Continuous-Flow N-Heterocyclic Carbene Generation and Organocatalysis

Article abstract of DOI:10.1002/chem.201505135

Two methods were assessed for the generation of common N-heterocyclic carbenes (NHCs) from stable imidazol(in)ium precursors using convenient and straightforward continuous-flow setups with either a heterogeneous inorganic base (Cs₂CO₃ or K₃PO₄) or a homogeneous organic base (KN(SiMe₃)₂). In-line quenching with carbon disulfide revealed that the homogeneous strategy was most efficient for the preparation of a small library of NHCs. The generation of free nucleophilic carbenes was next telescoped with two benchmark NHC-catalyzed reactions; namely, the transesterification of vinyl acetate with benzyl alcohol and the amidation of N-Boc-glycine methyl ester with ethanolamine. Both organocatalytic transformations proceeded with total conversion and excellent yields were achieved after extraction, showcasing the first examples of continuous-flow organocatalysis with NHCs.

Full text of DOI:10.1002/chem.201505135

DOI: 10.1002/chem.201505135
Full Paper
&
Carbenes in Microreactors |Hot Paper|
Continuous-Flow N-Heterocyclic Carbene Generation and
Organocatalysis
Lorenzo Di Marco,[a] Morgan Hans,[a] Lionel Delaude,[b] and Jean-Christophe M. Monbaliu*[a]
Abstract: Two methods were assessed for the generation of
common N-heterocyclic carbenes (NHCs) from stable imida-
zol(in)ium precursors using convenient and straightforward
continuous-flow setups with either a heterogeneous inor-
ganic base (Cs2CO3 or K3PO4) or a homogeneous organic
base (KN(SiMe3)2). In-line quenching with carbon disulfide re-
vealed that the homogeneous strategy was most efficient
for the preparation of a small library of NHCs. The genera-
tion of free nucleophilic carbenes was next telescoped with
two benchmark NHC-catalyzed reactions; namely, the trans-
esterification of vinyl acetate with benzyl alcohol and the
amidation of N-Boc-glycine methyl ester with ethanolamine.
Both organocatalytic transformations proceeded with total
conversion and excellent yields were achieved after extrac-
tion, showcasing the first examples of continuous-flow orga-
nocatalysis with NHCs.
Introduction
established reactive intermediates, for example, acyl anions,
acylazoliums, azolium enolates, azolium homoenolates, or acid-
base complexes.[10] Spectacular developments have been re-
ported in this field over the past few years, with a particular
emphasis on the design of enantiopure NHCs for stereoselec-
tive organocatalysis.[11]
Numerous studies revolving around the properties and reactiv-
ity of N-heterocyclic carbenes (NHCs) have flourished since Ar-
duengo’s seminal report on the isolation and full characteriza-
tion of 1,3-diadamantylimidazol-2-ylidene in 1991.[1] NHCs are
nowadays recognized as powerful nucleophiles[2] and strong
neutral Brønsted/Lewis bases,[3] in which the steric and elec-
tronic environments can be easily tuned by changing the nitro-
gen substituents and/or the nature of the central heterocycle.[4]
In coordination chemistry, NHCs behave as powerful s-donor
and moderate p-acceptor ligands for the whole gamut of
metals, whether they belong to the main group, the d-block
elements, or the lanthanides and actinides.[5] Consequently,
metal–NHC complexes are significantly more stable than their
metal–phosphine analogues and are generally endowed with
a superior catalytic activity.[6] For instance, Houk and Gard re-
cently reported the conversion of amides into esters by using
Ni–NHC catalysts to activate and cleave the amide bond.[7] Fol-
lowing the pioneering work of Ukai et al.[8] and Breslow[9] on
the benzoin condensation catalyzed by thiazol-2-ylidenes gen-
erated in situ, the synthetic utility of NHCs was further broad-
ened to organocatalysis. Indeed, several classes of organic sub-
strates can be activated through the formation of various well-
For all their synthetic versatility and vast range of applica-
tion, NHCs are oxygen- and moisture-sensitive species that
demand strictly controlled reaction conditions. Thus, their han-
dling and storage usually require an inert atmosphere, sophis-
ticated glassware and procedures, and costly equipment such
as glove-boxes.[12] To alleviate these experimental issues and to
avoid their isolation, free carbenes are usually generated in
situ from stable precursors.[13] In this respect, the most
common and convenient access to NHCs consists of the depro-
tonation of an azol(in)ium salt with a strong base.[14]
Over the last decade, microreaction technology and continu-
ous-flow chemistry have emerged as powerful tools to im-
prove chemical manufacturing.[15] The inherent properties of
microfluidic reactors in terms of fast mixing, efficient heat
transfer, and accurate control over the local stoichiometry con-
tribute to the development of more efficient, reliable, and
safer chemical processes.[16] Previous reports have illustrated
the successful implementation of transient,[17] moisture/
oxygen-sensitive,[18] and dangerous species[19] within micro-
and mesoreactors, even under intensified conditions.[20] Reac-
tion telescoping is a common strategy in continuous-flow
chemistry that aims to perform consecutive reactions continu-
ously within the same uninterrupted reactor network.[21] Multi-
step sequences involving hazardous, toxic, and transient spe-
cies benefit from process telescoping because it suppresses
manual intervention and stockpiling of chemical intermediates.
Instead, the troublesome species are immediately reacted
inside the microreactor network, thereby avoiding degradation,
decreasing chemical risk, and improving overall efficiency.[22]
[a] L. Di Marco, Dr. M. Hans, Dr. J.-C. M. Monbaliu
Center for Integrated Technology and Organic Synthesis
Department of Chemistry, University of Li›ge, 4000 Li›ge (Belgium)
http: www.citos.ulg.ac.be
E-mail: jc.monbaliu@ulg.ac.be
[b] Prof. L. Delaude
Laboratory of Organometallic Chemistry and Homogeneous Catalysis
Department of Chemistry, University of Li›ge, 4000 Li›ge (Belgium)
Supporting information and ORCID(s) from the author(s) for this article are
available on the WWW under http://dx.doi.org/10.1002/chem.201505135.
Chem. Eur. J. 2016, 22, 4508 – 4514
4508
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
Given that NHCs are highly reactive, oxygen- and moisture-
sensitive species, telescoping their preparation and their use in
the controlled and confined environment of a microreactor
could simplify their manipulation. In 2013, McQuade et al. re-
ported a continuous-flow setup for the preparation and cata-
lytic use of [Cu(NHC)Cl] complexes.[23] The copper(I)-carbene
active species were obtained by oxidative addition of azolium
salts onto Cu2O, rather than by generation of free NHCs and
coordination to the metal. More recently, Brown and co-work-
ers reported the NHC-mediated oxidative esterification of vari-
ous aldehydes by telescoping the formation of Breslow’s inter-
mediates from aldehydes and a thiazolium salt in the presence
of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU), their anodic oxida-
tion into the corresponding acylazolium derivatives, and a final
esterification reaction in an uninterrupted microreactor net-
work.[24]
ous organic bases, such as potassium bis(trimethylsilyl)amide
(KHMDS).[26] We first sought a suitable method to assess the ef-
ficiency of the NHC generation in flow. Imidazol(in)ylidenes are
known to form stable and isolable zwitterionic adducts with
a variety of allenes[27] and heteroallenes including CO2,[27]
COS,[28] CS2,[27] diimides,[27,29] ketenes,[30] and iso(thio)cyanates.[27]
Among these various electrophiles, we selected carbon disul-
fide because this simple, widely available, liquid reagent in-
stantaneously forms highly colored, shelf-stable zwitterions
with nucleophilic carbenes (Scheme 1).[31]
The first NHC generation method involved a thermostated
packed-bed reactor filled with an anhydrous inorganic base
(Figure 1a). Anhydrous and degassed solvents [tetrahydrofuran
(THF), MeCN, or N,N-dimethylformamide (DMF)] were em-
ployed to prepare feed solutions of imidazolium precursor 1a,
which served as a model substrate. They were conveyed to the
In this contribution, we report on the design of a convenient
and straightforward continuous-flow setup for the generation
of free NHC species, and their use as organocatalysts in two
benchmark NHC-promoted reactions; namely, a transesterifica-
tion and an amidation reaction. The implementation of in-line
IR monitoring and continuous extraction/separation is also
showcased for one example.
packed-bed reactor through
a
syringe pump set at
0.25 mLminÀ1 to get a 10 min residence time. The reactor efflu-
ent was mixed with a large excess of carbon disulfide
(0.25 mLminÀ1) through a T-mixer located immediately after
the outlet of the packed-bed reactor. A short perfluoroalkoxy
Results and Discussion
Continuous-flow NHC generation
Two distinct strategies were investigated for converting stable
imidazol(in)ium salts into NHCs by using a continuous-flow
setup. The first strategy used heterogeneous inorganic bases
to deprotonate the azolium precursors, whereas the second
approach relied on homogeneous organic bases. The batch an-
alogues of these two approaches are both very common for
the generation of NHCs in situ, either with heterogeneous inor-
ganic bases, such as cesium carbonate[25] or with homogene-
Figure 1. a) Heterogeneous and b) homogeneous strategies for the continu-
ous-flow generation and quenching of NHCs. Both setups include an anhy-
drous argon inlet to flush the system via a three-way valve prior to utiliza-
tion.
Scheme 1. Typical procedure for the transformation of imidazol(in)ium salts 1 into their stable zwitterionic NHC·CS2 adducts 3. The bases used were Cs2CO3
(2a), K3PO4 (2b), or KN(SiMe3)2 (2c).
Chem. Eur. J. 2016, 22, 4508 – 4514
www.chemeurj.org
4509
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
alkane (PFA) capillary loop (0.25 mL internal volume) further in-
creased the residence time to complete the formation of
IMes·CS2 (3a). The formation of this zwitterion could be visual-
ized through the appearance of an intense red coloration in
the PFA capillary loop. Samples were collected periodically and
analyzed by 1H NMR spectroscopy. Conversions were calculated
by comparing the characteristic signals of the imidazolium pre-
cursor 1a and the corresponding NHC·CS2 zwitterionic adduct
3a (Figure 2).[27,31]
MeCN (25 mLminÀ1). After a residence time of 10 min in the
PFA loop, the reactor effluent was quenched with a large
excess of CS2 (0.125 mLminÀ1) and analyzed by 1H and
13C NMR spectroscopies. Under these conditions, a 51% con-
version was reached. Samples were collected periodically over
2 h of operation and consistently gave the same conversion.
An important solvent effect was noted when KHMDS was dis-
solved in THF, in which case the conversion of 1a was com-
plete. Total conversion was preserved when working with a re-
duced excess of KHMDS (1.2 equiv, 0.012m in THF,
25 mLminÀ1). The best combination of solvents involved MeCN
to dissolve 1a, and THF to dissolve KHMDS, ensuring total con-
version and preventing the precipitation of KCl in the reactor.
A 10-fold increase in concentration also led to complete con-
version, although the zwitterionic NHC·CS2 product 3a accu-
mulated in the second PFA loop. The immersion of the reactor
in an ultrasonic bath prevented clogging and ensured a steady
operation of the system for more than 2 h with consistent
output.
In view of the good results obtained with 1a, the homoge-
neous method was extended to the generation and quenching
of NHCs derived from imidazolium salts 1b,c and imidazolini-
um salts 1d,e (Figure 1). In some instances, the excess of
KHMDS was increased to 1.5 equiv to ensure complete conver-
sion into zwitterionic adducts 3b–e. A significant loss of effi-
ciency was observed when IMes·HBF4 (1a’) was used as car-
bene precursor instead of IMes·HCl (1a). This observation is in
line with our failed attempts to deprotonate 1a’ by using
a packed-bed column of Cs2CO3. It can be correlated with a pre-
vious study on the deprotonation rate of various imidazolium
salts, which showed that the rate of C2-deprotonation in-
creased with the basicity of the associated counteranion.[32]
Indeed, the more basic chloride anion led to a significantly
higher rate of carbene formation than its tetrafluoroborate
counterpart. Despite some practical preparative assets over
their chloride analogues,[33] the slower C2-deprotonation rates
of imidazolium tetrafluoroborates may therefore hamper their
use as efficient NHC precursors for continuous-flow applica-
tions. McQuade et al.[23] solved this issue by using a packed-
bed column filled with an ion-exchange resin that allowed the
in situ conversion of imidazol(in)ium tetrafluoroborates into
more reactive imidazol(in)ium chlorides. Similarly, a packed-
bed column filled with Dowexꢁ 1X2 ion exchange resin (200–
400 mesh, ClÀ form) was successfully adapted to our microflui-
dic setup to enable the use of 1a’ as an efficient NHC precur-
sor (see the Supporting Information for details).
Figure 2. Representative conversions of imidazolium precursors 1a and 1a’
using the heterogeneous (squares) and homogeneous (circles) strategies
(see Figure 1).
Reactive packed-bed reactors are very convenient for per-
forming continuous-flow processes, but they inherently suffer
from the progressive disappearance of the packed reagent,
and their performance is dramatically affected over the course
of a reaction.[23] This phenomenon was clearly identified in our
experiments as the conversion toward zwitterionic adduct 3a
usually dropped after 20 min of operation (Figure 2). The first
set of results with the heterogeneous setup was collected with
Cs2CO3 (2a) as the inorganic base (5 g) at 258C and a 0.1m so-
lution of 1a in DMF. A mere 23% conversion was obtained at
steady state (after 20 min). The conversion then began to de-
crease and was limited to 5% after 60 min of operation. Even
lower conversions were obtained in MeCN or THF. Potassium
phosphate (2b) performed similarly to Cs2CO3, with 21% con-
version over 20 min of operation, which then dropped to 10%
after 60 min. A 10-fold decrease in concentration helped to
sustain a constant conversion of ca. 30% over 30 min, but did
not prevent it from dropping when the run was further ex-
tended. The conversion of imidazolium precursor IDip·HCl (1b)
did not exceed 5% under these conditions. IMes·HBF4 (1a’)
failed to react under these conditions, and the starting material
1a’ was recovered unchanged.
The continuous-flow setup was adapted for the second NHC
generation method (Figure 1b). The packed-bed reactor was
replaced by a PFA capillary loop (0.50 mL internal volume) and
a second inlet was connected upstream to inject a solution of
KHMDS (2c). Mixing of both reagent feeds was achieved
through a polyether ether ketone (PEEK) T-mixer. The PFA ca-
pillary loop was thermostated at 258C and the quench with
CS2 was implemented as described above for the heterogene-
ous procedure.
Organocatalytic applications
Having optimized the generation of NHCs under continuous-
flow conditions, we devised a straightforward microfluidic
setup to investigate two benchmark NHC-catalyzed reactions;
namely, the transesterification and the amidation of esters. In
2002, Nolan[34] and Hedrick[35] independently reported the
transesterification of esters with alcohols in the presence of
preformed or in situ generated NHCs. The procedure was sub-
sequently extended to other substrates such as phospho-
A 0.01m solution of 1a in MeCN was pumped at a flow rate
of 25 mLminÀ1 and reacted with a 0.02m solution of KHMDS in
Chem. Eur. J. 2016, 22, 4508 – 4514
www.chemeurj.org
4510
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
nates[36] and an asymmetric version was developed by
Suzuki[37] and Maruoka.[38] In this study, the NHC-catalyzed
transesterification of vinyl acetate with benzyl alcohol to afford
benzyl acetate was selected as a model reaction and imple-
mented in a continuous-flow setup (Figure 3). The upstream
section of the microfluidic network consisted of the assembly
designed for generating NHCs under homogeneous conditions
(cf. Figure 1b). The CS2 quench line was removed and replaced
by a premixing unit for benzyl alcohol (4; 1m in THF or tolu-
ene) and vinyl acetate (5; 1m in THF or toluene). The premix-
ing unit consisted of a PFA capillary loop (0.25 mL internal
volume) connected via a T-mixer to two syringe pumps set at
25 mLminÀ1 each. The stoichiometric mixture of 4 and 5 was
then conveyed in a 1 mL PFA capillary loop together with the
NHC catalyst. The outlet of the reactor was connected to an IR
cell for in-line reaction monitoring.
The transesterification reached completion after a residence
time of 10 min (total flow rate: 0.1 mLminÀ1) at room tempera-
ture with 1a as NHC precursor (0.01m in MeCN) and KHMDS
as base (0.012m in THF). The flow rate and reagent ratios were
adapted to have 1 mol% of NHC in the presence of substrates
4 and 5. The choice of solvents, reagents, and products ena-
bled the course of the reaction to be conveniently monitored
by using in-line IR spectroscopy (Figure 4). The disappearance
of 4 (n˜OH at 3438 cmÀ1) and 5 (n˜C at 1764 cmÀ1 and n˜C at
Figure 4. In-line IR monitoring of the NHC-organocatalyzed transesterifica-
tion of vinyl acetate with benzyl alcohol. The upper insert summarizes the
typical peaks for compounds 5–7 in the 1600–1800 cmÀ1 region. The dotted
line (a.) marks the disappearance of reagents 4 and 5 and the appearance of
products 6 and 7 approximately 10 min after injecting the catalyst.
=
=
C
O
1648 cmÀ1), as well as the appearance of benzyl acetate 6 (n˜C
Table 1. NHC-catalyzed transesterification of vinyl acetate (5) with benzyl
alcohol (4) in THF.
=
O
at 1743 cmÀ1) and acetaldehyde 7 (n˜C at 1730 cmÀ1) were de-
=
O
duced from the evolution of characteristic vibration bands.
Samples were collected, processed, and analyzed by NMR
spectroscopy. The system was operated over 6 h with consis-
tent results and high operational stability.
Entry
NHC precursor[a]
Carbene
Conversion [%][b]
1
2
3
4
1a
1b
1d
1e
IMes
IDip
SIMes
SIDip
>99[c]
63[d]
10[d]
6[d]
The model transesterification of vinyl acetate (5) with benzyl
alcohol (4) was then carried out in THF with various NHC pre-
[a] 1 mol%. [b] Determined from the 1H NMR signal integral ratio of the
benzylic protons in 4 and 6. [c] KHMDS 0.012m in THF. [d] KHMDS
0.015m in THF.
1
cursors and the conversions were determined by H NMR spec-
troscopic analysis (Table 1). The nature of the N-substituents
and of the central heterocyclic core (imidazolium or imidazoli-
nium) markedly affected the catalytic activity. Similar trends
were reported in batch by Nolan,[34] Hedrick,[35] and others.[39]
Aryl-substituted imidazol-2-ylidenes IMes and IDip (Table 1, en-
tries 1 and 2) afforded benzyl acetate in higher yield than the
corresponding aryl-substituted imidazolin-2-ylidenes SIMes and
SIDip (Table 1, entries 3 and 4). The highest conversion was ob-
tained with NHC precursor 1a, which led to a 96% yield of
benzyl acetate (6) after work-up. The lower activity displayed
by 1b can be attributed to the steric congestion induced by
its ortho-iPr substituents. Similar results were obtained by
using 1m solutions of 4 and 5 in toluene instead of THF. This
enabled the implementation of a simple yet efficient in-line
quench/liquid–liquid extraction/separation system. The outlet
Figure 3. Continuous-flow setup for the preparation and use of NHCs as organocatalysts for a transesterification reaction.
Chem. Eur. J. 2016, 22, 4508 – 4514 www.chemeurj.org ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
4511

Full Paper
of the reactor was connected to a cross junction for the con-
comitant injection of toluene (0.5 mLminÀ1) and an aqueous
solution of sodium bicarbonate (5%-wt, 0.5 mLminÀ1). The re-
sulting segments were then conveyed through a short PFA ca-
pillary section to a continuous liquid–liquid membrane separa-
tor.[40] The extraction/separation sequence afforded essentially
pure 6 in 92% yield.
The scope of our continuous-flow methodology for the
preparation and organocatalytic application of free NHCs was
further extended to another benchmark NHC-promoted reac-
tion; namely, the amidation of an unactivated ester.[41] In 2005,
Movassaghi et al. reported a catalytic, one-pot, two-step ami-
dation of unactivated esters using IMes as an organocata-
lyst.[41a] The reaction proceeded with the initial transesterifica-
tion of a methyl ester with an amino alcohol. The intermediate
amino ester then underwent an intramolecular O-to-N acyl
shift,[41a,42] leading to the formation of an amide bond. The mi-
crofluidic setup devised for transesterification (cf. Figure 3) was
easily modified to telescope the NHC generation sequence
with an amidation reaction instead. Thus, the reaction loop
was replaced by a 1.5 mL internal volume PFA capillary to in-
crease the residence time to 15 min (Figure 5). Feeds c and d
ous formation of stable, highly colored NHC·CS2 zwitterions to
trap the free carbenes. It was successfully extended to allow
NHC-catalyzed reactions. In a first example, the NHC genera-
tion sequence was telescoped with the organocatalytic trans-
esterification of vinyl acetate and benzyl alcohol. The microflui-
dic setup enabled the fast scouting of reaction parameters and
screening of various NHC precursors to determine structure-ac-
tivity relationships. In-line reaction monitoring, downstream
quench, and liquid–liquid extraction/separation techniques
were implemented. A similar strategy was applied to carry out
an amidation reaction catalyzed by IMes. This second example
showcased the formal direct telescoping of three chemical
events including i) the generation of a free carbene, ii) the
transesterification of N-Boc-glycine methyl ester with ethanol-
amine, and iii) a subsequent intramolecular O-to-N acyl shift
leading to tert-butyl (2-hydroxyethylcarbamoyl)methylcarba-
mate within an uninterrupted reactor network.
Experimental Section
General information
1H NMR spectra were recorded at 250 or 400 MHz and 13C NMR
spectra were recorded at 62.8 or 100.6 MHz with Bruker Avance
spectrometers. The chemical shifts are reported in ppm relative to
TMS as internal standard or to solvent residual peak. HRMS spectra
were recorded with a FTMS (ESI) apparatus (Thermo Scientific Q Ex-
active). Solvents purchased from Labotec were dried according to
standard procedures and deoxygenated prior to use (15 min Ar
flow). Commercially available chemicals (Sigma-Aldrich or TCI) were
used as received, except for benzyl alcohol, vinyl acetate, and
amino ethanol, which were distilled and degassed prior to use. Po-
tassium phosphate (K3PO4) and cesium carbonate (Cs2CO3) were
stored in an oven at 808C. Potassium bis(trimethylsilyl)amide
(KHMDS) was stored under a dry and inert atmosphere. Ion ex-
change resin Dowexꢁ 1X2 200–400 mesh (ClÀ form) was rinsed
with absolute ethanol and dried under high vacuum for 72 h prior
to use. It was stored in a dessicator. Imidazolium and imidazolini-
um salts 1a–e and 1a’ were prepared according to reported pro-
cedures[43] and stored in a dessicator. N-(tert-Butoxycarbonyl)gly-
cine methyl ester 8 was prepared according to standard proce-
dures[44] and purified by column chromatography on silica gel 60
(230–400 mesh, Biosolve).
Figure 5. Continuous-flow setup for the preparation and use of IMes as an
organocatalyst in the amidation of N-Boc-glycine methyl ester with ethanol-
amine.
were loaded with stock solutions of N-Boc-glycine methyl ester
(8; 2m in THF) and ethanolamine (9; 2m in THF), respectively.
Imidazolium salt 1a (0.1m in MeCN) served as a catalyst pre-
cursor along with KHMDS (0.12m in THF). The flow rate and re-
agent ratios were adapted to attain a 5 mol% proportion of
NHC vs. substrates 8 and 9. Under these experimental condi-
Experimental setup
The microfluidic device was constructed with high purity PFA coils
(0.25, 0.5, 1.0, and 1.5 mL internal volume) and straight sections (1/
16’’ o.d., 0.03“ i.d.), Upchurch Scientific Super Flangeless nuts (nat-
ural PEEK, 1/4–28 thread for 1/16” o.d. tubing), Upchurch Scientific
Super Flangeless ferrules (yellow ETFE for 1/16“ o.d. tubing), Up-
church Scientific T-mixers (natural PEEK for 1/16” tubing, 0.5 mm
through hole), and Upchurch Scientific cross junctions (natural
PEEK for 1/16“ tubing, 0.5 mm through hole). Dry argon (Air Liq-
uide) was used for flushing the microfluidic device through a four-
port bulkhead switching valve (Upchurch Scientific 1/4–28 ports
for 1/16” o.d. tubing). The feed solutions were conveyed to the re-
actor using Chemyx Nexus 3000 and/or Nexus 6000 syringe
1
tions, full conversion was achieved as determined by H NMR
spectroscopic analysis of the crude mixture, and glycine deriva-
tive 10 was isolated in 70% yield after purification by column
chromatography.
Conclusion
We have developed a convenient, straightforward, and versa-
tile methodology for the continuous-flow generation of NHCs
in the sealed and confined environment of a microfluidic reac-
tor. The procedure was first optimized by using the instantane-
pumps.
A continuous-flow liquid–liquid membrane separator
(Zaiput Flow Technologiesꢁ) with a PTFE membrane (0.5 mm pores)
was used for the continuous liquid–liquid extraction/separation ex-
Chem. Eur. J. 2016, 22, 4508 – 4514
www.chemeurj.org
4512
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
periments. In-line, real-time IR monitoring was carried out with
a FlowIRTM from Mettler-Toledo with a DTGS detector using Happ-
Genzel apodization, equipped with a SiComp (Silicon) probe con-
nected via a FlowIRꢂ sensor. Sampling was performed from 4000
to 650 cmÀ1 at 8 wavenumber resolution with 208 scans.
Liquid–liquid extraction and separation: A PEEK cross junction
was inserted after the IR cell for the consecutive injection of tolu-
ene (0.5 mLminÀ1) and an aqueous solution of sodium bicarbonate
(5%-wt, 0.5 mLminÀ1). The fluids were then conveyed to a liquid–
liquid PTFE membrane separator after a short PFA loop (0.25 mL in-
ternal volume). The aqueous stream was redirected to a waste
tank, and the organic stream (toluene) was collected and concen-
trated under reduced pressure. The extraction/separation sequence
afforded essentially pure 6 in 92% yield.
General procedure for the continuous-flow transformation
of imidazol(in)ium salts into NHC·CS2 zwitterions using the
homogeneous strategy
General procedure for the continuous-flow NHC organocata-
lyzed amidation of N-Boc-glycine methyl ester with ethanol-
amine
The syringe pump used to deliver the 0.1m solution of imidazol-
(in)ium precursor (1) in MeCN was set to 25 mLminÀ1 and the sy-
ringe pump used to deliver the solution of KHMDS in THF was set
to 25 mLminÀ1. Depending on the nature of the precursor, the ratio
[KHMDS]/[1] was adapted from 1.2 (1a) to 1.5 (1b–e). For precur-
sor 1a’, a packed-bed column filled with Dowexꢁ 1X2 200–400
mesh (ClÀ form) was inserted before the first PFA loop. Both
streams came in contact through a PEEK T-mixer using Super
Flangeless nuts and ferrules. They were reacted in a PFA capillary
loop (0.5 mL internal volume, 10 min residence time). The outlet of
the first PFA loop was attached to a PEEK T-mixer with Super
Flangeless nuts and ferrules. CS2 was delivered by a third syringe
pump, set to 0.125 mLminÀ1 and the mixture was redirected to
a second PFA capillary loop (0.25 mL internal volume). The outlet
of the second PFA loop was connected to a collection vial. The crit-
ical section of the setup, i.e., the sequence (0.5 mL PFA loop — T-
mixer — 0.25 mL PFA loop), was immerged in an ultrasonic bath
to avoid reactor clogging. The solvents were removed under re-
duced pressure and the red solid residues were analyzed by 1H
and 13C NMR spectroscopies. Experimental data matched those re-
ported previously.[31]
The syringe pumps delivering the solution of 1a (0.1m in MeCN)
and the solution of KHMDS (0.12m in THF) were set to 25 mLminÀ1
(feeds a and b). Both streams came in contact through a PEEK T-
mixer using Super Flangeless nuts and ferrules. They were reacted
in a PFA capillary loop (0.5 mL internal volume, 10 min residence
time). The outlet of the first PFA loop was attached to a PEEK T-
mixer with Super Flangeless nuts and ferrules, and connected to
a premixing unit (0.25 mL internal volume) for feeds c and d (2m
N-Boc-glycine methyl ester 8 in THF and 2m ethanolamine 9 in
THF, respectively). The two corresponding syringe pumps were set
to 25 mLminÀ1 each. The stoichiometric mixture of 8 and 9 was
then conveyed in the presence of the NHC catalyst in a third PFA
capillary loop (1.5 mL internal volume, 15 min residence time). The
flow rates and reagents ratio were adjusted to 5 mol% of NHC in
the presence of substrates 8 and 9. The conversion of substrates 8
and 9 was complete, leading to glycine derivative 10 in 70% yield
after purification by chromatography on silica gel.
Start-up procedure: The entire microfluidic setup was flushed
with dry argon. The reactor was operated with feed a=MeCN, and
feeds b,c,d=THF for 10 min. Feeds c,d were then switched to N-
Boc-glycine methyl ester 8 (2m in THF) and ethanolamine 9 (2m in
THF), and the reactor was equilibrated for 30 min at 258C. Feeds
a,b were then switched to the solution of 1a (0.1m in MeCN) and
the solution of KHMDS (0.12m in THF), respectively. The reactor
was equilibrated for another 30 min.
General procedure for the continuous-flow NHC-organocata-
lyzed transesterification of vinyl acetate with benzyl alcohol
The syringe pumps delivering the solution of 1a (0.01m in MeCN)
and the solution of KHMDS (0.012m in THF) were set to
25 mLminÀ1 (feeds a and b). Both streams came in contact through
a PEEK T-mixer using Super Flangeless nuts and ferrules. They were
reacted in a PFA capillary loop (0.5 mL internal volume, 10 min resi-
dence time). The outlet of the first PFA loop was attached to
a PEEK T-mixer with Super Flangeless nuts and ferrules. The PEEK
T-mixer was connected to a premixing unit for benzyl alcohol (4;
1m in THF) and vinyl acetate (5; 1m in THF) consisting of a PFA ca-
pillary loop (0.25 mL internal volume) connected via a PEEK T-
mixer to two syringe pumps set to 25 mLminÀ1 each (feeds c and
d). The stoichiometric mixture of 4 and 5 was then conveyed in
the presence of the NHC catalyst in a third PFA capillary loop
(1 mL internal volume, 10 min residence time). The outlet of the re-
actor was connected to an IR cell for in-line reaction monitoring.
The flow rate and reagent ratio were adjusted to 1 mol% NHC in
the presence of substrates 4 and 5. Evaporation of the solvents
and filtration of the crude material on a short silica gel pad afford-
Acknowledgements
J.-C.M.M. gratefully acknowledges financial support from the
University of Li›ge (WG-13/03) and the F.R.S.-FNRS (CDR
J.0147.15).
Keywords: carbenes
· continuous-flow · microreactors ·
organocatalysis · transesterification
[1] A. J. Arduengo III, R. L. Harlow, M. Kline, J. Am. Chem. Soc. 1991, 113,
361–363.
[2] B. Maji, M. Breugst, H. Mayr, Angew. Chem. Int. Ed. 2011, 50, 6915–6919;
Angew. Chem. 2011, 123, 7047–7052.
1
ed essentially pure benzyl acetate (6) in 96% yield. H and 13C NMR
data matched those reported previously.[34]
[3] a) T. L. Amyes, S. T. Diver, J. P. Richard, F. M. Rivas, K. Toth, J. Am. Chem.
Soc. 2004, 126, 4366–4374; b) E. M. Higgins, J. A. Sherwood, A. G. Lind-
say, J. Armstrong, R. S. Massey, R. W. Alder, A. C. O’Donoghue, Chem.
Commun. 2011, 47, 1559–1561; c) R. S. Massey, C. J. Collett, A. G. Lind-
say, A. D. Smith, A. C. O. Donoghue, J. Am. Chem. Soc. 2012, 134,
20421–20432.
[4] a) T. Drçge, F. Glorius, Angew. Chem. Int. Ed. 2010, 49, 6940–6952;
Angew. Chem. 2010, 122, 7094–7107; b) H. Clavier, S. P. Nolan, Chem.
Commun. 2010, 46, 841–861; c) D. J. Nelson, S. P. Nolan, Chem. Soc. Rev.
2013, 42, 6723–6753.
Start-up procedure: The entire microfluidic setup was flushed
with dry argon. The reactor was operated with feed a=MeCN and
feeds b,c,d=THF for 10 min. Feeds c,d were then switched to
benzyl alcohol (4; 1m in THF or toluene) and vinyl acetate (5; 1m
in THF or toluene), and the reactor was equilibrated for 20 min at
258C. Feeds a and b were then switched to the solution of 1a
(0.01m in MeCN) and the solution of KHMDS (0.012m in THF), re-
spectively. The reactor was equilibrated for another 20 min.
Chem. Eur. J. 2016, 22, 4508 – 4514
www.chemeurj.org
4513
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Full Paper
[5] a) For a review on non-transition-metal complexes, see: C. E. Willans in,
Organometallic Chemistry, Vol. 36 (Eds.: I. J. S. Fairlamb, J. M. Lynam),
Royal Society of Chemistry, Cambridge, 2010, pp. 1–28; b) For a review
on group I, II, and early-transition-metal complexes, see: S. Bellemin-La-
ponnaz, S. Dagorne, Chem. Rev. 2014, 114, 8747–8774; c) For a review
on palladium complexes, see: N. Marion, S. P. Nolan, Acc. Chem. Res.
2008, 41, 1440–1449; d) For a review on late-transition-metal com-
plexes, see: S. Díez-Gonzµlez, N. Marion, S. P. Nolan, Chem. Rev. 2009,
109, 3612–3676; e) For a review on coinage-metal complexes, see:
J. C. Y. Lin, R. T. W. Huang, C. S. Lee, A. Bhattacharyya, W. S. Hwang,
I. J. B. Lin, Chem. Rev. 2009, 109, 3561–3598; f) For a review on group
XII metal complexes, see: S. Budagumpi, S. Endud, Organometallics
2013, 32, 1537–1562; g) For a review on lanthanide and actinide com-
plexes, see: P. L. Arnold, I. J. Casely, Chem. Rev. 2009, 109, 3599–3611.
[6] M. N. Hopkinson, C. Richter, M. Schedler, F. Glorius, Nature 2014, 510,
485–496.
Mascia, J. M. B. Evans, T. F. Jamison, K. F. Jensen, Org. Process Res. Dev.
2014, 18, 402–409; d) D. R. Snead, T. F. Jamison, Angew. Chem. Int. Ed.
2015, 54, 983–987; Angew. Chem. 2015, 127, 997–1001; e) N. Lambor-
elle, J. F. Simon, A. Luxen, J.-C. M. Monbaliu, Org. Biomol. Chem. 2015,
13, 11602–11606.
[23] a) S. M. Opalka, J. K. Park, A. R. Longstreet, D. T. McQuade, Org. Lett.
2013, 15, 996–999; b) B. A. Ondrusek, S. M. Opalka, O. Hietsoi, M. Sha-
truk, D. T. McQuade, Synlett 2013, 24, 1211–1214.
[24] R. A. Green, D. Pletcher, S. G. Leach, R. C. D. Brown, Org. Lett. 2015, 17,
3290–3293.
[25] Y.-R. Zhang, L. He, X. Wu, P.-L. Shao, S. Ye, Org. Lett. 2008, 10, 277–280.
[26] N. Duguet, C. D. Campbell, A. M. Z. Slawin, A. D. Smith, Org. Biomol.
Chem. 2008, 6, 1108–1113.
[27] L. Delaude, Eur. J. Inorg. Chem. 2009, 1681–1699.
[28] M. Hans, J. Wouters, A. Demonceau, L. Delaude, Eur. J. Org. Chem. 2011,
7083–7091.
[7] L. Hie, N. F. Fine Nathel, T. K. Shah, E. L. Baker, X. Hong, Y.-F. Yang, P. Liu,
K. N. Houk, N. L. Garg, Nature 2015, 524, 79–83.
[29] L. M. Martínez-Prieto, C. Urbaneja, P. Palma, J. Cµmpora, K. Philippot, B.
Chaudret, Chem. Commun. 2015, 51, 4647–4650.
[8] T. Ukai, R. Tanaka, T. Dokawa, J. Pharm. Soc. Jpn. 1943, 63, 296–300.
[9] R. Breslow, J. Am. Chem. Soc. 1958, 80, 3719–3726.
[30] a) M. Regitz, J. Hocker, B. Weber, Angew. Chem. Int. Ed. Engl. 1970, 9,
375; Angew. Chem. 1970, 82, 394–395; b) Y.-G. Lee, J. P. Moerdyk, C. W.
Bielawski, J. Phys. Org. Chem. 2012, 25, 1027–1032; c) M. Hans, J. Wout-
ers, A. Demonceau, L. Delaude, Chem. Eur. J. 2013, 19, 9668–9676; d) B.
Maji, H. Mayr, Angew. Chem. Int. Ed. 2013, 52, 11163–11167; Angew.
Chem. 2013, 125, 11370–11374; e) M. Hans, J. Wouters, A. Demonceau,
L. Delaude, Chem. Eur. J. 2015, 21, 10870–10877.
[31] L. Delaude, A. Demonceau, J. Wouters, Eur. J. Inorg. Chem. 2009, 1882–
1891.
[32] S. Wei, X.-G. Wei, X. Su, J. You, Y. Ren, Chem. Eur. J. 2011, 17, 5965–
5971.
[10] a) V. Nair, R. S. Menon, A. T. Biju, C. R. Sinu, R. R. Paul, A. Jose, V. Sreeku-
mar, Chem. Soc. Rev. 2011, 40, 5336–5346; b) X. Bugaut, F. Glorius,
Chem. Soc. Rev. 2012, 41, 3511–3522; c) H. U. Vora, P. Wheeler, T. Rovis,
Adv. Synth. Catal. 2012, 354, 1617–1639; d) J. Douglas, G. Churchill,
A. D. Smith, Synthesis 2012, 44, 2295–2309; e) J. Mahatthananchai, J. W.
Bode, Acc. Chem. Res. 2014, 47, 696–707.
[11] a) D. Enders, O. Niemeier, A. Henseler, Chem. Rev. 2007, 107, 5606–
5655; b) D. M. Flanigan, F. Romanov-Michailidis, N. A. White, T. Rovis,
Chem. Rev. 2015, 115, 9307–9387.
[12] M. K. Denk, J. M. Rodezno, S. Gupta, A. J. Lough, J. Organomet. Chem.
2001, 617–618, 242–253.
[13] M. F›vre, J. Pinaud, Y. Gnanou, J. Vignolle, D. Taton, Chem. Soc. Rev.
2013, 42, 2142–2172.
[14] L. Benhamou, E. Chardon, G. Lavigne, S. Bellemin-Laponnaz, V. CØsar,
Chem. Rev. 2011, 111, 2705–2733.
[15] C. Wiles, P. Watts, Micro Reaction Technology in Organic Synthesis, CRC
Press, Taylor and Francis Group, Boca Raton, FL, 2011.
[16] a) R. L. Hartman, J. P. McMullen, K. F. Jensen, Angew. Chem. Int. Ed. 2011,
50, 7502–7519; Angew. Chem. 2011, 123, 7642–7661; b) K. S. Elvira, X.
Casadevall i Solvas, R. C. R. Wootton, A. J. DeMello, Nat. Chem. 2013, 5,
905–915; c) J.-C. M. Monbaliu, A. Cukalovic, C. V. Stevens in Flow
Chemistry, Vol. 2, (Eds.: F. Darvas, G. Dormµn, V. Hessel), Walter de Gruyt-
er, Berlin, Germany, 2014, pp. 253–275.
[33] In many cases, imidazolium tetrafluoroborates were found to be easier
to recrystallize and less hygroscopic than their chloride analogues. See
for example: X. Bantreil, S. P. Nolan, Nat. Protoc. 2011, 6, 69–77.
[34] a) G. A. Grasa, R. M. Kissling, S. P. Nolan, Org. Lett. 2002, 4, 3583–3586;
b) G. A. Grasa, T. Güveli, R. Singh, S. P. Nolan, J. Org. Chem. 2003, 68,
2812–2819; c) R. Singh, R. M. Kissling, M.-A. Letellier, S. P. Nolan, J. Org.
Chem. 2004, 69, 209–212.
[35] G. W. Nyce, J. A. Lamboy, E. F. Connor, R. M. Waymouth, J. L. Hedrick,
Org. Lett. 2002, 4, 3587–3590.
[36] R. Singh, S. P. Nolan, Chem. Commun. 2005, 5456–5458.
[37] Y. Suzuki, K. Yamauchi, K. Muramatsua, M. Sato, Chem. Commun. 2004,
2770–2771.
[38] T. Kano, K. Sasaki, K. Maruoka, Org. Lett. 2005, 7, 1347–1349.
[39] Y. Wang, Z. Li, Chin. J. Catal. 2012, 33, 502–507.
[17] J. Yoshida, Y. Takakashi, A. Nagaki, Chem. Commun. 2013, 49, 9896–
9904.
[40] A. Adamo, P. L. Heider, N. Weeranoppanant, K. F. Jensen, Ind. Eng. Chem.
Res. 2013, 52, 10802–10808.
[18] R. Munirathinam, J. Huskens, W. Verboom, Adv. Synth. Catal. 2015, 357,
1093–1123.
[19] P. B. Palde, T. F. Jamison, Angew. Chem. Int. Ed. 2011, 50, 3525–3528;
Angew. Chem. 2011, 123, 3587–3590.
[20] V. Hessel, D. Kralisch, N. Kockmann, T. Nol, Q. Wang, ChemSusChem
2013, 6, 746–789.
[21] D. Webb, T. F. Jamison, Chem. Sci. 2010, 1, 675–680.
[22] a) D. R. Snead, T. F. Jamison, Chem. Sci. 2013, 4, 2822–2827; b) S.
Mascia, P. L. Heider, H. Zhang, R. Lakerveld, B. Benyahia, P. I. Barton, R. D.
Braatz, C. L. Cooney, J. M. B. Evans, T. F. Jamison, K. F. Jensen, A. S. Myer-
son, B. L. Trout, Angew. Chem. Int. Ed. 2013, 52, 12359–12563; Angew.
Chem. 2013, 125, 12585–12589; c) L. Heider, S. C. Born, S. Basak, B. Be-
nyahia, R. Lakerveld, H. Zhang, R. Hogan, L. Buchbinder, A. Wolfe, S.
[41] a) M. Movassaghi, M. A. Schmidt, Org. Lett. 2005, 7, 2453–2456; b) H.
Guo, Y. Wang, G.-F. Du, B. Dai, L. He, Tetrahedron 2015, 71, 3472–3477.
[42] J.-C. M. Monbaliu, A. Katritzky, Chem. Commun. 2012, 48, 11601–11622.
[43] M. Hans, J. Lorkowski, A. Demonceau, L. Delaude, Beilstein J. Org. Chem.
2015, 11, 2318–2325.
[44] See for instance: a) F. Gaccioli, R. Franchi-Gazzola, M. Lanfranchi, L.
Marchiò, G. Metta, M. A. Pellinghelli, S. Tardito, M. Tegoni, J. Inorg. Bio-
chem. 2005, 99, 1573–1584; b) S. Aspin, L. López-Suµrez, P. Larini, A. S.
Goutierre, R. Jazzar, O. Baudoin, Org. Lett. 2013, 15, 5056–5059.
Received: December 22, 2015
Published online on February 16, 2016
Chem. Eur. J. 2016, 22, 4508 – 4514
www.chemeurj.org
4514
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim