(5R)-5-Alkyl-5,6-dihydroindolizines via stereospecific domino hydroformylation/cyclodehydration of (3R)-3-(pyrrol-1-yl)alk-1-enes

Article abstract of DOI:10.1016/j.tetasy.2004.04.045

(5R)-5-Alkyl-5,6-dihydroindolizines 3a-c having the same high enantiomeric excess (>92%) as the corresponding starting olefins (3R)-3-(pyrrol-1-yl)alk- 1-enes 1a-c were obtained via a highly regioselective and stereospecific domino hydroformylation/cyclodehydration reaction sequence. The reasons for this configurational stability were also analyzed in the light of the general accepted rhodium catalyzed hydroformylation mechanism.

Full text of DOI:10.1016/j.tetasy.2004.04.045

Tetrahedron:
Asymmetry
Tetrahedron: Asymmetry 15 (2004) 1821–1823
(5R)-5-Alkyl-5,6-dihydroindolizines via stereospecific
domino hydroformylation/cyclodehydration of
(3R)-3-(pyrrol-1-yl)alk-1-enes
Roberta Settambolo,a Giuditta Guazzelli,b Alessandro Mandolib
and Raffaello Lazzaronib,*
aICCOM-CNR, Sezione di Pisa, Dipartimento di Chimica e Chimica Industriale, Via Risorgimento 35, 56126 Pisa, Italy
bDipartimento di Chimica e Chimica Industriale, Universitꢀa di Pisa, Via Risorgimento 35, 56126 Pisa, Italy
Received 24 March 2004; accepted 13 April 2004
Abstract—(5R)-5-Alkyl-5,6-dihydroindolizines 3a–c having the same high enantiomeric excess (>92%) as the corresponding starting
olefins (3R)-3-(pyrrol-1-yl)alk-1-enes 1a–c were obtained via a highly regioselective and stereospecific domino hydroformylation/
cyclodehydration reaction sequence. The reasons for this configurational stability were also analyzed in the light of the general
accepted rhodium catalyzed hydroformylation mechanism.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1. Introduction
amounts, did not undergo the same reaction, remaining
unchanged at the end of the reaction.
The construction of pyrrole fused alkaloids of the ‘izine’
or ‘izidine’ types in a stereoselective manner is a topic of
current interest due to the wide ranging biological
activity associated with these systems. Among the many
synthetic approaches, the oxo process is still poorly
investigated. Nevertheless the hydroformylation of ole-
fins presents many attractive features: the reaction needs
only catalytic amounts of a metal catalyst and all atoms
of the starting materials remain incorporated into the
product in a very economic C–C-bond forming reaction.
When the oxo process is a part of a domino reaction
sequence and involves optically active species, its
potential is greatly enhanced.1 A critical aspect could
still be the control of the regio and enantioselectivity of
the reaction depending on unsymmetrically substituted
and/or configurationally unstable olefins as starting
material. Recently we set up a new synthetic application
of the oxo process giving 5-alkyl-5,6-dihydroindolizines
from 3-(pyrrol-1-yl)alk-1-enes via in situ cyclodehydra-
tion of the thus formed linear 4-pyrrolylbutanals.2 On
the contrary the branched aldehyde, present in minor
2. Results and discussion
Encouraged by the above successful application and
taking into account that optically active 5-alkyl-5,6-di-
hydroindolizines could be precursors of natural indo-
lizidines,3 we planned to hydroformylate the highly
enantiomerically enriched (3R)-3-(pyrrol-1-yl)alk-1-enes
1a–c, recently synthesized by us:4 the (5R)-5-alkylindo-
lizidines 3a–c having the same enantiomeric excess
(>92%) as the starting olefins were obtained via a highly
regioselective and stereospecific domino transformation
(Scheme 1, Table 1).
The (3R)-3-(pyrrol-1yl)alk-1-enes 1a–c (ee >92%) were
D
prepared via a multistep reaction sequence using
-a-
aminoacids as the source of chirality.4 The hydro-
formylation of the substrates 1a–c was carried out in the
presence of Rh4(CO)12 as the catalyst precursor
according to a typical procedure.5 Under these condi-
tions the formed linear aldehydes 2 undergo an in situ
intramolecular electrophilic substitution on position
two of the pyrrole nucleus giving 5,6-dihydroindolizines
3 via formation of a bicyclic alcohol followed by
* Corresponding author. Tel.: +39-050-2219227; fax: +39-050-22192-
60; e-mail: lazza@dcci.unipi.it
0957-4166/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2004.04.045

1822
R. Settambolo et al. / Tetrahedron: Asymmetry 15 (2004) 1821–1823
O
-H2O
OH
N
N
N
R
R
R
3
2
CO/H2 (1:1)
N
+
Rh4CO12,toluene
R
1
N
R
O
4
Scheme 1. Hydroformylation of (3R)-3-(pyrrol-1-yl)alk-1-enes 1a–c in the presence of Rh4(CO)12 at 100 ꢁC and 125 ꢁC.
Table 1. Hydroformylation of (3R)-3-(pyrrol-1-yl)alk-1-enes 1a–c with Rh4(CO)12 at 100 ꢁC
1
T (ꢁC)
P (atm)
Reaction
time (h)
Conversion
(%)
Products
Entry
Ee (%)a
3/4 (%)
3 Yield (%)b
3 Ee (%)a
a
a
a
b
c
98
98
98
92
92
100
100
125
125
125
100
100
30
0.2
1.5
0.5
0.5
0.5
25
99
97
99
99
57/43
59/41
85/15
87/13
84/16
––
55
73
70
75
98
98
98
92
92
30
30
a: R ¼ Me; b: R ¼ iso-Pr; c: R ¼ n-Pr.
a Determined by gas chromatography with the chiral capillary column CHIRALDEX G-TA (c-cyclodextrin trifluoroacetyl, 50 m · 0.25 mm).
b Isolated pure product.
dehydration (Scheme 1). Compound 1a was submitted
under the same experimental hydroformylation condi-
tions adopted for the corresponding racemic substrate,2c
showing a very similar behaviour. After 0.2 h the con-
version was 25% and the reaction mixture comprised
5-methyl-5,6-dihydroindolizine 3a and branched alde-
hyde 4a in a 57/43 molar ratio. This value remained
unchanged at total conversion (after 1.5 h, 59/41 regio-
isomeric ratio) (Table 1). The linear aldehyde 2a, the
precursor of 3a, was present only in trace amounts in the
reaction mixture both at partial and complete substrate
conversion, the cyclization reaction being faster than
hydroformylation. An evaluation of the enantiomeric
excess of both unconverted 1a and produced 3a was
carried out in order to test the configurational stability
of these structures under hydroformylation conditions.
Interestingly 1a showed, at all conversions, practically
the same ee, that is, the starting ee value (98%). A simi-
lar behaviour occurred for the dihydroindolizine 3a, its
ee value remaining the same as the corresponding olefin
1a (98%) at all reaction times (Table 1). Encouraged by
this result, we turned our attention to setting up hydro-
formylation conditions able to increase the formation of
optically active 5-methyl-5,6-dihydroindolizine, that is,
directing the regioselectivity towards the linear alde-
hyde. According to the well documented behaviour of
vinyl and allyl substrates under rhodium-catalyzed
tions. In particular the olefin 1b gave 5-i-propyl-5,6-di-
hydroindolizine 3b in 3b/4b7 ¼ 87/13 regioisomeric ratio.
In a similar manner 5-n-propyl-5,6-dihydroindolizine 3c
was obtained from 1c (3c/4c7 ¼ 84/16). No traces of the
linear aldehydes 2b and c, the precursors to 3b and c,
were observed in the reaction mixture both at partial
and complete substrate conversion. The dihydroindoli-
zines 3b and c8 are new compounds and were isolated
1
and characterized by H and 13C NMR spectroscopy,
GC–MS and specific rotation. The racemic dihydro-
indolizine 3a is known9 while its optically active form
has been prepared here for the first time. All the
obtained dihydroindolizines are stable enough to be
handled easily at room temperature without any
decomposition or change of enantiomeric excess and can
be stored at 0 ꢁC for long periods. The ee values of the
isolated 5,6-dihydroindolizines were determined by gas
chromatography with a chiral capillary column. The
chromatographic conditions were tested on the analo-
gous racemic substrates obtained via the same oxo
process starting from racemic olefins. In particular 1b
and c were prepared via the multistep synthetic pathway
adopted for the corresponding optically active sub-
strates,4 starting from the appropriate racemic amino
acids. Compound 1a was easily prepared via N-alkyl-
ation of pyrrole with 3-chloro-1-butene under basic
conditions.2c
hydroformylation,2a;6
a marked improvement was
achieved by carrying out the hydroformylation of 1a at
125 ꢁC and at lower pressure (30 atm; CO/H2 ¼ 1:1): the
molar ratio 3a/4a conveniently increased to 85/15. It is
to note that these forcing conditions do not affect the
enantiomeric excess of 3a (ee 98%). The other olefins 1b–
c were submitted to the same hydroformylation condi-
The very high ee values obtained for 3 indicate that the
hydroformylation conditions are perfectly compatible
with the optically active pyrrolylolefins employed
allowing a complete configurational stability also under
potentially isomerizing conditions (high temperature
and low pressure). Taking into account the general

R. Settambolo et al. / Tetrahedron: Asymmetry 15 (2004) 1821–1823
1823
2. (a) Lazzaroni, R.; Settambolo, R.; Caiazzo, A.; Pontorno,
L. J. Organomet. Chem. 2000, 601, 320; (b) Settambolo,
R.; Savi, S.; Caiazzo, A.; Lazzaroni, R. J. Organomet.
Chem. 2001, 619, 241; (c) Settambolo, R.; Caiazzo, A.;
Lazzaroni, R. Tetrahedron Lett. 2001, 42, 4045; (d)
Settambolo, R.; Miniati, S.; Lazzaroni, R. Synth. Com-
mun. 2003, 17, 2953.
3. (a) Amat, M.; Llor, N.; Hidalgo, J.; Escolano, C.; Bosch,
J. J. Org. Chem. 2003, 68, 1919, and references cited
therein; (b) Daly, J. W. Prog. Chem. Nat. Prod. 1982, 41,
205.
N
N
Rh
N
N
2
Rh
R
R
R
R
5
1
B
4
Scheme 2.
accepted mechanism of hydroformylation,10 we can
affirm that, under the above conditions, the branched
alkyl-rhodium intermediate B undergoes a b-hydride
elimination process2c not involving the stereogenic cen-
tre, generating the olefin 1 again and not 5 (Scheme 2).
4. Settambolo, R.; Guazzelli, G.; Mengali, L.; Mandoli, A.;
Lazzaroni, R. Tetrahedron: Asymmetry 2003, 14, 2491.
5. Hydroformylation of (3R)-3-(pyrrol-1-yl)alk-1-enes 1a–c.
General procedure.
A
solution of pyrrolyl olefin
(3.36 mmol) and Rh4(CO)12 (substrate/Rh ¼ 100/1) in tol-
uene (10 mL) was introduced by suction into an evacuated
25 mL stainless steel reaction vessel. Carbon monoxide
was introduced, the autoclave was then rocked, heated to
the reaction temperature and hydrogen was rapidly
introduced to give the desired total pressure (CO/H2 ¼ 1/
1). When the gas absorption reached the value corre-
sponding to the desired conversion the reaction vessel was
rapidly cooled, and the reaction mixture was siphoned out
and GLC was used to determine the composition of the
reaction mixture.
In fact no traces of the internal olefin 5 were observed in
the crude reaction mixture at both partial or total con-
version. Due to the influence of the electron withdraw-
ing heteroaromatic effect, the methinic hydrogen bonded
to the carbon vicinal to the annular nitrogen into B
probably does not have sufficient hydride character for
b-hydride elimination.
6. Lazzaroni, R.; Raffaelli, A.; Settambolo, R.; Bertozzi, S.;
Vitulli, G. J. Mol. Catal. 1989, 50, 1.
7. 4b: MS m=e 179 (Mþ 30), 161 (52), 151 (38), 136 (37), 122
(57), 108 (85), 95 (63), 80 (43), 68 (100). 4c: MS m=e 179
(Mþ 8), 151 (30), 122 (36), 108 (28), 95 (12), 80 (36), 68
3. Conclusion
In conclusion, the chemistry reported here is a synthetic
application of the rhodium-catalyzed hydroformylation
providing new optically active 5-alkyl-5,6-dihydroindo-
lizines. Suitable experimental conditions avoiding race-
mization and enhancing the regioselectivity were set up
making the protocol a general regioselective and ste-
reospecific method for optically active 5-alkyl-5,6-dihy-
droindolizines. Investigations into diastereoselective and
stereospecific hydrogenation of the obtained dihydro-
indolizines to 5-alkyloctahydroindolizines of natural
origin are in progress.
(100).
26
D
8. For 3a ½aꢀ ¼ ꢁ107:5 (c 1.18, CH2Cl2). Selected data for
3b: as a orange oil (SiO2, CH2Cl2/hexane ¼ 1/7) 70% yield.
26
D
½aꢀ ¼ ꢁ60:8 (c 1.08, CH2Cl2). 1H NMR: d 6.66 (t,
J ¼ 2:2 Hz, 1H), 6.41 (dd, J ¼ 2:3; 9.6 Hz, 1H), 6.12 (dd,
J ¼ 2:8; 3.4 Hz, 1H), 6.03 (dd, J ¼ 1:5; 3.7 Hz, 1H), 5.64
(m, 1H), 3.84 (m, 1H), 2.65 (m, 1H), 2.44 (m, 1H), 2.22 (m,
1H), 0.99 (d, J ¼ 6:6 Hz, 3H), 0.86 (d, J ¼ 7:0 Hz, 3H). 13
C
NMR d 18.51, 19.4, 25.8, 31.4, 60.0, 105.5, 106.8, 117.9,
119.8, 121.0, 128.9. MS m=e 161 (Mþ 80), 146 (3), 132 (2),
118 (100), 91 (22), 63 (5), 39 (10). Selected data for 3c: as a
orange oil (SiO2, CH2Cl2/hexane ¼ 1/7) 75% yield.
26
D
½aꢀ ¼ ꢁ43:5 (c 0.86, CH2Cl2) 1H NMR: d 6.68 (t,
J ¼ 1:8 Hz, 1H), 6.42 (d, J ¼ 9:8 Hz, 1H), 6.13 (t,
J ¼ 3:1 Hz, 1H), 6.03 (b s, 1H), 5.63 (m, 1H), 4.04 (q,
J ¼ 6:5 Hz, 1H), 2.68 (m, 1H), 2.28 (m, 1H), 1.85–1.25 (m,
4H), 0.95 (t, J ¼ 7:5 Hz, 3H). 13C NMR: d 13.6, 18.9, 28.9,
36.2, 53.8, 105.6, 107.1, 117.2, 119.3, 119.6, 127.9. MS m=e
161 (Mþ 40), 132 (6), 118 (100), 91 (8).
Acknowledgements
Financial support by MIUR is gratefully acknowledged.
9. The NMR and MS spectral data for our synthetic (ꢁ)-3a
were identical with those previously reported for our
synthetic racemic 3a.2c
References and notes
€
1. (a) Eilbracht, P.; Barfacker, L.; Buss, C.; Hollmann, C.;
10. Lazzaroni, R.; Settambolo, R.; Caiazzo, A. Hydroformyl-
ation with unmodified rhodium catalysts. In Rhodium-
Catalysed Hydroformylation; van Leeuwen, P., Claver, C.,
Eds.; Dortrecht: Kluwer CMC, 2000; pp 15–33.
Kitsos-Rzychon, B. E.; Kranemann, C. L.; Rische, T.;
Roggenbuck, R.; Schmidt, A. Chem. Rev. 1999, 99, 3329;
(b) Breit, B. Acc. Chem. Res. 2003, 36, 264.