Ultraviolet Photoelectron Spectroscopy:Practical Aspects and Best Practices

Ultraviolet Photoelectron Spectroscopy (UPS): Applications, Principles, and Best Practices

Although ultraviolet photoelectron spectroscopy (UPS) is not as commonly used as X-ray photoelectron spectroscopy (XPS), it plays a significant role in chemistry, physics, and materials science. Similar to XPS, UPS involves the irradiation of a sample with photons, which induces the ejection of photoelectrons. However, instead of using X-rays, UPS employs deep ultraviolet (UV) photons. Compared to X-rays, UV photons have lower energy, which results in the ejection of valence band electrons rather than core-level electrons.

Unlike XPS, which is used for elemental analysis and oxidation state identification, UPS is primarily used to measure the ionization energy of valence shell electrons. This is particularly important for understanding the interactions between chemisorbed species and surfaces or for characterizing the electronic band structure of inorganic and organic semiconductors. Additionally, UPS is useful for determining the work function (Φ) of metals or semiconductor surfaces.

However, unlike XPS, UPS cannot be used on insulating surfaces. The ejected electrons cause charge accumulation, leading to unpredictable electric fields that affect the kinetic energy and trajectory of low-energy electrons, making measurements impossible.

Three-Step Model of the UPS Process

Figure 1 illustrates the three-step model of the UPS process and defines relevant parameters.

1. Photoelectron Excitation

   In the first step, an incident photon excites an electron from the valence band to a state above the vacuum energy level (E_VAC). Since the excited electrons originate from different levels within the valence band, their energy distribution above E_VAC roughly reflects the valence band structure, as shown in Figure 1a.

2. Electron Transport to the Surface

   Before escaping, the electrons must travel to the surface (step two). Some electrons reach the surface elastically, maintaining their initial energy relative to the valence band, while others lose kinetic energy due to inelastic scattering. The energy loss depends on the details of their transport path, leading to the formation of secondary electrons.

   Secondary electrons extend from the Fermi level (EF) to above the vacuum energy level (Figure 1b). The Fermi level is the energy reference of the photoelectron spectrometer and the metal sample holder or stage it is connected to. This is a critical point, as the Fermi level of the analyzed sample may not be equal to the spectrometer’s unless they are in electrical equilibrium.

3. Electron Escape and Detection

   In the third step, electrons escape into the vacuum. The spectrometer detects both secondary electrons and directly emitted valence band electrons, but only those with energy above the vacuum level can be detected. This results in the observed UPS spectrum (Figure 1c), where contributions from both valence band electrons and secondary electrons are present. The secondary electrons primarily dominate the region extending toward the vacuum level.

Figure 1:Illustration of the processes leading to the UPS spectrum

For He I and He II radiation, Barrie et al. [1] proposed that emission originates from the atomic decay of neutral and singly ionized He atoms from the 2p to 1s states. The linewidth is approximately 3 meV, depending on discharge conditions.

In contrast, Mg Kα and Al Kα X-rays have full width at half maximum (FWHM) values of 0.70 eV and 0.85 eV, respectively. Even monochromatic Al Kα X-rays have a typical FWHM of about 0.35 eV. The energy resolution of photoemission experiments is determined by the energy spread of the photon source, the resolution of the electron energy analyzer, and the intrinsic width of the electronic states from which electrons are emitted.

Sample inhomogeneity may affect the latter. In XPS, the energy spread of X-ray radiation is often the limiting factor. In contrast, laboratory UPS sources, which are typically discharge lamps, generate radiation from atomic transitions. As a result, the energy spread is very narrow and does not limit the overall energy resolution.

However, emitted radiation may contain contributions from satellite lines, accounting for a few percent of the total intensity. These include He Iα, He Iβ, and He Iγ, located at 23.09 eV, 23.75 eV, and 24.05 eV, respectively, which may introduce weak satellite structures in UPS spectra. Typically, this is not an issue.

However, when analyzing states near the valence band maximum or states extending into the semiconductor band gap, Zhang et al. [2] suggested that these satellite contributions may cause significant interference. They must be subtracted to avoid measurement artifacts. A monochromator can be added to the discharge lamp output to eliminate this issue, but this increases complexity and cost while potentially reducing photon flux. It is also worth noting that dedicated monochromatic light sources for photoemission experiments have been developed.

Energy Calibration and Data Representation

In the following equation, V_Bias represents the voltage applied to the sample, with the negative sign retained:

E_Binding = hΓ - (E_Kinetic + Φ)

Equation assumes that the spectrometer’s work function Φ has already been accounted for in the instrument calibration, which is typically the case in XPS measurements.

For some studies, it is preferable to present data relative to the vacuum level. In this case, UPS peak energies represent orbital ionization energies, similar to the ionization energies of gas-phase molecules.

Thus, when data is plotted in this manner, the x-axis is often labeled "Ionization Energy." This representation is sometimes chosen when comparing adsorbed molecules to gas-phase molecules or when evaluating theoretical ionization energies to understand how adsorption influences molecular orbitals.

Figure 2 illustrates the relationship between UPS spectra plotted in terms of binding energy (relative to the Fermi level) and ionization energy (relative to the vacuum level). It is crucial for authors to explicitly state how data is presented in figure captions or text in research papers and reports.

Figure 2:Comparison of He I UPS spectra plotted relative to the Fermi level and the vacuum level

[1] A. Barrie, Instrumentation for electron spectroscopy, in: D. Briggs (Ed.), Handbook of X-Ray and Ultraviolet Photoelectron Spectroscopy, Heyden & Son Ltd, London, 1978, pp. 79–119.

[2] F. Zhang, S.H. Silver, N.K. Noel, F. Ullrich, B.P. Rand, A. Kahn, Ultraviolet photoemission spectroscopy and Kelvin probe measurements on metal halide perovskites: advantages and pitfalls, Adv. Energy Mater. 10 (2020) 1–7, 1903252.