.
Angewandte
Communications
DOI: 10.1002/anie.201403796
Ruthenium Catalysis
Ruthenium-Catalyzed Oxidative Transformations of Terminal Alkynes
to Ketenes By Using Tethered Sulfoxides: Access to b-Lactams and
Cyclobutanones**
Youliang Wang, Zhitong Zheng, and Liming Zhang*
Abstract: The oxidation of in situ generated Ru vinylidenes to
ketenes is realized with tethered sulfoxides. The result is a Ru-
catalyzed oxidative transformation of terminal alkynes to
highly valuable ketenes. Moreover, the ketenes generated here
were shown to undergo characteristic ketene [2+2] cyclo-
addition reactions with tethered alkenes and external imines,
yielding synthetically versatile bicyclic cyclobutanones and b-
lactams, respectively.
teristic ketene [2+2] cycloaddition[9] with tethered alkenes
and external imines.[7] This offers strong evidence for the
intermediacy of a ketene and yields synthetically versatile and
strained cyclobutanones and b-lactams,[10] respectively, in
mostly good yields.
At the outset, we chose the enynyl aryl sulfoxide 1a as the
substrate and anticipated that if its terminal alkyne could be
converted into a ketene moiety (as in B) an intramolecular
=
[2+2] cycloaddition with the tethered C C bond would afford
R
uthenium vinylidenes[1] are versatile intermediates in
the bicyclic cyclobutanone 2a (Table 1).[11] The substrate was
readily prepared from chloroacetone in four steps and was
obtained as a mixture of diastereoisomers (d.r. 1.1:1, see the
Supporting Information, SI). This mixture was treated with
various ruthenium catalysts under different conditions.
Whereas RuIII catalysts such as RuCl3 (entry 1) and [Ru-
(acac)3] (entry 2) did not promote the reaction at all, a RuII
catalyst, [Ru(CO)3]Cl2, did catalyze the reaction albeit 2a was
formed in only 20% yield (entry 3). [CpRu(PPh3)2]Cl,
a typical catalyst for the generation of ruthenium vinyl-
idenes,[12] was ineffective in the absence of a chloride
scavenger (entry 4).[13] However, when NaPF6 (10 mol%)
was added, the desired bicyclic cyclobutanone 2a was formed
in 36% yield (entry 5). A nearly equal amount of a bicyclic
aldehyde side product, i.e., 3a, was isolated as a single
diastereoisomer (for a proposed mechanism, see SI).
organic synthesis. Owing to their ready accessibility from
terminal alkynes, ruthenium vinylidenes have served as an
entry into a diverse range of efficient transformations,[2] in
which the carbene center reacts with various nucleophiles. To
date, the oxidation[3] of in situ-generated Ru vinylidenes to
synthetically useful ketenes[4] (Scheme 1), thereby realizing
Scheme 1. Ru-catalyzed oxidation of a terminal alkyne to a ketene and
subsequent [2+2] cycloaddition reactions.
the Ru-catalyzed oxidative transformation of a terminal
alkyne into a ketene, has had limited success. An exception
is the excellent work by Liu and co-workers[5] using tethered
epoxides or nitrones as the oxidants, which were, however,
confined in conformationally rigid systems. Recently, Lee
reported that a Rh complex could catalyze this transforma-
tion with either internal or external oxidants;[6] however, the
generated ketenes were trapped only by heteronucleophiles,
leading to the formation of esters, amides, and acids.[7] Herein,
we disclose a Ru-catalyzed oxidation of terminal alkynes into
ketenes with flexibly tethered sulfoxides[6,8] as internal
oxidants under mild reaction conditions. Most importantly,
besides being trapped by heteronucleophiles, these oxida-
tively generated ketenes were shown to undergo the charac-
To our surprise, no reaction occurred when the bulkier
[Cp*Ru(PPh3)2]Cl was employed (entry 6). The fact that 3a
was formed in a large amount in entry 5 suggested that the
generation of ruthenium vinylidenes was not efficient. The
phosphine ligand L1, an AZARPHOS in which the pyridine
nitrogen atom is sterically shielded from coordinating to
Ru,[14] is known to facilitate the Ru-catalyzed anti-Markovni-
kov hydration of terminal alkynes[15] by accelerating the
isomerization of the terminal alkyne to the corresponding Ru
vinylidene.[16] Indeed, when 10 mol% of L1 was added, the
yield of 2a increased to 81% and the formation of 3a (4%
yield) was significantly suppressed (entry 7). An even better
result was attained with the synthetically more accessible
AZARPHOS L2 (entry 8).[14] Lowering the ligand loading to
5 mol%,[17] however, decreased the yield (entry 9). Little
improvement was achieved when the chloride scavenger was
changed to AgNTf2 (entry 10), but NaBArF (entry 11)
[*] Y. Wang, Z. Zheng, Prof. Dr. L. Zhang
Department of Chemistry and Biochemistry, University of California
Santa Barbara, CA (USA)
E-mail: zhang@chem.ucsb.edu
[**] This work was financially supported by the NSF (CHE-1301343) and
the ACS Petroleum Research Fund (52040-ND1)
4
proved to be superior, and the bicyclic cyclobutanone 2a
was formed in 90% NMR yield. Acid side products were
observed in the above reactions and confirmed by a reaction
in the presence of 10 equiv of H2O (see SI for details);
however, when the reaction was run in the presence of 4 ꢀ
MS, 2a was isolated in excellent yield(entry 12). A lower yield
was detected when the reaction was run at a much higher
Supporting information for this article is available on the WWW
9572
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2014, 53, 9572 –9576