One-pot multistep synthetic strategies for the production of fenpropimorph using an ionic liquid solvent

Article abstract of DOI:10.1021/op050172m

The one-pot synthesis of the fungicide fenpropimorph has been achieved using two different synthetic strategies in an ionic liquid. The first pathway consists of a Heck coupling followed by reductive animation; the second pathway consists of an aldol condensation followed by hydrogenation/reductive animation. Homogeneous and heterogeneous palladium catalysts have been utilised in the ionic liquid to provide a catalyst/solvent system that is suitable for recycling and process optimisation.

Full text of DOI:10.1021/op050172m

Organic Process Research & Development 2006, 10, 94−102
One-Pot Multistep Synthetic Strategies for the Production of Fenpropimorph
Using an Ionic Liquid Solvent
Stewart A. Forsyth, H. Q. Nimal Gunaratne, Christopher Hardacre,* Angela McKeown, and David W. Rooney
School of Chemistry and Chemical Engineering/QUILL Research Centre, Queen’s UniVersity, Belfast,
BT9 5AG, Northern Ireland, UK
Abstract:
lipase-catalysed acylation of prochiral 2-(4-tert-butylbenzyl)-
1,3-propanediol and a lipase-catalysed kinetic resolution of
racemic 3-(4-tert-butylphenyl)-2-methylpropionic acid.7 The
research reported herein demonstrates the possibility of
“redesigning” the current synthetic pathway for this material
by using an ionic liquid as the solvent of choice to achieve
an improved process.
The one-pot synthesis of the fungicide fenpropimorph has been
achieved using two different synthetic strategies in an ionic
liquid. The first pathway consists of a Heck coupling followed
by reductive amination; the second pathway consists of an aldol
condensation followed by hydrogenation/reductive amination.
Homogeneous and heterogeneous palladium catalysts have been
utilised in the ionic liquid to provide a catalyst/solvent system
that is suitable for recycling and process optimisation.
Although ionic liquids have only been studied in detail
for organic synthesis/catalytic reactions in the past decade,
a wide range of reactions have been demonstrated,8 and a
number of industrial processes have been established, notably
by BASF and Degussa.9 In the case of catalytic processes,
the studies in ionic liquids have been dominated by inves-
tigations into homogeneous catalyst systems whereby the
ionic liquid immobilises the catalyst and allows good
separation of the solvent/catalyst system from the products.8
More recently, the combination of heterogeneous catalysts
in ionic liquids has been investigated. In particular, selective
hydrogenation,10,11 selective oxidation,12 Heck coupling,13-16
Friedel-Crafts,17-19 and cyclization reactions20 have been
reported. In virtually all cases the reactions, whether catalytic
or stoichiometric, have been single-step processes, and few
investigations have been performed whereby the same ionic
liquid is used in a number of consecutive reactions.21 It is
unrealistic to envisage a process whereby a number of steps
Introduction
Fenpropimorph, 4-[3-(4-tert-butylphenyl)-2-methylpro-
pyl]-2,6-dimethylmorpholine (3), is a pesticide, specifically
categorised as a morpholine fungicide. It finds major use as
the active ingredient in agricultural formulations for the
control of disease in cereal crops. Industrially, fenpropimorph
is produced as the racemic mixture of the cis isomer, although
it is known that (S)-(-)-enantiomer is known to have the
higher fungicidal activity.1 Currently fenpropimorph, and
other similar molecules, is produced in excess of 40 000 tons
per annum and is used extensively worldwide. Several patents
have been filed concerning the synthesis of fenpropimorph;
however, all use multistep, inefficient processes that produce
large quantities of waste and yield product that often requires
refining.2-5 For example, the current industrial process uses
3-(4-tert-butylphenyl)-2-methylpropenal as the starting mate-
rial and couples the aldehyde to 1-(2-hydroxypropylamino)-
propan-2-ol via a condensation reaction at 165 °C using a
solid acid catalyst. The resulting intermediate oxazolidine
product is then isolated by distillation before being hydro-
genated at 240 °C and 15-25 atm H2 using a 5%Pd/C
catalyst. In this process the overall yield is 52%. When taken
with the 50% yield for the initial aldol condensation to form
3-(4-tert-butyl-phenyl)-2-methylpropenal, the overall yield
from 4-tert-butylbenzaldehyde is <30%. Vinkovic´ and Sˇunjic´
also reported a stereocontrolled synthesis of cis-fenpropi-
morph using a six-step synthesis which affords only 6%
overall yield.6 In addition, two biocatalytic routes to form
enantiopure cis-fenpropimorph have been reported using a
(7) Avadagic, A.; Gelopujic, M.; Sˇunjic´, V. Synthesis 1995, 1427; Avadagic,
A.; Cotarca, L.; Ruzˇic´, K. S.; Gelo, M.; Sˇunjic´, V. Biocatalysis 1994, 9,
49.
(8) Welton, T., Ed. Ionic Liquids in Synthesis; Wiley-VCH: Weinheim, 2003.
(9) Seddon, K. R. Nat. Mater. 2003, 2, 363.
(10) Anderson, K.; Goodrich, P.; Hardacre, C.; Rooney, D. W. Green Chem.
2003, 5, 448.
(11) Carlin, R. T.; Fuller, J. Chem. Commun. 1997, 1345.
(12) Cimpeanu, V.; Paˆrvulescu, V. I.; Amoro´s, P.; Beltra´n, D.; Thompson, J.
M.; Hardacre, C. Chem. Eur. J. 2004, 10, 4640; Hardacre, C.; Mullan, E.
A.; Rooney, D. W.; Thompson, J. M. J. Catal. 2005, 232, 355; Seddon, K.
R.; Stark, A. Green Chem. 2002, 4, 119.
(13) Xie, X. G.; Lu, J. P.; Chen, B.; Han, J. J.; She, X. G.; Pan, X. F. Tetrahedron
Lett. 2004, 45, 809.
(14) Okubo, K.; Shirai, M.; Yokoyama, C. Tetrahedron Lett. 2002, 43, 7115.
(15) Hagiwara, H.; Shimizu, Y.; Hoshi, T.; Suzuki, T.; Ando, M.; Ohkubo, K.;
Yokoyama, C. Tetrahedron Lett. 2001, 42, 4349.
(16) Forsyth, S. A.; Gunaratne, H. Q. N.; Hardacre, C.; McKeown, A.; Rooney,
D. W.; Seddon, K. R. J. Mol. Catal. A: Chem. 2005, 231, 61.
(17) Hardacre, C.; Rooney, D. W.; Thompson, J. M.; Katdare, S. P. Queens
University, Belfast (Quill), WO 03028882.
* To whom correspondence should be addressed. Telephone: +44 28 9097
4592. Fax: +44 28 9097 4687. E-mail: c.hardacre@qub.ac.uk.
(1) Himmele, W.; Pommer, E.-H. Angew. Chem. 1980, 92, 176.
(2) Himmele, W.; Kohlmann, F.-W.; Herberle, W. BASF AG, DE 2822326,
1979.
(3) Hugener, H.; Roth, W. Hoffmann La Roche, EP 0111736, 1984.
(4) Vitomir, S.; Mirjana, G. Caffaro Spa Ind Chim, EP 0645458, 1995.
(5) Andreas, K.; Wolfgang, S. BASF AG, DE 19720475, 1998.
(6) Vinkovic´, V.; Sˇunjic´, V. Tetrahedron 1997, 53, 689.
(18) Hardacre, C.; Katdare, S. P.; Milroy, D.; Nancarrow, P.; Rooney, D. W.;
Thompson, J. M. J. Catal. 2004, 227, 44.
(19) Shen, H.-Y.; Judeh, Z. M. A.; Chiang, C. B.; Xia, Q.-H. J. Mol. Catal. A:
Chem. 2004, 212, 301.
(20) Yadav, J. S.; Reddy, B. V. S.; Reddy, M. S.; Niranjan, N. J. Mol. Catal. A:
Chem. 2004, 210, 99.
(21) Zulfiqar, F.; Kitazume, T. Green Chem. 2000, 2, 137; Rodriquez, M.; Sega,
A.; Taddei, M. Org. Lett. 2003, 5, 4029; Peng, Y.; Song, G.; Huang, F. J.
Chem. Res. 2004, 677.
94
•
Vol. 10, No. 1, 2006 / Organic Process Research & Development
10.1021/op050172m CCC: $33.50 © 2006 American Chemical Society
Published on Web 11/30/2005

Scheme 1. One-pot sequential pathways to produce fenpropimorph
are replaced by ionic liquid processes, particularly where a
number of different ionic liquids would have to be used. The
major advantage of ionic liquids lies in the changes in rate/
selectivity which may be achieved compared to those
achieved with organic solvents, with the added economic
benefit and the reduction in downstream processing costs.
One method to reduce the number of process steps is to
combine consecutive steps in the same solvent without
intermediate purification, leading to an increase in overall
product yield. In this way, the energy costs will be reduced
and purification should be simplified.
The two synthetic pathways used for the synthesis of
fenpropimorph are shown in Scheme 1. Each pathway
contains a carbon-carbon bond-forming reaction via either
a Heck coupling (Path 1) or an aldol condensation (Path 2).
This is followed by a reductive amination on the aldehyde
formed to yield the aminated product. Heck reactions have
been studied extensively in ionic liquids using a range of
aromatic substrates and olefins.22,23 In the present study the
coupling reagent is an allylic alcohol which has been studied
to a much smaller extent than the more reactive allylic esters
for example. In particular, the formation of the â-Heck
product is required. Ionic liquids have been shown to be
versatile in differentiating between the R and â forms; for
example, Xaio et al.24 have recently shown that, by varying
the ionic liquid and the phosphine used, the system may be
tuned to give almost total selectivity to either isomer. Aldol
reactions have also been studied using ionic liquids. For
example, both the stereoselective and nonstereoselective aldol
condensations have been described,25 showing increased
yields of oligomers >C9 and high enantiomeric excess (ee).
In general, the ionic liquid systems show little loss in activity
or ee on recycle of the catalyst/ionic liquid system. However,
the selectivity with respect to the cross-aldol product
compared with that of the self-condensation products has
not been addressed.
As part of the preparation for the research performed
herein, the specific Heck and aldol condensations leading
to Lilial, 3-(4-tert-butylphenyl)-2-methylpropanal (1), and
3-(4-tert-butylphenyl)-2-methylpropenal (2), respectively,
have been examined recently. The Heck coupling of 2-meth-
yl-prop-2-en-1-ol with 4-tert-butyl-iodobenzene was dem-
onstrated to proceed efficiently in ionic liquids, and after
extracting with organic and aqueous solvents, recycling of
the ionic liquid-homogeneous catalyst system was pos-
sible.16 In the case of the aldol condensation of 4-tert-
butylbenzaldehyde and propanal, a range of ionic liquids,
particularly those based on bis{(trifluoromethyl)sulfonyl}-
amide and trifluorotris(perfluoroethyl)phosphate anions, were
shown to promote the cross-aldol product over the self-
condensation of propanal.26 Although a range of homoge-
neously and heterogeneously catalysed hydrogenations have
also been performed in ionic liquids, only one report of
reductive amination reactions has been reported, using a
homogeneous iridium complex in a range of ionic liquids.
Compared with organic solvents, the ionic liquid-mediated
reactions showed higher selectivities at high conversions.27
This paper describes an overall catalytic process used to
form fenpropimorph using a single ionic liquid solvent,
optimised from the individual reaction steps. The effect of
recycling the ionic liquid-catalyst system is examined and
the details on how the scaled up process may be operated
are presented.
Experimental Section
Unless otherwise stated, all reagents, solvents and catalysts
were obtained from Aldrich and used without further
purification. 1-Butyl-1-methylpyrolidinium bis{(trifluoro-
(22) Farina, V. AdV. Synth. Catal. 2004, 346, 1553.
(23) Biffis, A.; Zecca, M.; Basato, M. J. Mol. Catal. A: Chem. 2001, 173, 249.
(24) Xiao, J.; Twamley, B.; Shreeve, J. Org. Lett. 2004, 6, 3845.
(25) Mehnert, C. P.; Dispenziere, N. C.; Cook, R. A. Chem. Commun. 2002,
1610.
(26) Davey, P.; Forsyth, S. A.; Gunaratne, H. Q. N.; Hardacre, C.; McKeown,
A.; McMath, S. E. J.; Rooney, D. W.; Seddon, K. R. Green Chem. 2005,
7, 224.
(27) Imao, D.; Fujihara, S.; Yamamoto, T.; Ohta, T.; Ito, Y. Tetrahedron 2005,
61, 6988.
Vol. 10, No. 1, 2006 / Organic Process Research & Development
•
95

methyl)sulfonyl}amide ([bmpyrr][NTf2]), 1-butyl-3-meth-
ylimidazolium tetrafluoroborate ([C4mim][BF4]), and 1-butyl-
3-methylimidazolium bis{(trifluoromethyl)sulfonyl}amide
([C4mim][NTf2]) were prepared in house using standard
literature methods.28-30 The [NTf2]- ionic liquids contained
<0.2 wt % water and the [C4mim][BF4] 0.66 wt % water as
determined by Karl Fischer analysis. For the [NTf2]- ionic
liquids, the residual halide content was <7.2 ppm, and for
[C4mim][BF4] the residual chloride content was <7.1 ppm
as determined by ion chromatography.31 The heterogeneous
palladium catalysts 5 wt % Pd/C (BET surface area 652 m2
g-1; metal dispersion 10%) and 5 wt % Pd/Al2O3 324 (BET
surface area 134 m2 g-1; metal dispersion 14%) were supplied
by Johnson Matthey. PdCl2 was supplied by Fluka and used
as received. The gases used were all research grade and were
supplied by BOC.
1H NMR spectra were recorded on a Bruker Advance
DRX 500 MHz NMR spectrometer. GC-FID samples were
analysed using a Hewlett-Packard 6890 GC fitted with an
RTX-5 column (30 m, 0.25 µm diameter). The retention
times of the peaks were compared against those of authentic
samples. GC-MS (Perkin-Elmer Turbo Mass) was per-
formed using a PE5MS column (length, 30 m; thickness,
0.25 µm; IDO, 0.32 mm). The conversions were adjusted
using calculated sensitivity factors for starting materials and
products.
Three reactors were used for screening, small-scale kinetic
studies, and scale-up investigations. The screening experi-
ments were carried out in a Baskerville mini-autoclave with
magnetic stirring, denoted as the Baskerville reactor through-
out the text. The small-scale kinetics were performed in a
Hazards Evaluation Laboratory 50 cm3 autoclave equipped
with a gasifying overhead impeller and four baffles, denoted
as the HEL reactor throughout the text. Scaled up reactions
were performed using a 300 cm3 Parr stainless steel autoclave
with a gasifying overhead impeller and four baffles, denoted
as the Parr reactor throughout the text.
iodobenzene (10 mmol), 2-methylprop-2-en-1-ol (12 mmol),
triethylamine (12 mmol), PdCl2 dissolved in 3 cm3 [bmpyrr]-
[NTf2] (0.05 mmol equivalent of palladium). The tube was
then loaded into the Baskerville reactor, and the reaction
mixture was stirred at 100 °C for 4 h under nitrogen to ensure
complete conversion of 4-tert-butyl-iodobenzene. The reac-
tion was cooled to room temperature, 2,6-dimethylmorpho-
line added, and the reaction mixture pressurised to 10 atm
hydrogen atmosphere. The reductive amination reaction was
performed at room temperature for 16 h. The reaction was
analysed following both the initial Heck coupling and the
reductive amination reaction.
Procedure for One-Pot Aldol-Reductive Amination
Reaction: Path 2. The Parr reactor was charged with
[bmpyrr][NTf2] (30 cm3), 4-tert-butylbenzaldehyde (400
mmol), 2,6-dimethylmorpholine (240 mmol), and 5 wt %
Pd/Al2O3 (8.5 g, 4 mmol equivalent of palladium). The
reaction mixture was stirred at 600 rpm whilst propanal (400
mmol) was pumped into the reactor over 4 h using a
peristaltic pump. The reaction was allowed to stir until GC
analysis sample showed complete conversion, typically after
12 h. A further quantity of 2,6-dimethylmorpoline (160
mmol) was added. Then the vessel was sealed, purged with
nitrogen, and pressurised with 10 atm hydrogen. The reactor
was stirred at 600 rpm at 60 °C for 20 h. The reaction was
analysed following both the initial aldol condensation and
the reductive amination reaction.
Recycle Procedure. For both the reductive amination
reactions and the one-pot syntheses using Path 2, the product
was removed from the ionic liquid-catalyst mixture by
decantation. Any catalyst entrained in the product layer was
filtered off and returned into the ionic liquid layer before
the reaction was restarted to maintain the catalyst loading.
This was typically <5% of the original catalyst mass. Fresh
starting materials were added to the catalyst and ionic liquid
reaction mixture and the initial reaction procedure was
repeated.
Reductive Amination Procedure. Typically, ionic liquid
or organic solvent (2 cm3), catalyst (27 mg), equimolar
substrate, and amine (2.5 mmol; substrate/catalyst ≈ 100)
were introduced to the autoclave and purged with nitrogen.
The reactor was pressurised with hydrogen and heated. Once
the set temperature had been reached, the stirring was started
and left for the desired time, after which the stirring was
stopped and the pressure released. Samples were taken
periodically during the reaction whilst still under pressure.
The ionic liquid samples were extracted with cyclohexane
in a volume ratio 1:5 (IL/cyclohexane) and filtered before
analysis by GC-FID and GC-MS.
Results
Reductive Amination. Table 1 summarises the reductive
amination of different aldehydes by different bases using
10% Pd/C, 5% Pd/Al2O3, and PdCl2 in the Baskerville mini-
autoclave performed in [bmpyrr][NTf2] over 16 h. Clearly
both homogeneous and heterogeneous catalysts result in high
yields of reductive amination products for a range of
aldehyde-secondary amine combinations in both ionic liquids
and conventional solvents. However, following reaction with
PdCl2 in ethanol and [bmpyrr][NTf2] a significant formation
of palladium black was observed. Although this was not
unexpected in the case of molecular solvents, it may be
surprising that the ionic liquid solvent did not stabilise the
cationic palladium in solution. This may be due to the
decomposition of the palladium chloride on dissolution in
the ionic liquid to form palladium nanoparticles which
aggregate under hydrogen. Nanoparticle formation has been
observed on dissolution of palladium salts by EXAFS and
TEM studies previously in ionic liquids.32 In these cases,
the ionic liquid stabilised small Pd nanoparticles (<2 nm),
Procedure for the One-Pot Heck-Reductive Amination
Reaction: Path 1. To a glass tube were added 4-tert-butyl-
(28) MacFarlane, D. R.; Meakin, P.; Sun, J.; Amini, N.; Forsyth, M. J. Phys.
Chem. B 1999, 103, 4164.
(29) Bonhoˆte, P.; Dias, A. P.; Papageorgiou, N.; Kalyanasundram, K.; Gra¨tzel,
M. Inorg. Chem. 1996, 32, 1168.
(30) Cammarata, L.; Kazarian, S. G.; Salter, P. A.; Welton, T. Phys. Chem. Chem.
Phys. 2001, 3, 5192.
(31) Villagra´n, C.; Deetlefs, M.; Pitner, W. R.; Hardacre, C. Anal. Chem. 2004,
76, 2118; Villagra´n, C.; Banks, C. E.; Deetlefs, M, Driver, G., Pitner, W.
R.; Compton, R. G.; Hardacre, C. ACS Symp. Ser. 2005, 902, 244.
96
•
Vol. 10, No. 1, 2006 / Organic Process Research & Development

Table 1. Percentage conversion with respect to the aldehyde and isolated yield of saturated reductive amination product
following the reaction of a range of aldehydes with secondary aminesa using PdCl2, 5%Pd/C, and 5%Pd/Al2O3 (100:1;
aldehyde:Pd) in ethanol and [bmpyrr][NTf2] at 60 °C and 10 atm hydrogen for 16 h in the Baskerville reactor
aldehyde
amine
catalyst, solvent
% conversion
% yield
cis/trans
1
1
1
1
1
1
1
2
2
2
2
2
2
2
8
9
A
A
A
A
A
B
C
none
A
Et3N
(iPr)2EtN
A
B
C
A
A
PdCl2, EtOH
PdCl2, [bmpyrr][NTf2]
Pd/C, EtOH
92
92
99
99
99
99
99
60
85
76
97
93
80
71
83
100
92
92
82
88
99
62/27
62/27
73/27
73/27
71/29
Pd/C, [bmpyrr][NTf2]
Pd/Al2O3, [bmpyrr][NTf2]
Pd/Al2O3, [bmpyrr][NTf2]
Pd/Al2O3, [bmpyrr][NTf2]
Pd/C, [bmpyrr][NTf2]
Pd/C, [bmpyrr][NTf2]
Pd/C, [bmpyrr][NTf2]
Pd/C, [bmpyrr][NTf2]
Pd/C, Toluene
99
99
58c
83
75/8
75c
93c
93
68/25
Pd/C, [bmpyrr][NTf2]b
Pd/C, [bmpyrr][NTf2]b
Pd/C, [bmpyrr][NTf2]
Pd/C, [bmpyrr][NTf2]
33d
51d
83
54/29
58/41
99
a 1 ) Lilial; 2 ) 3-(4-tert-butylphenyl)-2-methylpropenal; 8 ) 4-tert-butylbenzaldehyde; 9 ) 2-methyl-2-pentanal; A ) 2,6-dimethylmorpholine; B ) piperidine;
C ) morpholine. b Reaction temperature was 90 °C, at 60 °C the conversions were ∼10% after 16 h. c Formation of Lilial. d Balance Lilial.
Scheme 2. Reaction products originating from the reductive amination of 3-(4-tert-butylphenyl)-2-methylpropenal and
2,6-dimethylmorpholine
and a precipitate was not observed. The aggregation is not
associated with the reductive amination reaction as palladium
black is formed when PdCl2 dissolved in [bmpyrr][NTf2] is
exposed to 10 atm H2 in the absence of either substrate.
As expected, for the reductive amination of Lilial (1) and
4-tert-butylbenzaldehyde with 2,6-dimethylmorpholine only
fenpropimorph (3) and 4-(4-tert-butylphenyl)-2,6-dimethyl-
morpholine, respectively, are formed. In the case of 2-methyl-
2-pentenal and 3-(4-tert-butylphenyl)-2-methylpropenal (2),
a number of products are possible, as illustrated in Scheme
2 for the case of 2. For both R,â-unsaturated aldehydes, high
selectivity to the saturated reductive amination products were
observed with <1% of any other product. This is important
in the case of formation of fenpropimorph from the aldol
condensation derived unsaturated intermediate, 2, as, whilst
it is possible to reduce 2 to Lilial highly selectively (>98%)
using 10% Pd/C in [bmpyrr][NTf2] in the absence of base
followed by reductive amination, this is not required, and a
one-step process can be used. In addition to the products
shown in Scheme 2, isomers containing the cis- and trans-
2,6-dimethylmorpholinium species are present. The starting
(32) Hamill, N. A.; Hardacre, C.; McMath, S. E. J. Green Chem. 2002, 139;
Deshmukh, R. R.; Rajagopal, R.; Srinivasan, K. V. Chem. Commun. 2001,
1544.
Vol. 10, No. 1, 2006 / Organic Process Research & Development
•
97

Table 2. Percentage product distribution and conversion of
3-(4-tert-butylphenyl)-2-methylpropenal after the reductive
amination with 2,6-dimethylmorpholine using 5% Pd/Al2O3
(250:1 aldehyde/Pd) in a range of solvents at 60 °C and 10
atm hydrogen for 5 h in the HEL reactor
% product distributiona
solvent
toluene
1
4
5
3
%conversion
0
23
0
5
1
6
7
0
0
0
1
0
7
5
2
4
3
94
61
95
93
95
82
100
98
100
100
100
98
water
fenpropimorph
[bmpyrr][NTf2]
[C4mim][NTf2]
[C4mim][BF4]
12
a 1 ) Lilial; 3 ) fenpropimorph; 4 ) 3-(4-tert-butylphenyl)-2-methyl-
propanol; 5 ) 3-(4-tert-butylphenyl)-2-methyl-propenol.
Table 3. Percentage conversion after the Heck couplinga of
4-tert-butyliodobenzene with 2-methyl-prop-2-en-1-ol and
percentage yield of fenpropimorph following the reductive
aminationa reaction with 2,6-dimethylmorpholine during the
one-pot sequential reaction (Path 1) using PdCl2 and 5%
Pd/C in a range of solvents
Figure 1. Percentage conversion to fenpropimorph as a
function of time during the reaction of 3-(4-tert-butylphenyl)-
2-methylpropenal (2) with 2,6-dimethylmorpholine using 5%
Pd/Al2O3 (aldehyde/Pd, 250:1) in a toluene (4), fenpropimorph
(2) water (0), [C4mim][NTf2] (b), [C4mim][BF4] (9), and
[bmpyrr][NTf2] (O) at 60 °C and 10 atm hydrogen in the HEL
reactor.
Heck %
% yield
catalyst
solvent
conversion fenpropimorph
material contains the cis and trans forms in a 7:3 ratio which
is maintained at 100% conversion; however, as shown in
Table 1, at intermediate conversions the ratio is decreased,
showing the higher rate of reaction for trans-2,6-dimethyl-
morpholine compared with that for cis-2,6-dimethylmorpho-
line. Interestingly, the rate of reaction for the reductive
amination is much higher compared with the hydrogenation
of the R,â-unsaturated aldehyde in the absence of amine.
This is due to the presence of the base and not due to the
mechanism of the reductive amination reaction as a similar
rate increase is also seen if a tertiary amine is substituted
for the secondary amine. In the presence of diisopropylethyl-
amine (Hunig’s base), 2 is reduced to Lilial as expected since
Hunig’s base is a tertiary amine and therefore cannot form
the reductive amination product. Although this may be
associated with the amine modifying the surface of the
catalyst, this is unlikely in the case of Hunig’s base which
is generally considered as uncoordinating due to the steric
bulk of the branched alkyl chains. Reaction rate changes have
been observed previously for hydrogenation reactions in the
presence of N-containing bases; however, both decreased as
well as increased rates have been reported.33
Although similar selectivities are observed in both ionic
liquids and molecular organic solvents, as found with other
hydrogenations using heterogeneous catalysts in ionic liquids,
a significant decrease in the activity of the catalyst was
observed compared with that in molecular solvents (Figure
1 and Table 2).10 This decrease may be understood by
comparing the viscosity of the ionic liquid and the hydrogen
gas solubility in ionic liquids compared with those in
molecular solvents. The increased viscosity of the ionic
liquids decreases the mass transport in the solvent which is
exacerbated in the presence of a heterogeneous catalyst where
the solvent/substrate must penetrate the pores of the catalyst.
Pd/C
Pd/C
Pd/C
Pd/C
PdCl2
[bmpyrr][NTf2]
[N(C4H9)4]Br
No Solvent
N-methylpyrrolidinone
[bmpyrr][NTf2]
99
80
40
99
99
65
51
14
72
81
a The Heck coupling was performed at 100 °C for 4 h, and the reductive
amination was performed at room temperature for 16 h under 10 atm hydrogen
in a Baskerville reactor.
In addition, hydrogen has a lower solubility in ionic liquids
compared with that in many organic solvents; this results in
a reduction of the effective pressure of the system and lowers
the reaction rate accordingly. It is interesting to note that in
homogeneously catalysed hydrogenations such an effect is
not observed, and similar reaction rates are found in
molecular and ionic solvents.34
One-Pot Synthesis to Fenpropimorph: Path 1. The
combined Heck reaction-reductive amination one-pot syn-
thesis, i.e. performing the Heck reaction in the presence of
amine and hydrogen atmosphere, was not possible in either
molecular solvents or ionic liquids due to the rapid hydro-
dehalogenation of 4-tert-butyliodobenzene to tert-butylben-
zene. Using 5% Pd/C, no reductive amination or Heck
coupling products obtained. Similarly, using PdCl2 dehalo-
genation also occurred, but due to the slower rate of this
reaction compared with the Heck process, 15% conversion
to Lilial was achieved; however, again no reductive amina-
tion products were observed.
Table 3 summarises the yield of fenpropimorph resulting
from the sequential one-pot synthesis, i.e. the Heck coupling
of 4-tert-butyliodobenzene with 2-methyl-prop-2-en-1-ol
followed by the reductive amination using 2,6-dimethylmor-
pholine in a range of solvents. For each reaction mixture,
(34) Dyson, P. J.; Laurenczy, G.; Ohlin, C. A.; Vallance, J.; Welton, T. Chem.
Commun. 2003, 2418.
(33) Mallat, T.; Baiker, A. Appl. Catal., A 2000, 200, 3.
98
•
Vol. 10, No. 1, 2006 / Organic Process Research & Development

Figure 2. Percentage composition as a function of time of the reductive amination with 2,6-dimethylmorpholine using 5% Pd/
Al2O3 (100:1 aldehyde/Pd) in [bmpyrr][NTf2] at 60 °C and 10 atm hydrogen in the Parr reactor following the aldol condensation
of 4-tert-butylbenzaldehyde and propanal during (a) the first one-pot sequential reaction (Path 1) and (b) on the second recycle. 1
) Lilial; 2 ) 3-(4-tert-butylphenyl)-2-methylpropenal,; 3 ) fenpropimorph; 5 ) 3-(4-tert-butylphenyl)-2-methyl-propenol; 7 ) 3-[4-
(2,6-dimethylmorpholino)]-2-methyl-1-(4-tert-butylphenyl)-1-propene; 8 ) 4-tert-butylbenzaldehyde; 10 ) 4-(4-tert-butylphenyl)-
2,6-dimethylmorpholine. The Michael adduct has been removed in each case for clarity.
the reductive amination was performed on the reaction
mixture from the Heck reaction without isolation of the Lilial
Heck product from the solvent system. With the exception
of the solventless media, good yields of fenpropimorph were
found for both ionic liquids and molecular solvent systems.
For the solventless system, the mixture became highly
viscous following the Heck reaction due to the formation of
triethylammonium iodide which reduced the mass transport
in the reductive amination.
h with an overall yield for the aldol and reductive amination
reactions of 75%. Understandably, the major by-product was
4-[(4-tert-butylphenyl)]-2,6-dimethylmorpholine from the
reductive amination of unreacted 4-tert-butylbenzaldehyde.
However, unlike in the reductive amination of the isolated
2, the one-pot reaction also formed 3-(4-tert-butylphenyl)-
2-methyl-propenol (5) and 3-[4-(2,6-dimethylmorpholino)]-
2-methyl-1-(4-tert-butylphenyl)-1-propene (7), albeit with
yields of only 6% and 5%, respectively.
Although it was possible to perform the one-pot synthesis
using Path 1, this system was not recyclable. As described
above, the reductive amination using the homogeneous
catalysts formed palladium black which was unreactive for
the Heck reaction on subsequent reactions. This problem
should be alleviated using the heterogeneous catalyst;
however, during the Heck reaction the palladium is leached
from the support to form colloidal palladium in solution
which then precipitates from solution during the reductive
amination.16
One-Pot Synthesis to Fenpropimorph: Path 2. In Path
2, the propanal is added slowly during the aldol condensation
in the presence of 60 mol % 2,6-dimethylmorpholine as the
base. Following the aldol condensation, a further 40 mol %
of 2,6-dimethylmorpholine is added for the reductive ami-
nation; i.e. the catalyst in the first reaction is used as one of
the reactants in the second stage. Figure 2a shows the result
of the reductive amination following the aldol condensation
of 4-tert-butylbenzaldehyde with propanal using 2,6-di-
methylmorpholine as the base in the presence of 5% Pd/
Al2O3. The aldol condensation was stopped after the addition
of 100 mol % propanal so as to minimise the formation of
the self-aldol condensation product of propanal, 2-methyl-
2-pentenal. As found for the sequential reaction from Path
1, high conversion to fenpropimorph was observed after 20
The selectivity and rate of the aldol condensation reaction
was not substantially affected by the presence of the
heterogeneous catalyst. However, in addition to the formation
of the cross- and self-aldol products, small amounts (∼6%)
of the Michael addition product 3-(4-tert-butylphenyl)-3-(2,6-
dimethylmorpholin-4-yl)-2-methyl-propanal (6) was found
(Scheme 2); this was not observed in the absence of the
catalyst using a small-scale reaction.26 This may be associated
with the more efficient stirring in the larger-scale experiments
compared with the magnetically stirred reactions reported
previously.26 Aza-Michael additions, i.e. Michael additions
that are carried out with amine nucleophiles, are known to
be promoted by water.35 In the ionic liquid it has been
demonstrated that a solid is formed when the 2,6-dimethyl
morpholine is mixed with 4-tert-butylbenzaldehyde; this
precipitate is thought to be the iminium salt. On the small
scale, the reaction medium is poorly mixed, and therefore,
the water formed during the aldol reaction cannot effectively
promote the Michael addition reaction. In contrast, using an
overhead mechanical stirrer enables the solid to be broken
up more efficiently, releasing water as the reaction progresses,
thereby magnifying the effect of water. No Michael addition
(35) Naidu, B. N.; Li, W. Y.; Sorenson, M. E.; Connolly, T. P.; Wichtowski, J.
A.; Zhang, Y. H.; Kim, O. K.; Matiskella, J. D.; Lam, K. S.; Bronson, J. J.;
Ueda, Y. Tetrahedron Lett. 2004, 45, 1059.
Vol. 10, No. 1, 2006 / Organic Process Research & Development
•
99

Figure 3. Variation of the percentage conversion to fenpropi-
morph as a function of recycle following the reaction of 3-(4-
tert-butylphenyl)-2-methylpropenal (2) with 2,6-dimethylmor-
pholine using 5% Pd/Al2O3 (250:1 aldehyde/Pd) in [bmpyrr]-
[NTf2] at 60 °C and 10 atm hydrogen in the HEL reactor where
the aldehyde used is formed ex situ (0) and in situ (9), i.e. the
reduction is performed as part of the sequential one-pot process
in Path 2.
Figure 4. Images of the reaction mixture following the one-
pot sequential reaction (Path 2) using 5% Pd/Al2O3 in [bmpyrr]-
[NTf2] in the Parr reactor (a) directly following reaction and
(b) after catalyst removal.
though a drop in the catalyst BET surface area was observed
following the reaction in the present case, on washing the
catalyst with acetonitrile, the original surface area was
recovered. Furthermore, after washing the catalyst followed
by drying, redispersion in fresh ionic liquid resulted in >95%
of the activity found on the next reaction.
The reduction in rate on recycle cannot be explained by
palladium leaching during either the aldol or reductive
amination reactions. The average dissolved palladium content
was found to be less than 3 ppm by ICP from either
palladium supported on carbon or alumina in either the
fenpropimorph layer or in the ionic liquid after separation
of the catalyst.
Due to the method workup, i.e. simple decantation of the
fenpropimorph layer from the ionic liquid, the solubility of
the product in the [bmpyrr][NTf2] was measured and found
to be ∼0.144 M. The low level of solubility of fenpropi-
morph in the ionic liquid allows for simple product recovery
through layer separation. The presence of low concentrations
of the product in the ionic liquid may be the source of the
change in rate and selectivity on recycle; however, the
reductive amination of 2 performed in fenpropimorph as the
solvent resulted in similar activity as found in the ionic liquid
(Figure 1). Similarly, the other hydrocarbon by-products, in
particular 4-(4-tert-butylphenyl)-2,6-dimethylmorpholine which
is the least efficiently extracted by-product of fenpropimorph,
are not thought to be responsible for the changes observed
on recycle. It should be noted that polymeric material may
also be formed in the aldol condensation from multiple
additions and may be adsorbed strongly on the catalyst or
be retained in the ionic liquid. However, reuse of either the
catalyst with fresh ionic liquid or the ionic liquid with fresh
catalyst showed no change in the rate of the reductive
amination of isolated 2 compared with the reaction performed
with unused catalyst and ionic liquid.
product is observed after the reductive amination as this
reaction is reversible, and therefore, the Michael adducts
break down as the reductive amination proceeds and leading
to the formation of the desired product only. Although aza-
Michael addition reactions have also been shown to be
promoted by palladium and other transition metals,36 it is
not thought that the presence of the catalyst contributes to
the reaction as similar yields are observed in the absence of
catalyst using an overhead stirrer.
Figures 2b and 3 show the effect of recycling the
catalyst-ionic liquid system in the Parr and HEL reactors,
respectively. Following the initial aldol-reductive amination
reactions, the fenpropimorph formed was decanted from the
ionic liquid solvent. This was subsequently recharged with
4-tert-butylbenzaldehyde and the aldol condensation repeated
followed by the reductive amination. Clearly, a significant
reduction in the reaction rate and an increase in the number
of by-products is observed, as shown for the second recycle
data in the Parr reactor in Figure 2b. This does not reflect a
significant change in the selectivity/conversion achieved in
the aldol condensation as, with the exception of a small
increase in the Michael addition product, no changes in the
conversion versus selectivity profile compared with the initial
aldol reaction were found on recycle.
A reduction in rate is also observed on recycling the
catalyst-ionic liquid system for the reductive amination
reaction of isolated 2 using 2,6-dimethylmorpholine after 6
h (Figure 3). With an increasing number of reactions, the
conversion is found to drop gradually. Although a similar
deactivation was reported by Anderson et al. for the selective
hydrogenation of R,â-unsaturated aldehydes, in that case the
decrease was a step change, and following the first reaction
the rate remained constant for subsequent reactions.10 In this
case, the sharp decrease was attributed to pore blocking of
the catalyst, and the activity could not be recovered by
washing the catalyst or catalyst-ionic liquid system. Al-
Following separation of the catalyst from the reaction
mixture, unexpectedly three phases of liquid were observed,
shown in Figure 4. Using a hydrophobic ionic liquid, the
water formed in both the aldol condensation and the reductive
amination reactions separates from the ionic liquid and
(36) Kawatsura, M.; Hartwig, J. F. Organometallics 2001, 20, 1960.
100
•
Vol. 10, No. 1, 2006 / Organic Process Research & Development

Figure 5. Variation of the percentage conversion to fenpropi-
morph following the reaction of 3-(4-tert-butylphenyl)-2-meth-
ylpropenal with 2,6-dimethylmorpholine using 5% Pd/Al2O3
(250:1 aldehyde/Pd) in [bmpyrr][NTf2] at 60 °C and 10 atm
hydrogen in the HEL reactor with (9) and without 3 mol equiv
(O) of water with respect to the 3-(4-tert-butyl-phenyl)-2-
methylpropenal (2).
Figure 6. Schematic process diagram for the scale-up of Path
2. Aldehyde and amine are charged into the reactor containing
ionic liquid and catalyst while the second aldehyde is pumped
in slowly with stirring until complete conversion. Further amine
is introduced, hydrogen pressure is applied with heating and
stirring to completion. The reaction mixture is then allowed to
settle, the product layer is pumped off for filtration and
distillation. The process steps may then be repeated.
product phases. Commonly, the solvent makes up the
majority of the volume in the reaction, and thus, even if a
hydrophobic solvent is used, any water remains in the
solvent. In the present case, the ionic liquid is only required
to ensure that the aldol condensation has high selectivity and
to support the catalyst for the reductive amination; in both
cases, only a minimum volume is required. In these reactions,
the volume ratio between the water formed if 100% reaction
and the ionic liquid used is 1:2 (H2O/IL). Figure 5 shows
the effect of water on the reductive amination of unsaturated
Lilial using 5% Pd/Al2O3 in [bmpyrr][NTf2]. Using 3 mol
equiv of water to the amount of substrate, i.e. the quantity
of water formed during two aldol condensations plus a
reductive amination reaction, to represent the effect of a
recycle reaction, a 44% reduction in the initial rate of reaction
is observed. Similarly when water is used as the solvent, a
significant reduction in the rate is found compared with use
of the ionic liquids and toluene. Interestingly, in water, the
yield of 3-(4-tert-butylphenyl)-2-methyl-propenol is sub-
stantially greater than in ionic liquids or toluene, even when
the [bmpyrr][NTf2] was spiked with water, as shown in Table
2 and Figure 1. The fact that in the one-pot reaction so little
3-(4-tert-butyl-phenyl)-2-methyl-propenol is formed indicates
that the reaction still proceeds in the ionic liquid phase. The
reduction in reaction rate found in the hydrophobic ionic
liquid in the presence of water may, therefore, be due to the
water extracting the amine from the ionic liquid phase and
consequently reducing its concentration resulting in a
decrease in the overall reaction rate. This possibility has been
verified by measuring the partition coefficient between water
and [bmpyrr][NTf2] for 2,6-dimethylmorpholine, which has
a value for [amine]IL/[amine]H2O of 0.31. In addition, when
the catalyst/ionic liquid layer was dried between recycles,
conversions are significantly improved when compared with
the nondried system. For example, after three reactions in
the HEL reactor, the conversion for the dried system was
>10% higher than that found for the nondried system.
A combined aldol reaction-reductive amination one-pot
synthesis was also performed, i.e. pumping the propanal into
the reactor under 2.5 atm of hydrogen at room temperature
in the presence of 100 mol % 2,6-dimethylmorpholine; after
the addition the vessel was pressurised to 10 atm. In this
case, the reaction mixture became highly viscous; however,
a yield of 36% fenpropimoph was found after solvent
extraction using cyclohexane. In addition to fenpropimorph
a large number of by-products were also observed. As with
the combined reaction in Path 1, although the single process
step reaction is possible, there are significant disadvantages
of performing the combined reaction over the sequential one-
pot process.
Comparison of Process Paths. It is clear that although
both pathways are viable, Path 2 offers a more atom-efficient
process as well as the possibility of recycling the catalyst/
ionic liquid system. The major disadvantages of Path 2 are
the increased number of by-products and a small decrease
in overall yield of fenpropimorph compared with those of
Path 1. The atom efficiency for Path 1 was calculated to be
46% whilst that from Path 2 was 67%. In addition, the
E-factors were found to be 1.26 and 0.37 for Paths 1 and 2,
respectively.37 The high efficiency of Path 2 is due, in part,
to the use of one of the reagents as a catalyst in a previous
step but also that the main by-product is water. Path 1 is
inefficient because of its triethylammonium iodide salt by-
product formation. Furthermore, the reaction profile and rates
could be correlated on scales ranging from 10 to 150 cm3 in
Path 2. Figure 6 shows the proposed process diagram for
the scaled-up overall process where after reaction the product
is isolated by distillation and the ionic liquid/catalyst system
is dried in situ and recycled.
(37) The E-factor ) total waste (kg)/product (kg). The E-factors were calculated
without taking into account the water as waste and assuming 100% recovery
of the ionic liquid and catalyst. Sheldon, R. A. J. Mol. Catal. A: Chem.
1996, 107, 75.
Vol. 10, No. 1, 2006 / Organic Process Research & Development
•
101

Conclusions
solvents, the separation is more difficult and requires
evaporation and recondensation of the solvent.
Two routes to fenpropimorph have been demonstrated via
a Heck reaction or an aldol condensation and then a reductive
amination reaction. Although both Heck and aldol processes
proceed efficiently in the ionic liquid, only the aldol pathway
may be recycled. By using the amine catalyst from the aldol
condensation as the reagent in the subsequent reductive
amination reaction, a highly atom-efficient process is pos-
sible. A versatile multistep catalytic synthesis using a single
ionic liquid is possible with the ionic liquid able to be easily
recycled, albeit with a drop in conversion and a decrease in
the selectivity. In contrast, using conventional molecular
Acknowledgment
We thank the EPSRC manufacturing molecule initiative
(MMI) for funding this project under Grant GR/S19219 and
DELNI (AMcK) and Johnson Matthey for the loan of the
catalysts used in this study.
Received for review September 16, 2005.
OP050172M
102
•
Vol. 10, No. 1, 2006 / Organic Process Research & Development