Photoelectron spectroscopy is based on Einstein's
photoelectric effect.1 A photon can ionize an electron from a
molecule if the photon has an energy greater than the energy
holding the electron in the molecule. Any photon energy in
excess of that needed for ionization is carried by the outgoing
electron in the form of kinetic energy.
The kinetic energy of the ejected (photo)electron is measured
in the photoelectron spectrometer. As shown in equation (1), the
difference between the photon energy (hν), which is known, and
the electron kinetic energy (K.E.), which is measured, is the
ionization energy (I.E.).
For a neutral molecule in the gas phase, the photon ionizes
the molecule (M) and leaves the molecule in a positively charged
ion state (Mi+) according to equation (2).
Rearranging equation (2) and considering the energy
components of the process as in equation (3) shows that a
measure of the ionization energy is also a measure of the
difference in energy between the positive ion state Mi+ and the
initial state of the molecule M.
Thus, photoelectron spectroscopy measures the relative
energies of the ground and excited positive ion states that are
obtained by removal of single electrons from the neutral
molecule.
At the most fundamental level ionization energies are
well-defined thermodynamic quantities related to the heats of
protonation, oxidation/reduction chemistry, and ionic and
covalent bond energies. Ionization energies are closely related
to the concepts of electronegativity, electron-richness, and the
general reactivity of molecules. The energies and other
characteristic features of the ionization bands observed in
photoelectron spectroscopy provide some of the most detailed and
specific quantitative information regarding the electronic
structure and bonding in molecules. Photoelectron spectroscopy
has served as a particularly important basis for the bonding
models used to describe organic, inorganic, and organometallic
molecules because the energetics of ion formation from the
neutral ground state are directly related to orbital electron
configurations, oxidation states, charge distributions, and
covalency.
The Orbital Model of Ionization
Ionization is explicitly defined in terms of transitions between
the ground state of a molecule and ion states as shown in
equation (2) and as illustrated in the above figure.
Nonetheless, the information obtained from photoelectron
spectroscopy is typically discussed in terms of the electronic
structure and bonding in the ground states of neutral molecules,
with ionization of electrons occurring from bonding molecular
orbitals, lone pairs, antibonding molecular orbitals, or atomic
cores. These descriptions reflect the relationship of ionization
energies to the molecular orbital model of electronic structure.
Ionization energies are directly related to the energies of
molecular orbitals by Koopmans' theorem,2 which states that
the negative of the eigenvalue of an occupied orbital from a
Hartree-Fock calculation is equal to the vertical ionization
energy to the ion state formed by removal of an electron from
that orbital, provided the distributions of the remaining
electrons do not change. There are many limitations to Koopmans'
theorem, but in a first order approximation each ionization of a
molecule can be considered as removal of an electron from an
individual orbital. The ionization energies can then be
considered as measures of orbital stabilities, and shifts can be
interpreted in terms of orbital stabilizations or
destabilizations due to electron distributions and bonding. Koopmans' theorem is implicated whenever an orbital picture is
involved, but is not necessary when the focus is on the total
electronic states of the positive ions.
The direct ionization transition is not restricted by any
symmetry selection rules because the ejected electron can carry
any necessary angular momentum to make the process electric
dipole allowed. Therefore, ionization to any excited positive
ion state obtained by removal of a single electron and within
range of the photon energy can be observed. That is, ionizations
can be observed that correspond to removal of electrons from any
of the occupied orbitals. Photons with energies in the keV
(X-ray) range are able to ionize down to the core electrons of
atoms and molecules. Core ionizations of molecules are
associated with the individual atoms present and fall in
characteristic energy ranges related to the specific elements.
The exact ionization energy of a core level is influenced by the
charge potential around the atom and is useful for
distinguishing between atoms in different chemical environments.
Information Content in Photoelectron Spectroscopy
While the energy information that is contained in a
photoelectron spectrum is surely the most important information,
further insight into the electronic structure of molecules is
contained in other information within the spectrum. This
information includes:
Ionization Band Shape: Removal of an electron changes the
electronic structure and bonding in the molecule and results in
a shift of the equilibrium internuclear separations. If the
geometry changes are sufficiently great, the most probable
(vertical) transitions from the neutral ground state are to
excited vibrational levels of the final ion state. When these
vibrational levels are resolved, the change in vibrational
spacing from the initial state to the final state gives a
measure of the change in vibrational frequencies and force
constants associated with the excitation, and the intensity
pattern of the transitions to the excited vibrational levels
(the Franck-Condon factors) gives a measure of the change in
equilibrium bond distances.
Ionization Intensity: The probability of ionization is
dependent upon the energy of the ionizing photon, the atomic
character of the orbitals ionized, the angle at which
photoelectrons are collected, and the polarized nature of the
photons. These probabilities of photoionization are understood
in terms of photoionization cross-sections. The dependence of
photoionization cross-sections upon the energy of the ionization
photon and the character of the molecular orbital has been
particularly useful in the assignments of the spectra of
multi-atomic molecules. The general trends that have been
observed for photoionization behavior often lead to a clear and
convenient differentiation between ionization from metal-based
and ligand-based orbitals in the photoelectron spectra of
inorganic and organometallic molecules.
References
1. Einstein, A. "On a Heuristic Point of View Concerning the
Production and Transformation of Light." Ann. Phys.1905, 17,
549.
2. Koopmans, T. "Ordering of Wave Functions and Eigenvalues
to the Individual Electrons of an Atom." Physica1933,
1, 104.