Atom-economical synthesis of functionalized cycloalkanes via catalytic redox cycloisomerization of propargyl alcohols

Article abstract of DOI:10.1021/ol300278x

An atom-economical procedure for the direct synthesis of cycloalkanes from propargyl alcohols is reported. This high-yielding one-pot process involves a sequence consisting of a Ru-catalyzed redox isomerization of ynols into enones or an enal followed by

Full text of DOI:10.1021/ol300278x

ORGANIC
LETTERS
2012
Vol. 14, No. 7
1708–1711
Atom-Economical Synthesis of
Functionalized Cycloalkanes via Catalytic
Redox Cycloisomerization of Propargyl
Alcohols
Barry M. Trost,* Alexander Breder, and Bao Kai
Department of Chemistry, Stanford University, Stanford, California 94305-5080,
United States
bmtrost@stanford.edu
Received February 3, 2012
ABSTRACT
An atom-economical procedure for the direct synthesis of cycloalkanes from propargyl alcohols is reported. This high-yielding one-pot process
involves a sequence consisting of a Ru-catalyzed redox isomerization of ynols into enones or an enal followed by an intramolecular Michael
addition of a variety of carbon nucleophiles. Furthermore, an asymmetric variant of this protocol realized by the aid of a chiral nonracemic diamine
catalyst, which provides the cyclization products in up to 97% ee, is presented.
In light of the paramount relevance of cycloalkanes, i.e.
in the realm of natural product synthesis, pharmaceuticals,
and material sciences, a great body of effort has been
devoted to the development of new strategies for their
construction. An eminent method for the chemo- and
stereoselective assembly of functionalized cycloalkanes is
the intramolecular Michael reaction.1 Conventional pro-
tocols of this type usually rely on multistep procedures in
which the Michael acceptor is prepared in a separate
operation prior to the conjugate addition step. In particu-
lar, the synthesis of enones and enals generally requires
independent redox manipulations for the implementation
of the carbonꢀcarbon and carbonꢀoxygen double bond,
respectively.2,3 As an alternative, our research group has
focused on the direct and nondissipative conversion of
primary and secondary propargyl alcohols into R,β-
unsaturated carbonyl compounds via Ru-catalyzed redox
isomerization.4 In the course of our program we demon-
strated that the redox isomerization is suitable for the
design of novel consecutive and domino reactions5 such
as intramolecular conjugate additions of heteroatomic
nucleophiles,6 FriedelꢀCrafts/conjugate additions,7 and
cyclopropanations of unactivated olefins.8
On the basis of our previous discoveries, we became
interested in the question of whether propargyl alcohols 1,
which contain a distal CꢀH acidic carbon nucleophile, can
be redox cycloisomerized under ruthenium cocatalysis
to provide a variety of cycloalkanes 2 in a one-pot opera-
tion (Scheme 1). Such a process allows for the direct and
(4) (a) Trost, B. M.; Livingston, R. C. J. Am. Chem. Soc. 2008, 130,
11970. (b) Trost, B. M.; Livingston, R. C. J. Am. Chem. Soc. 1995, 117,
9586.
(1) Little, R. D.; Masjedizadeh, M. R. Org. React. 1995, 47, 315.
(2) For reductions of propargyl alcohols to E-configured allylic
alcohols, see: (a) Grob, C. A.; Gradient, F. Helv. Chim. Acta 1957,
40, 1145. (b) Jenny, E. F.; Druey, J. Helv. Chim. Acta 1959, 42, 401. (c)
Oroshnik, W.; Mebane, A. D. J. Am. Chem. Soc. 1954, 76, 5719. (d)
Marvel, C. S.; Hill, H. W. J. Am. Chem. Soc. 1951, 73, 481.
(5) Tietze, L. F. Chem. Rev. 1996, 96, 115.
(6) (a) Trost, B. M.; Maulide, N.; Livingston, R. C. J. Am. Chem. Soc.
2008, 130, 16502. (b) Trost, B. M.; Gutierrez, A. C.; Livingston, R. C.
Org. Lett. 2009, 11, 2539.
(7) Trost, B. M.; Breder, A. Org. Lett. 2011, 13, 398.
(8) Trost, B. M.; Breder, A.; O’ Keefe, B. M.; Rao, M.; Franz, A. W.
J. Am. Chem. Soc. 2011, 133, 4766.
(3) For chemoselective oxidations of primary and secondary allylic
alcohols to the corresponding R,β-unsaturated carbonyl compounds,
see: Gritter, R. J.; Wallace, T. J. J. Org. Chem. 1959, 24, 1051.
r
10.1021/ol300278x
Published on Web 03/14/2012
2012 American Chemical Society

atom-economical assembly of diversely functionalized
molecular frameworks from simple starting materials.
In addition, this concept offers the opportunity for the
design of an asymmetric process by the use of chiral
cocatalysts for the conjugate addition step. It is note-
worthy that the latter aspect bears significant difficulties,
since the catalytic Michael addition has to be compatible
with the reaction conditions necessary for the redox iso-
merization. Despite these challenges, we report herein an
efficient protocol for the catalytic redox cycloisomer-
ization9,10 of propargyl alcohols and an asymmetric
version thereof.
Table 1. Optimization of the Redox Cycloisomerization
entry
additive
conditions
yield [%]
1
2
3
none
none
24 °C, 16 h
65 °C, 24 h
65 °C, 16 h
0
0
Cs2CO3 (100 mol %)
complex
mixture
<5
4
5
TMGa (30 mol %)
TMGa (200 mol %)
65 °C, 16 h
65 °C, 48 h
98
Scheme 1. Synthetic Concept
a TMG = N,N,N0,N0-tetramethylguanidine.
When we looked at ynols 1jꢀl we made an interesting
observation. Upon subjection of these substrates to the
redox isomerization protocol, the corresponding cyclopen-
tane derivatives 2jꢀl were formed even in the absence of
additional base in yields ranging from 42% to 65%
(Scheme 2). Substrate 1m, on the other hand, required
the presence of TMG in order to convert into 2m (71%
yield). The cause for the difference in reactivity of 1jꢀl
compared to ynols containing a C4-tether may result from
the significantly faster rate of cyclization to 5-membered
rings wherein deprotonation becomes the rate determining
step. In the case of alcohol 1m, however, the steric demand
of the two phenylsulfone groups may hamper the direct
conjugate addition, which would explain the need for an
additional base. Further experiments to elucidate the
mechanistic aspects of the additive-free redox cycloisome-
rization are currently in progress.
In order to verify our hypothesis, investigations began
with a screening for appropriate cyclization conditions. In
initial experiments propargyl alcohol 1a was redox cycloi-
somerized to 2a using IndRu(PPh3)2Cl (Ind = indenide),
indium triflate, and camphorsulfonic acid (5 mol % each)
in THF (0.2 M) at 65 °C followed by addition of a series of
basic additives (Table 1). Inthe absence of any additives no
cyclization product 2a was observed at both room tem-
perature (entry 1) and 65 °C (entry 2). Addition of cesium
carbonate (100 mol %, entry 3) at 65 °C merely led to a
complex mixture of multiple products. Use of 30 mol % of
N,N,N0,N0-tetramethylguanidine (TMG)11 led only to
trace amounts of target structure 2a (entry 4). However,
when the amount of TMG was increased to 200 mol %,
cyclohexane derivative 2a was obtained in an excellent
yield of 98% (entry 5).
With an efficient set of conditions in hand, we continued
with the exploration of the scope of the redox cycloisome-
rization (Scheme 2). Consequently, we synthesized a series
of primary and secondary propargyl alcohols 1bꢀm and
subsequently subjected them to the new reaction protocol.
In general, the method proved efficient for a variety of
carbon nucleophiles such as malonates, β-ketoesters, and
bis-sulfones. In the case of ynols 1aꢀh the corresponding
cyclohexane derivatives were isolated in good to excellent
yields ranging from 64 to 98%. An exception was substrate
1i, which was converted into 2i only in a yield of 55% as a
1:1 mixture of diastereoisomers.
At this point our efforts focused on the design of an
asymmetric redox cycloisomerization protocol. As pre-
viously indicated, the feasibility of a catalytic, asymmetric
conjugate addition is strongly dependent on the compatibility
of the respective catalyst with the reaction conditions for
the redox isomerization. From previous work it was
known that THF was the superior solvent for this parti-
cular step.4 Consequently, we began to screen for appro-
priate catalyst systems that would allow for high levels of
asymmetry during the Michael reaction in THF. For this
purpose we synthesized enone 3e (79% yield) via redox
isomerization, which was subsequently used as our test
substrate for the Michael reaction (Table 2). The reason
why we focused on bis-sulfone nucleophiles was the fact
that none of the tested bis-sulfones displayed background
cyclization during the redox isomerization (cf. ynol 1m vs
1jꢀl). This aspect is important because in the case of C3-
tethered malonates or β-ketoesters the stereoinduction of
a chiral catalyst would presumably be negatively affected
by strong background reactivity.
(9) The term “redox cycloisomerization” refers to the fact that
cycloalkanes 2 are directly generated from their constitutional isomers
1 via redox isomerization and intramolecular Michael addition, ir-
respective of the fact that this transformation usually proceeds in two
stages.
(10) For transition-metal-catalyzed cycloisomerization reactions,
see: Trost, B. M.; Krische, M. J. Synlett 1998, 1.
(11) Ishikawa, T. Superbases for Organic Synthesis: Guanidines,
Amidines, Phosphazenes, and other Organocatalysts; John Wiley & Sons
Ltd: Hoboken, NJ, 2009.
In initial experiments, 9-amino-9-deoxyepiquinidine12
(9-AQD) was used in combination with various Brønsted
(12) Luo, J.; Xu, L.-W.; Hay, R. A. S.; Lu, Y. Org. Lett. 2009, 11, 437.
Org. Lett., Vol. 14, No. 7, 2012
1709

Scheme 2. Scope of the Redox Cycloisomerizationa
Table 2. Optimization of the Conjugate Addition
entry
Ru-/In-cat. [mol %]
additives
yield [%]
1a
2
0/0
5/5
5/0
3/3
3/3
3/3
none
none
none
H2O
PPh3
dppp
71
0
3
20
0
4
5
35
65
6
a 20 mol % of 9-AQD and 40 mol % of 2-CBA were used. 9-AQD =
9-amino-9-deoxyepiquinidine; 2-CBA = 2-chlorobenzoic acid; dppp =
1,3-bis(diphenylphosphino)propane.
3 mol % in order to minimize their malign impact on the
Michael reaction. Additionally, we screened for Lewis
basic additives that are capable of quenching the Ru-
and In-salts without negatively affecting the stereoinduc-
tion of 9-AQD.
Upon addition of 5 vol % of water, based on the amount
of THF, to the reaction mixture no product formation was
observed (entry 4). Addition of 30 mol % of triphenylpho-
sphine, however, led to an increased yield of 35% (entry 5).
Changing to 30 mol % of bidentate 1,3-bis(diphenyl-
phosphino)propane (dppp) afforded cyclohexane 2e in
an isolated yield of 65% (entry 6).
a Conditions: all reactions were performed in THF (0.2ꢀ0.3 M) for
16ꢀ20 h. For the intramolecular Michael reaction 1.1ꢀ2.0 equiv of
TMG were used. bCyclization occurred in the absence of TMG. TMG =
N,N,N0,N0-tetramethylguanidine.
Application of the optimized conditions for the Michael
reaction to our aspired asymmetric redox cycloisomeriza-
tionproved fruitful (Scheme 3). Subjection ofynol 1etothe
new protocol afforded cyclohexane derivative 2e in 60%
yield and in 84% ee. Notably, this result demonstrates that
the presence of dppp has only a marginal impact on the
enantioselectivity. Even better results regarding the ee were
obtained with substrates possessing a branched alkyl
group in R- and β-position to the carbinol center (ynols
1d and 1f). The corresponding cyclization product 2d was
isolated in 70% yield and with an excellent ee of 97%.
Homologous structure 2f was obtained with a somewhat
lower ee of 87% but in good yield (73%). Unfortunately,
when primary propargyl alcohol 1g was used, the ee
decreased drastically to 14%. We speculate that the low
stereoinduction is caused by a fast background reaction,
which is caused by neither the ruthenium nor the indium
catalyst.17 This hypothesis is supported by the fact that
conversion of 1g to 2g is complete within 24 h whereas the
reaction time for the analogous secondary ynols is 3 d.18
A similar observation was made for cyclopentane 2m, which
was isolated in 73% yield and 22% ee. In the cases of
acid cocatalysts.13,14 From these studies we learned that
20 mol % of the organocatalyst in combination with
40 mol % of 2-chlorobenzoic acid (2-CBA)15 in THF at
45 °C efficiently furnished 2e in 71% yield and 90% ee
(Table 2, entry 1).16 Unfortunately, when enone 3e was
first premixed with 5 mol % of both IndRu(PPh3)2Cl and
In(OTf)3 followed by addition of 9-AQD (20 mol %) and
2-CBA (40 mol %) no conversion was observed (entry 2).
When In(OTf)3 was omitted from the reaction mixture, we
observed some reactivity; however, target structure 2e was
isolated only in 20% yield (entry 3). From these observa-
tions we concluded that the presence of strong Lewis acids
is detrimental to the activity of the organocatalyst. Con-
sequently, we lowered the amount of metal catalysts to
(13) Sun, X.; Yu, F.; Ye, T.; Liang, X.; Ye, J. Chem.;Eur. J. 2011,
17, 430.
(14) For a screening of Brønsted acids for the optimization of
enantioselectivity, see Table 1 in the experimental section.
(15) In general, benzoic acid derivatives gave better ee’s than alipha-
tic carboxylic acids (see Supporting Information). However, the exact
reason for the superiority of 2-chlorobenzoic acid remains unclear at this
point.
(17) We did not observe any Michael addition product during the
redox isomerization of ynol 1g in the absence of base.
(16) During the investigations described in Table 2 weprimarilyfocused
on the impact of various Lewis bases on the isolated yield. Thus, we did not
determine the ee values of 2e for entries 3, 5, and 6 in Table 2.
(18) Although all redox cycloisomerization reactions described in
Scheme 3 were allowed to proceed for 3 d, TLC analysis for ynols 1g and
1m indicated that the reaction reached completion after 24 h.
1710
Org. Lett., Vol. 14, No. 7, 2012

Scheme 3. Scope of the Asymmetric Redox Cycloisomerizationa
Scheme 4. Mechanistic Proposal for the Redox Cycloisomeri-
zation
cycloalkanes 3dꢀf and 3h enone activation is presumably
the rate limiting step (Scheme 4, iminium ion III).19 For
compounds 2g and 2m, on the other hand, deprotonation
of the bis-sulfone entity is rate limiting. Consequently, the
decreased enantioselectivity could be explained by the low
stereocontrol exerted by the 9-amino-9-deoxyepiquini-
dinium counterion compared with the intramolecular
control present in iminium species III.
Based on our previous investigations on redox isomer-
ization reactions in combination with the present study
and recent work by You et al.,4,19 we propose the following
mechanistic hypothesis for the asymmetric redox cycloi-
somerization (Scheme 4). In the first step, propargyl
alcohol 1 is catalytically converted to R,β-unsaturated
carbonyl compound 3 via 1,2-hydride migration.4 Sub-
sequent intramolecular Michael addition by means of
9-AQD-mediated iminium catalysis generates carbocycle
IV, which upon hydrolysis converts into ketone 2.
broad substrate scope and generally furnishes the cycliza-
tion products in good to excellent yields. We have also
demonstrated that this procedure can be easily directed in
an asymmetric fashion by use of a cinchona alkaloid-
derived diamine catalyst, which offers levels of enantios-
electivity of up to 97% ee.
Acknowledgment. This work was supported in part by a
fellowship for A.B. from the German Academic Exchange
Service (DAAD). We thank the US National Science
Foundation (CHE 0948222) for their generous support
of our program. We thank Umicore for a generous gift of
ruthenium salts used to prepare the catalysts.
In summary, we have developed a highly atom-economical
and redox neutral protocol for the direct conversion of
propargyl alcohols into cycloalkanes via ruthenium and
Brønsted base cocatalysis. This novel method features a
Supporting Information Available. Experimental pro-
1
cedures, spectrocopic data, and spectra of H and 13C
NMR for the addition products. This material is available
free of charge via the Internet at http://pubs.acs.org.
(19) Cai, Q.; Zheng, C; Zhang, J.-W.; You, S.-L. Angew. Chem., Int.
Ed. 2011, 50, 8665.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 7, 2012
1711