PHIP NMR spectroscopy in ionic liquids: Influence of salts on the intensity of polarization signals

Article abstract of DOI:10.1021/ac502537z

Parahydrogen-induced dynamic nuclear polarization NMR spectroscopy (PHIP) in ionic liquids leads to weak or no polarization signals, depending on the type of experiment. We demonstrate that the intensity of polarization is directly correlated to the concentration of the ionic liquids. High ion concentration is connected to fast T₁ relaxation, resulting in annihilation of the polarization signals.

Full text of DOI:10.1021/ac502537z

Article
pubs.acs.org/ac
PHIP NMR Spectroscopy in Ionic Liquids: Influence of Salts on the
Intensity of Polarization Signals
Andreas Brohl and Ralf Giernoth*
̈
Department fur Chemie, Universitat zu Koln, Greinstr. 4, D-50939 Koln, Germany
̈
̈
̈
̈
S
* Supporting Information
ABSTRACT: Parahydrogen-induced dynamic nuclear polarization NMR
spectroscopy (PHIP) in ionic liquids leads to weak or no polarization signals,
depending on the type of experiment. We demonstrate that the intensity of
polarization is directly correlated to the concentration of the ionic liquids. High
ion concentration is connected to fast T1 relaxation, resulting in annihilation of
the polarization signals.
HIP NMR spectroscopy (parahydrogen-induced polar-
ization) is a powerful method for studying homogeneous
atmosphere. The NMR tube was connected to a 1.2 m glass
P
tube with the help of a small piece of rubber tubing as
connector. Degassed (“freeze-pump-thaw”) acetone-d6 and
degassed IL were then inserted into the NMR tube through
the rubber connector with the help of a syringe. The mixture
was homogenized with the help of a vortex mixer.
NMR Measurements. The construction containing a NMR
tube, rubber tubing connector, and a glass tube was provided
with two NMR spinners and transferred into the NMR magnet.
The end of the glass tube protruded approximately 10 cm
above the top of the magnet. Parahydrogen was produced by
letting H2 gas pass through a U-shaped metal tube that was
filled with activated charcoal and cooled with the help of liquid
nitrogen. The outlet of this construction was connected to an
HPLC tube which was used to bubble para-enriched hydrogen
directly into the NMR tube. A sketch of this experimental setup
can be found in the Supporting Information.
hydrogenation reactions.1−6 It is conducted by hydrogenating
inside the NMR magnet using para-enriched hydrogen. In this
way, polarization signals arise, given that the hydrogen
molecule is transferred to the substrate in a pairwise fashion.
The antiphase pattern of the resulting PHIP signals contains
information about the reaction mechanism, whereas a strong
enhancement of the signal intensity enables the detection of
intermediates and products even in very low concentration.
Advancements of the method are currently focusing on
selective signal enhancement for NMR tomography.7,8
Homogeneous hydrogenation reactions were among the first
being studied in ionic liquids (ILs) as the solvents.9,10 To quite
a surprise, even substrates with very limited solubility gave
remarkable results. In the following, ionic liquids proved
especially valuable for biphasic processes in which the catalyst
was heterogenized in the ionic phase. Consequently, hydro-
genation reactions in ILs should be perfectly suitable to be
studied using PHIP NMR spectroscopy (Figure 1).
In a typical PHIP experiment, 0.7 mL of acetone-d6 was
added to 60 mg of substrate, 1 mg of catalyst, and the respective
amount of salt. After placing the tube into the NMR magnet,
the experiment was conducted. The spectrum was recorded on
EXPERIMENTAL SECTION
■
Sample Preparation. Solids (MAC and the catalyst) and, if
applicable, salt were placed into an NMR tube in an argon
1
a Bruker AV400 spectrometer after a single 45° H pulse. T1
relaxation data have been acquired through a standard
inversion−recovery experiment.
All experimental procedures and additional data can be found
in the Supporting Information.
Received: July 10, 2014
Accepted: September 27, 2014
Figure 1. Structures of the ionic liquids that have been used in this
study.
© XXXX American Chemical Society
A
dx.doi.org/10.1021/ac502537z | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
RESULTS AND DISCUSSION
■
To validate our experimental setup, we first focused on the
standard hydrogenation substrate MAC (methyl-Z-α-acetamido
cinnamic acid). PHIP hydrogenation of MAC in acetone-d6
using [Rh(cod)(dppb)][BF4] as the catalyst gave the expected
polarization signals, as depicted in Figure 2.
Figure 3. Hydrogenation of MAC under PHIP conditions. Plot of the
total area of the antiphase signals (i.e., intensity of polarization signals)
versus the mole fraction of IL.
Because we had already performed experiments in diluted
solution, we know that IL viscosity cannot have anything to do
with the nonappearance of the polarization signals. To check
whether the presence of a large concentration of salt is
responsible for the disappearance of these signals, we repeated
the experiments with varying mole fractions of LiNTf2, a
crystalline salt. Indeed, the results were almost identical to the
ones with ILs (Table 2 and Figure 4).
Figure 2. PHIP NMR spectrum during the hydrogenation of methyl-
Z-α-acetamido cinnamic acid (MAC) in acetone-d6 with para-enriched
hydrogen.
In the following, we hydrogenated MAC under PHIP
conditions in varying acetone/IL mixtures. (The structures of
the ILs are depicted in Figure 1, above.) The results are listed in
Table 1. In Figure 3, we plotted the intensity of the polarization
Table 2. PHIP Hydrogenation of MAC in Acetone with
Varying Mole Fractions of LiNTf2
a
b
entry
m(salt) [mg]
χ(salt)
A
1
2
3
4
5
6
0
30
0
1.91
1.10
0.58
0.26
0.15
0
Table 1. PHIP Hydrogenation of MAC in Acetone with
Varying Mole Fractions of IL
0.011
0.036
0.082
0.144
0.230
a
100
246
464
910
V(IL)
[mL]
V(acetone)
[mL]
b
IL
entry
χ(IL)
A
c
[C1C2im][P(C2F5)3F3]
1
2
3
4
5
1
2
3
4
0
0.7
0.6
0.5
0.4
0.3
0.7
0.6
0.5
0.4
0
4.07
1.45
0.92
0.52
0
a
0.1
0.2
0.3
0.4
0
0.036
0.082
0.144
0.230
0
Reaction conditions: Varying amounts of LiNTf2 were added to 1 mg
of catalyst ([Rh(cod)(dppb)][BF4]), 60 mg of MAC and 0.7 mL of
acetone-d6 in a standard 5 mm NMR tube. Total area of the
antiphase signals (i.e., signal intensity)
b
d
[C1C2im][CH3SO3]
4.10
0.31
0.07
0
0.1
0.2
0.3
0.068
0.150
0.248
A combination of the data in one single graph in which the
total intensities are normalized demonstrates the close
resemblance of the two experiments (Figure 5).
a
Reaction conditions: 1 mg of catalyst ([Rh(cod)(dppb)][BF4]), 60
mg of MAC, and a mixture of acetone-d6 and IL given by χ in a
b
standard 5 mm NMR tube. Total area of the antiphase signals (i.e.,
c
signal intensity). 1-Ethyl-3-methylimidazolium tris(penta-
d
fluoroethyl)trifluorophosphate. 1-Ethyl-3-methylimidazolium methyl
sulfonate.
signals (represented by the total area of the antiphase signals)
versus the mole fraction of IL. The mole fraction is given as
n(IL)
χ(IL) =
n(IL) + n(acetone)
Obviously, the IL concentration has a strong influence on the
intensity of the polarization signals. With a mole fraction of
0.23, no polarization is detectable anymore.
The strong decay in the case of [C2C1im][CH3SO3] can be
attributed to the strong coordination ability of the anion,
deactivating the catalyst.11
Figure 4. Hydrogenation of MAC under PHIP conditions. Plot of the
total area of the antiphase signals (i.e., intensity of polarization signals)
versus the mole fraction of LiNTf2.
B
dx.doi.org/10.1021/ac502537z | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
this case, no PHIP signals were detectable at a mole fraction of
0.23. These findings can be aligned with the very similar signal
decay in the case of LiNTf2 as additive (cf. Figures 4 and 5).
In the view of the sum of all our findings, the reason for the
nonappearance of the PHIP signals most probably arises from
the relaxation behavior at high ion concentrations. The T1
relaxation times of the protons in MAC are in the range of 4.5−
2.5 s without any additives and in the range of 1.2 s−850 ms at
a LiNTf2 mole fraction of 0.23 (cf. Table 3 and Figure 7). The
Table 3. T1 Relaxation Times of MAC (cf. Figure 2) at
Different LiNTf2 Concentrations
χ(LiNTf2)
H-1′
H-6
H-4
H-1
H-3′
Figure 5. Combined plot of the normalized data from Figure 2 and 3
0
3.919 s
3.302 s
2.888 s
2.156 s
1.623 s
1.219 s
4.548 s
3.868 s
3.263 s
2.370 s
1.559 s
904.8 ms
4.191 s
3.309 s
3.234 s
2.370 s
1.559 s
904.8 ms
2.467 s
2.038 s
1.638 s
1.289 s
929.2 ms
967.4 ms
2.444 s
2.206 s
1.645 s
1.198 s
916.0 ms
857.4 ms
0.011
0.036
0.082
0.144
0.230
To elucidate the reasons for these results, we first examined
the conversion data. For this purpose, MAC was hydrogenated
under the same conditions as during the PHIP experiments
without any ionic additive and with a content of 23 mol %
[C2C1im][P(C2F5)3F3] and LiNTf2, respectively. This mole
fraction was the limit at which no polarization signals could be
detected at all.
We found that there is no crucial effect of the ionic additives
concerning the conversion of MAC (cf. Figure 6). Therefore, a
too low concentration of the hydrogenated product at high ion
concentrations cannot be responsible for the lack of the
polarization signals.
Figure 7. Decay of the T1 relaxation times of the protons in MAC, the
trend of the phenylic protons (H-6) is shown exemplarily.
T2 relaxation times do not seem to be affected as there is no
line broadening at higher ion concentrations observed at all.
Hence, we deduce that, at high ion concentrations, the
polarized spin system has completely relaxed into equilibrium
before the spectrum was recorded. Therefore, no NMR signals
are detectable for these groups.
Figure 6. Conversion of MAC in the absence and presence of ionic
additives, determined by H NMR spectroscopy. In each case, the
amount of hydrogenated product is high enough to be potentially
detected via PHIP NMR.
1
These findings are also suitable for the explanation of the
observations in recent literature, where three groups have been
trying to perform PHIP hydrogenations in ionic liquids. Stark
and Buntkowsky studied the hydrogenation of ethyl acrylate in
−
Hydrogenation experiments with Wilkinson’s catalyst
resulted either in a poor solubility of the catalyst (in acetone-
d6 and acetonitrile-d3), a poor solubility of MAC and IL/salt (in
benzene-d6) or in no conversion despite of complete solubility
(in THF-d8 and DCM-d2, in the case of THF probably because
of the strong coordination, deactivating the catalyst). As the
primarily chosen system acetone-d6/[Rh(cod)(dppb)][BF4]
resulted in good conversions, the other attempts were not
taken into further consideration.
ionic liquids containing the anion NTf2 .12 They were only able
to detect PHIP signals in cases where the IL was present in very
low concentration (substrate/IL ratio of 100:1). With high IL
concentrations (where the IL could accurately be called a
solvent), no PHIP was detected. The authors attributed this
behavior predominantly to the high viscosity of the ionic liquid
and the result of low mass transport.
In contrast, Koptyug and co-workers have studied supported
ionic liquid phase (SILP) hydrogenation of propyne,13 and
As imidazolium-based ionic liquids are known for undergoing
H/D exchange reactions, we noticed the possibility that
parahydrogen could possibly be “deactivated” (i.e., the para/
ortho equilibrium might be shifted to the ortho form) by such
exchange reactions. As a consequence, we performed the PHIP
measurements of MAC with [C4C1pyrr][BF4] (1-butyl-1-
methylpyrrolidinium tetrafluoroborate) as additive. Also in
Klankermayer and Blumich have investigated the SILP
̈
hydrogenation of propene using PHIP NMR.14 In SILP
catalysis, an ionic liquid containing dissolved catalyst is applied
to the surface of a solid support.15 In this fashion, poor mass
transport is circumvented, making catalytic reactions in ILs
more effective. The authors were indeed able to detect PHIP
signals, although it is important to note that these are gas-phase
C
dx.doi.org/10.1021/ac502537z | Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry
Article
(14) Gong, Q.; Klankermayer, J.; Blumich, B. Chem.Eur. J. 2011,
̈
reactions, and in their experimental setup, the NMR detection
took place outside the ionic phase. Still, the IL/catalyst ratio
had a strong influence on the PHIP intensity.
17, 13795−13799.
(15) Mehnert, C. P. Chem.Eur. J. 2005, 11, 50−56.
In the case of Stark and Buntkowsky, where no PHIP signals
could be detected at high IL concentrations, the hydrogenated
product was still in contact with the IL, leading to very fast
relaxation. The cases of Koptyug and Klankermayer,
respectively, both concern gas-phase hydrogenations where
the hydrogenated products were not in contact with the IL
during the acquisition. So, the T1 times were not shortened,
and PHIP signals could be observed.
CONCLUSIONS
■
In conclusion, we have demonstrated that the appearance of
polarization signals in PHIP NMR spectroscopy can be
suppressed in the presence of high ion concentrations, most
probably due to fast T1 relaxation.
ASSOCIATED CONTENT
* Supporting Information
Additional experimental data. This material is available free of
charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail: ralf.giernoth@uni-koeln.de. Phone: +49-221-4703094.
■
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We thank Deutsche Forschungsgemeinschaft (DFG) for
funding via the Special Priority Programme 1191 “Ionic
Liquids”, the “Fonds der Chemischen Industrie” for a graduate
fellowship for A.B., and Andrea Kuchenbuch for providing the
artwork for the table of contents.
REFERENCES
■
(1) Bowers, C. R.; Weitekamp, D. P. J. Am. Chem. Soc. 1987, 109,
5541−5542.
(2) Eisenschmid, T. C.; Kirss, R. U.; Deutsch, P. P.; Hommeltoft, S.
I.; Eisenberg, R.; Bargon, J.; Lawler, R. G.; Balch, A. L. J. Am. Chem.
Soc. 1987, 109, 8089−8091.
(3) Pravica, M. G.; Weitekamp, D. P. Chem. Phys. Lett. 1988, 145,
255−258.
(4) Giernoth, R.; Hubler, P.; Bargon, J. Angew. Chem. 1998, 110,
̈
2649−2651.
(5) Duckett, S. B.; Sleigh, C. J. Prog. Nucl. Mag. Res. 1999, 34, 71−92.
(6) Hubler, P.; Giernoth, R.; Kummerle, G.; Bargon, J. J. Am. Chem.
̈
̈
Soc. 1999, 121, 5311−5318.
(7) Reineri, F.; Santelia, D.; Viale, A.; Cerutti, E.; Poggi, L.; Tichy, T.;
Premkumar, S. S. D.; Gobetto, R.; Aime, S. J. Am. Chem. Soc. 2010,
132, 7186−7193.
(8) Waddell, K. W.; Coffey, A. M.; Chekmenev, E. Y. J. Am. Chem.
Soc. 2011, 133, 97−101.
(9) Chauvin, Y.; Mussmann, L.; Olivier, H. Angew. Chem. 1995, 107,
2941−2943.
(10) Suarez, P. A. Z.; Dullius, J. E. L.; Einloft, S.; deSouza, S. Z.
Polyhedron 1996, 15, 1217−1219.
(11) Lungwitz, R. Ph.D. Thesis, Chemnitz University of Technology:
Chemnitz, Germany, Feb. 11, 2011.
(12) Gutmann, T.; Sellin, M.; Breitzke, H.; Stark, A.; Buntkowsky, G.
Phys. Chem. Chem. Phys. 2009, 11, 9170−9175.
(13) Kovtunov, K. V.; Zhivonitko, V. V.; Kiwi-Minsker, L.; Koptyug,
I. V. Chem. Commun. 2010, 46, 5764−5766.
D
dx.doi.org/10.1021/ac502537z | Anal. Chem. XXXX, XXX, XXX−XXX