Selective hydroformylation of N-allylacetamide in an inverted aqueous two-phase catalytic system, enabling a short synthesis of melatonin

Article abstract of DOI:10.1039/b003715j

Water increases the selectivity in the Rh-phosphine catalysed hydroformylation of N-allylacetamide; an aqueous-organic biphasic system, containing a hydrophobic Rh-catalyst, provided facile catalyst/product separation, after which the aqueous product phase could be used in a one-pot synthesis of N-acetyl-5-methoxytryptamine (melatonin).

Full text of DOI:10.1039/b003715j

Selective hydroformylation of N-allylacetamide in an inverted aqueous
two-phase catalytic system, enabling a short synthesis of melatonin†‡
Göran Verspui, Guido Elbertse, Frank A. Sheldon, Michiel A. P. J. Hacking and Roger A. Sheldon*
Delft University of Technology, Laboratory for Organic Chemistry and Catalysis, Julianalaan 136, NL-2628 BL
Delft, The Netherlands. E-mail: r.a.sheldon@tnw.tudelft.nl
Received (in Liverpool, UK) 5th May 2000, Accepted 13th June 2000
Published on the Web 4th July 2000
Water increases the selectivity in the Rh–phosphine cata-
lysed hydroformylation of N-allylacetamide; an aqueous–
organic biphasic system, containing a hydrophobic Rh-
catalyst, provided facile catalyst/product separation, after
which the aqueous product phase could be used in a one-pot
synthesis of N-acetyl-5-methoxytryptamine (melatonin).
Olefin hydroformylation is one of the most commercially
important reactions that makes use of a homogeneous catalyst.1
Rh–phosphine complexes are active catalysts for this reaction
with high selectivities towards linear aldehydes and a number of
Scheme 1 Alternative synthesis of N-acetyl-5-methoxytryptamine.
successful methods for the recycling of these catalysts have
been developed.2 Unfortunately, the feedstock is mainly
restricted to linear, non-functionalised -olefins, such as
such as methanol. Since N-allylacetamide is also soluble in
propene, styrene, etc.
water, we decided to carry out the hydroformylation in a one-
Aldehydes bearing various functional groups are widely
phase aqueous medium, employing the water-soluble Rh–tppts
applied in the synthesis of, for example, flavours, fragrances
catalyst (tppts = P(C6H4-m-SO3Na)3). To our surprise we
and pharmaceuticals.3 Their preparation by hydroformylation
found that the reaction proceeded smoothly under mild
of a hetero-atom functionalised olefin, however, is not straight-
conditions. At 70 °C and a 10 bar pressure of a H2–CO (1 1)
forward since these substrates react slowly, require high catalyst
mixture, the reaction was completed in 45 min using as little as
loadings and harsh reaction conditions, which leads to the
0.04% catalyst (Table 1, Exp. 2), with an initial rate of 3891
formation of condensation products, such as acetals, hemiace-
turnovers h 1. The isomeric aldehydes were obtained in 99%
tals, imines, etc.4
selectivity. It was demonstrated by Mortreux and co-workers
In connection with an alternative synthesis of N-acetyl-
previously that the hydroformylation of acrylesters, such as
5-methoxytryptamine (melatonin), a human hormone which
methylacrylate, into their - and -formylesters also proceeds at
regulates sleep,5 we required 4-acetamidobutanal. When com-
a faster rate in a toluene–water two-phase system, employing
bined with 4-methoxyphenylhydrazine this aldehyde affords an
the Rh–tppts catalyst, compared to the rate of Rh–PPh3 in
intermediate hydrazone, which in the presence of an acid
1
toluene alone (initial TOF = 545 vs. 225 h (50 °C, 50
undergoes a Fischer-indole reaction (Scheme 1).6 4-Acet-
bar)).8
amidobutanal was previously prepared by hydroformylation of
In a series of MeOH–H2O mixtures we found that both the
N-allylacetamide in THF using 1% of HRh(CO)(PPh3)3.7 After
reaction rate and the selectivity towards the aldehydes increased
18 h at 80 °C and 83 bar (CO H2 = 1 1), a conversion of 76%
with the water content. A direct comparison of the two catalysts
was reached with the product distribution: 4-acetamidobutanal
in neat MeOH showed that Rh–PPh3 is slightly more active than
(11%), 3-acetamido-2-methylpropanal (63%), N-acetylpyrroli-
Rh–tppts and yields a product mixture with a slightly higher
dine (13%) and 2-formyl-N-acetylpyrrolidine (13%).
linear/branched (l/b) ratio (1.15 vs. 1.05, Table 1). The rather
We found that the Rh–PPh3 catalysed hydroformylation of N-
low regioselectivity compared to non-functionalised linear
allylacetamide gave the best results in a polar, protic solvent,
olefins, such as propene (l/b = 23),2 suggests that coordination
of the amide may take place during the insertion of the olefin
† Catalytic conversions in water, part 18. (Part 17: G. J. ten Brink,
into the Rh–hydride bond. The possible formation of a
6-membered chelate ring (Scheme 2) is expected to enhance the
formation of the branched aldehyde.
I. W. C. E. Arends and R. A. Sheldon, Science, 2000, 287, 1636.)
‡ Experimental details are available as electronic supplementary informa-
tion (ESI). See http://www.rsc.org/suppdata/cc/b0/b003715j/
Table 1 The hydroformylation of N-allylacetamide in organic, aqueous and biphasic media
1
Exp.a
Catalyst
Solvent
Time/min
Conv./%
Yield/%
1/b ratio
TOFf/h
1
2
3
4
Rh–PPh3
Rh–tppts
Rh–tppts
Rh–tppts
MeOH
H2O
MeOH–H2Od
MeOH
180
45
120
180
98.6
98.9
99.9
99.9
92.3
97.9
95.3
93.4
1.15
1.30
1.09
1.05
1342
3891
1459
1239
5
Rh–tppts
Rh–PPh3
Rh–Xantphos
Rh–Xantphos
Toluene–H2Oe
Toluene–H2Oe
Toluene–H2Oe
Toluene–H2Oe
60
270
1320
600
97.4
81.6
26.4
99.6
96.8
80.2
24.4
96.4
1.34
1.54
20.0
15.3
2946
578
31
6
7b
8c
177
a Reaction conditions: 0.010 mmol Rh(acac)(CO)2, 0.25 mmol ligand, 25 mmol N-allylacetamide, 70 °C, 10 bar H2O–CO (1/1), 1000 rpm, 125 ml total
reaction volume. b 0.010 mmol [Rh(acac)(CO)2], 0.050 mmol Xantphos. c 0.050 mmol [Rh(acac)(CO)2], 0.250 mmol Xantphos, 90 °C. d MeOH–H2O = 1/1
(v/v), 125 ml total. e 100 ml toluene, 100 ml H2O. f TOF = initial turnover frequency in mol aldehyde per mol catalyst per hour.
DOI: 10.1039/b003715j
Chem. Commun., 2000, 1363–1364
This journal is © The Royal Society of Chemistry 2000
1363

catalyst phase could be recycled in 5 consecutive runs without
loss in activity. Rh-analysis of the water layers by means of
atomic absorption spectrometry confirmed a quantitative
( > 99%) catalyst/product separation.
N-Acetyl-5-methoxytryptamine was subsequently prepared
in a one-pot procedure starting from allylamine (see ESI). By
successive acetylation of allylamine, selective hydroformyla-
tion, hydrazone formation with 4-methoxyphenylhydrazine
(HCl salt) and a Fischer indole reaction, melatonin was isolated
from the aqueous reaction mixture in 44% yield (not opti-
mised).
In conclusion, we have demonstrated that the hydro-
formylation of N-allylacetamide proceeds smoothly in a one-
phase aqueous medium, as well as in an aqueous–organic two-
phase protocol under mild reaction conditions. Although we
could not take full advantage of the rate accelerating effect of
water in the inverted two-phase catalyst system, the linear
aldehyde was obtained in high selectivity and efficient catalyst/
product separation was achieved. The organic soluble catalyst
was recycled and the aqueous product phase was applied
without purification in the synthesis of N-acetyl-5-methoxy-
tryptamine. Mechanistic details and the scope of the inverted
two-phase system are currently under investigation.
We thank Dr G. Papadogianakis and Dr G. Besenyei for the
fruitful discussions and Mr J. Padmos for the Rh-analysis.
Financial support by NWO-CW is also gratefully acknowl-
edged.
Scheme 2 Proposed intermediates in the Rh/tppts catalysed hydro-
formylation of N-allylacetamide.
The high partition coefficients of 4-acetamidobutanal and
3-acetamido-2-methylpropanal in the water layers forestalled
our attempts to extract them from the aqueous reaction mixtures
with an organic solvent. Especially in the synthesis of
pharmaceuticals the presence of traces of heavy metals is
undesirable. In order to achieve efficient product/catalyst
separation, we turned to an ‘inverse two-phase catalyst system’,
containing the hydrophobic Rh–PPh3 catalyst in a toluene–
water mixture. In such a system the catalyst remains dissolved
in the organic phase, while the products will move to the water
layer; the opposite of standard aqueous biphasic catalysis.
In the biphasic system, the Rh–PPh3 catalysed reaction (Exp.
6) proceeded considerably slower compared to Rh–tppts (Exp.
5), consistent with our observation (see above) that the reaction
rate decreases when the hydroformylation is carried out in an
apolar solvent, such as toluene. In addition, due to mass transfer
limitations, the reaction rate decreased substantially after ca.
50% conversion (see Fig. 1). Such a decrease was less
pronounced in the case of Rh–tppts, that operates in the aqueous
layer, where the substrate concentration remains sufficiently
high, resulting in a zero-order reaction profile until ca. 80%
conversion. Nevertheless, due to the presence of water, the
selectivity towards the aldehydes remained high (98%, Exp. 6)
and the Rh–PPh3 catalyst could conveniently be separated from
the aqueous product phase.
Notes and references
1 K. Weissermel and H. J. Arpe, Industrial Organic Chemistry, 3rd edn.
VCH, Weinheim, 1997; Applied Homogeneous Catalysis with Organo-
metallic Compounds, ed. B. Cornils and W. A. Herrmann, vol. 1, VCH,
Weinheim, 1996.
2 E. G. Kuntz, CHEMTECH, 1987, 570; Y. Chauvin, L. Mussmann and
H. Olivier, Angew. Chem., Int. Ed. Eng., 1995, 34, 2698; I. T. Horváth,
G. Kiss, R. A. Cook, J. E. Bond, P. A. Stevens, J. Rabai and E. J.
Mozeleski, J. Am. Chem. Soc., 998, 120, 3133; E. Lindner, T. Schneller,
F. Auer and H. A. Mayer, Angew. Chem., Int. Ed. Eng., 1999, 38,
2155.
3 H. H. Szmant, Organic building blocks of the chemical industry, Wiley,
New York, 1989; Ullmann’s Encyclopedia of Ind. Chem., Vol. A11,
VCH, Weinheim, Germany, p. 141.
4 C. Botteghi, R. Ganzerla, M. Lenarda and G. Moretti, J. Mol. Catal.,
1987, 40, 129; P. Eilbracht, L. Bärfacker, C. Buss, C. Hollmann, B. E.
Kitsos-Rzychon, C. L. Kranemann, T. Rische, R. Roggenbuck and A.
Schmidt, Chem. Rev., 1999, 99, 3329.
5 H. M. Hugel and D. J. Kennaway, Org. Prep. Proced. Int., 1995, 27,
1.
6 W. Marais and C. W. Holzapfel, Synth. Commun., 1998, 28, 3681.
7 I. Ojima and Z. Zhang, J. Org Chem., 1988, 53, 4422; I. Ojima and Z.
Zhang, J. Organomet. Chem., 1991, 417, 253; I. Ojima, A. Korda and
W. R. Shay, J. Org. Chem., 1991, 56, 2024; I. Ojima, Z. Zhang, A.
Korda, P. Ingallina and N. Clos, Adv. Chem., 1991, 56, 278; see also: S.
Sato, M. Takesada and H. Wakamatsu, Nippon Kagaku Zasshi, 1969,
90, 579; Y. Becker, A. Eisenstadt and J. K. Stille, J. Org. Chem., 1980,
45, 2145.
Fig. 1 Reaction profiles of Exp. 5 and Exp. 6.
To increase the regioselectivity towards the linear aldehyde,
generally, rigid diphosphine ligands with a large bite angle are
used.9 We tested the Rh–Xantphos combination10 (Xantphos =
4,5-bis(diphenylphosphino)-9,9-dimethylxanthene) in our in-
verted two-phase system and, indeed, the l/b ratio increased to
20. The reaction rate, however, decreased dramatically to 31
8 G. Fremy, E. Monflier, J. F. Carpentier, Y. Castanet and A. Mortreux,
Angew. Chem., Int. Ed. Engl., 1995, 34, 1474.
9 For the hydroformylation of functionalised olefins in THF: G. D. Cuny
and S. L. Buchwald, J. Am. Chem. Soc., 1993, 115, 2066.
10 M. Kranenburg, Y. E. M. van der Burgt, P. C. J. Kamer, P. W. N. M. van
Leeuwen, K. Goubitz and J. Fraanje, Organometallics, 1995, 14,
3081.
h
1 (Table 1, Exp. 7). At 90 °C, in the presence of 0.2% catalyst,
a nearly quantitative conversion was reached in 10 h reaction
time with a small compromise in the l/b ratio. The organic
1364
Chem. Commun., 2000, 1363–1364