Organic Letters
Letter
The utility of organosulfur compounds as precursors of
organozinc reagents is thus underdeveloped.
Scheme 3. Optimization Study of the Zincation Reaction
We are interested in the development of catalytic reactions
of organosulfur compounds via the cleavage of the C−S
bond.13−15 Recently, we have reported the nickel-catalyzed
carboxylation of arylsulfonium salts using zinc powder with
carbon dioxide.16 This study implied that an arylzinc species is
the intermediate formed prior to the carboxylation event,
although tangible evidence for this hypothesis had been
elusive17 or deemed unlikely18 in nickel-catalyzed carboxyla-
tion reactions, which use a metallic reductant. Our previous
success in preparing arylboronate esters from arylsulfonium
salts and bis(pinacolato)diboron14a via a palladium-catalyzed
reaction also encouraged us to develop another important ipso-
metalation of arylsulfonium salts. With this in mind, we report
here an efficient nickel-catalyzed preparation of arylzinc
reagents from a variety of arylsulfonium salts (Scheme 1c).
Encouraged by Liebeskind’s original idea,19−21 we prefer
using arylsulfonium salts as aryl pseudohalides. When
compared with neutral aryl sulfides, arylsulfonium salts have
several advantages for nickel-catalyzed zinc-insertion reactions
that can be explained on the basis of the possible reaction
mechanism5,7,16 (Scheme 2): (1) Arylsulfonium salts are
reproducibility. After some experimentation, we found that the
addition of 6.0 mol % of a supporting ligand ensures
reproducibility. Of the ligands tested, 2,9-dimethylphenanthro-
line (L1) proved to be the most effective.23 Whereas
monomethylated L2 showed comparable reactivity, neither
the parent phenanthroline (L3) nor 3,4,7,8-tetramethylphe-
nanthroline (L4) worked. The more flexible 2,2′-bipyridyl
ligands L5 and L6 did not facilitate the zincation. In these four
runs with L3−L6, low conversions were observed. Finally,
using 2.5 mol % of the complex NiBr2L1 and an additional 10
mol % of L1 provided 3a in 91% yield. A screening of solvents
revealed that N,N-dimethylacetamide (DMA) is the best,
whereas N-methylpyrrolidone (NMP), dimethyl sulfoxide
(DMSO), and N,N′-dimethylpropyleneurea (DMPU) are
almost comparable. Conversely, the zincation did not occur
in other polar solvents such as acetonitrile or tetrahydrofuran
(THF). N,N′-Dimethyl-2-imidazolidinone (DMI) and tetra-
methylurea (TMU) showed moderate performance. The
corresponding cobalt complex CoBr2L1 was not catalytically
active in the zincation. In this case, as well as in the absence of
any transition-metal complex, gradual demethylation of 1a was
observed, most likely via a SET directly from the zinc powder
to 1a.24 The neutral sulfide p-tBuC6H4SMe did not undergo a
similar zinc insertion reaction, discarding the possibility of an
aryl sulfide intermediate in the zincation.
With the optimized conditions in hand, we investigated the
reaction scope (Scheme 4). Electron-rich and -neutral
arylsulfonium salts 1a−d, 1h, and 1i displayed good
reactivity.25 As expected, 1c, which contains a methylsulfanyl
group, was selectively zincated at the sulfonium moiety. The
methoxy group at the ortho position in 1d has little effect on
the zinc insertion. Electron-deficient cyano-substituted 1e
suffered from competitive SET from either the zinc powder or
a low-valent nickel species,24 and thus the demethylation of 1e
competed to yield 4-methylsulfanylbenzonitrile, and the yield
of 3e was moderate.26 Notably, the zincation of the other
electron-deficient arylsulfonium salts 1f−g proceeded effi-
ciently under the standard conditions. The potentially reactive
carbonyl groups in 3f−h were unaffected by the reaction
conditions. Surprisingly, the acidic protons in 1h have no
detrimental effect on the reaction, and the corresponding zinc
Scheme 2. Mechanistic Working Hypothesis
electron-deficient and undergo smoother oxidative addition.
(2) The oxidative addition forms a cationic arylnickel
intermediate that is more susceptible to reduction by metallic
zinc.22 (3) Instead of anionic sulfur fragments, a neutral and
catalytically less poisonous sulfide is formed. (4) Arylsulfonium
salts, which can be readily prepared from aryl mercaptans or
sulfides on a large scale, usually exhibit good crystallinity as
well as bench stability. We also had to pay attention to the
possibility of the undesired degradation of the arylsulfonium
salts into neutral aryl sulfides via demethylation. This could
potentially occur via the action of a nucleophilic arylmetal
species generated in situ or by a single-electron transfer (SET)
from zinc or a low-valent nickel species.23 Suppressing the SET
is also important for controlling the regioselectivity of the C−S
bond cleavage, where sp2 C−S bond cleavage via two-electron
oxidative insertion predominates over SET-induced sp3 C−S
bond cleavage.
Our investigations began by evaluating the catalytic
zincation of 1a (Scheme 3). The efficiency of the zincation
was assessed using 1H NMR spectroscopy to analyze the
iodinated product 3a after iodolysis. Simply applying the
standard conditions from our previous carboxylation in the
absence of carbon dioxide16 resulted in a zincation with poor
B
Org. Lett. XXXX, XXX, XXX−XXX