Creation of highly congested quaternary centers via Cu-catalyzed conjugate addition of alkenyl alanates to β-substituted cyclic enones

Article abstract of DOI:10.1021/ol400378v

Easily prepared alkenylalanates proved to be excellent nucleophiles for the creation of highly congested quaternary centers via copper-catalyzed conjugate addition. In addition, functionalized cis-decaline systems can now be prepared in a simple two-step sequence involving Cu-catalyzed conjugate addition with functionalized alkenylalanates.

Full text of DOI:10.1021/ol400378v

ORGANIC
LETTERS
XXXX
Vol. XX, No. XX
000–000
Creation of Highly Congested Quaternary
Centers via Cu-catalyzed Conjugate
Addition of Alkenyl Alanates to
β‑Substituted Cyclic Enones
€
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 7, 2013
ABSTRACT
Easily prepared alkenylalanates proved to be excellent nucleophiles for the creation of highly congested quaternary centers via copper-catalyzed
conjugate addition. In addition, functionalized cis-decaline systems can now be prepared in a simple two-step sequence involving Cu-catalyzed
conjugate addition with functionalized alkenylalanates.
In the 1970s, Weiss and co-workers exploited lithium
(E)-1-alkenylalanates for the synthesis of prostaglandines
and congeners (Scheme 1).1 They showed that lithium
alkenylalanates undergo conjugate addition to cyclopen-
Scheme 1. Non-catalyzed Conjugate Addition of Alkenylala-
nates to Yield Prostaglandin Derivatives
tenones in the absence of a metal catalyst at room tem-
perature to give the addition products in poor to good
yields. Interestingly, only the alkenyl ligand on aluminum is
specifically transferred in a 1,4-manner despite the presence
of three alkyl ligands. The requisite alkenylalanates were
easily prepared either by hydroalumination of alkynes and
Schwartz stated in 1980:3 “..the use of these aluminum
species for conjugate addition offers no advantage over the
subsequent treatment with methyl lithium or by reaction of
trimethylaluminium with the (E)-1-alkenyllithiums.
With the exception of the MAD and MAT system,
alkenylcuprate-based routes since both routes rely on the
availability of the corresponding alkenyllithium reagent”.
introduced by Yamamoto in 1987,2 alkenylalanates were
no longer used for 1,4-addition reactions because as
During our work on the copper-catalyzed asymmetric
conjugate addition using alkenylaluminums to β-substi-
(1) (a) Weiss, M. J.; Bernady, K. F. Tetrahedron Lett. 1972, 40, 4083–
tuted cyclic enones we were confronted with the problem
4086. (f) Floyd, M. B.; Weiss, M. J. Prostaglandins 1973, 3, 921–924. (b)
that encumbered dimethylalkenylaluminums led to methyl
Bernady, K. F.; Poletto, J. F.; Weiss, M. J. Tetrahedron Lett. 1975, 48,
765–768. (c) Skotnicki, J. S.; Schaub, R. E.; Bernady, K. F.; Siuta, G. J.;
Poletto, J. F.; Weiss, M. J. J. Med. Chem. 1977, 20, 1551–1557. (d) Bernady,
K. F.; Floyd, M. B.; Poletto, J. F.; Weiss, A. H. J. Org. Chem. 1979, 44,
1438–1447. (e) Floyd, M. B.; Weiss, A. H. J. Org. Chem. 1979, 44, 71–75.
(2) Maruoka, K.; Nonoshita, K.; Yamamoto, H. Tetrahedron Lett.
1987, 28, 5723–5726.
instead of alkenyl-transfer (Scheme 2).4 Moreover, sterically
€
(4) (a) Muller, D.; Hawner, C.; Tissot, M.; Palais, L.; Alexakis, A.
€
Synlett 2010, 1694–1698. (b) Muller, D.; Tissot, M.; Alexakis, A. Org.
€
(3) Schwartz, J.; Loots, M. J.; Kosugi, H. J. Am. Chem. Soc. 1980,
402, 1333–1340.
Lett. 2011, 13, 3040–3043. (c) Muller, D.; Alexakis, A. Org. Lett. 2012,
14, 1842–1845.
r
10.1021/ol400378v
XXXX American Chemical Society

congested substrates gave low conversion and afforded
only traces of the desired product. In both cases, we were
unable to isolate and characterize the desired products and
hence focused on the racemic syntheses of these β-disub-
stituted cyclic enones.
with Me3Al and to employ them in situ for the Cu-cata-
lyzed conjugate addition (Scheme 3 and Table 1). Many
alkenylbromides, such as 4, are commercially available or
can be easily prepared.11
Scheme 3. Generation of Mixed Alkenylalanate A1
Scheme 2. Formation of Highly Encumbered Quaternary
Stereogenic Centers: A Challenging 1,4-Addition
Not surprisingly, when the reaction was performed un-
der the conditions reported by Weiss (No metal catalyst,
room temperature) the desired product 1b was observed
(Table 1 entry 1). However, GC-MS analysis showed
about 19% of side products, which could not be character-
ized. When we carried out the reaction at ꢀ10 °C in the
presence of 10 mol % of Cu(II)naphthenate or CuBr SMe2,
3
clean reaction occurred and 1b was observed as a single
product (entries 2ꢀ3). It is worthy to mention that this
represents the first metal-catalyzed conjugate addition
employing alkenyl alanates.12 Why alkenyl alanates react
in such a clean manner, in Cu-catalyzed conjugate addi-
tion, might be rationalized by the fact that after every
transmetalation of an alkenyl group to copper, one mole-
cule of Lewis acidic trialkylaluminum is released which
then activates the substrate by complexation to the carbo-
nyl oxygen. Hence, alkenylalanates are not only highly
nucleophilic reagents, comparable to Grignard reagents,
but are also a source of catalytic amounts of trialkylalu-
minum reagent which acts as a strong Lewis acid. The
double activating ability of alkenyalanates also explains
why conjugate additions with alanates do not require
additional Lewis acids such as BF3 or TMS-Cl.
Only a few conjugate additions of alkenylcuprate re-
agents to sterically congested β-substituted enones were
reported in the literature and most of them use unhindered
vinyl-cuprate reagents made from commercially available
vinylmagnesium bromide.5 Protocols using other alkenyl
cuprates often note the occurrence of low conversion,6
isomerization of configuration during the preparation of
the alkenyl Grignard reagent,7 the requirement of a Lewis
acidfor efficient 1,4-addition8 and the use of stoichiometric
amounts of toxic CuCN9 or malodorous P(n-Bu)3.10 Hence,
we were considering alternative nucleophiles for the con-
struction of highly congested quaternary centers via Cu-
catalyzed conjugate addition reactions. A first indication
that alkenyl alanates might be appropriate nucleophiles
came from the observation that alkenylalanes generated
from the corresponding alkenyl bromides via BrꢀLi ex-
change and reaction with Me2AlCl reacted better when a
slight excess of t-BuLi was used. Therefore, we decided to
synthesize alkenyl alanates by the reaction of alkenyllithiums
Table 1. Optimization of Reaction Conditionsa
(5) (a) Lipshutz, B. H.; Sengupta, S. Org. React. 1992, 41, 135–631.
(6) Barnes, D. M.; Bhagavatula, L.; DeMattei, J.; Gupta, A.; Hill,
D. R.; Manna, S.; McLaughlin, M. A.; Nichols, P.; Premchandran, R.;
Rasmussen, M. W.; Tian, Z. P.; Wittenberger, S. J. Tetrahedron:
Asymmetry 2003, 14, 3541–3551.
(7) Mechin, B.; Naulet, N. J. Organomet. Chem. 1972, 39, 229–236.
(8) (a) Horiguchi, Y.; Matsuzawa, S.; Nakamura, E.; Kuwajima, I.
Tetrahdron Lett. 1986, 27, 4025–4028. (b) Piers, E.; Yeung, B. W. A.
J. Org. Chem. 1984, 49, 4567–4569.
(9) Representative examples: (a) Lipshutz, B. H.; Ellsworth, E. L. J. Am.
Chem. Soc. 1990, 112, 7440–7441. (b) Delaloge, F.; Prunet, J.; Pancrazi, A.;
Lallemand, J. Tetrahdron Lett. 1997, 38, 237–240. (c) Aradjo, M. A.;
Barrientos-Astigarraga, R. E.; Ellensohn, R. M.; Comasseto, J. V. Tetra-
hdron Lett. 1999, 40, 5115–5118.
(10) Suzuki, M.; Suzuki, T.; Kawagishi, T.; Noyori, R. Tetrahdron
Lett. 1980, 21, 1247–1250.
entry
Cu-salt
T (°C)
convb [%]
1
þ20
ꢀ10
ꢀ10
ꢀ10
ꢀ10
ꢀ10
ꢀ10
92
>98
>98
>98
30
2
Cu(II)naphthenate
CuBr.Me2S
3
4c
5d
6e
7f
Cu(II)naphthenate
Cu(II)naphthenate
Cu(II)naphthenate
Cu(II)naphthenate
95
>98
a Reactions performed under Ar atmosphere on a 0.30 mmol scale.
b Determined by GC-MS. c Reaction performed with 1.5 equiv of Li-Alanate
A1. d Reaction time was 2 h. e Reaction performed in THF. f Preformation of
chiral complex with L1 (11 mol %); product was obtained as a racemic mixture.
(11) For some examples for the synthesis of various alkenyl bromides
see: (a) Al Dulayymi, J. R.; Baird, M. S.; Simpson, M. J.; Nyman, S.
Tetrahedron 1996, 52, 12509–12520. (b) Abbas, S.; Hayes, C. J.; Worden,
S. Tetrahedron Lett. 2000, 41, 3215–3219. (c) Wolfe, J. P.; Yang, Q.; Hay,
M. B.; Ney, J. E. Adv. Synth. Catal. 2005, 347, 1614–1620.
B
Org. Lett., Vol. XX, No. XX, XXXX

Further optimizationsofthe reaction conditionsshowed
that decreasing the amount of alkenyl Alanate 1b from 2.0
to 1.5 equivalents still afforded full conversion (entry 4).
Nevertheless, we continued to use 2.0 equivalents as for
other substrates such as 9 the conversion was low using less
than 2.0 equivalents (Table 3, entry 4). For the formation
of product 1b we noted that the conversion was only 30%
after 2 h, which shows that the reaction proceeds relatively
slowly (entry 7). No stereocontrol was observed in the
presence of a chiral copper complex, which typically
afforded high enantioselectivities for the conjugate addi-
tion with alanes (entry 7).13 Next, we envisaged the con-
jugate addition of a number of various bulky alkenyl
alanates whose dimethylalkenyl aluminum congeners af-
forded high levels of undesired Me-transfer (Table 2;
compare to Scheme 2). The exceptional preference for
the alkenyl-transfer in alkenylalanates is best illustrated
for the conjugate addition of highly encumbered Alanate
A2 (entry 1), which cleanly affords the desired product in
good isolated yields and without occurrence of Me-trans-
fer. The feasibility and practicality of the procedure needs
to be emphasized. For instance (Z)-bromopropene was
transformed by bromineꢀlithium exchange into the corre-
sponding (Z)-alkenyllithium with retention of configura-
tion. Subsequent treatment of the alkenyllithium with a
commercial solution of Me3Al in hexanes afforded Alanate
A2, which is configurationally stable at room temperature.
Then the suspension of the Alanate and lithium salts was
cooled to ꢀ10 °C and a hydrocarbon solution of Cu(II)-
naphthenate was added followed by addition of the neat
enone. Hence, all the required reagents are commercially
available solutions and no predrying of copper-salts or
the formation of a cyanocuprate reagent as the case for
00traditional00 cuprate chemistry is required.
Table 2. Cu-catalyzed Conjugate Addition with Various Alke-
nyl Alanatesa
a Reactions performed under Ar atmosphere on a 0.9 mmol scale.
b Estimated by GC-MS. c Isolated yield. d Performed on a 2.0 mmol scale.
Table 3. Copper-catalyzed Conjugate Addition of Alkenylalanate
A6 to β-Substituted Cyclic Enonesa
Though conjugate additions with vinylcuprate reagent
toafford congestedquaternary centers are well-known, the
high synthetic value of the CHdCH2 unit was reason
enough to test if lithium vinyl Alanate prepared from in-
expensive vinyl bromide would cleanly undergo 1,4-addition.
Indeed, the reaction workedperfectlywelland affordedthe
product in high yield (entry 4). In contrast, when vinyl
Grignard reagent was used for the preparation of Alanate A5
the yield was very low (entry 5). We reason that magnesium
salts or the presence of THF, which stems from the Grignard
reagent, might be the origin of the sluggish reaction.14
On the basis of the previous results with sterically en-
cumbered alkenyl nucleophiles we were confident that
alkenyl alanates would also efficiently undergo conjugate
addition to sterically congested enones bearing alkyl or
aryl-substituents in the β-position (Table 3).
a Reactions performed under Ar atmosphere on a 0.90 mmol scale.
b Estimated by GC-MS. c Isolated yield. d Reaction carried out on
0.45 mmol; 11% of 1,2-addition; estimated by GC-MS.
Remarkably, alkenylalanate A6 cleanly afforded the de-
sired addition-products in good yields for all encumbered
cyclohexenone substrates (entries 1ꢀ3). Only phenyl-
substituted substrate 3 gave lower yields, which are due
to the smaller scale (0.45 mmol) and 11% of 1,2-addition
product. We were delighted that even isophorone 6, a
notoriously unreactive Michael acceptor gave full conversion
with encumbered Alanate A6. Cyclopentenone substrate 7
was causing troubles in alane chemistry as the presence of
only a small excess of Me2AlCl for the generation of the
requisite dimethyl-2-propenylaluminum led to complete
(12) Copper and nickel catalyzed conjugate additions using LiAlMe4
was reported by Ashby: Ashby, E. C.; Heinsohn, G. J. Org. Chem. 1974,
39, 3297–3299.
(13) For an overview of Cu-catalyzed asymmetric conjugate addi-
€
tions with alkenylaluminums see: Muller, D.; Alexakis, A. Chem.
Commun. 2012, 48, 12037–12049.
(14) It is well-known that THF coordinates the Lewis acids such as
Liþ or AlR3 and hence slows down conjugate addition reactions.
Org. Lett., Vol. XX, No. XX, XXXX
C

Table 4. Use of Sterically Congested Alkenylalanates Made via
Ni-catalyzed Hydroaluminationa
Scheme 4. Generation of Mixed alkenyl Alanates A7 and A8
entry
Al-reagent
1e:1fb
conv.c [%]
yieldd[%]
1
9
n.d.e
n.d.e
n.d.e
n.d.e
91:9
87:13
98:2
100
100
96
n.i.f,g
n.i.f,g
n.i.f
n.i.f
67
2h
3
9
A7
A8
A7
A8
A8i
degradation of the substrate, which was probably due to
Lewis acid induced aldol-condensation reactions. In contrast
strongly basic Alanate reagent A6 favored the clean forma-
tion of addition-product 7a. However, volatility of product
7a and incomplete conversion led to low isolated yield.
Finally, we were interested if the developed methodol-
ogy would also allow for efficient 1,4-addition of highly
congested alkenyl alanates generated by Ni-catalyzed
hydroalumination and treatment of alane with MeLi or
n-BuLi as shown in Scheme 4. Alanates A7 and A8, as well
as alane 9, were tested for the Cu-catalyzed conjugate
addition to enone 5 (Table 4). Before investigating the
Cu-catalyzed conjugate addition with alkenylalanates A7 and
A8 (Table 4, entries 3ꢀ6) we were interested if alane 9 would
also be able to undergo clean 1,4-addition (entries 1ꢀ2).
Moreover, we evaluated the noncatalyzed conjugate addition
of Alanate A8 applying conditions described by Weiss and
co-workers (Entry 7).1
To our surprise alane 9 did not even afford traces of the
desired product 1e (entries 1ꢀ2).15 Even in the presence of
one equivalent of Me3Al only a complex reaction mixture
was observed by GC-MS; the observed unidentified pro-
ducts are probably resulting from 1,2-addition and elim-
ination of the tertiary alcohol under acid workup.16 To our
delight use of alanates A7 and A8 afforded desired product
1e with good conversion, albeit accompanied by about
7ꢀ9% of unidentified side-products (entries 3ꢀ4). Addition
4
97
5h
6h
7h
100
100
96
56
24j
a Reactions performed under Ar atmosphere on a 0.90 mmol scale.
b Determined by 1H NMR of the purified product; in accordance with
estimated GC-MS value of crude product. c Estimated by GC-MS
d Isolated yield; unseparable mixture of 1e and 1f. e Not determined.
f Not isolated. g Complex reaction mixture. h Co-activation with Me3Al
(1.0 equiv). i Reaction without copper, according to Weiss procedure.
j Product contains 6% of unidentified side-product; Crude product
contained 30% of unidentified side-products.
of one equivalent of Me3Al (entries 5ꢀ6) helped to increase
the conversion and to lower the amounts of noncharacterized
side-products to <3%. Intriguingly, we also detected small
amounts (8ꢀ12%) the β-addition product 1f. This is rather
puzzling as in the absence of catalyst this side product was
observed in only 2% of 1f (entry 7).17 The origin of the
conjugate addition of the β-alkenyl nucleophile is unclear.18
Nevertheless, the ease of the synthesis of 1e in good yield
from alkyne 8 is remarkable (entry 7). Piers in his previous
work described the synthesis of 1e by hydrostannation of
alkyne 8, followed by a SnꢀLi exchange and transmetala-
tion to copper to form the required cuprate reagent for the
1,4-addition.19,20
In conclusion, we developed an efficient and practical
protocol for the conjugate addition of alkenyl alanates to
form highly congested quaternary centers. The synthetic
utility of this methodology was demonstrated by the
conjugate addition of functionalized alkenyl Alanate A7
to afford 1e in good overall yield. The development of an
asymmetric version for the Cu-catalyzed conjugate addi-
tion employing alkenyl-alanates is among the objectives
being pursued in our laboratories.
(15) Observed by GC-MS; lack of signal at the corresponding reten-
tion time.
(16) Addition of Me3Al leads to stronger Lewis acid activation
compared to (i-Bu2)Al(alkenyl). This concept of coactivation was
successfully introduced for the Cu-catalyzed conjugate addition with
sterically unhindered alkenylalauminums; see ref 4b.
(17) Hoveyda reported that the β-hydroalumination product was
afforded in <2%: Gao, F.; Hoveyda, A. H. J. Am. Chem. Soc. 2010, 132,
10961–10963.
(18) From our previous work we know that β-alkenylaluminums
react faster than R-alkenyaluminums. Hence, under non-catalyzed
conditions one would expect 9ꢀ13% of β-alkenyl addition product
which is not the case. Observations concerning isomerization of alkenyl
nucleophiles were also made for Rh-catalyzed 1,4-additions: Corey,
E. J.; Lalic, G. Tetrahedron Lett. 2008, 49, 4894–4896.
(19) (a) Piers, E.; Yeung, B. W. A. J. Org. Chem. 1984, 49, 4567–4569.
(b) Piers, E.; Yeung, B. W. A.; Fleming, F. F. Can. J. Chem. 1993, 71,
280–286. 1e was isolated in 55% yield.
(20) The corresponding lithium reagent is only semistable, as it
decomposes quickly to methylenecyclobutane at temperatures higher
than ꢀ63 °C; see ref 19.
Acknowledgment. The authors thank the Swiss National
Research Foundation (grant No 200020-144344) for finan-
cial support.
Supporting Information Available. Analytical data and
NMR spectra for all compounds. This material is avail-
able 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