Direct trapping of sterically encumbered aluminum enolates

Article abstract of DOI:10.1021/ol400642y

The formation of chiral and sterically congested cyclohexanone derivatives has been achieved through a multistep sequence with one single purification step. (n-Butoxymethyl)-diethylamine was identified as a highly efficient reagent for the direct trapping of aluminum enolates. The Lewis acidic character of aluminum suffices to activate the α-aminoether to form in situ an electrophilic iminium species. In return the aluminum enolate is rendered more nucleophilic by coordination of the butoxy group and formation of an aluminate.

Full text of DOI:10.1021/ol400642y

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Direct Trapping of Sterically Encumbered
Aluminum Enolates
€
Christian Bleschke, Matthieu Tissot, Daniel Muller, and Alexandre Alexakis*
Department of Organic Chemistry, University of Geneva, Quai Ernest Ansermet 30,
CH-1211 Geneva 4, Switzerland
alexandre.alexakis@unige.ch
Received March 11, 2013
ABSTRACT
The formation of chiral and sterically congested cyclohexanone derivatives has been achieved through a multistep sequence with one single
purification step. (n-Butoxymethyl)-diethylamine was identified as a highly efficient reagent for the direct trapping of aluminum enolates. The
Lewis acidic character of aluminum suffices to activate the R-aminoether to form in situ an electrophilic iminium species. In return the aluminum
enolate is rendered more nucleophilic by coordination of the butoxy group and formation of an aluminate.
One of the most powerful and versatile reactions for the
creation of stereogenic centers in an enantioselective fash-
ion is the asymmetric conjugate addition (ACA).1 The use
of a copper catalyst together with a chiral ligand allows the
selective addition of several organometallic nucleophiles
namely zinc, aluminum, or Grignard reagents.2 The beauty
of the reaction lies, besides high enantioselectivities, in the
regioselective formation of a reactive enolate that can be
used for further transformations.3 This domino reaction
has been extensively used to build up multiply substituted
carbocycles with contiguous stereogenic centers.3,4 Several
different types of electrophiles such as aldehydes, acetals,
and alkyl halides have been utilized.4ꢀ6 However in many
cases additives are necessary. Lewis acids for example acti-
vate acetals to facilitate the trapping reaction.5 HMPA on the
other hand can be used to complex the cation to render the
enolate more nucleophilic.6 In recent years several new
methods appeared that allow the formation of enantiopure
quaternary stereogenic centers by ACA.1f This makes the
methodology even more valuable most notably for its appli-
cation in natural product synthesis. In contrast, subsequent
trapping reactions next to these quaternary all-carbon cen-
ters have been scarcely described.4 Mostly for steric reasons
the reactivity of the corresponding enolates is strongly
diminished when the adjacent substituents are bigger than
two methyl groups. The method of choice is in many cases a
preceding trapping as a silyl enol ether or O-acylation of the
enolate.7 These intermediates can be used to reform the
enolate in a separate step under different reaction conditions
(1) For reviews on asymmetric conjugate additions, see: (a) Krause,
N. Angew. Chem., Int. Ed. 1998, 110, 295. (b) Alexakis, A.; Benhaim, C.
Eur. J. Org. Chem. 2002, 3221. (c) Hayashi, T.; Yamasaki, K. Chem. Rev.
€
2003, 103, 2829. (d) Christoffers, J.; Koripelly, G.; Rosiak, A.; Rossle,
M. Synthesis 2007, 1279. (e) Harutyunyan, S. R.; den Hartog, T.; Geurts,
K.; Minnaard, A. J.; Feringa, B. L. Chem. Rev. 2008, 108, 2824. (f) Hawner,
C.; Alexakis, A. Chem. Commun. 2010, 46, 7295.
(2) (a) For reviews on copper catalyzed ACA, see: Alexakis, A.;
€
ꢀ
ꢁ
Backvall, J. E.; Krause, N.; Pamies, O.; Dieguez, M. Chem. Rev. 2008,
108, 2796. (b) Jerphagnon, T.; Pizzuti, M. G.; Minnaard, A. J.; Feringa,
B. L. Chem. Soc. Rev. 2009, 38, 1039.
(3) For reviews on domino processes, see: (a) Tietze, L. Chem. Rev.
1996, 96, 115. (b) Chapman, C. J.; Frost, C. G. Synthesis 2007, 1.
(4) For reviews on tandem transformations triggered by ACA, see:
(a) Guo, H.-C.; Ma, J.-A. Angew. Chem., Int. Ed. 2006, 45, 366.
ꢂ
ꢂ
ꢁ
(b) Galestokova, Z.; Sebesta, R. Eur. J. Org. Chem. 2012, 34, 6688.
(5) (a) Alexakis, A.; Trevitt, G. P.; Bernardinelli, G. J. Am. Chem.
Soc. 2001, 123, 4358. (b) Welker, M.; Woodward, S. Tetrahedron 2010,
66, 9954.
(7) (a) Smith, A. B., III; Nolen, E. G., Jr.; Shirai, R.; Blase, F. R.; Ohta,
M.; Chida, N.; Hartz, R. A.; Fitch, D. M.; Clark, W. M.; Sprengeler, P. A.
J. Org. Chem. 1995, 60, 7837. (b) Knopff, O.; Alexakis, A. Org. Lett. 2002,
4, 3835. (c) Alexakis, A.; March, S. J. Org. Chem. 2002, 67, 8753.
(d) Vuagnoux-d’Augustin, M.; Alexakis, A. Tetrahedron Lett. 2007, 48,
7408. (e) Welker, M.; Woodward, S. Tetrahedron 2010, 66, 9954.
(f) Mendoza, A.; Ishihara, Y.; Baran, P. S. Nat. Chem. 2012, 4, 21.
(8) (a) Vuagnoux-d’Augustin, M.; Alexakis, A. Chem.;Eur. J. 2007,
13, 9647. (b) Palais, L.; Alexakis, A. Chem.;Eur. J. 2009, 15, 10473.
(6) (a) Degrado, S. J.; Mizutani, H.; Hoveyda, A. H. J. Am. Chem.
Soc. 2001, 123, 755. (b) Kehrli, S.; Martin, D.; Rix, D.; Mauduit, M.;
Alexakis, A. Chem.;Eur. J. 2010, 16, 9890. (c) Ngoc, D. T.; Albicker,
M.; Schneider, L.; Cramer, N. Org. Biomol. Chem. 2010, 8, 1781.
ꢁ
€
ꢁ ꢁ
(d) Tissot, M.; Poggiali, D.; Henon, H.; Muller, D.; Guenee, L.;
Mauduit, M.; Alexakis, A. Chem.;Eur. J. 2012, 18, 8731.
r
10.1021/ol400642y
XXXX American Chemical Society

and above all with a countercation of choice such as lithium
which ensures high nucleophilicity.8,9 Nevertheless this step
requires additional synthetic effort and these intermediates
are often unstable and very difficult to purify.7f
compounds in our own laboratory.14 R-Aminoethers such
as 2 can be activated by Lewis acids and form in situ an
iminium salt similar to the Eschenmoser reagent. These
reagents are liquid, stable, and readily available. Regardless
of these interesting properties R-aminoethers have never
In the course of our research toward natural product
synthesis we desired to build up highly substituted cyclo-
hexanone derivatives (6) with a variety of substituents in
the R-position (Scheme 1). Yet, many of the corresponding
electrophiles were not available or not reactive enough to
be introduced as such in a trapping reaction next to a
quaternary all-carbon center. To circumvent this problem
we moved to a different strategy which implied the installa-
tion of an exocyclic methylene group by the trapping of a
small C1-electrophile (Scheme 1, bꢀc). This reactive Michael
acceptor should then allow the nucleophilic introduction of a
variety of substituents in the R-position that are otherwise not
accessible (Scheme 1, d).
ꢂ
been used for the reaction with enolates. However Sebesta
and co-workers reported recently the direct trapping of
magnesium enolates with R-amidoethers.15 Yet, stoichiom-
etric amounts of TiCl4 as an external Lewis acid had to
be added and only trapping next to tertiary centers was
described.
In recent years several methods for the metal catalyzed
ACA of trialkyl aluminum reagents were developed in our
laboratory.8,16 The distinct Lewis acidic character of these
reagents is crucial for the activation of the carbonyl group
and allows the formation of quaternary all-carbon stereo-
genic centers. We recognized that these reagents should
also be capable of activating an R-aminoether to release a
highly electrophilic iminium ion which can react with the
metal enolate (Scheme 1, b).
Scheme 1. Conception of the Multistep Sequence
Scheme 2. Conception of the New Trapping Reaction
The reagent of choice to install a methylene group next
to a carbonyl group is normally the Eschenmoser salt
(Scheme 1, b).10 Trapping of an enolate gives an amino-
methyl substituent which can be readily eliminated to furnish
the exocyclic double bond (Scheme 1, c). Nevertheless initial
experiments to trap encumbered enolates after a preceeding
ACA remained unsuccessful, and we assumed that maybe
the low solubility of the Eschenmoser salt was partially
responsible for the low reactivity.
Our search for alternatives led us to an old class of com-
pounds, R-aminoethers. These reagents were used for the
aminomethylation of Grignards reagents,11 organo zinc12
and organo copper compounds13 decades ago (Scheme 2, a).
In 1980 we used R-aminoethers in combination with
organocopper and for the first time with organoaluminum
To investigate the above-mentioned concept methyl-
cyclohexenone 3a was reacted with trimethylaluminum in
different solvents followed by quenching with the R-amino-
ether 2 (Table 1, entries 1ꢀ4). We were pleased to observe
full conversion and perfect selectivity for the transformation
of the aluminum enolate to the desired product 4a in diethyl
ether (entry 4). Also in all other tested solvents the alumi-
num species present in the mixture were able to activate the
R-aminoether and the trapped product 4a was obtained in
selectivities higher than 63%.
Next, the amount of trimethylaluminum and R-amino-
ether 2 could be decreased to 1.2 and 1.5 equiv respectively
(entries 5ꢀ6). This is remarkable because aluminum and
also zinc enolates are able to transfer the remaining alkyl
substituents on the metal in significant amounts.17 The fact
that no large excess of the electrophile is needed displays
(9) Brown, M. K.; Hoveyda, A. H. J. Am. Chem. Soc. 2008, 130,
12904.
(10) Schreiber, J.; Maag, H.; Hashimoto, N.; Eschenmoser, A.
Angew. Chem., Int. Ed. 1971, 10, 330.
(14) Germon, C.; Alexakis, A.; Normant, J. F. Tetrahedron Lett.
1980, 21, 3763.
ꢂ
k, F.; Drusan, M.; Marak, J.; Sebesta, R. J. Org. Chem.
ꢂ
ꢁ
(15) Bilcı
´
(11) (a) Robinson, G. M.; Robinson, R. J. Chem. Soc., Trans. 1923,
123, 532. (b) Theodore Stewart, A., Jr.; Charles, R. J. Am. Chem. Soc.
1955, 77, 1098. (c) Yankep, E.; Charles, G. Tetrahedron Lett. 1987, 28,
427. (d) Yankep, E.; Kapnang, H. Tetrahedron Lett. 1989, 30, 7383.
(12) Miginiac, L.; Mauze, B. Bull. Soc. Chim. Fr. 1968, 6, 2544.
(13) (a) Germon, C.; Alexakis, A.; Normant, J. F. Bull. Soc. Chim. Fr.
1984, 9ꢀ10, 377. (b) Epsztein, R.; Le Goff, N. Tetrahedron Lett. 1985,
26, 3203.
2012, 77, 760.
(16) (a) Alexakis, A.; Albrow, V.; Biswas, K.; d’Augustin, M.; Prieto,
€
O.; Woodward, S. Chem. Commun. 2005, 2843. (b) Hawner, C.; Muller,
D.; Gremaud, L.; Felouat, A.; Woodward, S.; Alexakis, A. Angew.
€
Chem., Int. Ed. 2010, 49, 7769. (c) Muller, D.; Tissot, M.; Alexakis, A.
Org. Lett. 2011, 13, 3040.
(17) Rathgeb, X.; March, S.; Alexakis, A. J. Org. Chem. 2006, 71,
5737.
B
Org. Lett., Vol. XX, No. XX, XXXX

with a selectivity of 54% (entries 9ꢀ10). Knowing that the
Lewis acidity of these reagents is much lower compared to
aluminum, 1.0 equiv of trimethylaluminum was added after
the addition of the R-aminoether. The examples before have
shown that this reagent can activate the R-aminoether and
furnish the desired product. Nevertheless still no product
was obtained from the magnesium enolate, and also in the
case of zinc no improvement of the result could be achieved
(entries 11ꢀ12). These surprising results lead us to the
assumption that not only the presence of aluminum species
but also especially the presence of an aluminum enolate is
detrimental to a highly selective reaction. It is known that
the reaction of a trialkyl aluminum compound with a Lewis
base forms a so-called aluminate where the nucleophilicity
of the substituents on the aluminum is strongly enhanced.
Cramer and co-workers have already shown that this works
also with aluminum enolates which can be activated by
stoichiometric amounts of methyllithium.6c Therefore we
think that our aluminum enolate plays a double role in the
reaction (Scheme 3). The coordination of the R-aminoether
2 to an aluminum enolate and the subsequent transfer of the
n-butoxy group to the aluminum could not only form the
desired electrophile but also activate the enolate in the same
step. The formation of a good electrophile and a good
nucleophile at the same time in close proximity could explain
the high efficiency of the transformation.
Table 1. Initial Trapping Test Reactions with the Aminoether 2a
entry
R
MR0X
solv.
2 (equiv)
GC yield (%)d
1
Me
Me
Me
Me
Me
Me
Me
i-Bu
Me
Me
Me
Me
AlMe3
AlMe3
AlMe3
AlMe3
AlMe3
AlMe3
AlMe3
AlMe3
MeMgBr
ZnEt2
THF
Tol
3
87
2
3
67
3
DCM
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
Et2O
3
63
4
3
>95
>95
>95
>95
93
0
5e
6e
7e,f
8e,f
9
3
1.5
1.5
1.5
3
10
11g
12g
3
54
MeMgBr
ZnEt2
3
0
3
47
a Reactions performed on a 0.3 mmol scale, under a N2 atmosphere
with 2 equiv of the organometallic reagent. b In the case of zinc or
Grignard reagents 5 mol % copper(II) triflate were used. In the case of
aluminum reagents a solution of copper(II) naphthenate in pentane was
used (5 mol %). c In the case of zinc or Grignard reagents 6 mol % of an
N-heterocyclic carbene were used (see Supporting Information). In the
case of aluminum reagents 6 mol % of SimplePhos ligand 7 were used
(see Table 2). d Yield after the second reaction step, determined by GC-MS
(see Supporting Information). e 1.2 equiv of the aluminum reagent was used.
f Trapping reaction: 3 h instead of 15 h. g 1.0 equiv of trimethylaluminum was
added after addition of the aminoether 2.
Scheme 3. Proposed Reaction Mechanism
the clear preference of aluminum to transfer the enolate
and not the methyl groups to the iminium electrophile. In
the following the reaction time could be decreased to 3 h
without any loss in conversion or selectivity (entry 7). The
optimized reaction conditions were then used with a new
substrate 3b that possesses a bulky isobutyl substituent in
the β-position. To our delight even the corresponding
sterically encumbered aluminum enolate reacted cleanly
to afford the desired product 4c. Full conversion and still a
high selectivity of 93% was observed after a short reaction
timeof 3 h whichunderlinesagain the high efficiencyof this
transformation (entry 8).
During our optimization the decrease of trimethylalu-
minum to 1.2 equiv drew our attention because there was
not more than 0.2 equiv of free trimethylaluminum left to
activate the R-aminoether 2. This means that the alumi-
num enolate itself is able to activate the R-aminoether,
since fullconversiontothe trapped product is observed. To
gain deeper insight into the reaction mechanism the trap-
ping reaction was repeated with zinc and magnesium enolates
which were created under similar reactions conditions.18 In
the case of the magnesium enolate no trapping product could
be detected whereas the zinc enolate gave the desired product
For the formation of the desired double bond we chose a
subsequent oxidation elimination sequence. First we tried
a one-pot procedure, but the elimination step was quite
sluggish and seemed to be hampered by the aluminum
species or amines in the mixture. Even changing the diethyl
ethertoDCM, whichisthestandardsolvent for thistype of
reaction, did not allow full conversion after several hours.
Since the oxidation and elimination in a separate reaction
needed only 30 min to reach full conversion, we decided
that a workup after the trapping reaction was the most
practical. The alternative methylation with MeI also gave
inferior results for this elimination. In the following, products
5 turned out to be unstable toward dimerization and could
not be purified or stored for a long time. Nevertheless
products 5 were obtained with just minor impurities after
the three reaction steps, which was confirmed by NMR and
GC-MS analysis. Therefore it was possible to carry out the
following conjugate addition to the exocyclic double bond
again without prior purification. All products 6 were finally
isolated by column chromatography after the last step which
constitutes an altogether four-step process with one single
purification.
(18) The generation of the enolates was tested to be complete and
with a GC yield of 80ꢀ95%. For further information, see the Supporting
Information.
Org. Lett., Vol. XX, No. XX, XXXX
C

reagents, and nucleophiles (Table 2). First different trial-
kylaluminum sources were used. The trapping reaction
worked equally well with larger alkyl groups although a
slightly longer reaction time was needed. The products
could be isolated in 27% to 41% yield, and the correspond-
ing average yields per step lay between 72% and 80%
(entries 1ꢀ3). Although the aluminum enolates play an
important role, the trapping step was not hampered by the
largersubstituentssuchasn-butyl(entry3). Inthenextstep
the Grignard reagent for the second conjugate addition
was changed. Alkyl and alkenyl nucleophiles worked
equally well as the phenyl Grignard before, and it was
even possible to install two vicinal isobutyl substituents on
the cyclohexanone 6e. The sequence delivered the highly
encumbered product in an acceptable overall yield of 24%
or 70% per step respectively (entry 5). Another valuable
compound is product 6f which can be further functiona-
lized on the protected double bond. In this case it was
crucial for the first ACA step to use a large silyl group
which can prevent the otherwise favored 1,6-addition. On
the other hand the bulky substituent is unfavorable for the
subsequent R-functionalization. Nevertheless trapping
with the R-aminoether 2 worked smoothly, and the final
product 6f was isolated in 30% yield over the four steps
(entry 6). Notably, high enantioselectivities between 96%
and 98% ee were reached for all products due to the high
efficiency of the SimplePhos ligand 7. The diastereomeric
ratios lay between 1:1 and 7:3 and increased with the
increasing size of the substituents. Finally the intermediate
magnesium enolate was reacted in a fifth step with acetic
anhydride, and compound 6g was obtained in a 35% yield
or an 81% average yield respectively. This extension of the
methodology shows already how to avoid the problem of
diastereoselectivity.
Table 2. Initial Trapping Test Reactions with the Aminoether 2a
In conclusion we developed a four-step sequence involv-
ing an unprecedented electrophilic trapping reaction with
an R-aminoether. This one-pot functionalization of alumi-
num enolates works with just 1.5 equiv of the reagent, it
does not need any additional Lewis acid, and it still works
efficiently next to quaternary stereogenic centers. After the
complete sequence highly functionalized cyclohexanone
derivatives 6 were obtained in good overall yields.
Acknowledgment. This research was supported by the
Swiss National Research Foundation (Grant No. 200020_
144344). Chiral amines were kindly provided by BASF.
a Reactions performed on a 1.0 mmol scale, under a N2 atmosphere.
b CuNaph = solution of copper(II) naphthenate in pentane. c Absolute
configuration according to published results (see Supporting Information).
d Isolated yield over four steps. Average yield per step in parentheses.
e Determined from a small sample after the asymmetric conjugate addition
by chiral GC. f Reaction time for the ACA was 6 h. g Ac2O was added after
the Grignard addition, 1 h at ꢀ78 °C.
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.
After having established the reaction sequence we in-
vestigated different combinations of substrates, aluminum
The authors declare no competing financial interest.
D
Org. Lett., Vol. XX, No. XX, XXXX