Shifting Chemical Equilibria in Flow - Efficient Decarbonylation Driven by Annular Flow Regimes

Article abstract of DOI:10.1002/anie.201407219

To efficiently drive chemical reactions, it is often necessary to influence an equilibrium by removing one or more components from the reaction space. Such manipulation is straightforward in open systems, for example, by distillation of a volatile product from the reaction mixture. Herein we describe a unique high-temperature/high-pressure gas/liquid continuous-flow process for the rhodium-catalyzed decarbonylation of aldehydes. The carbon monoxide released during the reaction is carried with a stream of an inert gas through the center of the tubing, whereas the liquid feed travels as an annular film along the wall of the channel. As a consequence, carbon monoxide is effectively vaporized from the liquid phase into the gas phase and stripped from the reaction mixture, thus driving the equilibrium to the product and preventing poisoning of the catalyst. This approach enables the catalytic decarbonylation of a variety of aldehydes with unprecedented efficiency with a standard coil-based flow device.

Full text of DOI:10.1002/anie.201407219

Angewandte
Chemie
DOI: 10.1002/anie.201407219
Continuous-Flow Reactions
Shifting Chemical Equilibria in Flow—Efficient Decarbonylation
Driven by Annular Flow Regimes**
Bernhard Gutmann, Petteri Elsner, Toma Glasnov, Dominique M. Roberge,* and
C. Oliver Kappe*
Abstract: To efficiently drive chemical reactions, it is often
necessary to influence an equilibrium by removing one or more
components from the reaction space. Such manipulation is
straightforward in open systems, for example, by distillation of
a volatile product from the reaction mixture. Herein we
describe a unique high-temperature/high-pressure gas/liquid
continuous-flow process for the rhodium-catalyzed decarbon-
ylation of aldehydes. The carbon monoxide released during the
reaction is carried with a stream of an inert gas through the
center of the tubing, whereas the liquid feed travels as an
annular film along the wall of the channel. As a consequence,
carbon monoxide is effectively vaporized from the liquid phase
into the gas phase and stripped from the reaction mixture, thus
driving the equilibrium to the product and preventing poison-
ing of the catalyst. This approach enables the catalytic
decarbonylation of a variety of aldehydes with unprecedented
efficiency with a standard coil-based flow device.
Scheme 1. Catalytic cycle for the rhodium-mediated decarbonylation of
aldehydes.[1,6]
T
he decarbonylation of aldehydes with stoichiometric
amounts of the Wilkinson complex ([RhCl(PPh3)3]) was
reported by Tsuji and Ohno in 1965.[1–3] The reaction proceeds
at or near room temperature and generates the decarbony-
lated product along with the stable inorganic complex trans-
[RhCl(CO)(PPh3)2]. Early studies by Walborsky and Allen as
well as others demonstrated that the reaction occurs with
a high degree of stereoselectivity and with overall retention of
the configuration of the carbon atom attached to the aldehyde
group.[4] Stereospecific decarbonylation has since matured
into a key strategic element in the synthesis of complex
natural products.[5]
quently reacts by a rapid, reversible oxidative addition into
À
the C(O) H bond of the aldehyde to form an acyl rhodium-
(III) hydride complex 3.[6] The reaction then proceeds by
a rate-limiting migratory extrusion of CO to give the 18e alkyl
or aryl rhodium(III) hydride 4, and finally produces the 16e
carbonyl complex trans-[RhCl(CO)(PPh3)2] (5) and the
decarbonylated product 7 by an irreversible reductive elim-
ination (Scheme 1).[6]
Early attempts to make the system catalytic failed because
À
the electronically depleted Rh CO complex 5 does not
À
participate in the oxidative addition into the aldehyde C H
Computational and experimental studies indicate that the
decarbonylation reaction starts with the dissociation of
a phosphine ligand from the square-planar rhodium complex
1 to generate a coordinatively unsaturated d8 complex 2
(Scheme 1).[6] The coordinatively unsaturated species subse-
bond, and loss of the CO ligand to regenerate the active
species does not proceed at preparatively useful reaction rates
at temperatures below 2208C.[1] Indeed, even heating of the
[RhCl(CO)(PPh3)2] complex to temperatures as high as
2608C with an excess of triphenylphosphane failed to convert
the rhodium carbonyl complex back into the tris(triphenyl-
phosphine)rhodium species.[1c] Significant progress toward
a more efficient decarbonylation reaction was made by the
introduction of cationic rhodium complexes of chelating
bisphosphines, such as [Rh(dppp)2]Cl (dppp = 1,2-bis(diphe-
nylphosphanyl)propane).[7] These complexes are less electron
rich and, accordingly, bind CO less strongly owing to
[*] Dr. B. Gutmann, Dr. T. Glasnov, Prof. Dr. C. O. Kappe
Christian Doppler Laboratory for Flow Chemistry and
Institute of Chemistry, University of Graz, NAWI Graz
Heinrichstrasse 28, 8010 Graz (Austria)
E-mail: oliver.kappe@uni-graz.at
Homepage: http://www.maos.net
Dr. P. Elsner, Dr. D. M. Roberge
Chemical Development, Lonza AG
3930 Visp (Switzerland)
À
decreased Rh CO p-back-bonding. More recently, a conven-
ient procedure for the decarbonylation of aldehydes was
reported for which the active catalyst was prepared in situ
from commercially available RhCl3·3H2O and the dppp
ligand.[8] The reaction was performed under an argon
atmosphere and required a reaction time of 16 h with catalyst
loadings of 0.4–10 mol% in boiling diglyme as the solvent
E-mail: dominique.roberge@lonza.com
[**] This research was supported by a grant from the Christian Doppler
Research Society (CDG).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201407219.
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

.
Angewandte
Communications
(b.p. 1628C).[8,9] However, harsh conditions and extended
reaction times, together with the requirement for high-boiling
solvents, which are difficult to remove from the product after
the reaction, make catalytic decarbonylation reactions gen-
erally incompatible with a high-yielding and sustainable
reaction procedure. Therefore, the majority of decarbon-
ylation protocols for the synthesis of complex molecules still
rely on stoichiometric or near-stoichiometric amounts of
a rhodium complex.[5]
Herein, we describe a unique biphasic gas–liquid contin-
uous-flow decarbonylation protocol that enables the efficient
decarbonylation of aldehydes in the presence of a catalytic
amount of a rhodium precursor within only 8–25 min. This
otherwise very challenging transformation is made possible
by inducing an annular flow regime in a coil-based micro-
reactor by using nitrogen as an inert carrier gas, which
efficiently removes CO from the equilibrium.
We chose 4-cyanobenzaldehyde (6a) as the model sub-
strate for initial optimization experiments (for optimization
details, see the Supporting Information). As it was clearly
evident from previously reported studies that a fast, catalytic
decarbonylation reaction would have to involve a high-
temperature regime, controlled sealed-vessel microwave
heating was used for process optimization.[10] The active
[Rh(dppp)]+ catalyst could be generated in situ from dppp
and various Rh sources, such as
pressure of the flow system was reduced, and the pressure
approached the vapor pressure of the solvent, did the
conversion increase from 30 to around 70% (see Table S5).
These experiments clearly demonstrate that the lack of
a headspace in the closed flow system, and the consequently
high concentration of CO in the liquid phase, inhibits the
decarbonylation reaction and prevents genuine catalysis. It is
well-known that, depending on temperature and the partial
pressure of CO, rhodium forms a range of different carbonyl
complexes, and the electronically depleted carbonyl com-
pounds generally do not participate in the decarbonylation
reaction.[1,13]
To prevent the poisoning of the catalyst by carbon
monoxide, we envisaged stripping the CO from the reaction
mixture with a stream of an inert gas fed into the flow system
(Figure 1).[14,15] Indeed, when N2 was fed into the flow reactor,
the conversion increased strikingly and approached or even
surpassed the results obtained under microwave batch con-
ditions on a small scale. Optimization of the process
parameters clearly illustrates that operation at the highest
temperature and lowest pressure is desired—namely, slightly
above the vapor pressure of the solvent at a given temper-
ature, to avoid its vaporization (Table 1; see also Tables S9–
S12).[14,16] Under such conditions, the experimentally deter-
mined residence time of the reaction mixture in the flow
RhCl3·3H2O or Rh(OAc)2 (see
Table S1 in the Supporting Informa-
tion). Initial optimization of the reac-
tion conditions provided best results
with 4 mol% of Rh(OAc)2 and
8 mol% of dppp in aprotic, inert
solvents, such as toluene. The reaction
was remarkably clean (ca. 95%
according to GC with a flame ioniza-
tion detector, GC–FID), and the only
detectable side product was 4,4’-dicya-
nobenzophenone, formed by a decar-
bonylative homocoupling of 4-cyano-
benzaldehyde.[11]
Figure 1. Gas–liquid continuous-flow decarbonylation of aldehydes with N2 as a stripping gas.
P=HPLC pump, SL=sample loop (poly(tetrafluoroethylene), 0.8 mm inside diameter), M=mixer
(1.0 mm i.d.), RT=residence tube (stainless steel, 1 mm i.d.), HE=heat exchanger (stainless
steel, 1 mm i.d.), BPR=back-pressure regulator).
Our initial batch experiments con-
firmed that the rhodium-catalyzed
decarbonylation of aromatic aldehydes
requires fairly high temperatures for
a reaction to occur at reasonable rates,
Table 1: Continuous-flow decarbonylation reactions at different temper-
or for the catalytic cycle to be completed at all. Furthermore,
even though a clean decarbonylation was possible at reaction
temperatures of around 2008C by sealed-vessel microwave
heating, the reaction rate greatly depended on the scale on
which the experiments were performed. With increasing
filling volume and consequently decreasing available head-
space in the sealed vessel, the reaction rate decreased
markedly. Whereas full conversion was observed on
a 0.6 mmol scale in a 10 mL microwave vial after 15 min at
1808C, the conversion gradually decreased to 26% when the
scale was increased to 1.4 mmol (see Table S4). Similarly,
conversions were around only 30% for initial continuous-flow
experiments in pressurized coil reactors, in which the head-
space is essentially eliminated.[12] Only when the back
atures and N2 flow rates (Figure 1; flow rate of liquid feed:
0.5 mLminÀ1).[a]
T/p
N2 [mLN minÀ1 [b]
]
[8C]/[bar]
8
15
25
180/6.8
200/9.1
220/12.5
65%
89%
92%
72%
93%
93%
89%
95%[c]
[a] The percentages given are the conversion of 6a into 7a under the
specified conditions of temperature, pressure, and nitrogen flow rate, as
determined by GC–FID. Flow reactions were performed in a 20 mL
stainless-steel reactor with a 0.2m solution of 4-cyanobenzaldehyde (6a)
in toluene (2 mL), Rh(OAc)2 (4 mol%), and dppp (8 mol%). [b] Gas flow
in mLN minÀ1 under normal conditions (Tn =08C and pn =1 atm). [c] The
pressure was increased from 9.1 to 9.7 bar.
2
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
system was essentially independent of the flow rate of the
gaseous feed, the back pressure, and the reaction temper-
ature. With flow rates of 15 mLN minÀ1 for the N2 feed and
0.5 mLminÀ1 for the liquid feed at 1808C and a back pressure
of around 6 bar, carbon monoxide could be detected at the
outlet of the tubing already 3 to 4 min after the reaction
mixture had been injected, whereas the product left the
reactor after roughly 8 to 9 min (see Table S9 and Figure S6 in
the Supporting Information). This result indicates that the
liquid travels as an annular film on the channel wall, whereas
the carbon monoxide is carried with the stream of nitrogen
through the center of the tubing (annular flow).[16] With an
increasing flow rate of the gaseous feed, as well as with
increasing temperature and/or decreasing pressure, a transi-
tion from annular flow or annular-dispersed flow (gas core
contains entrained liquid droplets) to dispersed flow was
apparent. In the case of dispersed flow, the gas core contains
the volatile liquid, and the nonvolatile catalyst is deposited on
the wall of the capillary (dry-out).[16]
Essentially full conversion of 4-cyanobenzaldehyde (6a)
into benzonitrile (7a) was observed on a 2 mmol scale with
Rh(OAc)2 (4 mol%) and dppp (8 mol%) at 2008C and flow
rates of 0.5 mLminÀ1 for the liquid feed and 25 mLN minÀ1 for
the N2 feed. Pure benzonitrile was isolated in 84% yield after
column chromatography (Table 2). As expected, benzalde-
hydes lacking electron-withdrawing groups reacted signifi-
cantly slower than cyanobenzaldehyde. With 2,5-dimethoxy-
benzaldehyde (6b), conversions of > 90% were observed
after the flow rate of the liquid feed was decreased to
0.3 mLminÀ1, and the pure product was isolated as described
above in 87% yield. The scale of this reaction was then
increased to 10 mmol (20 mL) to simulate steady-state
operation (the volume of the reaction mixture was about
5 times the effective residence volume). The decarbonylated
product 7b was formed with essentially identical conversion
on this scale (93% according to GC–FID), and the product
was isolated in 82% yield, thus demonstrating that the CO is
effectively removed from the reaction space. For some of the
other tested aromatic aldehydes, 6d–f, full conversion did not
readily occur under these conditions, and conversions could
not be improved by simply increasing the reaction time or by
increasing the reaction temperature above 2008C. Thus, the
catalyst loading was increased to 8 mol% to promote high
conversion. Generally, the reactions were very clean and the
decarbonylated compounds were the only products detected
in the reaction mixtures.
The stereospecific decarbonylation of a-alkyl trans-cin-
namaldehydes to give the corresponding cis alkenes has been
reported previously.[1b,4] Since the trans-cinnamaldehydes can
be readily prepared by an aldol condensation of benzalde-
hyde and an enolizable aliphatic aldehyde, the condensation/
decarbonylation sequence might constitute an attractive two-
step synthesis of cis alkenes. The flow-decarbonylation
procedure gave complete conversion of 4-cyano- and 4-
nitro-a-ethyl-trans-cinnamaldehyde (6g and 6h) into the
decarbonylated products 7g and 7h at temperatures as low
as 1808C (Table 2). With both substrates, a roughly 10:90
mixture of the cis and the trans alkene was obtained, with the
trans compound as the major component (Table 2). Similarly,
a-methyl-trans-cinnamaldehyde gave a roughly 30:70 mixture
of cis- and trans-b-methylstyrene. Furthermore, 2-methyl-
indanone was formed in significant amounts from the a-
methylcinnamaldehyde by an inter-
molecular hydroacylation (see the
Table 2: Decarbonylation of aldehydes under continuous-flow conditions (Figure 1).[a]
Supporting Information). Addition-
ally, decarbonylation of the ali-
phatic aldehyde 3-phenylpropional-
dehyde with Rh(OAc)2/dppp gave
a 1:1 mixture of ethylbenzene and
styrene (see Table S15). The Wil-
kinson complex generally affords
the respective alkanes from the
aliphatic aldehydes selectively.[1]
The efficiency of this novel flow
protocol was then highlighted by
the synthesis of various chromenes
by decarbonylation of the corre-
Substrate
Conversion [%]
(yield [%])[b]
Substrate
Conversion [%]
(yield) [%][b]
6a[c]
6c
94 (84)
85 (71)
6b
92 (87)
88 (82)
6d[d]
6e[d]
95 (76)
6 f[d]
69 (53)
sponding
2-formylchromenes.
100 (83)
cis/trans 11:89
100 (61)
cis/trans 9:91
Chromenes constitute an important
class of natural products and are
key intermediates in the synthesis
of many natural products, including
cannabinoids, anthocyanides, and
flavones.[9] A two-step synthesis of
chromenes by a Michael addition/
aldol condensation of a salicylalde-
hyde and an acrylaldehyde, fol-
lowed by subsequent rhodium-cata-
lyzed decarbonylation, was utilized
by Brꢀse and co-workers for the
6g[e]
6i
6h[e]
6j
100 (84)
100 (85)
[a] Flow reactions were performed at 2008C in a 25 mL stainless-steel reactor with 4 mL of a 0.5m
solution of aldehydes 6a–j, Rh(OAc)2 (4 mol%), and dppp (8 mol%) in toluene as the solvent; the
pressure was adjusted to approximately 10 bar, and the flow rates of the liquid and gaseous feed were
0.3 and 25 mLminÀ1, respectively (residence time: 8–15 min). [b] The conversion was determined by
GC–FID. The yield of the isolated product is given in parentheses. [c] The reaction was performed in
a 20 mL stainless-steel coil with flow rates of 0.5 and 25 mLminÀ1 for the liquid and gaseous feed,
respectively. [d] The reaction was performed with Rh(OAc)2 (8 mol%) and dppp (16 mol%). [e] The
reaction was performed at 1808C and 7 bar.
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

.
Angewandte
Communications
synthesis of eulatachromene and various related chromene
derivatives.[9] The decarbonylation reaction was performed by
a CSTR cascade may be feasible, but would significantly
increase complexity if operated under pressure. In this study,
these principles have been implemented in a novel continu-
ous-flow high-temperature/high-pressure process in a stan-
dard coil-based microreactor environment.[18] The unique
transport capabilities at these scales (i.e., large interfacial
areas and short diffusion paths) and the corresponding
efficient gas–liquid interaction enabled the desired trans-
formation to be performed with exceptional efficiency.
a
procedure developed by Madsen and co-workers
(RhCl3·3H2O (5 mol%), dppp (10 mol%), 16 h at the reflux
temperature of diglyme).[8] Our flow decarbonylation proto-
col provided complete decarbonylation of 2-substituted for-
mylchromenes in residence times of around 10 min at
a reaction temperature of 2008C. The 2-methyl- and 2,2-
dimethyl-2H-chromenes 6j and 6i were formed with good
selectivity and isolated by chromatography in 85 and 84%
yield, respectively (Table 2).
Finally, we performed the decarbonylation of the 16-
formylandrosta-5,16-diene derivative 6k to yield the C17-
heteroaryl steroidal CYP17 inhibitor/antiandrogen 7k
(Scheme 2). This and related compounds are experimental
Received: July 15, 2014
Published online: && &&, &&&&
Keywords: continuous-flow reactions · decarbonylation ·
.
homogeneous catalysis · process intensification · rhodium
[1] a) J. Tsuji, K. Ohno, Tetrahedron Lett. 1965, 6, 3969 – 3971; b) J.
Tsuji, K. Ohno, Tetrahedron Lett. 1967, 8, 2173 – 2176; c) K.
Ohno, J. Tsuji, J. Am. Chem. Soc. 1968, 90, 99 – 107.
[2] For a recent review on the rhodium-catalyzed decarbonylation,
see: A. Korotvicka, D. Necas, M. Kotora, Curr. Org. Chem. 2012,
16, 1170 – 1214.
[3] For examples of palladium- and iridium-catalyzed decarbon-
ylation reactions, see: Pd: a) D. Maiti, Green Chem. 2012, 14,
2314 – 2320; b) A. Modak, A. Deb, T. Patra, S. Rana, S. Maity, D.
Maiti, Chem. Commun. 2012, 48, 4253 – 4255; c) S. Dahoah, Z.
Nairoukh, M. Fanun, M. Schwarze, R. Schomꢀcker, J. Blum, J.
Mol. Catal. A 2013, 380, 90 – 93; Ir: d) T. Iwai, T. Fujihara, Y.
Tsuji, Chem. Commun. 2008, 6215 – 6217; e) E. P. K. Olsen, R.
Madsen, Chem. Eur. J. 2012, 18, 16023 – 16029.
Scheme 2. Synthesis of the C17-heteroaryl steroidal CYP17 inhibitor/
antiandrogen 7k.
drugs for the treatment of prostate cancer and are currently
under clinical development.[17] A scalable and cost-efficient
protocol for the generation of these compounds is therefore
of utmost importance. The C17-heteroaryl 6-formylandrosta-
diene can readily be synthesized in two steps by the
Vilsmeier–Haack formylation of commercially available
dehydroisoandrosterone 3-acetate and subsequent nucleo-
philic replacement of the C17 chloride substituent with
imidazole.[17] Continuous-flow decarbonylation under the
conditions outlined above with toluene/MeCN (3:1) as the
solvent proceeded with remarkable selectivity, and no sign of
double-bond isomerization was apparent. With a catalyst
loading of 4 mol% of Rh(OAc)2 at a reaction temperature of
2008C, the desired target compound 7k was formed as
virtually the sole product after a residence time of about
25 min. The product 7k was isolated by elution through
a short column of silica gel in 83% yield (1 mmol scale).
In conclusion, we have developed an efficient catalytic
high-temperature/high-pressure continuous-flow Tsuji–Wil-
kinson decarbonylation protocol that can be applied to
a variety of aldehydes. The carbon monoxide released
during the reaction was removed from the reaction mixture
by a continuous stream of N2 as a stripping gas, thus driving
the equilibrium to the product and preventing poisoning of
the catalyst. Separation techniques involving mass transfer
(e.g. distillation, stripping, and absorption) are common for
the selective removal of components from a reaction mixture
and shifting of an equilibrium in the desired direction.[14] In
a batch mode, this operation would have to be performed
under reflux conditions in a high-boiling solvent, and thus
under complex reaction conditions that are difficult to scale
up. In the flow mode, the use of a falling-film reactor or
[4] a) H. M. Walborsky, L. E. Allen, Tetrahedron Lett. 1970, 11,
823 – 824; b) H. M. Walborsky, L. E. Allen, J. Am. Chem. Soc.
1971, 93, 5465 – 5468.
[5] For selected examples of rhodium-catalyzed decarbonylation as
part of multistep syntheses of complex molecules, see: a) C.-M.
Zeng, M. Han, D. F. Covey, J. Org. Chem. 2000, 65, 2264 – 2266;
b) T. Kato, M. Hoshikawa, Y. Yaguchi, K. Izumi, Y. Uotsu, K.
Sakai, Tetrahedron 2002, 58, 9213 – 9222; c) B. J. Morgan, C. A.
Mulrooney, M. C. Kozlowski, J. Org. Chem. 2010, 75, 44 – 56;
d) R. M. Wilson, R. K. Thalji, R. G. Bergman, J. A. Ellman, Org.
Lett. 2006, 8, 1745 – 1747; e) J. P. Malerich, T. J. Maimone, G. I.
Elliott, D. Trauner, J. Am. Chem. Soc. 2005, 127, 6276 – 6283;
f) M. Harmata, S. Wacharasindhu, Org. Lett. 2005, 7, 2563 –
2565; g) H. Zhang, A. Padwa, Tetrahedron Lett. 2007, 47,
3905 – 3908.
[6] P. Fristrup, M. Kreis, A. Palmelund, P.-O. Norrby, R. Madsen, J.
Am. Chem. Soc. 2008, 130, 5206 – 5215.
[7] D. H. Doughty, L. H. Pignolet, J. Am. Chem. Soc. 1978, 100,
7083 – 7085.
[8] M. Kreis, A. Palmelund, L. Bunch, R. Madsen, Adv. Synth. Catal.
2006, 348, 2148 – 2154.
[9] M. C. Brçhmer, N. Volz, S. Brꢀse, Synlett 2009, 1383 – 1386, and
references therein.
[10] T. N. Glasnov, C. O. Kappe, Chem. Eur. J. 2011, 17, 11956 –
11968.
[11] The rhodium-catalyzed decarbonylative homocoupling of ben-
zaldehydes in the presence of an oxidant has been exploited for
the preparation of benzophenones: L. Yang, T. Zeng, Q. Shuai,
X. Guo, C.-J. Li, Chem. Commun. 2011, 47, 2161 – 2163.
[12] For a recent example highlighting the reverse effect of head-
space in microwave batch versus flow experiments, see: N.
Zaborenko, M. W. Bedore, T. F. Jamison, K. F. Jensen, Org.
Process Res. Dev. 2011, 15, 131 – 139.
4
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!

Angewandte
Chemie
[13] a) F. Abu-Hasanayn, M. E. Coldman, A. S. Goldman, J. Am.
Chem. Soc. 1992, 114, 2520 – 2524; b) T. C. Fessard, S. P.
Andrews, H. Motoyoshi, E. M. Carreira, Angew. Chem. Int.
Ed. 2007, 46, 9331 – 9334; Angew. Chem. 2007, 119, 9492 – 9495;
c) E. Taarning, R. Madsen, Chem. Eur. J. 2008, 14, 5638 – 5644.
[14] J. Benitez, Principles and Modern Applications of Mass Transfer
Operations, 2ednd edWiley, Hoboken, 2009.
[15] Besides thermal means to regenerate the catalytically active
species, photochemical and chemical methods to remove CO
from a carbonyl complex have been reported (see the Support-
ing Information for details).
Zhang, Transport Phenomena in Multiphase Systems, Elsevier,
Burlington, MA, 2006.
[17] a) V. C. O. Njar, K. Kato, I. P. Nnane, D. N. Grigoryev, B. J.
Long, A. M. H. Brodie, J. Med. Chem. 1998, 41, 902 – 912;
b) V. D. Handratta, T. S. Vasaitis, V. C. O. Njar, L. K. Gediya, R.
Kataria, P. Chopra, D. Newman, Jr., R. Farquhar, Z. Guo, Y.
Qiu, A. M. H. Brodie, J. Med. Chem. 2005, 48, 2972 – 2984; c) P.
Purushottamachar, A. M. Godbole, L. K. Gediya, M. S. Martin,
T. S. Vasaitis, A. K. Kwegyir-Afful, S. Ramalingam, Z. Ates-
Alagoz, V. C. O. Njar, J. Med. Chem. 2013, 56, 4880 – 4898.
[18] P. Plouffe, A. Macchi, D. M. Roberge in Flow Chemistry, Vol. 1
(Eds.: F. Darvas, V. Hessel, G. Dorman), De Gruyter, Berlin,
2014, pp. 139 – 152.
[16] Depending on the total flow rate and volumetric flow ratio,
several gas – liquid flow patterns are possible; see: A. Faghri, Y.
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
5
These are not the final page numbers!

.
Angewandte
Communications
Communications
Continuous-Flow Reactions
Swept away from the action: A biphasic
gas/liquid continuous-flow protocol ena-
bled the efficient decarbonylation of
aldehydes with a rhodium catalyst. This
transformation is made possible by the
induction of an annular flow regime in
a coil-based microreactor by using nitro-
gen as an inert carrier gas to remove CO
from the equilibrium (see picture; dppp=
1,2-bis(diphenylphosphanyl)propane).
B. Gutmann, P. Elsner, T. Glasnov,
D. M. Roberge,*
C. O. Kappe*
&&&& — &&&&
Shifting Chemical Equilibria in Flow—
Efficient Decarbonylation Driven by
Annular Flow Regimes
6
www.angewandte.org
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 1 – 6
These are not the final page numbers!