2
Wu et al. Sci China Chem
multaneously, giving 10.0/1.0 selectivity and 95% yield.
Encouraged by this result, we decided to focus on evaluation
of aryl sulfoxides (2i–2n) (Scheme 2). Although more steric
hindrance sulfoxide 2i led to lower reactivity, diphenyl
sulfoxide 2j provided its sulfonium salt in quantitative yield
with similar selectivity. Compared to diphenyl sulfoxide, less
steric hindrance dibenzothiophene sulfoxide (2k) and
thioxanthone sulfoxide (2l) resulted in lower selectivities.
Gratifyingly, phenoxathiine sulfoxide (2m) and thianthrene
sulfoxide (2n) significantly improved the selectivity to >76/1
(para/ortho) with high efficiency.
To shed light on the causation for remarkable para se-
lectivity using phenoxathiine and thianthrene derived sulf-
oxide, detailed mechanistic studies were carried out.
Although the sulfoxide screening confirmed that the re-
gioselectivity can be partially controlled by the steric effects,
the electron paramagnetic resonance (EPR) experiment of
the reaction system with various in situ generated sulfide
ditriflates indicated that the striking para-selecitivity might
also be related to the formation of phenoxathiine [10a] and
thianthrene radical cation intermediates [10b–10f,11]. In
comparison, thioxanthone sulfoxide (2l) with similar scaf-
fold without EPR response resulted in a lower para se-
lectivity (Figure 1(a)). However, the reaction did not proceed
Scheme 1 Background and synopsis for para-C–H functionalization of
monosubstituted arenes (color online).
viding the para-decorated aromatics. By carefully selecting
the transient mediator and conditions, para-functionalization
of monosubstituted benzene could be realized in a single
synthetic operation. In this case, highly electrophilic phe-
noxathiine or thianthrene dication intermediate [8], in situ
initiated by its sulfoxide with trifluoromethanesulfonic an-
hydride, was capable of generating the corresponding sul-
fonium salt with a high para-selectivity. Para-borylation of
monosubsituted benzenes was realized by combining the
sulfonium salt formation with further photocatalyzed bor-
ylation using phenoxathiine or thianthrene as the transient
mediator.
We started our investigation by searching a suitable tran-
sient mediator, which has the capability of being installed at
the para-position of the monosubstituted aromatics with high
selectivity, and being readily converted to other functional
groups. Being aware of the remarkable activity of sulfide
ditriflate, a highly active intermediate formed in situ by
at low concentration of sulfoxide and Tf O, although the
2
radical cation intermediate existed which confirmed by the
EPR experiment. We hence doubted whether the radical
cation intermediate is real reactive species for sulfonium
salts formation, and were intrigued by the origin of the
outstanding regioselectivity. To further understand this pro-
cess, preliminary computational studies were performed at
SMD-M062x/def2-TZVP level of theory, and three possible
intermediates, radical cation, radical cation dimer and sulfide
dication, were systematically investigated. The intermediates
of C–S formation with toluene and thianthrene radical cation
did not show apparently energy difference (<0.6 kcal/mol),
indicating the poor selectivity of para/ortho position by the
calculated energy scanning of forming C–S distance (Figure
S8, Supporting Information online). Based on Dunsch’s
study [10g], a dimer intermediate could be formed via a
reversible dimerization of thianthrene radical cation; how-
ever, the dimer can not lead to stable adducts for the rea-
sonable para/ortho selectivity in the electrophilic
substitution of toluene through calculations (Figure S9). We
next turned our attention to the plausible thianthrene dication
intermediate, and found the sulfide dication intermediate is
facile from heterolysis of sulfide ditriflate 2n-OTf with
slightly endothermic rather than disproportionation from
thianthrene radical cation. The large free energy difference
(>4.0 kcal/mol) between para- and ortho- intermediates
imply that the sulfide dication would be the possible active
species responsible for the high regioselectivity. And the
dispersedly positive NPA charge (Figure S11) on both
treating sulfide oxide with Tf O [9], we hypothesized that the
2
electrophilic sulfate ditriflate reagents might be a potential
handle for addressing this challenging problem by adjusting
the steric effects. In situ generated dimethyl sulfide ditriflate
(
DMSD) reacted with toluene providing the corresponding
sulfonium salts in 99% yield with an insufficient selectivity
p/o=1.5/1.0). The selectivity was slightly improved with the
increase of steric hindrances (2b–2e). Diisopropyl sulfoxide
2f) gave a remarkable para/ortho ratio of >20/1.0, whereas
(
(
the efficiency of sulfonium salt formation is unsatisfying and
cannot be further improved probably due to bulky isopropyl
substituent. It is noteworthy that methyl phenyl sulfoxide
(2h) maintained the high efficiency and high selectivity si-