Angewandte
Communications
Chemie
During the course of these studies, we recognized the
À
possibility of I2/I3 formation species, which are known to
quench the photoexcited state *[Ru(bpy)3]2+ (5) and would be
detrimental for the progress of the reaction.[13] In this sense,
we were pleased to find that by adding water and sodium
thiosulfate as a suitable iodine quencher to the reaction
mixture, the yield of 3a successfully increased up to 76%
(entry 2).
These conditions were adapted for a larger scale reaction
using 1 gram of (E)-anethole (1a) without compromising the
efficiency of the process (85% yield, entry 2). Control
experiments showed that i-Pr2EtN (entry 3) and visible light
(entry 4) are essential for this reaction, and also the necessity
of degassing the reaction mixture prior to irradiation
(entry 5).[14] More importantly, we found what E/Z-mixtures
of 1a (E/Z, 1:2) resulted in the exclusive formation of 3a, thus
demonstrating the feasibility of conducting a stereoconver-
gent cyclopropanation reaction.[15] Additional control experi-
ments showed that (Z)-1a to (E)-1a or cis-3a to trans-3a
isomerizations did not occur during the course of the reaction
and UV light (300 nm) does not promote the cyclopropana-
tion as previously described by Kropp for alkyl-substituted
alkenes (see the Supporting Information).[11a–c]
Scheme 1. Working hypothesis.
photoexcited state *[Ru(bpy)3]2+ (5; E1/2(II)*/(I) = 0.77 V vs.
SCE), which is reduced to [Ru(bpy)3]+ (6) with N,N-diiso-
propylethylamine by a well-established SET process. The RuI
intermediate is a strong reductant (E1/2(II)/(I) = À1.33 V vs.
SCE) and would donate an electron to CH2I2 (Ered = À1.44 V
vs. SCE),[10] thereby generating a transient anion radical
[CH2I2]ÀC, which evolves into iodomethyl radical (7).[11]
Addition to E,Z-alkenes 1 would conduct to intermediates
Having the optimized reaction conditions in hand, we next
investigated the scope of this cyclopropanation by examining
a wide range of styrenes 1a–x. As shown in Table 2, this
process was successfully applied in substrates functionalized
with electron-donating and electron-withdrawing groups (3b–
j) and also with heterocycles (3k,l).
À
int I and int II, which would be in equilibrium through a C C
bond rotation. Finally, a cyclization step would take place by
homolytic substitution preferentially on intermediate int II,
in which both R1 and R2 groups are in a favorable relative anti
orientation.[12] Our envisaged cyclopropanation was initially
evaluated using (E)-anethole (1a) as the alkene substrate. To
our delight, in a first round of experiments we identified that
the cyclopropanation reaction took place with 59% yield of
desired trans-cyclopropane 3a as a single diastereoisomer by
using 1 mol% Ru(bpy)3(PF6)2 (4), diiodomethane (2;
2.5 equiv), N,N-diisopropylethylamine (5 equiv), acetonitrile
(0.1m) and a 21 W compact fluorescent lamp (CFL) as visible
light source (Table 1, entry 1).
We noticed that styrenes bearing electron-rich substitu-
ents converted to the corresponding cyclopropane in a more
efficient manner (3a–c). It is possible that the nucleophilic
character of the benzyl radical int II plays an important role
in the cyclization event (Scheme 1, R1 = Ar). In fact, no
product (3h) or low yield (3i,j) was observed for styrenes
bearing moderate/strong electron-withdrawing groups, and
alkene decomposition was observed. Having established that
the electronics of the aromatic moiety have a direct impact in
the outcome of the cyclopropanation reaction, we next
evaluated the alkyl substitution for alkenes 1m–w. This
procedure was successfully applied to a wide range of alkyl
groups functionalized with aromatic rings (3m), cycloalkanes
(3n), acetals (3o), alcohols (3p–s), imides (3t,u), alkyl iodides
(3v), or alkenes (3w). The excellent level of efficiency for this
radical cyclization is notable because no other reaction
products have been observed for substrates bearing embed-
ded nucleophiles, which could potentially intercept a benzylic
radical intermediate (Table 2; 3m,s,w), as well as for substrate
Table 1: Optimization studies.[a]
Entry
Alkene 1a
Additive
Yield 3a [%][b]
À
1v bearing a C I bond, which could participate in a compet-
1
2
3
4
5
6
E
E
E
E
E
E/Z
–
59
Na2S2O3,H2O
Na2S2O3,H2O
Na2S2O3,H2O
Na2S2O3,H2O
Na2S2O3,H2O
76(85)[c]
0[d]
itive cyclization for the synthesis of a cyclobutane analogue.
However, only cyclopropane 3v was observed as only product
of the reaction. The excellent functional group tolerance of
this radical cyclopropanation is illustrated in substrates
bearing tertiary amines (3b,m–o) or sulfides (3c,u,w), which
generally react with iodomethylzinc reagents generating the
corresponding ylide.[16] Furthermore, directing groups were
not needed to ensure site-selectivity for a substrate bearing
two reactive alkenes (3w). Remarkably, only the alkene
conjugated with the aromatic group was selectively cyclo-
0[e]
45[f]
79
[a] Reaction conditions: 1a (0.10 mmol), 2 (0.25 mmol), i-PrEt2N
(0.50 mmol), CH3CN (1 mL); Na2S2O3 (0.50 mmol), H2O (0.4 mL).
Reactions were degassed prior to irradiation. [b] GC/MS yields calculated
using acetophenone as internal standard. [c] Yield of isolated product
using 1 gram of (E)-1a. [d] Reaction carried out without i-Pr2EtN.
[e] Reaction carried out in the dark. [f] Reaction mixture not degassed.
2
ꢀ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2016, 55, 1 – 5
These are not the final page numbers!