Letter https://doi.org/10.

1038/s41586-019-0982-0

Site-selective and versatile aromatic C−H

functionalization by thianthrenation
Florian Berger1, Matthew B. Plutschack1, Julian Riegger1, Wanwan Yu1, Samira Speicher1, Matthew Ho1, Nils Frank1 &
Tobias Ritter1*

Direct C–H functionalization can quickly increase useful structural targeted a reaction with a large absolute Hammett value, which could
and functional molecular complexity 1–3. Site selectivity can result from an endergonic radical reaction with a late transition state.
sometimes be achieved through appropriate directing groups or A large absolute Hammett value, by definition, would result in reaction
substitution patterns1–4—in the absence of such functionality, rates spanning several orders of magnitude for a broad spectrum of
most aromatic C–H functionalization reactions provide more electronically different arenes. A broad substrate scope, with resulting
than one product isomer for most substrates1,4,5. Development slow productive reaction rate for some arenes, requires that the reactive
of a C–H functionalization reaction that proceeds with high species not be consumed by other deleterious reaction pathways—such
positional selectivity and installs a functional group that can serve as hydrogen atom abstraction by a radical—even if the desired reaction
as a synthetic linchpin for further functionalization would provide with the substrate is slow. On the basis of this analysis, we evaluated
access to a large variety of well-defined arene derivatives. Here we electrophilic persistent sulfur-based radicals. A persistent radical does
report a highly selective aromatic C–H functionalization reaction not engage in deleterious side reactions to the extent that most radicals
that does not require a particular directing group or substitution do and could result in an endergonic radical reaction, and therefore a
pattern to achieve selectivity, and provides functionalized arenes late transition state.
that can participate in various transformations. We introduce a Thianthrene radical cations can be accessed in situ from the new
persistent sulfur-based radical to functionalize complex arenes thianthrene sulfoxide 1 (Fig. 2), which can be prepared on scale in
with high selectivity and obtain thianthrenium salts that are ready to two steps from 1,2-difluorobenzene and disulfur dichloride via
engage in different transformations, via both transition-metal and tetrafluorothianthrene 2 (TFT). As part of an extended study on the
photoredox catalysis. This transformation differs fundamentally reactivity of thianthrene in the 1970s, the reaction of isolated thian-
from all previous aromatic C–H functionalization reactions in that threnium radical cations with simple electron-rich arenes—such as
it provides direct access to a large number of derivatives of complex anisole—was evaluated, and the formation of arylthianthrenium cations
small molecules, quickly generating functional diversity with was observed14–16. However, the selectivity of the reactions was not
selectivity that is not achievable by other methods. investigated, and the synthetic utility of arylthianthreniums was
Bromo and boryl substituents are among the most useful linchpins not recognized, possibly owing to the reported incompatibility of
in organic chemistry, but introducing those substituents selectively can thianthrenium radical cations with many functional groups, such as
be problematic when molecules lack appropriate functional groups or pyridines17, amines18 and alcohols19. Instead of the moisture-sensitive,
substitution patterns. We evaluated two of the most selective transfor- explosive thianthrenium perchlorate radical salt prepared separately,
mations for bromination6 and borylation7 to show their selectivity for our new method uses the bench-stable, fluorinated sulfoxide-based
functionalization of ethylbenzene (Fig. 1). Electrophilic bromination is thianthrene reagent 1 which—together with the modified reaction
mostly controlled electronically and, although some substrates can be conditions—expands the substrate scope, probably owing to a higher
brominated selectively, even the most selective bromination reactions reduction potential and functional-group tolerance (see Supplementary
developed so far6,8–11 produce mixtures of isomers for most substrates Tables S2, S3). Moreover, we discovered that addition of thianthrene
(Fig. 1, Supplementary Table S6). Iridium-catalysed borylation reac- proceeds with excellent selectivity and that thianthreniums can func-
tions are mostly sterically controlled and can afford high selectivity tion as synthetically useful functional groups.
for certain substitution patterns, such as for 1,3-disubstituted arenes1. The tetrafluorothianthrene radical cation generated in situ
Of the few highly selective aromatic C–H functionalization reactions by comproportionation of 1 and 2 reacts chemoselectively to func-
that do not require particular directing groups or substitution patterns, tionalize arenes in preference to undergoing deleterious side reactions.
including our para-selective TEDAylation reaction12, none can intro- The high degree of chemoselectivity enables a large substrate scope
duce synthetic linchpins, and they are therefore not broadly useful (Fig. 2). Thianthrenation can proceed on arenes as electron-rich as
for introducing a variety of substituents. The C–H functionalization aniline derivatives to those as electron-poor as 1,2-dichlorobenzene.
reaction reported herein can proceed in >99% selectivity—not only Arenes that are more electron-rich than anisole undergo unproductive
for complex small molecules, but also for simple monosubstituted oxidation with the TFT-reagent 1, presumably through single-electron
arenes such as ethylbenzene—to afford novel synthetically useful oxidation. Therefore, we used an analogous non-fluorinated thian-
aryltetrafluorothianthrenium salts (Ar–TFT+; Fig. 1). threne reagent for all arenes more electron-rich than anisole, such as
Previously, we developed an aromatic C–H functionalization reac- meclofenamic acid (30) and famoxadone (31). Even more electron-rich
tion that gives access to a large variety of constitutional isomers to gen- arenes, such as indole, are not tolerated and undergo unproductive
erate structural diversity13. The design was based on achieving a highly oxidation; hence, they cannot be functionalized. However, once an
exergonic radical addition with an early transition state that, in accord electron-withdrawing group is installed, the desired reaction pathway
with Hammond’s postulate, would elicit little discrimination between proceeds (for example, 32). In addition, boronic acids are not tolerated
different positions of a given arene. Hammett analysis afforded a ρ in the reaction because they undergo ipso-substitution. The transforma-
value close to zero, as would be expected for a reaction with an early tion is highly tolerant of many other functional groups, such as amines,
transition state. To achieve highly selective arene functionalization, we amides, alcohols, ethers, esters, carboxylic acids and heterocycles.

1
Max-Planck-Institut für Kohlenforschung, Mülheim an der Ruhr, Germany. *e-mail: [email protected]

1 4 M A R C H 2 0 1 9 | V O L 5 6 7 | N A T U RE | 2 2 3
RESEARCH Letter

Versatile synthetic R Highly selective C−H

linchpins functionalizations

Cl
Me Me
Me Me F S F N+
Br
O O
B F S+ F N+
BF4–

Et
p:o = 6:1 Et Et F
p:m = 1:2 p:o > 500:1
p:o > 99:1
p:m > 200:1

FG

Fig. 1 | Selectivity of thianthrenation. Comparison of thianthrenation of isomers for the borylation of ethylbenzene. TEDAylation12 is highly site-
with halogenation6, borylation7 and TEDAylation12. FG, functional group selective, giving a para:ortho ratio of p:o > 99:1 for fluorobenzene (yellow
(for example, CO2Et, Cl, CN). Bromination6 and borylation7 produce area); however, aryl-TEDAs are of low synthetic value. Thianthrenation
synthetically useful arylbromides and arylboronic acids (blue area) as a 6:1 produces synthetically versatile arylthianthrenium salts in high selectivity
mixture of isomers for the bromination of ethylbenzene and a 1:2 mixture (p:o > 500:1, para:meta ratio of p:m > 200:1 for ethylbenzene; green area).

Alcohols are trifluoroacetylated under the reaction conditions but bias for regioselective electrophilic substitution reactions; for example,
trifluoroacetate esters hydrolyse during aqueous workup. In the case bromination of 1,2-dimethoxybenzene (3) and 2-methoxypyridine (28)
of sufficiently electron-rich arenes, olefins can be tolerated (for exam- proceed regioselectively20,21. However, regioselectivity is not achieved
ple, strychnine, 22). Basic functional groups are protonated under generally in arene C–H functionalizations, as shown in Fig. 1. In no
acidic reaction conditions and are thus protected from oxidation (for instance out of 48 evaluated cases was bromination more selective
example, 12, 21 and 27). HBF4OEt2 must sometimes be replaced by for the position of thianthrenation, and in 46 of those cases bromina-
Lewis acids—such as BF3OEt2 or TMSOTf—which enables success- tion was substantially less selective or afforded no product at all (see
ful functionalization of compounds that contain acid-sensitive func- Supplementary Table S6).
tionality, such as salicin pentaacetate (6), or of compounds that would The thianthrenium salts can behave as versatile linchpin electrophiles
otherwise be protonated and therefore deactivated for functionaliza- in both palladium-catalysed cross-coupling chemistry and photoredox
tion, such as electron-rich pyridine derivatives (for example, 28). The catalysis. We developed an initial set of 13 reactions, exemplified by
reaction is not sensitive to dioxygen or traces of water and can thus be the derivatization of the insecticide pyriproxyfen (16; Fig. 3a). Unlike
carried out under an ambient atmosphere in most cases. The majority alkylaryl sulfonium salts—which can engage in some cross-coupling
of the arylthianthrenium salts are soluble in organic solvents such as reactions22,23 but are strong alkylating reagents23—aryl thianthre-
acetonitrile, dichloromethane, 1,4-dioxane, dimethylformamide and nium salts are more resistant to nucleophiles such as cyanide, tertiary
dimethyl sulfoxide, and can be stored as solids under ambient condi- amines and organo zinc reagents, although treatment with strong bases
tions for years (see Supplementary Table S4). Although these salts can can result in benzyne formation24,25. The increased stability of
usually be used without purification (see below), chromatography on thianthrenium salts compared to alkylarylsulfonium salts towards
silica gel provides analytically pure compounds, as shown in Fig. 2. side-reactions makes them useful cross-coupling partners in many
High regioselectivity was observed in all cases, even for compounds palladium-catalysed reactions, such as Heck26, Sonogashira 27,
containing several reactive positions in different aromatic rings, such Negishi28, Suzuki29 and carbonylation30 reactions. The thianthre-
as in 2-fluorobiphenyl (5), meclofenamic acid (30), famoxadone (31) nium substituent can engage in faster oxidative addition reactivity
and bifonazole (7). The only substrate for which we could identify more than observed with bromide and triflate, which enables chemoselec-
than one isomer was mizolastine (27), which produced a mixture of tive cross-coupling (Fig. 3b). Palladium is inserted selectively into the
two products (ratio 16:1); the major isomer was obtained in pure form. desired of the three C–S bonds of the triarylsulfonium group of the
The selectivity of the reaction is assumed to be mostly dependent on thianthrenium salts—possibly a result of steric hindrance and rigidity
electronic effects, which can discriminate even between slightly dif- of the tricyclic thianthrene structure.
ferent positions (for example, 7) and can override steric biases (for In addition to palladium-catalysed cross-coupling chemistry, pho-
example, 26). Although thianthrenium is a large functional group, toredox catalysis31 enables the coupling of thianthrenium salts with
sterically hindered positions can be accessed—as exemplified by the several different nucleophiles at ambient temperature. For example,
thianthrenation of mesitylene (13)—presumably owing to the long C–S with nucleophiles such as chloride, trifluoromethylthiolate and triphe-
bond. The high selectivity for the most electron-rich position, com- nylphosphite, various carbon–heteroatom bonds can be accessed.
bined with the positive charge of the sulfonium salt, enables selective Photoredox functionalization of arylthianthrenium salts is also tolerant
monosubstitution, as illustrated by the selective functionalization of of arylhalides (for example, 51 and 52) and can access reactivity that
compound 29. Some of the arenes shown in Fig. 2 have an intrinsic is not currently achievable with aryl bromides, such as a Minisci-type

2 2 4 | N A T U RE | V O L 5 6 7 | 1 4 M A R C H 2 0 1 9
Letter RESEARCH

F S F
1–3 mol% 2 O
H
1.0 equiv. 1 F S F
3.0 equiv. (CF3CO)2O F S+ F F S F
1.1–4.5equiv. HBF4OEt2
F S F
F S F
MeCN, 0–25 °C, 1–24 h
R
R 1 2 (TFT)

H F
OMe
S OAc
OAc
OMe N
Br N
AcO O
O
H AcO
H Br
OAc
3 4 5 Salicin pentaacetate (6) H (±)-Bifonazole (7)
76%a 68% 85% 94%b 87%c

O
Me I
O H
H
O O N
N O O
Me
Et O Et NEt2
H I
Bu
H Br H
H H O
8 9 10 Atomoxetine (11) Amiodarone (12)
92% 86% 87% 58%a,d 92%
O O
H
S
Me NH N O

Me I O N O Me O
N F
H H N

H O H O

Me Me NO2
H O
13 14 Nimesulide (15) (±)-Pyriproxyphen (16) Ketanserin (17)
62%a 59% 93% 92% 60%

N
Me N
Me O H
H N NH H H
OTf N S
Cl N
O N N H O
O H Cl N N
H Me N H
H OH
Me O
H
18 19 Nefiracetam (20) Dasatinib (21) Strychnine (22)
86%c 66% 63%b 80%c 64%

Cl F

O
OEt
Cl O
HN
Cl H O
O H H N N
H N N
Me Me N N
O
H O Cl N Me
H
23 24 Boscalid (25) Etofenprox (26) Mizolastine (27)
20%c 64% 95% 84%a 62%g
OMe

OH O
HO O O MeO O
Cl
H H Me
N Me Me N NH
H N
O
N OMe O
H Cl O Cl
H
28 29 Meclofenamic acid (30) (±)-Famoxadone (31) H Indometacin methylester (32)
87%e 76%f 87%a 81%a 85%a

Fig. 2 | Substrate scope of thianthrenation. aNonfluorinated thianthrene d

The amino group was trifluoroacetylated. eTMSOTf was used instead of
S-oxide and nonfluorinated thianthrene were used instead of 1 and 2; no HBF4OEt2 and the reaction was carried out under an inert atmosphere
additional acid was used. The reaction was initiated at −78 °C. bBF3OEt2 with dry MeCN. f0.90 equiv. 1 was used. gSelectivity 16:1, isolated yield of
was used instead of HBF4OEt2. cTfOH was used instead of HBF4OEt2. major isomer. OTf, triflate; TMS, trimethylsilyl.

1 4 M A R C H 2 0 1 9 | V O L 5 6 7 | N A T U RE | 2 2 5
RESEARCH Letter

a Transition-metal-catalysed Photoredox catalysis

cross-couplings Borylation
Me
Me
Carbonylation Phosphonylation
O Me
O
B Me O
O OPh
OEt P
OPh
R
R
33, 76% R
Suzuki coupling 45, 91% Cyanation
34, 58%
N
a
b
l
R R
44, 90% 35, 69%
c
k
O N
n-C4H9 BF4–
SCF3
Me d
Sonogashira coupling Pseudohalogenation
j O TFT+
R
R
36, 53%
43, 95% O e
i 16a Cl
Alk
f
h R
R g
Negishi coupling 37, 53%
41, Alk = c-Pr, 59% Chlorination
I
42, Alk = Me, 78%
Ph O O
S R
Ph
R
38, 66%
40, 86% R
Heck reaction 39, 78%m Iodination

Sulfonylation

b
Competition with (pseudo-) halides in Suzuki coupling Minisci-type C–H arylation

OTf OTf
O
O
n BF4– q
N
BF4– +TFT Br
MeO Br
+TFT 9a
N 50, 86%
19a
46, 92%

Competition with aryliodides in photoredox catalysis

O I I
– n O
BF4
r
+TFT Br MeO
Br BF4–
9a
+TFT
47, 96% NC
14a 51, 83%
Functional-group tolerance NEt2
NEt2
I
N N I
O
O O
H H H O O
o EtO s
I
N H O N H O I Bu
H H Bu
Cl O
O H O
O
Amiodarone (12) 52, 50%, two steps
Strychnine (22) 48, 48%, two steps

Scalability (25 g scale) OMe

H
O O
O O

p OAc n OAc
NH Me N NH
Me N
O O AcO O
O OAc AcO O
AcO O OAc
O O AcO
OAc
OAc
(±)-Famoxadone (31) H 49, 62%, two steps CN Salicin pentaacetate (6) 53, 75%, two steps, 2 (TFT) recovered: 76%

Fig. 3 | See next page for caption.

2 2 6 | N A T U RE | V O L 5 6 7 | 1 4 M A R C H 2 0 1 9
Letter RESEARCH

Fig. 3 | Application of thianthrenation for functionalizing complex 50 °C. mNon-fluorinated thianthrenium salt used. nm-methoxyphenylboronic
arenes. a, Reaction scope of thianthrenium salts. Functionalization of acid (1.0–1.2 equiv.), K3PO4 (2.0–3.0 equiv.), Pd(dppf)Cl2, i-PrOH/dioxane
the pyriproxyfen-derived thianthrenium salt 16a. b, Compatibility of (1/1), 50 °C. o1) Thianthrenation, 2) CO (1.0 bar), N-methylmorpholine
thianthrenium salt reactions with aryl electrophiles, functional-group (2.0 equiv.), Pd(dppf)Cl2, EtOH, 50 °C. p1) Thianthrenation (non-
tolerance, Minisci-type reaction and scalability. aB2Pin2 (2.5 equiv.), fluorinated thianthrene), 2) NBu4CN (2.5 equiv.), Cu(MeCN)4BF4
pyridine (5.0 equiv.), LED. bP(OPh)3 (5.0 equiv.), pyridine (4.9 equiv.), (1.0 equiv.), MeCN/DMSO (2/1), LED. qK2CO3 (1.0 equiv.), [Ir{dF(CF3)
NaI (20 mol%), LED. cNBu4CN (2.5 equiv.), Cu(MeCN)4BF4 (1.2 equiv.), ppy}2(dtbpy)]PF6 (1 mol%), no Ru(bpy)3(PF6)2, DMSO/pyrazine (1/2),
LED. dNMe4SCF3 (1.1 equiv.), Cu(MeCN)4BF4 (1.0 equiv.), LED. eCuCl LED. rNBu4CN (3.3 equiv.), CuCN (2.3 equiv.), MeCN/DMSO (10/7),
(2.0 equiv.), NBu4Cl (2.5 equiv.), LED. fLiI (10 equiv.), Cu(MeCN)4BF4 LED. s1) Thianthrenation, 2) Cu(MeCN)4BF4 (1.6 equiv.), LiCl (4.8
(1.0 equiv.), MeCN/DMSO (3/2), LED. gPhSO2Na (3.1 equiv.), Pd2(dba)3 equiv.), CF3CO2H (1.2 equiv.), MeCN, LED. All photoredox processes:
(2.5 mol%), Xantphos (5 mol%), THF, 60 °C. hPd(OAc)2 (9.0 mol%), PPh3 LED (450 nm, 60 W), 22 °C, Ru(bpy)3(PF6)2 (2.0–4.0 mol%), MeCN,
(15 mol%), styrene (2.0 equiv.), NEt3 (3.0 equiv.), DMF, 100 °C. ic-PrZnBr 3–18 h, unless stated otherwise. All Pd-catalysed reactions: Pd-source
or MeZnCl (3.0 equiv.), Pd(PPh3)2Cl2, KOAc (6.0 equiv.), THF, 50 °C. (2.0–5.0 mol%), 16–48 h, unless stated otherwise. Alk, alkyl; dba,
j
1-hexyne (2.0 equiv.), CuI (20 mol%), Pd(dppf)Cl2, N-methylmorpholine dibenzylideneacetone; DMF, dimethylformamide; DMSO, dimethyl
(2.0 equiv.), dioxane, 40 °C. kCyclohexylvinylboronic acid (2.0 equiv.), sulfoxide; dppf, 1,1ʹ-bis(diphenylphosphino)ferrocene; dtbpy, 4,4′-
K2CO3 (4.0 equiv.), Pd(dppf)Cl2, EtOH, 50 °C. lCO (1.0 bar), di-tert-butyl-2,2′-dipyridyl; pin, pinacolato; ppy, phenylpyridyl; THF,
N-methylmorpholine (2.0 equiv.), Pd(dppf)Cl2, EtOH/dioxane (1/1), tetrahydrofuran.

pyrazine arylation reaction (Fig. 3b). We assume that the thianthrenium mostly by precipitation after the two-step reaction sequence to 53 on
salt is cleaved into thianthrene and aryl radicals after single-electron a 25-g scale.
reduction by a photosensitizer. The aryl radicals then abstract atoms We propose that thianthrenation proceeds in three steps: generation
or groups, react with copper complexes or add to arenes. of radical cations (I), formation of a dicationic adduct (II) and finally
Complex small molecules such as 49, 52 and 53 can be accessed irreversible deprotonation to form the product (Fig. 4a). The generation
directly from the parent C–H compounds in two steps, without puri- of radical cations by comproportionation of TFT S-oxide 1 and TFT
fication of the intermediate sulfonium salts, and TFT can be recovered 2 under the reaction conditions, observed by electron paramagnetic
after the reaction; for example, 76% of thianthrene 2 was recovered resonance spectroscopy (Fig. 4b), supports the proposed mechanism.

a
F S F
2 H

F S F
CF3CO2– R
I
F
O

S F TFT+
O CF3
F S+ F
F S F F S+ H –H+

F S F F
F S F
R
2
II R

b
(CF3CO)2O,
HBF4OEt2 F S F
1+2
MeCN, 25 °C
F S F
3,393 3,413 3,433 3,453
EPR
I
B (G)

c p:o ≈ 520:1
O p:m ≈ 230:1
F S F (CF3CO)2O,
HBF4OEt2
+
F S F
F S F MeCN,
Et 0 °C to 25 °C
1
F S+ F

Undivided cell
F S F electrolysis
HBF4OEt2 Et
+
F S F NBu4BF4,
MeCN, 0 °C p:o ≈ 500:1
Et 2 p:m ≈ 250:1

Fig. 4 | Proposed reaction mechanism and mechanistic experiments. S-oxide with arenes (reaction almost complete; top) and the purple
a, Proposed mechanism involving chain transfer. b, Comproportionation colour of TFT radical cations formed at the surface of the Pt anode
of TFT S-oxide and TFT under the reaction conditions (left); electron during electrolysis of TFT in the presence of HBF4OEt2, NBu4BF4 and
paramagnetic resonance spectrum (right; B, magnetic flux density). ethylbenzene in MeCN at 25 °C (bottom; the photograph was taken
c, Comparison of chemical and electrochemical thianthrenation. The approximately 10 s after turning the current on).
photographs show the typical colour of the reaction mixture of TFT
1 4 M A R C H 2 0 1 9 | V O L 5 6 7 | N A T U RE | 2 2 7
RESEARCH Letter

Moreover, TFT radicals could also be generated and observed by anodic 14. Shine, H. J. & Silber, J. J. Ion radical. XXII. Reaction of thianthrenium perchlorate
oxidation of TFT 2, and in the presence of arene the formation of TFT (C12H8S2•+ ClO4−) with aromatics. J. Org. Chem. 36, 2923–2926 (1971).
15. Kim, K., Hull, V. J. & Shine, H. J. Ion radicals. XXIX. Reaction of thianthrene cation
salts was observed to have selectivities similar to those observed under radical perchlorate with some benzene derivatives. J. Org. Chem. 39,
standard reaction conditions, providing evidence that both reactions 2534–2537 (1974).
proceed via the same reactive species (Fig. 4c). Radical formation from 16. Shin, S.-R. & Shine, H. J. Reactions of phenols with thianthrene cation radical.
J. Org. Chem. 57, 2706–2710 (1992).
1 can also be induced by reductants other than 2—for example, solvents 17. Shine, H. J., Silber, J. J., Bussey, R. J. & Okuyama, T. Ion radicals. XXV. The
or trace impurities—because radicals were also detected in the initial reactions of thianthrene and phenothiazine perchlorates with nitrite ion,
absence of 2. Under typical reaction conditions, radical concentrations pyridine, and other nucleophiles. J. Org. Chem. 37, 2691–2697 (1972).
18. Kim, K. & Shine, H. J. Ion radicals. XXX. Reactions of thianthrene cation
of more than 20% of the total TFT concentration were detected about radical perchlorate with amino compounds. J. Org. Chem. 39, 2537–2539
one hour after combining the reagents, and the deep purple colour (1974).
of such mixtures, which is indicative of I, did not fade away within 19. Shine, H. J. & Yueh, W. Reaction of thianthrene cation radical with alcohols:
one week at 25 °C. The inverse first-order dependence of the reac- cyclohexanols. Tetrahedron Lett. 33, 6583–6586 (1992).
20. Hirano, M., Monobe, H., Yakabe, S. & Morimoto, T. Kaolin-assisted aromatic
tion rate of thianthrene radical cations with anisole on the concentra- chlorination and bromination. J. Chem. Res. (S), 662–663 (1998).
tion of thianthrene15 rules out radical addition followed by hydrogen 21. Windscheif, P.-M. & Vögtle, F. Substituted dipyridylethenes and -ethynes and
atom abstraction. A small primary kinetic hydrogen isotope effect of key pyridine building blocks. Synthesis 1994, 87–92 (1994).
22. Srogl, J., Allred, G. D. & Liebeskind, L. S. Sulfonium salts. Participation par
1.7 and 1.9 for intra- and intermolecular competition experiments, excellence in metal-catalyzed carbon–carbon bond-forming reactions. J. Am.
respectively, together with this inverse first-order dependence, rules Chem. Soc. 119, 12376–12377 (1997).
out radical addition followed by deprotonation and subsequent oxida- 23. Minami, H., Otsuka, S., Nogi, K. & Yorimitsu, H. Palladium-catalyzed borylation
of aryl sulfoniums with diborons. ACS Catal. 8, 579–583 (2018).
tion. A Hammett analysis indicated that positive charge is accumulated 24. Kim, K. S., Ha, S. M., Kim, J. Y. & Kim, K. 5-arylthianthreniumyl perchlorates as
on the aromatic ring during the reaction, which supports a mecha- benzyne precursor. J. Org. Chem. 64, 6483–6486 (1999).
nism proceeding via a cationic intermediate such as II. The unusually 25. Yoon, K. & Ha, S. M. Kim, K. A convenient route to diverse heterocycles through
an addition of β-amino carbonyl compounds to 3-halo-4-methoxybenzenes.
steep slope of ρ = −11, obtained when using the σ+ values reported in J. Org. Chem. 70, 5741–5744 (2005).
ref. 32, means that the rate of reaction varies by more than nine orders 26. Heck, R. F. Palladium-catalyzed reactions of organic halides with olefins. Acc.
of magnitude between anisole and chlorobenzene. Dication II could be Chem. Res. 12, 146–151 (1979).
27. Sonogashira, K. Development of Pd–Cu catalyzed cross-coupling of terminal
formed from arene and I by radical addition and subsequent oxidation, acetylenes with sp2-carbon halides. J. Org. Chem. 653, 46–49 (2002).
by single-electron oxidation of arene followed by recombination with 28. Negishi, E.-I. Palladium- or nickel-catalyzed cross coupling. A new selective
radical I, or by addition of arene to a thianthrene dication, which could method for carbon-carbon bond formation. Acc. Chem. Res. 15, 340–348
result from disproportionation of I. Detailed additional mechanistic (1982).
29. Suzuki, A. Recent advances in the cross-coupling reactions of organoboron
investigations to distinguish between these three pathways for a better derivatives with organic electrophiles, 1995–1998. J. Organomet. Chem. 576,
appreciation of the source of selectivity are currently under investiga- 147–168 (1999).
tion in our laboratory. 30. Brennführer, A., Neumann, H. & Beller, M. Palladium-catalyzed carbonylation
reactions of aryl halides and related compounds. Angew. Chem. Int. Ed. 48,
4114–4133 (2009).
Online content 31. Prier, C. K., Rankic, D. A. & MacMillan, D. W. C. Visible light photoredox catalysis
Any methods, additional references, Nature Research reporting summaries, source with transition metal complexes: applications in organic synthesis. Chem. Rev.
data, statements of data availability and associated accession codes are available at 113, 5322–5363 (2013).
32. Brown, H. C. & Okamoto, Y. Electrophilic substituent constants. J. Am. Chem.
https://doi.org/10.1038/s41586-019-0982-0. Soc. 80, 4979–4987 (1958).
Received: 13 July 2018; Accepted: 23 January 2019;
Acknowledgements We thank M. van Gastel (Max Planck Institute for
Published online 13 March 2019. Chemical Energy Conversion) for the acquisition of electron paramagnetic
resonance spectra, as well as S. Marcus and D. Kampen (Max-Planck-Institut
1. Mkhalid, I. A. I., Barnard, J. H., Marder, T. B., Murphy, J. M. & Hartwig, J. F. C–H für Kohlenforschung) for mass spectrometry analysis. We thank G. Berger for
activation for the construction of C–B bonds. Chem. Rev. 110, 890–931 (2010). assistance with the construction of a photoreactor. We thank UCB Biopharma
2. Lyons, T. W. & Sanford, M. S. Palladium-catalyzed ligand-directed C–H and the Max-Planck-Institut für Kohlenforschung for funding.
functionalization reactions. Chem. Rev. 110, 1147–1169 (2010).
3. Leow, D., Li, G., Mei, T.-S. & Yu, J.-Q. Activation of remote meta-C–H bonds Reviewer information Nature thanks Kuangbiao Liao and the other anonymous
assisted by an end-on template. Nature 486, 518–522 (2012). reviewer(s) for their contribution to the peer review of this work.
4. Cheng, C. & Hartwig, J. F. Rhodium-catalyzed intermolecular C–H silylation of
arenes with high steric regiocontrol. Science 343, 853–857 (2014). Author Contributions F.B. designed reagent 1, developed the reaction chemistry
5. Olah, G. A. Aromatic substitution. XXVIII. Mechanism of electrophilic aromatic and investigated the mechanism. J.R., F.B. and M.H. explored the substrate
substitutions. Acc. Chem. Res. 4, 240–248 (1971). scope. F.B., M.B.P., W.Y. and J.R. optimized the cross-coupling and photoredox
6. Tang, R.-J., Milcent, T. & Crousse, B. Regioselective halogenation of arenes and reactions. M.B.P. investigated the selectivity of bromination. F.B., S.S., N.F. and
heterocycles in hexafluoroisopropanol. J. Org. Chem. 83, 930–938 (2018). J.R. developed the synthesis of reagent 1. T.R. and F.B. wrote the manuscript. T.R.
7. Saito, Y., Segawa, Y. & Itami, K. para-C–H borylation of benzene derivatives by a directed the project.
bulky iridium catalyst. J. Am. Chem. Soc. 137, 5193–5198 (2015).
8. Mo, F. et al. Gold-catalyzed halogenation of aromatics by N-halosuccinimides. Competing interests A patent application (number EP18204755.5, Germany),
Angew. Chem. Int. Ed. 49, 2028–2032 (2010). dealing with the use of thianthrene and its derivatives for C–H functionalization
9. Samanta, R. C. & Yamamoto, H. Selective halogenation using an aniline catalyst. and with reagent 1, has been filed and F.B. and T.R. may benefit from royalty
Chem. Eur. J. 21, 11976–11979 (2015). payments.
10. Maddox, S. M., Nalbandian, C. J., Smith, D. E. & Gustafson, J. L. A practical Lewis
base catalyzed electrophilic chlorination of arenes and heterocycles. Org. Lett. Additional information
17, 1042–1045 (2015). Supplementary information is available for this paper at https://doi.org/
11. Xiong, X., Tan, F. & Yeung, Y.-Y. Zwitterionic-salt-catalyzed site-selective 10.1038/s41586-019-0982-0.
monobromination of arenes. Org. Lett. 19, 4243–4246 (2017). Reprints and permissions information is available at http://www.nature.com/
12. Boursalian, G. B., Ham, W. S., Mazzotti, A. R. & Ritter, T. Charge-transfer-directed reprints.
radical substitution enables para-selective C–H functionalization. Nat. Chem. 8, Correspondence and requests for materials should be addressed to T.R.
810–815 (2016). Publisher’s note: Springer Nature remains neutral with regard to jurisdictional
13. Ham, W.-S., Hillenbrand, J., Jacq, J., Genicot, C. & Ritter, T. Divergent late-stage claims in published maps and institutional affiliations.
(hetero)aryl C–H amination by the pyridinium radical cation. Angew. Chem. Int.
Ed. 58, 532–536 (2019). © The Author(s), under exclusive licence to Springer Nature Limited 2019

2 2 8 | N A T U RE | V O L 5 6 7 | 1 4 M A R C H 2 0 1 9
Nature Research, brought to you courtesy of Springer Nature Limited (“Nature Research”)

Terms and Conditions

Nature Research supports a reasonable amount of sharing of content by authors, subscribers and authorised or authenticated users (“Users”), for
small-scale personal, non-commercial use provided that you respect and maintain all copyright, trade and service marks and other proprietary
notices. By accessing, viewing or using the nature content you agree to these terms of use (“Terms”). For these purposes, Nature Research
considers academic use (by researchers and students) to be non-commercial.

These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the terms, a site licence or a personal subscription (to the
extent of the conflict or ambiguity only). By sharing, or receiving the content from a shared source, Users agree to be bound by these Terms.

We collect and use personal data to provide access to the nature content. ResearchGate may also use these personal data internally within
ResearchGate and share it with Nature Research, in an anonymised way, for purposes of tracking, analysis and reporting. Nature Research will
not otherwise disclose your personal data unless we have your permission as detailed in the Privacy Policy.

Users and the recipients of the nature content may not:

1. use the nature content for the purpose of providing other users with access to content on a regular or large scale basis or as a means to
circumvent access control;
2. use the nature content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability,
or is otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by either Nature
Research or ResearchGate in writing;
4. use bots or other automated methods to access the nature content or redirect messages; or
5. override any security feature or exclusionary protocol.

These terms of use are reviewed regularly and may be amended at any time. We are not obligated to publish any information or content and may
remove it or features or functionality at our sole discretion, at any time with or without notice. We may revoke this licence to you at any time and
remove access to any copies of the shared content which have been saved.

Sharing of the nature content may not be done in order to create substitute for our own products or services or a systematic database of our
content. Furthermore, we do not allow the creation of a product or service that creates revenue, royalties, rent or income from our content or its
inclusion as part of a paid for service or for other commercial gain. Nature content cannot be used for inter-library loans and librarians may not
upload nature content on a large scale into their, or any other, institutional repository.

To the fullest extent permitted by law Nature Research makes no warranties, representations or guarantees to Users, either express or implied
with respect to the nature content and all parties disclaim and waive any implied warranties or warranties imposed by law, including
merchantability or fitness for any particular purpose.

Please note that these rights do not automatically extend to content, data or other material published by Nature Research that we license from
third parties.

If you intend to distribute our content to a wider audience on a regular basis or in any other manner not expressly permitted by these Terms
please contact us at

[email protected]
The Nature trademark is a registered trademark of Springer Nature Limited.