Physical Chemistry Foundations

The Photoelectric Effect

X-ray Photoelectron Spectroscopy (XPS) is fundamentally governed by the photoelectric effect. When a material is irradiated with monochromatic X-rays (commonly Al K$\alpha$ at 1486.6 eV or Mg K$\alpha$ at 1253.6 eV), photons interact with core-level electrons. If the photon energy $h\nu$ exceeds the electron’s binding energy $E_b$, the electron is emitted into the vacuum.

The kinetic energy $E_k$ of the emitted photoelectron is measured by the spectrometer’s hemispherical analyzer. By conservation of energy, the binding energy is calculated as:

$E_b = h\nu - E_k - \Phi$

Where $\Phi$ represents the work function of the spectrometer (typically around 4 to 5 eV). Because binding energies are quantized and unique to specific atomic orbitals, XPS provides direct elemental identification.

Chemical Shift

While the base binding energy identifies the element, slight variations in this energy—termed chemical shifts—provide information regarding the atom’s local chemical environment.

The binding energy of a core electron is influenced by the valence electron density. When an atom bonds with a highly electronegative ligand, valence electron density is drawn away, reducing the electrostatic shielding of the core electrons. Consequently, the effective nuclear charge increases, resulting in a higher binding energy.

For instance, in the case of Titanium, the $Ti^{0}$ metallic state exhibits a $2p_{3/2}$ binding energy of approximately 453.8 eV. In contrast, Titanium in a fully oxidized state ($Ti^{4+}$, as found in $SrTiO_3$ or $TiO_2$) experiences significant electron withdrawal from the oxygen lattice, shifting the $2p_{3/2}$ peak to approximately 458.5 eV. Intermediate oxidation states (like $Ti^{3+}$) appear between these extremes, allowing for the quantification of defect structures or mixed-valence oxides.

Spin-Orbit Coupling

For electrons originating from orbitals with an orbital angular momentum quantum number $l > 0$ (such as $p, d,$ and $f$ orbitals), the magnetic field generated by the electron’s orbit interacts with its intrinsic spin. This phenomenon, known as spin-orbit coupling, splits the energy level into two distinct states, resulting in a doublet in the XPS spectrum.

The total angular momentum quantum number $j$ is given by:

$j = |l \pm s|$

Where $s = 1/2$. For a $p$ orbital ($l=1$), the possible values for $j$ are $1/2$ and $3/2$. Therefore, the emission from a $2p$ orbital manifests as two peaks: $2p_{1/2}$ and $2p_{3/2}$.

The relative intensity of these doublet peaks is strictly governed by the quantum mechanical degeneracy ($2j + 1$) of each state. The theoretical area ratios are mathematically fixed:

• $p$ orbitals ($j=1/2, 3/2$): Area ratio 1:2
• $d$ orbitals ($j=3/2, 5/2$): Area ratio 2:3
• $f$ orbitals ($j=5/2, 7/2$): Area ratio 3:4

Furthermore, the energy separation between the two peaks ($\Delta E$) is a constant specific to the element and the specific orbital. For example, the Strontium $3d$ doublet exhibits a fixed separation of $\approx 1.8$ eV, while the Titanium $2p$ doublet exhibits a separation of $\approx 5.7$ eV. These physical constraints are utilized during the non-linear curve fitting process to ensure mathematically valid deconvolutions.

Inelastic Mean Free Path (IMFP)

XPS is an inherently surface-sensitive technique. While X-rays penetrate deeply into the bulk material (often several micrometers), the emitted photoelectrons have a high probability of undergoing inelastic collisions with the lattice. Only electrons that escape the solid without losing energy contribute to the characteristic photoemission peaks.

The probability of an electron escaping without inelastic scattering decays exponentially with depth $d$:

$I(d) = I_0 \exp\left(-\frac{d}{\lambda \cos \theta}\right)$

Where $\lambda$ is the Inelastic Mean Free Path (IMFP) and $\theta$ is the emission angle relative to the surface normal. For typical XPS kinetic energies (100 - 1000 eV), $\lambda$ ranges from 1 to 3 nm. Consequently, 95% of the unscattered photoelectron signal originates from within a depth of $3\lambda$ (approximately 5 to 10 nm), defining the analytical volume of the technique. Electrons that suffer energy losses before escaping contribute to the step-like background observed on the higher binding-energy side of each photoemission peak.