.
Angewandte
Communications
would, themselves, contain a homoallylic carbonate motif
suitable for further elaboration.
found that the use of the corresponding carbamate-derivative
1b afforded excellent yield of the desired product with similar
selectivity. Interestingly, close monitoring of the reaction
revealed that 3a was not formed in the crude reaction mixture
until a hydrolytic work up was applied to form the cyclic
carbonate. We found that the dimethylcarbamate gave
superior yields to those of other amine derivatives, and that
no variation of the solvent or the reaction temperature
improved the selectivity. Therefore, optimal reaction condi-
tions were concluded to involve treatment of 1b with
2 equivalents of 2a and 5 mol% of (CuIOTf)2·PhH as the
catalyst in a solution of dichloromethane at 408C for 4 hours,
and resulted in the isolation of 3a in 96% yield as a 5:1
mixture of syn and anti diastereoisomers.
We next assessed the scope of the new copper-catalyzed
oxy-alkenylation reaction of homoallylic carbamates
(Table 2). We were pleased to observe that the process
works well on a range of substrates, thus providing access to
a diverse array of functionalized syn carbonates. Substrates
bearing simple alkyl substituents proved to be high yielding
for the 1,3-carbonates 3b–d. Interestingly, the diastereoselec-
tivity observed across these products was relatively unaf-
fected by the changing steric demands of the substrate, and
marginal improvement was observed by increasing substitu-
ent size from methyl to isopropyl to tert-butyl (d.r. = 4:1 for
3c to 6:1 for 3d). A selection of synthetically versatile
functional groups were compatible with the reaction and
formed products in comparable yields and, in many cases,
higher diastereoselectivity: homoallyl (3e), phenyl (3 f),
chloromethyl (3g), and protected hydroxy motifs (3h and
3i) all proceeded well to provide useful cyclic carbonate
products with selectivities between 5.5 and 10:1. Homoallylic
carbonates with substituents displaying additional stereogenic
centers reacted smoothly, again in good yield and diastereo-
selectivity, to form the stereotriads 3j and 3k commonly
found in polyketide natural products.
Over the last seven years, our group has pioneered the use
of diaryliodonium reagents in combination with copper salts
as a catalytically generated source of aromatic electrophile
equivalent.[13] Proceeding via the intermediacy of a putative
copper(III)/aryl species, this highly electrophilic organome-
tallic reagent has been shown to undergo reaction with
a variety of latent carbon nucleophiles, thus leading to the
development of a range of previously unknown transforma-
tions. In the context of an electrophilic oxy-carbofunctional-
ization strategy towards synthetically versatile 1,3-diols, we
reasoned that the action of a copper catalyst on an alkenyl-
(aryl)iodonium salt would generate a copper(III)/alkenyl
species (Scheme 1c).[14] This reactive intermediate would
engage the alkene, thus polarizing the carbon–carbon double
bond and promoting attack of the pendant carbonyl motif in
such a way to form a carbon–oxygen bond and a new
copper(III)–carbon bond. Reductive elimination from the
high oxidation state copper(III) species would transfer the
alkenyl group leading to a cyclic carbonate product, which
was predicted to be the syn diastereoisomer in line with the
related iodine-mediated reaction.
Herein, we report the successful realization of such
a copper-catalyzed oxy-alkenylation strategy and show that
a range of readily available, substituted homoallylic alcohol
derivatives and alkenyl(aryl)iodonium salts can be combined
to form syn-1,3-carbonates in excellent yield and with high
selectivity (Scheme 1d). Moreover, a simple iteration of the
process affords highly complex polyketide-like fragments
which could be used to expedite the synthesis of natural
products.
At the outset of our studies, we selected the simple mixed
carbonate 1a to assess the oxy-alkenylation strategy
(Table 1). Based on our hypothesis, we first assessed simple
In addition to the primary homoallylic carbamates that
worked well to form the corresponding products, we were
pleased to find that substituents on the internal position of the
carbon–carbon bond worked well to form functionalized
cyclic carbonates containing a quaternary carbon atom (3m
and 3n; Table 2). The reaction could also be deployed on
allylic carbamates to form either the five-membered ring (3o)
or six-membered (3p) cyclic carbonates, depending on the
nature of the alkene substituent. The transferred alkenyl
group can also be varied without significantly affecting the
yield or diastereoselectivity (3q–s).
Table 1: Optimization of the carbonyl component of the substrate.[a]
Having established that terminal alkenes were excellent
reaction partners in the copper-catalyzed oxy-alkenylation,
we next investigated whether 1,2-disubstituted carbon–
carbon double bonds could also function as part of this
process. Methyl substituents appended to the alkene were of
particular interest because the product of the transformation
would form vicinal stereocenters reminiscent of propionate
motifs commonly found in polyketide natural products. We
were pleased to find that the Z-alkene (Z)-4 formed the anti-
isomer 5a with very good selectivity and in high yield
(Scheme 2). Importantly, the E-alkene (E)-4, under similar
reaction conditions, formed the corresponding syn-product
[a] Major diastereoisomer confirmed as syn by one-dimensional NOESY.
copper salts as catalysts with the pentenyl(triisoproylbenze-
ne)iodonium triflate 2a (readily available in our laboratories;
see the Supporting Information for full details of optimiza-
tion). We found that the desired reaction was best catalyzed
by the copper(I) salt (CuIOTf)2·PhH, thus affording the cyclic
carbonate 3a with good diastereoselectivity (d.r.), albeit in
low yield. However, under the same reaction conditions, we
7858
ꢀ 2015 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2015, 54, 7857 –7861
Holt, Dean
