the 2D IR data based on the response function formalism for both systems
((f) - (h) and (n) - (p)), which matched the experimental data almost perfectly. A
simple Kubo-picture [62] was sufficient to achieve close matching between
experimental and theoretical results, with only one component in the case of NMA,
but two components in the case of trialanine. Concomitantly performed MD
simulations indicated that the spectral diffusion in NMA is nearly entirely caused by
hydrogen bonding to the surrounding water and that the same timescale is observed
for trialanine. This showed that the long-time difference in the elongation of the 2D
IR spectra must originate from intramolecular structural fluctuations. Further, DFT
calculations on the sample systems indicated that there exist two distinct
conformations for trialanine, which differ in their dihedral angles of the central
peptide bond (/, w), and which interconvert on timescales much longer ([ [4 ps)
than the vibrational lifetime of the amide-I mode. Combining the results from the
experiment with the insights from MD and DFT calculations, a detailed
understanding of the structural dynamics alongside the intermolecular hydrogen
bonding dynamics of small peptides could be obtained and reliably traced back to
the spectroscopic observables that are directly extractable from 2D IR data. Since
then, that approach has been proven as very valuable regarding the general ultrafast
structural elucidation of molecules in the condensed phase [10].
For a broad range of samples, the observation of spectral diffusion has led to a
detailed interpretation of ultrafast molecular dynamics such as solvent-solute
interactions, hydrogen bonding excitation energy transfer, or molecular reorientation [105–108]. Spectral diffusion has been observed for even molecular samples as
large
as
proteins
by
employing
different
types
of
IR-labels
[30, 32, 40, 44, 96, 109, 110]. Such studies can sense the conformational dynamics
of the protein itself as well as the solvation environment, and these are generally
slowed down as compared to the spectral dynamics of isolated IR-labels. Extremely
slow spectral diffusion dynamics have been observed in liquid crystal samples
[111, 112]. In such cases, the dynamics can take place on timescales as slow as
hundreds of picoseconds and are likely to originate from density fluctuations rather
than from conformational re-orientations. In many cases, spectral diffusion can be
even quiet complex and originate from different types of interactions of a molecule
with its environment. Non- or multi-exponential spectral diffusion has been
observed in glass-forming liquids near the glass transition temperature [113], in
proteins [30, 109], hydrogen bond dynamics in liquid water [114], or IR-labels in
ionic liquids [115–117]. The origin of the various timescales is often the existence
of different molecular ensembles (e.g. aggregates), re-orientational and solvationinduced spectral diffusion, or structural spectral diffusion.
Overall, spectral diffusion is an important and readily extractable observable
from 2D IR spectra and its application is significantly broader than solely bulk
solution dynamics. Recently, a series of reports have appeared where 2D IR
spectroscopy has been used in conjunction with non-isotropic environments, e.g.
confined environments such as interfaces, metal-organic-frameworks and nanostructures. Examples from such applications will be considered in Sect. 3.2.
Top Curr Chem (Z) (2017) 375:86
123
136
Reprinted from the journal
((f) - (h) and (n) - (p)), which matched the experimental data almost perfectly. A
simple Kubo-picture [62] was sufficient to achieve close matching between
experimental and theoretical results, with only one component in the case of NMA,
but two components in the case of trialanine. Concomitantly performed MD
simulations indicated that the spectral diffusion in NMA is nearly entirely caused by
hydrogen bonding to the surrounding water and that the same timescale is observed
for trialanine. This showed that the long-time difference in the elongation of the 2D
IR spectra must originate from intramolecular structural fluctuations. Further, DFT
calculations on the sample systems indicated that there exist two distinct
conformations for trialanine, which differ in their dihedral angles of the central
peptide bond (/, w), and which interconvert on timescales much longer ([ [4 ps)
than the vibrational lifetime of the amide-I mode. Combining the results from the
experiment with the insights from MD and DFT calculations, a detailed
understanding of the structural dynamics alongside the intermolecular hydrogen
bonding dynamics of small peptides could be obtained and reliably traced back to
the spectroscopic observables that are directly extractable from 2D IR data. Since
then, that approach has been proven as very valuable regarding the general ultrafast
structural elucidation of molecules in the condensed phase [10].
For a broad range of samples, the observation of spectral diffusion has led to a
detailed interpretation of ultrafast molecular dynamics such as solvent-solute
interactions, hydrogen bonding excitation energy transfer, or molecular reorientation [105–108]. Spectral diffusion has been observed for even molecular samples as
large
as
proteins
by
employing
different
types
of
IR-labels
[30, 32, 40, 44, 96, 109, 110]. Such studies can sense the conformational dynamics
of the protein itself as well as the solvation environment, and these are generally
slowed down as compared to the spectral dynamics of isolated IR-labels. Extremely
slow spectral diffusion dynamics have been observed in liquid crystal samples
[111, 112]. In such cases, the dynamics can take place on timescales as slow as
hundreds of picoseconds and are likely to originate from density fluctuations rather
than from conformational re-orientations. In many cases, spectral diffusion can be
even quiet complex and originate from different types of interactions of a molecule
with its environment. Non- or multi-exponential spectral diffusion has been
observed in glass-forming liquids near the glass transition temperature [113], in
proteins [30, 109], hydrogen bond dynamics in liquid water [114], or IR-labels in
ionic liquids [115–117]. The origin of the various timescales is often the existence
of different molecular ensembles (e.g. aggregates), re-orientational and solvationinduced spectral diffusion, or structural spectral diffusion.
Overall, spectral diffusion is an important and readily extractable observable
from 2D IR spectra and its application is significantly broader than solely bulk
solution dynamics. Recently, a series of reports have appeared where 2D IR
spectroscopy has been used in conjunction with non-isotropic environments, e.g.
confined environments such as interfaces, metal-organic-frameworks and nanostructures. Examples from such applications will be considered in Sect. 3.2.
Top Curr Chem (Z) (2017) 375:86
123
136
Reprinted from the journal