relative intensity of an XPS peak corresponding to the methods.
N−1 electron ion being in a final state ΨN−1 α is simply the DFT is inherently a one configuration theory [82,83], and
probability of finding the system in this state. From Eq. (5), multiplets and multiconfiguration effects do not arise naturally
Irel(k,α) ¼ |〈ΨN−1
k |ΨN−1
α 〉| ; i.e., the square of the overlap of the
2
unless one uses time-dependent DFT (TDDFT) [84]. Further-
N−1 electron FO wavefunction with N−1 electron eigenstate more, it has been argued [85] that not all excited states are
ΨN−1
α . For the discussion of intensity losses from main XPS included in a TDDFT treatment and it is not clear how this
peaks, it is convenient to number the eigenstates in Eq. (5) so limitation will affect states relevant for XPS. In contrast, all
that α ¼ 0 is the lowest energy state with a hole in the kth shell excitations can, in principle, be included in a CI wavefunction.
and, hence, normally represents the “main” XPS peak. An It is possible to use DFT energies for different determinants to
important quantity is the deviation of Irel(k,0) from 1, which determine multiplet energies for open shell systems; see, for
indicates the loss of intensity from the main peak to satellites; example, Refs. [86,87]. The equations for the multiplet
it is common to describe this loss as a percent of the total energies obtained in this way are obtained from the energy
intensity, into all final ionic states, for ionization of the kth expressions for the angular momentum coupled wavefunctions
shell or [1-Irel(k,α)]  100. For PES calculations based on for a single configuration [88]. Thus the multiplet splittings
Green's function methods to treat satellites [75,76], the relative given by these equations may be expected to have limitations
intensities, Irel(k,α), are described as pole strengths. It is in their accuracy similar to the splittings determined with HF
important to point out that we use two different sets of orbitals, wavefunctions. While HF energies for the different multiplets
one variationally optimized for the initial state, where the core arising from an open shell configuration often give the correct
shells are filled, and the other for the final, core-hole, state order and magnitude of the multiplet splittings, additional
[10,40]. The use of two sets of orbitals gives a compact and, as many-body effects must be taken into account to obtain very
we show in the following sections, a chemically meaningful accurate values for these splittings. For many-body corrections
description of “closed shell” screening. Because the two to the HF multiplet splittings for pn open shell configurations
different sets of orbitals are not orthogonal to each other, the see Refs. [89–93]; for corrections for dn open shell configura-
evaluation of the overlap integral between ΨN−1 k and ΨN−1
α is tions see Ref. [94]. It is also possible to use the orbitals
not as simple as for the case where the orbital sets are mutually obtained as solutions of the DFT Kohn–Sham equations to
orthogonal [77]. However, the calculation is tractable [78]. carry out CI calculations for the energies and wavefunctions
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 279

for core-excited states. These CI wavefunctions can include perturbation, normally described as the level of (n+1)hole–(n)
full angular momentum coupling of the multiplets arising from particle excitations. A difficulty for application to the XPS of
the open valence and core open shells as well as other many- core-holes in transition and heavy metals is that high order
body effects. Such CI calculations using Kohn–Sham DFT hole-particle excitations will be required to give accurate BEs
orbitals have been carried out for core excited states arising in and Irel. These high order excitations are needed because
X-ray adsorption spectroscopy (XAS) by Ikeno et al. [95,96]. Green's function methods use a single set of orbitals where
Unfortunately, since it is necessary to compute CI wavefunc- closed-shell orbital screening is not taken into account.
tions with the Kohn–Sham orbitals to treat many-body effects
for the core excited states, some of the computational 3. One-electron features
advantages of a DFT treatment are lost. An advantage of HF
wavefunctions is that the HF orbital energies give initial state In this section, we define important concepts needed for the
BEs while this is not the case for Kohn–Sham orbital energies; analysis of XPS spectra and then, in Section 3.4, we consider
see the comparison of the meaning of HF and DFT orbital applications of these concepts to interpret two features of XPS
energies discussed in Section 3.1. spectra, BE shifts and vibrational, or Franck–Condon, broad-
A theoretical framework based on CI wavefunctions to ening of the XPS features. The concerns in this section are for
represent many-body effects is also the basis of the semi- one-electron features of the XPS; i.e., features that can be
empirical Anderson Model methods used to describe the XPS understood on the basis of a wavefunction that is a single
of metal oxides and halides [12–14,97]. However in the semi- determinant. In contrast, many-electron, or many-body, fea-
empirical formulation, it is assumed that orbitals used are tures require the use of wavefunctions described by several
either pure metal or pure ligand and there are fundamental determinants; many-body effects are described in later sec-
issues concerning the validity of this assumption. The metal tions. In Section 3.1, definitions are given to rigorously
and ligand orbitals have a non-zero overlap and, hence, cannot, distinguish initial and final state effects; final state relaxation,
as is done in the Anderson Model Hamiltonian calculations, be which screens a core hole is also defined and discussed. The
treated as though they were orthogonal. Especially in cases use of Koopmans' theorem as a means to separate initial
where the overlap is large, see the discussion in Section 5, and final state contribution to BEs is also discussed in this
errors may be introduced by neglecting the overlap. In sub-section. In Section 3.2, the different ways that a core-hole
principle, it could be possible to add parameters to the model may be screened are described and different theoretical models
Hamiltonian to take account of this orbital overlap; possibly in for this screening are compared and contrasted. Furthermore,
the context of a non-orthogonal CI [98,99]. However, this has the description of screening that can be obtained by replacing a
not been done. A second, but closely related, concern is that core-ionized atom with the next atom in the periodic table is
the assumption of pure metal or pure ligand orbitals neglects discussed and limitations of this model are considered. In
the covalent mixing of these orbitals to form filled bonding Section 3.3, the initial state electronic mechanisms that lead to
orbitals and open-shell or empty anti-bonding orbitals [10]; chemical shifts of BEs are presented; in particular, it is shown
also see Section 5. This covalent mixing is fundamental to our that hybridization can make substantial contributions to BE
understanding of chemical bonding. The consequences of shifts.
neglecting covalent mixing of orbitals could be avoided by
allowing all possible distributions of the electrons over the
nominally pure metal and ligand orbitals that are involved in 3.1. Initial and final state effects and Koopmans' theorem
the many-body treatment. This is because mixing of a
complete set of configurations is invariant to the rotation of It is very important to distinguish between initial and final
the orbitals among each other [77]. Unfortunately, such state contributions to the BEs, especially for the interpretation
complete mixings are not practical from a computational point of the physical and chemical meaning of BE shifts, ΔBE
of view since the number of the configurations to be mixed [24,25,58,59,100–103]. As we show later in this section, the
grows exponentially. In addition, parameters in the Anderson separation may also help to understand a class of XPS
model Hamiltonians are normally adjusted to fit experimental satellites. The initial state effects are obtained from the BE
data [12–14] and one can get good fits to XPS spectra for the for the FO wavefunction of Eq. (4) as
wrong chemical and physical reasons [47]. In principle, one BEk ðInitialÞ ¼ BEk ðFOÞ ¼ 〈Ψ kN−1 jH N−1 jΨ N−1 〉−〈Ψ N jH N jΨ N 〉;
k
could adjust the model Hamiltonian parameters to fit the
ð6Þ
results of rigorous, non-empirical calculations [97], so that
the meaning of the parameters could be placed on a sound where the subscript k indicates a hole is created in the k shell.
footing [97]. However, this is rarely done. If there are open valence shells, the index k may also represent
Green's function methods have also been used to simulate the angular momentum coupling of the core-hole with the
the PES [75,76] of isolated and chemisorbed molecules but valence electrons. Since, in the FO wavefunctions, the
not, to our knowledge, to the high spin oxides of transition, potentials seen by the core electrons are fixed using the initial
lanthanide, and actinide metals. Green's function methods are state orbitals, it is proper to describe the BEs as initial state
rigorous and, since they are based on perturbation theory, have BEs. It can be shown with Koopmans' theorem, and its
a path to exact solutions by increasing the effective order of the extensions for open shell systems [104], that BEk(FO) ¼ −εk,,
280 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

where εk is the orbital energy of the kth shell in the initial state significantly larger than the observed XPS BEs. It has been
Hartree–Fock wavefunction; we shall return shortly to discuss pointed out, correctly, that the BE(FO) have significant errors
Koopmans' theorem further. First, we contrast the initial state when compared to experiment; see, for example, Refs. [79,80].
BE(FO) with a “final” state BE. When we solve the variational In this sense, one may ask if the BE(FO) have any value or
equations for the core-hole state and allow the “passive” scientific use. The answer is yes because, in many cases, the
orbitals to relax in the presence of the core-hole, we obtain shifts of BEs are dominated or have, at least, major contribu-
the ΨN−1
α of Eq. (5) where the lowest energy core-hole state tions from initial state effects; several examples will be given
with α ¼ 0 is normally the state of interest. This wavefunction, later in this section, Furthermore, the initial state effects can be
since it has allowed the “passive” electrons to relax or respond directly related to the chemistry and the bonding in the
to the core hole, includes both initial and final state effects. The material being studied. A common assignment is that a shift
BE obtained with the relaxed N−1 electron wavefunction is to larger BE, ΔBE 4 0, indicates that the atom with the larger
called a ΔSCF BEα and is given by BE has a larger positive or a less negative effective charge
while a shift to lower BE, ΔBE o 0, indicates that the ionized
BEα ðΔSCFÞ ¼ 〈 Ψ N−1
α jH
N−1
jΨ αN−1 〉−〈Ψ N jH N jΨ N 〉: ð7Þ atom has more electronic charge and its effective charge is
The difference between BE(Initial) and BE(ΔSCF) is the either less positive or is negative. This relationship between the
relaxation energy ER, charge state of the ionized atom and the shift of the BE is
intuitive and it is supported by a large body of data for atomic
ER ¼ BEðInitialÞ− BEðΔSCFÞ≥ 0; ð8Þ ionization potentials [106]. However, it must be used with care
where ER is normally defined for the lowest energy core-hole because there are other reasons for shifts in BE. Indeed, using
state with α ¼ 0. The FO BE(Initial) is an important quantity only the relationship that ΔBE reflects the charge of the core-
since it reflects the potential arising from the physical and ionized atom can lead to contradictory conclusions. Thus,
chemical environment of the XPS core-ionized atom, and Wertheim [107], on the basis of BE shifts in the cation levels,
hence, it contains direct information about this environment. concluded that BaO had considerable covalent character, while
On the other hand, the variationally optimized wavefunctions Barr and Brundle [108], on the basis of BE shifts of the anion
for the hole state configurations, ΨN−1 levels, concluded that BaO was “very ionic”. This apparent
α , are, as discussed below,
considerably changed from the those for the initial state. With contradiction was resolved by the work of Pacchioni and
XPS, only the BE including both initial and final state coworkers [16,109], where other contributions to BE shifts, in
contributions, as given by ΒΕα(ΔSCF) in Eq. (7), can be addition to effective charges were taken into account.
measured; BE(FO) or ER must be obtained from the theory. It is common to use −ε from Hartree–Fock and Kohn–Sham
However, if one combines experimental XPS and Auger data, orbitals as an approximation to BEs and to describe this as a
it is possible to get information on the initial and final state Koopmans' theorem, KT, BE [9,70,104,110]. It is important,
contributions to the shifts of BEs [100,102,103]; this is however, to give a rigorous definition of the KT BEs. For HF
discussed at the end of this sub-section. wavefunctions of closed shell systems, it is straightforward to
Since orbital variation within the k hole subspace leads to show the initial state or FO BE is identical to the KT BE; i.e.,
decrease in the energy, the relaxation energy ER, Eq. (8), must BEk ðinitialÞ ¼ BEk ðFOÞ ¼ BEk ðKTÞ ¼ −εk : ð9Þ
be a positive quantity [70,105]. The ER may be very large and
it depends on the size of the system [58] and on how deep the For open shell systems, where there are several final state
core-hole is. To illustrate the magnitude of ER, we give values multiplets from the angular momentum coupling of the core-
of ER obtained for different core-holes on an isolated Cu+ hole with the valence open shell electrons, an additional
cation. The closed shell 1s22s2…3p63d10 configuration was qualification is required for Eq. (9). The KT BE is the
chosen to avoid the multiplets that would arise for the core weighted average of the BEs of the different final configuration
ions of open shell neutral Cu. The BE(ΔSCF) and the ER in multiplets [104]. If the multiplet splitting is not of direct
Table 1 are obtained from non-relativistic HF calculations for interest, then KT can be used for open as well as closed shell
the ground state and the core-hole states of Cu+. The ER are systems; if the multiplet splitting is of interest, the average KT
reasonably similar for a given principle quantum number and BE must be corrected with suitable exchange integrals [88].
increase as one goes from M to L to K shells. Even for light The important conclusion from Eq. (9) is that the KT BEs for
atoms, the initial state BE(FO) for the core levels are different systems, obtained from HF wavefunctions, directly
indicate the initial state contributions to the changes or shifts of
Table 1 the BEs between these systems. This is different from the
Non-relativistic BE(ΔSCF) and ER, in eV, for different shells of Cu+. interpretation of the DFT Kohn–Sham orbital energies [9,80].
Shell BE(ΔSCF) ER
The DFT orbital energies are an approximation to the exact
BEs, which include final state relaxation as well as initial state
1s 8907.9 64.5 effects; see Ref. [80] and references therein. In particular, the
2s 1094.4 24.7 exact Kohn–Sham orbital energy for the highest occupied
2p 951.2 26.4
molecular orbital (HOMO) obtained when the exact density
3s 138.2 6.7
3p 93.0 6.1 functional is used, is the exact first ionization potential of a
system. Unfortunately, except for a few special cases, the
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 281

general form of the exact density functional is not known [80]. and in the choice of Auger lines used [58,59,100]. Depending
The analysis that we have described above for BEs also applies on these choices one may obtain either a useful or a misleading
to electron affinities. In particular for DFT, the Kohn–Sham decomposition.
orbital energy for the lowest unoccupied molecular orbital
(LUMO) is a good approximation to the electron affinity [9].
3.2. Core-hole screening and relaxation effects
However, if one uses DFT orbital energies as representing only
initial state effects in the sense of Koopmans' theorem for HF
We have shown that the ER may be quite large and that this
wavefunctions, one may get misleading results.
implies large screening effects of the core-hole. One way that
We illustrate both the accuracy of DFT BEs [79–81]
screening has been described, especially in ionic materials, is
obtained with ΔSCF calculations and the limitations of
in terms of charge transfer (CT) from ligands or neighbors to
Kohn–Sham orbital energies as reflecting initial state effects
the core-ionized atom; see, for example, Refs. [7,8,12–
by considering the case of the closed shell CO molecule where
14,20,43,112]. Indeed, there is a nomenclature to describe
multiplet splittings are avoided. In Table 2, the O(1s) BEs for
the screening in terms of the CT from ligand to metal that we
the CO molecule from HF and DFT ΔSCF calculations are
illustrate with an example for the high-spin material MnO. The
compared with the values of the HF and DFT orbital energies
ground state is dominated by a configuration that is Mn2+ and
with sign changed, −ε, and with experiment. The differences
O2− and is denoted as
of the ΔSCF BEs and the values of −ε are discussed in terms
of the relaxation energy, ER, as defined in Eq. (8). The DFT …½cores3d5 ; ð10Þ
results are taken from Refs. [79,111] and the HF results have
been calculated using good quality basis sets; all calculations where the occupation of the O anions is assumed to be
are for the experimental C–O bond length. While both DFT 1s22s22p6. When a core electron is removed from the Mn
and HF BE(ΔSCF) give reliable values for the O(1s) BE, the several final state configurations are considered:
DFT value is much more accurate; this is not surprising since …½cores−13d5 ð11aÞ
the DFT result includes correlation effects. However, if we
treat the DFT O(1s) −ε as a KT initial state BE and use Eq. (8) …½cores−13d6 L ð11bÞ
to calculate ER then ER o 0, which is not physical. The
incorrect sign of ER is compelling evidence that Kohn–Sham
…½cores−13d7 L 2 and higher order excitations: ð11cÞ
orbital energies do not represent initial state effects and, thus,
cannot be used to separate initial and final state effects. In Eq. (11), [cores−1] indicates a core-hole. In the config-
In our discussion thus far, we have treated the separation of uration of Eq. (11a), an electron has been removed from Mn,
initial and final state effects through calculations of the which is now an Mn3+ cation and the O ligands are still O2− as
quantities BE(FO) and BE(ΔSCF). It is, however, possible to in the initial state. In Eq. (11b), an electron has been removed
deduce initial and final state contributions to BE shifts by from a core level but, in addition an electron has been
combining experimental shifts of XPS and Auger lines transferred from an O anion to the 3d shell, the Mn occupation
[100,102,103]. The experimental data is analyzed using an is now 3d6 and the CT of an electron from a ligand is denoted
Auger parameter and the separation between initial and final L. The CT of two electrons from the ligands to the Mn 3d,
state effects follows from arguments about the scaling of the Eq. (11c), is denoted L 2 . The charges on the Mn cation, including
relaxation energy for one hole, XPS, and two hole, Auger, final the core-hole, in Eqs. (11b) and (11c) are Mn2+ and Mn1+,
states. These arguments are based on an analysis of the respectively. In fact, it is possible to have higher orders of CT up to
difference of the extra-atomic relaxation energy for the singly configurations where 5 electrons have been transferred from the
charged XPS hole state and the doubly charged Auger hole ligands to Mn, …½cores−13d10 L 5 ; although, such high order
state [100–103]. However, there is strong evidence that the excitations are not normally used. The extension to other 3d
assumptions about the relaxation energy for the Auger state do occupations, to Lanthanides with an open 4f shell, and to actinides
not hold if the Auger decay involves an electron from a high with an open 5f shell is straightforward. However, the theory that
lying level in a valence or conduction band [59,100]. Thus, uses the CT model configurations of Eq. (11) is semi-empirical; see
care must be taken in the formulation of the Auger parameter references cited above. Indeed, the analysis of screening as CT
assumes that the orbitals are either pure metal or pure ligand in
Table 2 character whereas the chemical bonds in most oxides have some
BE(ΔSCF), orbital energies, and ER, in eV, for the O(1s) ionization of CO; all degree of covalent character, see Section 5. Consequently, rigorous
values in eV. methods cannot use the CT model, as described above. Since the
BE(ΔSCF) −ε(O1s) ER
configurations of Eq. (11) are allowed to mix, the use of CT as the
major form of screening belongs with many-electron effects and it
HF 541.6 562.4 20.8 will be discussed further in Section 6. The CT model is introduced
DFTa 542.5 513.5 −29.0 here because of its general use as a mechanism for the screening of
Exptb 542.55 — —
core-holes and so that we can contrast it with an alternative
a
See Refs. [80,111]. understanding of screening that explicitly recognizes and makes
b
See Ref. [5]. use of covalent chemical bonds.
282 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

The physical and chemical basis of screening through ground, state. The deepest lying 3s orbital is a bonding orbital
change in the covalent character between the initial and final, polarized toward O. The 4s is a lone pair on O with 〈z〉 slightly
core-hole, states is best understood using the ideas that form outside the position of the O nucleus while the 5s is a lone pair
the basis of the equivalent core approximation introduced by on C with 〈z〉 well outside of the position of the C nucleus. The
Jolly et al. [41,42]. The fundamental physical concept of this bonding 1π orbital is strongly polarized toward O. The 〈z〉 are
approximation is that the effective nuclear charge seen by the fully consistent with the usual view of bonding in the CO
valence electrons is increased by 1 when a hole is created in molecule [64]. When a C(1s) electron is ionized, all the 〈z〉
one of the core levels of an atom. Hence, we can replace the become closer to zero indicating a motion of charge toward the
nucleus by that of one atom higher in the periodic table but C center. This covalent screening of the core-hole has been
with the core shells all filled. Thus C, which has a nuclear described as “closed-shell” screening [10] to distinguish it
charge of Z=6, with a 1s core-hole and configuration 1s12s22p2 from the CT screening described above which involves
will be equivalent to N+, which has Z=7, and configuration moving an electron into an open shell. The direction of the
1s22s22p2. The equivalent core approximation is sometimes changes of 〈z〉 for the equivalent core molecule NO+ are the
referred to as the Z+1 model. Thus, CO with a C(1s) core-hole same as for the C(1s) core-hole ion of CO. While the 〈z〉 for the
is approximated by NO+ and with an O(1s) hole by CF+. It is high lying 5s and 1π orbitals are very similar for the C(1s)
entirely expected that the character of the orbitals in NO+ and core-hole state and for NO+, the changes in 〈z〉 for the deeper
CF+ is different from that in the parent molecule CO. We lying 3s and 4s as well as for the sum of the 〈z〉 are more
caution that there is penetration of the orbitals into the region different. Analogous comments can be made for the 〈z〉 of the
of the nucleus; in particular, orbitals that have atomic s O(1s) hole state and the Z+1 model of CF+. The center of
character have non-zero values at the nucleus. Hence, the charge moves toward O to provide closed shell screening of
Z+1 model is indeed an approximation but it does provide a the O(1s) core-hole and the Z+1 model qualitatively shows the
way of understanding how changes in the covalent character of same trends.
the “passive” orbitals screen a core-hole without the need for a The Z+1 model has been used to explain core-level BE
change of configuration as is explicitly required in the CT shifts between surface and bulk atoms using a Born–Haber
model described above. cycle to change the position of the Z+1 equivalent core atom
It is possible to examine the changes in the covalent between surface and bulk [23,114,115]. The Z+1 model has
character of the “passive” orbitals by examining how the 〈z〉 also been used to interpret resonances in X-ray adsorption
for the orbitals change for the CO molecule where the z-axis is spectra [116] where the electron that is excited by the incident
taken as the inter-nuclear axis and the C atom is at the origin light moves in the potential of the Z+1 model of the system
and the O atom at positive z. The 〈z〉 indicates the center of with a core-hole. The results in Table 3 suggest that while the
charge for an orbital or for a sum of orbitals. If the change is Z+1 model qualitatively describes the screening of the core-
toward a larger value, this indicates a motion of charge from C hole by the valence or conduction band electrons, there are
toward O. (With the use of 〈z〉 to indicate motion of charge, it quantitative limitations to the description of the screening.
is possible to avoid uncertainties associated with population Furthermore, since the core shell is filled in the equivalent core
analyses [113].) We show in Table 3, the 〈z〉 for the valence, atom, it is not possible to describe the multiplets arising from
3s, 4s, 5s, and 1π, orbitals for the ground state and for the C the angular momentum coupling of the open core level with
(1s) and O(1s) core-hole states; these expectation values are the open valence levels. As we show, especially in Section 6,
also given for the equivalent core molecules, NO+ and CF+. these multiplets must be taken into account for a correct
All calculations are for the CO experimental distance of description of XPS spectra.
1.13 Å. “Ideal” covalent orbitals, as for N2, would have the
center of charge at the center of the bond with 〈z〉¼ 0.57 Å. We 3.3. Initial state mechanisms leading to BE shifts
also give in the table the sum of 〈z〉 for these 10 valence
electrons where the individuals 〈z〉 are weighted by the orbital In this sub-section, our concern is to understand the origins
occupations. We consider first the properties for the initial, or of shifts in core-level BE for an atom in different environ-
ments. Such shifts are especially relevant for cases where the
Table 3 XPS spectra are dominated by a main peak and where satellites
〈z〉 in Å, for the valence orbitals of the ground, C(1s)-hole, and O(1s)-hole have low intensity. It is common to discuss these shifts in
states of CO, as well as for the Z+1 models of the core-hole states, NO+ and
CF+. The sum of the 〈z〉 weighted by the orbital occupations is also given; if terms of the charges of the atom where the atoms that have a
the center of charge were at the center of the molecule, this sum would be more negative charge or a less positive charge have smaller
5.65 Å. core level BEs. While core-level BEs do reflect the effective
charge of the ionized atom, there are other initial state
CO–GS CO+–C(1s) NO+ CO+–O(1s) CF+
mechanisms that can induce BE shifts. The electric fields
〈z〉3s 0.82 0.74 0.70 0.84 0.97 generated by charged species in a molecule or in a compound
〈z〉4s 1.14 1.08 0.82 1.21 1.02 have been recognized, since the early efforts in interpreting
〈z〉5s −0.26 −0.13 −0.18 −0.30 −0.26 XPS, as leading to ΔBE and estimates of the electrostatic
〈z〉1π 0.86 0.73 0.74 0.99 0.99
effects of charged species were developed to explain ΔBE in
Sum 6.83 6.28 5.62 7.48 7.31
molecules [5,6]. For ionic crystals, the Madelung potential
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 283

must be taken into account to properly describe BE shifts in Table 5

these crystals [16,109]. However, the hybridization of elec- Sizes, 〈r〉, 1/〈r〉, and 〈1/r〉, for the Cu atom; r in Å.
trons in deeper valence shells or, more generally, the involve- 〈r〉 1/〈r〉 〈1/r〉
ment of these electrons in the covalent bonding, has not been
recognized until rather recently [24,25,58,59]. In the following 3d 0.52 1.91 2.63
paragraphs, we will show why this hybridization can lead to 4s 1.76 0.57 0.72
4p 2.74 0.37 0.46
BE shifts and also show that it is an initial state effect. We will
also indicate how changes in the chemical environment can
induce changes in the hybridization. From the logic of Eq. (12), it would be expected that 〈r〉4p is
Core electrons are so close to the nucleus that, to a good larger than 〈r〉4s which, in turn, is larger than 〈r〉3d. The 〈r〉, 1/
approximation, contributions to core-level BEs due to valence 〈r〉, and 〈1/r〉 for these orbitals are given in Table 5; the values
and conduction band electrons are simply the potential of these for the 3d and 4s orbitals are taken from the configuration …
electrons at the nucleus. Thus the N electrons in a valence shell 3d104s1 while the 4p values are from …3d104p1. From these
orbital, φ, have a potential at the nucleus of atom A, VA(φ), sizes, it is clear that the BE order will be BE(…3d104s1)o BE
(…3d104p1)⪡ΔBE(…3d94s2).
V A ðφÞ ¼ −N〈φj1=r A jφ〉≈−N=〈φjr A jφ〉; ð12Þ The practical question is how we may a priori relate the
degree of hybridization to the chemical and physical environ-
where the minus sign signifies that the electrons in φ reduce
ment of atoms in the system. Hybridization involves promotion
the BE and the approximate equality on the right allows us to
of electrons in more strongly bound atomic shells, i.e. deeper
relate the potential to our intuitive view of the size of the
valence levels, into shells that are less strongly bound, i.e.
orbital. Thus, the larger a valence orbital is then the less it
shallow valence levels, in order to facilitate chemical bonding
reduces the BEs of core electrons and therefore the BE
with neighboring atoms. Often, one thinks of the promotion of
becomes larger. Of course, if an electron is removed entirely
an entire electron, as in the case of sp3 hybridization of the C
to form a cation, then the potential of Eq. (12) is no longer
atom in the CH4 molecule. However, hybridization can also
present and the BE is raised. Similarly, if an electron is added
involve promotion of only a fractional number of electrons
to form an anion, a new term of the form of Eq. (12) is added
from, for example, the 3d shell into the 4s and 4p shells of
and the BE is reduced. The magnitude of the effect of
transition metal atoms [24]. The essential fact is that hybridi-
hybridization is indicated in Table 4 where the shifts of the
zation enhances the chemical bonding between atoms. Thus if
deep core 1s BE of the isolated Cu atom are given for three
an atom has fewer neighbors, we would expect a smaller
low-lying states. These states are the ground state with
hybridization of the deeper valence to the shallow valence
configuration …3d104 s1, the excited state where the 4s
levels directly involved in the bonding. Since surface atoms
electron is promoted to the somewhat more diffuse 4p orbital,
have fewer neighbors than bulk atoms, hybridization should
…3d104p1, and the excited state where a 3d electron is
act to make the core-level BEs of surface atoms smaller than
promoted into the 4s shell, …3d94s2. These are the three
those of bulk atoms. The degree of hybridization should also
lowest states of the Cu atom. The ΔBEs, both the initial state,
increase as the bonds become shorter since stronger bonds are
ΔBE(KT), and the total shift including both initial and final
normally formed at shorter distances. For this reason, lattice-
state effects, ΔBE(ΔSCF), are given for ionization to the
strain in nanoparticles contributes to BE shifts as a function of
weighted average of the singlet and triplet 1s-hole multiplets.
particle size [58].
The excitation energies from the ground state to the excited
In contrast, the environmental charge density surrounding
states, ΔE(Total), are also given. All results are from Hartree–
the core-ionized atom can offset BE shifts due to hybridization
Fock calculations on the initial and final states. The Cu 1s KT
[24,117,118]. From the same reasoning that leads to Eq. (12),
BE shifts, ΔBE(KT) in Table 4 are 2.7 eV when the 4s
the environmental charge density surrounding an atom,
electron is promoted to the 4p shell, 4s-4p, and 6.8 eV when
neglecting any changes in the atom's electronic structure that
a 3d electron is promoted to the 4s shell, 3d-4s; the ΔBE for
might facilitate bonding, will create a potential that reduces the
the 3d-4s excitation is more than twice as large as for the
core-level BEs on the atom. The environmental charge density
4s-4p excitation. While the relaxation energies are somewhat
is affected by the both the number of neighboring atoms and
different for a 1s-hole in each of these states, the ΔBE are
by the distance of these atoms from the atom to be ionized.
dominated by initial state effects as can be seen from the fact
However, the effect of the environmental charge density on the
that the ΔBE(ΔSCF) are similar to the ΔBE(KT); see Table 4.
BE shifts cancels the effect of the hybridization. Thus, it is
Table 4 possible that this could lead to a net value of ΔBE which is
Excitation energies, ΔE(Total), and KT and ΔSCF BE shifts of the 1s BE, small. We shall examine how this cancellation occurs in detail
ΔBE, for three low-lying states of the Cu atom. All energies are in eV. in the following sub-section.
…3d104s1 …3d104p1 …3d94s2
3.4. BE shifts in molecules, clusters, and solids
ΔE(Total) — 3.1 0.4
ΔBE(KT) — 2.7 6.8
In this sub-section, we use the concepts developed above to
ΔBE(ΔSCF) — 2.2 7.8
analyze and interpret the chemical information that can be
284 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

obtained from observed ΔBE. The first case that we consider is where no constraints are imposed denoted Ψ[Full SCF]. We
the shift of the N(1s) BE in two closely related molecules, use the KT BEs since we have shown in Table 6 that the ΔBE
pyridine, C5H5N, and pyrolle, C4H5N [26,119], where the N is dominated by initial state effects described with KT. The
(1s) BE in pyrolle is ∼1 eV larger than the N(1s) BE in first indication that 2s-2p hybridization is more important in
pyridine. The original assignment for the origin of this ΔBE pyrrole than in pyridine comes from the energy improvement
was a charge transfer from the N in pyrolle to the C atoms in when hybridization is allowed; namely E[Frz N(s)]−E[Full
the ring. From a theoretical analysis, we are able to show that SCF]. This difference is 2.3 eV for pyridine but almost twice
this assignment is incorrect and to identify the correct origin of as large, 3.9 eV, for pyrrole. The consequences of this greater
the ΔBE. In Table 6, we compare the XPS measured N(1s) N 2s-2p hybridization for pyrrole for the BE shift are shown
BEs with the BE(KT) and BE(ΔSCF) from HF calculations on in Table 7. The Full SCF BE for both pyridine and pyrrole are
the two molecules. The absolute and relative values of the BE larger than the Frz N(s) BEs. This is entirely consistent with
(ΔSCF) are close to the measured XPS indicating that the the prediction of Eq. (12) for the contribution to the potential at
theoretical treatment is reasonably accurate. However, the the N center for a larger hybridized orbital than for an orbital
ΔBE(KT) is also close to the observed BE shift. Despite fixed to be a N(2s) orbital. However, the increase is much
the fact that the ER is large, it is similar for the two molecules. larger for pyrrole than for pyridine simply because there is a
The agreement of the ΔBE(KT) with ΔBE(ΔSCF) and with larger 2s-2p hybridization in pyrrole. We emphasize that
experiment establishes that initial state mechanisms are the even though CT from the N atom to the C atoms is allowed
dominant origin of the ΔBE; it does not, however, establish the with the Frz N(s) wavefunction, the N(1s) BEs are almost the
mechanism. Possible reasons for the initial state N(1s) BE shift same. It is only when 2s-2p hybridization is allowed that the
between pyridine and pyrolle are as follows: (1) there is charge Pyrolle BE becomes larger than the Pyridine BE. We now
transfer from the pyrolle N to the ring C atoms that leads to a consider two further examples to demonstrate the importance
larger BE than for the N atom in pyridine; (2) there is a of hybridization in determining BE shifts.
chemical bonding between N and H in the N–H unit in pyrolle The shift of core level BEs between surface atoms at the
while this unit is not present in pyridine; and (3) there is some (100) face of Cu from bulk atoms or surface core level shift
other difference in the chemical bonding of the N atom in these (SCLS) is such that the bulk atoms have a larger BE than the
two molecules. We describe a theoretical analysis that decom- surface atoms; ΔBE(B−S) 4 0. For Cu 2p3/2, the difference is
poses the different contributions to the chemical bonding 0.24 eV [24]. Usually [23], the ΔBE(B−S), for metal surfaces,
[44–46] and makes it possible to quantify how these different are small, well under an eV; however, they may be either
contributions affect the N(1s) BEs in the two molecules positive or negative depending on the metal being studied [23].
[26,119]. This analysis makes it possible to rule out charge The fact that the SCLS has different signs is consistent with a
transfer as the origin of the N(1s) BE shift; furthermore, the cancellation of positive and negative contributions. Indeed, the
analysis definitively identifies hybridization as the chemical theoretical analysis that we discuss [24,25] demonstrates that
bonding mechanism that leads to the N(1s) shift to higher BE there is just such a cancellation and it identifies the chemical
in pyrolle. In order to decompose the various individual mechanisms that are responsible for the cancellation. Our
contributions to the BE shift, we constrain the variation model for the SCLS of Cu(100) is shown in Fig. 3(a). We
allowed in the calculation of the wavefunctions to exclude N consider the deepest 1s Cu core level, although the SCLS of
2s-2p hybridization by fixing the N(1s) and N(2s) orbitals as other core levels are similar. The KT SCLS is ΔBE(B−S;
they are for the isolated N atom in its 4S ground state. This KT) ¼ +0.63 eV, while the ΔSCF SCLS is ΔBE(B−S;
wavefunction and related properties are described as Frz N(s). ΔSCF) ¼ +0.36 eV. Again, this is a strong indication that
Although N 2s-2p hybridization is excluded from Ψ[Frz N initial state effects dominate the SCLS. This is not surprising;
(s)], charge transfer from and covalent bonding between the N the relaxation, or screening, of a core-hole on a surface or a
(2p) and the C atoms is allowed. (The fixing of the N(1s) bulk atom should be similar since the conduction band
orbitals is simply a convenience in the application of the electrons can just as easily move to the location of the core
constraints within the CSOV formalism [44–46]. There is no hole at the surface or in the bulk. The difference between the
significant effect on the KT N(1s) BE since the N(1s) is not KT and ΔSCF SCLS, see the values given above, shows that
involved in the chemistry of either molecule.) The properties ER for a bulk atom is only 0.27 eV (¼ 0.63−0.36 eV) larger
of Ψ[Frz N(s)] are contrasted with those of the wavefunction than ER for a surface atom. Clearly, this difference in ER is a
very small fraction of the total ER≈50 eV for the Cu(1s) BE of
Table 6 either a bulk or a surface atom. Thus, we are able to use the KT
Experimental, XPS, and theoretical, BE(KT) and BE(ΔSCF), N(1s) BEs for
pyridine and pyrolle, all values in eV; see Ref. [26] for the sources of the data. Table 7
N(1s) BE(KT) for pyridine and pyrolle, all values in eV; see Ref. [26].
Pyridine Pyrolle ΔBE
Pyridine Pyrolle ΔBE
XPS 404.9 406.2 +1.2
BE(ΔSCF) 405.1 406.3 +1.2 Frz N(s) 422.5 422.4 −0.15
BE(KT) 423.9 424.8 +0.9 Full SCF 423.9 424.8 +0.9
ER 18.8 18.5 +0.3 Change +1.4 +2.4 –
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 285

BE to analyze individual chemical contributions to the SCLS. The situation is completely different with the Frz Ar Core
In order to avoid large numbers for the absolute values of the constraint when the Cu 3d electrons are allowed to hybridize
Cu(1s) BEs, we take the zero of BE as the KT BE of a free Cu and to participate in the chemical bonding. Now, the BEs
atom. All BEs are for multiplet weighted averages of the spin of both bulk and surface atoms are larger than the BE of an
couplings of the core-hole with the open shell valence level isolated atom and the SCLS changes sign and becomes
electrons. As well as the Full SCF SCLS, we also consider two ΔBE(B−S) 4 0 consistent with XPS measurements. Exactly
constrained variations. These constrained variations follow the as predicted by Eq. (12), the hybridization reduces 〈1/r〉 for the
logic used in our analysis of the BE shift between pyridine and orbitals that are dominantly 3d in character and hence allows
pyrrole; however, here the constraints are to either exclude or the BE to become larger. However, the 3d hybridization shift
include 3d hybridization to the 4sp “conduction band” of Cu. is differential; the BE of the bulk atom is raised by 3.9 eV but
The first constraint, denoted Frz Core, fixes the orbitals for the the BE of the surface atom, with fewer nearest neighbors and
28 electron core of the surface and bulk Cu atoms as they are hence a reduced 3d-4sp hybridization, is only raised by
for the isolated Cu atom. Only 1 electron for each of these 2.6 eV. The small SCLS arises from a cancellation of the
atoms is allowed to vary and interact with the neighboring Cu environmental SCLS of −0.6 eV and the hybridization SCLS
atoms in the Cu18 cluster model of Cu. For this Frz Core of +1.3 eV. This cancellation explains both the small SCLS
constrained variation, 3d-4sp hybridization is excluded. The and the fact that the sign of the SCLS may change for different
second constraint, denoted Frz Ar Core, includes the 3d elements [23].
orbitals and electrons of the surface and bulk Cu atoms in Changes in hybridization with the size of nanoparticles also
the variation and fixes only the orbitals of the 18 Ar core, contribute to shifts of core-level BEs [58,59,120]. However,
1s2…3p6, electrons to be as they are for the isolated Cu atom. the reason for the change in the hybridization with particle size
This variation allows the 3d orbitals of the representative is different from that discussed above for the SCLS. For
surface and bulk atoms to hybridize and to participate in the nanoparticles, theoretical analysis shows that changes in
chemical bonding with the other Cu atoms. The BEs obtained hybridization are due to bond distance variations as the particle
with the Frz Ar Core constraint are reasonably similar to the size changes rather than to the difference of coordination as
values for the full, unconstrained SCF calculation on the was the case for SCLSs. The core-level BEs shift by ∼1 eV to
cluster [24]. In Fig. 4, we compare the surface and bulk BEs lower BE as the metal particle size increases from very small
for these two constrained variations. When the 28 electron to large with several atomic layers. This BE shift was
cores are frozen and 3d hybridization and participation in the originally ascribed to a final state effect arising from the
bonding is excluded, both bulk and surface atoms have BEs charging of the nanoparticle during the XPS photo-ionization.
significantly smaller than the isolated Cu atom and the surface Regarding the metal particle as a perfect macroscopic con-
atom BE is larger than the bulk atom BE, ΔBE(B−S)¼ ductor, the charge must reside on the surface of the particle.
−0.6 eV. These are exactly the shifts that are expected from Furthermore, assuming that the particle is spherical, this
the potential due to the environmental charge density that positive charge at the surface would lead to an increase in
surrounds the bulk and surface atoms. This potential lowers the the BE∝1/R, where R is the radius of the particle [121].
BEs because the potential due to electrons lowers BEs, see Eq. However, the assumption that the particle can be treated as a
(12), and it has a larger magnitude for a bulk atom, which has macroscopic conductor is not valid for very small particles
12 nearest neighbors, than for a surface atom which has only 8 since it neglects the atomistic structure of matter [117,122].
nearest neighbors. However, the SCLS for this constraint has Thus, it is not possible to use the BE shifts of small
the opposite sign than measured or than predicted by theory nanoparticles as a measure of the particle size. In a recent
when no constraints are imposed. joint experimental and theoretical effort [58], the measured BE
shifts with particle size were compared with theoretical
predictions. The theoretical models were for Cu particles, up
to the large cluster shown in Fig. 3(b), where the final state
screening or relaxation energy has converged to within
∼0.1 eV of the bulk value. It was found that the shift to
smaller BE with increasing cluster size was only 2/3 of the
expected shift of ∼1 eV. However, an assumption made for
this theoretical study of cluster size was that the “lattice
constant” for the particles was the same as in the bulk. This
choice allowed the separation of the consequences of simply
changing cluster size from other effects. The ΔBE for the fixed
lattice constant show clearly that the final state charging
argument proposed earlier [121] is incomplete. From transmis-
sion electron microscopy studies, it is known that there is a
Fig. 4. The KT 1s BEs for Bulk Cu and Surface Cu atoms for the (100) crystal
5–10% contraction of the bond distances between atoms in
face; the BEs are for the Frz Core and Frz Ar Core wavefunctions where 3d- small nanoparticles relative to those in the bulk metal
4s hybridization is excluded or allowed, respectively. [123,124] .The change in bond distance for nanoparticles is
286 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

described as a “lattice strain” since the lattice constant changes The final topic in this section concerns another consequence
from the bulk value. It is reasonable that bond lengths are of changes in bond distances, in this case for vibrational
longer for higher coordinations where the bond strength is broadening of the MgO XPS. Since MgO is nearly an ideal
distributed over a larger number of bonds than for lower ionic oxide [35,126], when a hole is created on either the metal
coordinations where the bond strength is concentrated in fewer cation or the oxygen anion, there will be a significant change
bonds. For small particles, where the average coordination is in the electrostatic attraction of the counter-ions and hence a
less than for the bulk, it is logical that the bond distance will be change in the Mg–O bond distance. Qualitative studies [33,34]
contracted relative to bulk. have suggested that bond distance changes in ionic compounds
Lattice strain in nanoparticles was modeled by a breathing lead to a substantial Franck–Condon, FC, broadening of the
motion for the Cu18, Fig. 3(a) and the smaller Cu13 cluster, XPS lines because the final, ionic states will be in highly
which contains only one fully coordinated bulk Cu atom excited vibrational levels. This may also be described as the
[58,120]. The objective of this work was to identify the ionic state having large phonon excitations accompanying the
magnitudes and the chemical reasons for BE shifts arising core level ionization. A recent combined theoretical and
from changes in the bond distances. The BE(KT) and BE experimental study of the Mg 2p XPS for MgO thin films
(ΔSCF) were examined for changes in the bond distance shows that the line width depends on the film thickness [35].
ranging from an increase of 2% to a decrease of 6% with From the theoretical analysis, it was possible to identify the
respect to the bulk values, As for the analysis of the SCLS, we reason for the reduced broadening of the Mg 2p XPS lines in
also consider the KT BE shifts for the Frz Core and Frz Ar very thin films in terms of the character of the chemical
Core constrained variations. For the bulk atom in Cu13, the Cu bonding of the thin film to the substrate. This relationship
2s ΔBE are summarized in Table 8, where the BE for the between the broadening of the XPS peaks and chemical
bulk lattice constant is taken as reference for the shifts and bonding of the film adds a new dimension to the usefulness
ΔBE ¼ 0 for this distance. The small differences between the of FC broadening. For the Mg 2p ionization, bulk MgO is
ΔBE(ΔSCF) and ΔBE(KT) show that the changes in the BE described with an MgO6Mg18 cluster, embedded in a point
with lattice strain are dominated by initial state affects. The charge field; where each O is surrounded by six Mg cations.
ΔBE are almost linear with the change in bond distance, with This is distinct from clusters used in other studies of oxides
the BE decreasing for larger bond distances and increasing as where only one of the neighbors of the O anions is a real cation
the bond distances are increased. The ΔBE for the Frz Ar and the others are point charges. The additional extended
Core, not given in Table 8, are identical to the Full SCF ΔBE cations are needed in the present study because a new feature
(KT). This shows that as soon as the d hybridization and of the metal–oxygen interaction, not treated in the studies that
participation in the covalent bonding is included, there are no used clusters like that shown in Fig. 3(c), had to be
further significant changes in the ΔBE. As for the SCLS, the investigated in this study of FC broadening. The new feature
constrained variation that allows only the BE shifts due to the is the dependence of the metal–oxygen distance on whether a
environmental charge, ΔBE(Frz Core), leads to a reduction in core-hole is present on the central metal cation. If changes in
the BE as the atomic distances are decreased. This is to be the position of the O are to be studied, it is not adequate to
expected since the charge density surrounding the ionized move the O toward or away from point charges since point
atom increases as the inter-atomic distance is reduced. How- charges do not provide the steric repulsion that arises from the
ever, the increases in the BE due to the increase in the d extended charge distribution of the electrons associated with
hybridization as the inter-atomic distance is reduced are almost the cation. As is shown below, once the O's are fully
twice as large as the decrease due to the environmental charge. surrounded by real cations, this steric repulsion is properly
Thus, the hybridization contribution to ΔBE due to lattice represented and reliable metal–oxygen distances are obtained.
strain in small particles dominates and the shift is to larger BE. A potential curve was determined for the breathing motion of
The effects described above are general for noble metals and the 6 O nearest neighbors of the central Mg atom. For the
for systems where the d shell is nearly filled [58,120,125]. The ground state of MgO, the equilibrium re(Mg–O) is 2.09 Å,
SCLS and the ΔBE due to lattice strain show that hybridiza- which is quite close to the distance in the crystal, r(Mg–O) ¼
tion of deeper valence levels into higher lying, more diffuse 2.11 Å. When a Mg 2p electron localized on the central Mg
valence levels makes an important contribution to BE shifts. cation in the MgO6Mg18 cluster is ionized, the Mg cation
becomes effectively Mg3+ The minimum on the breathing
potential curve is now re(Mg–O) ¼ 1.97Å a decrease of 0.12 Å.
Table 8 There are almost identical reductions in re when an Mg 2s or
Shifts for the Cu 2s BE in Cu13 for changes in r(Cu–Cu), in %, from the bulk an Mg 1s electron is ionized providing confirmation that the
value. The ΔSCF, KT, and the Frz Core constrained KT ΔBE are given in eV; changes in the Mg–O bond distance are electrostatic in origin.
see Refs. [26,120].
Conversely, ionizing an O2− core electron using an O centered
Δr −6% −4% −2% 0(bulk) +2% cluster reduces its electrostatic attraction with Mg2+ which
increases the Mg–O distance by 0.1 Å. Fig. 5 is a schematic
ΔBE(ΔSCF) 0.67 0.41 0.19 0 −0.16 view of the potential curves for the ground state and for two
ΔBE(ΚΤ) 0.74 0.47 0.22 0 −0.20
ionic states. In case I, the Mg–O distance for the ion is shorter
ΔBE(KT; Frz Core) −0.67 −0.44 −0.22 0 +0.22
than in the ground state, as in the Mg 2p XPS, and in case II,
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 287

interpretation of the satellites in terms of the covalent character

of the substrate–adsorbate bond in both the initial and the final
states. In the second case, we focus on the character of the
screening as indicated by the positions of the center of charge
in the NO dimer. We have two objectives for this study. The
first is to establish the essential multiconfigurational character
of the dimer wavefunctions and the second is to distinguish
between two types of screening, namely inter-unit and intra-
unit screening. At the time the original work was performed, it
was necessary to use much smaller and simpler molecular
models for the wavefunctions. This could, in fact, have been an
Fig. 5. Schematic potential curves for the ground and hole-state potential advantage since it was necessary to carefully select the models
energy curves showing the FC vibrational broadening that arises when the Mg–O
that were used and to explicitly include the critical physical
distance changes in the ionic state.
and chemical mechanisms. The original papers used the
common jargon of the time. In particular, the distinction was
the Mg–O distance for the ion is longer than in the ground made between “screened” and “unscreened” final states. We
state, as in the O 1s XPS. The horizontal lines on the ground now know that there are different types of responses to core-
state curve mark the energy of the v¼ 0 vibrational level and holes, where inter-molecular or intra-atomic screening consti-
the classical turning points for this level are the intersections tute a more appropriate terminology.
with the potential curve. Since the electronic transitions are
much faster than the nuclear motion [31], the XPS transitions 4.1. XPS satellites of chemisorbed molecules
are vertical. However, the system may be at any point allowed
on the v¼ 0 ground state vibrational level. Excitations from The C(1s), N(1s), and O(1s) XPS of CO or N2 chemisorbed
the classical turning points are a good approximation to the on metal surfaces have a satellite at ∼5 eV higher BE than the
extremes of the vertical transitions. Using information for main peak where the intensity of the satellite increases as the
the harmonic vibrational frequencies [35], FC broadening of adsorption bond becomes weaker [127,132]. A general dis-
the Mg(2p) XPS and O(1s) XPS are predicted to be cussion of the XPS of CO in compounds and as an adsorbate
0.8 eV FWHM. on metal surfaces has been given by Freund and Plummer
In order to describe a monolayer of MgO/Ag(100), a point [133] as part of an effort to relate the features of the XPS to the
charge embedded MgO4Mg8 cluster modeling an MgO mono- character of the metal carbonyl interaction. The study of CO
layer was placed above an Ag25 cluster representing the Ag(100) on metal surfaces is of particular importance because it can be
surface. The calculated FC broadening of the Mg 2p XPS is 0.8 extended to the complex surfaces that are involved in catalytic
and 0.4 eV for the bulk and thin film MgO, respectively. processes [134]. The original theoretical study of the satellite
Furthermore, the theory indicates that it is the chemical interac- structure of weakly bound CO adsorbates was based on the
tion between the MgO film and the Ag(100) surface that affects model Hamiltonian calculations [135] where the relationship
the FC broadening. It was possible to exclude other mechanisms between the XPS and the character of the chemical interaction
such as a change in the final state screening of the Mg core-hole was indirect. Here, we review the work of Bagus and Hermann
by the metal support or a difference between the bulk and isolated [131] on the XPS of CO/Ni where the connection between the
monolayer MgO. The measured increase in FWHM going from 1 satellite features and the character of the metal–CO interaction
monolayer MgO to thick MgO films with bulk properties is could be established in a clear and direct fashion. Based on
0.15 eV measured with an unmonochromated laboratory XPS extensions of this early work [127,128,130] and on the
system [35]. If one takes into account an additional broadening of understanding of the CO XPS features [133], we expect that
1 eV for instrumental resolution and convolutes this with the the analysis that is described will have a general validity. The
calculated FC vibrational broadening, the theory would predict a model, used in this work, was a simple linear cluster with one
reduction in the broadening going from bulk to one monolayer Ni atom and one CO molecule, NiCO, which allowed an easy
MgO of 0.18 eV. This is close to the experimentally measured interpretation of the XPS in terms of covalent bonding between
difference and indicates that the assignment of the chemical origin the metal substrate and CO. The natural state to use as the
of the change is correct. Although the changes in FWHM due to initial state, before core ionization, is one of the low-lying
the chemical interaction of the film with its support are small, states that can be derived from the ground state 3d94s1
they are easily measurable. This work has demonstrated a new configuration of the Ni atom plus the closed shell CO
use of XPS to characterize the chemistry of materials. molecule. However, with this configuration back-donation
from the Ni conduction band to the CO(2πn) cannot occur
4. XPS satellites: covalent interactions and screening since the one Ni atom does not have any sp electrons with π
symmetry. Yet, as confirmed later [54,130], back-donation is
In this section, we consider two cases from early work: (1) quite important, especially for hole-states of chemisorbed CO
CO and N2 chemisorbed on metal surfaces [127–131] and (2) [130]. Consequently, for this cluster, an excited state, denoted
the NO dimer [129], (NO)2. In the first case, we focus on the GS(π), derived from the 3d94p1 configuration of Ni was used
288 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

as the initial state [131]. This state allowed back-donation or

covalent bonding between the metal sp conduction band and
CO(2πn) and provided a direct explanation of the origin of the
XPS satellite structure of chemisorbed CO.
The GS(π) initial state has a configuration:
……ðvπÞ1 …ð1sÞ2 ð2sÞ2 …ð1πÞ4 ; ð13Þ
where only the Ni and CO orbitals of interest are shown.
The orbital labeled vπ is bonding between Ni and CO and, for
Ni–C distances near that of CO from the Ni surface, vπ has
large amounts of both Ni 4pπ and CO 2πn character. Thus, vπ
represents the back-donation from the Ni sp conduction band
to CO(2πn). The orbitals labeled 1s and 2s are the O(1s) and
C(1s) orbitals of CO, respectively and 1π is the occupied CO π
orbital, albeit modified by the presence of the Ni. The
configurations for the XPS core-hole states are as follows:
……ðnπ′Þ1 …ð1sÞ1 ð2sÞ2 …ð1πÞ4 and ð14aÞ

……ðnπ′Þ1 …ð1sÞ2 ð2sÞ1 …ð1πÞ4 ; ð14bÞ

for each hole, the Hartree–Fock equations are solved for two
different states. For the lowest state the valence π orbital,
denoted 1π′, is occupied and for the excited state, the 2π′
orbital is occupied. We use the notation nπ′ to avoid confusion
with the 1π orbital associated with CO. The main difference
between the two states is for the 1π′ and 2π′ orbitals. The
character of the 1π′ and 2π′ orbitals are best understood by first
considering CO at a large distance from Ni so that there is little
interaction. The ionization potential of a Ni atom, as well as
the work function of a Ni surface, are ∼5 ev and the electron
affinity of CO with an O(1s) or a C(1s) core-hole, based on the Fig. 6. BEs and relative intensities, I, for the two C(1s) hole states as a
function of the Ni–CO distance, dNi–C; the states are indicated with subscripts
energies of the equivalent core approximation molecules as BE1, BE2, I1, and I2. (a) Energies where the left scale is for the BE and the
[41,42], is ∼10 eV; see Ref. [131]. Thus, at large separation right for ΔBE¼ BE2−BE1. (b) Relative intensiies where the left scale is for the
1π′ is simply the CO(2πn) orbital in the presence of the CO individual I1 and I2 and the right scale is for the ratio I1/I2.
core hole and 2π′ is the Ni 4pπ orbital. From the energy
balance above, the difference in energy of the two states is two states will be Irel(1) ¼ Irel(2) ¼ 0.5. The Irel are shown as a
∼5 eV. As the distance between Ni and CO is shortens, the 1π′ function of CO dNi–C in Fig. 6(b). The main conclusion is that
orbital acquires Ni 4pπ character and is bonding between Ni one must compare the covalent character of the ground state
(4pπ) and CO(2πn). On the other hand, the 2π′ orbital acquires and the core-hole state orbitals to understand the intensities of
CO(2πn) character and is anti-bonding between Ni and CO. In main and satellite peaks. It is not sufficient to look only at the
Fig. 6(a), we give the C(1s) BEs for the two final states, covalency in the initial state. We will return to the importance
BE1 and BE2, as a function of the distance between Ni and C, of considering both initial and final state covalency in a later
dNi–C; ΔBE ¼ BE2−BE1 is also given. For convenience, section concerning the satellite intensity in the XPS of oxides.
dNi−C ¼ 0 is chosen as a distance near the distance of CO
above Ni(100); thus dNi−C o 0 and dNi−C 4 0 correspond to 4.2. XPS satellites for the NO dimer
distances shorter and longer than equilibrium. The BEs and the
ΔBE depend only weakly on dNi−C. Similar weak distance Thin films of NO adsorbed on Ag surfaces and nanoparticles
dependence is found for the O(1s) BEs [131]. The situation is are known to form dimers; see, for example, Refs. [129,
quite different for the XPS relative intensities [38]. Neglecting 136–139] and references therein. There is a significant and
the relaxation of the other passive orbitals, Irel(n)≈|〈vπ|nπ′〉|2 ongoing interest in the photochemistry of NO on Ag; in
where vπ is the GS(π) ground state valence π orbital defined in particular, since these studies have implications for catalytic
Eq. (13) and the nπ′ orbitals are the final, core-hole state activity [136–139]. Since the NO dimer is weakly bound
orbitals defined in Eq. (14). At large dNi−C, the limiting values [140], it is expected, as discussed above, that the XPS of the
are Irel(1) ¼ 0 and Irel(2) ¼ 1. For small separations, Irel(1) dimer will be complex with intense satellites. Furthermore,
becomes close to 1 since both vπ and 1π′ are bonding between there is a potential that the XPS, properly interpreted, can help
Ni 4pπ and (2πn) and Irel(2) becomes close to 0 since 2π′ is to identify presence of dimers. To demonstrate this potential,
anti-bonding. At an intermediate distance, the intensities of the we describe an early combined theoretical and experimental
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 289

XPS study [129] that explained the origin of the complex O states in order to gain an understanding of the character of the
(1s) and N(1s) spectra of (NO)2 in terms of the screening of the screening in each of the core-hole states. It was relatively
core-hole. The XPS measurements discussed are for condensed straightforward to do this for the previous example of CO/Ni
multilayers of NO on Ag(111) where dimers are formed. The because it was possible to focus on a single orbital and to show
theory considers an isolated (NO)2 dimer at the experimental how the metal and CO character of this orbital changed as a
geometry where the core-hole is localized on the N or O atoms function of the distance between the CO and the Ni surface.
on one of the two NO units [72]. In contrast to the NiCO For (NO)2, there are two electrons, one associated with each
model of CO/Ni, where we could examine changes in the one of the in-plane 2πn orbitals on each NO unit and it is not
covalent character of a single valence electron, the interaction possible to follow a single orbital. For this reason, the
and the covalent bond between NO units in (NO)2 involves positions of the centers of charge are used to distinguish the
two electrons arising from the 2π orbitals on each of the NO characters of the different final states. These positions indicate
units. The wavefunctions that describe the weak NO–NO the character of the response or screening of the “passive”
interaction for both the initial and final core-hole configura- electrons to the core hole. The significance of the centers of
tions are inherently multiconfigurational. They involve mix- charge can be understood by considering the FO final state of
tures of configurations with different distributions of the two Eq. (4). Since the core hole is localized on one of the N or O
valence electrons over the two 2π orbitals. This mixing is large atoms, the center of charge will be essentially on the nucleus of
because the configurations are nearly degenerate. Recall that the core-ionized atom. Of course, once the response of the
near degeneracy is not measured by the difference in diagonal passive orbitals is taken into account, the position of the center
energies but by the ratio of the off-diagonal matrix element, of charge will change significantly, As we discuss below, this
Hij, to the difference of the diagonal matrix elements, Hii−Hjj. displacement of the center of charge for the various states
In Section 6, we describe other cases where configuration indicates the character of the screening of the core-hole in the
mixing in multiconfigurational CI wavefunctions, Eq. (5), is different states. When we discuss the covalent bonding and the
essential to properly describe XPS. screening in metal oxides in the next section, it will be
We found that there are two states, separated by ∼4 eV, that necessary to use still another measure to distinguish the
carry significant intensity for both N(1s) and O(1s). There is a character of the states.
third final state for the N(1s) XPS that carries modest intensity In Fig. 8, the centers of charge for the O(1s) and N(1s) hole
but is very near the main satellite. In Fig. 7, we compare the states are indicated by a ■ and a Δ, respectively, and the core
XPS of condensed NO on Ag(111) with our theoretical hole is on the right NO denoted β in the figure. For the O(1s)
predictions. The theory gives intense satellites and correctly holes, the center of charge for the first excited state is shown
predicts greater satellite intensity for N(1s) compared to O(1s) even though its XPS Irel is small and it is not plotted in Fig. 7.
The main limitation of the theory is that the satellite BEs are The positions of the centers of charge for the O(1s) holes are at
∼1 eV too high. It is possible that this error arises because the the level of the N atoms, even though the hole is created on the
calculations do not include the environmental effects present in O atom in the lower right of the figure. This shows that the
the condensed NO dimers on the Ag surface. The close charge moves from the N atoms to fully screen the core hole
correspondence between theory and experiment is a clear on O. The equivalent core molecule of NO with an O(1s) core-
indication that NO dimers are formed in the overlayers. hole is NF+, and the position of the center of charge being at
Furthermore, and perhaps even more important, it is possible the level of the N atoms, indicates that the NF+ molecules is
to examine the character of the wavefunctions for the different best described as being N+ and F0 which is not surprising since
it is easier to remove an electron from N than from F. The
center of charge for the 2nd excited O(1s) hole state,
corresponding to the ∼3 eV XPS satellite is almost at the N
atom of NO β the core-ionized NO, and shows, for this state,
that the electrons of NO α, which is not core-ionized, are not
significantly involved in the screening of the O(1s) core-hole

Fig. 7. XPS for condensed NO/on Ag(111) compared to multiconfigurational

theory for isolated (NO)2; see Ref. [129]. Fig. 8. Centers of charge for the various hole-states of (NO)2; see Ref. [129].
290 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

on NO β. The screening in this case is purely intra-unit and the strong dependence on the parameters of the basis sets used
electrons on the other NO are essentially passive. The screen- to expand the orbitals in the wavefunctions for oxides and
ing of the low intensity 1st excited O(1s) hole state is also other systems [113]. Another approach used to characterize the
intra-unit. On the other hand, the center of charge of the lowest distribution of charge in a molecular system is based on the
O(1s) hole state is almost halfway between the two NO units. analysis developed by Bader [147,148]. The Bader analysis
This shows there is significant involvement of the electrons on assigns the charge in certain regions of space to be assigned to
the un-ionized NO α to screen the core-hole on NO β; i.e., a particular atom in the system based on the topological
there is significant intra-unit screening. For the N(1s) core- properties of the charge distribution. However, charge in a
hole, the y coordinates, see Fig. 8, of the centers of charge for molecule or in a condensed system does not belong to
all three states are roughly mid-way between the N and O individual atoms but is shared between atoms; this is the
atoms. Since the equivalent core molecule for this case is O+2, a essence of a chemical bond. Thus, as well as assigning an
homopolar molecule, it is expected that the center of charge effective charge to an atom, it would be desirable to be able to
would be near the center between the N and O atoms. For the estimate the uncertainty in the assignment of the atomic
N(1s) hole states, the centers of charge are either near the charge. Here, we focus on the use of orbital projections
center between the two NO units or, in the case of the lowest [10,149] to estimate the number of electrons or, equivalently,
state, even closer to the NO α unit than to the core-ionized NO the effective charge to be associated with an atom or a
β unit indicating considerable inter-unit screening. The lowest fragment of the complete system. This approach is taken
N(1s) state, corresponding to the main XPS peak is very close because, as we show below, projection often allows us to
to NO α indicating that NO α has almost lost all of its 2π estimate the uncertainty of the assignments of charge [60]. The
electron and is close to NO+, while the 2π occupation on core- projection involves constructing an projection operator, φiφ†i ,
ionized NO β is almost 2. The centers of charge provide a where φi is an orbital of an atom or a fragment whose
better understanding of the inter-unit screening of a core-hole occupation is to be determined. The expectation values of this
than the simplified picture of charge transfer, CT, where a operator are then taken for the wavefunctions of the total
binary view that there is CT or there is not CT, see Eq. (11), is system to determine the desired occupation; the methodology
often used and the covalent character of the orbitals is is described in more detail below in connection with applica-
neglected. tions to selected oxides. It is appropriate to address why a
reliable quantification of the departures from nominal oxida-
5. Metal oxides and covalency tion states is a key piece of information for understanding the
properties of this class of materials, in general, and the analysis
It is common to describe oxides, as well as other ionic of their XPS, in particular. The nominal charges obtained from
compounds, in terms of their oxidation states; thus, for MnO, the stoichiometry of an oxide, or other ionic compound, carry
one describes the Mn oxidation state as +2 with a 3d5 no information about the covalent mixing of metal and ligand
occupation. For high spin oxides, the open shell electrons orbitals. Indeed, these nominal charges neglect the overlap of
are commonly assumed to have a maximum spin alignment metal and cation orbitals, which is critical in the formation of
although, for heavy metal oxides, this assumption, which chemical bonds. However, a knowledge of the covalent
neglects the spin–orbit coupling in the open shell, has been mixings has the potential to give direct insight into the
questioned [141]. Our principle concern in this section is to reactivity and catalytic activity of the material. For the
demonstrate that the purely ionic view is misleading and that, interpretation of the XPS features, we show, in this and the
very often, there is a significant covalent character to the following sections, how the covalent character is related to
cation-oxygen bonds in metal oxides. Moreover, we will show the distribution of intensity between main peaks and satellites.
that the covalent character may be strongly different for initial This opens the way to use the XPS spectra directly to draw
states and for core-hole states [10,22,60]. The discussion of the inferences about the electronic structure and the chemical and
covalent character of oxides is key to understanding the physical properties of the material.
significance of the XPS of these materials, especially their We first use graphical views of the charge densities and
satellite structure [52,53]. A knowledge of the covalent orbitals to contrast the properties of MgO, a nearly ideal ionic
character of the interactions in oxides is also needed in oxide [35], and MnO, where there is some covalent character
properly characterize the electronic structure of these ionic [10]. Then, we consider quantitative estimates of the cation
systems. charge states, which are obtained by the projection of orbital
First, it is necessary to establish theoretical methods and character on the oxide wavefunctions. As well as these simpler
criteria that are suitable to characterize and to quantify the oxides, we also consider lanthanide and actinide oxides to
covalent character of an interaction. The covalent character of show that their covalent character is unexpectedly large; in
bonds have commonly been quantified using population particular for core-hole configurations. In Section 3, we
analyses [142–145] to determine cation shell occupations and introduced the term closed-shell screening to describe the
charges in oxides; however, these analyses can give misleading changes in covalent character for core-hole configurations.
information; for a particularly serious error in the charges from Here, we discuss how the large closed shell screening in CeO2
a Mulliken population analysis, see Ref. [146]. Clearly, the affects the XPS Franck–Condon vibrational broadening (first
atomic charges obtained from population analyses have a introduced in Section 3 for MgO). In Section 6, we consider
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 291

other, more general, effects of covalency on the XPS of metal from the superposed spherical atoms to the HF Ψ is over
oxides. 10 eV, almost 4 times the change for the MgO wavefunctions.
Charge density difference plots are shown in Fig. 9 for An important cause for this large energetic improvement for
embedded cluster models of the cubic oxides MgO and MnO MnO is the covalent character of the MnO cluster orbitals. We
[60], where experimental geometries and bond distances are give in Fig. 10, contour plots of two eg orbitals taken from HF
used [150]. The plots are through a plane (denoted xy) that wavefunctions for the embedded MnO6 cluster [10]; the
contains a central metal cation and 4 nearest neighbor O anions orbitals for the initial state configuration are plotted in
near the corners of the square. The density difference, Δρ, is Fig. 10(a) and (b) and the orbitals for the 2p-hole configuration
the difference between ρ for the HF Ψ and an anti-symmetric are plotted in Fig. 10(c) and (d). The same conventions are
Ψ formed from superposed spherical atoms, which are HF used for Fig. 10 as for the plots in Fig. 9 except that the signs
solutions for the isolated ions: Δρ ¼ ρ(HF)−ρ(Spherical). The shown by the solid, dashed, and dotted lines now refer to the
wavefunctions for the superposed atoms take the Pauli exclu- orbital, φ, rather than the density difference, Δρ. The two
sion into account [151,152] and, hence, the Δρ shown in Fig. 9 orbitals plotted are the closed shell, e4, orbitals, Fig. 10(a) and
represent chemical changes in ρ(HF). The solid lines in Fig. 9 (c), that are dominantly of O(2p) character and the open shell,
are for contours of constant Δρ4 0, the dashed lines for e2, orbitals, Fig. 10(b) and (d), that are dominantly of Mn(3d)
contours of constant Δρ o 0, and the dotted lines are for character. We have plotted the xy component of the degenerate
Δρ¼ 0. The contours are for uniform steps of Δρ and there are pair of e orbitals since these components can form covalent
cutoffs for large magnitudes of Δρ. The simpler case of MgO, combinations with the O ligands in the plane shown. In the
in Fig. 9(a), is discussed first. There is a polarization of the semi-empirical Anderson Model formulation [12–14], these
spherical O towards the Mg2+ cation at the center. There is orbitals are treated as though they were pure metal or ligand
also a departure of the O charge density from spherical away orbitals. Clearly this is not the case. Although, the closed shell
from the central Mg and toward the neighboring O anions orbitals, shown in Fig. 10(a) and (c), are dominantly p orbitals
indicating the formation of the O(2p) band. The small changes on O, there are contours that extend to the Mn atom. Further,
that occur in the immediate region of the central Mg arise as a the combination is a bonding combination of O(2p) and Mn
result of the incomplete basis sets [153,154] used and do not (3d) in this closed shell orbital. The open shell orbitala, shown
represent actual charge flow. There is, however, no buildup of in Fig. 10(b) and (d), are dominantly Mn(3d) and the d(xy)
charge between O and Mg as would occur if a covalent bond character of these orbitals are obvious. However, there is some
were formed. Another indication of the ideal ionicity of MgO O(2p) character and the node, or zero, between O and Mn
comes from the difference in the energies of the HF and demonstrates that the orbitals are anti-bonding. Before turning
superposed spherical atoms wavefunctions; the energy of the to quantitative estimates of the covalent mixing, we use orbital
HF wavefunction is lower by only 2.6 eV. Further, a CSOV contour plots, to show that the covalent character between O
decomposition [44–46] shows that 90% of this energy change (2p) and Mn(3d) is different for the initial and the final, core-
comes from the reorganization of the O charge due to the hole configurations. Comparing Fig. 10(c) and (d) with Fig. 10
presence of the Mg2+ cation and not from a departure from the (a) and (b), it is clear that the covalent character, for both the
perfect ionicity assumed in the superposition of spherical filled bonding orbital and the half filled anti-bonding orbital, is
charges. The situation is very different for MnO, Fig. 9(b), larger for the 2p-hole configuration than for the initial state
where there is a buildup of O charge between O and Mn. This configuration. The orbital plots in Fig. 10 are for non-
buildup appears to have a symmetry that would arise from relativistic HF wavefunctions [10]. Although, these plots are
orbitals of e symmetry in the Oh point group of MnO. only for eg orbitals, there is also closed shell bonding and open
Furthermore, the energy change for the wavefunctions going shell anti-bonding covalent character for the t2g orbitals [10].

Fig. 9. Charge density difference contours for (a) MgO and (b) MnO; see text.
292 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

Fig. 10. Orbital contour plots of the dominantly O(2p) closed shell, denoted e4, and the dominantly Mn(3d) open shell, denoted e2, orbitals for the embedded MnO6
cluster model of MnO: (a)e4 orbital for the initial state MnO6 configuration; (b) e2 orbital for the initial state; (c) and (d) e4 and e2 orbitals for the final state 2p-hole
configuration.

In order to quantify the covalent character of the metal– in Eq. (15) can be limited to a subset of the orbitals or even to
oxygen interaction, we turn to a discussion of the projection of a single orbital; as we show below, the properties of subsets of
the orbitals of the isolated cation on the cluster model orbitals the orbitals may help explain the chemistry. Second, the
to determine the occupation of the cation orbitals in the occupation NP depends weakly on the choice of configuration
wavefunctions for the condensed phase [149]. The specific for the isolated atom [149,155] and all reasonable choices will
summation to obtain the 3d occupation in MnO from projec- give the same physical picture. Third, as with all assignments
tion, NP(3d) is of atomic charges, the value of NP is influenced by the overlap
of orbitals on different atoms [142–145,149]. In other words,
N P ð3dÞ ¼ ∑i occ ðiÞ〈φi jφð3dÞφ† ð3dÞjφi 〉; ð15Þ
because chemical bonding involves the sharing of electrons
†
where the projection operator, φ(3d)φ (3d), is for a 3d orbital between atoms, it is not possible to uniquely assign electrons
obtained for an isolated atom and the matrix element is simply to atoms when covalent bonds are present. We describe this an
the square of the overlap integral 〈φi|φ(3d)〉. The sum is uncertainty in the assignment of charges to the atoms in a
weighted by the occupation of the orbitals in the MnO compound and we describe below ways to use internal
configuration; the sum over the different atomic 3d orbitals consistencies in order to obtain estimates of the uncertainties.
is not given explicitly in Eq. (15) but, in practice, is always The NP(3d) for MnO obtained from projections on relati-
taken. As written, Eq. (15) is strictly valid for a single vistic and non-relativistic wavefunctions for the ground and
configuration or for a single CSF. The projections could also 2p-hole configurations of an embedded MnO6 cluster are listed
be made for the more general CI wavefunctions of Eq. (3) but in Table 9 [10]. The projections are grouped into the closed
we restrict ourselves to the simpler formalism for a single shells of the cluster where the nominal number of d electrons is
configuration since our primary interest is to identify the zero and the open shells where the nominal number of five d
covalent character of the bonding. The generalization of electrons is distributed, in the ground state, into 3t2g and 2eg.
Eq. (15) to other projections and other systems is straightfor- For both the relativistic and non-relativistic wavefunctions, the
ward [149]. Three comments are in order. First, the summation projections are for the average of configurations with orbitals
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 293

Table 9 although, we would expect to see similar exaggerations for

Projections, NP(3d), on relativistic and non-relativistic wavefunctions of an both the initial and the final state configurations. An estimate
MnO6 cluster model of MnO for Initial, un-ionized, and Final, 2p-hole,
of the effect of the overlap on the value of NP(3d) can be made
configurations. The difference of the initial and iinal NP are given as Δ.
by projecting the Mn 3d orbitals on the wavefunction for O6
Relativistic Non-relativistic [10,158]. This projection is 0.17 and it is very likely to be an
upper bound for the “error” of the value of NP(3d) [10,158].
Initial Final Δ Initial Final Δ
While there are formulations which use projections based on
Closed shells 0.46 1.07 +0.61 0.45 1.12 +0.67 the overlaps to form a population analysis [159,160], we prefer
Open shells to use the overlap projection as an estimate of the “uncer-
t2g 2.93 2.84 −0.09 2.93 2.83 −0.10 tainty” of assigning 3d electrons to the cation because the
eg 1.83 1.58 −0.25 1.83 1.56 −0.28
electrons are shared between the cation and ligand. Although,
Sum 4.75 4.41 −0.34 4.77 4.39 −0.38
Total 5.22 5.48 +0.26 5.22 5.51 +0.29 this uncertainty reduces the covalent character of the interac-
tion, a modest amount remains. There is clearly a significant
covalent character to the interaction for both the initial and
optimized for this average [68]. For the relativistic case, the final, core-hole, configurations. Furthermore, the increase of
division into t2g and eg is approximate since, in the Oh double the total NP(3d) for the core-hole configuration, shows that
group [156], the appropriate representations are the doubly there is closed shell screening of the Mn 2p-hole by the O
degenerate γ7 and the four fold degenerate γ8, where t2g and eg electrons. The overall view is that MnO is dominantly an ionic
can mix in the γ8 spinors; however, the mixing of eg and t2g is system with modest covalent character. In the following
small for MnO. paragraphs, it will be shown that closed shell screening is
Not surprisingly, the relativistic and non-relativistic projec- even larger for heavy metal oxides and contributes to the
tions are quite similar and our discussion focuses on the complexity of their XPS.
relativistic projections. For the initial state configuration, the We next consider closed shell CeO2 and UO3 modeled with
covalent bonding in the t2g and eg closed shells leads to an Mn embedded CeO8 (Fig. 3c) and UO6 (with the δ-UO3 coordina-
3d occupation of almost half an electron in these shells. tion according to Ref. [161]) clusters and discuss covalency
Because of the anti-bonding character of the t2g and eg open only for relativistic wavefunctions [60]. The nominal oxidation
shells, the open shells have an NP(3d) ¼ 4.75, reduced from the states and occupations are Ce(IV)–(4f)0 (5d)0 and U(VI)–(5f)0
nominal value of 5. As shown in Table 9 for the open shell (6d)0, respectively. Since these are closed shell systems, the
orbitals, the covalent character of the t2g orbitals is only about occupations given for the Ce 4f and 5d and for the U 5f and 6d
25% of that for the eg orbitals. The smaller covalent mixing of can only arise from closed shell orbitals that have bonding
the t2g orbitals follows because the t2g and eg orbitals are character between the metal and ligand. A novel approach is
directed between and towards the O atoms, respectively used to estimate the uncertainty of the occupations, NP,
[10,157]. Summing the t2g and eg covalent mixings, the open obtained with the projection analysis [60]. As well as project-
shells have lost ∼0.25 d electrons because they now have anti- ing the cation orbitals on the embedded clusters, we also
bonding mixing with the O electrons. However, the total project the orbitals of an O6 cluster (for UO3) or an O8 cluster
number of Mn 3d electrons has increased by 0.2 electrons over (for CeO2). Bare On clusters were also used in earlier work to
the nominal value of 5, see Table 9, because of the covalent determine the energetic importance of the O polarizations and
bonding character of the fully occupied closed shell orbitals. bonding in oxides with lighter metals [109,126,162]. The
This leads to a reduction of the net charge on the Mn cation present interest is to use the projections of the O(2p) character
below the nominal value of +2. Similar comments apply to the as a guide to counting electrons involved in the covalent
2p-hole state orbitals except the covalent mixing of these interactions [60]. Since the On clusters also have cubic
orbitals is substantially larger and there are now almost 5.5 3d symmetry, the orbitals, or spinors, belong either to g or u
electrons associated with Mn. The 0.26 increase of NP(3d) for point group representations. By symmetry, only the O(2pg)
the 2p core-hole configuration is a “closed shell” screening of electrons can mix with the central metal atom d orbitals and
the 2p core-hole. Because of this closed shell screening, there only the O(2pu) electrons can mix with the central metal atom f
may not be a need to invoke a change of configuration with a orbitals. For the O(2p) projections, the quantity ΔNP(O2p), the
transfer of one or more electrons from the ligands to the metal number of 2p electrons lost from the full 2p occupation in the
3d; see Eq. (10), as is done in the semi-empirical Anderson bare On cluster, is used rather than the total number of 2p
model approach. electrons, NP(O2p). If there were no uncertainties in the
When the metal 3d and ligand orbitals overlap, the bonding projections, then NP(metal cation orbital) ¼ −NP(O2p), where
and anti-bonding character of orbitals affects the apparent the metal cation orbital is one of the d or f orbitals listed above
occupations, NP, determined from projection [149]. Specifi- and the O2p projection is restricted to either g or u orbitals
cally, the overlap of the ligand and metal orbitals will act to depending on whether the metal orbital is d or f. In other word,
increase the NP(3d) for the dominantly ligand bonding orbitals the number of electrons gained by the cation orbital is equal to
and decrease NP(3d) for the dominantly metal open shell the number lost from the O 2p ligand orbitals. Because there
orbitals. The net effect is that the overlap of the metal 3d and are uncertainties in the projection due to the overlap of the
ligand 2p orbitals will exaggerate the covalent character; metal and ligand orbitals, this equality will not be satisfied
294 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

exactly. The extent of the uncertainty in the assignment of Table 9. When a 3d or 4d core-hole is made on Ce, the closed
occupation based on the projections can be estimated from the shell screening is large as judged from both NP(4f) and
extent to which the magnitudes of the two projections are not ΔNP(O2pu). Furthermore, the closed shell screening is larger
equal. We consider first the projections, in Table 10, related to for the deeper 3d core-hole than for the 4d core-hole; see
the occupation and covalent bonding with the outermost metal Table 10. The different screening occurs naturally with our
nf orbital for both the initial and the final core-hole config- wavefunction based theory [21,22,60] while for semi-empirical
urations. The difference of the NP between the initial and the Anderson model theories, shallower or deeper core-holes are
final configurations is given as Δ(Initial) and it provides a distinguished by empirically adjusting parameters to fit
direct measure of the closed shell screening. observed XPS spectra [112,163]. Furthermore, the core-hole
For the initial state of UO3, the occupation of the “nominally is screened through a change in the chemistry of the core-
empty” 5f shell is 1.4 electrons and the loss of the O(2pu) ionized atom with its surroundings rather than a change in
electrons is −0.8. Although these two estimates of the covalent configuration, see Eq. (10), as required in semi-empirical
interaction of the U(5f) are different, they both indicate that the theories [8,12–14,112]. Finally, we point out that the uncer-
covalent mixing in bonding orbitals leads to a large 5f tainty in the assignment of charges to atoms is a natural
occupation of nearly 1 electron. It is critical to stress that this consequence of chemical bonding and interactions; when
5f electron character is not in an open shell; the UO6 cluster bonds form, electrons are shared between atoms and cannot
does not have open shells. Rather, the 5f character arises from be uniquely assigned at one atom or another.
the covalent bonding in the dominantly O(2p) closed shells. The uncertainty of the assignments of metal character due to
For the 4f-hole state, there is a large increase of NP(5f) by 1.2 covalent mixing with ligands is seen very clearly for the
electrons and a corresponding reduction of ΔNP(O2pu) by 0.9 analogous projections for the outermost d shells of U and Ce
electrons. The magnitudes of the changes in NP(5f) and and the O(2pg) shells shown in Table 11 [60]. The projection
ΔNP(O2pu) occupations between the initial and 4f-hole con- of the nominally unoccupied cation 5d or 6d metal orbitals on
figurations, Δ(Initial), are more similar than the agreement of the initial state wavefunctions for the embedded cluster models
the absolute values of NP and ΔNP for either the initial or 4f of UO3 and CeO2, give large NP(nd)≈3. However, the
core-hole configurations. As we noted in the discussion above magnitude of ΔNP(O2pg)≈−0.6 is much smaller and indicates
for the MnO 3d occupation, the errors or uncertainties in the that there is substantial uncertainty in the actual occupation of
NP will be similar for the initial and final, core-hole config- this d shell. This uncertainty arises because of the larger spatial
urations. We can view this as a cancellation of errors in the NP extent of the d orbitals than of the f orbitals and, thus, a much
which give more accurate values for the closed shell screening; larger overlap of the d and the O(2p) orbitals. While, the d
there is a similar cancellation for CeO2, see below. The orbitals do have some participation in the covalent interaction
|Δ(Initial)| differ by ∼0.3, which is less than half of the with oxygen, it is not possible to give a precise value for the d
differences between |NP(5f)| and |ΔNP(O2pu)|; see Table 10. occupation. The important fact is that the NP(nd) and
The Δ(Initial) values give strong support to a closed shell ΔNP(O2pg) for the core-hole states are very similar to the
screening of the 4f-hole by∼1 electron. The large initial state values for the initial states; see Table 11. The changes between
participation of the U(5f) in covalent mixing with the O(2pu) initial and final states are not especially large and are
orbitals is, at least in part, a consequence of the large nominal considerably smaller than for the projections of the nf and O
positive charge of +6 on the U cation. For CeO2, where the (2pu) shells shown in Table 10. The participation of the d
nominal Ce charge is +4, NP(4f) ¼ 0.32 is smaller than the shells in the closed shell screening of core holes is much
cation occupation for UO3 but 50% larger than the cation smaller than that for the f shells. There is a simple chemical
occupation in MnO where the nominal ionicity is only +2; see explanation for this difference. The outer f shells are more

Table 10 Table 11
Projections, NP or ΔNP, of outer f and O(2pu) shells for initial and hole-states Projections, NP or ΔNP, of outer d and O(2pg) shells for initial and hole-states
of UO3 and CeO2. of UO3 and CeO2.

System State Projection System State Projection

Shell NP/ΔNP Δ(Initial) Shell NP/ΔNP Δ(Initial)

UO3 Initial 5f +1.40 – UO3 Initial 6d +2.99 –

O(2pu) −0.77 – O(2pg) −0.61 –
4f-hole 5f +2.59 +1.20 4f-hole 6d +3.11 +0.12
O(2pu) −1.69 −0.92 O(2pg) −0.55 +0.06
CeO2 Initial 4f +0.32 – CeO2 Initial 5d +2.51 –
O(2pu) −0.10 – O(2pg) −0.52 –
4d-hole 4f +1.05 +0.73 4d-hole 5d +2.84 +0.33
O(2pu) −0.57 −0.47 O(2pg) −0.55 −0.03
3d-hole 4f +1.47 +1.16 3d-hole 5d +2.63 +0.12
O(2pu) −1.08 −0.98 O(2pg) −0.48 +0.04
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 295

contracted than the outer d shells [164]. Since, there is an valence open shell to give a total multiplet the same as when
increase in the number of f electrons by ∼1, the effective the core-hole is coupled to the ground state multiplet of the
charge seen by the nd electrons is, to a large extent, valence open shell [30,166,167].While these angular momen-
unchanged. Thus, the closed shell screening by covalent tum re-coupled multiplets are XPS forbidden, they are able to
mixing to form bonding orbitals between the ligand and nf mix with the XPS allowed multiplets and thus lead to
orbitals cancels the increase in charge due to the core-hole and “satellites” with large intensity [30]. An example of the
the electrostatic driving force for closed-shell screening into many-body mixing of XPS forbidden with XPS allowed
the nd orbitals is largely absent. multiplets in given in Section 6.2. For all these reasons, we
The large closed shell screening in the heavy metal oxides chose to group the discussion of XPS multiplets with many-
has important consequences for the Franck–Condon vibra- body effects. In Section 6.1, we consider the multiplet splitting
tional broadening of the XPS peaks. In a previous section, it of the N(1s) and O(1s) XPS of NO [17]. We also examine the
was shown that a core-hole on the Mg cation in MgO leads to Mn 3s XPS [4], where multiplet theory is not sufficient. In
a 0.12 Å decrease in the Mg–O distance and this leads to a Section 6.2, angular momentum recoupling is introduced as a
vibrational broadening of the Mg 2p FWHM by ∼0.8 eV [35]. major correction to the simple angular momentum coupling
Because, as discussed above, the closed shell screening of a theory. This recoupling is important because the re-coupled
core-hole in MnO is only modest in size, ∼0.2 electrons, the CSFs are nearly degenerate with the XPS allowed multiplets.
Mn–O distance for a 2p core-hole on Mn is reduced by 0.13 Å Other types of near degeneracy are also discussed that provide
and this leads to a broadening of the Mn 2p FWHM by 0.9 eV a more complete description of “atomic” processes than
[165]. On the other hand, calculations of the Ce–O distance for multiplet theory alone. In this section, the effects largely
a core-hole on Ce lead to only a small change in the Ce–O depend on orbitals with dominantly atomic character but
distance. In fact, for a 4s core-hole on Ce, the Ce–O distance which can be modified by ligand field splittings and covalent
for the core-hole state is actually larger, but by only 0.01Å, bonding. In Section 6.3, we consider very recent results
than for the initial state [165]. This is because, as suggested by [21,22,52,53] for the XPS of the two closed shell oxides,
the projections, the core-hole is actually over-screened by the CeO2 and UO3 whose covalent character was discussed in
final state covalent interaction between the O(2p) and the Ce Section 5.
(4f). With this small change in the Ce–O distance, the
vibrational broadening must be small and there must be a 6.1. Multiplet splitting
different origin for the large broadening observed in the CeO2
XPS [20–22]. NO is an open shell molecule with a filled bonding 1π
orbital and a single electron in the anti-bonding 2π orbital. The
6. XPS satellite features and many-body effects configurations of the 2Π ground state and the N(1s) and O(1s)
hole states are
In prior sections, we primarily considered PES main peaks
except for Section 4 where satellites were considered specifi- 1s2 2s2 ⋯1π4 2π1 ð15aÞ
cally in relation to covalent bonding and anti-bonding orbitals,
especially for cases of weak chemical interactions. In the 1s1 2s2 ⋯1π4 2π1 ð15bÞ
present section we expand this earlier treatment to consider a
1s2 2s1 ⋯1π4 2π1 ; ð15cÞ
wide variety of many-body effects that can lead to satellites,
often with very large intensities. A suitable definition for where 1s is the O(1s) shell and 2s is the N(1s) shell. The spins
many-body effects is that they require the use of wavefunc- of the single 1s electron and the single 2π electron can couple
tions that cannot be represented by a single configuration or a parallel to 3Π or anti-parallel to 1Π multiplets [17,71,88]. In the
single CSF. In particular, for closed shell systems, this means HF approximation, the energy difference, ΔE between 3Π and
that we must go beyond a single determinant description of the 1
Π multiplets is 3K(ns,2π) [88], where K is the exchange
wavefunction. Final states that correspond to multiplets are integral between the core-hole and the open valence 2π shells.
also considered a many-body effect, although the reasoning Since there are 6 states in the 3Π and but only 2 states in
requires explanation. A multiplet can be regarded as a the 1Π multiplet, the ratio of the intensities is expected to be
degenerate set of states that arise from the angular momentum I(3Π)/I(1Π)¼ 3 [17,168]. The 3Π−1Π multiplet splittings in
coupling of the open shells in a single configuration [88] and, NO were calculated with Hartree–Fock (HF) wavefunctions
hence, not a many-body effect. On the other hand, wavefunc- determined separately for each multiplet, Eq. (15) [71]. The
tions that are eigenfunctions of the orbital and spin angular measured (calculated) multiplet splittings are ΔE(O1s) ¼ 1.41
momentum operators normally are combinations of determi- (1.35) eV and ΔE(N1s) ¼ 0.53(0.48) eV [71], where the HF
nants [88] and, hence, multiplets can be grouped with many- multiplet splittings are just outside of experimental error. With
body effects. Furthermore, there is a type of angular momen- the resolution available using a laboratory XPS system for the
tum coupling that is a pure many-body effect. This occurs measurements on NO [17], the multiplet, or exchange, splitting
when the valence open shell can couple to multiplets other of the N(1s) XPS peaks could not be resolved as two separate
than the ground state multiplet of the initial state configuration. peaks and the splitting of 0.48 eV was determined by fitting
It is often possible to couple the core-hole with the re-coupled the measured spectra with two Gaussian peaks [17,71]. On the
296 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

other hand, the 1.35 eV separation of the multiplet split O(1s) suggested from the results for NO; see above. The ∼8 eV error
peaks could be resolved in the measured XPS spectra [17]. If in the multiplet splitting is also much greater than would be
the multiplet splittings are to be used as a guide to the chemical expected from the errors of BEs calculated from atomic HF
bonding in NO, it is essential that the very different multiplet calculations [51,70]. Clearly, many-body effects other than
splittings be understood in terms of the characters of the multiplets must make important contributions as discussed in
bonding NO 1π and the anti-bonding NO 2πn orbitals. Indeed, the following sub-section.
based on the character of the 2πn orbital, there is a simple
reason that the O(1s) splitting is ∼3 times larger than the N(1s) 6.2. Near-degeneracy effects: angular momentum recoupling
splitting, The 2πn orbital is dominantly on N; for the initial and FACs
state, 2πn is ∼70%, on N with only ∼30% on O. The
magnitude of the exchange integral, K(ns,2πn), depends Before turning to the Frustrated Auger Configurations,
largely the local character of the 2πn orbital at the center of FACs, that are needed to address the Mn 3s XPS multiplet
the ns 1s orbital and, thus, K(2s,2πn)4 K(1s,2πn). Of course, splitting discussed above, we consider the angular momentum
the distribution of the 1π orbital does not contribute to the recoupling needed to treat core-holes in non-spherically sym-
multiplet splitting since the 1π shell is filled and makes metric shells. Such open core shells lead to a much richer
identical contributions to the energies of both the 1Π and 3Π angular momentum coupling than the s or s levels discussed
final XPS multiplets [88]. However, once we know that the previously. Freeman et al., FBM [37], considered the 3p holes
2πn orbital is dominantly on the N atom, then, from orthogon- for Mn2+ where 3p5(2P) couples with 3d5(6S) to give 7P and 5P
ality, it follows that the 1π orbital must be dominantly on the O multiplets. However, it is possible to find total 5P multiplets by
atom. The experimental and theoretical XPS intensity ratios, I recoupling the d5 shell to either 4D or 4F. These configurations
(3Π)/I(1Π), have also been compared with each other and with can be viewed as a spin flip within the d shell coupled to a
the statistical ratio of 3. The calculated and observed intensity simultaneous spin flip within the 3p shell. While the spin-flip
ratios are both somewhat larger than the expected statistical configurations are XPS forbidden, they can and do mix with
value of 3 but the deviations are less than 25%. This small the one allowed configuration where the d5 electrons remain
deviation from the statistical is relevant for the Mn 3s XPS coupled to 6S. This leads to XPS peaks beyond the two XPS
multiplet splitting, discussed next. Other examples of multiplet allowed 7P and 5P multiplets. Hence, it is properly described as
splitting in molecules can be found in Ref. [6]. an angular momentum recoupling. Bagus et al. [30] also used
The dominantly ionic character of MnO, Section 5, where the atomic model of Mn2+ to describe MnO but they extended
the 5 open shell electrons are ∼90% Mn 3d, even for a deep 2p the FBM treatment to include relativistic effects with a full 4
core-hole, suggests that an isolated Mn2+ cation with a 3d5 component treatment including spin–orbit coupling, which
open shell coupled to the high spin, 6S, multiplet would be a enabled study of both the 2p and 3p XPS. The treatment of
suitable model for the Mn XPS. A direct comparison of the angular momentum recoupling was made in the following
measured 3s XPS in Mn atoms and in MnF2 and MnO crystals way: All determinants were formed by distributing the 5
shows that the spectra are very similar for all these systems, electrons in the ionized NP shell in all possible ways over
which justifies the belief that Mn2+ is a good model for the the 6 np1/2 and np3/2 spinors and the 5 electrons in the 3d shell
XPS of MnO [169]. It should be noted that the atomic XPS in all possible ways over the 10 3d3/2 and 3d5/2 spinors. A
data in Ref. [169] was for Mn atoms and not the Mn2+ cations complete configuration mixing was made over this set of
appropriate for ionic crystals and that the energy separation of determinants and all the wavefunctions were determined by
the first two peaks is slightly larger for Mn0 than for MnF2 and diagonalizing the CI Hamiltonian matrix. The intensities were
MnO. This slightly larger separation is reproduced in the determined using the sudden approximation and broadened
calculated separation of the first two peaks between Mn0 [47] with a Voigt [36] convolution of a 1.0 eV FWHM Gaussian
and Mn2+ [3]. One expects that the energy separations and the and a 0.7 eV FWHM Lorentzian. [30] The theory for Mn2+
relative intensities of the 3s XPS peaks should be similar for and the XPS experiment for MnO are compared in Fig. 11
Mn2+ and Mn0 since the 4s electrons, present in Mn0, are where individual peaks with large intensity as well as the sum
spectators for the core-level ionization, However, there will be of all contributions are shown. No adjustments of the relative
some differences, which arise because a portion of the 4s energies, Erel, or the relative intensities, Irel, were made. Note
charge density penetrates toward the core of the Mn atom. It is that the doublet predicted for the leading edge is also found
important that the differences between the theoretical results in the higher resolution experimental XPS of Ref. [28].
for Mn0 and for Mn2+ are mirrored in the differences between Considering that the theory takes no solid state effects into
the XPS for the MnO or MnF2 crystals and the Mn atom. account, the agreement with experiment is remarkable. In
Following the logic for NO, discussed above, the 3s XPS particular, there is no need to invoke, as have others [14,170],
should have two peaks. 7S and 5S split by 6  K(3s,3d) [88]. configurations for charge transfer from O to Mn, see Eq. (11),
HF theory predicts a splitting of ∼14 eV and an intensity ratio in order to explain the main features of the spectra, which arise
of 1.4:1 while the XPS measurements for several ionic crystals from a purely atomic model of MnO. The two most intense
give a splitting of ∼6 eV with an intensity ratio of ∼2:1 composite peaks are due almost entirely to ionization of 2p3/2
[4,169]. This ratio is larger, by over 40%, than the statistical and 2p1/2. The intermediate satellite has roughly equal con-
ratio of the two multiplets and is outside the range of deviation tributions from both 2p3/2 and 2p1/2. The origin of the broad
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 297

good first order model for Mn in an ionic crystal, it is

appropriate to ask if there are atomic many-body effects that
have been neglected in the multiplet analysis. Consider the
configuration, 3s23p43d6, formed by the “excitation” of two 2p
electrons, one to fill the 3s shell and one promoted into the
half-filled 3d shell, denoted 3p2-3s3d. For this configuration,
several angular momentum couplings are possible to the low
spin 5S multiplet but not to the high spin 7S multiplet
[3,37,88,171]. Thus, mixing this excitation with the XPS
allowed 3s13p63d5 configuration can lower the energy of
the 5S 3s-hole multiplet but it cannot change the energy of
the high-spin 7S multiplet. This would change the Erel in the
correct direction from the pure multiplet splitting results. In
recent work [47], the acronym FAC, frustrated Auger config-
Fig. 11. Comparison of the theoretical Mn 2p XPS for Mn2+ with experiment
uration, has been applied to this type of excitation because of
for MnO. The dotted line is experiment and the solid lines are from the theory;
the dark solid line is the sum of the intensities for all final states while the light its similarity to an Auger transition except an electron is
solid lines are for individual states with large intensity. The curves are replotted promoted into a bound state not into the continuum. This is a
from the data in Ref. [30]. useful acronym since it can direct us to identify configurations
that may lead to low lying, nearly degenerate, excited states.
leading edge is the multiplet splitting of the coupling of 2p3/2 Indeed, the energies of the angular momentum coupled CSFs
with the 3d5 shell high spin coupled to 6S with J=5/2. This for the 3p2-3s3d excitation are ∼15 eV above the …
coupling leads to the 4 multiplets shown with J=4, 3, 2, and 1 3s13p63d5 CSF, which is comparable to the non-zero
with intensities that almost follow the (2J+1) statistical ratio. off-diagonal Hamiltonian matrix elements [3,37]. Including
This is an excellent example of how unresolved multiplets can this FAC in the CI wavefunctions for the 3s-hole states,
lead to major broadening of XPS features. On the other hand, a originally reported in Ref. [3], leads to a major reduction in the
parallel analysis of the main XPS Mn 2p1/2 peak at Erel∼15 eV, separation of the first two XPS peaks as well as two satellites
see Fig. 11, in terms of the multiplet splitting of the 2p1/2 and that are observed [3]. The comparison of the theory including
3d open shells does not hold. Based on the coupling of j ¼ 1/2 the 3p2-3s3d FAC with the experiment of Ref. [169] is
for the hole in the 2p1/2 shell with j ¼ 5/2 for the 3d5shell, we shown in Fig. 12 taken from Ref. [47]. The dark vertical bars
would expect 2 peaks with J ¼ 3 and 2 with an intensity ratio represent the 1973 theoretical results from Ref. [3]. The
of 7:5. It is clear from Fig. 11 that several individual final improved agreement over the multiplet splittings without the
states contribute to this peak, not just 2; within the peak, there
are large contributions from 5 individual final states. The
difference of this 2p1/2 peak from the leading edge 2p3/2 peak
shows that many-body effects make a much larger contribution
for the 2p1/2 than for the 2p3/2 peak. These effects may well
include recoupling of the 3d5 electrons away from the XPS
allowed 6S5/2 coupling to other, XPS forbidden, multiplets.
The effect of covalency in MnO, see Section 4, on the 2p
XPS was studied with an embedded MnO6 cluster [10]. The
main effect is to reduce the splittings of the different peaks
largely because the partial ligand-character of the Mn 3d
orbitals decreases the exchange integrals. The reduction of the
multiplet splitting improves agreement between the theoretical
simulations and measurement [10]. With the MnO6 cluster
modified to include the steric repulsion of the next shell of
cations, Franck–Condon vibrational broadening, see Section 4,
contributes a large fraction of the ad hoc broadening of each
multiplet used to compare theory with experiment [165]. This
is important since it provides an independent justification for
the choice of FWHM for the calculated multiplets.
In the previous sub-section, we showed that the angular
momentum coupling of the Mn 3s-hole with the open 3d5
valence shell to form high and low spin multiplets did not
correctly describe either the relative energy or the relative Fig. 12. Experiment for the Mn 3s XPS for MnO, MnF2, and Mn gas, from
intensity of the leading peaks in the Mn 3s XPS. Since there is Ref. [169], compared to theory, solid vertical bars from Ref. [3], and dotted
very good reason to expect that the isolated cation should be a vertical line from Ref. [47].
298 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

FACs is evident but the separation of the first two peaks is Although there are exceptions, [174] in most cases where
below experiment by ∼2 eV. Okada and Kotani [13] were able rigorous, non-empirical theory is used, the lowest state does
to reproduce the experimental splitting by reducing a key not have a significant shake contribution [10,21,22,52,53,175].
atomic interaction integral to 75% of the value obtained from a In the remainder of this section, we consider two cases of
rigorous atomic calculation but they did not introduce any new closed shell oxides. One is the 4s XPS of CeO2 [21,22], where
many-body effects beyond the 3p2-3s3d FAC of Ref. [3]. In both a FAC as well as shake excitations contribute to the
2004, Bagus et al. [47] recognized that there is another satellites. Here, we examine the different many-body contribu-
important many-body effect that involves a FAC where a 3p tions to the XPS. The second case is the 4f XPS of UO3 [53],
electron still fills the 3s shell but now a 3d electron is promoted where we could not identify FACs that might contribute to the
to a 4f shell. The 3p3d-3s4f FAC has a differential affect and XPS and only contributions from shake excitations to the
lowers the high spin state by 2 eV more than the excited low satellite structure were considered. In this case, we examine the
spin 3s-hole state bringing the low spin 3s-hole state to a influence of bond distance on the satellite intensity. In both
correct relative energy without the need to empirically adjust cases, we use the sudden approximation (SA) [38,39] to
directly computed integrals. The position of the lowest 5S determine the Irel, where important quantities are the overlap
3s-hole state is show by a dotted vertical line in Fig. 12. This integrals between the orbitals of the initial and final, core-hole
example shows that empirical adjustments to bring theory into states. As discussed in the previous section, the extent of
agreement with experiment may get the “right” answer for the covalency is quite different for the initial and final states.
“wrong” reason. Hence the SA intensities should be strongly affected; this is
analyzed in connection with the XPS satellites for UO3.
6.3. Satellites in closed shell oxides The comparison of theory and experiment for the Ce 4s XPS
of CeO2 [21] is shown in Fig. 13 where the theory includes
Shake excitations have been the traditional explanation for FAC and shake many-body effects. The main peak (A) is
an important class of XPS satellites in atomic and molecular aligned so that it is at the same energy as the leading XPS peak
systems [5,6,19,38]; although this interpretation may also be and it is broadened by a Gaussian with a FWHM chosen to
applied to condensed systems, including oxides [21,22,39] and match experiment; the same FWHM is used to broaden all the
to CO chemisorbed on metal surfaces [133,172,173]. The logic theoretical final states. No adjustments to the relative theore-
of the name shake arises because the photoelectron, as it exits tical energies or intensities have been made. It is important to
the system, has some probability to excite, or shake, another recognize that, in general, several rather than a single final
electron from an occupied orbital into a partially occupied or state contribute to each of the resolved XPS peaks. An
an unoccupied level; from this excitation, the photoelectron important reason for this distribution of intensity is that
loses some of its kinetic energy. There is a probability that the many-body effects combined with spin–orbit and ligand field
electron leaves the system without making this excitation of a splittings distribute the allowed XPS intensity over several
valence electron and there are probabilities for exciting the final states [10,21,22,50]. In particular for CeO2, except for the
valence electrons into different excited levels. A more rigorous main peak, A, in Fig. 13, several individual final states
treatment than this simple view involves mixing the shake contribute to each of the features B, C, and D. The peaks B
excited configurations with the configurations for the other and C are at an Erel that is larger than experiment by ∼3 eV.
many-body effects discussed in the previous sections. If the This error arises from the choice of a single set of orbitals
covalent mixing of metal cation and ligand orbitals is optimized for the main peak, A [21,22]. However, the
neglected, then shake is equivalent to including the configura- wavefunctions allow the composition of the final states to be
tions of Eqs. (11b) and (11c) into the hole-state wavefunctions. analyzed in terms of the many-body effects that contribute to
This is the reason that these excitations have been called these states. The different determinants in the wavefunctions
charge transfer or CT [7,8,12–14,43]. However, the covalent are placed into three groups: Shake(0), Shake(1), and Shake(2)
mixing may be very large [60]. Regardless, it is more
fundamental to describe this many-body process as shake
rather than as CT because the electron is excited from a
molecular bonding orbital and not strictly from a pure ligand
orbital. Since it normally costs energy to make this excitation,
the outgoing electron loses kinetic energy and the shake
satellites are normally to higher BE than the main line,
However, it is possible to have shake-down satellites on the
low BE side of the main XPS line, especially when more than
one electron is involved in the shake excitation [21]. In this
regard, papers based on semi-empirical Anderson model
Hamiltonian treatments of the CT or shake effects often claim
Fig. 13. Experiment, upper curve, and theory, lower curves, for the Ce 4s XPS
that the lowest energy core-hole states are dominated by the of CeO2 [21]. The darker theoretical curve is the envelope of all contributions
shake or CT configuration of Eq. (11b) rather than by the while the contributions of individual states with large Irel are shown below the
configuration without shake of Eqs.(11a) [7,8,12,20,43]. envelope. The letters A–D indicate the four main theoretical features.
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 299

which are distinguished by the occupations of bonding,

dominantly O(2p) orbitals, and anti-bonding, dominantly Ce
4f and Ce 5d orbitals. When there are no shake excitations,
Shake(0), the occupations are N(bonding)=48 and N(anti-
bonding)=0. The significance of 48 is that the bonding orbitals
arise from the 6 2p electrons associated with each of the eight
O anions in the CeO8 cluster model of CeO2, Fig. 3(c), before
the covalent mixing that forms occupied bonding and unoccu-
pied anti-bonding orbitals. The Shake(1) group of determinants
have occupations N(bonding) ¼ 47 and N(anti-bonding) ¼ 1,
and the Shake(2) group have N(bonding) ¼ 46 and N(anti-
bonding)¼ 1. We now determine the contributions of these
determinants to the wavefunctions of the different final states.
This is done by summing, for a given wavefunction, the
Fig. 14. Experimental and theoretical U 4f XPS for UO3.
squares of the coefficients of the determinants in each of the
groups giving Weight(i) for i¼ 0, 1, and 2; where the sum of
the three weights is exactly 1.0 since the wavefunctions are is because the 5f7/2 level is split by the ligand field into three
normalized. (A different notation of charge transfer neglecting closely spaced levels with two doubly and one fourfold
the covalent character of the orbitals was used for these degenerate level. The fourfold and one of the twofold
weights in Ref. [22].) The weights for representative states degenerate levels are grouped into a single peak leading to
in each of groups A–D are given in Table 12. The state in peak the 3:1 intensity ratio. Similar arguments apply to the multiple
A is 95% composed of determinants that have no shake peaks under the main 4f5/2 peak but they are more involved
character, Shake(0). This is the reverse of the semi-empirical [52]. The composition of the main peaks shows that they are
assignments [20,112] where the main peak at Erel=0 is claimed dominated by the determinants where there are no shake
to have almost no Shake(0) weight; instead, Shake(0) weight is excitations, while the shake excitations dominate the states
placed in a highly excited peak that corresponds to peak C in that compose the satellite peaks. This is consistent with the
Fig. 13. It would be interesting to see what choices of assignments for the CeO2 XPS states, see above and Refs.
adjustable parameters would lead the semi-empirical model [21,22], and lends confidence to the assignments made for the
Hamiltonian calculations [112] to give the same composition character of the final states contributing intensity to the
for the final states as the rigorous many-body theory primary and satellite peaks.
(Table 12). Given the large extent of covalency between U and O in
For the U 4f XPS of UO3, [52,53] only shake excitations UO3, see Section 4, this is an excellent system to investigate
were considered since other redistributions of electrons gave the role of covalency on the relative intensity of satellites and
rise to considerably higher energies which did not mix with the main peaks [53]. In order to make this connection, it is
single 4f-hole configurations. Only Shake(1) configurations (or necessary to develop a measure for the loss of intensity to
determinants) were mixed with the one shake(0) configuration. satellites and to use an approach that allows the degree of
The materials model for the theoretical XPS studies was an covalency to be varied independently of other changes in the
embedded UO6 cluster model of δ–UO3. [161] The theoretical U–O interaction. For a simple measure of losses to satellites,
results at the experimental distance of d(U–O) ¼ 2.09 Å are we determine the SA Irel for the Dirac–Hartree–Fock wave-
compared in Fig. 14 to the XPS for Schoepite where a functions for a 4f-hole in the closed shell configuration of the
Tougaard background [176,177] has been removed from the initial state of UO6 neglecting shake many-body effects [53].
experimental data [52]. The theory reproduces the main XPS There are 14 wavefunctions that can be formed with these
features including the satellites at ∼3–4 eV. A careful exam- optimized spinors with an ∼11 eV spin–orbit splitting into a
ination shows that under the leading edge 4f7/2 peak there are group of 6 and 8 states with further small splittings due to the
two closely spaced peaks with an intensity ratio of ∼3:1. This ligand field in UO3; see discussion above. If the orbitals of
these final states were the same as those for the initial state,
Table 12 where there is no core-hole, the SA Irel of each of the final
Weights of the different groups of shake determinants for representative hole
states would be 1. This follows from the fact that the matrix of
states in the 4s XPS of CeO2 from each of the peaks A–D in Fig. 2; the Erel, in
eV, of the selected states are also given; see text. the overlaps of the initial and final state orbitals is, by
construction, a unit matrix [17,38]. In other words, in the
Peak Erel(eV) Weights absence of closed-shell screening of the core-hole, there is no
Shake(0) Shake (1) Shake (2)
intensity lost to shake satellites. When shake configurations are
allowed to mix with the determinants that do not include
A 0.0 0.95 0.01 0.05 shake, the intensity of the main peak will change. However,
B 8.6 0.01 0.85 0.14 since the main peak has only relatively small contributions
C 16.8 0.01 0.83 0.16
from determinants where there is a shake excitation, the
D 24.4 0.06 0.66 0.34
definition of the loss to shake satellites given above is a useful
300 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

Table 13 significance of XPS spectra. There have been several recurring

Projection of U(5f) occupation for initial and 4f-hole state configurations and themes. One has been to provide clear definitions for various
XPS intensity losses from the main 4f5/2 and 4f7/2 peaks for different d(U–O),
terms used in the analysis and interpretation of XPS spectra.
in Å. The intensity losses to satellites, given as a percentage of the total
intensity, is the same for 4f5/2 and 4f7/2 hole configurations. These terms include: initial and final state effects; core-hole
screening; and covalency. Another recurring theme has been to
d(U–O) NP(5f) 4f intensity relate the quantities measured in XPS to the chemical and
loss (%) physical properties of materials. The review has drawn a
Initial state 4f-hole Δ
distinction between one-body and many-body effects; in this
2.06 1.40 2.60 +1.20 29 context, an important objective has been to clarify the concepts
2.09 1.36 2.59 +1.23 35 of many-body interactions and to provide a unifying descrip-
2.12 1.32 2.62 +1.30 45
tion of different many-body mechanisms.
The different mechanisms that can lead to BE shifts were
guide to the intensity lost and it is an easy quantity to calculate. discussed. An important conclusion is that an attempt to relate
In order to separate the change in covalency from other effects, ΔBE to the effective charge on an atom is an oversimplifica-
the distance d(U–O) in the UO6 cluster is changed and the tion and that several other chemical properties, in addition to
distance to the embedding point charges is also changed; in charge, play important roles in determining the signs and
effect, the lattice constant of UO3 is artificially changed so that magnitudes of BE shifts. In terms of materials properties, one
the covalency can be changed. The distance variations are kept needs to take into account the fact that BE shifts depend on
small to insure that the interaction remains dominantly ionic bond distances and on the coordination, or number of nearest
since, if d(U–O) is increased by too large an amount, the neighbors, of the ionized atom. One of the reasons that these
Madelung potential will not be large enough to support a materials properties are important for the ΔBE is because they
dominantly ionic interaction. We expect that the initial state influence the electronic structure. In particular, they modify the
covalency will decrease as d(U–O) is increased since the extent of hybridization of inner valence orbitals, which, in turn,
covalent interaction depends, to first order, on the overlap of modifies the electrostatic potential that a core electron sees.
the metal cation and O anion orbitals [64]. Following from the These different effects are often canceling and the net BE shifts
analysis of XPS satellites of CO chemisorbed on metal may be small, as is often the case for the shifts of the core-level
surfaces discussed in Section 3, it would seem reasonable to BEs of surface atoms compared to bulk atoms.
expect that as the covalent interaction between U and O is Many-body effects may lead to complex XPS spectra with
decreased then the loss to satellites will also decrease. intense satellites, often sufficiently intense that it is not possible to
However, this analogy turns out to be misleading. identify a single main line with satellites. We have considered
In Table 13, the projected numbers of U(5f) electrons, NP(5f), satellites, especially for oxides as representative of ionic crystals.
see Section 4, and the XPS intensity losses from the “main” 4f5/2 For these oxides, we showed that covalent mixing of the metal
and 4f7/2 peaks are given for different values of d(U–O) about the cation and the ligand orbitals may be important and we gave
experimental value of 2.09 Å. The NP are given for the initial and examples that related this covalent character to the XPS of the
final, 4f-hole, configurations. As expected, the initial state cova- oxides. For the oxides, three types, or groups, of many-body
lency, as measured by NP(5f), decreases as d(U–O) becomes larger; effects have been distinguished: (1) multiplets and generalized
see Table 13. However, there is the surprising result that the multiplets; (2) other near degeneracy effects; and (3) shake
intensity loss to XPS satellites actually increases, rather than excitations. Examples were presented to show that the multiplet
decreases, when the initial state covalency decreases. Further, the splittings in the XPS spectra may be large. This is true for
covalent character of the 4f-hole configuration remains almost condensed systems as well as for atoms and molecules. Indeed,
constant. This is not surprising since the equivalent core atom sometimes, an isolated atom will actually be an acceptable model
[41,42] for the core-ionized U is Nd and the greater electron for important aspects of the XPS of a condensed phase system.
affinity of Nd increases the covalent character of the interaction By generalized multiplets, we mean the angular momentum
independent of the relatively small changes in d(U–O). The correct recoupling of the electrons in the open valence shells, which then
correlation to satellite losses is the difference of the covalency of couple to the core-hole. While these angular momentum cou-
the initial and final state configurations since it is this difference plings are XPS forbidden, they can steal intensity by mixing with
that changes the overlap integrals between the initial and hole-state the XPS allowed multiplets. We have also discussed how ligand
sets of orbitals. For the geometric structure of UO3, it is clear that field splittings modify the multiplets of an isolated atom. Many-
this difference increases as d(U–O) becomes larger and, thus, the body effects in groups (2) and (3) above are closely related since
XPS losses to satellites become larger. It would be interesting to they both lead to satellites at relatively low BEs, especially when
examine data for various UOx compounds to determine if there is they involve configurations that are nearly degenerate with the
this correlation between d(U–O) and satellite intensity. lowest core-hole configuration. The distinction between them is
how the rearrangements of electrons, or excitations, occur. The
7. Conclusions shake excitations involve excitations of electrons from levels
whose character is bonding between the metal cation and the
This review has presented a wide range of material on the ligand orbitals to levels whose character is anti-bonding between
interpretation and analysis of the physical and chemical metal and ligand. The effects described as near degeneracy
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 301

normally involve levels localized about the cation that are also “passive” electrons are fixed or frozen as they are for the
normally semi-core levels. In treatments where the covalent initial state before ionization.
character of the orbitals is neglected, the shake excitations are 13. FWHM: full width at half maximum.
referred to as charge transfer from ligand to metal. While the 14. HF: Hartree–Fock. This is used to describe the wavefunc-
shake excitations do involve a net motion of charge from ligand tions and other properties that result from Hartree–Fock
to metal, their description as shake stresses the importance of their calculations.
covalent character. It has been pointed out that there is an 15. Irel: relative intensity. In this context, the intensities of the
interplay between these different effects so that no single one can different XPS peaks arising from the ionization of the
be considered to the exclusion of others. same primary electron.
It is the interplay of all these different effects that complicates 16. KT: Koopmans' theorem. Used to describe energies
the interpretation and the analysis of XPS. However, it is necessary obtained by freezing orbitals of the “passive” electrons
to properly treat this complexity in order to draw the correct when an electron is ionized; closely related to FO
inferences about a material's electronic structure from XPS spectra. properties.
The ability to draw these inferences greatly expands the value of 17. PES: photoemission spectroscopy
XPS beyond the simple identification of elemental composition. 18. SA: sudden approximation. An approximation for the
calculation of the relative intensity of the XPS ionization
Acknowledgments into a particular ionic state; the approximation is exact in
the limit of ionization by very high energy photons.
We acknowledge support by the Geosciences Research 19. SCF: self-consistent field.
Program, Office of Basic Energy Sciences, U.S. DOE. We 20. SCLS: surface core level shift. This is used to describe the
also wish to express our gratitude to Prof. Hajo Freund for his BE shift between photoelectrons arising from the surface
support and encouragement in the preparation of this paper, and from the bulk of a crystal.
which was essential for the successful completion of the work. 21. XPS: X-ray photoelectron spectroscopy.

Appendix A. List of acronyms and abbreviations

References
1. BE: binding energy. The binding energy of an electron
associated with a peak in a photoelectron spectra. [1] T.L. Barr, The Principles and Practice of X-Ray Photoelectron Spectro-
2. CI: configuration interaction. The common use is to scopy, Taylor and Francis, London, 1994.
[2] C.S. Fadley, Surface and Interface Analysis 40 (2008) 1579.
describe many-body wavefunctions that are the mixing [3] P.S. Bagus, A.J. Freeman, F. Sasaki, Physical Review Letters 30 (1973)
of several determinants for different configurations. 850.
3. CSF: configuration state function. A determinant or [4] C.S. Fadley, D.A. Shirley, A.J. Freeman, P.S. Bagus, J.V. Mallow,
combination of determinants that is an eigenfunction of Physical Review Letters 23 (1969) 1397.
[5] K. Siegbahn, C. Nordling, A. Fahlman, R. Nordberg, K. Hamrin,
the angular momentum operators. Normally CSFs are
J. Hedman, G. Johansson, T. Bergmark, S.E. Karlsson, I. Lindgren,
formed for Russell–Saunders, L–S coupling but they can B. Lindberg, ESCA-Atomic, Molecular, and Solid State Structure Studied
also be formed for j–j coupling. by Means of Electron Spectroscopy, Almqvist and Wiksells, Uppsala,
4. CT: charge transfer. Usually refers to the transfer of an Sweden, 1967.
electron from a ligand orbital into an unoccupied or [6] K. Siegbahn, C. Nordling, G. Johansson, J. Hedman, P.F. Hedén,
partially occupied, metal orbital. K. Hamrin, U. Gelius, T. Bergmark, L.O. Werme, R. Manne, and Y.
Baer, ESCA-Applied to Free Molecules, North-Holland, Amsterdam,
5. CSOV: constrained space orbital variation. A theoretical 1969.
procedure for decomposing the contributions to various [7] G. van der Laan, C. Westra, C. Haas, G.A. Sawatzky, Physical Review B
properties by constraining the space of orbitals varied and 23 (1981) 4369.
the space of basis functions in which they are varied. [8] J. Zaanen, C. Westra, G.A. Sawatzky, Physical Review B 33 (1986)
6. ΔSCF: delta self-consistent field. Normally refers to the 8060.
[9] D.J. Tozer, F. De Proft, Journal of Physical Chemistry A 109 (2005)
difference in the properties of two states, an initial and a 8923.
final state, each determined from separate self-consistent [10] P.S. Bagus, E.S. Ilton, Physical Review B 73 (2006) 155110.
field variations. [11] P.S. Bagus, F. Illas, in: P.v.R. Schleyer et al.(Ed.), Encyclopedia of
7. DFT: density functional theory. Computational Chemistry, Vol. 4, Wiley, Chichester, United Kingdom,
8. DHF: Dirac–Hartree–Fock. 1998, p. 2870.
[12] K. Okada, A. Kotani, B.T. Thole, Journalof Electron Spectroscopy and
9. ER: relaxation energy. Related Phenomena 58 (1992) 325.
10. Erel: relative energy. Normally of the BE of a peak relative [13] K. Okada, A. Kotani, Journal of the Physical Sociey of Japan 61 (1992)
to some reference BE taken as zero. 4619.
11. FC: Franck–Condon. Normally used in connection with [14] M. Taguchi, T. Uozumi, A. Kotani, Journal of the Physical Sociey of
the vibrational broadening of peaks in photoemission Japan 66 (1997) 247.
[15] A. Lewera, W.P. Zhou, R. Hunger, W. Jaegermann, A. Wieckowski,
spectra. S. Yockel, P.S. Bagus, Chemical Physical Letters 447 (2007) 39.
12. FO: frozen orbital. Used to describe wavefunctions and [16] P.S. Bagus, G. Pacchioni, C. Sousa, T. Minerva, F. Parmigiani, Chemical
other properties obtained when the orbitals of the Physical Letters 196 (1992) 641.
302 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

[17] P.S. Bagus, M. Schrenk, D.W. Davis, D.A. Shirley, Physical Review A 9 [56] P.S. Bagus, F. Illas, Chemical Physical Letters 224 (1994) 576.
(1974) 1090. [57] E.S. Ilton, W.A. de Jong, P.S. Bagus, Physical Review B 68 (2003) 125106.
[18] P.S. Bagus, R. Broer, C. de Graaf, W.C. Nieuwpoort, Journalof Electron [58] B. Richter, H. Kuhlenbeck, H.J. Freund, P.S. Bagus, Physical Review
Spectroscopy and Related Phenomena 99 (1999) 303. Letters 93 (2004).
[19] R. Manne, T. Åberg, Chemical Physical Letters 7 (1970) 282. [59] P.S. Bagus, A. Wieckowski, H.J. Freund, Chemical Physical Letters 420
[20] D.R. Mullins, S.H. Overbury, D.R. Huntley, Surf. Sci. 409 (1998) 307. (2006) 42.
[21] P.S. Bagus, C.J. Nelin, E.S. Ilton, M. Baron, H. Abbott, E. Primorac, [60] C.J. Nelin, E.S. Ilton, and P.S. Bagus, Journal of Chemical Physics, in
H. Kuhlenbeck, S. Shaikhutdinov, H.J. Freund, Chemical Physical Letters preparation.
487 (2010) 237. [61] CLIPS, is a program system to compute ab initio SCF and correlated
[22] C.J. Nelin, P.S. Bagus, E.S. Ilton, S.A. Chambers, H. Kuhlenbeck, wavefunctions for polyatomic systems. It has been developed based on
H.J. Freund, International Journal of Quantum Chemistry 110 (2010) the publicly available programs in the ALCHEMY package from the IBM
2752. San Jose Research Laboratory by P. S. Bagus, B. Liu, A. D. McLean, and
[23] W.F. Egelhoff, Surface Science Rep. 6 (1987) 253. M. Yoshimine.
[24] P.S. Bagus, G. Pacchioni, F. Parmigiani, Physical Review B 43 (1991) [62] DIRAC, a relativistic ab initio electronic structure program, Release
5172. DIRAC08, 2008, written by L. Visscher, H. J. Aa. Jensen, and T. Saue,
[25] P.S. Bagus, G. Pacchioni, Physical Review B 48 (1993) 15274. with new contributions from R. Bast, S. Dubillard, K. G. Dyall, U.
[26] P.S. Bagus, F. Illas, J. Casanovas, Chemical Physical Letters 272 (1997) Ekström, E. Eliav, T. Fleig, A. S. P. Gomes, T. U. Helgaker, J.
168. Henriksson, M. Iliaš, Ch. R. Jacob, S. Knecht, P. Norman, J. Olsen,
[27] D. Briggs, M.P. Seah (Eds.), Practical Surface Analysis, vol. 1, Wiley, M. Pernpointner, K. Ruud, P. Sałek, and J. Sikkema (see the URL at
1983. http://dirac.chem.sdu.dk..
[28] V. Bayer, R. Podloucky, C. Franchini, F. Allegretti, X. Bo, G. Parteder, [63] H. Umeyama, K. Morokuma, Journal of American Chemical Society 99
M.G. Ramsey, S. Surnev, F.P. Netzer, Physical Review B 76 (2007) (1977) 1316.
165428. [64] I.N. Levine, Quantum Chemistry, Prentice-Hall, Upper Saddle River, NJ, 2000.
[29] J. Stöhr, NEXAFS Spectroscopy (Springer-Verlag, Berlin, 1992). [65] Y.S. Lee, W.C. Ermler, K.S. Pitzer, Journal of Chemical Physics 67
[30] P.S. Bagus, R. Broer, W.A. de Jong, W.C. Nieuwpoort, F. Parmigiani, (1977) 5861.
L. Sangaletti, Physical Review Letters 84 (2000) 2259. [66] P.A. Christiansen, Y.S. Lee, K.S. Pitzer, Journal of Chemical Physics 71
[31] G. Herzberg, Molecular Spectra and Molecular Structure, Van Nostrand, (1979) 4445.
Princeton, 1950. [67] P.S. Bagus, C.W. Bauschlicher Jr., C.J. Nelin, B.C. Laskowski, M. Seel,
[32] U. Gelius, S. Svensson, H. Siegbahn, E. Basilier, A. Faxalv, K. Siegbahn, Journal of Chemical Physics 81 (1984) 3594.
Chemical Physical Letters 28 (1974) 1. [68] L. Visscher, O. Visser, P.J.C. Aerts, H. Merenga, W.C. Nieuwpoort,
[33] M. Iwan, C. Kunz, Physical LettersA 60A (1977) 345. Computer Physics Communications 81 (1994) 120.
[34] P.H. Citrin, P. Eisenberger, D.R. Hamann, Physical Review Letters 33 [69] C.C.J. Roothaan, P.S. Bagus, Methods in Computational Physics,
(1974) 965. Academic Press, New York47.
[35] C.J. Nelin, P.S. Bagus, M.A. Brown, M. Sterrer, H.-J. Freund, Ange- [70] P.S. Bagus, Physical Review 139 (1965) A619.
wandte Chemie International Edition 50 (2011) 10174. [71] P.S. Bagus, H.F. Schaefer, The Journal of Chemical Physics 55 (1971)
[36] J.A. Gubner, Journal of Physics A (Mathematical and General) 27 (1994) 1474.
745. [72] P.S. Bagus, H.F. Schaefer III, Journal of Chemical Physics 56 (1972)
[37] A.J. Freeman, P.S. Bagus, J.V. Mallow, International Journal of 224.
Magnetism 4 (1973) 49. [73] P.S. Bagus, B. Liu, A.D. McLean, M. Yoshimine, in: W.C. Price et al.
[38] T. Aberg, Physical Review 156 (1967) 35. (Ed.), Wave Mschanics: The First Fifty Years, Butterworths, London,
[39] L. Sangaletti, F. Parmigiani, P.S. Bagus, Physical Review B 66 (2002) 1973, p. 99.
115106. [74] B.D. Hermsmeier, C.S. Fadley, B. Sinkovic, M.O. Krause, J. Jimenez-Mier,
[40] P.S. Bagus, F. Illas, G. Pacchioni, F. Parmigiani, Journalof Electron P. Gerard, T.A. Carlson, S.T. Manson, S.K. Bhattacharya, Physical Review
Spectroscopy and Related Phenomena 100 (1999) 215. B 48 (1993) 12425.
[41] W.L. Jolly, D.N. Hendrickson, Journal of American Chemical Society 92 [75] W. von Niessen, J. Schirmer, L.S. Cederbaum, Computer Physics Reports
(1970) 1863. 1 (1984) 57.
[42] W.L. Jolly, in: D.A. Shirley (Ed.), Proceedings of the International [76] V.G. Zakrzewski, J.V. Ortiz, J.A. Nichols, D. Heryadi, D.L. Yeager, J.
Conference on Electron Spectroscopy, North-Holland, Amsterdam, Neth- T. Golab, International Journal of Quantum Chemistry 60 (1996) 29.
erlands, 1972, p. 629. [77] P.O. Löwdin, Physical Review 97 (1955) 1474.
[43] M.A. van Veenendaal, G.A. Sawatzky, Physical Review Lett. 70 (1993) 2459. [78] F. Prosser, S. Hagstrom, Journal of Chemical Physics 48 (1968) 4807.
[44] P.S. Bagus, K. Hermann, C.W. Bauschlicher Jr., Journal of Chemical [79] D.P. Chong, Journal of Chemical Physics 103 (1995) 1842.
Physics 80 (1984) 4378. [80] D.P. Chong, O.V. Gritsenko, E.J. Baerends, Journal of Chemical Physics
[45] P.S. Bagus, K. Hermann, C.W. Bauschlicher Jr., Journal of Chemical 116 (2002) 1760.
Physics 81 (1984) 1966. [81] M. Segala, Y. Takahata, D.P. Chong, Journalof Electron Spectroscopy
[46] P.S. Bagus, F. Illas, Journal of Chemical Physics 96 (1992) 8962. and Related Phenomena 151 (2006) 9.
[47] P.S. Bagus, R. Broer, E.S. Ilton, Chemical Physical Letters 394 (2004) [82] P. Hohenberg, W. Kohn, Physical Review 136 (1964) B864.
150. [83] W. Kohn, L.J. Sham, Physical Review 140 (1965) A1133.
[48] E.S. Ilton, P.S. Bagus, Physical Review B 71 (2005) 195121. [84] E. Runge, E.K.U. Gross, Physical Review Lett. 52 (1984) 997.
[49] P.S. Bagus, E.S. Ilton, Theoritical Chemistry Accounts 118 (2007) 495. [85] J. Schirmer, A. Dreuw, Physical Review A 75 (2007) 022513.
[50] E.S. Ilton, P.S. Bagus, Surface Science 602 (2008) 1114. [86] P.S. Bagus, B.I. Bennett, International Journal of Quantum Chemistry 9
[51] P.S. Bagus, R. Broer, F. Parmigiani, Chemical Physical Letters 421 (1975) 143.
(2006) 148. [87] F. Gilardoni, J. Weber, K. Bellafrouh, C. Daul, H.U. Gudel, Journal of
[52] P. Bagus, C. Nelin, and E.S. Ilton, Physical Review B, in preparation. Chemical Physics 104 (1996) 7624.
[53] P. Bagus, C. Nelin, E.S. Ilton, A. Allahgholi, and I.J. Flege, Chemical [88] J.C. Slater, Quantum Theory of Atomic Structure, McGraw-Hill, New
Physical Letters, in preparation . York, 1960.
[54] K. Hermann, P.S. Bagus, C.J. Nelin, Physical Review B 35 (1987) 9467. [89] D.R. Hartree, W. Hartree, B. Swirles, Philosophical Transactions of the
[55] P.S. Bagus, G. Pacchioni, M.R. Philpott, Journal of Chemical Physics 90 Royal Society of London, A 238 (1939) 229.
(1989) 4287. [90] P.S. Bagus, N. Bessis, C.M. Moser, Physical Review 179 (1969) 39.
 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304 303

[91] P.S. Bagus, A. Hibbert, C. Moser, Journal of Physics B 4 (1971) 1611. [129] C.J. Nelin, P.S. Bagus, J. Behm, C.R. Brundle, Chemical Physical
[92] P.S. Bagus, C. Moser, Physical Review 167 (1968) 13. Letters 105 (1984) 58.
[93] P.S. Bagus, C.M. Moser, Journal of Physics B 2 (1969) 1214. [130] P.S. Bagus, M. Seel, Physical Review B 23 (1981) 2065.
[94] K. Pierloot, E. Tsokos, B.O. Roos, Chemical Physical Letters 214 (1993) [131] P.S. Bagus, K. Hermann, Surface Science 89 (1979) 588.
583. [132] J.C. Fuggle, E. Umbach, D. Menzel, K. Wandelt, C.R. Brundle, Solid
[95] H. Ikeno, I. Tanaka, T. Miyamae, T. Mishima, H. Adachi, K. Ogasawara, State Communications 27 (1978) 65.
Materials Transactions 45 (2004) 1414. [133] H.J. Freund, E.W. Plummer, Physical Review B 23 (1981) 4859.
[96] H. Ikeno, I. Tanaka, Y. Koyama, T. Mizoguchi, K. Ogasawara, Physical [134] H.J. Freund, Angewandte Chemie (International Edition in English) 36
Review B 72 (2005) 75123. (1997) 452.
[97] L. Sangaletti, L.E. Depero, P.S. Bagus, F. Parmigiani, Chemical Physical [135] K. Schönhammer, O. Gunnarsson, Solid State Communications 23
Letters 245 (1995) 463. (1977) 691.
[98] R. Broer, W.C. Nieuwpoort, Theoretica Chimica Acta 73 (1988) 405. [136] W.A. Brown, P. Gardner, M. Perez Jigato, D.A. King, Journal of
[99] R. Broer, A.B. Van Oosten, W.C. Nieuwpoort, Reviews of Solid State Chemical Physics 102 (1995) 7277.
Science 5 (1991) 27. [137] W.A. Brown, P. Gardner, D.A. King, Surface Science 330 (1995) 41.
[100] G. Hohlneicher, H. Pulm, H.J. Freund, Journalof Electron Spectroscopy [138] W.A. Brown, P. Gardner, D.A. King, Journal of Physical Chemistry 99
and Related Phenomena 37 (1985) 209. (1995) 7065.
[101] C.D. Wagner, J.A. Taylor, Journalof Electron Spectroscopy and Related [139] D. Mulugeta, K. Watanabe, D. Menzel, H.J. Freund, Journal of
Phenomena 28 (1982) 211. Chemical Physics 134 (2011) 164702.
[102] C.D. Wagner, A. Joshi, Journalof Electron Spectroscopy and Related [140] J. Billingsley, A.B. Callear, Transactions of the Faraday Society 67
Phenomena 47 (1988) 283. (1971) 589.
[103] C.D. Wagner, Journalof Electron Spectroscopy and Related Phenomena [141] P.S. Bagus, E.S. Ilton, R.L. Martin, H.J.r.A. Jensen, S. Knecht,
10 (1977) 305. Chemical Physical Letters 546 (2012) 58.
[104] P.A. Cox, Molecular Physics 30 (1975) 389. [142] R.S. Mulliken, The Journal of Chemical Physics 23 (1955) 2338.
[105] J.K.L. MacDonald, Physical Review 43 (1933) 830. [143] R.S. Mulliken, The Journal of Chemical Physics 23 (1955) 2343.
[106] C.E. Moore, Atomic Energy Levels, Natl. Bur. Stand. No. 467, U. S. [144] R.S. Mulliken, The Journal of Chemical Physics 23 (1955) 1833.
GPO, Washington, D. C. 1952; see also, URL 〈http://physics.nist.gov/ [145] R.S. Mulliken, The Journal of Chemical Physics 23 (1955) 1841.
cgi-bin/AtData/main_asd〉. [146] H. Chang, J.F. Harrison, T.A. Kaplan, S.D. Mahanti, Physical Review B
[107] G.K. Wertheim, Journalof Electron Spectroscopy and Related Phenomena 49 (1994) 15753.
34 (1984) 309. [147] J. Hernandez-Trujillo, R.F.W. Bader, Journal of Physical Chemistry A
[108] T.L. Barr, C.R. Brundle, Physical Review B 46 (1992) 9199. 104 (2000) 1779.
[109] C. Sousa, T. Minerva, G. Pacchioni, P.S. Bagus, F. Parmigiani, [148] R.F.W. Bader and H. Essen, Journal of Chemical Physics 80, 1943
Journalof Electron Spectroscopy and Related Phenomena 63 (1993) (1984).
189. [149] C.J. Nelin, P.S. Bagus, M.R. Philpott, Journal of Chemical Physics 87
[110] R.S. Mulliken, J. chim. Phys. 46 (1949) 497. (1987) 2170.
[111] O.V. Gritsenko, B. Braida, E.J. Baerends, Journal of Chemical Physics [150] R.W.G. Wyckoff, Crystal Structures, Wiley, New York, 1963.
119 (2003) 1937. [151] P.S. Bagus, K. Hermann, C. Wöll, Journal of Chemical Physics 123
[112] A. Kotani, T. Jo, J.C. Parlebas, Advances in Physics 37 (1988) 37. (2005) 184109.
[113] C.W. Bauschlicher Jr., P.S. Bagus, Journal of Chemical Physics 81 [152] P.S. Bagus, V. Staemmler, C. Woll, Physical Review Letters 89 (2002)
(1985) 5889. 096104.
[114] B. Johansson, N. Martensson, Physical Review B 21 (1980) 4427. [153] S.F. Boys, F. Bernardi, Molecular Physics 19 (1970) 553.
[115] A. Rosengren, B. Johansson, Physical Review B 23 (1981) 3852. [154] T. Van Mourik, A.K. Wilson, K.A. Peterson, D.E. Woon,
[116] O. Plashkevych, T. Privalov, H. Agren, V. Carravetta, K. Ruud, T.H. Dunning, Advances in Quantum Chemistry 31 (1999) 105.
Chemical Physics 260 (2000) 11. [155] P.S. Bagus, C.J. Nelin, W. Muller, M.R. Philpott, H. Seki, Physical
[117] F. Parmigiani, E. Kay, P.S. Bagus, C.J. Nelin, Journalof Electron Review Letters 58 (1987) 559.
Spectroscopy and Related Phenomena 36 (1985) 257. [156] G. Burns, Introduction to Group Theory With Applications, Academic
[118] P.S. Bagus, D. Coolbaugh, S.P. Kowalczyk, G. Pacchioni, Press, New York, 1977.
F. Parmigiani, Journalof Electron Spectroscopy and Related Phenomena [157] P.A. Cox, Transition Metal Oxides: An Inroduction to their Electronic
51 (1990) 69. Structure and Properties, Clarendon Press, Oxford, 1992.
[119] P.S. Bagus, F. Illas, J. Casanovas, J.M. JimenezMateos, Journalof [158] P.S. Bagus, F. Illas, Physical Review B 42 (1990) 10852.
Electron Spectroscopy and Related Phenomena 83 (1997) 151. [159] E.R. Davidson, Journal of Chemical Physics 46 (1967) 3320.
[120] P.S. Bagus, A. Wieckowski, H. Freund, Computational and Theoretical [160] J. Knut Faegri, T. Saue, The Journal of Chemical Physics 115 (2001)
Chemistry 987 (2012) 22. 2456.
[121] G.K. Wertheim, S.B. DiCenzo, D.N.E. Buchanan, Physical Review B [161] M.T. Weller, P.G. Dickens, D.J. Penny, Polyhedron 7 (1988) 243.
33 (1986) 5384. [162] C. Sousa, F. Illas, G. Pacchioni, Journal of Chemical Physics 99 (1993)
[122] P.S. Bagus, C.J. Nelin, E. Kay, F. Parmigiani, Journalof Electron 6818.
Spectroscopy and Related Phenomena 43 (1987) C13. [163] A. Kotani, H. Mizuta, T. Jo, J.C. Parlebas, Solid State Communications
[123] M. Klimenkov, S. Nepijko, H. Kuhlenbeck, M. Baeumer, R. Schloegl, 53 (1985) 805.
H.J. Freund, Surface Science 391 (1997) 27. [164] J.P. Desclaux, Atomic Data and Nuclear Data Tables 12 (1974) 311.
[124] S.A. Nepijko, M. Klimenkov, M. Adelt, H. Kuhlenbeck, R. Schlogl, [165] C.J. Nelin, E.S. Ilton, R. Egdell, H.J. Freund, and P. Bagus, Physical
H.J. Freund, Langmuir 15 (1999) 5309. Review B, in preparation.
[125] P.S. Bagus, C. Woll, A. Wieckowski, Surface Science 603 (2009) [166] R.P. Gupta, S.K. Sen, Physical Review B 12 (1975) 15.
273. [167] R.P. Gupta, S.K. Sen, Physical Review B 10 (1974) 71.
[126] G. Pacchioni, C. Sousa, F. Illas, F. Parmigiani, P.S. Bagus, Physical [168] H.A. Bethe, E.W. Salpeter, Quantum Mechanics of One- and
Review B 48 (1993) 11573. Two-Electron Atoms, Academic Press, 1957.
[127] C.R. Brundle, P.S. Bagus, D. Menzel, K. Hermann, Physical Review B [169] B. Hermsmeier, C.S. Fadley, M.O. Krause, J. Jimenez-Mier, P. Gerard,
24 (1981) 7041. S.T. Manson, Physical Review Letters 61 (1988) 2592.
[128] K. Hermann, P.S. Bagus, C.R. Brundle, D. Menzel, Physical Review B [170] A. Kotani, T. Yamazaki, Progress of Theoretical Physics Supplement
24 (1981) 7025. (1992) 117.
304 P.S. Bagus et al. / Surface Science Reports 68 (2013) 273–304

[171] E.-K. Viinikka, Y. Öhrn, Physical Review B 11 (1975) 4168. [174] P.S. Bagus, H.J. Freund, T. Minerva, G. Pacchioni, F. Parmigiani,
[172] H.J. Freund, E.W. Plummer, W.R. Salaneck, R.W. Bigelow, Journal of Chemical Physical Letters 251 (1996) 90.
Chemical Physics 75 (1981) 4275. [175] P. Bagus and E.S. Ilton, Physical Review Letters, in preparation.
[173] H.J. Freund, W. Eberhardt, D. Heskett, E.W. Plummer, Physical Review [176] M.P. Seah, I.S. Gilmore, S.J. Spencer, Surface Science 461 (2000) 1.
Lett. 50 (1983) 768. [177] M.P. Seah, Surface Science 420 (1999) 285.