Organic Letters
Letter
Recent progress in mechanochemical synthesis and ball-
milling has led to the investigation of several organic reactions
using this unique reactor environment. Very recently, a few
general trends have started to emerge about the potential
benefits that ball-milling might offer, such as new reactivity,
decreased reaction times, or alternative selectivity.8 In the
context of Pd-catalyzed cross-coupling reactions by ball-
milling, we and others have found a seemingly decreased
sensitivity to air and moisture for both Buchwald−Hartwig and
Negishi cross-coupling processes (Scheme 1B).8k,9a,c,f We
hypothesized that the use of mechanochemistry could allow
us to overcome several of the limitations encountered by
solution-based, Pd-catalyzed C−S approaches, while simulta-
neously allowing us to further explore and understand the
method of ball-milling (Scheme 1C). Our results are reported
herein.
Table 1. Optimization of Mechanochemical C−S Coupling
a
entry
variation from standard conditions
yield of 3 (%)
1
2
3
none
no grinding agent
sand as the grinding agent (2 mass equivalents)
76
0
94 (90)
Entries below this line contain sand (2 mass equivalents)
2 h reaction time
4
5
95
1.5 h reaction time
Entries below this line run for 1.5 h
96 (92)
Studies commenced by employing our recent reaction
conditions developed for a Buchwald−Hartwig amination
reaction.9a Iodobenzene (1) and thiophenol (2) could be
coupled successfully using Pd−PEPPSI−IPent (1 mol %) as
the catalyst, potassium tert-butoxide (2 equiv) as the base, and
sand (3 mass equiv) as the grinding auxiliary. Grinding
auxiliaries are often used when there are one or more liquid
reagents and provide a solid surface to host the liquid materials
and permit more efficient energy transfer from the balls to the
reactants.8 Mass equivalents of grinding auxiliaries are
measured as the mass of all reagents combined multiplied by
the corresponding number of equivalents. Notably, the
materials were added to the milling jar under an air
atmosphere; i.e., no precaution was taken for the exclusion
of air or moisture from the reaction mixture. The jars were
milled for 3 h at 30 Hz. Under these conditions, the reaction
afforded the arylated thiol product (3) in 76% yield as
6
7
8
9
no catalyst
0
96
95 (92)
31
catalyst A (2 mol%)
catalyst A (0.5 mol%)
catalyst A (0.25 mol%)
Entries below this line contain 0.5 mol% Catalyst A
tBuOK (1 equiv)
tBuOK (3 equiv)
tBuOK (1.5 equiv)
tBuONa (2 equiv)
Cs2CO3 (2 equiv)
10
11
12
13
14
32
5
49
13
6
a
Yield determined by 1H NMR using mesitylene as the internal
standard; numbers in parentheses represent isolated yields.
pound 8) as well as other ortho-substituted substrates (Scheme
2, compounds 6 and 7). 1-Bromonaphthalene was also
successfully coupled in 76% yield, demonstrating the steric
tolerance of this reaction. These conditions could also be
applied to a range of electron rich and electron deficient aryl
halides.
It is worth noting that for electron poor examples 13 and 14,
the reaction was also performed in the absence of a catalyst
whereby no desired product was observed, demonstrating that
the reaction does not proceed by an SNAr mechanism under
these conditions. Notably, 3-bromopyridine was tolerated,
affording product 10 in a moderate 51% yield. Substrates
containing acidic sites (carboxylic acid and phenol) were
unsuccessful in the coupling reaction.
The reaction scope with respect to the thiol component was
first explored for aryl thiols using iodobenzene as the coupling
partner. A range of aryl thiols were coupled successfully under
the standard conditions. It was shown that for 4-amino-
thiophenol, C−S coupling was favored over C−N coupling of
the aniline functionality, leading to thioether 19 in 71% yield.
However, it was found that some examples (Scheme 2,
compounds 23, 24, 18, and 7) did not afford the desired
thioether products but instead gave the corresponding disulfide
of the thiol starting material. This was investigated further
using 4-ethylthiophenol (Scheme 2) as the model substrate.
Under the standard reaction conditions, milling 4-ethyl-
thiophenol with iodobenzene afforded none of the thioether
product and 84% of the disulfide (22) (Scheme 2). We
hypothesized that inclusion of a reductant might enable the
desired thioether formation.10 To our delight, addition of Zn
metal as a reductant led to the observation of no disulfide and
72% of the desired arylated thiol product (23) (SI Table 4,
entry 2). After a short optimization, it was found that 2.5 equiv
1
determined by H NMR (Table 1, entry 1). The reaction
parameters were then refined to further optimize the reaction.
Screening commenced by evaluating the grinding agent used
where it was found that 2 mass equivalents of sand provided an
improvement, giving the desired cross-coupled product in 90%
isolated yield (Table 1, entry 3). After a small range of six
NHC-Pd catalysts had been screened (SI Table 1), it was
found that the Pd−PEPPSI−IPent catalyst was optimal and
the loading of this catalyst could be decreased to 0.5 mol %
(Table 1, entry 8). Notably, omission of the palladium catalyst
resulted in no reaction (Table 1, entry 6), which is a key
observation given that trace metals (such as iron and cobalt)
from the stainless steel balls and jar have previously been
implicated as non-innocent in some ball-milled processes.8l
The reaction time could also be decreased to 90 min, which
was sufficient for the coupling of these model substrates giving
an isolated yield of 92% for the cross-coupled product 3. No
improvements were observed when the base or the number of
equivalents of base was changed (Table 1, entries 10−14).
With optimal conditions identified (Table 1, entry 8),
application to a range of aryl halides was explored (Scheme 2).
It was quickly discovered that many examples required >90
min; the reactions were therefore run for 3 h as the standard
procedure. In all cases, reactions were run under an air
atmosphere and afforded the desired products in moderate to
excellent yield [43−92% isolated yields (Scheme 2)]. Initial
testing found that the reaction was tolerant of aryl bromides
and the more electronically challenging aryl chlorides in
addition to aryl iodides. The reaction is tolerant of the
sterically demanding mesityl functionality (Scheme 2, com-
B
Org. Lett. XXXX, XXX, XXX−XXX