Copper-catalyzed asymmetric 1,4-addition of alkenyl alanes to N-substituted-2-3-dehydro-4-piperidones

Article abstract of DOI:10.1021/ol3004436

Readily available vinyl alanes are used in the Cu-catalyzed asymmetric conjugate addition reaction to N-substituted-2-3-dehydro-4-piperidones. The enhanced reactivity of recently developed and easily prepared phosphine amine ligands in combination with inexpensive Cu(II)naphtenate (CuNp) allows the introduction of a great variety of alkenyl, alkyl, and aryl aluminums in high enantioselectivity.

Full text of DOI:10.1021/ol3004436

ORGANIC
LETTERS
2012
Vol. 14, No. 7
1842–1845
Copper-Catalyzed Asymmetric
1,4-Addition of Alkenyl Alanes to
N-Substituted-2-3-dehydro-4-piperidones
€
Daniel Muller and Alexandre Alexakis*
Department of Organic Chemistry, University of Geneva 30, quai Ernest Ansermet
CH-1211 Geneva 4, Switzerland
alexandre.alexakis@unige.ch
Received February 23, 2012
ABSTRACT
Readily available vinyl alanes are used in the Cu-catalyzed asymmetric conjugate addition reaction to N-substituted-2-3-dehydro-4-piperidones.
The enhanced reactivity of recently developed and easily prepared phosphine amine ligands in combination with inexpensive Cu(II)naphtenate
(CuNp) allows the introduction of a great variety of alkenyl, alkyl, and aryl aluminums in high enantioselectivity.
Since the pioneering work of Hayashi in 2004 related
to the asymmetric conjugate addition (ACA) of arylzinc
reagents to 2,3-dihydro-4-piperidones, these kinds of com-
pounds have become benchmark substrates for ACA.1
Various aryl and alkyl nucleophiles such as arylzinc,1
arylboronic acids,2 arylboroxines,3 tetraarylborates,4 aryl-
[2-(hydroxymethyl)-phenyl]dimethylsilanes,5 arylsiloxanes,6
aryltitanium tri-isopropoxides, and alkylzinc7 and alkylalu-
minum reagents8 were successfully introduced affording
highly enantioenriched products. However, all reported
attempts to introduce the highly functionalizable vinyl
group failed.1,2b This and the fact that the corresponding
piperidone products represent valuable building blocks in
the synthesis of pharmaceutically active piperidines caught
our attention.9 Over the past four years, we showed
particular interest in the introduction of the alkenyl group
to various Michael acceptors. This includes β-substituted
cyclic enones, which upon addition ofa carbonnucleophile
form an all carbon quaternary stereogenic center.10 There-
fore, we reasoned that two recently developed reac-
tion protocols might solve the longstanding challenge of
alkenyl addition to 2,3-dihydro-4-piperidones such as 1a
(Scheme 1).10e,f Herein, we present an efficient set of
protocols for catalytic ACA reactions for a wide range
of alkenyl aluminum reagents with unactivated sub-
strates 1a and 1b. The requisite aluminum-based reagents
can be prepared efficiently via hydroalumination or
(1) Shintani, R.; Tokunaga, N.; Doi, H.; Hayashi, T. J. Am. Chem.
Soc. 2004, 126, 6240–6241.
(2) (a) Zhang, T.; Shi, M. Chem.;Eur. J. 2008, 14, 3759–3764. (b)
Xu, Q.; Zhang, R.; Zhang, T.; Shi, M. J. Org. Chem. 2010, 75, 3935–
3937. (c) Thaler, T.; Guo, L.-N.; Steib, A. K.; Raducan, M.; Karaghios-
off, K.; Mayer, P.; Knochel, P. Org. Lett. 2011, 13, 3182–3185.
(3) Jagt, R. B. C.; de Vries, J. G.; Feringa, B. L.; Minnaard, A. J. Org.
Lett. 2005, 7, 2433–2435.
(4) Zhang, X.; Chen, J.; Han, F.; Cun, L.; Liao, J. Eur. J. Org. Chem.
2011, 1443–1446.
(5) Nakao, Y.; Chen, J.; Imanaka, H.; Hiyama, T.; Ichikawa, Y.; Duan,
W.; Shintani, R.; Hayashi, T. J. Am. Chem. Soc. 2007, 29, 9137–9143.
(6) Gini, F.; Hessen, B.; Feringa, B. L.; Minnaard, A. J. Chem.
Commun. 2007, 710–712.
(7) Sebesta, R.; Pizzuti, M. G.; Boersma, A. J.; Minnaard, A. J.;
Feringa, B. L. Chem. Commun. 2005, 1711–1713. (b) Endo, K.; Ogawa,
M.; Shibata, T. Angew. Chem. 2010, 122, 2460–2463. Angew Chem. Int.
Ed. 2010, 49, 2410–2413.
(8) Pizzuti, M. G.; Minnaard, A. J.; Feringa, B. L. Org. Biomol.
Chem. 2008, 3464–3466.
(9) Weintraub, P. M.; Sabol, J. S.; Kane, J. M.; Borcherding, D. R.
Tetrahedron 2003, 59, 2953–2989 and references therein.
(10) (a)Alexakis, A.;Albrow, V.;Biswas, K.;d’Augustin, M.; Prietob, O.;
Woodward, S. Chem. Commun. 2005, 2843–2845. (b) Vuagnoux-d’Augustin,
M.; Alexakis, A. Chem.;Eur. J. 2007, 13, 9647–9662. (c) Hawner, C.;
Li, K.; Cirriez, V.; Alexakis, A. Angew. Chem., Int. Ed. 2008, 120, 8334–
8337. Angew. Chem., Int. Ed. 2008, 47, 8211–8214. (d) Palais, L.;
€
Alexakis, A. Chem.;Eur. J. 2009, 15, 10473–10485. (e) Muller, D.;
Hawner, C.; Tissot, M.; Palais, L.; Alexakis, A. Synlett 2010, 1694–
€
1698. (f) Muller, D.; Tissot, M.; Alexakis, A. Org. Lett. 2011, 13,
3040–3043.
r
10.1021/ol3004436
Published on Web 03/15/2012
2012 American Chemical Society

Table 1. Initial Examination of Various Chiral Cu-Complexesa
Scheme 1. Reported and This Work on the ACA to 2,3-Dihydro-
4-piperidones
transmetalation reactions. Reactions, promoted in the
presence of a chiral monodentate phosphine amine
copper complex, afford the products with high enan-
tioselectivity (up to 97%) whatever the nature of the
nucleophile (aryl, alkenyl, and alkyl).
We began our investigations by examining the ability of
highly efficient chiral copper complexes, previously reported
in the literature in the context of Cu-catalyzed ACA with alu-
minum organyls, to promote the ACA of alkenyl aluminum
reagents to 2À3-dehydydro-4-piperidones.8,10b,10c,10dÀ10f,11
We selected 1a as a representative substrate as it can be
easily prepared in a simple one step procedure from com-
mercially available 4-methoxypyridine.12
convb
[%]b
eec
entry
L
Cu salt
solvent
[%]
1d
2
L1
L1
L1
L2
L2
L2
L3
CuNpe;20
CuNpe;10
CuNpe;5
CuNpe;10
CuNpe;10
CuNpe;10
CuTC; 10
CuTC; 10
Cu(OTf)2; 10
CuTC; 10
Et2O
Et2O
Et2O
toluene
toluene
Et2O
Et2O
THF
21
n.d.
91
100
96
3
82
4d
5e
6
94
71
As shown in Table 1 (entries 1 and 2) the recently
reported coactivation with AlMe3 as a Lewis acid was
absolutely critical for the outcome of the reaction.10e Only
electron rich phosphine amine ligand L2 was able to
achieve almost full conversion without additional activa-
tion with AlMe3 (entry 4). Interestingly, the presence of
ethereal solvent for Me3Al coactivation is absolutely vital
as its absence results in mere methyl transfer (entry 5).
The results depicted in Table 1 underline the very high
reactivity of phosphine amine ligands in comparison to
phosphoramidite, phospine, or ferrocene based ligands.
Cu(II)naphtenate (CuNp), the most inexpensive organic
copper salt, in combination with L1 achieved the highest
enantioselectivity (91%). Although the ligand loading of
11 mol % could not be further decreased without loss of
enantioselectivity (Table 1, entry 3), this represents the
lowest amount of ligand loading for the Cu-catalyzed
ACA employing unprotected vinyl alanes made via
hydroalumination.13 Interestingly, no 1,2-addition pro-
duct was formed which represented a common byproduct
for the ACA of alkenyl alanes made via hydroalumination
to β-substituted cyclic enones.10b,f We rationalize this ob-
servation by the fact that substrate 1a does not contain a
β-substituent and therefore no steric preference for the 1,2-
addition is present. It is well-known that 1a is less reactive in
ACA reactions compared to simple cyclohexenones.1,7a
However, we were surprised to see that its reactivity is
66f
100
78
n.d.
60
7
26
8
(R)-binap
62
27
9g
10
L4
L5
Et2O
Et2O
31
n.d.h
n.d.
n.d.
a Reactions performed under an Ar atmosphere on a 0.30 mmol scale.
b Determined by GC-MS. c Determined by chiral Supercritical Fluid
Chromatography. d No AlMe3 added. e Solution of Cu(II)naphtenate
(CuNp) in pentane. f Only methyl transfer observed. g 20 mol % of ligand
used.h Complex reaction mixture.
even lower compared to 3-methylcyclohex-2-enone when
exposed to our standard protocol.14
With the optimized reaction conditions in hand we
envisaged the introduction of a great variety of alkenyl
groups. The simplest way to generate an alkenylalane is
hydroalumination of a terminal alkyne (Table 2, entry 1).
Usually such reactions proceed cleanly by syn-addition of
HÀAl(i-Bu)2 to the triple bond and only small amounts
of Al-acetylides are formed.15 A directing group such as
tert-butoxy in close proximity to the alkyne changes the
mechanism, and anti-addition is observed instead (Table 2,
entry 2).16 Conjugation of the triple bond with an aromatic
system, double or triple bonds, or an electron-withdrawing
substituent greatly increases the acidity of the acetylenic
hydrogen. This allows the formation of significant amounts
of metalation, instead of a hydroalumination product; e.g.
(11) (a) d’Augustin, M.; Palais, L.; Alexakis, A. Angew. Chem. 2005,
117, 1400–1402. Angew. Chem., Int. Ed. 2005, 44, 1376–1378. (b)
Vuagnoux-d’Augustin, M.; Kehrli, S.; Alexakis, A. Synlett 2007, 13,
(14) Without activation with AlMe3, L1 gave 98% conversion of
the addition of (E)-hex-1-enyldiisobutylaluminum to 3-methylcy-
clohex-2-enone at À30 °C whereas the addition of the same reagent
to substrate 1a at À10 °C gave only 21% conversion under the same
conditions.
(15) Zweifel, G.; Miller, J. A. Org. React. (N.Y.) 1984, 32, 375.
(16) Alexakis, A.; Duffault, J. M. Tetrahedron Lett. 1988, 29, 6243–
6246.
€
2057–2060. (c) Tissot, M.; Muller, D.; Belot, S.; Alexakis, A. Org. Lett.
2010, 12, 2770–2773. (d) Gremaud, L.; Alexakis, A. Angew. Chem. 2012,
124, 818–821. Angew. Chem., Int. Ed. 2012, 51, 794–797.
(12) See Supporting Information for details.
(13) Previous reports employed between 22À30 mol % of ligand
(cf. ref 10b, 10f).
Org. Lett., Vol. 14, No. 7, 2012
1843

Table 2. Representative Protocols for the Generation of Various
Alkenyl Alanes
Table 3. Cu-Phosphine Amine Catalyzed ACA of Vinylalumi-
nums to Michael Acceptor 1aa
a Contains 6% of metalation and 4% of bis-hydroalumination
product.15 b Directing groups such as OtBu afford purely the syn
hydroalumination product.16 c The β-hydroalumination product and
the metalation product are formed in <2%.17 d 7% of the R-hydro-
alumination product is formed; the metalation product is formed in
<2%.17 e Control of the exact ratio of tBuLi to Me2AlCl is of great
importance, as an excess of either tBuLi or Me2AlCl can lead to a
decrease in enantioselectivity.10e
a Reactions performed under Ar atmosphere on a 0.30 mmol scale;
conversion = 100% determined by GC-MS or TLC. b The number
indicated refers to the protocol of alane generation (Table 2) which was
used. c Yields of isolated vinyl addition products. d Determined by chiral
Supercritical Fluid Chromatography. e 22 mol % of L1 and 20 mol % of
Cu(II)naphtenate (CuNp) were used. f 3 equiv of alane added. g Addition of
0.2 mL of THF. h No Cu(II)naphtenate (CuNp) added, conversion <100%.
29% of Al-acetylide is obtained for the hydroalumination
of phenylacetylene.15 Therefore, Hoveyda et al. developed
a protocol for the Ni-catalyzed hydroalumination which
suppresses this side reaction and even allows, by choice
of ligand, a selective R or β hydroalumination (Table 2,
entries 3À4).17 In contrast, we reported the synthesis of
conjugated alkenyl alanes from the corresponding bro-
mides by a lithiumÀbromine exchange with tBuLi, followed
by transmetalation with Me2AlCl (Table 2, entry 5).10e This
approach is advantageous for the synthesis of conjugated
β-alkenyl alanes because compared to the Ni-catalyzed
hydroalumination the obtained alanes do not suffer from
contamination by R-alkenyl alanes.
Under optimized conditions alkyl substituted β-vinyl
aluminums (Table 3, entries 1À7) undergo ACA with good
yields and high enantioselectivity (84À96% ee). This is
particularlyinteresting given thatCu-catalyzedACAusing
dialkylzinc reagents gave only low yields and enantioselec-
tivity when employing longer alkyl chains such as nBu.7a
Moreover, the length of the alkyl chain did not have an
influence on the outcome of the reaction (entries 1À3).
This also holds true for functionalized alkynes (entries
6À7) which are very useful for further transformations
(vide infra). Although the increased steric bulk of the alkyl
substituent led to a decrease in enantioselectivity, high
levels of stereoinduction are still maintained (entries 4À5).
To compare the influence of the preparation of the alkenyl
aluminums, products 2hÀj were made from alanes pre-
pared according to protocols 4 and 5 (Table 2). Both alanes
usually gave the same levels of enantioselectivity (Table 3,
entries 8À13, with the exception of 2j), and in all cases
yields were greatly improved when the alane was prepared
according to protocol 5 (Table 2). We rationalize this
observation by the fact that the Ni-catalyzed hydroalumi-
nation affords alkenyl alanes which are less pure (cf.
footnote d, Table 2) than alanes made from vinyl bro-
mide by a lithiumÀbromine exchange, transmetalation
sequence. To exclude the possibility that the ACA reaction
in the presence of nickel salts is catalyzed by a chiral nickel
complex, generated through ligand transfer from copper to
nickel, we performed the reaction in the absence of copper
salt. Since nickel was reported to catalyze 1,4-addition
reactions with alanes it came to no surprise that the
reaction still proceeded, although less efficiently (Table 3,
entry 14).18 The fact that in the absence of copper salt no
enantioselectivity was observed emphasizes the efficiency
of the chiral copper catalyst which in the presence of this
racemic background reaction was still able to afford
products with high optical purity (97% ee for 2h). In
contrast to the Ni-catalyzed β-hydroalumination the R-
hydroaluminationproceedscleanly(cf.footnoted,Table2)
and the products 2k and 2l were cleanly formed in high
optical purity. It is noteworthy that product 2k represents
(18) First reports of nickel catalyzed conjugate addition; using alkyl
aluminums: (a) Jeffery, A. E.; Meisters, A.; Mole, T. Aust. J. Chem. 1974,
27, 2569–2576. (b) Aryl aluminums: Westermann, J.; Imbery, U.;
Nguyen, A. T.; Nickisch, K. Eur. J. Inorg. Chem. 1998, 295–298. (c)
Alkyne aluminums: Hansen, R. T.; Carr, D. B.; Schwartz, J. J. Am.
Chem. Soc. 1978, 100, 2244–2245.
(17) Gao, F.; Hoveyda, A. H. J. Am. Chem. Soc. 2010, 132, 10961–10963.
1844
Org. Lett., Vol. 14, No. 7, 2012

tions were also applicable for other substrates than 1a we
submitted substrate 1b containing a Boc group instead of a
Cbz group to the ACA reaction. We were pleased to see
that the high enantioselectivity of 97% which was pre-
viously achieved for substrate 1a (Table 3, entry 9) was
maintained. The yield and enantioselectivity were slightly
higher when hydroalumination protocol 5 was used in-
stead of 4 (Table 4, entries 3À4). Having introduced a large
variety of vinyl nucleophiles to substrate 1a and 1b we were
interested if the methodology was general and would also
allow for the introduction of alkyl and aryl nucleophiles.
To our delight nonoptimized reaction conditions afforded
products 2q, 2r, and 2s in high enantioselectivities (>90%)
and good yields. The absolute configuration of addition
products containing an alkenyl group was assigned R in
analogy with the general bottom face attack observed for
products 2q, 2r, and 2s.12
Table 4. Extension of the Methodologya
a Reactions performed under Ar atmosphere on a 0.30 mmol scale;
conversion = 100% determined by GC-MS and TLC. b The number
indicated refers to the protocol of alane generation (Table 2) which was
used. c Yields of isolated and vinyl addition products. d Determined by
chiral Supercritical Fluid Chromatography. e Vinylalane prepared from
vinylmagnesium bromide (see Supporting Information for details). f 22
mol % of L1 and 20 mol % of Kubas salt [Cu(CH3CN)4]BF4 were used; 3
equiv of alane added; Et2O replaced by THF. g Aryl alane prepared from
aryl bromide (see Supporting Information for details). h Commercially
available, only addition of 2.0 equiv of Me3Al. i Commercially available,
only addition of 2.0 equiv of Et3Al.
In order to show the high synthetic potential of the
developed methodology we submitted chlorine containing
product 2f to standard hydrogenation conditions. In one
pot we carried out deprotection of the Cbz group, hydro-
genation of the double bond, and an intramolecular SN2
reaction (Scheme 2).23
Scheme 2. One Pot Synthesis of Bicyclic Amine 3a
the first example of an ACA employing an alkyl substituted
R-alkenyl alane made via Ni-catalyzed hydroalumination;
the same reaction failed for 3-methylcyclohex-2-enone as
the substrate.19 Addition of isopropenylalane to substrate
1a afforded product 2m in high yield and enantioselectivity.
The simple vinyl group is certainly one of the most
valuable substituents as it can undergo a wide range of
further transformations such as the cross-metathesis
reaction.20 This motivated us to investigate the use of vinyl
alane prepared in situ from inexpensive and commercially
available vinyl magnesium bromide. Although yield and
enantiomeric excess were moderate (Table 4, entry 1), 2n
represents the first example of a vinyl aluminum reagent
employed for metal catalyzed ACA. The preparation of 2o
presented a challenge as standard conditions afforded
the product (25%) as a mixture with the Me-transfer
(23%) and i-Bu-transfer products (52%). Replacement
of Et2O by THF suppressed methyl as well as i-Bu transfer
to <5%, but the enantioselectivity was very low (14%
ee).21 Change of the copper salt to Kubas salt [Cu-
(CH3CN)4]BF4 increased the enantioselectivity of the
reaction (49%) while maintaining low levels of methyl
and i-Bu transfer.22 To check if the experimental condi-
In conclusion, we showed that the recently developed class
of chiral phosphine amines,11a prepared in one or two steps
from commercially available amines, achieved far superior
results compared to all other ligands tested.11a,10d,10f
In addition, we were able to use alkenyl alanes generated
by hydroalumination from unprotected alkynes or by
a simple lithiumÀbromine exchange, transmetalation
sequence from vinyl bromides. The developed methodology
proved to be particularly efficient for the ACA of aluminum
organyls to unactivated substrates, and further applications
toward other challenging substrates are among the objec-
tives being pursued in our laboratories.
Acknowledgment. The authors thank the Swiss National
Research Foundation (Grant No. 200020-126663) and COST
action D40 (SER Contract No. C07.0097) for financial sup-
port, as well as BASF for the generous gift of chiral amines.
€
(19) Muller, D.; Alexakis, A. Unpublished results.
(20) For an extensive article concerning cross metathesis reactions,
see: Chatterjee, C.; Choi, T.; Sanders, D. P.; Grubbs, R. H. J. Am. Chem.
Soc. 2003, 125, 11360–11370.
(21) The addition of THF was necessary also for the products 2f and
2j: for 2f, to suppress about 20% of methyl and i-Bu transfer and, for 2j,
to ensure total conversion.
Supporting Information Available. Experimental pro-
cedures, NMR spectra and chiral separations for all
compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
(22) We reported the use of Kubas salt [Cu(CH3CN)4]BF4 in the
context of ACA employing vinyl alanes before; cf. ref 10b.
(23) Under similar conditions product 2g did not cyclize.
The authors declare the following competing financialinterest(s): The
authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 7, 2012
1845