Photoelectron Spectroscopy

Photoelectron spectroscopy (PES) is an experimental quantum-chemical technique that determines the binding energies of electrons in atoms, molecules, and condensed-phase materials by irradiating a sample with monochromatic vacuum-ultraviolet (VUV) or X-ray photons and analyzing the translational kinetic energy of the emitted photoelectrons. Grounded in Albert Einstein's 1905 quantum theory of the photoelectric effect, photoelectron spectroscopy provides the most direct empirical validation of canonical molecular orbital theory and electron configuration shell structure. When a photon of frequency ν and energy hν exceeds the ionization threshold of a bound electron in orbital k, a single-photon photoionization event ejects the electron into the continuum:

M + hν → M_k^+• (v′, J′) + e⁻ (E_k)

By conservation of total energy, the measured translational kinetic energy E_k of the outgoing photoelectron moving at velocity v_e (where E_k = ½m_ev_e²) is related to the ionization energy I_k required to produce the radical cation M_k^+• in vibrational state v′ and rotational state J′:

E_k = hν − I_k − E_vib⁺(v′) − E_rot⁺(J′)

Because each canonical molecular orbital possesses a distinct binding energy and symmetry, a photoelectron spectrum plots photoelectron count rate against ionization energy I_k = hν − E_k, revealing discrete ionization bands whose positions, relative integrated intensities, and vibrational fine structures map the electronic structure of the target molecule.

Photon Sources and Core vs. Valence Regimes

Photoelectron spectroscopy is divided into two primary operational regimes according to the energy of the ionizing radiation source:

1. Ultraviolet Photoelectron Spectroscopy (UPS): Developed in the 1960s by David W. Turner for gas-phase molecules, UPS employs a windowless helium gas-discharge lamp emitting monochromatic He Iα radiation at 58.43 nm (hν = 21.22 eV) or He IIα radiation at 30.38 nm (hν = 40.81 eV) with a narrow natural line width below 3 meV. Because photon energies of 21.22–40.81 eV match the binding energies of outer-shell valence electrons (typically 7–35 eV), UPS selectively ionizes valence bonding, non-bonding, and antibonding molecular orbitals while resolving individual vibrational quanta of the resulting molecular ion.
2. X-Ray Photoelectron Spectroscopy (XPS): Pioneered by Kai Siegbahn (who termed it Electron Spectroscopy for Chemical Analysis, or ESCA), XPS uses soft X-ray anodes—most commonly aluminum Kα (hν = 1486.6 eV) or magnesium Kα (hν = 1253.6 eV)—or tunable synchrotron radiation. These high-energy photons eject tightly bound inner-shell core electrons (such as C 1s at ~284.8 eV, N 1s at ~399 eV, and O 1s at ~532 eV). Although core orbitals do not participate directly in covalent bond formation, their binding energies exhibit reproducible "chemical shifts" of 0.5–10 eV that reflect the local electrostatic potential and oxidation state of the emitting atom, governed by the electronegativity of bonded substituents.

Koopmans' Theorem and Electronic Relaxation Corrections

The theoretical link between experimental ionization energies I_k and Hartree-Fock self-consistent field (SCF) orbital eigenvalues ε_k is established by Koopmans' theorem (formulated by Tjalling Koopmans in 1934). Within the frozen-orbital approximation—which assumes that the spatial wavefunctions of the remaining N − 1 electrons do not relax upon sudden removal of an electron from spin-orbital φ_k—the vertical ionization energy equals the negative of the canonical Hartree-Fock orbital energy:

I_k^KT = E_HF(N − 1, frozen) − E_HF(N) = −ε_k

In real polyatomic molecules, two opposing physical effects modify Koopmans' theorem:

1. Orbital Relaxation (ΔE_relax < 0): When an electron is removed, the remaining N − 1 electrons experience reduced interelectronic screening and contract toward the positively charged nuclei, stabilizing the cation and lowering the true ionization energy relative to −ε_k by 1 to 4 eV.
2. Electron Correlation (ΔE_corr > 0): Because the N-electron neutral ground state contains more electron pairs than the (N − 1)-electron cation, dynamic electron correlation stabilizes the neutral molecule more than the ion, increasing the true ionization energy by 0.5 to 3 eV.

Thus, the vertical ionization energy is given by I_k = −ε_k + ΔE_relax + ΔE_corr + ΔE_rel. For most main-group molecules, relaxation and correlation corrections partially cancel within 0.5–1.5 eV, allowing Koopmans' theorem to predict the ordering and symmetry assignments of valence photoelectron bands accurately, with well-documented exceptions in compact lone-pair orbitals such as the 1π_u / 3σ_g pair of N₂.

Franck-Condon Envelopes in Molecular Ionization

Because photoionization occurs on an attosecond timescale (~10⁻¹⁶ s) that is far faster than nuclear vibrational motion (~10⁻¹⁴ to 10⁻¹³ s), the transition intensity into each vibrational level v′ of the molecular cation is governed by the Franck-Condon principle. The intensity of the (v′ ← v″ = 0) vibronic peak is proportional to the square of the vibrational overlap integral, |⟨χ_v′⁺ | χ₀⟩|². The shape of a UPS vibrational progression directly diagnoses the bonding character of the ionized molecular orbital:

1. Non-Bonding (Lone-Pair) Orbitals: Removing an electron from a non-bonding orbital (such as the 1b₁ out-of-plane oxygen 2p_x lone pair of H₂O at 12.62 eV) causes minimal change in equilibrium bond length (Δr_e ≈ 0) or vibrational frequency (ν⁺ ≈ ν₀). The Franck-Condon overlap concentrates almost exclusively in the 0–0 adiabatic transition, producing a single, sharp vertical spike with negligible vibrational tailing.
2. Strongly Bonding Orbitals: Ionizing a bonding orbital (such as the 1π_u orbital of C₂H₄ or N₂) weakens the chemical bond, lengthening the equilibrium bond distance (r_e⁺ > r_e) and reducing the vibrational stretching frequency of the cation (ν⁺ < ν₀). The vertical transition reaches a high vibrational state (v′ > 0), producing a broad, multi-peak progression with reduced vibrational spacing compared to the neutral ground state.
3. Antibonding Orbitals: Ejecting an electron from an antibonding orbital (such as the 1π_g* HOMO of O₂ at 12.07 eV) increases the bond order from 2.0 to 2.5 in O₂⁺, shortening the bond length (r_e⁺ < r_e) and increasing the vibrational spacing of the cation progression (ν⁺ = 1905 cm⁻¹ in O₂⁺ vs. ν₀ = 1580 cm⁻¹ in neutral O₂).

Resolving the 2a₁ and 1t₂ Bands of CH₄

A classic pedagogical benchmark in quantum chemistry is the valence photoelectron spectrum of methane (CH₄, tetrahedral point group T_d). In a naive localized orbital hybridization model, carbon forms four equivalent sp³ hybrid orbitals that overlap with four hydrogen 1s orbitals to create four identical C–H σ bonds, which might superficially suggest a single valence ionization energy for all eight valence electrons.

In contrast, canonical symmetry-adapted molecular orbital theory classifies the four hydrogen 1s orbitals under T_d symmetry into a totally symmetric 2a₁ combination (overlapping with carbon 2s) and a triply degenerate 1t₂ set (overlapping with carbon 2p_x, 2p_y, and 2p_z), yielding the ground-state valence configuration (2a₁)²(1t₂)⁶.

Experimental He II (40.81 eV) and X-ray photoelectron spectra of gaseous CH₄ confirm the canonical MO prediction: methane exhibits two distinct valence ionization bands separated by more than 10 eV, alongside the core C 1s (1a₁) peak at 290.8 eV:

• Band 1 (1t₂ HOMO): A broad, Jahn-Teller split band centered near 12.6–14.4 eV corresponding to ionization of the six electrons in the triply degenerate 1t₂ orbital. Removal of one electron produces a ²T₂ CH₄^+• cation that undergoes a Jahn-Teller effect distortion from T_d to C_2v or D_2d symmetry, splitting the band into a double-humped progression.
• Band 2 (2a₁ Inner Valence): A single symmetric progression at 22.9–23.0 eV corresponding to ionization of the two electrons in the non-degenerate carbon 2s-derived 2a₁ orbital. After correcting for photoionization cross sections, the integrated areas of the 1t₂ and 2a₁ bands approach the 3:1 degeneracy ratio predicted by canonical group theory. Note that a unitary transformation of the canonical (2a₁)²(1t₂)⁶ Slater determinant yields four localized sp³ bond orbitals while leaving the total N-electron wave function and overall electron density invariant; however, because those four localized bond orbitals have non-zero off-axis exchange integrals ⟨σ_i | F̂ | σ_j⟩ ≠ 0 with one another, only the diagonalized canonical molecular orbitals correspond to stationary states of the cation observed in photoelectron spectroscopy.

Why does methane show two valence ionization peaks in PES if all four C–H bonds are equivalent?

Although the total ground-state electron density of CH₄ can be represented equally well by four localized sp³ bonds or by canonical delocalized molecular orbitals, photoionization produces an (N − 1)-electron stationary state of the CH₄^+• cation. Because the four localized C–H bond orbitals interact strongly through off-diagonal Fock matrix elements, eigenstates of the ionized cation correspond to the symmetry-adapted canonical orbitals: the triply degenerate 1t₂ set (ionized near 12.6–14.4 eV) and the non-degenerate 2a₁ orbital (ionized at 23.0 eV).

How can you distinguish a bonding orbital from an antibonding orbital in a UPS spectrum?

Both bonding and antibonding orbitals produce broad Franck-Condon vibrational progressions because removing an electron alters the equilibrium bond length of the cation. However, ionizing a bonding orbital weakens the bond and decreases the vibrational spacing ν⁺ relative to the neutral ground-state frequency ν₀, whereas ionizing an antibonding orbital strengthens the bond and increases the vibrational spacing (ν⁺ > ν₀).

Why does X-ray photoelectron spectroscopy (XPS) detect oxidation states if core electrons do not form chemical bonds?

Electronegative ligands withdraw valence electron density from an atom, reducing the electrostatic shielding experienced by its inner-shell core electrons. This net positive electrostatic potential deepens the core orbital eigenvalues, shifting the measured XPS binding energy (such as C 1s in CF₄ at 293.0 eV vs. CH₄ at 290.8 eV) toward higher energy in direct proportion to the formal oxidation state and ligand electronegativity.

1. Turner, D. W., Baker, C., Baker, A. D., & Brundle, C. R. (1970). Molecular Photoelectron Spectroscopy: A Handbook of He 584 Å Spectra. Wiley-Interscience. https://doi.org/10.1002/bbpc.19710750619
2. Koopmans, T. (1934). "Über die Zuordnung von Wellenfunktionen und Eigenwerten zu den Einzelnen Elektronen Eines Atoms." Physica, 1(1–6), 104–113. https://doi.org/10.1016/S0031-8914(34)90011-290011-2)
3. Brundle, C. R., Robin, M. B., & Basch, H. (1970). "Electronic energies and electronic structures of the fluoromethanes." The Journal of Chemical Physics, 53(6), 2196–2213. https://doi.org/10.1063/1.1674313
4. Hüfner, S. (2003). Photoelectron Spectroscopy: Principles and Applications (3rd ed.). Springer. https://doi.org/10.1007/978-3-662-09280-4