Orthogonal cross-coupling through intermolecular metathesis of unstrained C(aryl)–C(aryl) single bonds

Article abstract of DOI:10.1038/s41557-021-00757-4

While metathesis reactions involving carbon–carbon double bonds, namely olefin metathesis, have been well established with broad utility in organic synthesis and materials science, direct metathesis of kinetically less accessible C–C single bonds is extremely rare. Here we report a ruthenium-catalysed reversible C–C single-bond metathesis reaction that allows redox- and pH-neutral biaryl synthesis. Assisted by directing groups, unstrained homo-biaryl compounds undergo aryl exchanges to generate cross-biaryl products, catalysed by a well-defined air-stable ruthenium(II) complex. Functional groups reactive under typical cross-coupling reactions, such as halogen, silyl and boronate moieties, are compatible under the metathesis conditions. Mechanistic studies disclose an intriguing ‘olefin-metathesis-like’ pathway that involves an unexpected heptacoordinated, 18-electron closed-shell intermediate. The distinct reaction mode discovered here is expected to inspire the development of more general C–C single-bond metathesis and orthogonal cross-coupling reactions. [Figure not available: see fulltext.].

Full text of DOI:10.1038/s41557-021-00757-4

Articles
https://doi.org/10.1038/s41557-021-00757-4
Orthogonal cross-coupling through intermolecular
metathesis of unstrained C(aryl)–C(aryl)
single bonds
Jun Zhu1,2, Rui Zhang1,2 and Guangbin Dongꢀ ꢀ
1
ꢀ✉
While metathesis reactions involving carbon–carbon double bonds, namely olefin metathesis, have been well established with
broad utility in organic synthesis and materials science, direct metathesis of kinetically less accessible C–C single bonds is
extremely rare. Here we report a ruthenium-catalysed reversible C–C single-bond metathesis reaction that allows redox- and
pH-neutral biaryl synthesis. Assisted by directing groups, unstrained homo-biaryl compounds undergo aryl exchanges to gen-
erate cross-biaryl products, catalysed by a well-defined air-stable ruthenium(II) complex. Functional groups reactive under
typical cross-coupling reactions, such as halogen, silyl and boronate moieties, are compatible under the metathesis conditions.
Mechanistic studies disclose an intriguing ‘olefin-metathesis-like’ pathway that involves an unexpected heptacoordinated,
18-electron closed-shell intermediate. The distinct reaction mode discovered here is expected to inspire the development of
more general C–C single-bond metathesis and orthogonal cross-coupling reactions.
he past three decades have witnessed enormous advance- could be better described as an initiator, as the mechanism exhib-
ment of organic metathesis reactions, which are defined its chain-reaction character20. Inspired by our recent mechanistic
T
as the exchange of chemical bonds between similar reac- study of the Ru-catalysed reductive cleavage of unstrained C(aryl)–
tants to afford identical or similar bonds in products1. Among C(aryl) bonds22 (for selected examples and reviews of catalytic acti-
many chemical bonds that have been known to undergo metath- vation of non-polar and unstrained C–C bonds see refs. 16,17,23–28), in
esis reactions2–5, olefin metathesis (Fig. 1a), redistributing frag- which a Ru(II) monohydride species undergoes oxidative addition
ments of alkenes through cleaving and reforming carbon–carbon into the aryl–aryl bond to give a Ru(IV) intermediate that subse-
double bonds (C=C), represents one of the most powerful tools quently undergoes aryl–H reductive elimination and hydrogenoly-
with broad utility in organic synthesis and materials science6,7. sis to regenerate the Ru(II) catalyst, we questioned whether C–C
By contrast, C–C single bonds are typically considered inert and single bond metathesis could be realized when the hydride ligand
non-functional and rarely undergo direct metathesis reactions. on Ru(II) is replaced with an aryl ligand. The hypothesis is that,
In 1990, a nickel-catalysed ring-opening dimerization of biphe- using a Ru(II)−Ar species as the catalyst/initiator, the biaryl sub-
nylenes to access cyclooctatetraenes was reported by Vollhardt strate would first coordinate with Ru and undergo oxidative addi-
and co-workers8 and later the detailed mechanism, involving tion to cleave the C(aryl)–C(aryl) bond (Fig. 1b). The subsequent
metathesis-type rearrangement of C–C bonds, was disclosed by reductive elimination of the resulting triaryl–Ru complex29,30 could
Jones and co-workers9. As strain release was the essential driv- lead to new C(aryl)–C(aryl) bond formation and reform a Ru(II)−
ing force, this reaction was known to be irreversible. An inter- Ar species that could react with another biaryl substrate via a simi-
esting benzyl-exchange reaction of polar C(alkyl)–C(carbonyl) lar sequence. As a result, C–C single bond metathesis between two
single bonds in dibenzyl ketones was discovered by Arisawa biaryl species could be realized (Fig. 1c).
et al. in 2012, although the reaction mechanism remains unclear10.
On the other hand, efficient metathesis of alkanes, which still Results and discussion
mainly relies on olefin metathesis processes, has been realized Model study and density functional theory calculation. To test
with either heterogeneous11 or homogeneous catalysis12–15. To the this hypothesis, two biaryl substrates (1a and 2a) were employed
best of our knowledge, metathesis of unstrained and non-polar as the model substrates (Fig. 2a). An easily prepared and bench sta-
C–C single bonds16,17, such as bonds in biaryl compounds, has ble Ru(II) p-cymene complex (cat 1)31 that contains an aryl ligand
not been reported. In this Article, we describe our initial stud- similar to the monomers of the substrates was used as the precata-
ies of a Ru-catalysed reversible intermolecular metathesis reac- lyst30,32–35. After an extensive survey of various reaction conditions,
tion of unstrained C(aryl)–C(aryl) bonds. This reaction possesses the desired metathesis product (3a) was indeed obtained albeit only
an intriguing ‘olefin-metathesis-like’ pathway and provides an in 17% yield. Careful analysis of side-products revealed formation
unusual approach to accessing cross-biaryl compounds through of monomers (4a and 5a) via a formal reductive cleavage reaction
aryl exchanges of two homo-biaryl substrates.
and formation of oligomers (6–8 and so on) via ortho C–H cou-
The general Ru-catalysed olefin-metathesis mechanism includes plings36,37 between 1a and 2a or the aryl fragment on the catalyst.
steps of catalyst activation, substrate coordination, π-bond cleavage The production of these oligomers was likely to provide the hydro-
and re-formation, substrate dissociation and ‘catalyst’ regenera- gen source for the monomer (4a and 5a) formation. In addition,
tion (Fig. 1a)18–21. Considering that the regenerated Ru–alkylidene these oligomers were found to be active for the metathesis reaction
complex may not be the same as the initial one, the ‘catalyst’ (Supplementary Section 8.3).
1
2
✉
Department of Chemistry, University of Chicago, Chicago, IL, USA. These authors contributed equally: Jun Zhu, Rui Zhang. e-mail: gbdong@uchicago.edu
NatuRe ChemistRy | www.nature.com/naturechemistry

NATure CHeMIsTry
Articles
barrier. In the C–H activation pathway, the metallation occurs
through concerted metallation deprotonation (CMD) with assis-
tance of the neighbouring pivalate ligand (TS1a-CH). The gener-
ated PivOH does not dissociate and instead undergoes a retro-CMD
process to give a monomer product. The overall activation barrier is
33.7kcal mol−1, which is 2.1kcal mol−1 lower than the C–C activa-
tion pathway. This explains the experimentally observed low yield
of 3a and the formation of side products.
a
L
R
[Ru]
R1
R2
R2
R2
R1
R2
R1
R1
Catalyst
activation
–L
R
R
R2
[Ru]
[Ru]
R1
R1
R1
R1
R2
It was noted that in the C–H activation transition state
(TS1a-CH), the C4 (meta to the DG) C–H bond on the phenyl
ring is directly oriented towards the pyridine ring, while in TS1a
(the C–C activation transition state), with a backbone flip, the same
bond points outwards and no significant steric hindrance with Ru
is expected. Thus, we envisioned that the C–H activation pathway
could be inhibited by installing substituents at the C4 (meta to the
DG) position. This was further supported by DFT calculations on
the meta tert-butyl-substituted substrate 2b. The C–H activation
barrier with 2b is increased to 46.8kcal mol−1 and the transition
state (Fig. 2c, TS1a-CH-Bu) is highly distorted, in which the Rh−C
bond is elongated from 2.11 to 2.22Å and bent outwards by 12.3°.
Meanwhile, the C–C activation transition state is almost unaffected
(with a 35.4kcal mol−1 activation barrier), which makes C4-blocked
substrates suitable for further study. The C3-position blocked sub-
strates were also tested, but they could only give trace product (see
Supplementary Scheme 2.1 for details).
R1
R2
R1
R
R
R
[Ru]
R1
[Ru]
R1
[Ru]
R1
R1
R1
b
L
[Ru] Ar
Ar1
Ar1
Ar1
Ar2 Ar1
Ar2
Catalyst
activation
–L
[Ru] Ar
Ar1
Ar1
Ar2
Ar2
[Ru] Ar
Ar1
Ar1
Ar1
Ar2
Ar
Ar1
Ar
Ar1
[Ru]
Ar1
[Ru] Ar
Ar1
[Ru]
Ar1
Reaction optimization and additional mechanistic studies.
Indeed, the C4-substituted substrates 1b and 2b showed enhanced
metathesis reactivity (Table 1). Further, a modified Ru(II) complex
(cat 4), possessing an ortho-methylpyridyl meta-isopropylphenyl
ligand, was found to be a better precatalyst; the cross-dimer prod-
uct (3b) was obtained in good yield (entry 1). Given the revers-
ibility of the reaction (see below), the theoretical maximum yield
is 67.7%. Replacing the isopropyl group with a methyl group in
the catalyst (cat 3) gave a comparable result, but removal of the
isopropyl substituent (cat 2) or removal of both the isopropyl
and the methyl substituents (cat 1) or introducing an ortho OMe
group on the phenyl ligand (cat 5) resulted in slightly lower effi-
ciency (entries 2–5). Use of the corresponding Ru(II) iodide (cat
6) instead led to much lower yields (versus cat 2) (entry 6). Other
known ruthenium complexes, such as [RuCl2(p-cymene)]2 and
[Ru(OAc)2(p-cymene)]2, gave low yields while a Ru(0) complex,
Ru(1,5-cod)(1,3,5-cot) (cod, cyclooctadiene; cot, cyclooctatriene),
was not reactive (entries 7–9). AgOPiv was found to be a critical
additive; only 12% yield was observed without AgOPiv (entry 10).
Ar1
Me
Me
c
Me
Ru
Ar1
Ar2
Ar1
Ar2
Ar2
Cl
Ar
[Ru] cat.
DG1
DG1
DG2
DG2
DG
DG1
DG2
+
DG = N-heterocycles
Ar1
Fig. 1 | metathesis reactions of carbon−carbon double and single bonds.
a, The general process of a Ru-catalysed olefin metathesis is described,
involving a Ru–alkylidene intermediate. L, ligand. b, A strategy of C–C single
bond metathesis is proposed, which is analogous to the olefin-metathesis
mechanism. c, This work describes a Ru-catalysed metathesis of
unstrained, non-polar C(aryl)–C(aryl) single bonds. Cat., catalyst.
To better understand the competition between the C–H and C–C By contrast, use of less basic AgOAc or AgTFA (TFA, trifluoroac-
activation pathways and to provide insights on further optimiza- etate) decreased the yields (entries 11 and 12); low coordinative
−
−
tion, density functional theory (DFT) calculations were conducted. anions, such as SbF6 and BF4 , shut down the reactivity, which is
The energy profiles for both pathways are shown in Fig. 2b. First, consistent with the DFT-calculated role of the pivalate anion. While
cat0 is generated via ligand substitution between the precatalyst and 20mol% of the catalyst was found to be optimal (entries 13 and 14),
the AgOPiv (OPiv, pivalate) additive. The replacement of p-cymene considering that both substrates contain nearly identical C–C bonds
by substrate 2a is highly exothermic and gives a pair of isomers to be activated, the ‘actual’ catalyst loading based on the total reac-
Int0a and Int0b as the resting states. The two pyridine directing tive moieties is 6.5mol%. Besides toluene, fluorobenzene was also
groups (DGs) in both Int0a and Int0b adopt a cis geometry and no an excellent solvent (entry 15). While increasing the concentration
significant interaction between the C(aryl)–C(aryl) bonds and Ru did not significantly affect the yield, diluted conditions reduced the
is observed, possibly because of the bulkier coordination environ- reaction conversion (entries 16 and 17). Optimal yield was obtained
ment. In the C–C activation pathway, the pivalate ligand partially with a ratio of 1:2.1 between the two reactants (entries 18 and 19).
2
1
dissociates from an η to an η coordination, leaving room for the The somewhat diminished efficiency with excess 2b could be due to
agnostic C–C/Ru interaction. The subsequent oxidative addition a lower ‘actual’ catalyst loading based on the total reactive moieties.
(TS1a) is the rate-determining step with an overall activation barrier Finally, adding either substrate in portions did not much alter the
of 35.8kcal mol−1, which leads to a heptacoordinated, 18-electron reaction outcomes (entries 20 and 21).
closed-shell intermediate Int2a (ref. 38); the structures without OPiv
With the optimized conditions in hand, some mechanistic
or pyridine coordination are much less stable (see Supplementary aspects of the reaction were further explored. First, consistent
Scheme 9.1 for details). The following reductive elimination with the DFT calculation, the p-cymene ligand on the Ru(II)
(TS2a) is the reverse process of TS1a, though the two DGs are now precatalyst was found to be innocent and to quantitatively dis-
trans to each other, which accounts for the much lower activation sociate (Supplementary, Section 8.1). Second, the reversibility of
NatuRe ChemistRy | www.nature.com/naturechemistry

NATure CHeMIsTry
Articles
a
DG1
DG1
1
Side products:
H
DG2
+ other oligomers
DG1
Cat 1 (20 mol%)
AgOPiv (20 mol%)
2
DG1
4′
4
+
1′
2
2
Tol. 130 °C
24 h
2′
DG2
DG2
2a
DG1
DG2
3a, 17%
4a
1a
DG2
Me
N
DG1
DG2
Me
DG1
Me
H
Me
Ru
DG1
=
DG2
=
Cl
N
DG2
DG2
F
N
N
6
7
8
Cat 1
5a
Ar
Ar
Ar
b
∆Gsol in kcal mol–1
F
N
H
N
F
O
(∆Hsol
)
N
Ru
N
F
H
tBu
N
Ru
O
O
N
Ru
O
O
Ar =
Ar
tBu
Ar
N
tBu
H
O
TS1a-CH
13.9
(8.3)
TS1a
F
Int1a-CH
TS2a-CH
N
F
iPr
F
N
HO
N
Ru
9.8
(8.4)
N
Ru
N
Ru
O
8.4
(5.5)
tBu
O
O
O
O
tBu
Int1a-CH
H
TS1a-CH
11.8
TS2a-CH
tBu
Int2a-CH
cat0
(9.7)
Int3a-CH
DG1
DG1
–0.9
(–3.8)
0.0
Ru
O
(0.0)
–4.3
(–6.3)
Int2a-CH
DG2
Ar
N
O
TS2a
0.1
(–6.5)
DG2
O
2a + cat0
DG2
Ru
Int3a-CH
tBu
Int4b-CH
N
Ru
DG2
1a
Int3a
–4.6
(–9.3)
Int2a
–4.4
OPiv
O
N
tBu
(–9.6)
TS1a
DG2
Ru
Int0a
1a
Monomer
–18.6
(–22.7)
Int0a
DG2
OPiv
Int1a
–14.8
(–19.7)
N
N
–22.8
DG2
Ru
DG2
3a
Int3a
(–28.4)
Monomer
Int4b-CH
Int0b
Ru
OPiv
DG2
DG2
OPiv
–21.9
N
DG1
Int4b
–22.8
(–27.6)
DG1
N
(–26.1)
DG2
DG2
DG2
O
Int2a
TS2a
Ru
O
Ru DG2
O
Ru
N
OPiv
Int1a
N
tBu
N
O
tBu
Int4b
Int0b
F
c
2.27
2.33
2.29
1.67
1.66
2.11
2.22
2.38
112.8°
100.5°
TS1a
DG1 = 35.8 kcal mol–1
TS1a-CH
DG1 = 33.7 kcal mol–1
TS1a-Bu
DG1 = 35.4 kcal mol–1
TS1a-CH-Bu
DG1 = 46.8 kcal mol–1
Fig. 2 | exploration of a Ru-catalysed biaryl metathesis reaction. a, The initial reaction discovery. The desired product was formed in 17% yield, with the
formation of some monomer and oligomer side-products from C–H activation pathways. b, Computed energy profiles of the Ru-catalysed C–C and C–H
activation pathways at the M06/SDD-6-311ꢀ+ꢀG(d,p)/SMD(toluene)//B3LYP/LANL2DZ-6-31G(d) level of theory. The C–C activation pathway is in black
and the C–H activation pathway is in red. The rate-determining step is the oxidative addition, with a heptacoordinated transition state ts1a. For substrate
2a, the C–H activation is slightly favoured over the C–C activation. c, Optimized structures of the C–C and C–H activation transition states of substrates 2a
and 2b. Distances are in Å. The additional tBu group in 2b destabilizes the C–H activation transition state and the C–C activation pathway is thus favoured.
Tol, toluene.
the metathesis reaction was further investigated. Both forward These results support the conclusion that the metathesis reac-
and reverse reactions were conducted under the optimal condi- tion is reversible and that the equilibrium of the reaction could be
tions (Fig. 3a,b). After 24h, the two reactions gave a mixture of reached. Moreover, kinetic monitoring of the model reaction under
the same homo-dimers and cross-dimers in very similar ratios. the standard conditions showed that the reaction exhibited no
NatuRe ChemistRy | www.nature.com/naturechemistry

NATure CHeMIsTry
Articles
(TMS) moieties (3f–3h). The TMS group can potentially be removed
or transformed to other functional groups through reactions with
electrophiles. Only trace metathesis product was observed with-
out the C4 substituent when simple pyridine was used as the DG
(3e); however, use of 3-methylpyridyl as the DG (DG1) instead sig-
nificantly increased the yield for those with a small C4 substituent
(3i and 3j)39,40. Based on DG1, the functional group tolerance was
examined. Electron-donating OMe (3m) and CH2OBn (3n) per-
formed well and moderate yields were observed for substrates that
contain terminal and internal olefins (3o and 3p). Phenyl group
(3q), redox-active chloride (3r), bromide (3s) and iodide (3t) and
electron-withdrawing groups such as ester (3v) and amide (3w) at
the C4 position, were compatible. The lower conversions to prod-
ucts 3o, 3p and 3w were likely to be due to the coordinative capabil-
ity of the olefin and amide moieties, which hindered the metathesis
process to some extent. Notably, homo-biaryls with TMS and bro-
mide substituents still reacted to afford the metathesis product that
contains both a nucleophilic and an electrophilic moiety (3u).
The effect of using different DGs has also been investigated.
Introducing a C5-ester moiety on the DG reduced the overall effi-
ciency (3y), although the exact reason is unclear. By contrast, the
C3-F-pyridyl group gave a remarkably improved yield (3z). Finally,
3-fluoro-4-methylpyridine (DG3) was found to be most efficient
and the yield of the metathesis product (3aa) reached 63%, which
is close to the theoretical statistical yield (67.7%) without consider-
ing the reactivity difference between the reactants and the product.
Note that keeping a fluorine substituent on the DG of the excess
reactant (2) was important for easy separation of cross-biaryl 3
from homo-biaryl 2. Use of other DGs instead of 5-fluoropyridine
resulted in lower efficiency (3ab–3ad). Gratifyingly, substitution
at the C5′ position (para to the DGs) can be tolerated (3ae–3ah).
Functional groups including ester (3v, 3y), methyl (3ae) and ben-
zyl (3af) ethers, furan (3ah) and ketone (3ai) at the C5′ position
were also compatible. Besides pyridine, isoquinoline also served as
a good DG (3aj). While the C3 and C6 substitutions were generally
inhibitive, a C3-methyl-C5-fluoro-substituted substrate surpris-
ingly gave the desired product (3ak) in decent yield. Notably, in
certain cases (3k, 3l, 3y, 3ac, 3ai, 3ak), the less bulky cat 2 exhibited
better reactivity than cat 4.
Table 1 | selected optimization studies
t Bu
2
Cat 4 (20 mol%)
AgOPiv (20 mol%)
Py
DG2
DG2
+
Tol. (0.7 ml)
130 °C, 24 h
Standard conditions
(theoretical max. 67.7%)
Py
tBu
t Bu
1b
2b
3b
t Bu
Entrya
Change from the ‘standard’ condition
None
Yieldb
54% (50%)
50%
Mass balancec
82%
1
2
Cat 1 instead of cat 4
89%
3
Cat 2 instead of cat 4
48%
82%
4
Cat 3 instead of cat 4
53% (48%)
44%
80%
5
Cat 5 instead of cat 4
91%
6
Cat 6 instead of cat 4
18%
19%
86%
82%
7
[RuCl2(p-cymene)]2 instead of cat 4
[Ru(OAc)2(p-cymene)]2 instead of cat 4
Ru(1,5-cod)(1,3,5-cot) instead of cat 4
No AgOPiv
8
14%
85%
9
Trace
12%
81%
10
11
12
13
14
15
16
17
18
19
20
21
77%
AgOAc instead of AgOPiv
AgTFA instead of AgOPiv
10 mol% Cat 4 + 10 mol% AgOPiv
30 mol% Cat 4 + 30 mol% AgOPiv
PhF as solvent
48%
79%
31%
34%
68%
93%
46%
70%
52%
83%
0.5 ml Tol.
52%
82%
2.0 ml Tol.
34%
88%
1.5 equiv. 2b
2.4 equiv. 2b
1b added in three portions over 8 h
2b added in three portions over 8 h
46%
48%
84%
79%
48%
80%
46%
72%
Me
Me
Me
Me
Me
Me
Me
Me
Me
Ru
Ru
Ru
Cl
Cl
Cl
N
N
N
Me
Cat 3
Cat 1
Cat 2
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Me
Aryl bromides and boronates are among the most common reac-
tive moieties used in various types of cross-coupling reaction, thus
cross-biaryl synthesis that can tolerate both bromide and boronate
functional groups is extremely rare. The orthogonal reactivity of the
biaryl metathesis approach can be seen in the cross metathesis of
the di-bromo and di-pinacolborane substrates (Fig. 4a). Under the
standard conditions except at an elevated temperature, the desired
cross-biaryl product was formed in good yield. In addition, prelimi-
nary success has been achieved using pyrazoles as DGs. These are
Ru
Ru
Ru
Cl
I
Cl
N
N
N
iPr
OMe
Me
Me
Cat 4
Cat 5
Cat 6
aUnless otherwise noted, the reaction was run with 1b (0.05ꢀmmol), 2b (0.105ꢀmmol), cat 4
(0.01ꢀmmol) and AgOPiv (0.01ꢀmmol). bDetermined by 1H NMR using 1,1,2,2-tetrachloroethane
as internal standard; the numbers in parentheses are isolated yields. cMass balance includes the
remaining starting materials, the metathesis products and the monomers generated. Tol, toluene.
Py, 2-pyridyl.
known to be removable and can serve as an amine surrogate25,41
.
Reaction of pyrazole-containing biaryl 1aa with stoichiometric cat
7 and AgOPiv afforded the aryl-transfer product 3am in 40% NMR
yield (Fig. 4b). Moreover, cross-metathesis reactivity was observed
induction period. This is in line with the fast dissociation of between biaryl 1aa and pyridine-based substrate 2b at higher reac-
p-cymene. The reaction was fast in the first 5h with an initial rate tion temperatures. Efforts to understand the reactivity difference
of 11.2mM h−1 but started to decrease afterwards. The equilibrium between various DGs and improve reactivity with non-pyridine
was reached after 24h (Fig. 3c).
DGs through better catalyst design are ongoing.
Substrate scope. The substrate scope was next explored (Table 2). Conclusion
Key questions include what substitution patterns/functional groups Insummary, thediscoveryofareversibleC–Csinglebondmetathesis
could be tolerated and what DGs could be more effective. First, the of unstrained biaryl compounds is described. This provides a unique
steric hindrance on the phenyl C4/4′ positions can indeed inhibit the approach to accessing cross-biaryls in a redox- and pH-neutral man-
undesired C–H activation pathway. For example, replacing the C4 ner. The reaction is catalysed by a well-defined air-stable Ru catalyst,
tBu group with a CF3 or CH3 group in the limiting substrate 1 gave involving an ‘olefin-metathesis-like’ pathway. The involvement of a
notably reduced yields of cross-biaryl products (3c and 3d) and for- heptacoordinated, 18-electron closed-shell intermediate, disclosed
mation of the corresponding monomers has been observed; by con- by the computational study, could inspire future catalyst design.
trast, good reactivity was observed for those bearing trimethylsilyl Orthogonal reactivity to typical cross-couplings has been illustrated
NatuRe ChemistRy | www.nature.com/naturechemistry

NATure CHeMIsTry
Articles
Table 2 | substrate scope of the biaryl metathesis reaction
Me
Me
4′
Me
Ar1
Ar1
Ar2
Ar2
Ar2
Cat 4 (20 mol%)
AgOPiv (20 mol%)
Ru
2′
DG
DG
DG2
DG2
DG2
DG
1′
Cat 4
+
Cl
1
Tol. 130 °C
24 h
DG = N-heterocycles
N
2
iPr
Ar1
5
1
2
3
4
Me
tBu
tBu
tBu
tBu
tBu
tBu
TMS
TMS
Py
DG2
DG1
DG2
DG2
DG2
N
DG2
DG2
DG2
N
DG2
Py
Py
Py
N
Py
Py
Py
5
F
3
Me
tBu
TMS
tBu
Me
H
TMS
H
3i, 27%
[83%b]
CF3
3b, 50%
[82%b
3c, 34%
[72%b]
3d, 13%
[44%b]
3e, trace
3f, 45%
[84%b]
3g, 38%
[85%b]
3h, 36%
[77%b]
]
tBu
tBu
tBu
tBu
tBu
tBu
tBu
tBu
DG2
DG2
DG2
DG2
DG2
DG2
DG2
DG2
DG1
DG1
DG1
DG1
DG1
DG1
DG1
DG1
Me
tBu
OMe
Ph
Me
Me
OBn
3n, 45%
[70%b]
3j, 52%
[74%b]
3k, 45%b
[75%b]
3l, 50%a
[76%b]
3m, 45%
[95%b]
3o, 23%
[74%b]
3p, 23%
[77%b]
3q, 38%
[55%b]
tBu
tBu
tBu
tBu
tBu
tBu
TMS
DG2
DG2
DG2
DG2
DG2
DG2
DG2
DG1
DG1
DG1
DG1
DG1
DG1
DG1
F
Br
Cl
Br
I
CO2Me
Me
O
N
3r, 43%
[68%b]
3s, 37%
[66%b]
3t, 34%
[66%b]
3u, 31%
[55%b]
3v, 53%
[92%b]
3w, 29%
[62%b]
3x, 43%
[79%b]
tBu
tBu
tBu
tBu
tBu
tBu
DG3
F
F
E
5
N
DG2
N
DG2
N
DG2
DG3
N
DG3
N
DG3
N
Me
tBu
F
Cl
CF3
E = COOMe
F
F
tBu
tBu
tBu
3aa, 63%
[89%b]
tBu
tBu
tBu
3y, 41%a
[90%b]
3ac, 50%a
[96%b]
3z, 62%
[91%b]
3ab, 40%
[85%b]
3ad, 58%
[90%b]
tBu
tBu
tBu
tBu
tBu
tBu
MeO
BnO
Ph
O
DG2
DG2
DG2
DG2
DG2
N
DG2
DG2
DG3
DG3
DG3
DG3
DG3
Py
Me
Me
F
tBu
tBu
tBu
tBu
tBu
O
tBu
3ae, 61%
[93%b]
3af, 42%
[82%b]
3ag, 40%
[82%b]
3ah, 40%
[67%b]
3ai, 50%a
[88%b]
3aj, 51%
[80%b]
3ak, 42%a
[67%b]
Unless otherwise noted, the reaction was run with biaryl 1 (0.05ꢀmmol), biaryl 2 (0.105ꢀmmol), cat 4 (0.01ꢀmmol) and AgOPiv (0.01ꢀmmol); all yields are isolated yields. Theoretical statistical yield is 67.7%.
acat 2 (0.01ꢀmmol) was used instead of cat 4. bMass balance includes the remaining starting materials, the metathesis products and the monomers generated. Tol, toluene. Bn, benzyl.
NatuRe ChemistRy | www.nature.com/naturechemistry

NATure CHeMIsTry
Articles
a
tBu
c
100
80
60
40
20
0
F
F
N
2
N
2
Standard
conditions
F
Me
N
Me
1b
3b
+
N
F
tBu
1w
0.050 mmol
0.014 mmol
tBu
2b
3aa
–
tBu
0 h
24 h
0.050 mmol
0.017 mmol
0.043 mmol
tBu
b
F
F
N
N
2
2
Standard
conditions
F
Me
N
Me
+
N
F
tBu
tBu
0
5
10
15
20
25
30
35
40
3aa
0.10 mmol
1w
–
2b
–
tBu
Time (h)
0 h
24 h
0.049 mmol
0.012 mmol
0.016 mmol
Fig. 3 | Preliminary mechanistic study. a, Forward reaction—the metathesis reaction using homo-biaryls 1w and 2b as the substrates. b, Reverse reaction—
the metathesis reaction using hetero-biaryl 3aa as the substrate. Similar product distributions with these two reactions were observed, which indicates
that the reaction is reversible. c, Kinetic monitoring of the model reaction with substrates 1b and 2b under the standard conditions, which shows that the
equilibrium was reached after 24ꢀh. The study was only carried out once. For raw data, see Supplementary Section 6. Standard conditions: 1 (0.05ꢀmmol), 2
(0.105ꢀmmol), cat 4 (20ꢀmol%, 0.01ꢀmmol), AgOPiv (20ꢀmol%, 0.01ꢀmmol) in 0.7ꢀml toluene, 130ꢀ°C, 24ꢀh.
a
DG1
Bpin
OH
Fig. 4 | Further exploration of the reaction scope and limitation. a,
Orthogonal reactivity shown in the synthesis of the cross-biaryl product
that contains both bromide and pinacol boronate moieties. b, Attempts at
utilizing pyrazoles as removable DGs with stoichiometric Ru cat 7. c, Cross
metathesis of pyrazole-based substrate 1aa with pyridine-based substrate
2b. Standard conditions: 1 (0.05ꢀmmol), 2 (0.105ꢀmmol), cat 4 (20ꢀmol%,
0.01ꢀmmol), AgOPiv (20ꢀmol%, 0.01ꢀmmol) in 0.7ꢀml toluene, 130ꢀ°C, 24ꢀh.
Br
2
Standard
NaOH
H2O2
conditions
1p
DG2
DG2
0.105 mmol
except at
160 °C
DG1
DG1
Bpin
3al
2
45% over
two steps
Br
Br
DG2
2k
0.05 mmol
by the tolerance of various redox-active functional groups such as
those in aryl bromides, iodides, silyls and boronates. While at the
current stage this work only presents an initial fundamental study,
given the redox/pH-neutral reaction conditions and high chemose-
lectivity, as well as easy access to homo-biaryls42, one may imagine
that practical applications could ultimately be found when common
functional groups are employed as DGs or handles. Moreover, the
distinct mode of activation uncovered here could have broad impli-
cations in developing other C–C single-bond metathesis reactions.
iPr
b
Cat 7 (100 mol%)
AgOPiv (100 mol%)
Tol. 180 °C, 24 h
N
N
2
N
N
N
Me
N
Me
tBu
Me
Ru
1aa
tBu
3am
40% NMR yield
Cl
Online content
N
N
iPr
Any methods, additional references, Nature Research report-
ing summaries, source data, extended data, supplementary infor-
mation, acknowledgements, peer review information; details of
author contributions and competing interests; and statements of
data and code availability are available at https://doi.org/10.1038/
s41557-021-00757-4.
Cat 7
c
N
N
tBu
tBu
tBu
2
Received: 19 November 2020; Accepted: 22 June 2021;
Published: xx xx xxxx
1aa
0.05 mmol
Cat 7 (20 mol%)
DG2
N
N
N
N
DG2
AgOPiv (20 mol%)
Tol. 200 °C
24 h
+
+
References
1. McNaught, A. D. & Wilkinson, A. IUPAC Compendium of Chemical
Terminology: Te Gold Book 2nd edn (Blackwell Scientifc, 1997);
https://doi.org/10.1351/goldbook
tBu
2
tBu
3an
8.2%, based on 1aa
iPr
3ao
37%, based on cat 7
DG2
2. Fürstner, A. & Davies, P. W. Alkyne metathesis. Chem. Commun.
2307–2320 (2005).
2b
0.105 mmol
NatuRe ChemistRy | www.nature.com/naturechemistry

NATure CHeMIsTry
Articles
25. Onodera, S., Ishikawa, S., Kochi, T. & Kakiuchi, F. Direct alkenylation of
allylbenzenes via chelation-assisted C–C bond cleavage. J. Am. Chem. Soc.
140, 9788–9792 (2018).
3. Zhang, W. & Moore, J. S. Alkyne metathesis: catalysts and synthetic
applications. Adv. Synth. Catal. 349, 93–120 (2007).
4. Diver, S. T. & Gies sert, A. J. Enyne metathesis (enyne bond reorganization).
Chem. Rev. 104, 1317–1382 (2004).
26. Zhu, J., Wang, J. & Dong, G. Catalytic activation of unstrained C(aryl)–
C(aryl) bonds in 2,2′-biphenols. Nat. Chem. 11, 45–51 (2019).
27. Onodera, S., Togashi, R., Ishikawa, S., Kochi, T. & Kakiuchi, F. Catalytic,
directed C–C bond functionalization of styrenes. J. Am. Chem. Soc. 142,
7345–7349 (2020).
5. Bhawal, B. N. & Morandi, B. Catalytic isofunctional reactions—expanding the
repertoire of shuttle and metathesis reactions. Angew. Chem. Int. Ed. 58,
10074–10103 (2019).
6. Schuster, M. & Blechert, S. Olefn metathesis in organic chemistry. Angew.
Chem. Int. Ed. 36, 2036–2056 (1997).
28. Song, F., Gou, T., Wang, B.-Q. & Shi, Z.-J. Catalytic activations of unstrained
C–C bond involving organometallic intermediates. Chem. Soc. Rev. 47,
7078–7115 (2018).
7. Ogba, O. M., Warner, N. C., O’Leary, D. J. & Grubbs, R. H. Recent advances
in ruthenium-based olefn metathesis. Chem. Soc. Rev. 47, 4510–4544 (2018).
8. Schwager, H., Spyroudis, S. & Vollhardt, K. P. C. Tandem palladium-, cobalt-,
and nickel-catalyzed syntheses of polycyclic π-systems containing
cyclobutadiene, benzene, and cyclooctatetraene rings. J. Organomet. Chem.
382, 191–200 (1990).
29. Saavedra-Díaz, O., Cerón-Camacho, R., Hernández, S., Ryabov, A. D. & Le
Lagadec, R. Denial of tris(C,N-cyclometalated) ruthenacycle: nine-membered
η -N,N-trans or η -N,N-cis RuII chelates of 2,2′-bis(2-pyridinyl)-1,1′-biphenyl.
6
2
Eur. J. Inorg. Chem. 2008, 4866–4869 (2008).
9. Edelbach, B. L., Lachicotte, R. J. & Jones, W. D. Mechanistic investigation of
catalytic carbon–carbon bond activation and formation by platinum and
palladium phosphine complexes. J. Am. Chem. Soc. 120, 2843–2853 (1998).
10. Arisawa, M., Kuwajima, M., Toriyama, F., Li, G. & Yamaguchi, M.
Rhodium-catalyzed acyl-transfer reaction between benzyl ketones and
thioesters: synthesis of unsymmetric ketones by ketone CO–C bond cleavage
and intermolecular rearrangement. Org. Lett. 14, 3804–3807 (2012).
11. Basset, J.-M., Coperet, C., Soulivong, D., Taoufk, M. & Cazat, J. T. Metathesis
of alkanes and related reactions. Acc. Chem. Res. 43, 323–334 (2010).
12. Goldman, A. S. et al. Catalytic alkane metathesis by tandem alkane
dehydrogenation-olefn metathesis. Science 312, 257 (2006).
13. Ahuja, R. et al. Catalytic ring expansion, contraction, and
metathesis-polymerization of cycloalkanes. Chem. Commun. 253–255 (2008).
14. Jia, X., Qin, C., Friedberger, T., Guan, Z. & Huang, Z. Efcient and selective
degradation of polyethylenes into liquid fuels and waxes under mild
conditions. Sci. Adv. 2, e1501591 (2016).
30. Rogge, T. & Ackermann, L. Arene-free ruthenium(II/IV)-catalyzed bifurcated
arylation for oxidative C-H/C-H functionalizations. Angew. Chem. Int. Ed. 58,
15640–15645 (2019).
31. Djukic, J.-P. et al. Syntheses of nonracemic ortho-mercurated and
6
ortho-ruthenated complexes of 2-[tricarbonyl(η -phenyl)chromium]pyridine.
Organometallics 23, 5757–5767 (2004).
32. Hofmann, N. & Ackermann, L. Meta-selective C–H bond alkylation with
secondary alkyl halides. J. Am. Chem. Soc. 135, 5877–5884 (2013).
33. Moselage, M., Li, J., Kramm, F. & Ackermann, L. Ruthenium(II)-catalyzed
C–C arylations and alkylations: decarbamoylative C–C functionalizations.
Angew. Chem. Int. Ed. 56, 5341–5344 (2017).
34. Ruan, Z. et al. Ruthenium(II)-catalyzed meta C–H mono- and
difuoromethylations by phosphine/carboxylate cooperation. Angew. Chem.
Int. Ed. 56, 2045–2049 (2017).
35. Wang, X.-G. et al. Tree-component ruthenium-catalyzed direct
meta-selective C-H activation of arenes: a new approach to the alkylarylation
of alkenes. J. Am. Chem. Soc. 141, 13914–13922 (2019).
36. Arockiam, P. B., Bruneau, C. & Dixneuf, P. H. Ruthenium(II)-catalyzed C–H
bond activation and functionalization. Chem. Rev. 112, 5879–5918 (2012).
37. Yang, Y., Lan, J. & You, J. Oxidative C–H/C–H coupling reactions between
two (hetero)arenes. Chem. Rev. 117, 8787–8863 (2017).
38. Duan, L., Fischer, A., Xu, Y. & Sun, L. Isolated seven-coordinate Ru(IV)
dimer complex with [HOHOH]− bridging ligand as an intermediate for
catalytic water oxidation. J. Am. Chem. Soc. 131, 10397–10399 (2009).
39. Zeng, R. & Dong, G. Rh-catalyzed decarbonylative coupling with
alkynes via C–C activation of isatins. J. Am. Chem. Soc. 137,
1408–1411 (2015).
40. Zeng, R., Chen, P.-h & Dong, G. Efcient benzimidazolidinone synthesis via
rhodium-catalyzed double-decarbonylative C–C activation/cycloaddition
between isatins and isocyanates. ACS Catal. 6, 969–973 (2016).
41. Wang, H., Choi, I., Rogge, T., Kaplaneris, N. & Ackermann, L. Versatile and
robust C–C activation by chelation-assisted manganese catalysis. Nat. Catal.
1, 993–1001 (2018).
42. Vasconcelos, S. N. S., Reis, J. S., de Oliveira, I. M., Balfour, M. N. & Stefani,
H. A. Synthesis of symmetrical biaryl compounds by homocoupling reaction.
Tetrahedron 75, 1865–1959 (2019).
15. Haibach, M. C., Kundu, S., Brookhart, M. & Goldman, A. S. Alkane
metathesis by tandem alkane-dehydrogenation–olefn-metathesis catalysis and
related chemistry. Acc. Chem. Res. 45, 947–958 (2012).
16. Dermenci, A., Coe, J. W. & Dong, G. Direct activation of relatively unstrained
carbon–carbon bonds in homogeneous systems. Org. Chem. Front. 1,
567–581 (2014).
17. Xia, Y. & Dong, G. Temporary or removable directing groups enable
activation of unstrained C–C bonds. Nat. Rev. Chem. 4, 600–614 (2020).
18. Jean-Louis Hérisson, P. & Chauvin, Y. Catalyse de transformation des oléfnes
par les complexes du tungstène. II. Télomérisation des oléfnes cycliques en
présence d’oléfnes acycliques. Makromol. Chem. 141, 161–176 (1971).
19. Sanford, M. S., Love, J. A. & Grubbs, R. H. Mechanism and activity of
ruthenium olefn metathesis catalysts. J. Am. Chem. Soc. 123, 6543–6554 (2001).
20. Vougioukalakis, G. C. & Grubbs, R. H. Ruthenium-based heterocyclic carbene-
coordinated olefn metathesis catalysts. Chem. Rev. 110, 1746–1787 (2010).
21. Liu, P., Taylor, B. L. H., Garcia-Lopez, J. & Houk, K. N. in Handbook of
Metathesis (eds Grubbs, R. H. & Wenzel, A. G.) 199–252 (Wiley-VCH, 2015).
22. Zhu, J., Chen, P.-h, Lu, G., Liu, P. & Dong, G. Ruthenium-catalyzed reductive
cleavage of unstrained aryl–aryl bonds: reaction development and
mechanistic study. J. Am. Chem. Soc. 141, 18630–18640 (2019).
23. Gozin, M., Weisman, A., Ben-David, Y. & Milstein, D. Activation of a
carbon–carbon bond in solution by transition-metal insertion. Nature 364,
699–701 (1993).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
24. Gozin, M. et al. Transfer of methylene groups promoted by metal
complexation. Nature 370, 42–44 (1994).
© The Author(s), under exclusive licence to Springer Nature Limited 2021
NatuRe ChemistRy | www.nature.com/naturechemistry

NATure CHeMIsTry
Articles
methods
University of Chicago for assistance with the calculations carried out in this work. P. Liu
from the University of Pittsburgh is acknowledged for DFT advice.
General method for the C(aryl)–C(aryl) metathesis reaction: To a flame-dried
4.0ml vial charged with a stir bar, biaryl 1 (0.05mmol) and biaryl 2 (0.105mmol)
were added. The vial was then transferred into a glovebox. The ruthenium
catalyst (for example cat 4, 20mol%, based on 1), AgOPiv (2.1mg, 20mol%,
based on 1) and toluene (0.7ml) were added in the glovebox. The vial was sealed
and stirred on a pre-heated heating block at 130°C for 24h. To avoid substrate
or product precipitation on top of the vial, the vial was covered by aluminum
foil to stabilize the temperature. On completion of the reaction, the solvent was
removed under reduced pressure and the residue was analysed using 1H NMR with
1,1,2,2-tetrachloroethane as the internal standard. The reaction mixture was then
directly purified by chromatography on silica gel to afford the desired product.
author contributions
J.Z. and G.D. conceived and designed the experiments. J.Z. performed experiments. R.Z.
performed additional experiments and the DFT calculations. J.Z., R.Z. and G.D. co-wrote
the manuscript.
Competing interests
The authors declare no competing interests.
additional information
Data availability
Supplementary information The online version contains supplementary material
available at https://doi.org/10.1038/s41557-021-00757-4.
The data supporting the findings of this study are available within the paper and its
Supplementary Information.
Correspondence and requests for materials should be addressed to G.D.
Peer review information Nature Chemistry thanks the anonymous reviewers for their
acknowledgements
contribution to the peer review of this work.
We thank the University of Chicago and the NIH (2R01GM109054, G.D.) for research
support. We are grateful for the support of the Research Computing Center at the
Reprints and permissions information is available at www.nature.com/reprints.
NatuRe ChemistRy | www.nature.com/naturechemistry