Development of a stepwise [3 + 3] annelation to functionalized piperidines

Article abstract of DOI:10.1021/ol050965t

(Chemical Equation Presented) A stepwise formal [3 + 3] cycloaddition sequence via a Grignard addition-cyclization reaction leads to a much improved piperidine synthesis. This methodology provides improved flexibility in both the aziridine substrate and T

Full text of DOI:10.1021/ol050965t

ORGANIC
LETTERS
2005
Vol. 7, No. 14
2993-2996
Development of a Stepwise [3
Annelation to Functionalized Piperidines
+ 3]
Katharine M. Goodenough,† Piotr Raubo,‡ and Joseph P. A. Harrity*,†
Department of Chemistry, UniVersity of Sheffield, Brook Hill, Sheffield S3 7HF, U.K.,
and The Neuroscience Research Centre, Merck, Sharp and Dohme, Terlings Park,
Harlow, Essex CM20 2QR, U.K.
j.harrity@sheffield.ac.uk
Received April 29, 2005
ABSTRACT
A stepwise formal [3
+ 3] cycloaddition sequence via a Grignard addition−cyclization reaction leads to a much improved piperidine synthesis.
This methodology provides improved flexibility in both the aziridine substrate and TMM equivalent.
We have recently been investigating a new approach to
enantiopure piperidines via a formal [3 + 3] cycloaddition
reaction of aziridines with Pd-TMM species.1,2 This tech-
nique provides an expedient route to enantiopure 2-pip-
eridines because the precursor aziridines are readily made
in enantiomerically pure form from the corresponding amino
acids.3 During the course of our studies on the reaction scope
and the employment of this methodology in target synthesis,
we have uncovered several instances where the Pd-TMM
reagent fails to add to the aziridine or does so with low levels
of conversion. Accordingly, we have attempted to uncover
a more nucleophilic alternative to the Pd-TMM for use with
less reactive aziridines and report herein our preliminary
findings.
(2-[(trimethylsilyl)methyl]-2-propen-1-yl acetate) 3, which
is in turn prepared from methallyl alcohol 1 (Scheme 1).5
Scheme 1
Our piperidine forming strategy comprises the formal
addition of trimethylenemethane4 to an aziridine (Figure 1).
In practice, we have employed Trost’s conjunctive reagent
† University of Sheffield.
While the conversion of 3 to piperidines 4 takes place in
one pot, the low conversions observed in some of our [3 +
‡ Merck, Sharp and Dohme.
(1) (a) Hedley, S. J.; Moran, W. J.; Prenzel, A. H. G. P.; Price, D. A.;
Harrity, J. P. A. Synlett 2001, 1596. (b) Hedley, S. J.; Moran, W. J.; Price,
D. A.; Harrity, J. P. A. J. Org. Chem. 2003, 68, 4286.
(2) For reviews of alternative [3 + 3] approaches to piperidine systems,
see: (a) Harrity, J. P. A.; Provoost, O. Org. Bio. Chem. 2005, 1349. (b)
Hsung, R. P.; Kurdyumov, A. V.; Sydorenko, N. Eur. J. Org. Chem. 2005,
23.
(3) Craig, D.; Berry, M. B. Synlett 1992, 41.
(4) For a review of Pd-TMM cycloadditions see: (a) Trost, B. M. Angew.
Chem., Int. Ed. Engl. 1986, 25, 1.
(5) Trost, B. M.; Chan, D. M. T.; Nanninga, T. M. Org. Synth. 1984,
62, 58.
Figure 1. [3 + 3] Cycloaddition to piperidines.
10.1021/ol050965t CCC: $30.25
© 2005 American Chemical Society
Published on Web 06/15/2005

Table 1. Optimization of Aziridine Ring Openinga
Table 2. Ring Closure to Piperidine 8a
entry
conditionsa
5 mol % of PdCl2, THF
5 mol % of Pd(OAc)2, 20 mol % of PBu3
60 mol % of Et3B, THF
10 mol % of Pd(OAc)2, 40 mol % of PPh3
25 mol % of Ti(OPr-i)4, 4 Å MS, benzene
10 mol % of Pd(OAc)2, 40 mol % of PPh3
25 mol % of Ti(OPr-i)4, 4 Å MS, toluene
yield (%)
entry
equiv of 1
additive
conditions
yield (%)
1
2
0
75b
1
2
3
4
5
6
10
3
4
5
5
25 °C, 2 h
25 °C, 18 h
25 °C, 2 h
25 °C, 2 h
25 °C, 18 h
25 °C, 2 h
19
53
65
87
88
75
MgBr2
MgBr2
MgBr2
MgBr2
MgCl2
3
4
72
100
5
a Organomagnesium reagent prepared by addition of TMEDA (2.6 equiv)
and n-BuLi (3.0 equiv) to 1 followed by MgX2 (2.6 equiv).
a All reactions run at reflux over 18 h. b Reaction run for 48 h and product
contaminated by endocyclic isomer.
3] cycloadditions suggested that the in situ generated Pd-
TMM reagent 5 was only moderately reactive toward
aziridines. In an effort to address this limitation, we
contemplated the use of more nucleophilic TMM equivalents.
As depicted in Scheme 1, reagent 3 ultimately emanates from
the allylmetal reagent 2. This raised an intriguing question
as to the feasibility of using 2 in a stepwise addition-
cyclization process that we envisaged could provide an
alternative method of piperidine synthesis with hitherto
unreactive aziridines.6
cyclization process, we were attracted by the prospect of
employing Pd catalysis. Specifically, Hirai and Makabe have
reported that Pd(II) catalysts mediate the closure of carbam-
ates onto allylic alcohols without need for preactivation of
the hydroxyl group.8 Additionally, Ti-9 and B-mediated10
amination of allylic alcohols have been reported that are
proposed to proceed via the intermediacy of a Pd π-allyl
intermediate. We decided to employ these procedures in
our cyclization reaction, and the results are shown in Table
2.
To test this hypothesis, we set out to investigate the
addition of 1 to aziridine 6; our results are shown in Table
1. Double deprotonation of methallyl alcohol and addition
of the organolithium reagent to 6 only succeeded in furnish-
ing 7 after a large excess of reagent was employed (entry
1). We therefore opted to perform a transmetalation to the
organomagnesium reagent by addition of MgBr2. Pleasingly,
addition of the Grignard reagent to 6 provided 7 in moderate
yield over an extended reaction time (entry 2). Further
optimization showed that excellent yields of product could
be obtained when aziridine 6 was added to 5 equiv of
Grignard reagent and that the reaction was effectively
complete within 2 h (entries 4 and 5). Unsurprisingly, the
use of MgCl2 also delivered an efficient addition reaction,
although 7 was generated in slightly lower overall yield in
this case (entry 6).
Our next task was to develop an efficient method for the
cyclization process. Early studies toward the synthesis of
Nuphar alkaloids showed that this ring closure could be
carried out in a single step using Mitsunobu conditions.7 This
method provided the piperidine product in a single step when
tributylphosphine and ADDP were used; however, this
stoichiometric route was rather expensive and resulted in
purification problems. In an attempt to find a more efficient
Attempts to promote the cyclization using PdCl2 were
unsuccessful, and starting material was recovered from the
reaction mixture (entry 1). In contrast, the B-promoted
amination proceeded with good conversion although 8 was
contaminated with alkene isomerization product (entry 2).
Pleasingly, the Ti-promoted reaction proceeded smoothly to
give the desired piperidine 8 in high yield with the optimal
conditions resulting from overnight reflux in toluene (entries
3 and 4).
We were now in a position to investigate the scope of the
stepwise piperidine forming reaction and, in particular, to
compare the efficiency of this process with the Pd-TMM
cycloaddition methodology. We began by investigating the
scope of 2-alkyl-substituted aziridine substrates, our results
are outlined in entries 1-5 of Table 3. We first investigated
the effect of N-protecting group as the Pd-TMM technique
was restricted to arylsulfonamide containing aziridines.
Indeed, the stepwise annelation of SES-protected aziridine
12 generated piperidine 14 albeit in modest overall yield
(entry 3). Phenyl-substituted aziridine 15 had performed
efficiently in the Pd-TMM [3 + 3] cycloaddition but
provided an almost equal mixture of regioisomeric products.
Interestingly, the use of the organomagnesium reagent
resulted in much more useful levels of regiocontrol in favor
of the 3,5-disubstituted piperidine, allowing 17 to be isolated
(6) Addition of a related Grignard reagent to epoxides has been shown
to produce pyrans: van der Louw, J.; van der Baan, J. L.; Out, G. J. J.; de
Kanter, F. J. J.; Bickelhaupt, F.; Klumpp, G. W. Tetrahedron 1992, 48,
9901.
(7) (a) Moran, W. J.; Goodenough, K. M.; Raubo, P.; Harrity, J. P. A.
Org. Lett. 2003, 5, 3427. (b) Goodenough, K. M.; Moran, W. J.; Raubo,
P.; Harrity, J. P. A. J. Org. Chem. 2005, 70, 207.
(8) (a) Yokoyama, H.; Otaya, K.; Kobayashi, H.; Miyazawa, M.;
Yamaguchi, S.; Hirai, Y. Org. Lett. 2000, 2, 2427. (b) Makabe, H.; Kong,
L. K.; Hirota, M. Org. Lett. 2003, 5, 27.
(9) Yang, S.-C.; Hung, C.-W. J. Org. Chem. 1999, 64, 5000.
(10) Kimura, M.; Futamata, M.; Shibata, K.; Tamaru, Y. Chem. Commun.
2003, 234.
2994
Org. Lett., Vol. 7, No. 14, 2005

Table 3. Stepwise Annelation of Aziridines
in high yield. Finally, aziridine 18 had been found to be
surprisingly sluggish in the [3 + 3] cycloaddition chemistry
but was converted to 20 in excellent overall yield using this
modified procedure (entry 5). Moreover, we were pleased
to find that this technique could also be extended to provide
bicyclic piperidines (entries 6 and 7) and spirocyclic products
(entry 8) in high overall yield.
would allow us to successfully prepare more heavily
substituted piperidines by this route.
One final limitation of the Pd-TMM methodology that
we hoped to overcome was in the addition of substituted
conjunctive reagents to aziridines. While Trost and co-
workers have shown that substituted Pd-TMM reagents 30
can add to Michael acceptor systems,11 we found that these
reagents would not participate in [3 + 3] cycloadditions. We
therefore set out to investigate if the stepwise addition process
As outlined in Scheme 2, double deprotonation of meth-
allyl alcohol 31, transmetalation with MgBr2, and addition
to aziridine 6 gave the corresponding adduct 32 in excellent
yield. Initial attempts to convert 32 into the desired piperidine
33 using the Pd-catalyzed cyclization furnished a mixture
of products that resulted from addition of the sulfonamide
unit at either end of the putative8 Pd π-allyl intermediate.
(11) Trost, B. M.; King, S. A. J. Am. Chem. Soc. 1990, 112, 408.
Org. Lett., Vol. 7, No. 14, 2005
2995

In conclusion, we report a complementary and stepwise
annelation procedure that rivals our previously developed
Pd-TMM-mediated cycloaddition reaction toward pip-
eridines. This technique expands the scope of the [3 + 3]
strategy both in terms of participating aziridines and con-
junctive reagents and promises to allow a broader range of
heterocycle targets to be made available.
Scheme 2
Acknowledgment. We thank the EPSRC and MSD for
financial support.
Supporting Information Available: Full experimental
details for the syntheses reported are provided. This material
is available free of charge via the Internet at http://pubs.acs.org.
Accordingly, we returned to use of the Mitsunobu reaction
to complete the cyclization and were pleased to find that
this took place efficiently to provide the trisubstituted
piperidine 33 as an equal mixture of diastereomers at C-6.
OL050965T
2996
Org. Lett., Vol. 7, No. 14, 2005