1 3
Top Curr Chem (Z) (2018) 376:24
3.2 Coupling Accurate Electronic Structure Computations to 2DES
Two-dimensional electronic spectroscopy targets the manifold of (localized and
delocalized) excited states of the chromophoric units in the sample. A thorough
characterization of the electronic structure of the chromophoric systems is thus
required to properly interpret 2D electronic spectra, which generally represents a
great challenge even for the most advanced computational techniques, despite a
plethora of quantum mechanics (QM) methods currently available. In particular,
the QM treatment of isolated large chromophoric moieties (usually absorbing in
the Vis) or rather small UV-chromophores that still involve high-energy electronic
levels and thus encompass a large number of excited states in the target manifolds,
might already push the limits of what is currently computationally feasible. Treating multiple interacting chromophores, i.e. molecular aggregates, at the ab initio
level then quickly becomes prohibitive with increasing number or size of chromophoric units. At the same time, as shown in Eqs. 12 and 13, TEs and TDMs are the
fundamental ingredients for simulating the third-order nonlinear response recorded
in 2DES maps, and even if largely approximate, their estimations cannot be circumvented. A common strategy for coping with these limitations is to adopt Frenkel
exciton models that make simulations of 2DES spectra of realistic model systems
computationally feasible [36, 48]. Exciton modeling works well for vibrational
excited-state absorptions in 2DIR spectroscopy (namely overtone absorptions),
since anharmonicity for overtones of a chromophore (local overtones) or between
coupled chromophores can be reliably computed perturbatively (without explicit
knowledge of the high-lying energy level manifold). In general, such a perturbative treatment is less effective for electronic transitions, since local excited states
behave quite differently with respect to local overtones, showing large anharmonic
couplings that cannot be described perturbatively. Still, when dealing with coupled
electronic excitations between interacting chromophores, the TEs and TDMs of
each isolated chromophore can be initially calculated at the QM level (usually in
the gas phase and for a given geometry) and then employed as parameters to build
the exciton Hamiltonian. The electronic couplings in multi-chromophoric systems
can then be estimated, while neglecting electron exchange between chromophores.
Clearly, the limited description provided by Frenkel exciton modeling has computational advantages that become evident only when just a few excited states (i.e.
energy levels) are considered in the model Hamiltonian (i.e. few QM computations are initially performed). This implies that many of the excited states related
to (potential) ESA signals in 2DES maps are often neglected in most conventional
spectroscopy simulation protocols. This assumption generally holds if ESA signals
are expected to fall outside the spectral window of interest. However, broadband
transient absorption [49] and, more recently, also 2D [16, 50, 51] spectra show that
this assumption breaks down regularly, especially in UV-active bio-chromophores.
In fact, as we will show in the following sections, ESAs are ubiquitous, systemdependent, state-specific spectroscopic fingerprints that, for instance, make it possible to selectively study the excited-state dynamics of different decay channels. In
cases such as charge-transfer states, ESAs might represent the only spectroscopic
signature of a given state, being of particular interest for studying its photophysics.
73
Reprinted from the journal
Top Curr Chem (Z) (2018) 376:24
3.2 Coupling Accurate Electronic Structure Computations to 2DES
Two-dimensional electronic spectroscopy targets the manifold of (localized and
delocalized) excited states of the chromophoric units in the sample. A thorough
characterization of the electronic structure of the chromophoric systems is thus
required to properly interpret 2D electronic spectra, which generally represents a
great challenge even for the most advanced computational techniques, despite a
plethora of quantum mechanics (QM) methods currently available. In particular,
the QM treatment of isolated large chromophoric moieties (usually absorbing in
the Vis) or rather small UV-chromophores that still involve high-energy electronic
levels and thus encompass a large number of excited states in the target manifolds,
might already push the limits of what is currently computationally feasible. Treating multiple interacting chromophores, i.e. molecular aggregates, at the ab initio
level then quickly becomes prohibitive with increasing number or size of chromophoric units. At the same time, as shown in Eqs. 12 and 13, TEs and TDMs are the
fundamental ingredients for simulating the third-order nonlinear response recorded
in 2DES maps, and even if largely approximate, their estimations cannot be circumvented. A common strategy for coping with these limitations is to adopt Frenkel
exciton models that make simulations of 2DES spectra of realistic model systems
computationally feasible [36, 48]. Exciton modeling works well for vibrational
excited-state absorptions in 2DIR spectroscopy (namely overtone absorptions),
since anharmonicity for overtones of a chromophore (local overtones) or between
coupled chromophores can be reliably computed perturbatively (without explicit
knowledge of the high-lying energy level manifold). In general, such a perturbative treatment is less effective for electronic transitions, since local excited states
behave quite differently with respect to local overtones, showing large anharmonic
couplings that cannot be described perturbatively. Still, when dealing with coupled
electronic excitations between interacting chromophores, the TEs and TDMs of
each isolated chromophore can be initially calculated at the QM level (usually in
the gas phase and for a given geometry) and then employed as parameters to build
the exciton Hamiltonian. The electronic couplings in multi-chromophoric systems
can then be estimated, while neglecting electron exchange between chromophores.
Clearly, the limited description provided by Frenkel exciton modeling has computational advantages that become evident only when just a few excited states (i.e.
energy levels) are considered in the model Hamiltonian (i.e. few QM computations are initially performed). This implies that many of the excited states related
to (potential) ESA signals in 2DES maps are often neglected in most conventional
spectroscopy simulation protocols. This assumption generally holds if ESA signals
are expected to fall outside the spectral window of interest. However, broadband
transient absorption [49] and, more recently, also 2D [16, 50, 51] spectra show that
this assumption breaks down regularly, especially in UV-active bio-chromophores.
In fact, as we will show in the following sections, ESAs are ubiquitous, systemdependent, state-specific spectroscopic fingerprints that, for instance, make it possible to selectively study the excited-state dynamics of different decay channels. In
cases such as charge-transfer states, ESAs might represent the only spectroscopic
signature of a given state, being of particular interest for studying its photophysics.
73
Reprinted from the journal