Synthesis of α-fluoro-α,β-unsaturated esters monitored by 1D and 2D benchtop NMR spectroscopy

Article abstract of DOI:10.1002/mrc.4843

For optimization and control of pharmaceutically and industrially important reactions, chemical information is required in real time. Instrument size, handling, and operation costs are important criteria to be considered when choosing a suitable analytica

Full text of DOI:10.1002/mrc.4843

Synthesis of α-Fluoro-α,β-unsaturated esters monitored by 1D and 2D
benchtop NMR Spectroscopy
Dennis Weidener, Kawarpal Singh and Bernhard Blümich
Institut für Technische und Makromolekulare Chemie, RWTH Aachen University,
Worringerweg 2, D-52074 Aachen, Germany
Abstract
For optimization and control of pharmaceutically and industrially important reactions
chemical information is required in real time. Instrument size, handling and operation
costs are important criteria to be considered when choosing a suitable analytical
method apart from sensitivity and resolution. This present study explores the use of
a robust and compact NMR spectrometer to monitor the stereo-selective formation of
α-fluoro-α,β-unsaturated esters from α-fluoro-β-keto esters via deprotonation and
deacylation in real-time. These compounds are precursors of various
pharmaceutically active substances. The real-time study revealed the deprotonation
and deacylation steps of the reaction. The reaction was studied at temperatures
1
1
ranging from 293 to 333 K by interleaved 1D H, 19F and 2D H-1H COSY
experiments. The kinetic rate constants were evaluated using a pseudo-first order
kinetic model. The activation energies for the deprotonation and deacylation steps
were determined to (-28 ± 2) and (63.5 ± 8) kJ/mol, respectively. This showed that
the deprotonation step is fast compared to the deacylation step and that the
deacylation step determines the rate of the overall reaction. The reaction was
repeated three times at 293 K to monitor the repeatability and stability of the system.
The compact NMR spectrometer provided detailed information on the mechanism
and kinetics of the reaction which is essential for optimizing the synthetic routes for
stepwise syntheses of pharmaceutically active substances.
Keywords: Reaction monitoring, Flow NMR, Compact NMR spectrometer,
Deacylation, Deprotonation
This article has been accepted for publication and undergone full peer review but has not
been through the copyediting, typesetting, pagination and proofreading process which may
lead to differences between this version and the Version of Record. Please cite this article as
doi: 10.1002/mrc.4843
This article is protected by copyright. All rights reserved.

1.
Introduction
For understanding, optimization and cost reduction of chemical processes in industry
and research it is important to understand their reaction mechanisms and kinetics.
Different spectroscopic and chromatographic analytical tools are used for reaction
monitoring and quality control such as Raman and ultraviolet spectroscopy [1,2],
near Infrared (NIR) [3, 4] and Fourier transform infrared (FT-IR) spectroscopy [5],
gas chromatography (GC) [6], high performance liquid chromatography (HPLC) [7],
and more. Every technique provides specific information depending upon its
sensitivity, resolution, stability and ability to probe into the chemical environment.
Nuclear magnetic resonance (NMR) spectroscopy has been used as an
analytical instrument for reaction monitoring, process control and structure
elucidation to obtain information about the various chemical groups involved in the
reaction [8-9]. High-resolution NMR spectroscopy at high-field provides quantitative
and qualitative information for monitoring chemical reactions and processes without
prior calibration. But high-field NMR instruments are large and require cryogenic
cooling gases, sophisticated laboratories, and experienced professionals for their
maintenance and operation. On the other hand current state-of-the-art high-
resolution compact NMR spectrometers have opened new possibilities for monitoring
chemical processes at the production sites and at the chemical workbench [10-11].
These instruments are maintenance free and cost one-tenth of a high-field
instrument.
The low-field compact NMR spectrometer used in this study operates with a
1
Halbach magnet at a field strength of 1 Tesla corresponding to a H NMR frequency
of 42.5 MHz and a 19F NMR frequency of 40.1 MHz. The magnet and the
spectrometer are incorporated in one console of approximate size 40  70  50 cm³
[12]. The magnetic field strength of the instrument is low compared to that of high-
field spectrometers whose proton resonance frequencies go up to 1 GHz. In spite of
this, the sensitivity and resolution of compact NMR spectrometers are sufficient to
analyze a wide range of chemical reactions [13-14] and to elucidate the structures of
small complex molecules [15]. Deuterated solvents are not required for the field
frequency lock as the lock is external to the sample bore. The helps to avoid any
potential kinetic isotope effects in the reaction which directly affect the reaction rate
and activation energy. Moreover, shorter feed lines help to transport the reaction
This article is protected by copyright. All rights reserved.

mixtures in less time through the magnet than needed for high-field spectrometers
which require long feed lines. Different reaction conditions can easily be realized,
and experiments can be performed with a click of mouse button.
α-fluoro-α,β-unsaturated esters are used as precursors to many biologically
important compounds in the syntheses of medicines, pheromones and herbicides
[16]. The Wittig, Thia-Wittig and Julia olefinations are the main approaches used for
synthesizing these compounds [17]. These approaches require expensive metal
catalyst and starting materials, their selectivity is low, and the reaction conditions are
harsh. But a new method has emerged to obtain the required product with high
selectivity and with inexpensive reaction materials [18].
In the present work the kinetics and the percentage yield of the reaction were
studied in real-time with a compact 1 T NMR spectrometer at temperatures varying
1
from 293 to 333 K by interleaving 1D H and 19F and with 2D 1H-1H COSY
experiments. The results obtained from the different experimental approaches are
compared to assess the accuracy and versatility of the method. Because the
reaction time was long, of the order of hours, one experiment was repeated three
times to assess the repeatability of the reaction and the stability of the spectrometer.
2.
Materials and Methods
Ethyl benzoyl acetate, selectfluor, benzaldehyde, cesium carbonate, sodium sulfate,
sodium hydrogen carbonate, acetonitrile, ethyl acetate, hexane and silica gel 60
were purchased from Sigma Aldrich Chemie GmbH, Steinheim, and Alfa Aesar,
Thermo Fisher (Kandel) GmbH, Karlsruhe, Germany.
3.
Experimental
3.1
Synthesis of ethyl 2-fluoro-3-oxo-3-phenylpropanoate
Selectfluor ® (1 eq., 42.3 mmol, 15.0 g) was added portion wise to a solution of ethyl
benzoyl acetate (1 eq., 42.3 mmol, 8.14 g) in 40 mL MeCN, at 293 K, and the
resulting solution was stirred for two hours. The complete conversion of the starting
materials was monitored using thin layer chromatography (TLC). The reaction
mixture was diluted with dicholoromethane (DCM) (40 mL) to separate the organic
and aqueous layers. The resulting organic layer was washed two times with
saturated sodium bicarbonate (sat. NaHCO3) solution. After separation, the organic
This article is protected by copyright. All rights reserved.

layer was dried over sodium sulphate (NaSO4) and the solvent was removed under
vacuum. The resulting yellow oil was purified with column chromatography (ethyl
1
acetate: hexane, 1:1) yielding a colorless oil. 1D H NMR spectra of the resultant
product were measured at 9.4 and 1 Tesla. They showed the peaks of the phenyl
group at around 7 ppm, the proton at the fluorinated position at 5.64 ppm and the
signals corresponding to the CH2 and CH3 groups of the ethyl ester at 3.63 and 0.57
ppm.
1H NMR (400 MHz, CDCl3): δ = 7.61 – 6.98 (m, 5H), 5.64 (d, J = 47.7 Hz, 1H), 3.63
(q, J = 7.1 Hz, 2H), 0.57 (t, J = 7.0 Hz, 3H) ppm.
1H NMR (43 MHz, CDCl3): δ = 8.51 – 7.25 (m, 5H), 5.93 (d, J = 48.7 Hz, 1H), 4.28
(q, J = 7.2 Hz, 2H), 1.22 (t, J = 7.1 Hz, 3H) ppm.
3.2
Synthesis of ethyl (Z)-2-fluoro-3-phenylacrylate
A solution of benzaldehyde and ethyl 2-fluoro-3-oxo-3-phenylpropanoate in MeCN (9
mL) was placed inside a 100 ml reactor. Cesium carbonate (Cs2CO3) was added to
the resultant mixture under constant stirring at 200 rpm to avoid mass transfer
limitations due to inappropriate mixing. The temperature of the reaction was varied
accordingly. After completion of the reaction, the reaction mixture was filtered, and
the solvent was removed by rotary evaporation to obtain the raw product. The
residue was purified by column chromatography on silica gel (ethyl acetate: hexane,
2:1) yielding the pure product. NMR spectra of the product were acquired at 9.4 and
1 Tesla, showing the magnetically different chemical groups.
1H NMR (400 MHz, CDCl3): δ = 7.63 (d, J = 6.9 Hz, 2H), 7.43 – 7.35 (m, 3H), 6.91
(d, J = 35.0 Hz, 1H), 4.34 (q, J = 7.1 Hz, 2H), 1.38 (t, J = 7.1 Hz, 3H) ppm.
1H NMR (43 MHz, CDCl3): δ = 7.81 – 6.45 (m, 5H), 6.03 (s, 1H), 3.77 (q, J = 6.9 Hz,
2H), 0.79 (t, J = 7.1 Hz, 3H) ppm.
3.3
Online monitoring procedure for real-time monitoring
A compact Magritek SpinSolve 1 T NMR spectrometer (Fig. 1) was employed for
studying the reaction mechanism, kinetics and activation energy in real-time in the
temperature range from 293 to 333 K.
This article is protected by copyright. All rights reserved.

3.3.1 Flow NMR setup and temperature
Each reaction was monitored by passing the mixture, from the round bottomed flask
through the flow cell inserted in the bore of the NMR magnet with the aid of the
three-channel peristaltic pump (ISMATEC® Relgo) at a flow rate of 1 ml/min. The
flow cell with 4 mm inner diameter near the sensitive region of the NMR magnet and
1 mm inner diameter at the other regions away from the sensitive region was
connected to polytetrafluoroethylene (PTFE) tubing forming a closed loop (Fig. 1a).
A stainless steel filter was attached to the PTFE tubing to filter the reaction mixture
before entering into the tubing. Each reaction was performed at the required
temperature inside the round bottomed flask placed outside the NMR spectrometer.
A mechanical stirrer was inserted on the top hole of the flask to agitate the mixture at
the desired stirring speed. The reaction mixture was transferred to the NMR
spectrometer at the temperature of the magnet by cooling down the reaction mixture
by means of a cold-water bath placed between the reactor and the magnet. The
reaction kinetics was assumed to be hardly affected because the reaction time was
long between 6 to 24 hours compared to the transfer time of a few seconds. The
temperature fluctuation inside the reactor was 0.9 K due to heat exchange with the
surroundings. The PTFE tubing and flow cell were rinsed three times with methanol,
water and acetone before each new use to remove any reaction residual adsorbed
on the walls.
3.3.2 Delay times
Different characteristic times need to be considered to find the optimum flow rate for
studying the reaction kinetics. The time difference between a change in the reaction
mixture and the corresponding change in the signal amplitude of the reaction mixture
depends on the different characteristic times encountered in the flow setup [19-20].
The transfer time ttrans, delay time tdelay, dwell time tdwell and residence-time-
distribution (RTD) curves for the same setup are discussed elsewhere [21-24].
3.3.4 Data acquisition
1
Each 1D H NMR spectrum at 1 T was acquired with single scan using a π/2
excitation pulse of 7 μs, an acquisition time of 6.6 s, an acquisition bandwidth of 5
kHz and a dwell time of 200 μs with 32 768 data points. Each 1D 19F NMR
spectrum was obtained with 4 scans. Every scan was recorded using a π/2
This article is protected by copyright. All rights reserved.

excitation pulse of 27 μs, an acquisition bandwidth of 20 kHz, an acquisition time
of 1.6 s and a dwell time of 50 μs with 32 768 data points. The delay before each
1
1
19
1D H and 2D H-1H COSY was 15 seconds and before each F scan was 10
1
seconds. The delays were optimized to be in quantitative mode. Each 2D H-1H
COSY spectrum was obtained in 9 minutes using 256 t1 increments and one scan.
1
The 1D H NMR spectra at 9.4 T were measured in 1.5 minutes with a Bruker
(AVANCE III) spectrometer using a 30° pulse flip angle and 16 scans. The data
were processed with automatic phase and baseline corrections.
3.3.5 Processing of the low-field data
The FIDs were recorded and saved by the SpinSolve (Magritek GmbH, Aachen)
software. They were processed and analyzed using the MestreNova (Version 11.0,
Mestrelab Research S. L., Santiago de Compostela, Spain) software. Exponential
windowing of 0.6 Hz was applied to increase the sensitivity of the spectra. The
baseline of each spectrum was corrected with a zero-order polynomial. Both the 1D
1H and 19F spectra were processed with same processing parameters. The peaks of
interest were integrated to obtain the variation in signal amplitude as a function of
time. The integration ranges were selected depending upon line width and peak
overlap. The derived values were exported to the Origin (version 9.0, OriginLab,
Northampton, MA) for plotting. The values of the rate constants were evaluated by
fitting a kinetic model to the signal-amplitude vs. time curves. The activation energy
graphs were plotted in Origin with the obtained rate constants.
4.
Results and discussion
4.1
Reaction scheme and mechanism
The reaction of ethyl 2-fluoro-3-oxo-3-phenylpropanoate (A) with benzaldhyde (B) in
the presence of cesium carbonate leads to the formation of ethyl (Z)-2-fluoro-3-
phenylacrylate (C) along with benzoic acid (D) as a side product (reaction scheme).
The reaction consists of three steps (Fig. 2). The first step involves the deprotonation
of α acetic proton present at the fluorinated position of ethyl 2-fluoro-3-oxo-3-
phenylpropanoate (A) with cesium carbonate acting as a base. The second step
involves the coupling of the deprotonated compound (I) with benzaldehyde (B). The
This article is protected by copyright. All rights reserved.

third step involves a rearrangement of bonds between species (II) and (III) with
simultaneous formation of a double bond at the fluorinated site along with the
elimination of benzoic acid (D) and formation of the product ethyl (Z)-2-fluoro-3-
phenylacrylate (C) [17]. The deprotonation step 1 and the formation of the final
product ethyl (Z)-2-fluoro-3-phenylacrylate (C) in step 3 were monitored with the
compact NMR spectrometer in real time. The lifetime of the intermediate involving
elimination and bond formation was very short compared to the NMR measurement
time, therefore, it cannot be observed with NMR. The formation of the short-lived
intermediate in the deacylation step is the rate determining step of the reaction which
is responsible for the final outcome of the reaction.
4.2
1D NMR spectroscopy
The NMR spectra of the starting materials were recorded in acetonitrile (MeCN) at
1
the same concentrations as those used for real-time monitoring. The 1D H NMR
spectrum of ethyl 2-fluoro-3-oxo-3-phenylpropanoate (A) (Fig. 3a) showed a triplet at
0.8 ppm from the (CH3-) group, a quartet centered at 3.8 ppm from the (-CH2-) group
of the (-CH2-CH3) ethyl group. The doublet centered at 5.8 ppm arises from the
2
coupling of fluorine with proton through J coupling with a large coupling constant of
47.5 Hz. The superimposed multiplet centered at 7 ppm arises from aromatic ring.
The 1D 1H NMR spectrum of benzaldehyde (B) (Fig. 3b) showed a single peak from
the benzaldehyde proton at 9.6 ppm. The multiplet centered at 7 ppm arises from the
1
aromatic ring. The 1D H NMR spectra of the mixture of both compounds (A+B)
before (Fig. 3c) and after catalyst addition (Fig. 3d) show the impact of the
deprotonation reaction of compound A with the base cesium carbonate. The doublet
present in the Fig. 3c is not present in Fig. 3d because of the loss of the proton from
1
the fluorinated site. The 1D H NMR spectrum of the reaction mixture (Fig. 3e)
contains mainly the unreacted benzaldehyde (B) and ethyl (Z)-2-fluoro-3-
phenylacrylate (C) (product) and benzoic acid (D) obtained at the end of the reaction
after 22 hours. The spectrum shows a single peak from the double bond proton at
6.1 ppm. The deprotonation causes the compound to precipitate due to low solubility,
19
leaving no signal of the starting material (A) in the NMR spectrum. The 1D F NMR
spectra of the starting material (A) and the product (C) both showed one doublet
centered at -189.3 ppm and -124.7 ppm, respectively [25].
This article is protected by copyright. All rights reserved.

4.3
2D NMR spectroscopy
1
The reaction was also monitored at 293 K by 2D H-1H COSY NMR spectroscopy.
Because the reaction was very slow, the time-dependent concentration changes
could also be observed by 2D NMR spectroscopy during the course of the reaction.
The 2D 1H-1H COSY NMR spectrum (Fig. S1) of the reaction mixture observed after
3
22 hours showed excellent signals from the J coupling between the CH2 and CH3
groups of the ester group. These off-diagonal peaks were used to monitor the
progress of the reaction because they vanish after the catalyst addition due to
precipitation and again appear with the formation of the product.
4.4
Online monitoring of the reaction
The reaction was monitored in real-time by interleaved 1D 1H-19F NMR
1
spectroscopy. The 3D stack plot of the H NMR spectra (Fig. 4) obtained as a
function of time shows the change in the concentrations of the different chemical
groups involved in the reaction. The change in the signal amplitudes was used to
determine the reaction kinetics. The signal intensity of the single peak from the
benzaldehyde (B) proton at 9.6 ppm decreases, whereas a singlet at 6.1 ppm from
the double bond proton and the quartet at 3.8 ppm from the -CH2 group of ethyl (Z)-
2-fluoro-3-phenylacrylate (C) increase with time. The multiplet observed in the
aromatic region from 6.5 to 8 ppm changes its shape from start to end of the
reaction. The maximum of the multiplet at the beginning of the reaction is at higher
chemical shifts and decreases over time. This shows the absence of the more
shielding carbonyl group present in the reactant (A) which deshields the protons via
resonance leading to higher chemical shifts of the proton. Simultaneously a new
maximum in the aromatic region of the product (C) arises at lower chemicals shift.
4.5
Evaluation of Reaction Kinetics
The reaction kinetics were evaluated from fitting the signal amplitude-time plots with
the kinetic model. The reaction consists of two main steps. The first deprotonation
step consists of the reaction of ethyl 2-fluoro-3-oxo-3-phenylpropanoate (A) with the
cesium carbonate catalyst. It is followed by the deacylation step consisting of the
reaction of the deprotonated reactant (A) and benzaldehyde (B).
This article is protected by copyright. All rights reserved.

The first deprotonation step was very fast compared to second deacylation
step. Even at the lowest temperature of 293 K, the deprotonation time was 4 minutes
compared to the deacylation which was 22.5 hours. This suggests that the
deprotonation step has no influence on the second reaction step. Therefore, both
steps were assumed to be separate reaction steps and the rate constants and
activation energies of both steps were calculated independently.
The deprotonation step proceeds with first order kinetics because the
concentration of the catalyst does not vary with time. The kinetic expression obtained
for the deprotonation steps becomes,
(1)
.
Here, k is the rate constant of the reaction. [A] and [A]0 are the concentrations of
ethyl 2-fluoro-3-oxo-3-phenylpropanoate (A) at times t and 0, respectively. Equation
(1) was fitted to the signal amplitude-time plot (Fig. 5a) of the first step to obtain the
kinetic rate constant. The plot of the residuals illustrates the good quality of the fit
(Fig. 5b).
The deprotonated compound (I) precipitates and its concentration diminishes
relative to that of the other reactant benzaldehyde (B) so that the reaction becomes
pseudo-first order. Therefore, the rate of change of benzaldehyde (B) determines the
kinetics of the reaction.
The deprotonation step proceeds with first order kinetics because the
concentration of catalyst does not vary with time. The kinetic expression obtained for
the deprotonation steps becomes,
(2)
.
Here, k is the rate constant of the reaction. [B] and [B]0 are the concentrations of
benzaldehyde (B) at time t and 0, respectively. Equation (2) was fitted to the signal
amplitude-time plot (Fig. 6a) of first step to obtain the kinetic rate constant. The plot
of the residuals shows the good quality of the fit (Fig. 6b).
4.6
Repeatability measurements
The instrument stability is essential for obtaining good quantitative and qualitative
results. Three experiments were carried out with same initial reaction conditions at
This article is protected by copyright. All rights reserved.

1
1
293 K with real-time monitoring by 1D H, 19F and 2D H-1H COSY NMR
spectroscopy in the interleaved mode. Each series of 1H-19F-(1H-1H-COSY)
experiments took 10 minutes. Therefore, 6 points were monitored with each method
in one hour. This time resolution was sufficient to monitor the reaction with more 22
hours reaction time. As The peaks used for integration were quite distinct from the
overlapping peaks, so that the effect of changing integration limit on the repeatability
was not considered. The same integration limits were used for every reaction to
check the repeatability. The reaction showed excellent repeatability and stability
which is manifested by the 1H NMR spectra (Fig. 7a) but also by the 19F spectra (Fig.
1
S2). The 2D H-1H experiments showed comparatively poor repeatability compared
1
19
to the H and F due to the long experiment time of 9 min during which changes in
the component concentrations affect the quantification more strongly than in the 1D
19
1H and F experiments. Also, the signal-to-noise ratio of 2D experiments was less
compared to the 1D experiments leading to differences in the rate constants and
repeatability obtained from the 2D experiment. However, the 2D experiments can
serve as a tool to understand the coupling interactions of various chemical groups.
1
The reaction rate constants obtained with 1D H and 19F were within the error range
and have a low coefficient of variation (Tab. 1). The integration of the whole
spectrum showed excellent field stability and mass balance during the total reaction
time of more than 22 hours (Fig. 7a).
4.7
Effect of temperature on the reaction rate
The reaction was performed at different temperatures ranging from 293 to 333 K to
determine the impact of temperature on the reaction rate using the stoichiometric
concentration (Fig. 7b). The variation in signal amplitude of the benzaldehyde (B)
proton showed faster conversion of benzaldehyde to product at higher temperatures
than at lower temperatures. The rate constants (Tab. 2) of each reaction step were
1
evaluated by fitting the kinetic equations (1) and (2) to the data obtained with H
NMR spectroscopy.
The percentage yield for each reaction was calculated from the spectral data and
is within the range of literature values [18]. The determined values of the rate
constants are within the standard range for reactions in solution, which can also be
appreciated after calculating the Q10 factor which describes the change of the
reaction rate for a temperature change of 10 K. The averaged Q10 factor for the
This article is protected by copyright. All rights reserved.

deacylation calculated from the proton measurements is Q10 = 2.7 which is in the
range for reactions performed in solution.
Arrhenius plots were obtained for the separate deprotonation and deacylation
steps of the reaction using the rate constants obtained from fitting the kinetic model.
The activation energy of each step was calculated from the slope of the Arrhenius
plots. A negative activation energy of (-28 ± 2) kJ/mol was calculated for the
deprotonation step (Fig. 8a) while a positive activation energy of (63.5 ± 8.3) kJ/mol
was evaluated for the deacylation step (Fig. 8b). This suggests that the
deprotonation step is very fast compared to the deacylation step confirming that the
deacylation is the rate determining step of the reaction.
5
Conclusion
The synthesis of α-fluoro- α β-keto unsaturated esters was monitored in real-time by
1
NMR with a tabletop 1 T spectrometer. A series of interleaved H-19F-(1H-1H 2D
COSY) experiments was executed to study the mechanistic details and kinetics at
different temperatures under stoichiometric conditions. The deprotonation step is
difficult to monitor in batch and inline modes because of the required time resolution.
The deprotonation step is very fast compared to the deacylation step. The compact
NMR spectrometer provides sufficient sensitivity and resolution to monitor the
reaction under the given reaction conditions at low concentrations and without
deuterated solvents. The longest experiment of 22 hours was repeatedly monitored
by interleaved experiments to assess the repeatability and stability of the system.
1
1
The results obtained show excellent agreement between H, 19F and 2D H-1H
COSY spectroscopy as different reaction monitoring methods. The activation
energies of the deprotonation and deacylation steps were obtained separately. The
activation energy of the deprotonation step was obtained to be (-28 ± 2) kJ/mol and
for the deacylation steps was (63.5 ± 8) kJ/mol. This shows that the deacylation step
is the rate determining step of the reaction while the deprotonation step is very fast.
Furthermore, the rate constants obtained from the deprotonation reaction are one
order of magnitude higher than those for the diacylation reaction. A negative
activation energy emerges from the exothermic solubility behavior of the cesium
carbonate reactant in the deprotonation step. NMR spectroscopy with a compact
instrument has the potential to explore not only the kinetics of the various reaction
This article is protected by copyright. All rights reserved.

components but also the mechanistic details and the behavior of different
components during the various steps of the reaction.
Conflict of interest
BB is on the board of directors of Magritek.
Acknowledgements
The authors gratefully acknowledge financial support from DFG Gerätezentrum
Pro2NMR, a DFG supported joint instrumental NMR facility of RWTH Aachen
University and KIT Karlsruhe.
References
1. J. B. Cooper, Chemometr. Intell. Lab. Syst. 1999, 46, 231.
2. J. van den Broeke, G. Langergraber, Spectroscopy Europe 2006, 18, 15.
3. T. De Beer, A. Burggraeve, M. Fonteyne, L. Saerens, J. P. Remon, C.
Vervaet, Int. J. Pharm. 2011, 417, 32.
4. G. Knothe, J. Am. Oil Chem. Soc. 2000, 77, 489.
5. N. Doki, H. Seki, K. Takano, H. Asatani, M. Yokota, N. Kubota, Cryst. Growth
Des. 2004, 4, 949.
6. G. Guiochon, C.L. Guillemin, Quantitative gas chromatography for laboratory
analyses and on-line process control, Elsevier, 1988.
7. L. Zhu, R.G. Brereton, D.R. Thompson, P.L. Hopkins, R.E. Escott, Anal.
Chim. Acta 2007, 584, 370.
8. H. Friebolin, Basic One- and Two-Dimensional NMR Spectroscopy, 5th ed.,
Wiley-VCH, Weinheim, 2011.
9. H. Günther, NMR Spectroscopy: Basic Principles, Concepts, and Applications
in Chemistry, 3rd ed., Wiley-VCH, Weinheim, 2013.
10.K. Singh, B. Blümich, Trends Anal. Chem. 2016, 83, 12.
11.B. Blümich and K. Singh, Angew. Chem. 2018, 57, 6996.
12.B. Blümich, S. Haber-Pohlmeier, W. Zia, Compact NMR, de Gruyter, Berlin,
2011.
13.E. Danieli, J. Perlo, A. Duchateau, G. Verzijl, V. Litvinov, B. Blümich, F.
Casanova, ChemPhysChem 2014, 15, 3060.
14.M. Killner, Y.G. Linck, E. Danieli, J. Rohwedder, Fuel 2015, 139, 240.
This article is protected by copyright. All rights reserved.

15.K. Singh, B. Blümich, Analyst 2017, 142, 1459.
16.B. Davies, F. Maesen, C. Baur, Ciprofloxacin, Springer, 1986, pp. 97-102.
17.D.B. Longley, D.P. Harkin, P.G. Johnston, Nat. Rev. Cancer 2003, 3, 330.
18.J. Qian, W. Yi, M. Lv, C. Cai, Synlett. 2015, 26, 127.
19.N. Zientek, C. Laurain, K. Meyer, M. Kraume, G. Guthausen, M. Maiwald, J.
Magn. Reson. 2014, 249, 53.
20.F. Dalitz, L. Kreckel, M. Maiwald, G. Guthausen, App. Magn. Reson. 2014, 45
411.
21.K. Singh, B. Blümich, Analyst 2018, 143, 4408.
22.M. Maiwald, H. H. Fischer, Y. K. Kim, K. Albert, H. Hasse, J. Magn. Reson.
2004, 166, 135.
23.M. V. S. Elipe, R. R. Milburn, Magn. Reson. Chem. 2016, 54, 437.
24.F. Dalitz, M. Maiwald, G. Guthausen, Chem. Eng. Sci. 2012, 75, 318.
25.V. Sans, L. Porwol, V. Dragone, L. Cronin, Chem. Sci. 2015, 6, 1258.
This article is protected by copyright. All rights reserved.

Table 1. Rate constants obtained for the deacylation reaction step repeated three
times using three measurement methods at 293 K.
Coefficient of
variation [%]
k 105 [s-1]
4.74 ± 0.08
4.28 ± 0.26
3.98 ± 0.29
k  105 [s-1]
4.73 ± 0.07
4.05 ± 0.66
4.21 ± 0.36
k  105 [s-1]
4.59 ± 0.08
4.61 ± 0.55
2.94 ± 0.40
Method
1H
1.8
19F
6.5
18
COSY
Table 2. Rate constants for both reaction steps at temperatures ranging from
293 K to 333 K.
kDeprot.  102 [s-1] kdeacyl  104 [s-1]
Yield [%]
S. No. Temperature [K]
1
2
3
4
293
303
318
333
2.36 ± 0.19
1.50 ± 0.39
0.98 ± 0.17
0.61 ± 0.69
0.47 ± 0.01
1.23 ± 0.03
3.52 ± 0.06
8.71 ± 0.24
50.0
59.8
66.0
72.5
This article is protected by copyright. All rights reserved.

Figure 1: a) Real-time reaction monitoring scheme. b) Compact Magritek 1 T NMR
spectrometer
This article is protected by copyright. All rights reserved.

Figure 2: Mechanism of the reaction.
This article is protected by copyright. All rights reserved.

Figure 3: 1D 1H NMR spectra of the different compounds involved in the reaction. a) Ethyl
2-fluoro-3-oxo-3-phenylpropanoate (A). b) Benzaldehyde (B). c) Mixture of A and B. d)
Mixture of (A), (B) and cesium carbonate catalyst. e) Reaction mixture containing
benzaldehyde (B), ethyl (Z)-2-fluoro-3-phenylacrylate (C), benzoic acid (D) after 22 hours
for the reaction performed at 293 K.
This article is protected by copyright. All rights reserved.

Figure 4: 3D stack plot (top) of 1D 1H NMR spectra and its projection onto the x-y plane
(bottom) for the reaction performed at 293 K.
This article is protected by copyright. All rights reserved.

Figure 5. Kinetic model fitted to the experimental data obtained for the deprotonation
reaction. a) Pseudofirst order kinetic model fitted to the signal amplitude-time plot of
fluorinated proton of ethyl 2-fluoro-3-oxo-3-phenylpropanoate (A). b) Residuals of the fit.
This article is protected by copyright. All rights reserved.

Figure 6. Kinetic model fitted to the experimental results obtained for the deacylation
reaction. a) Pseudofirst order kinetic model fitted to the signal amplitude-time plot of
benzaldehyde (B) proton. b) Residuals of the fit
This article is protected by copyright. All rights reserved.

Figure 7: Variation of the signal amplitude of the aldehyde proton of benzaldehyde (B) with
time from 1D 1H NMR spectroscopy. a) At 293 K for three identical reactant concentrations
along with mass balance. b) At temperatures ranging from 293 to 333 K.
This article is protected by copyright. All rights reserved.

Figure 8: Arrhenius plots for both reaction steps. a) Deprotonation. b) Deacylation.
This article is protected by copyright. All rights reserved.

Scheme 1
This article is protected by copyright. All rights reserved.