X-Ray Photoelectron Spectroscopy (XPS):

Principles and Application for the Analysis 10

of Photoactive Materials

Pardis Simon , Víctor G. Baldovino-Medrano , and Robert Wojcieszak

Contents
10.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 249 results often poses problems because of the complexity
of the XPS analysis and existence of various phenomena
10.2 X-ray Photoelectron Spectroscopy (XPS) . . . . . . . . . . . . . 250
that have to be taken into account. Thus the aim of this
10.3 Surface Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 250 work is to provide a general introduction on fundamental
10.3.1 Surface Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
principles of XPS analysis and to help researchers in
10.3.2 Initial State Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253
10.3.3 Final Sate Effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 254 developing a good practice in XPS results interpretation
10.3.4 Background Selection for Quantification . . . . . . . . . . . . . . . . 255 especially for complex XPS spectra. The overview of
10.3.5 Limit of the Detection of the Technique . . . . . . . . . . . . . . . . . 255 most important aspects and theoretical models is pre-
10.3.6 Influence of Adsorbed Contaminants on the Surface . . . . 256
sented. Examples of XPS application for analysis of
10.3.7 Particle Size Determination from XPS . . . . . . . . . . . . . . . . . . . 256
10.3.8 Photon Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258 photoactive materials are discussed.
10.4 XPS Analysis of Photoactive Materials . . . . . . . . . . . . . . . . 258
10.4.1 Titanium Oxide (TiO2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 259 Keywords
10.4.2 Tantalum (V) Oxide (Ta2O5) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 262 XPS · Photoactive · TiO2 · Ta2O5 · Peak fitting · Particle
10.4.3 Cadmium Sulfide (CdS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 264
size · Charge effects
10.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
10.1 Introduction

Abstract X-ray photoelectron spectroscopy (XPS) is a powerful tech-

nique largely applied to materials such as minerals, hetero-
X-ray photoelectron spectroscopy (XPS) plays a crucial
geneous catalysts, metals and oxides, polymers and
role in the surface analysis of different organic and inor-
biopolymers, biomaterials, pharmaceutical molecules, and
ganic materials. The information that could be obtained
microbial cells. The principles of this spectroscopy are
through XPS analysis (chemical state, coordination, parti-
based on photoelectric effect [1, 2]. Nowadays, the XPS
cle size) is very important from the material science point
becomes one of the most applied techniques in materials
of view. However, the exact interpretation of the XPS
science because it gives an attractive compromise between
qualitative, quantitative, and chemical information. XPS can
be applied to all types of systems but by paying particular
P. Simon · R. Wojcieszak (*) attention to volatile samples as the analysis is carried out in
Univ. Lille, CNRS, Centrale Lille, Univ. Artois, UMR 8181 - UCCS -
Unité de Catalyse et Chimie du Solide, Lille, France high vacuum chamber.
e-mail: [email protected]; [email protected] XPS is already largely applied to photoactive materials.
V. G. Baldovino-Medrano However, considering the properties of different materials
Centro de Investigaciones en Catálisis, CICAT-UIS (@CICATUIS), several aspects have to be kept in mind such as chemical
Laboratorio Central de Ciencia de Superficies, SurfLab-UIS, Parque composition of the analyzed samples in terms of elements,
Tecnológico de Guatiguará, Universidad Industrial de Santander, Km texture of the material (especially particle size, specific sur-
2 vía El Refugio, Universidad Industrial de Santander, Piedecuesta,
Santander, Colombia face area, and porosity), homogeneity of the sample (chem-
e-mail: [email protected] ical composition of the surface can differ from that of the

© Springer Nature Switzerland AG 2022 249

D. W. Bahnemann, A. O. T. Patrocinio (eds.), Springer Handbook of Inorganic Photochemistry, Springer Handbooks,
https://doi.org/10.1007/978-3-030-63713-2_10
250 P. Simon et al.

bulk), and physicochemical properties of the materials (espe-

Sample surface Spectrometer
cially the adsorption of the small molecules on the surface

Kinetic energy
and hydrophobicity).

Kinetic energy
The aim of this work is to provide an introduction on
general principles of XPS analysis and to help researchers
E’k Ek
in developing a good practice in XPS results interpretation
especially for complex XPS spectra (example of tantalum 0 (ΦS – Φ)
ΦS hv Ev
(V) oxide). The overview of most important aspects and Φ
theoretical models is presented. The work is divided in 0 EF 0
M, etc

Binding energy
three general parts. The first one concerns the basic informa-
tion of XPS principles, the second one presents the most L2,3
important aspects of the XPS analysis, and the third part
L1
gives practical examples of the XPS analysis for some photo-
active materials such as TiO2, Ta2O5, and CdS. The Eb
deconvolution of experimental spectra for these materials is K
critically discussed and practical examples of particle size
determination using XPS are given. Fig. 10.1 Schematic energy level diagram for a metallic sample in
electrical contact with the electron spectrometer. ΦS is the sample
work function, E’k is the initial kinetic energy at the surface of the
sample and Ev is the vacuum level
10.2 X-ray Photoelectron Spectroscopy (XPS)
emission of a secondary electron via an Auger process
The principles of XPS are based on the photoelectric effect. which yields to a doubly charge ion. Both decay mecha-
They have been discussed in detail by Briggs [1] and Fadlay nisms are schematized in Fig. 10.2. Although the two decay
[2]. A solid emits photoelectrons from its surface when it is mechanisms are in competition, the probability of relaxa-
illuminated with a photon of energy hν. The kinetic energy tion via an Auger process is favored for core levels below
of the emitted photoelectrons is measured by a spectrome- about 2 keV.
ter, which provides a spectrum consisting in a series of In the Auger process, an electron from a higher electronic
discrete bands. The kinetic energy of these photopeaks level fills the core level vacancy while transferring its energy
reflects the electronic structure of the atoms present at the to a secondary electron. The latter is emitted as an Auger
solid surface. electron, leaving the atom with two holes. Specific Auger
Figure 10.1 presents a schematic diagram of the electronic transitions are denoted using the X-ray nomenclature of the
transitions involved during an XPS experiment. After the three electronic levels involved. For example, for an initial
photoionization process, the binding energy Eb of the elec- hole on a K level, filled by an electron from the L1 shell and
tron belonging to a given core level and involved in the resulting in the emission of an electron from the L3 shell, the
transition can be estimated from the measured kinetic energy Auger transition is named KL1L2,3. The energy of the
Ek of the emitted photoelectron through the Eq. (10.1): corresponding Auger electron is approximately given by
Eq. (10.2):
Eb ¼ hν  Ek  Φ ð10:1Þ

where hν is the energy of the incoming photon and Φ is the EKL1 L2,3 ¼ EK  EL1  EL2,3 ð10:2Þ
work function of the spectrometer. In the case of metallic
where Ei is the binding energy of the atomic levels i and EL2,3
samples in electrical equilibrium with the spectrometer, Eb is
relative to the sample Fermi EF level that is aligned to the represents the binding energy of the L2,3 atomic level modi-
spectrometer Fermi level. fied due to the presence of a hole in level L1 [1].
After photoionization, the atom remains in an excited
state with the creation of a single core hole; this can lead to
relaxation effects, which will be discussed later on. Core- 10.3 Surface Analysis
hole lifetime is governed by the processes that follow pho-
toemission, by which the excess energy of the ion state The knowledge of the exact chemical composition of the
decays. Two decay mechanisms are possible: either the studied material is necessary in developing active mate-
emission of a photon, that is, X-ray Fluorescence, or the rials. XPS permits to obtain important information such as
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 251

Ev Ev Ev

EF EF EF
M, etc M, etc M, etc

L2,3 L*2,3 L*2,3

hv
(EK–EL1)
L1 L1 L1

K K K 10
Photoemission Fluorescence Auger electron emission

Fig. 10.2 Schematic energy diagram of the decay mechanism after photon in a fluorescence process, or can be given up to another electron.
photoemission process. A photon of energy hν has created a hole in The latter, in this example in the L2,3 level, is then ejected in an Auger
the core level K. The core-hole is then filled by an electron from the L1 process
level, releasing an energy EK  EL1 . This energy can appear either as a


chemical composition, oxidation state, coordination num- z
I ¼ I 0 : exp ð10:3Þ
ber, and particle size. However, to do this the knowledge λ cos θ
of the various phenomenon that rule XPS analysis is man-
datory. In this chapter we present the most important where I0 is the intensity of the photoelectrons emitted from a
aspects of the XPS analysis and present the theoretical solid with infinite thickness and θ is the angle of emission of
models that permit to determine the layer thickness or the photoelectrons from the surface with respect to the sur-
average particle size. face normal. Thereby, most of the signal comes from within a
few atomic layers of the surface. However, the relationship
presented in Eq. 10.3 gives a rough estimation of the surface
sensitivity of XPS. Better estimates are obtained by taking
10.3.1 Surface Sensitivity into account the elastic-scattering properties of the
samples [5].
In XPS, photo-ionized electrons as well as Auger elec- An X-ray photoelectron spectrum contains features
trons are analyzed by the spectrometer; the resulting resulting from elastic processes and secondary electrons fea-
spectra thus contain both of corresponding peaks. XPS tures stemming from Auger and inelastic processes. As an
surface sensitivity originates from the photoelectrons example, Fig. 10.3 shows the XPS spectrum of a TiO2 crys-
kinetic energy, which lies between ca. 10 and 1000 eV. talline nanopowder obtained with a monochromatized Al Kα
With this range of kinetic energies, the electron’s inelas- X-ray source. The spectrum is dominated by a strong signal
tic mean free path (IMFP, λ) – the average distance that from photo peaks originating from titanium and oxygen
an electron with a given energy travels between succes- orbitals. Al Kα radiation (1486.6 eV) is only energetic
sive inelastic collisions – is only a few atomic layers in enough to probe the core levels of titanium up to Ti 2s orbital.
condense matter. It is established that λ obeys a universal When the photoemission process is elastic, this gives rise to a
curve with a minimum for kinetic energies around series of discrete photoelectron peaks with kinetic energies
100 eV [3]. The value of λ can be calculated from a determined by Eq. (10.1). The photoemission process is
predicted formula (the TTP-2M formula) derived by inelastic when the photoelectron undergoes one or several
Tanuma et al. [4]. Considering λ, the intensity of the inelastic scattering events in the solid before reaching its
emitted photoelectrons I decreases with the sample surface and being detected by the analyzer. These inelastic
depth z according to Eq. (10.3): scattering events correspond to an energy loss in the solid.
252 P. Simon et al.

4.0×105 O 1s TiO2

3.5×105 Ti 3p
Ti 3s
3.0×105 O 2s
Intensity (counts.s–1)

2.5×105 Ti LMM Ti 2p 80 60 40 20 0
O KLL
2.0×105 Ti 2s

1.5×105

1.0×105 C 1s

5.0×104

0.0
1200 1000 800 600 400 200 0
Binding energy (eV)

Fig. 10.3 X-ray photoelectron spectrum of TiO2 excited by monochromatized Al Kα. The inset shows a magnification of the low binding energy
region. At low kinetic energy (high binding energy) the background becomes dominated by secondary electrons

This results in steps in the low kinetic energy side of the energy difference ΔEj between these two states depends on
photo-peaks of the spectra. Inelastic scattering also produces the atomic subshell and the atomic number, and ranges
secondary electrons. At low kinetic energy, the intensity of between many eV. For a given subshell, ΔEj is expected to
the secondary electrons will increase and they will dominate increase with Z and increase when l decreases. As an exam-
the spectrum background (see Fig. 10.3 as an example). ple, Fig. 10.4 shows the Ti 2p core level for a TiO2
When an outgoing photoelectron passes through a solid, it compound.
can excite collective oscillations in the electrons belonging to The energy splitting between Ti 2p3/2 and Ti 2p1/2 is
the conduction band. As these oscillations have characteris- 5.7 eV and the FWHM is different for the two orbitals. On
tics frequencies, the electron will suffer a discrete energy loss this figure, two additional features are observed about 13 eV
called a plasmon loss (or several losses of energy and 19 eV above the main peak and correspond to satellites
corresponding to a multiple of the characteristic plasmon peak due to a charge transfer mechanism [6].
frequency). If the energy loss takes place in the solid, it is The peak width is defined by the Full Width at Half Max-
then called a bulk plasmon. If the energy loss takes place at imum (FWHM). It is a convolution of three parameters: the
the surface, it is then called a surface plasmon and will have a width of the photon source; the analyzer resolution; and the
lower energy than for the solid, as the plasmon frequency is natural width of the core level. The line width of a core level
no longer the same at the surface. varies with the inverse of the core-hole lifetime, in the
The core level peaks for each element have different remaining ion state after photoionization process. Core hole
intensities and widths and the core levels other than the lifetimes are governed by the decay mechanism that follows
s core levels are doublets. Such a feature is clearly illustrated photoionization process, that is, X-ray fluorescence or emis-
in Fig. 10.3. sion of a secondary electron by Auger process or by Coster-
Doublet peaks arise from the spin-orbit ( j-j) coupling Kronig process [7]. A Coster-Kronig process is a special type
when the core-level quantum number l is >0. The reader of Auger transition in which one of the final state holes is in the
may recall that each state is characterized by the quantum same shell, for example, same principal quantum number, as
number j and that j ¼ l + s. Thus, the spin-orbit doublets are the initial-state hole [8]. As these transitions are extremely fast,
designated by their nlj quantum number. The relative inten- the core levels influenced by Coster-Kronig processes suffer
sities of the nlj doublet peaks are determined by the electron from lifetime broadening. For example, for Ti the 2p1/2 line is
population, provided by the orbital degeneracies (2j + 1). The wider than the 2p3/2 line, due to a broadening of the Ti 2p1/2
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 253

In XPS, surface composition can be determined from the

TiO2 measurements of photoelectron peaks intensities. For the
4×104 ∆Ej simple case of homogeneous samples, the intensity IA from
Ti 2p3/2
an element A in a solid is proportional to the molar fraction
Intensity (counts.s–1)

3×104 XA within the analysis depth:

I A =I 1
2×104 XA ¼ P A1 ð10:4Þ
Satellites I i =I i
Ti 2p1/2 i

1×104
where I 1A is the intensity from pure A.
The terms contributing to IA in XPS have been extensively
0
discussed by Powell and Jablonski [21]. Key to XPS quanti-
fication is the estimation of adequate relative sensitivity fac-
490 480 470 460 450
tors (RSF); defined by Wagner et al. in [22] as RSF ¼
Binding energy (eV) σφyATλ (where σ [cm2/atom] is the photoelectron cross sec-
tion for a given electronic transition, φ is the angular effi-
10
Fig. 10.4 Ti 2p core level XPS spectrum of a TiO2 compound. Spec- ciency factor of the instrument, y is the efficiency of
trum recorded at Surface Analysis Platform of the University of Lille, production of photoelectrons of normal energy in the photo-
France electric process, A is the area of the sample from which the
photoelectrons are detected, T is the efficiency of the detec-
tion of the photoelectrons emitted by the sample, and λ is the
line by L2L3M45 Coster–Kronig processes (Fig. 10.4). This mean free path of the photoelectrons in the sample, since this
broadening of Ti 2p1/2 line causes difficulty in accurately peak parameter is characteristic of the spectrophotometer).
fitting Ti 2p region when multiple chemical states are present.
The relative intensity of the core level peaks is governed by
the atomic photoionization cross-section, σ, which depends on
the exciting energy. Calculated values of σ have been tabulated 10.3.2 Initial State Effects
for all of the atomic subshell [9, 10]. Readers should keep in
mind that the transmission characteristic of the electron ana- Beyond its ability to quantitatively analyze the elemental and
lyzer, which is a function of the kinetic energy of the photo- chemical composition with high surface sensitivity, XPS can
electrons, affects the relative intensity. provide information about chemical environments. As a mat-
As could be seen on Fig. 10.3, in addition to the Ti and O ter of fact, the technique was originally named electron
lines, there is also a strong contribution from carbon 1s spectroscopy for chemical analysis (ESCA) by Kai Siegbahn
orbital. Unless samples are prepared for analysis under con- [23]. The atomic core level binding energies depend to some
trolled atmospheres, this contribution is always present at the extent on the chemical state of the atoms. A change in peak
surface. It corresponds to adventitious contamination. The energy, also called chemical shift, arises from
component of this peak at around 284.8–285.0 eV is ascribed non-equivalency in atoms due to differences in oxidation
to C-[C,H] (hydrocarbon) species and routinely used for state, molecular environment or lattice sites.
static charge referencing up to the point that it has been As an example, Fig. 10.5 presents the 4f core level spectra
included as an ASTM standard [11]. However, static charge for a Ta in different oxidation states. For an increasing oxi-
referencing with this peak must be adopted with care (or often dation state, the Ta core level electrons feel a stronger bind-
avoided) for heterogeneous mixtures composed of substances ing, and consequently, the XPS spectra shift at higher binding
with different electric conductivities [12–14]. energies. In this case, the chemical shift reaches around
Recalling that the photoelectric process can be schema- 5.5 eV from Ta0 to Ta5+.
tized as: hν + Z ! Z+ + e, static charging during XPS It can be mentioned here that the NIST XPS Database [25]
analyses arises when the loss of electrons from the atoms of gives easy access to the energies of many line positions,
a nonconductive sample is not compensated by the electrical chemical shifts, doublet splitting, and energy separations of
connection of the sample with the instrument. As a conse- photoelectron and Auger-electron lines.
quence, a positive charge builds up at the surface of the The chemical shift can be interpreted in the simple model
sample causing a decrease in the photoelectrons kinetic of charge potential. In this model, the interaction of a given
energy. Careful analysis of static charging has cast light core electron with all other electrons and nuclei in a molecule
upon the mechanisms behind the phenomenon and further or solid is divided into an intra-atomic term and an extra-
exploitation for identifying chemical species [15–20]. atomic term. The binding energy Ei of an atom i is then
254 P. Simon et al.

Ta5+ CuO
Ta4+ Ta3+ Cu 2p3/2
Ta4+ Ta1+ 3×104

Intensity (counts.s–1)
Ta5+ Ta2+ Ta3+
Cu 2p1/2
Ta0
Ta2+
Ta1+ Satellites
2×104
Ta0

32 31 30 29 28 27 26 25 24 23 22 21 20 1×104
970 960 950 940 930
Binding energy (eV) Binding energy (eV)

Fig. 10.5 XPS Ta 4f peak taken from a depth of approximately 20 nm

Fig. 10.6 Cu 2p core level spectrum from CuO, showing the shake-up
into a 30 nm Ta2O5foil, fitted with Ta5+, Ta4+, Ta3+, Ta2+, Ta1+, and Ta0
satellite structures. Spectrum recorded at Surface Analysis Platform of
components. (Adapted from reference [24])
the University of Lille, France

described by its electrical charge qi plus the potential field from different spin distributions of the electrons in the band
due to the surrounding atoms j at a distance rij: structure. After photoionization, if the spin of the resulting
unpaired electron is parallel to that of the unpaired electron in
0
Xq
Ei ¼ Ei þ kqi þ i
ð10:5Þ the valence levels, then exchange interaction can occur
r ij resulting in a doublet peak with a peak in a lower energy
i6¼j
than for the case of antiparallel spin. In the case of non-s
In this equation, E0i is a binding energy reference for an levels, the orbital-angular momentum coupling gives rise to
atom i and k is a physical constant that corresponds, for a more complex multiplet splitting. Multiplet splitting effects
given core level, to the coulomb repulsion between the core are, for example observed for Mn and Cr 3s levels, Co, Fe
orbital and the valence orbitals. and Ni 2p levels, and the 4s levels of the rare earth elements.
When the atom i is placed in a different environment, a After the photoionization process, the presence of a core-hole
decrease in the valence electron charge (an increase on positive gives rise to a relaxation of the valence electrons. This can
charge), Δqi, on atom i increases its binding energy. This model result in the excitation of one of them to a higher unfilled
explains qualitatively and semi-quantitatively some of the basic level. Such a transition is called a “shake-up” and the energy
features of chemical shifts [2]. However, the last term of required for this transition is no longer available to the pri-
Eq. (10.5) should also be considered, as its variation will be mary photoelectron. As a consequence, a discrete “satellite”
opposite in sign to Δqi. Furthermore, Eq. (10.5) considers only structure will appear at the low kinetic energy side of the
the initial state effects. Another term should thus be considered photoelectron main peak. These satellite structures arise for
due to the relaxation in the final state. Indeed, after the photoion- certain transition metal and rare earth compounds, which
ization process, the atom will be left with a core-hole, which will have unpaired electrons in 3d or 4f shells. It is well illustrated
polarize the inner and valence electrons, as well as the surround- by the Cu 2p spectra of CuO compounds, where a strong
ing atoms [26–29]. Generally speaking, researchers can compare satellite structure is found around 8.5 eV higher than the main
their spectra to those collected by the NIST XPS Database [25] in peak (Fig. 10.6). However, as Cu2O and metallic copper do
order to identify chemical species. Crist has reviewed XPS not present unpaired electrons in the 3d shell, these shake up
databanks that can be found in open literature [30]. satellites structures are absent from the spectra. In transition
metal compounds, such features can also be very strong due
ligand-metal charge transfer [31, 32]. In aromatic organic
systems, shake-up satellite can also appear due to π ! π*
10.3.3 Final Sate Effects transition [33].
In the case of solid metals, the presence of an unfilled level
For a system with unpaired electrons in the valence levels, distribution above the Fermi level allows shake-up type
multiplet splitting of core level peaks may occur. This arises events. Thus, instead of having discrete satellites at higher
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 255

Ta0(7/2)
Ta0(5/2)

Shirley

31 30 29 28 27 26 25 24 23 22 21 20
Binding energy (eV)
10
Tougaard Linear
–U3–*
Fig. 10.7 XPS Ta 4f peaks with the metallic Ta0 4f7/2 and 4f5/2 peaks
fitted. Both peaks in the doublet show the high binding energy tail Fig. 10.8 Possible backgrounds that can be used for quantification
associated with shake-up events for the electrons close to the Fermi from a high-resolution spectra for the Ti 2p core level of a sample of
level. (Adapted from reference [24]) TiO2 (Degussa P 25) recorded with a power of the Al Kα source of
100 W and an Epass of 15 eV. No charge compensation. In all cases:
energy step ¼ 0.050 eV; spot area ~2  2 mm; analysis chamber
pressure ~1  107 Pa
binding energy, the metal core level peak will present an
asymmetry in its high binding energy side [34]. In this case,
asymmetrical line shapes have to be used in curve fitting of contribution. Therefore, an LA (1,1.2,180) line shape was
photoelectron spectra [35, 36]. An example of this type of selected for peak fitting with the CasaXPS software [38].
fitting is presented in Fig. 10.7 for metallic tantalum, where
the spectrum can be curved fitted with an asymmetrical
10.3.5 Limit of the Detection of the Technique
Gaussian-Lorentzian sum function.
In a similar way, a valence electron can be ejected to an
It is agreed that the limit of detection in XPS corresponds to
unbound continuum state in a “shake-off” process, leaving a
the concentration at which the given peak is three times as
doubly ionized state. The shake-off peaks are rarely seen as
high as the noise of the background [39–41]. For recent
discrete structures; they generally appear as broad shoulders
spectrometers equipped with a monochromatized X-ray
or contribute in the inelastic background. Indeed, the energy
source this limit is about 2 mol % for light elements such as
separation with the photoelectron main peak is higher than in
Li or B and around 0.02 mol % for heavier elements such as
the case of shake-up transitions.
lead or gold [1, 42]. Generally, the detection limit for the
analysis of oxygen or carbon is at the order of 0.1 mole%
and this limit is quite the same for S, P, and K. In general, the
10.3.4 Background Selection for Quantification limit of detection in XPS is not fixed since it depends on the
required precision. When reporting relative concentrations by
For quantification, the background must be selected according XPS, it is advisable to report very low values of concentration
to the sample, the common backgrounds used in the literature of metals as “below detection limit” instead of a zero value
are the Linear, Shirley, and Tougaard backgrounds. Figure 10.8 [42]. For the XPS peaks with a complex shape (multiplet
shows the use of these backgrounds for fitting the peaks of a Ti splitting or doublets), the determination of the detection limit
2p core level. In this case, the U3 Tougaard [37] background from a formal signal to noise ratio is difficult. To solve this
was preferred for peak fitting from considering the type of issue, it is best to consider a peak that is still clearly visible in
spectrometer employed in the experiment. the XPS spectra and then measure its area and evaluate the
The line shape must be selected according to the spectrom- factor by which the intensity should be decreased to lose the
eter used for the experiments. In the case of the example peak in the noise. Under such circumstances, the detection
illustrated above, the peaks recorded by the spectrometer limit is the concentration deduced from the peak considered,
were fitted better using a line shape with a strong Lorentzian divided by the evaluated factor [42].
256 P. Simon et al.

10.3.6 Influence of Adsorbed Contaminants authors the relative error for peak separations of up to 900 eV
on the Surface will be only 2% in case of the overlayer organic thickness
lower than 0.2 nm.
The contamination of the surface of high surface-energy However, for an overlayer contamination of about 1 nm,
solids (such as metals and oxides) by adsorbed organic com- a 300 eV peak separation would already give rise to a 2%
pounds [42–44] can occur very rapidly under ultrahigh vac- relative composition error. Even worst, at peak separation of
uum (UHV) conditions due, in most cases, to the high about 700 eV the relative error will reach 8%. The authors
proportion of organic contaminants in the residual gas. We presented a practical example using zinc sulfide for the
can identify three main sources of organic contamination in study. For this kind of compounds, the use of the Zn 2p3/2
UHV. First, the most important source is the pumping system and the S 2p core lines at 1022 and 163 eV, respectively, is
that can release hydrocarbons and constituents of lubricant typically admitted for quantification. As the separation
oils [45]. Second, surface contamination can occur via resid- between these two peaks is ca. 859 eV, the presence of a
ual gases that can easily adsorb on the surface [46–48]. And, 0.5 nm adlayer of organic contamination gives rise to rela-
third, contamination is due to the history of the spectrometer tive errors of approximately 5% for each element. This
utilization. Indeed, in some cases the molecules originated implies that if the results remain uncorrected, the chemical
from the analyzed samples before can desorb from the walls analysis would give an apparent stoichiometry of
of the vacuum chamber and then re-adsorb on the new sample Zn47.5S52.5 quite far from the stoichiometric ZnS value.
[43]. The contamination is a very important issue especially The same was observed for the majority of the compounds
during the analysis of low concentrated samples because studied in that work. In all cases the oxygen concentration
the surface contamination adds new elements (O, C, N) to was underestimated when compared to the expected for the
the composition of the surface. This may strongly affect the stoichiometric values [49].
apparent concentration ratio of elements on the surface. It is well known that obtaining exact concentration ratios
Moreover, the overlayer of the analyzed surface with new of given elements on the surface requires particular
contaminants may provoke changes in the thickness of the approaches especially if the photoelectrons differ strongly
analyzed surface and standards, which may in turn modify in kinetic energies (different IMFP in the contamination
the determination of sensitivity factors. This is especially overlayer). One possibility to minimize the errors in the
well illustrated in the work of Smith [49]. As could be seen analyzed samples is to assume that the contamination over-
from Fig. 10.9 the quantification errors increase with increas- layer is the same for the standards used for determination of
ing peak separation and overlayer thickness. The errors will the sensitivity factors. Another possibility is to follow the
increase also with increasing emission angle. As stated by the first-principles method considering a reduced thickness deter-
mined from the ratio of the intensities of the C 1s peak and of
the Auger peak of carbon (KL23L23) [50]. However, the
method based on deducing the overlayer thickness from the
Emission angle = 45° mole fraction of carbon is quite approximate [49]. In the cited
1.0
Relative % error in composition

14 publication, its application reduced the average error from

0.9 1000
12
0.8 about 8.2% to 5.6% but the relative error was not strongly
900
10 0.7 reduced. Sometimes the relative errors were increased for
Carbon overlayer 0.6
8 thickness (nm) 700 some elements even for the same compound. Another way
0.5 Line is to deduce a corrected intensity ratio computed as the slope
6 0.4 separation
0.3 500 (eV) of the plot of given element intensity as a function of stan-
4 0.2 dards intensity and samples on which the contamination
300
2 0.1 overlayer thickness is controlled [51].
100
0
0 5 10 15 20 25 30 35 40 45 50
Apparent carbon concentration (atom%)
10.3.7 Particle Size Determination from XPS

Fig. 10.9 The relative error in the calculated composition of a stoichio- XPS measurements are very useful for studying metal and
metric alloy AB as a function of the apparent surface carbon concentra- oxide particles of 1–5 nm, especially when coupled with
tion due to the presence of the hydrocarbon contamination overlayer, other chemical or physical methods. Taking advantage of
after normalization to exclude carbon for various values of the separa-
tion of the photoelectron lines of elements A and B, for hydrocarbon,
the improvements in data treatment, various XPS models
overlayer thicknesses from 0.05 to 1.0 nm. (Reprinted with permission can be applied to obtain an estimation of the average particle
from reference [49]) size or layer thickness. Some very good articles have explained
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 257

in detail the basic aspects for the latter application and have Ip  
¼ ðp=sÞb σ p =σ s ðβ1 =2Þ
claimed the crucial role that XPS can have for this purpose Is ð10:8Þ
h

i
[52–55]. In the case of the analysis of heterogeneous catalysts, β1 β2 α  
1þe = 1e 1  e =α
two main models are largely discussed in the literature. Davis
introduced his model in 1989 [56] and Kerkhof-Moulijn did where ( p/s)b is the bulk atomic concentration, σ is the pho-
the same in 1979 [57]. The most important difference between toelectron cross sections, β is the dimensionless support
these two models concerns the assumptions around the role of thickness, and α is the dimensionless crystallite size. The
the catalytic support in the analyzed samples. parameters β and α are defined as (10.9):
Davis model is based on the intensity ratio of two peaks of
the same metal coming from two different levels and suffi- β1 ¼ t=λss , β2 ¼ t=λps , α ¼ d=λpp ð10:9Þ
ciently distant to ensure a marked difference in their kinetic
energy. For his model, the physical properties of the support; where t is the thickness of the support sheets and λ is the
namely, its weight fraction, specific surface area, and pore inelastic mean free path (IMFP) for the electrons from the
structure, are not determining factors of the analysis support (s) or metal (p) escaping through the support (ss) or
[56]. The use of two different peaks of the same dispersed through the metal (pp). The intensity ratio of monolayer
phase also permits to study the influence of other parameters catalyst can be predicted (very small α) using Eq. (10.8). 10
such as particle shape and surface roughness. The intensity The relation between the monolayer and crystallite catalysts
ratio (10.6) of two dispersed phase core levels is given as [57] is given by (10.10):
follows [52, 56]:
 
Ip Ip ð1  eα Þ
= ¼ ð10:10Þ
I 1 ðdÞ σ 1 T 1 λ1 βðd, λ1 Þ Is crystallites
Is monolayer
α
¼ ð10:6Þ
I 2 ðdÞ σ 2 T 2 λ2 βðd, λ2 Þ
By plotting this relationship, an estimate of α can be
where σ is the photoionization cross section, T is an instru- calculated and when λpp is known, the particle size can be
mental factor that reflects the basic detection efficiency, λ is calculated [57].
an attenuation length of the primary photoelectrons, and β is For the d-transition metals, the main problems are linked
an attenuation factor, which is characteristics of the particle to the use of the Auger peaks with high kinetic energy. As an
shape and attenuation length. The subscripts correspond example, Pd supported on TiO2 systems were studied
to the two dispersed phase XPS peaks. Among these param- [52]. For the estimation of the particle size of palladium by
eters, the β attenuation factor strongly depends on the shape the Davis model, two peaks were used for the analysis; the Pd
of the particle and may be derived for different particle 3d peak, recorded at a kinetic energy of 1149 eV, and the
shapes using the functions given by Davis [56]. In the M45N45N45 Auger peak recorded at a kinetic energy of
case of simple spherical particles, this factor is given as 325 eV. Both peaks had high cross sections, sharp shapes,
follows (10.7): and their photoelectric peak areas were determined after the
subtraction of a nonlinear “Shirley” type background [52].

 In the case of the Auger peak recorded from a pure Pd foil,
2 2
βðd, λÞ ¼ 1  2λ =d ½1  exp ðd=λÞ
peak asymmetry was observed due to the two differently
þ ð2λ=dÞ exp ðd=λÞ ð10:7Þ populated levels in which the deep hole is created before
de-excitation by the Auger process [58]. The maximum of
where d is the particle size. this peak can be determined with sufficient precision for
Contrary to the Davis model, the Kerkhof-Moulijn model samples of the pure Pd metal. This was not the same for the
[52, 57] takes into account the density, metal loading, and palladium catalysts [52]. Due to the low concentration of Pd,
specific surface area of the catalysts. Kerkhof-Moulijn used a the Auger peaks were not well resolved. Moreover, the frac-
slabs model to represent supported oxide systems. This tion of PdO present on the particle surface could be respon-
assumption was contrary to the so-called diamond model sible for peak broadening. In consequence, the exact position
(proposed by Davis). For the latter, the considered emission of the peak amplitude could not be determined with sufficient
angle is equal to Θ ¼ 45 , whereas in the slabs model the precision to permit the calculation of the modified Auger
electrons are considered to leave the sample only in the parameter. The main requirement for this type of analysis is
direction perpendicular to the surface (Θ ¼ 0 ). Assuming to know the exact areas of the peaks and the value of the
these conditions, the relationship between the intensity of the inelastic mean free path (IMFP, λ). The latter was calculated
peak of the metal (p) to the intensity of the peak of the support using the Quases-IMFP-TPP2M program based on the
ratio, Ip/Is, is given by Eq. (10.8) [57]: Tougaard model [59, 60]. The same parameters of the energy
258 P. Simon et al.

band gaps, the escape depth, and Scofield cross sections were example, the O KLL line featured at ca. 2488 eV in
used for the calculations from the Kerkhof-Moulijn model Fig. 10.10 appears at ca. 983.7 eV when recorded with an
[4, 60]. Al Kα anode.
With present-day high-energy sources, hard X-ray pho-
toelectron spectroscopy (HAXPES) experiments have
10.3.8 Photon Sources become available [63, 64]. Thanks to the high kinetic
energy of the employed sources, it is possible to reach
In XPS, the most common X-ray sources are magnesium and deeper electronic layers. Moreover, it is possible to probe
aluminum anodes whose X-rays are emitted with mean ener- materials at higher depths, which opens the way to the study
gies of 1253.6 eV (Mg Kα) and 1489.6 eV (Al Kα), respec- of buried layers or interfaces [65, 66]. Other photon sources
tively. These elements present the advantage of having a can be used for photoemission experiments. Gas discharge
small natural line width below 1 eV. The addition of a lamps are used for Ultra-Violet Photoelectron Spectros-
monochromator increases the energy resolution by selecting copy. He(I), He(II), and Ne are the most used gases and
an individual X-ray line, while increasing the signal to back- give rise to radiation of 21.22 eV, 40.81 eV, and 16.67 eV,
ground ratio. However, their energies are not sufficient to respectively, well adapted for studying band structure and
excite the deepest core levels of the majority of the elements measuring highest occupied molecular orbital (HOMO)
in the periodic table, for example, the Al 1s and Pt 3d core energy levels [67, 68].
levels. In addition, they are able to access only a very limited Synchrotron radiation is another photon source used in
number of the Auger lines of most elements, namely, the photoemission experiments. Due to the high brightness and
Auger lines of solely 24 elements can be recorded with an tunability of the radiation energy over a wide frequency
Al Kα source [61]. In the past, several research groups range, synchrotron sources offer the possibility of very
investigated the use of alternative X-ray sources with certain detailed surface electron spectroscopy. Since for any discrete
success [61]. Among them, only the Ag Lα line at 2984.3 eV X-ray line, the photoionization cross section may be at its
continues to be employed and provided by XPS manufac- maximum for one particular core level, it may be at its
turers. The spectrum in Fig. 10.10 was recorded for a sample minimum for another core level. It is thus very convenient
of platinum-rhenium nanoparticles dispersed on an alumina to be able to select a particular photon energy while
carrier where the Al 1s, Pt 3d, and Re 3d photoelectron lines maintaining a high photon intensity, yielding to an optimum
and the Auger Al KLL line are measured besides the lines in the photoionization cross section of a given core level.
that are recorded with an Al Kα anode [62]. An additional Because the mean free path of an electron in a solid depends
advantage of the use of the Ag Lα source (and other sources on its kinetic energy, another advantage of the synchrotron
of higher energy) is the shift of the Auger lines that allow source energy tunability is the possibility to carry out photon
their separation from overlapping photoelectron lines, for energy dependent depth profiling experiment [69]. Indeed,
the escape depth of the photoelectrons will vary with the
incident photon energy.
Pt 3d Re 3d
Al 1s
10.4 XPS Analysis of Photoactive Materials
Counts per second

Al KLL The photocatalytic properties of inorganic oxides depend on

O 2s different factors such as chemical composition and geomet-
Pt 3d C 1s Al 2p4f
Re rical microstructure [70–72]. Several methods can be applied
Re 3d O 1s to increase the photoactivity of a given material. For exam-
O KLL Al 2s
ple, many authors have performed chemical doping with
Cl 2o metals [73–75] or other elements such as N, F, or S
F 1s
[76]. Another possibility is to change the microstructural
properties of the oxides [76]. For example, Jami et al. [70]
showed that the photoactivity of a TiO2 with an ordered
2500 2000 1500 1000 500 0 porosity is higher than that of titanium oxide nanoparticles.
Binding energy (eV) This behavior can be explained considering the structure of
the materials. Indeed, ordered materials have higher surface
Fig. 10.10 General spectrum for a sample of platinum-rhenium nano-
area and develop interconnectivity between their pores. The
particles dispersed on an alumina matrix (molar ratio: Pt(0.5)Re(0.9)) and adsorption and diffusion of the starting substrates and the
recorded with a Ag Lα X-ray source. (Adapted from reference [62]) products increase as well as the efficiency for capturing light
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 259

due to the light scattering properties [77–80]. All these Therefore, residuals from the above processes can persist in
parameters should be taken into account during the XPS commercial samples of TiO2. Certain samples can also con-
analysis. Below, we show examples of the XPS analysis of tain phosphorus [87].
various photoactive oxides (TiO2 and Ta2O5) and CdS hoping Defective titanium oxides are ionic amphoteric semicon-
to illustrate how to proceed when analyzing the properties of ductors [88]. The latter means that the material can be either
such materials. electron donors (n-type) or receptors ( p-type). The defects
are related to changes in the surface geometry of the crystal-
line structure of the oxide [84]. The defects and
10.4.1 Titanium Oxide (TiO2) non-stoichiometry in titania may modify the occupancy of
the Ti 3d orbital of titanium. Therefore, the geometrical
Titanium oxide is one of the most (if not the most) studied distortion of the crystal structure will modify the electronic
oxide systems using XPS analysis [81]. As a reducible oxide, configuration of the oxide as following the Jahn-Teller effect
TiO2 is widely used in various technological applications [89]. Particularly, the molecular orbitals experience transi-
such as heterogeneous catalysts support, photocatalysts, or tions from the t2g and eg orbitals of the defective titanium
gas sensors [76, 82]. Chemically, titanium oxide can have dioxide will decrease the difference in energy between the d
different stoichiometries and several of them can coexist orbitals of the oxide, that is, crystal field splitting. The bond 10
when metallic Ti is oxidized with oxygen. The oxide in the oxide shifts from covalent to ionic. And, the occupancy
known with the formula TiO2 has several polymorphs of the t2g molecular orbital increase in the order: TiO2 has
among which anatase, rutile, and brookite are found in nature none, Ti2O3 has one, and TiO has two electrons.
and produced at the industrial scale. Among these three poly- Taking this into account, the knowledge of the accurate
morphs, rutile and anatase are the most investigated owing XPS BE of various oxidation states of Ti is crucial to under-
their use in catalysis and photochemistry [76, 82, 83]. But, stand the results of its analysis by XPS. Figure 10.11 shows a
titanium forms multiple nonstoichiometric oxides. The crys- typical general spectrum recorded for a sample of Degussa
tallographic structure of the titanium oxides changes as a P25 titanium dioxide. This sample was selected because this
function of their stoichiometry [84]. Magnéli phases are material is the most employed in scientific studies for the
nonstoichiometric oxides of titanium that obey the formula: technological application of the oxide [76, 90, 91]. Titanium
TinO2n1 for n ¼ 4–10 possessing a triclinic structure. There- shows peaks from the 2s, 2p, 3s, and 3p core levels. The 1s
fore, Magnéli phase have O/Ti ratios ranging from 1.9 to peak cannot be detected with the Al Kα source since it
1.75. Besides, there are oxides with stoichiometric O/Ti ratios appears at a binding energy (BE) of 4966 eV for the element
from 1.67 (Ti3O5) to 0.5 (Ti2O). These oxides are considered [92]. Besides the peaks from the core lines, signals from the
defective structures with promising applications in electro- Ti LMM Auger lines of the element appear. Accompanying
chemistry, photocatalysis, and photovoltaic panels [76, the above peaks, those corresponding to the O 1s and C 1s
82, 83]. The surface stoichiometry of titanium oxides might core levels are also present. The Auger structures of these
differ from the stoichiometry of the bulk. Therefore, a rigor- elements are also visible in the spectrum. Finally, peaks
ous analysis of XPS data is important for assessing the produced by the emission of electrons from the valence
properties of titanium oxides. band of the oxide and of the O 2s level of oxygen are detected
Carbon and traces of sulfur, chlorine, nitrogen, calcium, as well.
and sodium can be found on the surface of commercial For the TiO2 sample presented in Fig. 10.11, the peaks from
samples of TiO2. Their presence is due to industrial the core levels of titanium appeared in the following order,
manufacturing, namely, the sulfate process and the chlorine using the O 1s peak at 530.0 eVas a reference for correcting the
route, Kroll process [85]. In the sulfate process, ilmenite, BE scale of the spectra: Ti 3p (BE ≈ 37.2 eV) < Ti 3s (BE ≈
FeTiO3, the native ore used as raw material, is dissolved in 62.1 eV) < Ti 2p3/2 (BE ≈ 458.7 eV) < Ti 2p1/2 (BE ≈
sulfuric acid and the resulting titanyl sulfate is further hydro- 464.4 eV) < Ti 2s (BE ≈ 565.3 eV). The first peak of the
lyzed and precipitated as hydrated titanium dioxide [86]. This Auger Ti LMM structure appears at ca. 1071.0 eV. The BE
process has been replaced by the less expensive chlorine values for the peaks Ti 2p3/2 and Ti 2p1/2 recorded for this
route, which also yields a purer product with more uniform sample fall within 99.0% t-Student confidence intervals for the
particle size distributions. In the Kroll process, the dry ilmen- means (μ) of these peaks built from the values reported by the
ite ore is reacted with chlorine and coke to yield liquid NIST database, namely, μTi 2p3/2 ¼ 458.70  0.09 eV (t0.01;90)
titanium chloride that is further burned with oxygen, carbon and μTi 2p1/2 ¼ 464.55  0.17 eV (t0.01;11). For the Ti 3p core
monoxide, or a suitable combustible mixture. Titanium diox- level, the NIST database reports four values of BE whose mean
ide is recovered in the gas stream of the processing unit after is within the 99.0% t-Student confidence interval:
cooling with chlorine gas. The product is washed to eliminate μTi 3p ¼ 37.45  0.16 eV. On the other hand, the Ti 3s core
chlorine [86]. Washing is normally performed with alkalis. level was reported to be separated by a gap of 394.9 eV by Kim
260 P. Simon et al.

Fig. 10.11 General survey for a

sample of TiO2 (Degussa P 25).
120,000

O 1s
Acquisition parameters: Al Kα
source operated at 200 W, pass
energy ¼ 100 eV; Energy
step ¼ 1.000 eV; Spot area 100,000
~2  2 mm; analysis chamber
pressure ~1  107 Pa; Flood Gun

Ti 2p
operated at 58 μA and 2 eV. 80,000

Counts per second

Spectrum gently provided by
SurfLab-UIS and recorded in the

O KLL

Ti 2s
ACenteno-XPS/ISS/UPS Surface
Science Platform 60,000

C 1s
O 2s

+ CKVV
Ti LMM
40,000

Valence band
Ti 3p
20,000

Ti 3s
0
1200 1000 800 600 400 200 0
Binding energy (eV)

and Winograd [93] and 393.6 eV by Scrocco [94] from the Ti XPS. To calculate the O/Ti ratio in the studied sample, the
2p3/2 peak. For the sample taken as an example herein, the gap fitting of the peak from the O 1s core level is necessary. At
between Ti 2p3/2 and Ti 3s was 396.6 eV. For Ti 2s, the gaps least two contributions in the peak must be identified; one
from the Ti 2p3/2 peak were 106.4 eV [93], 106.8 eV [94], and from oxygen bonded to titanium (O-Ti) and another from
106.6 for our sample. oxygen bonded to carbon, that is, organic oxygen (OOrg).
The concentration of titanium in the sample can be calcu- The distinction between these two species depends on the
lated from any of the peaks of its core levels and even from its analysis conditions. Figure 10.13 shows that lowering the
Auger peak if the corresponding Relative Sensitivity Factors power of the Al Kα source from 200 W to 100 W and the
are available. Figure 10.12 shows high-resolution Epass from 60 eV to 15 eV led to a spectrum where the
(HR) spectra for the Ti 3p, Ti 3s, Ti 2p, and Ti 2s core levels contributions from O-Ti and OOrg were best resolved. The
and for the Ti LMM Auger line recorded for our sample. The O/Ti molar ratio can thus be calculated as the quotient
most intense peaks are those of the Ti 2p core level; notice between concentration of the selected Ti line and the O-Ti
their highest resolution, that is, lowest FWHM (e.g., FWHM component of the O 1s core line.
for Ti 2p3/2 ¼ 0.99 eV compared to 5.15 for Ti 2s). Therefore, In addition, a simultaneous peak fitting of C 1s, O 1s, and
the Ti 2p line is normally used for elemental quantification. Ti 2p core levels must be carried out with chemical sense. As
For example, for Ti the 2p1/2 line is wider than the 2p3/2 line, demonstrated above, this implies distinguishing between the
due to a broadening of the Ti 2p1/2 line by L2L3M45 Coster– organic and inorganic contributions of the O 1s core level.
Kronig processes (Fig. 10.12) [95, 96]. Very recently, Bagus The organic contribution must be correlated to the oxidized
and coworkers [97] performed theoretical calculations and species in the C 1s core level while the inorganic oxygen can
analyzed the mechanism for the broadening of the of Ti 2p1/2 be correlated to titanium. It is thus necessary to acquire high-
core line and concluded that it is due to many-body effects resolution spectra under the maximum resolution of the spec-
involving mixing of several configurations of the molecular trometer. For the case study presented herein, the following
orbitals of the oxide which imply electronic transitions fitting of the Ti 2p, O 1s, and C 1s spectra was carried out
between filled and empty orbitals. From the practical point (Fig. 10.14).
of view, the broadening of the Ti 2p1/2 line causes difficulty in Accordingly, the following species were identified and
making an accurate peak fitting of the Ti 2p region when quantified (Table 10.1).
multiple chemical states are present. As mentioned before, titanium oxides are reducible.
A first approach to identify the type of oxide at the Therefore, several treatments leading to the formation of
surface of the sample is to calculate the O/Ti ratio from Ti3+, Ti2+, and Ti0 species have been proposed in the
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 261

500 4500
Ti 3p 4000
400 Ti 2p3/2
3500

Counts per second

Counts per second

3000
300 Δ ≈ 13.6 eV
Ti 3s 2500

Ti 2p1/2
FWHM FWHM 2000
200 Δ ≈ 13.8 eV = 2.70 eV = 2.15 ev
Δ ≈ 13.8 eV FWHM
1500 = 1.95 eV FWHM
Δ ≈ 13.6 eV = 0.99 eV
100 1000

O 2peg Ti 3deg§ 500

O 2peg Ti 3deg§
0 0
75 65 55 45 35 25 505 495 485 475 465 455 445
1800
Binding energy (eV)
2300
Binding energy (eV) 10
1700 Δ ≈ 13.6 eV Ti 2s
2200

Ti LMM
1600 2100
Counts per second
Counts per second

FWHM
1500 = 5.15 ev 2000 FWHM
= 9.29 eV
1400 1900
O 2peg Ti 3deg§
1300 1800
1200 1700
1100 1600
1000 1500
610 600 590 580 570 560 550 540 1150 1130 1110 1090 1070 1050 1030 1010
Binding energy (eV) Binding energy (eV)

Fig. 10.12 High-resolution spectra taken for the peaks from the Ti 3p, analysis chamber pressure ~1  107 Pa; No charge compensation.
Ti 3s, Ti 2p, and Ti LMM lines of a sample of TiO2 (Degussa P 25). Spectra gently provided by SurfLab-UIS and recorded in the
Acquisition parameters: Al Kα source operated at 100 W, pass ACenteno-XPS/ISS/UPS Surface Science Platform
energy ¼ 15 eV; Energy step ¼ 0.050 eV; Spot area ~ 2  2 mm;

literature [76]. When TiO2 samples are submitted to such concentration of the species assigned during peak fitting to
treatments complex Ti 2p spectra are found. Unfortunately, those associated to organic and inorganic oxygen. Therefore,
most authors then perform peak fitting without considering in this sense, their attempt lacked chemical rigor.
chemical principles and end up making nice mathematical Such a lack of rigor for performing peak fitting in XPS
fittings without sound physical and chemical meaning or analysis is regrettable since the formation of partially reduced
incomplete correlations that allow verifying the stoichiome- moieties on TiO2 is key for improving its photoactivity.
try of the compounds they attribute to the XPS components. Furthermore, the relative concentration Ti0, Ti2+, Ti3+, and
4+
In the latter case, Biesinger et al. [100] made an interesting Ti species within the material defines its photoelectronic
and carefully thought proposal for decomposing the Ti 2p and structural properties. The formation of the above chem-
core level from samples of anatase and rutile TiO2 and a heat- ical moieties has been reported to be achieved by plasma
treated (though they did define how this treatment was made) treatment after XPS analysis of the treated samples by Bharti
Ti-apatite composite (Fig. 10.15). The authors used several et al. [101]. However, the peak decomposition technique
constraints for performing the proposed peak fitting based on reported by these authors is quite unusual. For example,
a thorough literature review and on physical principles they unjustifiably used a linear background, they did not
behind the XPS spectrum. However, they neither presented present peak constraints, and also introduced a single peak,
the spectra for carbon and oxygen nor correlated the relative Ti 2p3/2, for the component that they attributed for the Ti3+
262 P. Simon et al.

a
Peak Position (eV) FWHM (eV) %Area
O-Ti 4000 Ti 2p3/2 459.2 0.99 66.37
3.0 Oorg Gap Ti 2p1/2 465.0 1.97 33.33

*CPS+2.5 Ti 2p3/2
*Counts per second

Counts per second

a
3000
Oorg Gap O-Ti
2.0 *CPS+1.5
b
Gap 2000
*CPS+0.8
c Ti 2p1/2
1.0
Oorg Gap 1000
d
0.0 0
537 535 533 531 529 527 485 480 475 470 465 460 455 450 445
Binding energy (eV)
Binding energy (eV)
b Peak Position (eV) FWHM (eV) %Area
5000 O-Ti 530.00 1.13 60.60
Fig. 10.13 High-resolution spectra of the O 1s core level recorded for a C-OH* 532.32 1.76 24.57
[C=O]-O-C* 533.88 1.48 14.83
sample of TiO2 (Degussa P 25) recorded with following power of the Al

Counts per second

O-Ti
Kα source and Epass: (a) 100 W and 15 eV; (b) 100 W and 30 eV; (c) 4000
100 W and 16 eV; and (d) 200 W and 60 eV. (a–c) No charge compen-
sation. (d) Charge compensation with Flood Gun operated at 58 μA and
2 eV. In all cases: energy step ¼ 0.050 eV; Spot area ~ 2  2 mm; 3000
analysis chamber pressure ~1  107 Pa. *Normalized Counts per
second ¼ *CPS. Normalization carried out dividing CPS/max(CPS) C-OH
2000 [C=O]-O-C
for each spectrum. Spectra gently provided by SurfLab-UIS and
recorded in the ACenteno-XPS/ISS/UPS Surface Science Platform
1000

species regardless of the fact that a well-defined Ti 2p1/2 540 535 530 525
should accompany the Ti 2p3/2 for each attributed chemical Binding energy (eV)
species. Unfortunately, the literature is plagued by similar c 2500
careless analyses. Peak Position (eV) FWHM (eV) %Area
C-[C,H] 284.80 1.26 70.51
C-OH* 286.16 1.77 19.40
[C=O]-O-C* 288.76 1.06 10.09
2000
Counts per second

10.4.2 Tantalum (V) Oxide (Ta2O5)

1500
Ta2O5 is one of the most promising candidates to deal with
modern technological challenges. Due to its specific proper-
1000
ties such as low leakage current and high dielectric constant,
Ta2O5 has been used as a storage capacitor, insulator, cata-
[C=O]-O-C C-OH
lyst, and gas detector. Perez et al. [102] have investigated the 500
chemical states of the Ta2O5 surface of various samples. The
binding energies of all elements were calibrated with respect 0
to the O 1s peak (at BE of 532 eV). The analysis of the wide 295 290 285 280
spectra revealed the presence of ubiquitous carbon contami- Binding energy (eV)
nation. The Ta 4f core-level for the samples without etching
authors observed only a spin-orbit doublet corresponding to
the levels 4f7/2 and 4f5/2. The BE for the doublets is located at Fig. 10.14 High-resolution spectra for the (a) Ti 2p, (b) O 1s, and (c) C
1s core levels for a sample of TiO2 (Degussa P 25) recorded with a
27.75  0.10 eV with a spin-orbit splitting of 1.89  0.01 eV power of the Al Kα source of 100 W and an Epass of 15 eV. No charge
in case of Ta5+ in stoichiometric amorphous tantalum oxide. compensation. In all cases: energy step ¼ 0.050 eV; Spot area ~ 2  2 mm;
It is expected taking into account that oxygen-rich environ- analysis chamber pressure ~1  107 Pa. *Organic species attributed
ment that the surface would be more oxidized than the rest of following the proposal by Rouxhet and his group (Genet, Dupont-
Gillain, and Rouxhet 2008; Rouxhet and Genet 2011). The binding
the film. This was further checked by sputtering the surface of
energy scale of each core level was adjusted by comparison with values
the samples for 30 s [102] and exposed them to ambience air reported in the literature (Genet, Dupont-Gillain, and Rouxhet 2008;
for 7 days. As reported by authors the closer look at the Rouxhet and Genet 2011) and in the NIST database
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 263

Table 10.1 Chemical species quantification in a TiO2, Degussa P25, sample analyzed as received
Core level Peak Species Molar% Molar ratio
C 1s 284.80 C-[C,H]a 37.17 –
286.16 C-OHa 10.23 C-OH/C-OH ≈ 0.92
288.76 [C¼O]-O-(C,H)a 5.32 [C¼O]-O-(C,H)/[C¼O]-O-(C,H) ≈ 0.92
O 1s 530.00 O-Tib 22.08 O-Ti/O-Ti ≈ 2.02
532.32 C-OHa 9.38 C-OH/C-OH ≈ 1.09
533.88 [C¼O]-O-(C,H)a 4.88 [C¼O]-O-(C,H)/[C¼O]-O-(C,H) ≈ 1.09
Ti 2p3/2 459.20 O-Tib 7.30 O-Ti/O-Ti ≈ 2.02
Ti 2p1/2 465.00 3.65
a
Species attribution as suggested by the group of Rouxhet [42, 44]
b
Species attribution based on the NIST database. For titanium, species attribution is also in agreement with [98, 99]

oxygen present in the atmosphere and near environment of the

surface, these phases can reorganize (the surface species tend
Ti (IV) 57%
Ti (III) 8% to adsorb oxygen) to the most stable phase of Ta2O5. This was 10
Ti (II) 6% confirmed for the studied samples. The surface of the materials
Ti (0) 29% was reached in amorphous Ta2O5 as compared to the bulk
reach in TaOx. Unfortunately the XPS analysis is not adequate
to distinguish between crystalline and amorphous
phases [102].
The deep analysis of the XPS spectra revealed the pres-
IV ence of six sets of peak splitting (see Figs. 10.16 and 10.17).
This is in good agreement with the ab-initio calculations for
the δ phase of Ta2O5 films [103–105]. For Ta2O5 the spec-
IV III trum was deconvoluted with four doublets with the 4f7/2
II
III II binding energies of 22.0, 23.2, 24.6, and 26.1 eV and spin-
0 0
orbit splitting of 1.9 eV. These peaks were attributed respec-
tively to Ta1+, Ta2+, Ta3+/Ta4+, and Ta5+, originated by
induced oxygen vacancies generated during the ion bombard-
ment. In the case discussed by authors, they were able to
470 465 460 455 distinguish six doublets corresponding to various chemical
Binding energy (eV) states Ta0, Ta1+, Ta2+, Ta3+, Ta4+, and Ta5+ with binding
energies at (21.7  0.3, 23.0  0.2, 23.8  0.4, 25.2  0.4,
26.2 and 27.8 eV), respectively, and peak splitting of 1.89 eV
Fig. 10.15 Peak fitting proposed by Biesinger et al. for a “heat-treated”
each. Some discrepancies observed in the BE of each doublet
Ti-apatite sample. (Reprinted with permission from reference [100])
may be due to differences in the calibration criteria adopted
for the measuring instruments [102].
Gonçalves et al. [106] also studied the Ta2O5 pentoxide
spectrum for zero etching of F0 and the spectra of F05 and F33 (nanotubes) to explain their activity in the preferential oxida-
for 30 s etching indicated the presence of a satellite in the low tion of CO. Very interesting observations were done in the
energy region (24.6 eV  0.3 eV). This satellite according to case of Ta4f peak. As shown in Fig. 10.16, Ta 4f peak of
authors was caused by screening of 5d electrons in amorphous amorphous sample (before calcination) was decomposed in
TaOx (x ¼ 1.86, 2.00) thus suggesting the presence of TaOx in five main components. Two 4f7/2 and 4f5/2 peaks with
the studied samples. Authors assumed the presence of Ta2O5 corresponding binding energy of 26.06 and 27.96 eV were
on the surface and expected immediately after sputtering the assigned to Ta5+ (Ta2O5 oxide). Moreover two additional
depletion of O atoms and thus the generation of O vacancies. peaks were observed at about 22.60 and 24.50 eV, which
This in turn forces the reordering of the surface atoms and in were assigned to the Ta2+ components. Additionally, one
the same time favors the reconstruction and/or relaxation pro- peak at 29.20 eV was assigned to the F contamination (F 2s
cesses to guarantee the structural stability of the superstructure peak due to the preparation method).
at the surface but also electroneutrality. The formation of At the first sight the XPS spectra of the calcined samples
metastable phases (TaO, Ta2O, TaO2, TaOx, or Ta2O3) was differ from that of an anodized sample (Figs. 10.17 and
promoted by this atomic reconfiguration. However, due to the 10.18). Six main components could be detected in their
264 P. Simon et al.

Ta2O5 as-anodized 4f5/2 4f7/2 Ta2O5 800 °C 4f5/2 4f7/2 Data

Data
Fit 30 min Fit
Ta5+
Ta5+
Intensity (a.u.)

Ta4+
Ta2+

Ta2+

Intensity (a.u.)
F2s
4f5/2 4f7/2 4f5/2 4f7/2
Ta2O5 800 °C 4f5/2 4f7/2
34 32 30 28 26 24 22 20
Binding energy (eV) 60 min
Ta5+
Fig. 10.16 XPS spectra of Ta 4f and F 2s regions for as-anodized
Ta2O5 NTs. (Adapted from Ref. [106] with permission from the PCCP Ta4+
Owner Societies)
Ta2+

Ta2O5 1000 °C 4f5/2 4f7/2 Data

Fit 4f5/2 4f7/2

Ta5+ 34 32 30 28 26 24 22 20
Intensity (a.u.)

Ta3+ Binding energy (eV)

Fig. 10.18 XPS spectra of Ta 4f region for Ta2O5 NTs calcined at

Ta2+ 800  C for 30 and 60 min. (Adapted from Ref. [106] with permission
from the PCCP Owner Societies)

In the case of samples calcined in air at 800  C

4f5/2 4f7/2
(Fig. 10.18) for 30 and 60 min the 4f peak was also divided
34 32 30 28 26 24 22 20 into three different doublets. High-energy peaks at 26.16 and
Binding energy (eV) 28.06 eV were assigned to 4f7/2 and 4F5/2 peaks of Ta5+.
These values corresponded well to the BE of the bare Ta5+
oxide reported in the literature [107–109].
Fig. 10.17 XPS spectra of Ta4f region for Ta2O5 NTs calcined at However, as could be seen from Fig. 10.18, the low energy
1000  C. (Adapted from Ref. [106] with permission from the PCCP
Owner Societies) peaks at 25.36 and 27.26 eV were shifted compared to that
observed in the case of calcined (at 1000  C) sample. The quite
large shift of about 1.19 eV was assigned to the different
spectra. In the case of sample calcined at 1000  C the XPS oxidizing state of tantalum (Ta4+ instead of Ta3+). Other two
spectra of 4f peak showed fix main components low energy peaks at BE of 22.33 and 24.23 eV were assigned to
(Fig. 10.17). High energy peaks at 26.00 and 27.90 eV 4f7/2 and 4f5/2 peaks of Ta2+, respectively.
corresponding to 4f7/2 and 4f5/2 components of Ta5+. Four
low energy peaks at 22.40 and 24.17 eV should be assigned
to 4f7/2 peaks of Ta3+ and Ta2+, respectively. Other two low 10.4.3 Cadmium Sulfide (CdS)
energy peaks at 26.07 and 24.30 eV are the 4f5/2 compo-
nents of Ta3+ and Ta2+. As claimed by authors the shift in CdS is frequently used to sensitize TiO2 photocatalysts for
the BE compared to the amorphous sample (0.2 eV) is due different applications such as photosensitized solar cells with
to the different crystal structure of calcined material and/or high quantum yields [110–114] and the photocatalytic degrada-
different particle size. tion of pollutants [115, 116]. One of the most popular
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 265

a 2.0 μm a 2.2
1.5 2.0 μm
1.0 2.0
10.9 nm 0.5
7
3.1 nm 1.5 1.8 1×CdS/TiO2 3×CdS/TiO2
2.0 μm 9
1.6 16

NIR
1.0 5×CdS/TiO2 7×CdS/TiO2
1.5 1.4 28
33
0.5 1.2 9×CdS/TiO2 80
1.0
1.0 15×CdS/TiO2
0.5 0.8
2.0 μm
1.5 0 1 3 7 20 55 148
1.0
0.5 Particle size (Å)
5 μm
4 5 μm
2
3 10
b 1 b 16
nm
nm 4
5 μm 14
3 12

Deposition-cycles
4
10
2
3 8
1 6
2
4
1 5 μm 2
4
3
2 0
1 0 10 20 30 40 50 60 70 80 90
2.0 μm
1.5 2.0 μm Particle size (Å)
c 1.0
13.0 nm 0.5
2.8 nm
2.0 μm 1.5 Fig. 10.20 (a) CdS particle size calculated by Davis model vs NIR
(normalized intensity ratio calculated from the intensity ratio of pure
metal foil and studied material). Particle size evolution vs (b) number of
1.5 1.0 deposition cycles. (Reprinted with permission from Ref. [114])

1.0 0.5
bands [116, 117]. The strong dependence of the photocatalytic
0.5 efficiency and photovoltaic conversion yield on the TiO2 parti-
2.0 μm cle size was already observed [118]. Moreover, varying CdS
1.5
1.0 particles size permits tuning of the band gap energy to reach
0.5
required values. In general, the conventional methods generally
used to estimate the average particle size of CdS are UV–vis
Fig. 10.19 AFM images showing size evolution of CdS particles grown [119] spectroscopy, X-ray diffraction (XRD) [116], and the
on mesoporous TiO2 with different number of deposition cycles (a) most adapted transmission electron microscopy (TEM)
5 Å ~ CdS/TiO2, (b) 7 Å ~ CdS/TiO2, and (c) 15 Å ~ CdS/TiO2. The
band gap decreased gradually in function of size and reached 2.46 eV for
[114, 118]. However, it is still very difficult to obtain good
7 Å ~ CdS/TiO2 and further decreased to 2.39 eV for 15 Å ~ CdS/TiO2, results when the very small particles are analyzed [119]. In the
which is close to the band gap of bulk CdS. (Reprinted with permission case of UV–vis spectroscopy the values of particle size are
from Ref. [114]) usually strongly overestimated [120]. However, this technique
is very useful to explain the dependence of the band gap on
applications of CdS is the visible-light sensitizer in the case of quantum size effects [121, 122]. Ghazzal et al. [114] have
small band gap (2.4 eV) semiconductors. It is generally admit- studied QDs-CdS prepared by successive ionic layer adsorption
ted that the electron transfer between CdS and TiO2 is due to the and reaction processes which permitted to obtain different par-
differences in the level of energy for conduction and valence ticles size (Fig. 10.19). The XPS analysis was used to determine
266 P. Simon et al.

a b
Ti 2p3/2 O 1s
O-Ti
Intensity (a.u.)

Intensity (a.u.)
5.7 eV
Ti 2p1/2 O 1s (O-H)
O 1s (H2O)
NaKL1L23

475 470 465 460 455 450 540 538 536 534 532 530 528 526
Binding energy (eV) Binding energy (eV)

c d
C1s Cd3/2

C-(C,H) Cd5/2
Intensity (a.u.)

Intensity (a.u.)

C=O C-O
O-C-O
COO

292 290 288 286 284 282 280 415 410 405 400
Binding energy (eV) Binding energy (eV)

S2p1/2
(CdS)
S2p3/2
Intensity (a.u.)

(CdS)

S2p (S-O)

175 170 165 160 155

Binding energy (eV)

Fig. 10.21 XPS analysis. Spectra of Ti 2p (a), O 1s (b), C 1s (c), Cd 3d (d), and S 2p (e), and core peaks for 15 Å ~ CdS/TiO2 sample. (Reprinted
with permission from reference [114])
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 267

Table 10.2 Binding energy and parameters taken for the Davis model calculations
Sample Cd 3s BE [eV] Cd 4s BE [eV] 3s/4s ratio IMFP [nm]
CdS1 770.7 109.0 3.288 λ3s λ4s
CdS2 770.4 108.8 3.165 Cd 0.959 Cd 2.047
CdS3 770.4 108.7 2.831 CdS 1.556 CdS 2.557
CdS4 770.4 108.7 2.468
CdS5 770.3 108.6 2.378
CdS6 770.3 108.7 2.017
CdS reference 770.2 108.6 1.886
Cd Scofield cross section [eV] 3.040 0.692
Adapted from the reference [114]

the chemical states and also the average particle size and then information could be obtained for photoactive materials.
compared to the results obtained from AFM analysis [114]. The chemical state studied by XPS could be easily corre-
From the AFM analysis very small CdS particles smaller lated to the photoactive properties of studied materials.
than 1 nm were observed (Fig. 10.19). In contrast, the samples Moreover, the XPS can also be very useful for the deter- 10
prepared with 15 deposition cycles the biggest particle were mination of the averaged particle size of very small parti-
observed (8 nm). Authors stated that the CdS particle size could cles and layer thickness of multilayer systems using
be controlled by the deposition method. They used XPS to different models discussed in this work.
confirm this statement [114]. The XPS analysis of the CdS
particle size was performed using Davis model (Fig. 10.20).
The smallest particles were formed in the case of the sample References
after one deposition cycle, whereas the biggest particles were
1. Briggs, D., Seah, M.P. (eds.): Practical Surface Analysis, 2nd
observed for 15 deposition cycles. A good correlation between Ed. Volume 1 - Auger and X-Ray Photoelectron Spectroscopy.
XPS analysis and AFM studies was observed. Wiley, Chichester (1990)
For the XPS analysis (Fig. 10.21) the authors used two 2. Brundle, C.R., Baker, A.D. (eds.): Electron Spectroscopy: Theory,
Experiments and Applications. Academic Press, New York (1978)
peaks originated from two different electronic levels of CdS
3. Seah, M.P., Dench, W.A.: Quantitative electron spectroscopy of
sufficiently separated in kinetic energy: Cd 3s and Cd 4s. surfaces: a standard data base for electron inelastic mean free paths
Moreover, they assumed that the electrons leave the sample in solids. Surf. Interface Anal. 1(1), 2–11 (1979)
under an emission angle of 45 [52]. 4. Tanuma, S., Powell, C.J., Penn, D.R.: Calculations of electron
inelastic mean free paths. Surf. Interface Anal. 21, 165–176 (1993)
The normalized intensity ratio (NIR) was calculated from
5. Powell, C.J., Jablonski, A.: Surface sensitivity of X-ray photoelec-
the intensity ratio of pure CdS and the prepared samples and tron spectroscopy. Nucl. Instrum. Methods Phys. Res. Sect.
using the parameters given in Table 10.2. A Accel. Spectrom. Detect. Assoc. Equip. 601(1–2), 54–65 (2009)
6. Okada, K., Kotani, A.: Theory of core level X-ray photoemission
photoabsorption in Ti compounds. J. Electron Spectros. Relat.
Phenomena. 62, 131–140 (1993)
10.5 Conclusion 7. Borman, V.D., Lai, X.C., Pushkin, M.A., Tronin, V.N., Troyan, V.
I.: The Coster-Kronig process used to study the transition of metal
XPS analysis is a powerful technique, which permits to nanoclusters into a nonmetallic state. J. Exp. Theor. Phys. Lett. 76,
444–449 (2002)
obtain huge quantity of useful information not only in the
8. Weightman, P.: X-ray-excited Auger and photoelectron spectros-
case of inorganic materials but also in case of biological copy. Reports Prog. Phys. 45(7), 753–814 (1982)
systems or complex heterogeneous catalysts. XPS permits 9. Scofield, J.H.: Hartree-Slater subshell photoionization cross-
also to obtain the best compromise between chemical sections at 1254 and 1487 eV. J. Electron Spectros. Relat. Phenom-
ena. 8(2), 129–137 (1976)
information, qualitative, and/or quantitative values of ana- 10. Yeh, J.J., Lindau, I.: Atomic subshell photoionization cross sec-
lyzed depth. We presented the most important aspects of tions and assymetry parameters. At. Data Nucl. Data Tables. 32, 1–
the XPS analysis, starting with the principles and theoret- 155 (1985)
ical models to practical applications for the most important 11. ASTM E2735–14, Standard Guide for Selection of Calibrations
Needed for X-ray Photoelectron Spectroscopy (XPS) Experiments,
materials.
ASTM International, West Conshohocken, PA, 2014, www.
As it was shown in this work the chemical information astm.org, (accessed on 28th of April, 2019) https://doi.org/10.
obtained via XPS permits to distinguish chemical species 1520/E2735-14
of each element present in the analyzed sample. This was 12. Baldovino-Medrano, V.G., Thang Le, M., Van Driessche, I.,
Bruneel, E., Gaigneaux, E.M.: Influence of graphite as a shaping
especially demonstrated for two largely studied photo-
agent of bi molybdate powders on their mechanical, physicochem-
active oxides: TiO2 and Ta2O5. The review presented ical, and catalytic properties. Ind. Eng. Chem. Res. 50(9), 5467–
above shows that very interesting and important 5477 (2011)
268 P. Simon et al.

13. Jacquemin, M., Genet, M.J., Gaigneaux, E.M., Debecker, D.P.: 35. Desimoni, E., Casella, G.I., Cataldi, T.R.I., Malitesta, C.: A
Calibration of the X-ray photoelectron spectroscopy binding comparison of some asymmetricalline shape for XPS data
energy scale for the characterization of heterogeneous catalysts: is analysis. J. Electron Spectros. Relat. Phenomena. 49,
everything really under control? ChemPhysChem. 14(15), 3618– 247–261 (1989)
3626 (2013) 36. Sobczak, I., Szrama, K., Wojcieszak, R., Gaigneaux, E., Ziolek,
14. Greczynski, G., Hultman, L.: C 1s peak of adventitious carbon M.: CuxCryOz mixed oxide as a promisiong support for gold. The
aligns to the vacuum level: dire consequences for material’s bond- effect of Au loading method on the effectiveness in oxidation
ing assignment by photoelectron spectroscopy. ChemPhysChem. reactions. Catal. Today. 187, 48–55 (2012)
18(12), 1507–1512 (2017) 37. Walton, J., Alexander, M.R., Fairley, N., Roach, P., Shard, A.G.:
15. Kelly, M.: Historical perspectives on charging issues in XPS. Film thickness measurement and contamination layer correction
J. Electron Spectros. Relat. Phenomena. 176, 5–7 (2010) for quantitative XPS. Surf. Interface Anal. 48(3), 164–
16. Suzer, S., Sezen, H., Ertas, G., Dâna, A.: XPS measurements for 172 (2016)
probing dynamics of charging. J. Electron Spectros. Relat. Phe- 38. CasaXPS Software. http://www.casaxps.com/
nomena. 176, 52–57 (2010) 39. Shard, A.G.: Detection limits in XPS for more than 6000 binary
17. Cazaux, J.: Secondary electron emission and charging mechanisms systems using Al and Mg Kα X-rays. Surf. Inter. Anal. 46, 175–
in Auger Electron Spectroscopy and related e-beam techniques. 185 (2014)
J. Electron Spectros. Relat. Phenomena. 176, 58–79 (2010) 40. Ghosh, T.B., Sreemany, M.: On the selection of an integration
18. Tielsch, B.J., Fulghum, J.E.: Differential charging in XPS. Part I: limit for quantitative XPS analysis. Appl. Surf. Sci. 64(1), 59–
demonstration of lateral charging in a bulk insulator using imaging 70 (1993)
XPS. Surf. Inter. Anal. 24(1), 28–33 (1996) 41. Korin, E., Froumin, N., Cohen, S.: Surface analysis of nano-
19. Vereecke, G., Rouxhet, P.G.: Surface charging of insulating sam- complexes by X-ray photoelectron spectroscopy (XPS). ACS
ples in x-ray photoelectron spectroscopy. Surf. Inter. Anal. 26(7), Biomater Sci. Eng. 3, 882–889 (2017)
490–497 (1998) 42. Genet, M.J., Dupont-Gillain, C.C., Rouxhet, P.G.: XPS analysis of
20. Greczynski, C., Hultma, L.: X-ray photoelectron spectroscopy: biosystems and biomaterials. In: Matijevic, E. (ed.) Medical Appli-
towards reliable binding energy referencing. Prog. Mater. Sci. cations of Colloids, 1st edn, pp. 177–305. New York, Springer
107, 100591–100637 (2020) (2008)
21. Powell, C.J., Jablonski, A.: Progress in quantitative surface analy- 43. Piao, H., McIntyre, N.S.: Adventitious carbon growth on alumin-
sis by X-ray photoelectron spectroscopy: current status and per- ium and gold–aluminium alloy surfaces. Surf. Interface Anal. 33,
spectives. J. Electron Spectros. Relat. Phenomena. 178–179(C), 591–594 (2002)
331–346 (2010) 44. Rouxhet, P.G., Genet, M.J.: XPS analysis of bio-organic systems.
22. Wagner, C.D., Davis, L.E., Zeller, M.V., Taylor, J.A., Raymond, R. Surf. Inter. Anal. 43(12), 1453–1470 (2011)
H., Gale, L.H.: Empirical atomic sensitivity factors for quantitative 45. Desimoni, E., Casella, G.I., Salvi, A.M., Cataldi, T.R.I., Morone,
analysis by electron spectroscopy for chemical analysis. Surf. Inter. A.: XPS investigation of ultra-high-vacuum storage effects on
Anal. 3(5), 211–225 (1981) carbon fibre surfaces. Carbon. 30, 527–531 (1992)
23. Siegbahn, K.: Electron spectroscopy - an outlook. J. Electron 46. Cousens, D.R., Wood, B.J., Wang, J.Q., Atrens, A.: Implications of
Spectros. Relat. Phenomena. 5(1), 3–97 (1974) specimen preparation and of surface contamination for the mea-
24. Simpson, R., White, R., Watts, J., Baker, M.: XPS investigation of surement of the grain boundary carbon concentration of steels
monatomic and cluster argon ion sputtering of tantalum pentoxide. using x-ray microanalysis in an UHV FESTEM. Surf. Interface
Appl. Surf. Sci. 405, 79–87 (2017) Anal. 29, 23–32 (2000)
25. NIST. NIST X-ray photoelectron spectroscopy database. National 47. Miller, D.J., Biesinger, M.C., McIntyre, N.S.: Interactions of CO2
Institute of Standards and Technology. Available at: https://srdata. and CO at fractional atmosphere pressures with iron and iron oxide
nist.gov/xps/ (2012) surfaces: one possible mechanism for surface contamination? Surf.
26. Wagner, C.D.: Chemical shifts of Auger lines, and the Auger Interface Anal. 33, 299–305 (2002)
parameter. Faraday Discuss. Chem. Soc. 60, 291–300 (1975) 48. Castle, J.E., Salvi, A.M., Guascito, M.R.: Substrate-related feature
27. Swingle II, R., Riggs, W., Amy, J.: ESCA. Crit. Rev. Anal. Chem. in the loss structure of contamination C 1s. Surf. Interface Anal. 27,
5(3), 267–321 (1975) 753–760 (1999)
28. Fujikawa, T.: ‘Reprint of’ new developments in theory of X-ray 49. Smith, G.C.: Evaluation of a simple correction for the hydrocarbon
photoemission from solids. J. Electron Spectros. Relat. Phenom- contamination layer in quantitative surface analysis by XPS.
ena. 178–179, 33–60 (2010) J. Electron Spectrosc. Relat. Phenom. 148, 21–28 (2005)
29. Bagus, P.S., Ilton, E., Nelin, C.J.: Extracting chemical information 50. Ebel, M.F., Schmid, M., Ebel, H., Vogel, A.: Reduced thickness of
from XPS spectra: a perspective. Catal. Lett. 148, 1785– contamination layers determined from C 1s- and CKVV-lines.
1802 (2018) J. Electron Spectrosc. Relat. Phenom. 34, 313–316 (1984)
30. Vincent Crist, B.: A review of XPS data-banks. XPS Reports. 1, 1– 51. Vereecke, G., Rouxhet, P.G.: New method to correct for the influ-
52 (2007) ence of organic contamination on intensity ratios in quantitative
31. Kim, K.S.: Charge transfer transition accompanying x-ray photo- XPS. Surf. Interface Anal. 27, 761–769 (1999)
ionization in transition-metal compounds. J. Electron Spectros. 52. Wojcieszak, R., Genet, M., Eloy, P., Ruiz, P., Gaigneaux, E.:
Relat. Phenomena. 3(3), 217–226 (1974) Determination of the size of supported Pd nanoparticles by X-ray
32. Nesbitt, H.W., Legrand, D., Bancroft, G.M.: Interpretation of Ni2p photoelectron spectroscopy. Comparison with X-ray diffraction,
XPS spectra of Ni conductors and Ni insulators. Phys. Chem. transmission electron microscopy, and H2 chemisorption methods.
Miner. 27(5), 357–366 (2000) J. Phys. Chem. C. 114, 16677–16684 (2010)
33. Clark, D.T., Dilks, A.: ESCA studies of polymers. VII. Shake-up 53. La Surface Data Base. XPS database. www.lasurface.com
phenomena in some alkane-styrene copolymers. J. Polym. Sci. 54. Wojcieszak, R., Jasik, A., Monteverdi, S., Ziolek, M., Bettahar, M.
Polym. Chem. Ed. 14(3), 533–542 (1976) M.: Nickel niobia interaction in non-classical Ni/Nb2O5 catalysts.
34. Hüfner, S., Wertheim, G.K.: Core-line asymmetries in the x-ray- J. Mol. Catal. Chem. 256, 225–233 (2006)
photoemission spectra of metals. Phys. Rev. B. 11(2), 678– 55. Militello, M.C., Simko, S.J.: Palladium oxide (PdO) by XPS. Surf.
683 (1975) Sci. Spectra. 3(4), 387–394 (1997)
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 269

56. Mark Davis, S.: Particle size information from dispersed phase 74. Castro, C., Osorio, P., Sienkiewicz, A., Pulgarin, C., Centeno, A.,
photoemission intensity ratios. J. Catal. 117, 432–446 (1989) Giraldo, S.: Photocatalytic production of 1O2 and •OH mediated by
57. Kerkhof, F.P.J.M., Moulijn, J.A.: Quantitative analysis of XPS silver oxidation during the photoinactivation of Escherichia coli
intensities for supported catalysts. J. Phys. Chem. 83, 1612– with TiO2. J. Hazard. Mat. 211-212, 172–181 (2012)
1619 (1979) 75. Castro-Lopez, C., Centeno, A., Giraldo, S.: Fe-modified TiO2
58. Brun, M., Berthet, A., Betolini, J.C.: XPS, AES and Auger param- photocatalysts for the oxidative degradation of recalcitrant water
eter of Pd and PdO. J. Elect. Spectr. Related Phenom. 104(55–60) contaminants. Catal. Today. 157, 119–124 (2010)
(1999) 76. Liu, L., Chen, X.: Titanium dioxide nanomaterials: self-structural
59. Tougaard, S.: QUASES-IMFP-TPP2M Program. Quases-Tougaard modifications. Chem. Rev. 114, 9890–9918 (2014)
Inc. (2002) 77. Tschirch, J., Bahnemann, D., Wark, M., Rathouský, J.: A compar-
60. Strehlow, W.H., Cook, E.L.: Compilation of energy band gaps in ative study into the photocatalytic properties of thin mesoporous
elemental and binary compound semiconductors and insulators. layers of TiO2 with controlled mesoporosity. J. Photochem. Photo-
J. Phys. Ref. Data. 2, 163–199 (1973) bio. A: Chem. 194, 181–188 (2008)
61. Wagner, C.D., Joshi, A.: The auger parameter, its utility and 78. Ismail, A.A., Hakki, A., Bahnemann, D.W.: Mesostructure
advantages: a review. J. Elect. Spect. Related Phen. 47, 283– Au/TiO2 nanocomposites for highly efficient catalytic reduction
313 (1988) of p-nitrophenol. J. Mol. Catal. A Chem. 358, 145–151 (2012)
62. J. Rodríguez-Pereira, MSc Thesis, Universidad Industrial de San- 79. Zhao, Y., Zhang, X., Zhai, J., Jiang, L., Liu, Z., Nishimoto, S.,
tander, Bucaramanga, Colombia, 2018, available at: http://tangara. Murakami, T., Fujishima, A., Zhu, D.: Ultrastable TiO2 foams
uis.edu.co/biblioweb/tesis/2018/173089.pdf derived macro-/meso-porous material and its photocatalytic activ-
63. Wojcik, J.C.: Hard X-ray photoelectron spectroscopy ity. Microp. Mesop. Mater. 116, 710–714 (2008) 10
(HAXPES). In: Springer Series in Surface Science, vol. 59. 80. Ismail, A.A., Bahnemann, D.W., Rathousky, J., Yarovyi, V., Wark,
Springer (2016) M.: Multilayered ordered mesoporous platinum/titania composite
64. Regoutz, A., Mascheck, M., Wiell, T., Eriksson, S., Lijenberg, C., films: does the photocatalytic activity benefit from the film thick-
Tetzner, K., Williamson, B., Scanlon, D., Palmgren, P.: A novel ness? J. Mater. Chem. 21, 7802–7810 (2011)
laboratory-based hard X-ray photoelectron spectroscopy system. 81. Diebold, U.: The surface science of titanium dioxide. Surf. Sci.
Rew. Sci. Instr. 89, 073105 (2018) Rep. 48, 53–229 (2003)
65. Sotelo-Vazquez, C., Quesada-Cabrera, R., Ling, M., Scanlon, D. 82. Chen, X., Mao, S.: Titanium dioxide nanomaterials: synthesis,
O., Kafizas, A., Thakur, P.K., Lee, T.-L., Taylor, A., Watson, G.W., properties, modifications, and applications. Chem. Rev. 107,
Palgrave, R.G., Durrant, J.R., Blackman, C.S., Parkin, I.P.: Evi- 2891–2959 (2007)
dence and effect of photogenerated charge transfer for enhanced 83. Nowotny, J.: Oxide Semiconductors for Solar Energy Conversion:
photocatalysis in WO3/TiO2 heterojunction films: a computational Titanium Dioxide, 1st edn. CRC Press, Boca Ratón (2017)., ISBN
and experimental study. Adv. Funct. Mater. 27(18), 9781138074873
1605413 (2017) 84. Jayashree, S., Ashokkumar, M.: Switchable intrinsic defect chem-
66. Rumaiz, A.K., Woicik, J.C., Cockayne, E., Lin, H.Y., Jaffari, G.H., istry of titania for catalytic applications. Catalysts. 8(12), 601–
Shah, S.I.: Oxygen vacancies in N doped anatase TiO2: experiment 626 (2018)
and first-principles calculations. Appl. Phys. Lett. 95(26), 85. Wiley-VCH (Ed.): Ullmann’s Encyclopedia of Industrial Chemis-
262111 (2009) try. 7th ed. (Wiley-VCH, 2011)
67. Harwell, J.R., Baikie, T.K., Baikie, I.D., Payne, J.L., Ni, C., Irvine, 86. Grzmil, B., Grela, D., Kic, B.: Formation of hydrated titanium
J.T.S., Turnbull, G.A., Samuel, I.D.W.: Probing the energy levels dioxide from seeded titanyl sulphate solution. Chem. Papers.
of perovskite solar cells: via Kelvin probe and UV ambient pressure 63(2), 217–225 (2009)
photoemission spectroscopy. Phys. Chem. Chem. Phys. 18(29), 87. Baldovino-Medrano, V.G., Pollefeyt, G., Bliznuk, V., van Driessche,
19738–19745 (2016) I., Gaigneaux, E., Ruiz, P., Wojcieszak, R.: Synergetic behavior of
68. Snook, J.H., Samuelson, L.A., Kumar, J., Kim, Y.G., Whitten, J.E.: TiO2-supported Pd(z)Pt(1-z) catalysts in the green synthesis of
Ultraviolet photoelectron spectroscopy of nanocrystalline TiO2 methyl formate. ChemCatChem. 8(6), 1157–1166 (2016)
films sensitized with (2,2′-bipyridyl)ruthenium(II) dyes for photo- 88. Nowotny, M.K., Bak, T., Nowotny, J.: Electrical properties and
voltaic applications. Org. Electron. Physics Mater. Appl. 6(2), 55– defect chemistry of TiO2 single crystal. I. Electrical conductivity.
64 (2005) J. Phys. Chem. B. 110(33), 16270–16282 (2006)
69. Bluhm, H., Hävecker, M., Knop-Gericke, A., Kleimenov, E., 89. Jahn, H.A., Teller, E.: Stability of polyatomic molecules in degen-
Schlögl, R., Teschner, D., Bukhtiyarov, V.I., Ogletree, D.F., erate electronic states. I. Orbital degeneracy. Proc. R. Soc.
Salmeron, M.: Methanol oxidation on a copper catalyst investi- A. 161(A905), 220–235 (1937)
gated using in situ X-ray photoelectron spectroscopy. J. Phys. 90. Ohtani, B., Prieto Mahaney, O., Li, D., Abe, R.: What is Degussa
Chem. B. 108(38), 14340–14347 (2004) (Evonik) P25? Crystalline composition analysis, reconstruction
70. Jami, M., Dillert, R., Suo, Y., Bahnemann, D., Wark, M.: Photo- from isolated pure particles and photocatalytic activity test.
activity of titanium dioxide foams. Int. J. Photoenergy. 2018, J. Photochem. PhotoBio. A: Chem. 216, 179–182 (2010)
5057814 (2018) 91. Hurum, D., Agrios, A., Gray, K., Rajh, T., Thurnauer, M.:
71. Vogelbaum, H., Sauvé, G.: Recently developed high-efficiency Explaining the enhanced photocatalytic activity of Degussa P25
organic photoactive materials for printable photovoltaic cells: a mixed-phase TiO2 using EPR. J. Phys. Chem. B. 107, 4545–
mini review. Synth. Metals. 223, 107–121 (2017) 4549 (2003)
72. Carius, R.: Structural and optical properties of microcrystalline silicon 92. van der Heide, P.: X-ray Photoelectron Spectroscopy: An Introduc-
for solar Celle applications. In: Marshall, J., Dimova-Malinovska, tion to Principles and Practices, ISBN 978-1-118-06253-1, p. 264.
D. (eds.) Photovoltic and Photoactive Materials, pp. 93–108. Springer, Wiley (2011)
Netherlands (2002, ISBN: 978-1-4020-0824-5) 93. Kim, K.S., Winograd, N.: Charge transfer shake-up satellites in
73. Castro, C., Centeno, A., Goraldo, S.: Iron promotion of the TiO2 X-ray photoelectron spectra of cations and anions of SrTiO3, TiO2
photosensitization process towards the photocatalytic oxidation of and Sc2O3. Chem. Phys. Lett. 31(2), 312–317 (1975)
azo dyes under solar-simulated light irradiation. Mat. Chem. Phys. 94. Scrocco, M.: X-ray photoelectron spectra of Ti4+ in TiO2. Evidence
129, 1176–1183 (2011) of band structure. Chem. Phys. Lett. 61(3), 453–456 (1979)
270 P. Simon et al.

95. Coster, D., Kronig, R.D.L.: New type of Auger effect and its X-ray photoelectron spectroscopy and AFM. Beilstein
influence on the x-ray spectrum. Physica. 2, 13–24 (1935) J. Nanotechnol. 5, 68–76 (2014)
96. Ohno, M.: Effects of Coster–Kronig fluctuation and decay on 115. Malashchonak, M.V., Poznyak, S.K., Streltsov, E.A., Kulak, A.I.,
X-ray photoelectron spectroscopy spectra. J. Elect. Spectr. Related Korolik, O.V., Mazanik, A.V.: Photoelectrochemical and Raman
Phenom. 131–132, 3–28 (2003) characterization of In2O3 mesoporous films sensitized by CdS
97. Bagus, P., Nelin, C., Brundle, C., Chambers, S.: A new mechanism nanoparticles. Beilstein J. Nanotechnol. 4, 255–261 (2013)
for XPS line broadening: the 2p-XPS of Ti(IV). J. Phys. Chem. 116. Bessekhouad, Y., Chaoui, N., Trzpit, M., Ghazzal, N., Robert, D.,
C. 123, 7705–7716 (2019) Weber, J.V.: UV–vis versus visible degradation of Acid Orange II
98. Diebold, U.: TiO2 by XPS. Surf. Sci. Spectra. 4, 227 (1996) in a coupled CdS/TiO2 semiconductors suspension J. Photochem.
99. Mayer, J.T., Garfunkel, E., Madey, T.E., Diebold, U.: Titanium and Photobiol. A. 183, 218–224 (2006)
reduced titania overlayers on titanium dioxide (110). J. Electron 117. Yu, P., Zhu, K., Norman, A.G., Ferrere, S., Frank, A.J., Nozik, A.:
Spectrosc. Relat. Phenom. 73, 1–11 (1995) Nanocrystalline TiO2 solar cells sensitized with InAs quantum
100. Biesinger, M., Lau, L., Gerson, A., Smart, R.: Resolving surface dots. J. Phys. Chem. B. 110, 25451–25454 (2006)
chemical states in XPS analysis of first row transition metals, 118. Ahmed, R., Will, G., Bell, J., Wang, H.: Size-dependent photo-
oxides and hydroxides: Sc, Ti, V, Cu and Zn. Appl. Surf. Sci. degradation of CdS particles deposited onto TiO2 mesoporous films
257, 887–898 (2010) by SILAR method. J. Nanopart. Res. 14, 1140–1153 (2012)
101. Bharti, B., Kumar, S., Lee, H., Kumar, R.: Formation of oxygen 119. Besson, S., Gacoin, T., Ricolleau, C., Jacquiod, C., Boilot, J.-P.: 3D
vacancies and Ti3+ state in TiO2 thin film and enhanced optical quantum dot lattice inside mesoporous silica films. Nano Lett. 2,
properties by air plasma treatment. Sci. Rep. 6, 32355 (2016) 409–414 (2002)
102. Perez, I., Sosa, V., Perera, F., Galindo, J., Carrejo, J., Gonzalez, P., 120. Lippens, P., Lannoo, M.: Calculation of the band gap for small
Rodriguez, C.: XPS depth profiling analysis of crystalline tantalum CdS and ZnS crystallites. Phys. Rev. B. 39, 10935–
pentoxide films. Appl. Surf. Sci.. arXiv:1804.02067, 10 (2018) 10942 (1989)
103. Atanassova, E., Spassov, D.: X-ray photoelectron spectroscopy of 121. Wang, Y., Herron, N.: Quantum size effects on the exciton energy
thermal thin Ta2O5 films on Si. Appl. Surf. Sci. 135, 71–82 (1998) of CdS clusters. Phys. Rev. B. 42, 7253–7255 (1990)
104. Ivanov, M.V., Perevalov, T.V., Aliev, V.S., Gritsenko, V.A., 122. Wang, Y., Suna, A., Mahler, W., Kasowski, R.: PbS in polymers.
Kaichev, V.V.: Electronic structure of δ-Ta2O5 with oxygen From molecules to bulk solids. J. Chem. Phys. 87, 7315–
vacancy: ab initio calculations and comparison with experiment. 7322 (1987)
J. Appl. Phys. 110, 024115 (2011)
105. Ivanov, M.V., Perevalov, T.V., Aliev, V.S., Gritsenko, V.A.,
Kaichev, V.V.: Ab initio simulation of the electronic structure of
δ-Ta2O5 with oxygen vacancy and comparison with experiment.
J. Exp. Theo. Phys. 112, 1035–1041 (2011)
106. Gonçalves, R.V., Wojcieszak, R., Uberman, P.M., Teixeira, S.R.,
Rossi, L.M.: Insights into the active surface species formed on
Ta2O5 nanotubes in the catalytic oxidation of CO. Phys. Chem.
Chem. Phys. 16, 5755–5762 (2014)
107. Wang, K., Liu, Z., Cruz, T.H., Salmeron, M., Liang, H.: In situ
spectroscopic observation of activation and transformation of tan-
talum suboxides. J. Phys. Chem. A. 114, 2489–2497 (2010)
108. Roginskaya, Y.E., Morozova, O.V., Loubnin, E.N., Popov, A.V.,
Ulitina, Y.I., Zhurov, V.V., Ivanov, S.A., Trasatti, S.: X-ray diffrac-
tion, transmission electron microscopy and X-ray photoelectron
spectroscopic characterization of IrO2 + Ta2O5 films. J. Chem.
Soc. Faraday Trans. 89, 1707–1715 (1993) Pardis Simon is a research engineer of the Centre National de la
109. Wang, J., Su, S., Liu, B., Cao, M., Hu, C.: One-pot, Recherche Scientifique (CNRS) at the University of Lille, France. She
low-temperature synthesis of self-doped NaTaO3 nanoclusters for received her Ph.D. in material science from the University of Paris-
visible-light-driven photocatalysis. Chem. Commun. 49, 7830– Saclay in 2011. She specializes in X-ray Photoelectron Spectroscopy
7832 (2013) and her research focuses mainly on the in situ study of heterogeneous
110. Vogel, R., Hoyer, P., Weller, H.: Quantum-sized PbS, CdS, Ag2S, catalysts.
Sb2S3, and Bi2S3 particles as sensitizers for various nanoporous wide-
bandgap semiconductors. J. Phys. Chem. 98, 3183–3188 (1994)
111. Jin-nouchi, Y., Naya, S., Tada, H.: Quantum-dot-sensitized solar
cell using a photoanode prepared by in situ photodeposition of CdS
on nanocrystalline TiO2 films. J. Phys. Chem. C. 114, 16837–
16842 (2010)
112. Baker, D.R., Kamat, P.V.: Photosensitization of TiO2 nano-
structures with CdS quantum dots: particulate versus tubular sup-
port architectures. Adv. Funct. Mater. 19, 805–811 (2009)
113. Lee, H.-J., Leventis, H.C., Moon, S.-J., Chen, P., Ito, S., Haque, S. Víctor G. Baldovino-Medrano completed Ph.D. studies at
A., Torres, T., Nüesch, F., Geiger, T., Zakeeruddin, S.M., Grätzel, Universidad Industrial de Santander -UIS-, studying catalysts for the
M., Nazeeruddin, M.K.: PbS and CdS quantum dot-sensitized hydrodesulfurization of dibenzothiophenes. Afterwards, he was a post-
solid-state solar cells: “old concepts, new results”. Adv. Funct. doc at UCLouvain for 5 years working on understanding the effects of
Mater. 19, 2735–2742 (2009) shaping on the properties of catalysts for oxidation. He is Assistant
114. Ghazzal, M., Wojcieszak, R., Raj, G., Gaigneaux, E.: Study of Professor at UIS where he researches on heterogeneous catalysis and
mesoporous CdS-quantum-dot-sensitized TiO2 films by using applied surface science.
10 X-Ray Photoelectron Spectroscopy (XPS): Principles and Application for the Analysis of Photoactive Materials 271

Dr. Robert Wojcieszak is the senior researcher of CNRS at the Uni-

versity of Lille, France. He obtained his Ph.D. in Molecular Chemistry
from Henri Poincare University in 2006 and subsequently held a Post-
doctoral Fellowship in the Institute of Condensed Matter and Nano-
materials at the Catholic University of Louvain, Belgium. His research 10
interests cover materials science and catalytic materials design.