analytical data from 2D NMR spectroscopy [10]. Cross peaks are very sensitive to
molecular structure and geometry and are, therefore, very useful to determine
distances and angles between transition dipoles [10]. However, it is important to
specify on a case-by-case basis what one refers to as an interaction, since there are
multiple possibilities from a molecular point of view how cross peaks can be
generated. Distinguishing these cases is important, since the origin is very
meaningful for a correct interpretation of the spectra and the dynamics. The most
important contributions are based on coupled oscillators, which can give rise to a
spatially delocalized excitation in the molecule (a vibrational exciton, in analogy to
the well-known Frenkel exciton). Vibrational coupling may not only be restricted to
intramolecular cases, since functional groups of different molecules can be coupled
as well, e.g. in closely packed aggregates. Alternatively, even in the case of largely
localized excitations and weak coupling, again both in intra- and intermolecular
cases, the excitation energy can be transferred between different oscillators, when a
donor and an acceptor are spatially close enough and exhibit a proper orientation
with respect to each other. It is, therefore, well established that the intensity and the
shape of the cross peaks reflects the strength of the coupling between two functional
groups, and this directly relates to molecular structure (Sect. 3.1.1). Strong coupling
and energy transfer are not the exclusive mechanisms by which cross peaks can be
generated in a 2D IR spectrum. Moreover, once an ultrashort laser pulse excites a
molecular vibration, the molecule can undergo a chemical reaction within the
lifetime of the vibrational excitation, thereby possibly influencing the vibrational
frequency of the initially excited bond. In addition, changes in the molecular
properties can give rise to cross peaks and the whole concept is generally referred to
as ‘‘chemical exchange’’. There are even more possibilities for cross peaks to appear
in a 2D IR spectrum, e.g. Fermi resonances [10, 73]. These cases are rather special
and are not further discussed here.
In a more detailed physical description, there are several ways in which two
coupled oscillators can be described, and these have been outlined in detail in Ref.
[10]. The distinction is made based on the coupling mechanism (e.g. electrostatic,
electrodynamic, or mechanical), as well as their relative contributions, since
multiple mechanisms can be active at the same time. Mechanical coupling exists in
many molecules, is a very efficient mechanism for energy delocalization in a
molecule, is generally interpreted with the analogy of interacting springs (‘‘throughbond’’) and has been well characterized with 2D IR spectroscopy [1, 10, 74, 75]. In
contrast, ‘‘through space’’ electrostatic coupling between two oscillators does not
require chemical connectivity, but can be very strong for molecules with very large
transition dipole moments. However, it is short-ranged due to the strongly nonlinear dependence of the coupling terms on the inter-group distance (Sect. 3.1.1)
and often only the nearest neighbors need to be considered. In addition to this, the
representation of oscillators as dipoles is often a rough approximation. Higher-order
multipoles can contribute to the coupling as well. Moreover, orbitals are often
strongly delocalized over an entire molecule. This establishes a charge density
distribution that responds to the vibration of certain oscillators. Such charge-flow
effects are not included in the simple picture of transition dipole coupling and thus
impose limitations to the applicability. However, simple transition dipole coupling,
Top Curr Chem (Z) (2017) 375:86
123
125
Reprinted from the journal
molecular structure and geometry and are, therefore, very useful to determine
distances and angles between transition dipoles [10]. However, it is important to
specify on a case-by-case basis what one refers to as an interaction, since there are
multiple possibilities from a molecular point of view how cross peaks can be
generated. Distinguishing these cases is important, since the origin is very
meaningful for a correct interpretation of the spectra and the dynamics. The most
important contributions are based on coupled oscillators, which can give rise to a
spatially delocalized excitation in the molecule (a vibrational exciton, in analogy to
the well-known Frenkel exciton). Vibrational coupling may not only be restricted to
intramolecular cases, since functional groups of different molecules can be coupled
as well, e.g. in closely packed aggregates. Alternatively, even in the case of largely
localized excitations and weak coupling, again both in intra- and intermolecular
cases, the excitation energy can be transferred between different oscillators, when a
donor and an acceptor are spatially close enough and exhibit a proper orientation
with respect to each other. It is, therefore, well established that the intensity and the
shape of the cross peaks reflects the strength of the coupling between two functional
groups, and this directly relates to molecular structure (Sect. 3.1.1). Strong coupling
and energy transfer are not the exclusive mechanisms by which cross peaks can be
generated in a 2D IR spectrum. Moreover, once an ultrashort laser pulse excites a
molecular vibration, the molecule can undergo a chemical reaction within the
lifetime of the vibrational excitation, thereby possibly influencing the vibrational
frequency of the initially excited bond. In addition, changes in the molecular
properties can give rise to cross peaks and the whole concept is generally referred to
as ‘‘chemical exchange’’. There are even more possibilities for cross peaks to appear
in a 2D IR spectrum, e.g. Fermi resonances [10, 73]. These cases are rather special
and are not further discussed here.
In a more detailed physical description, there are several ways in which two
coupled oscillators can be described, and these have been outlined in detail in Ref.
[10]. The distinction is made based on the coupling mechanism (e.g. electrostatic,
electrodynamic, or mechanical), as well as their relative contributions, since
multiple mechanisms can be active at the same time. Mechanical coupling exists in
many molecules, is a very efficient mechanism for energy delocalization in a
molecule, is generally interpreted with the analogy of interacting springs (‘‘throughbond’’) and has been well characterized with 2D IR spectroscopy [1, 10, 74, 75]. In
contrast, ‘‘through space’’ electrostatic coupling between two oscillators does not
require chemical connectivity, but can be very strong for molecules with very large
transition dipole moments. However, it is short-ranged due to the strongly nonlinear dependence of the coupling terms on the inter-group distance (Sect. 3.1.1)
and often only the nearest neighbors need to be considered. In addition to this, the
representation of oscillators as dipoles is often a rough approximation. Higher-order
multipoles can contribute to the coupling as well. Moreover, orbitals are often
strongly delocalized over an entire molecule. This establishes a charge density
distribution that responds to the vibration of certain oscillators. Such charge-flow
effects are not included in the simple picture of transition dipole coupling and thus
impose limitations to the applicability. However, simple transition dipole coupling,
Top Curr Chem (Z) (2017) 375:86
123
125
Reprinted from the journal