Enantioselective synthesis of substituted piperidines by addition of aryl grignard reagents to pyridine N-oxides

Article abstract of DOI:10.1021/ol303085q

The synthesis of optically active piperidines by enantioselective addition of aryl Grignard reagents to pyridine N-oxides and lithium binolate followed by reduction is reported for the first time. The reaction results in high yields (51-94%) in combinatio

Full text of DOI:10.1021/ol303085q

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Enantioselective Synthesis of Substituted
Piperidines by Addition of Aryl Grignard
Reagents to Pyridine N‑Oxides
Munawar Hussain,†,§ Thomas Sainte-Luce Banchelin,†,§ Hans Andersson,†
Roger Olsson,*,‡ and Fredrik Almqvist*,†
˚
Department of Chemistry and Umea Centre for Microbial Research, Umea University,
SE-90187 Umea, Sweden, ACADIA Pharmaceuticals Inc., 3911 Sorrento Valley
˚
˚
Boulevard, San Diego, California 92121, United States, Department of Chemistry and
Molecular Biology, University of Gothenburg, SE-41296 Gothenburg, Sweden, and
Department of Experimental Medical Science, Lund University, SE-22100, Lund, Sweden
fredrik.almqvist@chem.umu.se; roger@acadia-pharm.com
Received November 9, 2012
ABSTRACT
The synthesis of optically active piperidines by enantioselective addition of aryl Grignard reagents to pyridine N-oxides and lithium binolate
followed by reduction is reported for the first time. The reaction results in high yields (51ꢀ94%) in combination with good ee (54ꢀ80%). Some of
these products were subsequently recrystallized, affording enhanced optical purities (>99% ee).
As a consequence of its prevalence in many materials,
natural products, and therapeutic agents, the piperidine
framework has continued to capture the interest of
chemists worldwide.1 Although 4-substituted piperidines
are very common in many of the top-brand drugs in the
world (e.g., Risperdal2a and Aricept2b), their 2,4-substituted
analogues have not been as thoroughly investigated. One
explanation for this is the lack of practical methods that give
access to optically pure materials.2
Pyridine N-oxides and Grignard reagents are easily
accessible in most flavors, either from vendors or by
synthesis.3 In our search for a cheap, general, and
scaleup friendly method for the synthesis of six-membered
N-heterocycles, we revisited the addition of Grignard
reagents to pyridine N-oxides in 2007.4 A reaction pre-
viously studied by Kato and Kellogg in 1965 (Figure 1),5
and that had been lying largely dormant since, as a con-
sequence of inconclusive results. This reaction was developed
into an efficient and high yielding synthesis of substituted
pyridines, dihydropyridines, and dienal oximes.4,6 However,
† The Umeå University.
‡ ACADIA Pharmaceuticals Inc., University of Gothenburg, and Lund
University.
§ These authors contributed equally.
(1) (a) Fang, A. G.; Mello, J. V.; Finney, N. S. Org. Lett. 2003, 5, 967.
For review, see: (b) Buffat, M. G. P. Tetrahedron 2004, 60, 1701.
(c) Laschat, S.; Dickner, T. Synthesis 2000, 1781. (d) Comins, D. L.;
Abdullah, A. H. J. Org. Chem. 1982, 47, 4315. (e) Comins, D. L.; Joseph,
S. P.; Goehring, R. R. J. Am. Chem. Soc. 1994, 116, 4719. (f) Wang, X.;
Kauppi, A. M.; Olsson, R.; Almqvist, F. Eur. J. Org. Chem. 2003, 4586.
(g) Ahmed, A.; Molvia, K. I.; Nazim, S.; Baig, I.; Memon, T.; Rahil, M.
J. Chem. Pharm. Res. 2012, 4, 872.
€
(4) (a) Andersson, H.; Wang, X.; Bjorklund, M.; Olsson, R.; Almqvist, F.
Tetrahedron Lett. 2007, 48, 6941. (b) Andersson, H.; Olsson, R.; Almqvist, F.
Org. Biomol. Chem. 2011, 9, 337. (c) Andersson, H.; Sainte-Luce Banchelin,
T.; Das, J.; Olsson, R.; Almqvist, F. Chem. Commun. 2010, 46, 3384.
(d) Andersson, H.; Almqvist, F.; Olsson, R. Org. Lett. 2007, 9, 1335.
(e) Andersson, H.; Gustafsson, M.; Olsson, R.; Almqvist, F. Tetrahedron
Lett. 2008, 49, 6901. (f) Andersson, H.; Gustafsson, M.; Das, S.; Olsson, R.;
Almqvist, F. Tetrahedron Lett. 2010, 51, 4218.
(5) (a) Kato, T.; Yamanaka, H. J. Org. Chem. 1965, 30, 910.
(b) Kellogg, R. M.; Van Bergen, T. J. J. Org. Chem. 1971, 36, 1705.
€
(6) Andersson, H.; Gustafsson, M.; Bostrom, D.; Olsson, R.; Almqvist,
(2) (a) Schotte, A.; Janssen, P. F. M.; Gommeren, W.; Luyten, W. H.
M. L.; Van Gompel, P.; Lesage, A. S.; De Loore, K.; Leysen, J. E.
Psychopharmacology 1996, 124, 57. (b) Briks, J.; Harvey, R. J. Cochrane
Database of Systematic Reviews 2006, Art. No. CD001190. (c) Watson,
P. S.; Jiang, B.; Scott, B. Org. Lett. 2000, 23, 3679.
(3) (a) Campeau, L.-C.; Stuart, D. R.; Leclerc, J.-P.; Bertrand-Laperle,
M.; Villemure, E.; Sun, H.-Y.; Lasserre, S.; Guimond, N.; Lecavallier, M.;
Fagnou, K. J. Am. Chem. Soc. 2009, 131, 3291. (b) Coperet, C.; Adolfsson,
H.; Khuong, T.-A. V.; Yudin, A. K.; Sharpless, K. B. J. Org. Chem. 1998, 63,
1740. For review, see: (c) Bull, J. A.; Mousseau, J. J.; Pelletier, G.; Charette,
A. B. Chem. Rev. 2012, 112, 2642.
F. Angew. Chem., Int. Ed. 2009, 48, 3288.
r
10.1021/ol303085q
XXXX American Chemical Society

due to the preferential isolation of the ring-opened products
in the initial studies in mid-1960, pyridine N-oxides have not
been viewed as starting materials for the synthesis of sub-
stituted piperidines.7 Instead several methods were devel-
oped by using other activated pyridines, most notably by
Comins,2,8 Charette,7,9aꢀ9c Shibasaki,9d Feringa,9e and
Arndtsen.9fꢀh Challenged by this we developed the Grignard
addition to pyridine N-oxides, into a productive methodol-
ogy for the synthesis of substituted piperidines. Although,
this reaction is efficient, e.g., high yielding and highly stereo-
and regioselective, the products were racemic.6
and Frejd (BODOL)12c employing diols as chiral ligands.12
In our evaluation of different reaction conditions BINOL stood
out as the most promising chiral ligand for this reaction.13
4-Phenyl pyridine N-oxide (1a) was used in the initial
studies of the reaction. The substrate selection was based
on the rationale that control of additional stereocenters in
the reaction is necessary to allow development of more
complex substitution patterns. It is also important that the
4-substitution is not masking a substrate limitation, as is
the case with additions to N-acyl activated pyridines, where
additions to the 4-position have been seen.1d
The initial results emphasized the importance of the addi-
tion order of the reagents. A significant increase in yields and
enantiomeric excess was observed when pyridine-N-oxide 1a
and the BINOL-ate complex was formed prior to the addition
of the Grignard reagent. As it is likely that the first 2 equiv of
the Grignard reagent resulted in magnesium binolates, we
decided to investigate if the counterion was important
Hence, BINOL was deprotonated using n-BuLi (2.0 equiv)
in THF prior to the addition of the Grignard reagent, which
resulted in a trend toward an increase in both yield and
enantiomeric purities.13 However, the results were not repro-
ducible and switching the counterion to sodium or potassium
by deprotonating with NaH or KH, respectively, gave no
significant effect.13 In addition to being a strong base, n-BuLi
is also a potential nucleophile, and the crude LC-MS spec-
trum of the n-BuLi generated Li-binolate complex showed
a byproduct with m/z 343 (20ꢀ30%) corresponding to
BINOLþbutyl. Additional organolithium bases were there-
fore screened. In contrast to LDA, which gave no improve-
ments, phenyl lithium in THF gave a significant increase in
both yields, 51ꢀ94% (from 22 to 75%), and % enantiomeric
excess, 46ꢀ80% (from 48 to 68%).13 More importantly the
robustness of the protocol improved considerably, resulting in
reproducible results and cleaner reactions. The subsequent
reduction was performed at low temperature to avoid ring
opened byproducts as previously observed.6,14 Reaction at a
larger scale gave better yields (Table 1). The chiral ligand
BINOL was both easily removed and recovered from the
crude reaction mixture by extraction.15
Figure 1. Grignard addition to pyridine N-oxides.
Herein we report the first enantioselective synthesis of
substituted piperidines by the addition of Grignard re-
agents to pyridine N-oxides. The high reactivity between
pyridine N-oxides and Mg reagents enables mild reaction
conditionsthat increase the functionalgrouptoleranceand
scope of the reaction. However the high reactivity makes
stereoselective additionsdifficult. Asa consequenceof this,
there are few examples reported on the use of reactive
Grignard reagents in enantioselective additions. Initially,
we were inspired by Fu and Shintani;10 however using
Grignard reagents in combination with (ꢀ)-sparteine gave no
enantiomeric excess on addition to pyridine N-oxides. No-
tably, in our hands (ꢀ)-sparteine afforded optically active
piperazines in the reaction between Mg reagents and pyrazine
N-oxides.11 Other sources of inspiration were the work from
the groups of Seebach (TADDOL),12a Tejero (BINOL),12b
(12) (a) Seebach, D.; Weber, B. Angew. Chem., Int. Ed. 1992, 84.
(b) Merchan, F. L.; Merino, P.; Rojo, I.; Tejero, T. Tetrahedron: Asymmetry
1996, 7, 667. (c) Almqvist, F.; Torstensson, L.; Gudmundsson, A.; Frejd, T.
Angew. Chem., Int. Ed. 1997, 36, 376.
(13) See Tables 1 and 2, Supporting Information.
(14) General procedure for the synthesis of 2aꢀj using PhLi
(Table 1): In a 25 mL round-bottom flask R (þ) BINOL (1.2 equiv)
was dried under vacuum for 30 min and dissolved in THF (10 mL). The
mixture was cooled to ꢀ78 °C, and PhLi (2.4 equiv) was added dropwise
under inert atmosphere. Then the mixture was warmed to ambient
temperature and stirred for 30 min. The colorless reaction mixture was
cooled again to ꢀ78 °C, and 4-phenyl-pyridine-N-oxide (1a) or pyridine-
N-oxide (1b) (1.0 equiv) was added. The mixture was allowed to reach
room temperature, where it was stirred for another 60 min. A yellow
mixture was obtained and cooled to ꢀ78 °C. To this cooled mixture,
ArMgBr (3.0 equiv) in THF was added dropwise and stirred for 40 min
(followed by LC-MS). A fresh NaBH4 (3.0 equiv) suspension in MeOH
(2 mL) was added to the mixture, which thereafter was allowed to reach
room temperature. CH2Cl2 (20 mL) and H2O (10 mL) were added to the
reaction mixture, and the organic layer was shaken with NaOH 10%
(3 ꢁ 10 mL) and then washed with H2O, dried (Na2SO4), and concen-
trated under reduced pressure. The crude reaction mixture was purified
by flash silica column chromatography (AcOEt/heptane).
(7) Legault, C.; Charette, A. B. J. Am. Chem. Soc. 2003, 125, 6360.
(8) Comins, D. L.; Kuethe, J. T.; Hong, H.; Lakner, F. J.; Concolino,
T. E.; Rheingold, A. L. J. Am. Chem. Soc. 1999, 121, 2651.
(9) (a) Charette, A. B.; Grenon, M.; Lemire, A.; Pourashraf, M.; Martel,
J. J. Am. Chem. Soc. 2001, 123, 11829. (b) Legault, C. Y.; Charette, A. B.
J. Am. Chem. Soc. 2005, 127, 8966. (c) Mousseau, J. J.; Bull, J. A.; Charette,
A. B. Angew. Chem., Int. Ed. 2010, 49, 1115. (d) Ichikawa, E.; Suzuki, M.;
Yabu, K.; Albert, M.; Kanai, M.; Shibasaki, M. J. Am. Chem. Soc. 2004,126,
ꢀ
ꢀ
ꢀ~
ꢀ
11808. (e) Fernandez-Ibanez, M. A.; Macia, B.; Pizzuti, M. G.; Minnaard,
A. J.; Feringa, B. L. Angew. Chem., Int. Ed. 2009, 48, 9339. (f) Beveridge,
R. E.; Arndtsen, B. A. Synthesis 2010, 1000. (g) Beveridge, R. E.; Black,
D. A.; Arndtsen, B. A. Eur. J. Org. Chem. 2010, 3650. (h) Black, D. A.;
Beveridge, R. E.; Arndtsen, B. A. J. Org. Chem. 2008, 73, 1906.
(10) Shintani, R.; Fu, G. C. Angew. Chem., Int. Ed. 2002, 41, 1057.
(11) Andersson, H.; Sainte-Luce Banchelin, T.; Das, S.; Gustafsson,
M.; Olsson, R.; Almqvist, F. Org. Lett. 2010, 12, 284.
(15) Surendra, K.; Corey, E. J. J. Am. Chem. Soc. 2012, 134, 11992.
B
Org. Lett., Vol. XX, No. XX, XXXX

Optically active N-hydroxyl tetrahydropyridines are
generally interesting intermediates for further synthesis,
but our main goal was to develop a robust method to
obtainopticallyactivesubstitutedpiperidines. The remain-
ing reduction to the desired piperidines had to be stereo-
specific in the case of the 4-substituted analogues, since a
new stereocenter would be formed. In the absence of a
substituent in the 4-position we envisioned a high chance
for double bond migration to the conjugated derivative,
which would destroy the previously formed stereocenter.
Indeed, although palladium on charcoal and hydrogen gas
gave complete reduction of both the double bond and the
N-hydroxyl functionality, these conditions proved detri-
mental to optical purity. For the 4-substituted derivatives
2d and 2e the enantiomeric excess dropped from >98% to
90% of both the resulting disubstituted piperidines 3a and
3b (Table 2, entries 1 and 2). However, the reduction was
stereospecific, yielding the cis-diastereomer as the sole
product. As suspected, the reduction of N-hydroxyl tetra-
hydropyridine 2i with this method gave 2-substituted
piperidine 3c as a racemate (entry 3).
Table 1. Enantioselective Grignard Addition to Pyridine
N-Oxides
entry
N-oxidea
R
R1
yield (%)c
ee(%)c
1
1a
1a
1a
1a
1ab
1a
1ab
1a
1a
1b
1bb
1b
1bb
1a
1a
1a
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
Ph
H
Ph
2a (86)
2b (76)
2c (76)
2d (68)
2d (70)
2e (94)
2e (94)
2f (51)
2g (71)
2h (62)
2h (94)
2i (68)
2i (71)
2j (trace)
2k (00)
2l (30)
70
2
4-MePh
62
3
3,5(Me)2Ph
4-biphenyl
4-biphenyl
4-(MeO)Ph
4-(MeO)Ph
2-MePh
46
4
70
5
76(99)d
64
6
7
74(98)d
56
8
9
4-ClPh
80
10
11
12
13
14
15
16
4-biphenyl
4-biphenyl
4-(MeO)Ph
4-(MeO)Ph
Me
72
H
80
H
51
H
55
Ph
Ph
Ph
ꢀ
PhCC
ꢀ
Table 2. One-Step Reduction to Optically Active Piperidines
vinyl
33
a 0.17 mmol scale. b 2.0 mmol scale. c Yields and ee when PhLi was
used as base. d % ee after enantiomeric enrichment by recrystallization
from ethanol.
Thereafter a set of different aryl Grignard reagents were
used as nucleophiles (Table 1). This resulted in most cases
in yields and ee’s above 70% and 60% respectively, with
the best example giving 80% ee (entry 9). However, the
more sterically demanding aryl Grignard reagents (i.e.,
2-methyl-phenylmagnesium bromide) and the 3,5-disub-
stituted phenyl Grignard gave lower yields, % ee, or both
(entries 3 and 8). Notably scaling up of the reaction
demonstrated a significant increase of yields and enantio-
meric excesses (entries 4ꢀ7 and 10ꢀ13). When the Grignard
reagent was added to a mixture of lithium BINOL-ate and
the unsubstituted pyridine-N-oxide (1b), the products 2h
and 2i were obtained in good yields (94% and 71%, respec-
tively). In addition, comparable enantioselectivities as seen
for the 4-phenyl substituted pyridine-N-oxide were obtained
(entries 11 and 13). Further enantiomeric enrichment via
crystallization from ethanol of the optically active N-hydroxyl
tetrahydropyridines 2d and 2e resulted in >98% ee in both
cases (entries 5 and 7).
Next alkyl, alkenyl, and alkynyl Grignard reagents were
studied. As for the addition of alkyl Grignards to pyridine-
N-oxide without binolate complexation,4e the result was depro-
tonation instead of nucleophilic attack resulting in uncon-
sumed starting material after workup (Table 1, entry 14). The
same result was seen when the phenylacetylene Grignard
reagent was used (Table 1, entry 15). However, when a vinyl
Grignard was used as the nucleophile the desired product 2l
was formed, albeit in low yield and modest enantiomeric
purity (30% and 33%, respectively) (Table 1, entry 16).
entry
Ra
R1
yield (%)
ee (%)
1
2
3
Ph
Ph
H
4-biphenyl
4-(MeO)Ph
4-(MeO)Ph
3a (82)
3b (88)
3c (94)
90
90
Rac
a Starting with ee (2d) 99%, (2e) 98%, and (2i) 55%.
A solution to this problem could be a milder stepwise
reduction, which besides solving the racemization issue
also would result in new interesting optically active inter-
mediates for further synthesis. Indeed, reduction of 2e with
PtO2/EtOH/H2 provided N-hydroxyl intermediate 4a
(84%), and subsequent reduction of the hydroxyl amine
to the corresponding amine with In/Zn in sat. aq. NH4Cl/
EtOH16 afforded the cis-2,4-disubstituted piperidine 3b in
94% yield and with a retained ee of 98% (Table 3, entry 1).
Fortunately, this stepwise approach was also mild enough
to be applied for the reduction of analogues 3c and 3d,
which racemized under the previous conditions. As a
bonus, the intermediate 2-substituted hydroxyl amines 4b
and 4c obtained after reduction (PtO2/EtOH/H2) were
easy to purify via crystallization from ethanol resulting
in enantiomeric enrichment (from 80% ee to 98% ee for
4c and from 55% ee to 90% ee for 4b (entries 2 and 3)).
Subsequent reduction of the hydroxyl amines was
(16) Cicchi, S.; Bonanni, M.; Cardona, F.; Revuelta, J.; Goti, A. Org.
Lett. 2003, 5, 1773.
Org. Lett., Vol. XX, No. XX, XXXX
C

Table 3. Two-Step Reduction to Optically Active Piperidines
Scheme 1. Two-Step Reduction in Reverse Order
entry Ra
R1
yield (%) ee (%) yield (%) ee (%)
b
reagents to pyridine N-oxides gave N-hydroxyltetrahydro-
pyridines in good optical purities, 54ꢀ80% ee. Crystal-
lization, stepwise reduction, or both gave optically active
piperidines with an enantiomeric purity up to 99% in high
yields. The enantioselectivity was obtained simply by adding
recycable chiral lithium binolate ligands; no chiral auxiliary
that would require eventual removal is needed in the process.
1
2
3
Ph 4-(MeO)Ph
4a (84)
4b (94)
4c (96)
ꢀ
3b (94)
3c (71)
3d (94)
98
90
98
H
H
4-(MeO)Ph
4-biphenyl
55, 90c
80, 98c
a Starting with ee (2e) 98%, (2h) 80%, and (2i) 55%. b At this stage ee
was not measurable due to poor baseline separation. c ee after enantio-
meric enrichment by crystallization in ethanol.
straightforward, and the piperidines3cand 3dweresynthe-
sized with retained optical purities (entries 2 and 3).
If desired, the reduction sequence could be reversed
without affecting the optical purity. Piperidine 3d was
successfully obtained from 2h via tetrahydropiperidine 5
in a 96% overall yield (Scheme 1) with the same ee as
obtained previously (80% ee) (Table 1, entry 11).
In conclusion, we have reported an efficient synthesis
of optically active 2- and 2,4-disubstituted piperidines,
N-hydroxyltetrahydropyridines, N-hydroxylpiperidines, and
tetrahydropyridines. Enantioselective addition of Grignard
Acknowledgment. We thank the Swedish Natural Re-
search Council (FA) and the Knut and Alice Wallenberg
Foundation (FA) for financial support. M.H. thanks the
Carl Tryggers Foundation for Scientific research for pro-
viding a postdoctoral fellowship.
Supporting Information Available. Experimental de-
tails. This material is available free of charge via the
Internet at http://pubs.acs.org.
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX