A modular, efficient, and stereoselective synthesis of substituted piperidin-4-ols

Article abstract of DOI:10.1002/anie.201004712

The pied piper-idinol of Hamelin: The one-pot synthesis of piperidin-4-ols by sequential gold-catalyzed cyclization, chemoselective reduction, and spontaneous Ferrier rearrangement has a broad substrate scope and shows excellent diastereoselectivity in th

Full text of DOI:10.1002/anie.201004712

Communications
DOI: 10.1002/anie.201004712
Nitrogen Heterocycles
A Modular, Efficient, and Stereoselective Synthesis of Substituted
Piperidin-4-ols**
Li Cui, Chaoqun Li, and Liming Zhang*
Piperidine is a key structural motif in
various alkaloids and a variety of com-
pounds studied in medicinal chemistry.
Though many methods have been
developed for their construction,
there is still
a
need for novel
approaches, especially those with high
efficiency, good modularity, and excel-
lent stereoselectivity.
Recent intense research in gold
catalysis[1] has provided several novel
methods for the synthesis of piperi-
dines.[2,3] For example, we have previ-
Scheme 1. Our design for a modular synthesis of N-unsubstituted piperidines.
ously reported that piperidin-4-ones could be prepared in a
two-step, [4+2] manner;[2a] however, the products were
limited to those in which the ring nitrogen atoms contained
hard-to-remove aliphatic substituents or benzyl groups which
afforded low regioselectivities. To address this deficiency and
develop a general and modular synthesis of N-unsubstituted
piperidines, we envisioned that a gold-catalyzed cyclization of
N-homopropargyl amide 2 would offer cyclic imidate 3, which
could be chemoselectively reduced to afford a-amino ether A
(Scheme 1). We anticipated that A would undergo sponta-
neous Ferrier rearrangement to furnish piperidin-4-one B,
which might be further reduced in situ to the corresponding
alcohol (4). Several aspects of this design are noteworthy:
1) the sequence is highly modular and flexible; it is an overall
{[2+3]+1} annulation from readily available imines, a prop-
argyl Grignard, and carboxylic acids or their derivatives.
2) The enantiomeric synthesis is readily achievable as chiral
amine 1 would be easily prepared from chiral sulfinyl
imines.[4] 3) This sequence constitutes an alternative to an
aza-Petasis–Ferrier rearrangement, which has not yet been
realized.[5,6] The lack of reported aza-Petasis–Ferrier rear-
rangement is surprising as the Petasis–Ferrier rearrange-
ment[7] has been applied with much success in total syntheses
of complex natural products.[8] 4) The piperidine nitrogen
atom is free and could be readily derivatized. 5) The gold
catalysis is not the key transformation step but is instead
employed to deliver the requisite intermediate (i.e., 3) for
subsequent processes. This sequential combination of gold
catalysis and other distinctively different transformations in a
one-pot process offer new opportunities to develop versatile
synthetic methods with high efficiency.
We began our investigation by examining the feasibility of
the gold catalysis[9] using amide 5 as the substrate. To our
delight, the gold-catalyzed cyclization proceeded quantita-
tively in either dichloromethane or tetrahydrofuran at
ambient temperature [Eq. (1); M.S. = molecular sieves]. The
keys to this reaction were the addition of MsOH (1.2 equiv) in
order to prevent the nitrogen atom of imidate 6 from
coordinating to, and thus deactivating, the gold catalyst[10]
and the use of molecular sieves to minimize hydrolysis.
Owing to the sensitivity of imidate 6 to hydrolysis, we
decided to study its reduction in a one-pot process. Hence,
upon the complete consumption of amide 5 from the gold-
catalysis step, various reductants were added. To our delight,
the all-cis-isomer 7 was formed selectively when borane was
used as the reductant (Table 1, entry 1). Screening other
reductants, especially boron-based ones, revealed that cat-
echolborane (5 equiv) worked the best (Table 1, entries 5–9),
and the product was isolated in 80% yield by running the
reaction in dichloromethane at À788C. Notably, 7 was formed
with excellent diastereoselectivity, and other potential dia-
[*] Dr. L. Cui, Dr. C. Li, Prof. Dr. L. Zhang
Department of Chemistry and Biochemistry
University of California, Santa Barbara, CA (USA)
Fax: (+1)805-893-4120
E-mail: zhang@chem.ucsb.edu
Homepage: http://www.chem.ucsb.edu/~zhang/index.html
[**] We thank the NSF CAREER (CHE 0748484), the NIGMS (R01
GM084254), and the UCSB for generous financial support, and Dr.
Guang Wu for helping with X-ray crystallography.
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.201004712.
9178
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 9178 –9181

Angewandte
Chemie
Table 1: One-pot, sequential gold catalysis and reduction: Optimization
We then tested a range of secondary amides as substrates
of the reduction reaction conditions.[a]
following the optimized one-pot sequence with some fine-
tuning of reaction temperatures and the amount of the
borane. Different acyl groups, including aliphatic (Table 2,
entries 1–7) and aromatic substituents (Table 2, entries 8–11)
were well-tolerated. Sterically bulky substituents (Table 2,
entry 4) also successfully afforded the cyclized product,
although it required a more-prolonged reaction time. Differ-
À
ent functional groups, including a nonconjugated C C double
bond (Table 2, entry 6), fluoro groups (Table 2, entries 7 and
9), a carboxylate (Table 2, entry 11), and a naphthyl group
(Table 2, entry 10) were also tolerated. However, trifluoroa-
cetamide 8 (R’ = CF3, R = Cy) was not a suitable substrate as
its weakly nucleophilic carbonyl group failed to undergo gold-
catalyzed cyclization. Notably, high to excellent diastereose-
lectivities were observed in the cases of aliphatic amides.
However, the catecholborane reductions of aromatic amides
(Table 2, entries 8–11) were very slow at À408C, and the
reactions were run at ambient temperature in order to achieve
completion in 24 hours; as a result, low diastereoselectivities
were observed in most of these cases. To our delight, whilst
the major isomers were the expected all-cis isomers, the minor
component in each case appeared to be the 4-OH epimer,
which was confirmed by oxidation of the separated isomers of
9j to afford a common piperidin-4-one.[12] These results
indicated that the piperidine-ring-forming step was highly
diastereoselective, even at room temperature. In contrast to
Table 2, entry 6, the vinyl group in acrylamide 8l (Table 2,
entry 12) was reduced during the reaction, and piperidine 9l
with an ethyl group instead was isolated in 47% yield.
Entry Solvent Reductant (equiv)
Conditions
T [8C]
Yield [%][b]
t [h]
7
isomer
1
2
3
4
5
6
7
8
9
CH2Cl2 BH3·Et2O (5)
CH2Cl2 NaBH3CN (5) AlCl3 (2)
CH2Cl2 DIBAL-H (5)
CH2Cl2 9-BBN
À40
0 to RT
À78 to RT
À40
2
41
–
–
5
76
72
–
–
–
–
5
4
[c]
8
2
2
2
2
4
CH2Cl2 catecholborane (5)
À40
THF
catecholborane (5)
À40
toluene catecholborane (5)
CH2Cl2 catecholborane (5)
CH2Cl2 catecholborane (5)
À40
60 10
0
77
7
3
À78
86[d]
[a] [5]=0.05m. [b] Estimated by 1H NMR spectroscopy, using diethyl
phthalate as an internal reference. [c] 50% of 6 leftover. [d] 80% yield of
isolated product. DIBAL-Hdiisobutylaluminum hydride; 9-BBN=9-
borabicyclo[3.3.1]nonane; THF=tetrahydrofuran.
stereomers were formed in negligible amounts. The relative
stereochemistry of 7 was initially established by NMR
spectroscopy studies and later confirmed by X-ray crystallo-
graphic analysis of 9p (see the Supporting Information).[11]
Table 2: Scope of the one-pot sequential gold catalysis, chemoselective reduction, and Ferrier rearrangement.[a]
Entry
8
R, R’
t
[h]
T
[8C]
Yield of 9
[%][b]
Entry
8
R, R’
t
[h]
T
Yield of 9
[%][b]
[8C]
(d.r.)
(d.r.)
1
2
3
4
5
nHept, Me
Cy, nPr
Cy, Cy
4
6
À78
À40
À40
À40
À40
9a
9b
9c
9d
9e
72
(>25:1)
10[c]
11[c]
12[c]
13
Cy,
1-naphthyl
Cy,
4-MeO2CC6H4
Cy,
vinyl
24
24
24
8
RT
RT
9j
9k
67
(4:1)
78
(2:1)
47
(4:1)
77
(>25:1)
77
(>25:1)
83
(11:1)
80
(>25:1)
74
71
(10:1)
73
(10:1)
68
(13:1)
79
(15:1)
72
(17:1)
52
(>25:1)
70
(13:1)
61
(1.5:1)
6
À40 to RT
À78
9 l[d]
9m
9n
Cy, tBu
Cy, Bn
Cy,
12
6
Ph, Me
14
Ph, iPr
12
À40
6
24
24
À40
À40
9 f
15
16
Ph, tBu
12
12
À40
À40
9o
9p
7[c]
Cy, FCH2
Cy, Ph
9g
4-MeOC6H4,
Me
o-tolyl
Me
4-CF3C6H4,
Me
8[c]
9[c]
24
24
RT
RT
9h
9i
17
18
16
12
À40
À40
9q
9r
(>25:1)
81
(>25:1)
Cy,
3-FC6H4
[a] [8]=0.1m. [b] Yield of isolated product. [c] 8 equiv of catecholborane. [d] R’=Et. Bn=benzyl; Cy=cyclohexyl.
Angew. Chem. Int. Ed. 2010, 49, 9178 –9181 ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9179

Communications
Substrates containing an aryl group at the homopropargylic
In conclusion, we have developed a one-pot synthesis of
piperidin-4-ols by sequential gold-catalyzed cyclization, che-
moselective reduction, and spontaneous Ferrier rearrange-
ment. This reaction has a broad substrate scope and shows
excellent diastereoselectivities in the ring-formation step; in
combination with a routine amide formation, it constitutes a
highly flexible and diastereoselective [5+1] cycloaddition
approach to piperidines. As homopropargylic amines can be
readily prepared with excellent ee values from chiral sulfinyl
imines and propargylmagnesium bromide, this overall
{[2+3]+1} modular approach offers an ideal solution to the
enantioselective synthesis of various substituted piperidines.
Importantly, the piperidine nitrogen atom is unsubstituted
and can be readily derivatized. By coupling with one-pot
intramolecular alkylation reactions, this chemistry provides a
rapid access to quinolizidines and indolizidines, and allowed
us to complete a succinct enantioselective synthesis of (+)-
subcosine II.
position worked equally well in this one-pot process (Table 2,
entries 13–18), and at À408C the diastereoselectivities were
mostly excellent. Again, steric bulk (Table 2, entries 15 and
17) in the substituents was readily tolerated.
A key feature in these piperidin-4-ol products was that the
nitrogen atom in the ring was free and could be readily
derivatized. For example, subsequent one-pot intramolecular
alkylation (Scheme 2) provided quick access to the quinoli-
zidine and indolizidine skeletons, which can be found in the
structures of a range of alkaloids.[13]
As a demonstration of the synthetic utility of this
procedure, an enantioselective synthesis of (+)-subcosine
II[14,15] was achieved in six steps in a 22% overall yield
(Scheme 3). Notably, homopropargyl amine 10 was easily
prepared in greater than 94% ee using the sulfinyl imine
chemistry reported by Ellman et al.,[4a] and neither the gold
catalysis nor the reduction/Ferrier rearrangement compro-
mised the stereochemical integrity of the original chiral
carbon center.
Experimental Section
General procedure for the preparation of piperidin-4-
ols: 4 ꢀ M.S. (100 mg) were added to an oven-dried
Schlenk tube. The tube was flamed dried under
vacuum and flushed with N2 three times. Under a N2
atmosphere, an amide (0.1 mmol), [Au(PPh3)NTf2]
(5 mol%), freshly distilled CH2Cl2 (1.6 mL), and a
freshly prepared solution of MeSO3H in CH2Cl2
(0.4 mL, 0.03m) were sequentially added to the
reaction tube. The reaction mixture was stirred at
room temperature for 1 h before the reduction at the
indicated reaction temperature. Catecholborane
(0.6 mmol) was added to the reaction vessel and the
progress of the reduction was monitored by TLC.
Upon completion, the reaction was quenched with
MeOH and then stirred at room temperature for
15 min. A saturated disodium tartrate aqueous solu-
tion was added to the reaction, and the reaction
mixture was stirred for another 15 min. The resulting
mixture was treated with 10% NaOH (saturated with
NaCl) and extracted three times with CH2Cl2. The
combined organic layers were dried with anhydrous
MgSO4, filtered, and concentrated under vacuum.
The residue was purified by flash chromatography on
silica gel.
Scheme 2. a) Et3N, CH2Cl2; b) 1. [Au(PPh3)NTf2] (5 mol%), MsOH (1.2 equiv), 4 ꢀ M.S.,
1 h; 2. catecholborane (6 equiv), À40 8C, 24 h; 3. MeOH, K2CO3, reflux, 8 h. Ms=me-
thanesulfonyl; Tf=trifluoromethanesulfonyl.
Received: July 30, 2010
Published online: October 15, 2010
Keywords: cyclization · gold ·
.
nitrogen heterocycles · piperidine · reduction
[1] For selected reviews, see: a) A. S. K. Hashmi,
Chem. Rev. 2007, 107, 3180; b) A. Fꢁrstner, P. W.
Davis, Angew. Chem. 2007, 119, 3478; Angew.
Chem. Int. Ed. 2007, 46, 3410; c) S. Wang, G.
Zhang, L. Zhang, Synlett 2010, 692; d) A.
Arcadi, Chem. Rev. 2008, 108, 3266; e) Z. G.
Li, C. Brouwer, C. He, Chem. Rev. 2008, 108,
3239; f) R. Skouta, C.-J. Li, Tetrahedron 2008,
64, 4917; g) S. M. Abu Sohel, R.-S. Liu, Chem.
Scheme 3. Six-step, enantioselective total synthesis of (+)-subcosine II. a) (R)-tBu-
SONH2, CuSO4, CH2Cl2; propargylmagnesium bromide; b) conc. HCl, MeOH;
c) Cl(CH2)4COCl, CH2Cl2, Et3N; d) 1. [Au(PPh3)NTf2] (5 mol%), MsOH (1.2 equiv), 4 ꢀ
M.S., 1 h; 2. catecholborane (8 equiv), À40 8C, 24 h; 3. MeOH, K2CO3, reflux, 4 h;
=
e) Ph3P (2.5 equiv), DEAD (2.5 equiv), trans-3,4-(MeO)2C6H3CH CHCO2H (2 equiv),
toluene, 12 h. DEAD=diethyl azodicarboxylate.
9180 www.angewandte.org
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2010, 49, 9178 –9181

Angewandte
Chemie
Soc. Rev. 2009, 38, 2269; h) D. J. Gorin, B. D. Sherry, F. D. Toste,
Chem. Rev. 2008, 108, 3351.
[2] a) L. Cui, Y. Peng, L. Zhang, J. Am. Chem. Soc. 2009, 131, 8394;
b) C. Kim, H. J. Bae, J. H. Lee, W. Jeong, H. Kim, V. Sampath,
Y. H. Rhee, J. Am. Chem. Soc. 2009, 131, 14660; c) L. Cui, L.
Zhang, Sci. China Ser. B 2010, 53, 113.
[3] For the gold-catalyzed azepanone preparations, see: a) L. Cui, L.
Ye, L. Zhang, Chem. Commun. 2010, 46, 3351; b) L. Cui, G.
Zhang, Y. Peng, L. Zhang, Org. Lett. 2009, 11, 1225.
reactions, see: b) E. Tayama, S. Otoyama, W. Isaka, Chem.
Commun. 2008, 4216; c) H. Frauenrath, T. Arenz, G. Raabe, M.
Zorn, Angew. Chem. 1993, 105, 74; Angew. Chem. Int. Ed. Engl.
1993, 32, 83; d) A. P. Kozikowski, P. U. Park, J. Org. Chem. 1984,
49, 1674; The aza-Ferrier rearrangement was previously known
as the aza-Petasis–Ferrier rearrangement: e) M. Terada, Y. Toda,
J. Am. Chem. Soc. 2009, 131, 6354.
[9] A. S. K. Hashmi, M. Rudolph, S. Schymura, J. Visus, W. Frey,
Eur. J. Org. Chem. 2006, 4905.
[4] a) J. A. Ellman, T. D. Owens, T. P. Tang, Acc. Chem. Res. 2002,
35, 984; b) P. Zhou, B.-C. Chen, F. A. Davis, Tetrahedron 2004,
60, 8003.
[5] a) N. A. Petasis, S. P. Lu, Tetrahedron Lett. 1996, 37, 141; b) N. A.
Petasis, S.-P. Lu, J. Am. Chem. Soc. 1995, 117, 6394.
[6] For examples of its applications in the total synthesis of natural
products, see: a) A. B. Smith III, T. Bosanac, K. Basu, J. Am.
Chem. Soc. 2009, 131, 2348; b) A. B. Smith, K. Basu, T. Bosanac,
J. Am. Chem. Soc. 2007, 129, 14872; c) A. B. Smith, E. F.
Mesaros, E. A. Meyer, J. Am. Chem. Soc. 2005, 127, 6948.
[7] For a related case, see Ref. [2b]. For a review of aza-Prins
cyclization reactions in piperidine synthesis, see: C. Olier, M.
Kaafarani, S. Gastaldi, M. P. Bertrand, Tetrahedron 2010, 66, 413.
[8] For an analogous case, see: a) L. V. Adriaenssens, R. C. Hartley,
J. Org. Chem. 2007, 72, 10287; For examples of aza-Ferrier
[10] Without using the acid, the reaction was sluggish and did not
proceed to completion.
[11] CCDC 795487 (9p) contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
from The Cambridge Crystallographic Data Centre via www.
ccdc.cam.ac.uk/data_request/cif.
[12] See the Supporting Information.
[13] For selected reviews, see: a) J. P. Michael, Alkaloids, Vol. 55,
Academic Press, New York, 2001, p. 91; b) J. P. Michael, Nat.
Prod. Rep. 2008, 25, 139.
[14] K. Fuji, T. Yamada, E. Fujita, H. Murata, Chem. Pharm. Bull.
1978, 26, 2515.
[15] P. Chalard, R. Remuson, Y. Gelas-Mialhe, J. C. Gramain,
Tetrahedron: Asymmetry 1998, 9, 4361.
Angew. Chem. Int. Ed. 2010, 49, 9178 –9181
ꢀ 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
9181