The phenyl vinyl ether–methanol complex: A model system for quantum chemistry benchmarking

Article abstract of DOI:10.3762/bjoc.14.140

The structure of the isolated aggregate of phenyl vinyl ether and methanol is studied by combining a multi-spectroscopic approach and quantum-chemical calculations in order to investigate the delicate interplay of noncovalent interactions. The complementa

Full text of DOI:10.3762/bjoc.14.140

The phenyl vinyl ether–methanol complex: a model system for
quantum chemistry benchmarking
Dominic Bernhard1, Fabian Dietrich1, Mariyam Fatima2,3, Cristóbal Pérez2,3,
Hannes C. Gottschalk4, Axel Wuttke4, Ricardo A. Mata*4, Martin A. Suhm*4,
Melanie Schnell*2,3,5 and Markus Gerhards*1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2018, 14, 1642–1654.
1Fachbereich Chemie & Research Center Optimas, Technische
Universität Kaiserslautern, Erwin-Schrödinger-Str. 52, D-67663
Kaiserslautern, Germany, 2Max Planck Institute for the Structure and
Dynamics of Matter, Luruper Chaussee 149, D-22761 Hamburg,
Germany, 3Deutsches Elektronen Synchrotron (DESY), Notkestrasse
85, D-22607 Hamburg, Germany, 4Institut für Physikalische Chemie,
Georg-August-Universität Göttingen, Tammannstrasse 6, D-37077
Göttingen, Germany and 5Institute of Physical Chemistry,
Christian-Albrechts-Universität zu Kiel, Max-Eyth-Strasse 1, D-24118
Kiel, Germany
doi:10.3762/bjoc.14.140
Received: 16 March 2018
Accepted: 02 June 2018
Published: 02 July 2018
This article is part of the Thematic Series "Dispersion interactions".
Guest Editor: P. Schreiner
© 2018 Bernhard et al.; licensee Beilstein-Institut.
License and terms: see end of document.
Email:
Ricardo A. Mata* - rmata@gwdg.de;
Martin A. Suhm* - msuhm@gwdg.de;
Melanie Schnell* - melanie.schnell@desy.de;
Markus Gerhards* - gerhards@chemie.uni-kl.de
* Corresponding author
Keywords:
dispersion interactions; IR spectroscopy; quantum-chemical
calculations; rotational spectroscopy; structure determination; weak
hydrogen bonds
Abstract
The structure of the isolated aggregate of phenyl vinyl ether and methanol is studied by combining a multi-spectroscopic approach
and quantum-chemical calculations in order to investigate the delicate interplay of noncovalent interactions. The complementary
results of vibrational and rotational spectroscopy applied in molecular beam experiments reveal the preference of a hydrogen bond
of the methanol towards the ether oxygen (OH∙∙∙O) over the π-docking motifs via the phenyl and vinyl moieties, with an additional
less populated OH∙∙∙P(phenyl)-bound isomer detected only by microwave spectroscopy. The correct prediction of the energetic
order of the isomers using quantum-chemical calculations turns out to be challenging and succeeds with a sophisticated local
coupled cluster method. The latter also yields a quantification as well as a visualization of London dispersion, which prove to be
valuable tools for understanding the role of dispersion on the docking preferences. Beyond the structural analysis of the electronic
ground state (S0), the electronically excited (S1) state is analyzed, in which a destabilization of the OH∙∙∙O structure compared to the
S0 state is observed experimentally and theoretically.
1642

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
Introduction
The balance of different noncovalent interactions is crucial for indicating a comparatively weak hydrogen bond, which is also
chemical and biochemical processes as it controls molecular reflected in calculated binding energies [24,25].
recognition and aggregation [1-6]. In order to gain a deeper
understanding of these processes, knowledge on exact struc- With the herein presented work, we now extend our overall
tural arrangements and the respective role of different intermo- multi-spectroscopic study to mixed aromatic olefinic ethers: in
lecular forces such as electrostatic, dispersion and induction the case of phenyl vinyl ether (PVE), there is an ethenyl moiety
forces is needed. Thus, experimental examination as well as the replacing one of the phenyl rings compared to DPE. This intro-
precise prediction of a preferred molecular docking site for dif- duces a localized π system along with the delocalized phenyl π
ferent molecules is of crucial importance. Despite the remark- system as hydrogen bond acceptor sites. Thereby, the complexi-
able progress made in experiments and theory/computational ty of the system is increased, as now three qualitatively differ-
chemistry, there is still a need for improvement and bench- ent basic binding motifs have to be regarded instead of only two
marking [7].
for DPE. This also provides an enhanced challenge for theory,
with no clear preference for one of the motifs to be expected.
Many aromatic solute–solvent complexes have been studied in As shown, e.g., in the case of DPE–t-BuOH [20], there is a need
the gas phase (cf. [8-10] and references therein). Studied for benchmarking systems in order to improve and develop
systems involving methanol as attached solvent molecule better theoretical approaches especially for non-covalently
include the works on benzene–methanol clusters by the Zwier bound complexes. The study on PVE–MeOH is meant to
group [11] and on fluorobenzene–methanol clusters by the present a further benchmark system, probably even more chal-
Brutschy group [12], to mention only two examples. Com- lenging than DPE–t-BuOH.
plexes of aromatic ethers with polar solvent molecules are of
special interest due to the presence of different competing For an experimental elucidation of structural arrangements and
hydrogen bond acceptor sites. An extensive study on diphe- energetic preferences, investigations on a molecular level are re-
noxyethane–water clusters was performed by the Zwier group quired on isolated molecular aggregates, allowing for an ideal
[13-15] including studies in the excited S1 and S2 states. Con- comparison with gas phase calculations. This can be achieved
cerning aggregates of aromatic ethers with alcohols, there is a by molecular beam experiments, which can be combined with a
work of Pietraperzia et al. [16] on the anisole–phenol complex variety of spectroscopic techniques. For our multi-spectroscop-
in which an OH∙∙∙O structure was identified. In a systematic ic studies, we utilize FTIR spectroscopy, mass- and isomer-
study by the Suhm group on complexes of anisole derivatives selective IR/UV techniques (IR/R2PI, for methodical develop-
with methanol, a balance between OH∙∙∙O and OH∙∙∙π structures ments, cf., e.g., [8,26-29] and UV/IR/UV spectroscopy, cf., e.g.,
being very sensitive to the substitution pattern at the anisole [30-43]) and chirped-pulse Fourier transform microwave (CP-
moiety was identified [17,18]. In previous multi-spectroscopic FTMW) spectroscopy. Comparing spectroscopic results with
studies by the Schnell, Suhm and Gerhards groups on diphenyl quantum-chemical calculations is often mandatory for the inter-
ether (DPE)–solvent complexes [19-22], the influence of pretation of experiments. Furthermore, such comparison enables
different attached solvent molecules on the structural prefer- a critical evaluation of the approximations used, comparing the
ence was compared. It could be shown that the balance be- relative stability of different binding motifs.
tween OH∙∙∙π- and OH∙∙∙O-bound structures is very sensitive to
the size of the attached alcohol. Torsional balances in solution In this paper, the first structural investigation on the complex of
have been used to probe aromatic OH∙∙∙π interactions and to phenyl vinyl ether with methanol is presented. An established
show that these interactions remain important at room tempera- multi-spectroscopic approach [19,20] is used, coupling FTIR,
ture [23].
IR/UV and microwave spectroscopy with theoretical treatments
including dispersion-corrected density functional theory (DFT-
In such aromatic solute–solvent systems, one frequently en- D3) [44,45], spin-component-scaled approximated coupled
counters hydrogen bonds formed towards oxygen or nitrogen cluster-singles-doubles (SCS-CC2) [46] as well as explicitly
lone pairs, or R–H∙∙∙π binding motifs (R = O, N, C, S,…) in- correlated local coupled cluster theory (LCCSD(T0)-F12) [47]
volving aromatic π systems. Less often, R–H∙∙∙π bound com- calculations, the latter allowing for a quantification and visuali-
plexes are found involving nonconjugated, localized C=C zation of London dispersion interactions [48]. The aim of the
double bonds. Exceptions include the ethene–methanol com- presented study is the unambiguous experimental identification
plex [24] as well as bulky olefin–tert-butyl alcohol complexes of the preferred binding site of a first methanol solvent mole-
[25] investigated by jet FTIR spectroscopy. The observed OH cule to the multivalent hydrogen bond scaffold of phenyl vinyl
stretching red-shifts compared to the free alcohols are small, ether, followed by a classification of theoretical methods in
1643

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
terms of success or failure to predict this preference. Visualiza- radiation in the range of 3520–3750 cm−1 is generated by
tion of possible reasons for the subtle preference is a valuable difference frequency mixing (DFM) in a LiNbO3 crystal using
additional asset.
the fundamental (1064 nm) of a seeded Nd:YAG laser (Quanta-
Ray Pro-230, Spectra-Physics) and the output of a further dye
laser (PrecisionScan, Sirah), which is pumped by the second
harmonic (532 nm) of the same Nd:YAG laser. Amplification
Experimental Setup
FTIR setup
For the FTIR experiments, the so-called filet-jet setup, as de- of the resulting IR radiation is obtained by an optical para-
scribed in detail in [49], was used. In this setup, the scans of a metric amplification (OPA) process in a further LiNbO3 crystal
Bruker IFS 66 v/s spectrometer (80 kHz, resolution 2 cm−1) are using the output of the DFM process and the fundamental
synchronized to a pulsed supersonic expansion through a (1064 nm) of the Nd:YAG laser.
600 × 0.2 mm2 slit nozzle. Using two separate cooled satura-
tors, low concentrations (<0.1%) of PVE (Sigma-Aldrich, 97%, For the IR/R2PI spectra, the IR laser was fired 50 ns prior to the
used as purchased) and methanol (Sigma-Aldrich, ≥99.8%, used UV excitation laser, whereas for the UV/IR/UV spectra the IR
as purchased) were added to the carrier gas helium (Linde, laser was fired 2.0–3.0 ns after the UV excitation laser. The
99.996%) and premixed at a pressure of 0.75 bar in a 67 L time delay between UV excitation and ionizing laser was
reservoir before being expanded through the slit nozzle. The 4.0–4.5 ns.
pulsed operation with waiting times of 30–90 s between 150 ms
long pulses combined with a buffer volume of 12–23 m3 and a CP-FTMW setup
pumping capacity of 500–2500 m3/h resulted in background The rotational spectroscopy measurements were performed with
pressures of less than 0.1 mbar before expansions. This facili- the Hamburg chirped-pulse Fourier transform microwave (CP-
tated measurements of clusters of methanol and PVE in the FTMW) spectrometer COMPACT covering the 2–8 GHz fre-
zone of silence of the expansion at an average distance of quency range, which has been described in detail in [52]. The
10 mm to the nozzle. A calcium fluoride beam splitter, lenses molecules were seeded into a supersonic expansion with neon
and windows were used in combination with a 150 W tungsten as the carrier gas by using a pulse nozzle (Parker General
filament and an optical filter (4200–2450 cm−1) to maximize Valve, Series 9, 0.9 mm diameter orifice) equipped with a heat-
the signal-to-noise ratio in the OH stretching range of the vibra- able reservoir close to the valve orifice, operating at 8 Hz. PVE
tional spectra. For the final spectra, 150 to 775 pulses were was synthesized as described above and used without further
co-added to further improve signal-to-noise.
purification.
IR/UV setup
The liquid sample was held in the reservoir at room tempera-
The experimental setup for the IR/UV experiments is described ture, which resulted in sufficient vapor pressure (standard
in detail elsewhere [29,50], thus only a brief description is given boiling point of about 155 °C) for recording the rotational spec-
here. All experiments were carried out in a molecular beam trum. MeOH was kept in a separate reservoir. PVE–MeOH
apparatus consisting of a differentially pumped linear time-of- clusters were generated by first flowing the carrier gas (neon)
flight (TOF) mass spectrometer with a pulsed valve (Series 9 through the reservoir containing methanol that was external to
and pulse driver Iota One, General Valve, 500 µm orifice) for the chamber, followed by picking up PVE vapor. After super-
skimmed jet expansion. PVE was synthesized according to the sonic expansion into vacuum using neon at 3 bar, the molecular
procedure reported in [51] (cf. Supporting Information File 1 jet was polarized with a 4 µs chirp spanning 2–8 GHz. The
for details). MeOH (Sigma-Aldrich, ≥99.7%) and PVE were chirp was generated with an arbitrary waveform generator,
both supplied via separate cooled reservoirs (approx. −8 °C and amplified to 300 W with a traveling wave tube amplifier, and
−13 °C, respectively) and co-expanded with the carrier gas neon transmitted into the vacuum chamber via a horn antenna.
(2.5–3.0 bar).
Following excitation, 40 µs of the free induction decay (FID) of
the macroscopic ensemble of polarized molecules was recorded.
For the one- and two-color R2PI, the IR/R2PI and the UV/IR/ The fast frame capability [53] of the Tektronix DPO 71254C
UV experiments up to three tunable nanosecond laser systems was used in which eight consecutive excitation chirps, each fol-
were necessary, including two independent UV laser systems lowed by 40 µs during which the FID could be collected, were
and one IR laser system. The UV laser radiation is obtained via recorded and averaged. This resulted in an effective repetition
second harmonic generation in a BBO crystal using the output rate of 64 Hz.
of a dye laser (Cobra-Stretch and PrecisionScan, Sirah). They
are pumped by the second harmonic (532 nm) of a Nd:YAG For the spectrum of the PVE–MeOH dimer, 3 million FIDs
laser (SpitLight 600 and SpitLight 1000, Innolas). The IR laser were co-added. A resulting signal-to-noise ratio of about 500:1
1644

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
to 600:1 for the stronger transitions of the dominant complex method will be denoted as LCCSD(T0)-F12/CBS[T:Q]. In all
allowed us to determine the positions of the carbon atoms with correlated calculations the 1s electrons were removed from the
respect to the center of mass of the overall complex (see below) treatment (frozen-core approximation). Furthermore, we
exploiting the presence of 13C isotopologues in natural abun- analyzed the relative impact of dispersion interactions in the
dance and using the Kraitchman approach [54]. Fourier trans- different complexes through a local orbital analysis of the
formation of the averaged time domain FID, recorded at point CCSD (connected) doubles energy terms. The latter discussion
spacings of 10 ps, resulted in a frequency domain rotational is complemented with dispersion interaction density (DID) plots
spectrum with frequency resolution of 25 kHz.
[48]. The coupled cluster calculations were carried out with
Molpro 2015.1 [66].
The assignment was performed with the program JB95 [55],
then the fits to an asymmetric-rotor Hamiltonian were per-
Results and Discussion
formed using SPFIT/SPCAT. The experimental results were Theoretical results
complemented by and compared with the results of electronic In contrast to the already studied diphenyl ether–alcohol clus-
structure calculations. B3LYP-D3/aug-cc-pVTZ calculations ters [19,20,22], phenyl vinyl ether offers three different binding
were performed using the Gaussian 09, rev. D.01 program suite sites for possible interactions with small solvent molecules: the
[56] to guide the assignment.
ether oxygen, the phenyl ring and the vinyl moiety. Since both
the phenyl ring and the vinyl moiety interact with the solvent
via a π cloud, preferred binding sites are indicated using the
Computational Methods
Various input structures for the PVE–MeOH complex were following nomenclature: P (phenyl) and E (ethenyl), respective-
generated by using the MMFF94s force field [57] as imple- ly. The optimizations using B3LYP-D3(BJ)/def2-TZVP yield
mented in Avogadro [58]. Afterwards, geometry optimizations six different structures, representing each binding motif with
were performed by applying the Berny optimization algorithm two isomers (cf. Figure 1).
of Gaussian 09 [56] with energies and gradients obtained
from Turbomole 7.0 [59]. The DFT functional B3LYP with In order to verify the structures of the complexes, a second level
Grimme's two-body D3 corrections and Becke–Johnson damp- of theory was applied, namely SCS-CC2/def2-TZVP. Similar
ing [45] was used in combination with the def2-TZVP basis set minima were found in the latter calculations, confirming the
based on the documented performance of this level of theory for rich variety of binding motifs. However, distinct differences
the similar diphenyl ether–methanol system [19]. Furthermore, were found between the two methods: While B3LYP-D3 pre-
the obtained structures were re-optimized with the SCS-CC2 dicted the OH–E conformer as the most stable complex SCS-
method using the def2-TZVP basis set, both in the electronic CC2 gave OH–P as the lowest minimum (cf. Table S1 in Sup-
ground (S0) and first excited state (S1). The ricc2 module in porting Information File 1). This is in contrast to our results of
Turbomole 7.0 requires an auxiliary Coulomb fitting basis set diphenyl ether–alcohol clusters [19,20,22], where both compu-
(cbas) for the resolution-of-identity approximation (RI) for tational levels predicted the same energetic order of the isomers.
which def2-TZVP-cbas was chosen [60]. All obtained geome- The two structures correspond to quite different docking posi-
tries were confirmed as minima by harmonic frequency calcula- tions, reflecting well the demanding test this system imposes on
tions.
quantum chemical methods. Several minima are separated by
energy differences of 1 kJ/mol or less. The complete energetic
In order to evaluate the relative stability of the different analysis at both levels of theory is presented in Supporting
conformers found on the potential hypersurface, density fitted Information File 1 (cf. Tables S1 and S2). The reasons behind
explicitly correlated local coupled cluster with singles and the discrepancies are manifold, ranging from the method to the
doubles excitations and perturbative triples (DF-LCCSD(T0)- small basis set used. In order to obtain a more reliable theoreti-
F12) calculations were carried out [47]. In order to converge the cal prediction, LCCSD(T0)-F12/CBS[T:Q] calculations were
energies relative to the one particle basis, the VTZ-F12 and carried out on top of the DFT-optimized geometries. The results
VQZ-F12 basis sets [61,62] were used together with a are presented in Table 1, with and without zero-point vibra-
Schwenke style basis set extrapolation, as proposed in [63]. The tional energy (ZPVE) corrections.
orbitals were Pipek–Mezey [64] localized and orbital domains
determined by natural population analysis with a threshold of The coupled cluster results show a clear energetic preference
TNPA = 0.03 [65]. Defaults were used for the pair classifica- for the OH–O and OH–O’ isomers. Observing the intermolecu-
tion, with all pairs included in the F12 treatment. Furthermore, lar contacts, which may or may not be designated as weak
the intermolecular pairs were classified as strong (meaning that hydrogen bonds but are expected to stabilize the complexes, the
they were treated at the highest level of theory). The latter main difference between the two structures is a phenyl vs
1645

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
Figure 1: Minimum structures of the most stable PVE–MeOH dimers obtained at the B3LYP-D3(BJ)/def2-TZVP level; dashed colored lines indicate
the different primary docking motifs, dashed gray lines illustrate secondary CH–O contacts; values in parentheses correspond to the relative, zero-
point-corrected energies E0,rel with respect to the OH–O isomer, calculated at the LCCSD(T0)-F12/CBS[T:Q]//B3LYP-D3/def2-TZVP level of theory
(cf. Table 1).
ethenyl CH to methanol O contact (cf. dashed gray lines in Also featured in Table 1 are the computed O–H stretch funda-
Figure 1). Both are separated by only a few tenths of a mentals together with the IR intensity at the two different levels
kJ/mol, which is within the error of the method used of theory used in the optimizations. The frequencies were scaled
(considering that the coupled cluster expansion is truncated according to the experimental value of the OH–π isomer of
at triples excitations and the neglect of core-valence correlation DPE–MeOH [19]. Based on the computational results, the
effects, which should be the largest sources of error along vibrational spectral signals of the OH–O and OH–O’ isomers
with the harmonic B3LYP ZPVE error). It also confirmed will be extremely hard to distinguish, as they lie less than
the subtle difference between the six conformers, with an 3 cm−1 apart, with very similar intensities. The same can be
energy span of approximately 4–5 kJ/mol (≈1 kcal/mol, the asserted for the less stable OH–E and OH–E’ structures. This is
commonly accepted definition of chemical accuracy) among all not surprising, given the similarities of the OH binding pattern
structures.
for both sets of structures.
Table 1: Comparison of different structures for PVE–MeOH dimers in the S0 state with LCCSD(T0)-F12/CBS[T:Q]//B3LYP-D3/def2-TZVP electronic
energies Erel and B3LYP zero-point corrected energies E0,rel relative to the minimum OH–O structure. The scaled wavenumbers of the
OH-stretching vibration together with the respective IR intensity I are presented for two levels of theory: B3LYP-D3 (scaling factor: 0.9600) and SCS-
CC2/def2-TZVP (scaling factor: 0.9635).
B3LYP-D3
I [km/mol]
SCS-CC2
I [km/mol]
Erel [kJ/mol]
E0,rel [kJ/mol]
[cm−1]
3597
[cm−1]
3619
OH–O
OH–O’
OH–P
OH–P’
OH–E
OH–E’
0.0
−0.3
1.4
3.9
1.5
4.8
0.0
0.2
1.0
2.4
2.0
4.7
219
193
112
127
187
197
160
144
67
3600
3619
3631
3567
3567
3621
3631
3636
3607
3606
110
121
128
1646

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
In order to gain further insight into the energetic order of the sion, with a maximum energy difference of 2.6 kJ/mol when
different isomers, we conducted an analysis of the dispersion summed all together, the relative weight of the different molec-
interactions present in the system by decomposing the CCSD ular fragments varies quite significantly. The moiety with the
energy terms obtained with the largest basis set (VQZ-F12). largest potential as dispersion energy donor (DED) is the phe-
The latter procedure is based on the classification of the inter- nyl ring. This results in the strongest stabilization for the two
molecular excitation classes as detailed in [48,67]. The results conformers whereby the methanol is closest to the ring (OH–P
are shown in Table 2. Beyond the total dispersion contributions, and OH–P’). The other conformers have much more spread out
we also made use of the local analysis to separate the contribu- contributions. What is surprising is that even for the ethenyl
tion of different molecular moieties in the PVE molecule (phe- binding complexes the contribution of the phenyl ring is size-
nyl, ether oxygen and ethenyl). Shared orbitals are split up ac- able. Geometrically, this seems unlikely, given that the metha-
cording to their NPA (natural population analysis) charges as nol moiety is not oriented favorably relative to the ring. The
described in [68].
effect can, however, be understood by inspecting the respective
dispersion interaction densities (DIDs, cf. Figure 2), which
The dispersion interaction energies show an interesting pattern. allow for an even finer-grained analysis. There, one can observe
Although all structures are significantly stabilized by disper- that the major contributor is not the π-system of the phenyl ring,
Table 2: Comparison of different structures for PVE–MeOH dimers, with dispersion energies calculated at the LCCSD/VQZ-F12 level of theory (the
parentheses contain the percentage of the fragment’s dispersion relative to the total dispersion energy).
∆Edisp(total) [kJ/mol]
∆Edisp(phenyl) [kJ/mol]
∆Edisp(O) [kJ/mol]
∆Edisp(ethenyl) [kJ/mol]
OH–O
OH–O’
OH–P
OH–P’
OH–E
OH–E’
−14.3
−15.7
−16.9
−16.0
−15.6
−13.1
−6.2 (43.2)
−6.1 (38.7)
−11.9 (70.4)
−14.0 (87.6)
−6.9 (44.3)
−4.5 (34.0)
−5.1 (35.8)
−4.9 (31.5)
−1.5 (9.2)
−1.3 (8.0)
−2.1 (13.3)
−0.9 (6.7)
−3.0 (21.1)
−4.7 (29.8)
−3.5 (20.5)
−0.7 (4.4)
−6.6 (42.4)
−7.8 (59.3)
Figure 2: Dispersion interaction density (DID) plots calculated at the LCCSD/VQZ-F12 level. The brown zones indicate regions of electron density in
a monomer which interact strongly by dispersion interactions with the other molecule. Blue stands for weaker/diffuse contributions. For example, in
the top left figure one can observe that the OH group of methanol interacts strongly with the ether oxygen, with some dispersion energy coming as
well from a CH orbital in the phenyl close to the methanol.
1647

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
but a C–H contact to the methanol (a similar effect had already If one were to trust the relative harmonic wavenumber predic-
been observed in diphenyl ether–methanol complexes [19]). tions from the preceding subsection (cf. Table 1), this would
This contact is reminiscent of stabilization effects observed in imply a single docking motif, as different docking motifs are
coupled diamondoids [69] or supramolecular complexes [70], predicted to lead to larger spectral separations. However, differ-
where such interactions can be found in large numbers.
ent extents of anharmonicity do not allow to completely ruling
out overlapping docking motifs. Therefore, conformationally
selective methods are desirable to investigate this possibility.
Finally, the actual docking site has to be identified by structural
Electronic ground state spectra
FTIR spectroscopy
The results of an FTIR exploration of the conformational diver- or electronic excitation spectroscopy.
sity of this system are shown in Figure 3. Besides methanol
monomer, methanol dimer and a signal clearly attributed to a IR/R2PI spectroscopy
larger cluster, only a single, reasonably narrow absorption at Additional insight can be gained by using the mass- and isomer-
3625 cm−1 is observed. It can be attributed to mixed dimers of selective IR/R2PI technique. This method requires knowledge
MeOH with PVE and allows for a single rigorous conclusion, on electronic excitation energies of the PVE–MeOH complex.
due to the linearity of the technique and the comparable IR For this reason, one-color R2PI spectra were recorded in the
absorption cross section of all predicted dimer conformations range of 36100–37600 cm−1 (cf. Figure S1 in Supporting Infor-
(cf. Table 1 and Table S1, Supporting Information File 1): the mation File 1). While the R2PI spectrum of the PVE monomer
global minimum structure and any other, higher lying isomers shows well-resolved vibrational progressions (cf. Figure S1a,
which are initially formed and impeded from relaxation to the Supporting Information File 1), the spectrum of the solvent
global minimum due to broad or high interconversion barriers aggregate is broadened and affected by ionization-induced frag-
must have their OH stretching fundamental at 3625 ± 5 cm−1 or mentation of larger clusters (cf. Figure S1b, Supporting Infor-
be significantly less abundant.
mation File 1). This is also reflected in the recorded IR/R2PI
Figure 3: FTIR spectra of the supersonic expansion of methanol (MeOH) and phenyl vinyl ether (PVE) at different concentrations in helium. The
spectra are spread out along the ordinate to improve visualization. Only one dominant mixed dimer band is visible in the spectra, lying at 3625 cm−1
(marked MeOH + PVE) between the methanol monomer at 3686 cm−1 (MeOH) and methanol dimer at 3575 cm−1 ((MeOH)2). By comparing the spec-
trum at the top with the other two spectra recorded at reduced concentrations of methanol (middle) or methanol and PVE (bottom), the further down-
shifted band at 3466 cm−1 can be attributed to a higher cluster, probably a methanol-rich mixed trimer ((MeOH)2 + PVE (?)), due to its scaling with the
variation of the concentrations.
1648

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
spectra (cf. Figure S2, Supporting Information File 1), yielding (37907 cm−1), which would coincide with the fragmentation-
solely the spectrum shown in Figure 4 via the excitation energy dominated region of the R2PI spectrum, where, however, only
of 36885 cm−1 containing an OH stretching vibration of a signatures of larger clusters could be identified. These consider-
PVE–MeOH dimer.
ations strengthen the arguments for the presence of an OH∙∙∙O
structure laid before, on the basis of the computed coupled
cluster energies and the SCS-CC2 fundamental stretch frequen-
cies. Additional experimental insight will be gained from the
UV/IR/UV spectrum of the S1 state as well as the microwave
investigations in the following section.
Chirp pulse Fourier transform microwave
(CP-FTMW) spectroscopy
From the broadband CP-FTMW spectra obtained with neon as a
carrier gas, we assigned two PVE–MeOH complexes with sig-
nificantly different intensities. Complex 1 is about ten times
more intense than complex 2. The experimental rotational con-
stants (Table 3) for the two isomers agree the best with the
values calculated for the OH–O’ isomer (as also indicated in the
FTIR and the IR–UV investigations, which are, however,
unable to distinguish OH–O from OH–O’) and the OH–P
isomer, respectively. The identification of the two complexes to
Figure 4: The IR/R2PI spectrum in the range of 3520–3750 cm−1 was
obtained via the excitation energy of 36885 cm−1 using the carrier gas
neon; the asterisk (*) indicates ionization-induced fragmentation from
larger clusters (cf. Figure S2 in Supporting Information File 1).
Comparing the calculated OH stretching frequencies for the dif- the OH–O’ and the OH–P isomers is guided by the absolute and
ferent isomers obtained at the DFT-D3 and SCS-CC2 levels (cf. relative values of the B and C rotational constants. Generally,
Table 1) to the experiment does not allow for a clear structural the rotational constants calculated at the SCS-CC2/def2-TZVP
assignment: the DFT-D3 calculations show the best agreement level of theory agree somewhat better with the experimental
for the OH–P structure (3619 cm−1, scaling factor 0.9600). values than the B3LYP-D3(BJ)/def2-TZVP values (note that we
Regarding the relative electronic energies, the OH–O structure compare experimental B0 rotational constants with theoretical
is somewhat favored, with OH–O’ coming as a close second. Be rotational constants here). For the OH–P complex, however,
The latter frequencies are 3597 and 3600 cm−1, respectively, at we find that the B3LYP-D3(BJ)/def2-TZVP level of theory
the same level of theory. On the other hand, the computed SCS- provides a better prediction of the magnitudes of the dipole-
CC2 frequencies would provide a coincident assignment, as moment components. Experimentally, we only observe a-type
both O-docking isomers would have the closest fundamentals transitions for this complex, which points to rather low values
compared to the measured frequency (3619 and 3621 cm−1). for μb and μc. SCS-CC2 calculations predict all three
The assignment, however, would be tentative at best with this dipole-moment components to be of comparable magnitude. At
information alone. The OH–E isomers on the other hand can be the B3LYP-D3(BJ) level, μa is predicted to be significantly
excluded due to their lower OH stretching frequencies as well stronger than μb and μc. This change in magnitude for the
as the energetic disadvantage at the LCCSD(T0)-F12/CBS[T:Q] dipole-moment components for different levels of calculation is
level (cf. Table 1).
more often observed for weakly bound complexes because the
exact arrangement of the two monomers with respect to each
In order to elucidate this problem, the electronic excitation ener- other can have a major influence on the dipole-moment compo-
gies can serve as a further indication for the binding motif, as nents. Also note that in none of the spectroscopic experiments,
shown for DPE–alcohol clusters before [20-22]. Comparing the we observe the OH–E isomer that is also predicted to be of rela-
vertical excitation energies for the different isomers with the ex- tively low energy (cf. Table 1).
perimental excitation energy of 36885 cm–1 yields the best
agreement for the OH–O or OH–O’ isomer, which also show a The rotational spectra of the two isomers are qualitatively dif-
significantly blue-shifted S1←S0 transition compared to the ferent. For the OH–O’ isomer (complex 1), we observe a char-
PVE monomer (adiabatic excitation energies of 38291 and acteristic line splitting into so-called A and E components (cf.
38164 cm−1, respectively, compared to 38034 cm−1 for the PVE Figure 5) arising from internal rotation of the methyl group of
monomer, cf. Table S2, Supporting Information File 1), as ob- methanol, similar to the case of the DPE–MeOH complex. For
served experimentally. In contrast to that, a red-shifted S1←S0 the OH–P isomer (complex 2), no line splitting due to internal
transition compared to PVE is predicted for the OH–P isomer rotation was observed. This is consistent with the higher barrier
1649

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
Table 3: Experimental rotational constants of the two observed complexes, using neon as carrier gas, that are assigned to the OH–O’ and the OH–P
isomers, respectively. The experimental rotational parameters for the OH–O’ isomer (called Exp 1) are the results of a fit to a rigid-rotor asymmetric
Hamiltonian including solely the A lines of the internal rotation splitting. Rotational parameters of a global fit (XIAM) including both A and E levels due
to internal rotation for the OH–O’ isomer are presented in the Table S9 of Supporting Information File 1.
Complex 1
Complex 2
(OH–O’ isomer)
(OH–P isomer)
Exp 1
SCS-CC2/def2-TZVP Exp 1
SCS-CC2/def2-TZVP
A [MHz]
1466.59120(26)
697.48965(11)
572.109900(95)
0.72697(62)
1501.94
697.58
589.94
1275.7623(49)
818.45271(73)
640.2184(11)
0.070(14)
2.19(10)
–
1297.89
818.01
646.81
B [MHz]
C [MHz]
∆J [kHz]
∆JK [kHz]
−0.6669(26)
5.6217(62)
∆K [kHz]
δJ [kHz]
0.15121(11)
–
δK [kHz]
2.5783(29)
–
A state transition
Dipole moment (D) (μa/μb/μc)
σ [kHz]
213 (49/104/60)
20(20/0/0)
2.2/1.9/1.2
0.8/0.4/0.8
6.7
7.9
for this motion due to the secondary interactions of the methyl often a good approximation, is summarized (Exp 1). In addition,
group with PVE (cf. Figure 2).
we used the program XIAM to perform a global fit including
both A and E lines. This global fit does not only provide the
rotational constants, but also parameters of the internal rotor, in
this case the methyl group. This includes the barrier height for
internal rotation as well as the geometrical arrangement of the
rotor with respect to the overall rotating molecule, as also dis-
cussed for the DPE–MeOH complex [19]. For PVE–MeOH, the
barrier height was determined to be 261 cm−1, as summarized in
Table S9 of Supporting Information File 1. This value is in
agreement to barrier heights observed for other complexes with
methanol [19]. It is somewhat lower than in the case of free
methanol (373 cm−1) and also lower than the calculated barrier
height of 341 cm−1 (cf. Table S9, Supporting Information
File 1). This somewhat lower methyl group internal rotation
barrier for the OH–O’ isomer could point to a softening of the
C–O bond of methanol due to the hydrogen bond. The DID
plots in Figure 2 also indicate that the methyl group is basically
free from other interactions, so that no additional hindering is
expected.
Figure 5: A section of the experimental 2–8 GHz spectrum using a
mixture of PVE and MeOH (3 million acquisitions). The upper experi-
mental trace in black is compared with simulations, based on fitted pa-
rameters that can be assigned to the OH–O’ isomer (complex 1, red)
for the PVE–MeOHcomplex. The observed complex has a clear split-
ting pattern due to the internal rotation of the methyl group of metha-
nol, labeled with A and E. The experimental 13C positions (blue atoms)
(rs substitution structure) deduced from a Kraitchman analysis are
compared to the calculated structure at the SCS-CC2/def2-TZVP level
of theory and further confirm the observation of the OH–O’ isomer.
Furthermore, the transition intensities for the OH–O’ isomer are
strong enough (with a signal-to-noise (SNR) of about 500:1 to
600:1 for the stronger transitions) to assign rotational transi-
tions arising from all nine singly substituted 13C isotopologues
in natural abundance (about 1%, cf. Figure S3, Supporting
Information File 1). The additional data sets of rotational
Two different ways of analyzing the rotational spectrum of the constants are summarized in Supporting Information File 1
OH–O’ isomer (complex 1) were performed. In Table 3, the (Table S12) together with line lists of the main isotopologues
results from a fit to an asymmetric-top Hamiltonian of only the (Tables S10–S11) and the 13C isotopologues (Tables S13–S21).
A state species of the internal rotation splitting pair, which is They allow us, using Kraitchman’s equations, to determine the
1650

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
carbon substitution structure, rs, of the complexes, which are
the positions of the respective substituted carbon atoms with
respect to the center of mass of the complex and thus the car-
bon backbone structure. The obtained rs structure for complex 1
(cf. Figure 5) further confirms the assignment of complex 1 as
the OH–O’ isomer, where the methyl group of the methanol
moiety points towards the phenyl ring.
As mentioned, the OH–O’ isomer is about ten times more
intense than the OH–P isomer. The intensity observed in
CP-FTMW spectroscopy directly depends on the number of
molecules, i.e., the population of the respective isomers, as well
as the square of the transition dipole moments. Since the μa
values for the two isomers differ by a factor of two (cf. Table
S11 in Supporting Information File 1), the OH–P isomer can be
considered to be about 2.5 times less populated than the OH–O’
isomer, as an upper estimate. Taking the predicted energy
difference of 0.8 kJ/mol for granted, this ratio would corre-
Figure 6: UV/IR/UV spectrum of PVE–MeOH in the range of
3520–3750 cm−1; excitation laser: 36741 cm−1, ionizing laser:
31847 cm−1, carrier gas helium; the lower trace shows the calculated
OH stretching frequencies at the SCS-CC2/def2-TZVP level for the op-
timized S0 and S1 structure of the OH–O’ isomer scaled by 0.9635.
spond to a plausible [18] conformational freezing temperature the ground state at 3619 cm−1 (cf. Table 1) for the OH–O’
of 100 K. A three-fold lower or three-fold higher conformation- isomer, which is also reflected in an increase of the H∙∙∙O
al temperature appears unlikely, and thus a tentative experimen- hydrogen bond distance from 2.068 to 2.168 Å from S0 to S1
tal energy penalty for OH–P relative to OH–O’ ranges from 0.3 state geometry. This destabilization of the OH∙∙∙O hydrogen
to 2 kJ/mol. This contradicts both inexpensive approaches bond is further reflected in the calculated binding energies of
(B3LYP-D3 and SCS-CC2 with def2-TZVP) and suggests that the PVE–MeOH complex obtained at the SCS-CC2/def2-TZVP
these methods somewhat underestimate the stability of OH∙∙∙O level, which are reduced by 0.9 kJ/mol regarding D0 and
contacts.
1.6 kJ/mol regarding De in the S1 state compared to the S0 state
for OH–O’ (cf. Table S3, Supporting Information File 1). The
spectral shift can be explained by regarding the HOMO and
Electronically excited state spectrum
For the investigation of the electronically excited state by using LUMO orbitals involved in the S1←S0 transition, which is pre-
the UV/IR/UV technique, a two-color R2PI signal is required. dicted to be mainly a π–π* transition with a small charge
For this reason, the one-color R2PI signal was suppressed by transfer contribution from the ether oxygen to the phenyl ring.
attenuating the laser power of the excitation laser. On the other The latter leads to a slightly decreased electron density at the
hand, higher pulse energies were used for the ionizing laser. binding site for the methanol molecule and therefore weakens
The latter was set to 31847 cm−1 for the UV/IR/UV experiment the hydrogen bond. These findings are in line with observations
in order to yield the best two-color R2PI signal. Figure 6 shows in previous studies on diphenyl ether–alcohol complexes
the recorded UV/IR/UV spectrum for the PVE–MeOH mass [20,21].
trace.
In principle, as the OH–O’ isomer has been identified in the S0
Due to temporally overlapping laser pulses, the spectrum state, the observation of a respective OH∙∙∙O-bound structure
contains transitions from the S0 state, the electronically excited can be expected in the S1 state as well. However, the OH–P
(S1, at 3637 cm−1) and also the ionic D0 state (at 3667 cm−1). isomers are predicted to be significantly stabilized in the S1
This could not be avoided, as the lifetime of the excited state, state (cf. Table S2, Supporting Information File 1). Neverthe-
which is estimated to be in the order of 5–7 ns, is shorter than less, due to the predicted red-shifts of the OH stretching
the laser pulse-widths of 7–10 ns. The OH stretching vibration frequencies of the OH–P isomers (indicating an increased
at 3637 cm−1 originating from the electronically excited state of hydrogen bond strength compared to the S0 state), their pres-
the PVE–MeOH complex is blue-shifted compared to the ence, i.e., by a rearrangement reaction from the OH–O’ isomer,
ground state, which indicates a decrease of the hydrogen bond can be excluded. By exciting the electronic origin of the OH–O’
strength in the S1 compared to the S0 state. A comparison with isomer the formation of OH–E isomers can also be excluded as
SCS-CC2 calculations shows a good agreement of a blue- their expected excitation energies are higher than the one for
shifted OH stretching frequency at 3642 cm−1 (cf. Table S2, OH–O’ and in addition they are energetically less stable (cf.
Supporting Information File 1, scaled by 0.9635) compared to Table S2, Supporting Information File 1).
1651

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
Conclusion
a conclusion might be that methanol needs the interaction with a
In this paper, the first spectroscopic and theoretical investiga- second phenyl ring in order to prefer the OH∙∙∙π motif over
tion on the isolated phenyl vinyl ether–methanol complex is OH∙∙∙O, as observed for diphenyl ether. The secondary interac-
presented. From the FTIR spectra, the existence of one isomer tion of methanol with a smaller ethenyl moiety being present in
is concluded, which is confirmed by IR/UV spectroscopy in the phenyl vinyl ether instead of a phenyl ring seems to be insuffi-
electronic ground state (S0). The combined vibrational and elec- cient to favor the phenyl docking site.
tronic spectroscopic investigations, including a comparison of
vibrational frequencies and electronic excitation energies, allow
Supporting Information
for an assignment of an OH∙∙∙O-bound structure. Broadband
rotational (CP-FTMW) spectroscopy ultimately identifies
Supporting Information File 1
OH–O’ as the observed isomer, ruling out the presence of the
nearly isoenergetic OH–O. One explanation for its elusiveness
would be a low interconversion barrier. However, rotational
spectroscopy further reveals the presence of the OH–P isomer
as a second isomer, being less populated, which is not observed
with the less sensitive FTIR technique and might be superim-
posed by fragmentation of larger clusters in the usually more
sensitive IR/UV experiments or it is even not populated due to
different expansion conditions. No evidence was found for an
OH∙∙∙ethenyl-bound structure, which is in agreement with the
more pronounced energetic discrimination of OH–E isomers
compared to the other binding motifs predicted at the
LCCSD(T0)-F12/CBS[T:Q] level of theory.
Additional computational and experimental data.
[https://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-14-140-S1.pdf]
Acknowledgements
The authors thank the Deutsche Forschungsgemeinschaft (DFG,
Ge 961/9-1, Schn 1280/4-1, Su 121/5-1, and Ma 5063/3-1) for
financial support in context of the priority program SPP 1807
on the Control of London dispersion interactions in molecular
chemistry. M.F. acknowledges support from the Hamburg Inter-
national Max Planck Research School UFAST. This work is
part of the Ph.D. theses of D.B., M.F., H.G. and A.W.
ORCID® iDs
In the electronically excited state (S1), the OH stretching vibra-
tion of the attached methanol undergoes a blue-shift compared
to the S0 state. This indicates a weakening of the OH∙∙∙O bond
upon electronic excitation compared to the ground state and is
in good agreement with the calculated frequency shift for the S0
and S1 state structures obtained at the SCS-CC2/def2-TZVP
level and is furthermore in line with findings for similar
diphenyl ether–alcohol complexes from previous investigations
[20,21].
Dominic Bernhard - https://orcid.org/0000-0002-6148-0445
Fabian Dietrich - https://orcid.org/0000-0001-5813-5248
Cristóbal Pérez - https://orcid.org/0000-0001-5248-5212
Hannes C. Gottschalk - https://orcid.org/0000-0003-4343-9542
Ricardo A. Mata - https://orcid.org/0000-0002-2720-3364
Martin A. Suhm - https://orcid.org/0000-0001-8841-7705
Melanie Schnell - https://orcid.org/0000-0001-7801-7134
Markus Gerhards - https://orcid.org/0000-0002-8748-2940
References
1. van der Waals, J. D. Ph.D. Thesis, Leiden, 1873.
2. London, F. Z. Phys. 1930, 63, 245–279. doi:10.1007/BF01421741
3. Müller-Dethlefs, K.; Hobza, P. Chem. Rev. 2000, 100, 143–168.
doi:10.1021/cr9900331
In summary, we present a multi-spectroscopic analysis on a mo-
lecular complex with a very delicate balance between, for the
first time, three different binding motifs. This provides an
excellent benchmark system for theory, since DFT-D3 as well
as SCS-CC2 methods fail in predicting the correct energetic
order, whereas LCCSD(T0)-F12 succeeds in the preferred
docking motif. These differences are in the range of only
2 kJ/mol, when considering relative electronic energies, but that
is already enough to tip the scales in the wrong direction.
Comparing VTZ-F12 and VQZ-F12 results, we observe that the
electronic energies are well converged for the smaller basis
(Table S6, Supporting Information File 1). This would place the
main accuracy bottleneck in the electronic structure method
(i.e., functional, correlation truncation) chosen.
4. Lehn, J.-M. Angew. Chem., Int. Ed. Engl. 1988, 27, 89–112.
doi:10.1002/anie.198800891
5. Černý, J.; Hobza, P. Phys. Chem. Chem. Phys. 2007, 9, 5291–5303.
doi:10.1039/b704781a
6. Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int. Ed.
2003, 42, 1210–1250. doi:10.1002/anie.200390319
7. Mata, R. A.; Suhm, M. A. Angew. Chem., Int. Ed. 2017, 56,
11011–11018. doi:10.1002/anie.201611308
8. Zwier, T. S. Annu. Rev. Phys. Chem. 1996, 47, 205–241.
doi:10.1146/annurev.physchem.47.1.205
9. Ebata, T.; Fujii, A.; Mikami, N. Int. Rev. Phys. Chem. 1998, 17,
331–361. doi:10.1080/014423598230081
10.Brutschy, B. Chem. Rev. 2000, 100, 3891–3920.
doi:10.1021/cr990055n
Finally, regarding the docking preference in comparison to the
previously investigated diphenyl ether complex with methanol,
11.Pribble, R. N.; Hagemeister, F. C.; Zwier, T. S. J. Chem. Phys. 1997,
106, 2145. doi:10.1063/1.473784
1652

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
12.Djafari, S.; Barth, H.-D.; Buchhold, K.; Brutschy, B. J. Chem. Phys.
1997, 107, 10573–10581. doi:10.1063/1.474221
13.Buchanan, E. G.; Gord, J. R.; Zwier, T. S. J. Phys. Chem. Lett. 2013, 4,
1644–1648. doi:10.1021/jz400641p
36.Chen, Y.; Topp, M. R. Chem. Phys. 2002, 283, 249–268.
doi:10.1016/S0301-0104(02)00613-4
37.Ishiuchi, S.-i.; Daigoku, K.; Saeki, M.; Sakai, M.; Hashimoto, K.;
Fujii, M. J. Chem. Phys. 2002, 117, 7077–7082.
doi:10.1063/1.1508103
14.Walsh, P. S.; Buchanan, E. G.; Gord, J. R.; Zwier, T. S. J. Chem. Phys.
2015, 142, 154304. doi:10.1063/1.4917307
38.Stearns, J. A.; Das, A.; Zwier, T. S. Phys. Chem. Chem. Phys. 2004, 6,
2605–2610. doi:10.1039/b313831c
15.Walsh, P. S.; Buchanan, E. G.; Gord, J. R.; Zwier, T. S. J. Chem. Phys.
2015, 142, 154303. doi:10.1063/1.4917305
39.Fricke, H.; Bartl, K.; Funk, A.; Gerlach, A.; Gerhards, M.
ChemPhysChem 2008, 9, 2592–2600. doi:10.1002/cphc.200800499
40.Bartl, K.; Funk, A.; Gerhards, M. J. Chem. Phys. 2008, 129, 234306.
doi:10.1063/1.3037023
16.Pietraperzia, G.; Pasquini, M.; Mazzoni, F.; Piani, G.; Becucci, M.;
Biczysko, M.; Michalski, D.; Bloino, J.; Barone, V. J. Phys. Chem. A
2011, 115, 9603–9611. doi:10.1021/jp200444a
17.Gottschalk, H. C.; Altnöder, J.; Heger, M.; Suhm, M. A.
Angew. Chem., Int. Ed. 2016, 55, 1921–1924.
41.Bartl, K.; Funk, A.; Gerhards, M. ChemPhysChem 2009, 10,
1882–1886. doi:10.1002/cphc.200900097
doi:10.1002/anie.201508481
42.Stamm, A.; Weiler, M.; Brächer, A.; Schwing, K.; Gerhards, M.
Phys. Chem. Chem. Phys. 2014, 16, 21795–21803.
doi:10.1039/C4CP02546F
18.Poblotzki, A.; Gottschalk, H. C.; Suhm, M. A. J. Phys. Chem. Lett.
2017, 8, 5656–5665. doi:10.1021/acs.jpclett.7b02337
19.Medcraft, C.; Zinn, S.; Schnell, M.; Poblotzki, A.; Altnöder, J.;
Heger, M.; Suhm, M. A.; Bernhard, D.; Stamm, A.; Dietrich, F.;
Gerhards, M. Phys. Chem. Chem. Phys. 2016, 18, 25975–25983.
doi:10.1039/C6CP03557D
43.Miyazaki, M.; Fujii, M. Phys. Chem. Chem. Phys. 2017, 19,
22759–22776. doi:10.1039/C7CP03480F
44.Grimme, S. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2011, 1,
211–228. doi:10.1002/wcms.30
20.Bernhard, D.; Dietrich, F.; Fatima, M.; Perez, C.; Poblotzki, A.;
Jansen, G.; Suhm, M. A.; Schnell, M.; Gerhards, M.
Phys. Chem. Chem. Phys. 2017, 19, 18076–18088.
doi:10.1039/C7CP02967E
45.Grimme, S.; Ehrlich, S.; Goerigk, L. J. Comput. Chem. 2011, 32,
1456–1465. doi:10.1002/jcc.21759
46.Hellweg, A.; Grün, S. A.; Hättig, C. Phys. Chem. Chem. Phys. 2008,
10, 4119–4127. doi:10.1039/b803727b
21.Bernhard, D.; Holzer, C.; Dietrich, F.; Stamm, A.; Klopper, W.;
Gerhards, M. ChemPhysChem 2017, 18, 3634–3641.
doi:10.1002/cphc.201700722
47.Adler, T. B.; Werner, H.-J. J. Chem. Phys. 2009, 130, 241101.
doi:10.1063/1.3160675
48.Wuttke, A.; Mata, R. A. J. Comput. Chem. 2017, 38, 15–23.
doi:10.1002/jcc.24508
22.Dietrich, F.; Bernhard, D.; Fatima, M.; Perez, C.; Schnell, M.;
Gerhards, M. Angew. Chem., Int. Ed. 2018, in press.
doi:10.1002/anie.201801842
49.Suhm, M. A.; Kollipost, F. Phys. Chem. Chem. Phys. 2013, 15,
10702–10721. doi:10.1039/c3cp51515j
23.Maier, J. M.; Li, P.; Vik, E. C.; Yehl, C. J.; Strickland, S. M. S.;
Shimizu, K. D. J. Am. Chem. Soc. 2017, 139, 6550–6553.
doi:10.1021/jacs.7b02349
50.Unterberg, C.; Jansen, A.; Gerhards, M. J. Chem. Phys. 2000, 113,
7945. doi:10.1063/1.1315610
51.Iwasaki, T.; Miyata, Y.; Akimoto, R.; Fujii, Y.; Kuniyasu, H.; Kambe, N.
J. Am. Chem. Soc. 2014, 136, 9260–9263. doi:10.1021/ja5043534
52.Schmitz, D.; Alvin Shubert, V.; Betz, T.; Schnell, M. J. Mol. Spectrosc.
2012, 280, 77–84. doi:10.1016/j.jms.2012.08.001
53.Pérez, C.; Lobsiger, S.; Seifert, N. A.; Zaleski, D. P.; Temelso, B.;
Shields, G. C.; Kisiel, Z.; Pate, B. H. Chem. Phys. Lett. 2013, 571,
1–15. doi:10.1016/j.cplett.2013.04.014
24.Heger, M.; Mata, R. A.; Suhm, M. A. Chem. Sci. 2015, 6, 3738–3745.
doi:10.1039/C5SC01002K
25.Medel, R.; Heger, M.; Suhm, M. A. J. Phys. Chem. A 2015, 119,
1723–1730. doi:10.1021/jp508424p
26.Mons, M.; Robertson, E. G.; Simons, J. P. J. Phys. Chem. A 2000, 104,
1430–1437. doi:10.1021/jp993178k
27.Page, R. H.; Shen, Y. R.; Lee, Y. T. J. Chem. Phys. 1988, 88, 4621.
doi:10.1063/1.453775
54.Kraitchman, J. Am. J. Phys. 1953, 21, 17–24. doi:10.1119/1.1933338
55.JB95 Spectral fitting program; NIST.
28.Riehn, C.; Lahmann, C.; Wassermann, B.; Brutschy, B.
Chem. Phys. Lett. 1992, 197, 443–450.
Plusquellic, D.
56.Gaussian 09, revision D.01; Gaussian, Inc.: Wallingford, CT, USA,
2009.
doi:10.1016/0009-2614(92)85798-F
29.Gerhards, M.; Unterberg, C. Phys. Chem. Chem. Phys. 2002, 4,
1760–1765. doi:10.1039/b110029g
57.Halgren, T. A. J. Comput. Chem. 1996, 17, 490–519.
doi:10.1002/(SICI)1096-987X(199604)17:5/6<490::AID-JCC1>3.0.CO;
2-P
30.Loquais, Y.; Gloaguen, E.; Habka, S.; Vaquero-Vara, V.; Brenner, V.;
Tardivel, B.; Mons, M. J. Phys. Chem. A 2015, 119, 5932–5941.
doi:10.1021/jp509494c
58.Hanwell, M. D.; Curtis, D. E.; Lonie, D. C.; Vandermeersch, T.;
Zurek, E.; Hutchison, G. R. J. Cheminf. 2012, 4, 17.
doi:10.1186/1758-2946-4-17
31.Ebata, T.; Fujii, A.; Mikami, N. Int. J. Mass Spectrom. Ion Processes
1996, 159, 111–124. doi:10.1016/S0168-1176(96)04445-X
32.Gerhards, M.; Schumm, S.; Unterberg, C.; Kleinermanns, K.
Chem. Phys. Lett. 1998, 294, 65–70.
59.Furche, F.; Ahlrichs, R.; Hättig, C.; Klopper, W.; Sierka, M.;
Weigend, F. Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2014, 4,
91–100. doi:10.1002/wcms.1162
doi:10.1016/S0009-2614(98)00823-9
60.Weigend, F.; Köhn, A.; Hättig, C. J. Chem. Phys. 2002, 116, 3175.
doi:10.1063/1.1445115
33.Palmer, P. M.; Chen, Y.; Topp, M. R. Chem. Phys. Lett. 2000, 318,
440–447. doi:10.1016/S0009-2614(00)00036-1
61.Peterson, K. A.; Adler, T. B.; Werner, H.-J. J. Chem. Phys. 2008, 128,
84102. doi:10.1063/1.2831537
34.Robertson, E. G. Chem. Phys. Lett. 2000, 325, 299–307.
doi:10.1016/S0009-2614(00)00661-8
62.Yousaf, K. E.; Peterson, K. A. J. Chem. Phys. 2008, 129, 184108.
doi:10.1063/1.3009271
35.Chen, Y.; Palmer, P. M.; Topp, M. R. Int. J. Mass Spectrom. 2002, 220,
231–251. doi:10.1016/S1387-3806(02)00772-8
1653

Beilstein J. Org. Chem. 2018, 14, 1642–1654.
63.Hill, J. G.; Peterson, K. A.; Knizia, G.; Werner, H.-J. J. Chem. Phys.
2009, 131, 194105. doi:10.1063/1.3265857
64.Pipek, J.; Mezey, P. G. J. Chem. Phys. 1989, 90, 4916–4926.
doi:10.1063/1.456588
65.Reed, A. E.; Weinstock, R. B.; Weinhold, F. J. Chem. Phys. 1985, 83,
735–746. doi:10.1063/1.449486
66.MOLPRO, version 2015.1; a package of ab initio programs,
http://www.molpro.net.
67.Schütz, M.; Rauhut, G.; Werner, H.-J. J. Phys. Chem. A 1998, 102,
5997–6003. doi:10.1021/jp981168y
68.Andrejić, M.; Mata, R. A. Phys. Chem. Chem. Phys. 2013, 15,
18115–18122. doi:10.1039/c3cp52931b
69.Schreiner, P. R.; Chernish, L. V.; Gunchenko, P. A.; Tikhonchuk, E. Y.;
Hausmann, H.; Serafin, M.; Schlecht, S.; Dahl, J. E. P.;
Carlson, R. M. K.; Fokin, A. A. Nature 2011, 477, 308–311.
doi:10.1038/nature10367
70.Löffler, S.; Wuttke, A.; Zhang, B.; Holstein, J. J.; Mata, R. A.;
Clever, G. H. Chem. Commun. 2017, 53, 11933–11936.
doi:10.1039/C7CC04855F
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
(http://creativecommons.org/licenses/by/4.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
(https://www.beilstein-journals.org/bjoc)
The definitive version of this article is the electronic one
which can be found at:
doi:10.3762/bjoc.14.140
1654