Heteronuclear NOE spectroscopy of ionic liquids

Article abstract of DOI:10.1002/cphc.201100622

¹⁹F,¹H HOESY experiments with three ionic liquids ([bmim]BF₄, [bmim]PF₆ and [emim]BF₄) were run in two different solvents and neat. The results give preferred probabilities of presence and enable us to systematically study interactions between the cations and the anions in the ionic liquid phase by NMR spectroscopy. The influence of different solvents and of the presence or absence of air (i.e. oxygen) is discussed. This enabled us to substantially speed up the NMR experiments and to develop a more precise method for the investigation of liquid-phase structures in ionic liquids. Copyright

Full text of DOI:10.1002/cphc.201100622

DOI: 10.1002/cphc.201100622
Heteronuclear NOE Spectroscopy of Ionic Liquids
Yves Lingscheid, Sven Arenz, and Ralf Giernoth*[a]
19F,1H HOESY experiments with three ionic liquids ([bmim]BF4,
[bmim]PF6 and [emim]BF4) were run in two different solvents
and neat. The results give preferred probabilities of presence
and enable us to systematically study interactions between the
cations and the anions in the ionic liquid phase by NMR spec-
troscopy. The influence of different solvents and of the pres-
ence or absence of air (i.e. oxygen) is discussed. This enabled
us to substantially speed up the NMR experiments and to de-
velop a more precise method for the investigation of liquid-
phase structures in ionic liquids.
1. Introduction
Within the past twenty years ionic liquids (ILs) have developed
from curiosities to commodities. To date they are omnipresent
in nearly every aspect of modern chemistry.[1–3] ILs are fre-
quently referred to as designer solvents, for some authors
claim that their properties can be designed to suit any particu-
lar need. But of course, it is not their properties that can be
changed, it is the particular IL that can be exchanged. To be
able to do so, it is vital to know about the governing supra-
molecular structures and interactions in the ionic-liquid phase.
A useful method for investigations in the liquid phase is
NMR spectroscopy. For the last eight years our group has fo-
cussed on the establishment and application of NMR spectros-
copy in neat ILs.[4–8]
longitudinal portion, the factors RH,1 and RF,1 can be derived
from the longitudinal (spin-lattice) relaxation rates (T1). These
can be determined by means of the inversion-recovery experi-
ment.[14]
Several groups were already able to demonstrate the useful-
ness of NOE spectroscopy for the investigation of ion pairs,
particularly for metal complexes in solution.[15–20] Consequently,
the next step was the investigation of ILs in solution through
NOE spectroscopy. Dupont was the first to apply homonuclear
H,H HOESY NMR spectroscopy on ionic liquids in organic sol-
1
vents to study ion pairing.[21] Pochapsky determined the H,1H
ꢀ
and 11B,1H NOEs in the (C4H9)4N+BH4 ion pair using HOESY
NMR.[18] The first investigations in neat ILs were performed by
Mele for [bmim]BF4 and [bdmim]BF4 using the rotating frame
NOE experiment (ROESY).[22] The influence of a co-solvent was
studied by Nama for ion pairs in solution (methanol) and
neat.[23]
It is long established that the nuclear Overhauser effect
(NOE)[9] is a powerful tool for the determination of intermolec-
ular interactions.[10–12] This effect arises due to dipolar cross re-
laxation between two spatially neighbouring nuclear spins
(typically below 5 ꢀ). To be able to observe an NOE between
two atoms within a complex or an ion pair, the lifetime of this
structure needs to be long enough so that there is a notable
probability for cross relaxation. The time dependence of the
NOE build-up can be derived from the fundamental Solomon
equations[13] and results in the following exponential function
[Eq. (1) here shown exemplary for an H,F NOE]:[11]
As a consequence of these studies, application of NOE NMR
spectroscopy to neat ILs is needed. Zuccacia et al. already in-
vestigated neat [bmim]BF4 and [bdmim]BF4 but only consid-
ered cation–cation interactions for their structural analysis and
mentioned anions only briefly. Their assertions—most interac-
tions being due to p stacking and H-bonding—are in contrast
to the recent literature which neglects the influence of p stack-
ing and H-bonding in favor of Coulombic interactions.[24] Thus,
it seems reasonable to study the anion–cation interactions in
detail. Due to the absence of protons in the commonly used
1
2
HFÞt
HFt
NOEHfFgðtÞ ¼ eꢀðRꢀs ð1 ꢀ eꢀ2s
Þ
ð1Þ
ꢀ
ꢀ
anions BF4 and PF6 the use of 19F,1H NOE experiments is nec-
essary. It would also be desirable to obtain information on the
overall distances between each proton and the corresponding
anion. To that end, investigation of the NOE kinetics is necessa-
ry. It is important to note that for most ILs each individual nu-
cleus needs an individual treatment, because the relaxation
This equation assumes that the total relaxation rate R is ap-
proximately equal for all protons and fluorine nuclei. A more
precise treatment [Eqs. (2) and (3)] considers the different re-
laxation rates:
RH;1 þR
2 sin Dt
ꢀ
ð
F;1 t
Þ
2
ð2Þ
ð3Þ
NOEHfFgðtÞ ¼ M0s
e
D
[a] Y. Lingscheid, Dr. S. Arenz, Dr. R. Giernoth
Universitꢀt zu Kçln
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
R2H;1 ꢀ 2RH;1RF;1 þ R2F;1 þ 4sH2F
D ¼
Department fꢁr Chemie
Greinstr. 4, 50939 Kçln (Germany)
E-mail: ralf.giernoth@uni-koeln.de
2
Here the relaxation of each individual nucleus is treated sepa-
rately. Since the major part of the relaxation stems from the
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/cphc.201100622.
ChemPhysChem 2012, 13, 261 – 266
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
261

R. Giernoth et al.
rates of, for example, the rigid imidazolium ring protons differ
substantially from the ones of the flexible alkyl side chain. Up
to date, no-one seems to have taken these differences into ac-
count.
The aim of this work is to survey the applicability of NOE
spectroscopy for the investigation of macromolecular struc-
tures in the ionic-liquid phase. Our test substrates were the
ionic liquids 1-butyl-3-methylimidazolium tetrafluoroborate
([bmim]BF4) 1, 1-butyl-3-methylimidazolium hexafluorophos-
phate ([bmim]PF6) 2 and 1-ethyl-3-methylimidazolium tetra-
fluoroborate ([emim]BF4) 3 (Figure 1). These ILs are among the
Figure 2. 19F,1H HOESY NMR spectrum of neat [bmim]BF4; tm =600 ms.
Figure 1. The three ionic liquids analyzed.
most cited in the chemical literature and their anions are
spherical. Since H,F interactions are based on directional
forces, the geometrical differences in anion/cation alignment
can be neglected.
For each IL degassed as well as non-degassed samples were
examinated in [D6]DMSO, in CD2Cl2 and completely neat (with-
out any solvent) in the following combinations:
[bmim]BF4
Figure 3. 19F,1H HOESY built-up curves for neat [bmim]BF4.
*
in CD2Cl2 (degassed and non-degassed)
*
in [D6]DMSO (non-degassed)
*
neat (degassed and non-degassed)
respect to each proton, the intensities finally have to be
weighted by the number of the corresponding nuclei.
The relative intensities were then treated three times in a
different way. The range of tm <300 ms was fitted linearly and
the whole range was fitted according to the exponential Equa-
tions (1) and (2). Figure 4 shows all three fits in one diagram
for neat 1.
[bmim]PF6
*
in CD2Cl2 (non-degassed)
*
in [D6]DMSO (non-degassed)
*
neat (degassed and non-degassed)
[emim]BF4
*
neat (degassed)
2. Results and Discussion
2.1. 19F,1H HOESY Experiments
The results of our experiments are shown exemplarily for a
neat, non-degassed sample of [bmim]BF4 (see Supporting In-
formation for the other experiments). Figure 2 shows the
HOESY spectrum, the cross signals of the HOESY spectrum be-
tween the protons of the cation and the fluorine nuclei of the
anion are recognisable. The individual signal intensities repre-
sent the qualitative order of magnitude of the particular cross
relaxation and thus the intensity of interaction. Plotting these
intensities against the mixing time leads to the NOE build-up
curves shown in Figure 3. To obtain the right proportions with
Figure 4. Mathematical fits of the relative signal intensities for the H-2
proton of neat [bmim]BF4.
262
www.chemphyschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2012, 13, 261 – 266

HOESY Measurements of Ionic Liquids
The approximation of a linear behaviour is only appropriate
for very short mixing times. A better fit is obtained with the
standard textbook exponential Equation (1). However, this
function also does not satisfactorily describe the progression
of the function. The reason for this lies in the difference in the
relaxation times of the individual nuclei. Nonetheless, in many
cases this description might be sufficient for a qualitative inter-
pretation of inter-ionic interactions.
ever, under otherwise identical conditions the signals are more
intense in CD2Cl2 as opposed to those in [D6]DMSO. This has
also been reported by Dupont et al. who ran homonuclear H,H
HOESY NMR experiments on [bmim]BPh4.[21] Qualitatively, these
findings correspond to the literature. The solvent of lower po-
larity (CD2Cl2) leads to more and/or stronger ion pairing, result-
ing in a stronger signal.[9] The neat samples again are more in-
tense, and the differences between the individual cross signals
are larger. Thus, in comparison to diluted samples, a higher
quality HOESY spectrum can be recorded with neat samples in
less time.
Precise results (that are needed for the determination of reli-
able probabilities of presence) can only be provided by fitting
the relative intensities to the exponential Equation (2). This
function considers all individual relaxation rates R of the pro-
tons and of the fluorine nuclei. Since the relaxation rates are
equivalent to the longitudinal (spin-lattice) relaxation rates, the
factors RF,X and RH,X can be obtained from the standard spin-
echo experiment. Thus, the fit can be adjusted exactly to the
experimental data (Figure 4).
Relative Intensities and Cross Relaxation
The relative intensities of the cross relaxation in solution as
well as neat between the protons of 1 and 2 and their corre-
sponding anions are very similar. For both ILs the strongest
NOE (i.e. the shortest inter-ionic distance) is observable for the
H-2 proton (see Figure 5 for the labeling of protons). The
second strongest cross relaxation takes place between H-1’
and the anion. The protons H-4 and H-5 also show a strong in-
teraction, but it is weaker than the first two. The weakest inter-
action can be observed for the protons of the alkyl chain. An
overview is given in Figure 6.
2.2. Interpretation of the Build-Up Curves
Effect of Oxygen on the NOE Build-Up
A selection of our 19F,1H HOESY experiments has been run both
under degassed and non-degassed conditions. It is widely re-
cognized that oxygen from the air as an additional source for
relaxation can significantly interfere with cross relaxation meas-
urements and therefore minimize the NOE.[25] Our results show
that it is possible to record NOE build-up curves in the pres-
ence of oxygen, although the overall relaxation in non-de-
gassed samples is faster. This in turn leads to much shorter ac-
cumulation times for the experiments.
The reason for a weaker overall NOE in the presence of air
can be understood by assuming that paramagnetic oxygen
provides just another relaxation pathway. This leads to a faster
decay of the measured overall NOE but does not influence the
relative NOE between the individual nuclei. The resulting
weaker overall signal intensities are thus not relevant for inves-
tigations in neat ILs. The relaxation for each nucleus can be de-
termined experimentally and is used already as a coefficient in
the mathematical analysis of the
Figure 5. General assignment of the proton signals for imidazolium deriva-
tives.
For the interpretation of these data one important fact
needs to be considered. ILs are liquid and therefore have a
flexible structure. The relative distances are rather preferred
probabilities of presence between the anion and cation. None-
NOE build-up curves. In summa-
ry, there is no need for further
adjustments since the formulas
in use are already derived from
the Solomon–Redfield equa-
tions[26] considering the influ-
ence of longitudinal relaxation
on NOE build-up.
Influence of the Solvent and
Comparison with the Neat
Samples
The measurements of the ILs 1
and
2 in [D6]DMSO and in
CD2Cl2 show equal relative in-
Figure 6. Relative cross relaxation rates between the protons and the anion. In each case the strongest interaction
tensities in both solvents. How- (H-2) has been used for normalization. * Unrealistic high value, most possibly due to an experimental error.
ChemPhysChem 2012, 13, 261 – 266
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
263

R. Giernoth et al.
theless, they agree well with the structures determined by
quasielastic neutron scattering[27] and MD simulations.[28]
For [emim]BF4 (3) the results are different, as depicted in
Figure 7. Here, the interactions between all protons and the
anion are almost equal.
the same group extended their study, combining proton NMR,
X-ray diffraction, conductivity and microcalorimetry measure-
ments. Dupont then concluded that these ILs were best de-
scribed as a supramolecularly organized polymeric structure.[30]
Early classical Monte Carlo calculations on [bmim]PF6 already
showed a high degree of long-range order in the IL.[31] The au-
thors attributed this result to the long-range nature of Cou-
lombic forces. Ab initio density functional theory (DFT) calcula-
tions throughout demonstrated anion–cation proximities that
resemble our HOESY results.[32] The same holds true for the MD
simulations.[28,29] In all cases the anion showed preferred prox-
imity to H-2, H-4 and H-5 of the imidazolium ring compared to
the protons of the alkyl chain. Solid-state structures that are
available for some ILs point in the same direction.[30]
Figure 7. Relative cross relaxation rates between the protons and the anion
in neat [emim]BF4. The strongest interaction has been used for normaliza-
tion.
Dielectric spectra with [bmim]Tf2N showed no sign of long-
lived ion pairs.[33] Nonetheless, the authors report “transient ion
pairs in the cage of counter ions”. A more recent study by Na-
kamura also varied the anion.[34] By comparison with 13C relaxa-
tion data the authors were able to assign the different vibra-
tional modes. The fastest of these resembled the interaction
between the anion and cation while the slowest and some of
the intermediate modes were due to the rotation of ion pairs.
Aggregate formation of ionic liquids in organic solvents has
also been studied by mass spectrometry. Kragl and co-workers
reported a decrease in the size of the formed aggregates with
an increase in solvent polarity and a decreased IL concentra-
tion.[35] Bini et al. reported mixed networks of cations and
anions[36] and were able to build a scale of the strength of the
interaction solely based on ESI-MS measurements. Recently,
Kennedy and Drummond have observed aggregate formation
with protic ILs using the same technique. They claimed that
this contributes considerably to the solvent structuring.
This finding is due to the symmetrical structure of the 1-
ethyl-3-methylimidazolium cation. The MD simulations done
by Qiao et al.[29] again exactly match our results, especially the
slightly higher interactions between the H-1’ protons in com-
parison to the protons of the alkyl chain (Figure 8).
3. Conclusions
Figure 8. Three-dimensional spatial distribution functions (SDFs) of anions
around a cation. The average distances are r=0–0.59 nm for dark isosurfaces
and 0–0.73 nm for light ones. (Reprinted with permission from F. Dommert,
J. Schmidt, B. Qiao, Y. Zhao, C. Krekeler, L. D. Site, R. Berger, C. Holm, J. Chem.
Phys. 2008, 129, 224501. Copyright 2008, American Institute of Physics.)
We studied the inter-ionic interactions of three ionic liquids by
means of 19F,1H HOESY NMR spectroscopy for the first time. For
the neat ILs preferred probabilities of presence for the anions
relative to the cations were found, confirming earlier predic-
tions through MD simulations.[28,29] Measurements of dilute
samples demonstrated the increase of ion pairing in less polar
solvents in the expected way.[37] Even in the highly polar sol-
vent DMSO a sufficiently large number of ion pairs were
found. The agreement of the results with a whole range of lit-
erature data proves the validity of our method, even though
we are dealing with a completely different time scale.
Since the strongest interaction between the cation and
anion is based on Coulombic forces,[24] these interactions are
not restricted to the plane of the cation (which would be the
case if H-bonding would be the dominant force). The iso-surfa-
ces show a nearly spherical spatial distribution function (SDF)
that is also in agreement with our findings that the strongest
interactions are in a defined radius.
Comparative studies with degassed and non-degassed sam-
ples indicate that the influence of oxygen only leads to a faster
relaxation of the whole system while the relative intensities
stay unchanged. However, the faster relaxation times enable
us to shorten the duration of the experiments significantly.
This increases the resolution of the experiments despite the
shorter measurement time (hours instead of days for full build-
up curves or minutes instead of hours for single experiments,
respectively). The loss in signal intensity is irrelevant when in-
vestigating neat ionic liquids.
Comparison of our Findings with Literature Results
As already mentioned, the first to study ion-pairing of imidazo-
lium ILs in solution by NOE NMR spectroscopy was the Dupont
group.[21] They found H,H NOEs as a proof for ion pairs and
they also registered stronger signals in dichloromethane in
contrast to dimethyl sulfoxide. The results match our findings,
but they were naturally limited to H–H contacts. In ref. [30],
264
www.chemphyschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2012, 13, 261 – 266

HOESY Measurements of Ionic Liquids
1-Ethyl-3-methylimidazolium tetrafluoroborate 3 ([emim]BF4) was
synthesized from 1-ethyl-3-methylimidazolium bromide and
sodium tetrafluoroborate. 11.2 g (58.6 mmol, 1 equiv) 1-ethyl-3-
methylimidazolium bromide and 6.59 g (60.0 mmol, 1.02 eq)
sodium tetrafluoroborate were dissolved in acetone. After stirring
for 12 h the resulting precipitate was filtered off and the solvent
was removed under reduced pressure. The crude product was dis-
solved in dichloromethane and stirred with activated charcoal.
After filtration the solvent was removed to give the product in
50% yield. The water content (234 ppm) was determined by Karl-
Fischer titration. Ion chromatography showed no residue halides or
Application of a more precise exponential fit of the NOE
build-up curves resulted in a much higher quality of the quan-
titative data. By this we found a high degree of order in the
ionic-liquid phase (at least for the first two shells with respect
to a given ion). The dominant mode of interaction is hydrogen
bonding between the C2 proton of the imidazolium ring and
the anion but we see no hint of p stacking.
In summary, we presented a tool for the investigation of
liquid-ionic phases with the help of NMR spectroscopy only,
which should be available in every chemistry laboratory. There-
fore, the method is applicable by virtually everyone to all kinds
of questions regarding interactions in ionic liquid phases (e.g.
of chiral ion pairs used for catalysis[38]). We are currently work-
ing on the quantitative interpretation of the NOE data and the
implementation of new NMR experiments for this purpose.
1
amines. H NMR ([D6]DMSO, 300 MHz; numbers refer to protons as
labeled in Figure 5; d is in ppm in all cases): d=9.29 (s, br, 1H, CH,
H-2), 7.84 (s, 1H, H-4), 7,74 (s, 1H, H-5), 4.21 (quart, 2H, J=7.35 Hz,
H-1“), 3.86 (s, 3H, H-1’), 1.41 (t, 3H, J=7.35 Hz, H-2”). 13C NMR
([D6]DMSO, 75 MHz; numbers refer to carbons as labeled in
Figure 5):) d=136.1 (s, C-2), 123.5 (s, C-4),122.0 (s, C-5), 44.1 (s,
C-1“), 35.8 (s, C-1), 15.1 (s, C-2”).
NMR experiments: All NMR experiments were run on a Bruker
Avance 400 spectrometer equipped with an L. T. TBI 5 mm four-nu-
Experimental Section
2
1
cleus probe with z-gradient (1H, 19F, H, X). H chemical shifts were
Synthesis of the Ionic Liquids: The ionic liquid 3-butyl-1-methylimi-
dazolium tetrafluoroborate 1 ([bmim]BF4) was synthesized from 3-
butyl-1-methylimidazolium bromide through anion exchange with
tetrafluoroboric acid. The exchange was carried out by dissolving
15.15 g (50 mmol) 3-butyl-1-methylimidazolium bromide in water
and the addition of 4.04 mL (52.00 mmol) tetrafluoroboric acid.
After stirring for 24 h the solvent was removed and the residue
was washed with water until no free halides and amine residues
could be detected anymore (determined by ion chromatography).
After drying under reduced pressure, the crude product was dis-
solved in dichloromethane and dried with magnesium sulfate.
After filtration the solvent was removed and the product was ob-
tained in 59% yield. The water content (134 ppm) was determined
by Karl-Fischer titration. 1H NMR ([D6]DMSO, 400 MHz; numbers
refer to protons as labeled in Figure 5, d is in ppm in all cases): d=
9.04 (s, br, 1H, CH, H-2), 7.73 (s, 1H, H-4), 7,66 (s, 1H, H-5), 4.16 (t,
2H, J=7.19 Hz, H-1“), 3.85 (s, 3H, H-1), 1.77 (quint, 2H, J=7.42 Hz,
H-2”), 1.265 (sext, 2H, J=7,5 Hz, H-3“), 0.89 (t, 3H, J=7,39, H-4”).
13C NMR ([D6]DMSO, 100 MHz; numbers refer to carbons as labeled
in Figure 5): d=136.6 (s, C-2), 123.7 (s, C-4),122.3 (s, C-5), 48.7 (s, C-
1“), 35.8 (s, C-1), 31.4(s, C-2”) 18.87 (s, C-3“) 13.30 (s, C-4”). 19F NMR
([D6]DMSO, 376 MHz) d=ꢀ148.3 (m, BF4).
3-butyl-1-methylimidazolium hexafluorophosphate 2 ([bmim]PF6)
was synthesized accordingly through anion exchange of 3-butyl-1-
methylimidazolium bromide with hexafluorophosphoric acid.
4.02 mL (45.50 mmol) hexafluorophosphoric acid were added to a
solution of 5.48 g (25 mmol) 1-butyl-3-methylimidazolium bromide
in water. The solvent was decanted and the remaining crude prod-
uct was washed with water until no free halides and amine resi-
dues could be detected anymore (determined with ion chromato-
graphy). After drying under reduced pressure the product was dis-
solved in dichloromethane and dried with magnesium sulfate.
After filtration and removal of the solvent the product was ob-
tained in 79% yield. The water content (152 ppm) was determined
by Karl-Fischer titration. 1H NMR ([D6]DMSO, 400 MHz; numbers
refer to protons as labeled in Figure 5, d is in ppm in all cases): d=
9.07 (s, br, 1H, CH, H-2), 7.72 (s, 1H, H-4), 7,66 (s, 1H, H-5), 4.18 (t,
2H, J=7.19 Hz, H-1“), 3.86 (s, 3H, H-1), 1.79 (quint, 2H, J=7.42 Hz,
H-2”), 1.29 (sext, 2H, J=7,5 Hz, H-3“), 0.92 (t, 3H, J=7,39, H-4”).
13C NMR ([D6]DMSO, 100 MHz; numbers refer to protons as labeled
in Figure 5): d=136.2 (s, C-2), 123.7 (s, C-4),122.3 (s, C-5), 48.7 (s,
C-1“), 35.8 (s, C-1), 31.4 (s, C-2”) 18.87 (s, C-3“) 13.30 (s, C-4”).
19F NMR ([D6]DMSO, 376 MHz) d=69.15, 71.24 (d, PF6).
1
referenced to tetramethyl silane with respect to the H resonance
of the respective solvent (dimethylsulfoxid-d6, [D2]dichloro-
methane). Neat samples were referenced to an external standard
using a coaxial outer tube. The 19F signals are reported relative to
the resonance of a neat sample of CFCl3. Each experiment was car-
ried out at 298 K if not stated otherwise. T1 relaxation times were
measured using the standard t1rpg inversion-recovery pulse se-
quence of Bruker. T1 data were fitted to an exponential decay in
the form of Equation (4):
MðtÞ ¼ ða0 ꢀ a2Þeðꢀt=T Þ þ a2
1
ð4Þ
where a0 corresponds to the equilibrium magnetization and a2 to
the magnetization immediately following the inversion pulse,
using the XWINNMR software package provided by Bruker.
For the 19F,1H HOESY experiments the hoesygp pulse sequence was
used. All experiments were run under fluorine detection. An in-
verse experiment showed identical results except for a slightly
longer experiment duration. The resulting two-dimensional data
were processed using a Gaussian transformation function (imple-
mented in the Xwinnmr software) in both dimensions prior to
Fourier transformation.
The experiments on solutions of 1 and 2 were performed using a
sample of 28 mgmLꢀ1 in the respective deuterated solvent. The
ionic liquid
1 was investigated in [D6]dimethylsulfoxid and
[D2]dichloromethane. The latter experiment was performed twice,
degassed (using the freeze-pump-thaw technique) and non-de-
gassed.
Neat samples of 1, 2 and 3 were sealed in a coaxial inner tube. The
outer tube (a regular 5 mm bore NMR tube) was filled with
[D6]DMSO as the source for the lock signal.
For the 19F,1H HOESY experiments 4 K data points were acquired in
the directly detected dimension. The mixing time was chosen be-
tween 25 and 5000 ms for the non-degassed and between 0 and
1200 ms for the degassed samples. The delay between two experi-
ments was between 5 to 10 s, chosen after determination of the T1
relaxation times to ensure complete relaxation.
ChemPhysChem 2012, 13, 261 – 266
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemphyschem.org
265

R. Giernoth et al.
[19] T. C. Pochapsky, P. M. Stone, S. S. Pochapsky, J. Am. Chem. Soc. 2002,
124, 1460–1462.
Acknowledgements
[20] P. S. Pregosin, Pure Appl. Chem. 2009, 81, 615–633.
[21] J. Dupont, P. S. Souza, R. Burrow, J. Kintzinger, Chem. Eur. J. 2000, 6,
2377–2381.
[22] A. Mele, G. Romanꢂ, M. Giannone, E. Ragg, G. Fronza, G. Raos, V.
Marcon, Angew. Chem. 2006, 118, 1141–1144; Angew. Chem. Int. Ed.
2006, 45, 1123–1126.
We would like to thank Dr. Nils Schlçrer for technical support
with the NMR spectrometers, F. Dommert, Prof. R. Berger and
Prof. C. Holm for fruitful discussions and DFG for funding via the
Priority Programme 1191 “Ionic Liquids”.
[23] D. Nama, P. Kumar, P. Pregosin, T. Geldbach, P. Dyson, Inorg. Chim. Acta
2006, 359, 1907–1911.
[24] H. Weingꢃrtner, Angew. Chem. 2008, 120, 664–682; Angew. Chem. Int.
Ed. 2008, 47, 654–670.
Keywords: ionic liquids
relaxation · solvent effects
· NMR spectroscopy · NOE ·
[25] T. D. W. Claridge, High-Resolution NMR Techniques in Organic Chemistry,
2nd ed., Elsevier, Oxford, 2009, p. 256.
[26] A. G. Redfield, Adv. Magn. Reson. 1965, 1, 1.
[1] P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis, 2nd, completely re-
vised edition, Wiley-VCH, Weinheim, 2007.
[2] N. V. Plechkova, K. R. Seddon, Chem. Soc. Rev. 2008, 37, 123–150.
[3] R. Giernoth, Angew. Chem. 2010, 122, 5740–5741; Angew. Chem. Int. Ed.
2010, 49, 5608–5609.
[4] R. Giernoth, D. Bankmann, N. Schloerer, Green Chem. 2005, 7, 279–282.
[5] R. Giernoth, D. Bankmann, Eur. J. Org. Chem. 2005, 4529–4532.
[6] R. Giernoth, D. Bankmann, Tetrahedron Lett. 2006, 47, 4293–4296.
[7] D. Bankmann, R. Giernoth, Prog. Nucl. Magn. Reson. Spectrosc. 2007, 51,
63–90.
[8] R. Giernoth, D. Bankmann, Eur. J. Org. Chem. 2008, 2881–2886.
[9] A. W. Overhauser, Phys. Rev. 1953, 92, 411–415.
[10] D. G. Davis, J. Am. Chem. Soc. 1987, 109, 3471–3472.
[11] D. Neuhaus, M. P. Williamson, The Nuclear Overhauser Effect in Structural
and Conformational Analysis, 2nd ed., Wiley, New York, 2000.
[12] C. Yu, G. C. Levy, J. Am. Chem. Soc. 2002, 124, 6533–6537.
[13] I. Solomon, Phys. Rev. 1955, 99, 559–566.
[27] C. Hardacre, J. D. Holbrey, M. Nieuwenhuyzen, T. G. A. Youngs, Acc.
Chem. Res. 2007, 40, 1146–1155.
[28] F. Dommert, J. Schmidt, B. Qiao, Y. Zhao, C. Krekeler, L. D. Site, R. Berger,
C. Holm, J. Chem. Phys. 2008, 129, 224501.
[29] B. Qiao, C. Krekeler, R. Berger, L. D. Site, C. Holm, J. Phys. Chem. B 2008,
112, 1743–1751.
[30] C. Consorti, P. Suarez, R. de Souza, R. Burrow, D. Farrar, A. Lough, W.
Loh, L. da Silva, J. Dupont, J. Phys. Chem. B 2005, 109, 4341–4349.
[31] J. K. Shah, J. F. Brennecke, E. J. Maginn, Green Chem. 2002, 4, 112–118.
[32] T. I. Morrow, E. J. Maginn, ACS Symp. Ser. 2003, 856.
[33] C. Daguenet, P. Dyson, I. Krossing, A. Oleinikova, J. Slattery, C. Wakai, H.
Weingꢃrtner, J. Phys. Chem. B 2006, 110, 12682–12688.
[34] K. Nakamura, T. Shikata, ChemPhysChem 2010, 11, 285–294.
[35] S. Dorbritz, W. Ruth, U. Kragl, Adv. Synth. Catal. 2005, 347, 1273–1279.
[36] R. Bini, O. Bortolini, C. Chiappe, D. Pieraccini, T. Siciliano, J. Phys. Chem.
B 2007, 111, 598–604.
[14] R. L. Vold, J. S. Waugh, M. P. Klein, D. E. Phelps, J. Chem. Phys. 1968, 48,
3831–3832.
[37] J. T. Gerig, J. Org. Chem. 2003, 68, 5244–5248.
[15] C. Zuccaccia, G. Bellachioma, G. Cardaci, A. Macchioni, Organometallics
1999, 18, 1–3.
[38] P. S. Schulz, N. Mꢄller, A. Bçsmann, P. Wasserscheid, Angew. Chem. 2007,
119, 1315–1317; Angew. Chem. Int. Ed. 2007, 46, 1293–1295.
[16] C. Hofstetter, P. S. Wilkinson, T. C. Pochapsky, J. Org. Chem. 1999, 64,
8794–8800.
[17] C. Zuccaccia, A. Macchioni, I. Orabona, F. Ruffo, Organometallics 1999,
18, 4367–4372.
Received: August 12, 2011
[18] T. C. Pochapsky, P. M. Stone, J. Am. Chem. Soc. 1990, 112, 6714–6715.
Published online on October 14, 2011
266
www.chemphyschem.org
ꢁ 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemPhysChem 2012, 13, 261 – 266