Photosensitized oxidative addition to gold(i) enables alkynylative cyclization of o-alkylnylphenols with iodoalkynes

Article abstract of DOI:10.1038/s41557-019-0295-9

The well-established oxidative addition–reductive elimination pathway is the most followed one in transition metal-catalysed cross-coupling reactions. While readily occurring with a series of transition metals, gold(i) complexes have shown some reluctance to undergo oxidative addition unless special sets of ligands on gold(i), reagents or reaction conditions are used. Here we show that under visible-light irradiation, an iridium photocatalyst triggers—via triplet sensitization—the oxidative addition of an alkynyl iodide onto a vinylgold(i) intermediate to deliver C(sp)²–C(sp) coupling products after reductive elimination. Mechanistic and modelling studies support that an energy-transfer event takes place, rather than a redox pathway. This particular mode of activation in gold homogenous catalysis was applied in several dual catalytic processes. Alkynylbenzofuran derivatives were obtained from o-alkynylphenols and iodoalkynes in the presence of catalytic gold(i) and iridium(iii) complexes under blue light-emitting diode irradiation.

Full text of DOI:10.1038/s41557-019-0295-9

Articles
https://doi.org/10.1038/s41557-019-0295-9
Photosensitized oxidative addition to gold(i)
enables alkynylative cyclization of o-alkylnylphenols
with iodoalkynes
1
1
1
1
2
2
Zhonghua Xia , Vincent Corcéꢀ ꢀ , Fen Zhao , Cédric Przybylski , Agathe Espagne , Ludovic Jullien ,
2
1,3
1
1
1
Thomas Le Saux , Yves Gimbert , Héloïse Dossmannꢀ ꢀ , Virginie Mouriès-Mansuy *, Cyril Ollivierꢀ ꢀ *
1
and Louis Fensterbankꢀ ꢀ *
The well-established oxidative addition–reductive elimination pathway is the most followed one in transition metal-catalysed
cross-coupling reactions. While readily occurring with a series of transition metals, gold(i) complexes have shown some reluc-
tance to undergo oxidative addition unless special sets of ligands on gold(i), reagents or reaction conditions are used. Here
we show that under visible-light irradiation, an iridium photocatalyst triggers—via triplet sensitization—the oxidative addi-
2
tion of an alkynyl iodide onto a vinylgold(i) intermediate to deliver C(sp) –C(sp) coupling products after reductive elimina-
tion. Mechanistic and modelling studies support that an energy-transfer event takes place, rather than a redox pathway. This
particular mode of activation in gold homogenous catalysis was applied in several dual catalytic processes. Alkynylbenzofuran
derivatives were obtained from o-alkynylphenols and iodoalkynes in the presence of catalytic gold(i) and iridium(iii) complexes
under blue light-emitting diode irradiation.
ver the past two decades, homogeneous gold catalysis relevance in visible-light catalysis involving organometallic
2
7
has been extensively used to efficiently and selectively complexes . Useful photophysical guidelines for Dexter versus
1
–3
O
promote a variety of cyclization processes . The typical Förster and exergonic versus endergonic energy transfers have
28
casting involves bifunctional substrates bearing an unsaturation been drawn .
prompt to electrophilic activation and a judiciously positioned
Extending the scope of possible partners in these transforma-
internal nucleophile. Protodemetalation of the organogold inter- tions is also highly desirable, and we aimed to develop alkynyla-
2
mediates to afford hydrofunctionalized products generally termi- tive cyclization processes that correspond to a formal C(sp) –C(sp)
4
nates the catalytic cycles . To pursue the step economy principle cross-coupling reaction, which has little precedent in this type of
1
3,29–31
and aim for higher levels of molecular complexity, some in situ dual catalytic transformation
. Thus, replacing aryl diazoni-
5
postfunctionalization reactions of the organogold intermedi- ums with alkynyl iodide partners in our recent dual photoredox/
ate have been devised, such as electrophilic halogenation or gold-catalysed arylative cyclization of o-alkynylphenols, which
3
2
cross-coupling reactions. Although palladium-catalysed cross- leads to benzofurans , would constitute an appropriate base for
coupling from an organogold(ꢀ) intermediate has been rendered exploration as well as provide valuable scaffolds. We were also
6
,7
possible , this transformation still needs generality and most of aware that alkynyl iodides are much less reactive than aryl diazoni-
the described coupling reactions have transited through a Au(ꢀ) ums and that we would probably have to devise a distinct mode of
to Au(ꢀꢀꢀ) oxidation, necessitating an oxidant in stoichiometric activation of gold(ꢀ) complexes to promote the C–C bond forma-
8
,9
quantity, transmetalation and a reductive elimination cycle . A tion step (Fig. 1c). Indeed, gold(ꢀ) complexes are notoriously resis-
1
0,11
33
notable breakthrough in this area was achieved by Glorius
tant to oxidative addition . This can be rendered feasible only by
1
2,13
34–36
and Toste , who bypassed the burden of stoichiometric oxi- using special sets of electrophilic reagents
and/or certain con-
3
7
dants by merging gold catalysis with photoredox catalysis, ensur- ditions. For instance, Toste and collaborators showed that CF I
3
1
4–16
ing that the oxidation states shuttle . Arylative cyclization and adds to arylgold(ꢀ) complexes under UV irradiation. Substrates
3
8,39
related transformations, as well as cross-coupling processes that bearing a directing group or with inherent ring strain
can also
rely on the use of easily reduced aryl diazonium salts, have been undergo oxidative addition to provide cyclometallated gold(ꢀꢀꢀ)
1
7–24
40,41
42
devised (Fig. 1a) . Two mechanism pathways have been pro- intermediates. Recently, Amgoune and Bourissou
and Russell
posed (Fig. 1b) that both feature the reductive elimination from have demonstrated that bidentate ligands on gold(ꢀ) with par-
a vinylgold(ꢀꢀꢀ) intermediate of type D but differ by the stage of ticular features promote oxidative addition and that the result-
addition of the radical on gold (intermediate A versus F). Pathway ing gold(ꢀꢀꢀ) intermediate can react with a nucleophile to provide
4
2
41
I has been recently supported by stoichiometric reactions cross-coupling products , notably through a catalytic cycle .
9
,25,26
and calculations
. A ground-breaking advance in these In this work, we uncover a mode of C–C bond formation via pho-
reactions would be to promote the oxidative addition step by tosensitized energy transfer that promotes oxidative addition of a
energy transfer (photosensitization) as this has found more gold(ꢀ) complex (Fig. 1c).
1
2
Sorbonne Université, CNRS, Institut Parisien de Chimie Moléculaire, Paris, France. PASTEUR, Département de Chimie, École Normale
3
Supérieure, PSL University, Sorbonne Université, CNRS, Paris, France. Université Grenoble Alpes and CNRS, DCM (UMR 5250), Grenoble, France.
*
e-mail: virginie.mansuy@upmc.fr; cyril.ollivier@upmc.fr; louis.fensterbank@upmc.fr
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

Articles
NATure CHeMisTry
a
L
Ar
X
+ [PC]n+
R1
Nu
Ar
L
Au
X
R1
Au
R2
R1
R2
+
Ar N2
+
Nu
H
Via pathway I or II
Visible light
R2
Nu
b
NuH
NuH
c
R
Nu
AuIII
Nu
R
+ X
L
AuII
L
R
Visible
light
R
Cl
Cl
F
[PC](n+1)+
–
H
AuI
L
AuIII
C
E
[
PC]n+
L
SET
R
B
NuH
Gold
catalysis
X
–
H
Photoredox
catalysis
Nu
SET
PC]n+*
R
[
Gold
catalysis
Photoredox
catalysis
SET
L
[PC]n+*
X
AuIII
D
L
R
R
D
Nu
Cl
Cl
L
AuI Cl
SET
PC]n+
[
A
PC](n+1)+
AuII
R
[
Visible
light
L
AuIII
Cl
R
R
+ X
Nu
AuI Cl
L
Nu
R
R
R
Pathway I
Pathway II
O
d
O
O
R1
R1
OH
L
Au
X
R1
I
R2
Ir
+
base
Au
Au
L
R2
R1
L
R2
I
Photosensitized
gold catalysis
Fig. 1 | Gold-catalysed additions to alkynes. a, General scheme of the dual gold-photoredox catalysed difunctionalization of alkynes. b, Proposed
mechanistic pathway I in dual gold-photoredox catalysis. The key Au(iii) intermediate b is generated via radical addition on the Au(i) precatalyst and the
resulting Au(ii) complex A is oxidized through single electron transfer (SET) by the oxidized photocatalyst. Pathway I has been supported by calculations
and the isolation of intermediates. c, Alternate mechanistic pathway II implies the initial formation of a vinyl Au(i) intermediate E. d, Dual gold and
photoinduced alkynylative O-cyclization via photosensitized oxidative addition.
ꢂesults and discussion
role of 1,10-phenanthroline were also performed. A stoichiomet-
Optimization studies. We surveyed the feasibility of such a trans- ric amount of 1,10-phenanthroline proved detrimental to the yield
formation by examining the model reaction between 2-(p-tolyl- (entry 6); however, other amines, such as quinuclidine (entry 7)
ethynyl)phenol 1a and iodoethynyl benzene 2a under various or tetramethylethylenediamine (TMEDA), 1,4-diazabicyclo[2.2.2]
conditions (see Supplementary Section III for detailed conditions octane (DABCO) and 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU)
optimization). Preliminary negative results based on the previous (see Supplementary Section III. 4), could also be used to improve the
arylation protocol using a catalytic mixture of Ru(bpy) Cl and reaction. In sharp contrast, no desired product could be obtained in
3
2
PPh AuCl in MeOH drove us to use other conditions. The first hits the absence of the base (K CO , entry 11), [Au-CF ] (entry 10) or
3
2
3
3
in the formation of benzofuran 3aa (structure confirmed by X-ray light (entry 12). Finally, it is worth noting that the iodoetherifica-
diffraction analysis, Cambridge Crystallographic Data Centre tion product 4a and diyne 5a were the side compounds in almost
(
CCDC) accession code no. 1850903) were obtained in acetoni- all conditions. Diyne 5a was present in a lower quantity (<20%) and
trile and by adding a base (see Supplementary Section III. 2 and its formation presumably requires gold catalysis (see Supplementary
Table 1, entry 1). This finding was consolidated by using Ir[dF(CF ) Section III. 6). When (bromoethynyl)benzene 2a-Br was subjected
3
ppy] (dtbbpy)PF ([Ir-F]) as a photocatalyst (entry 3, 26% of 3aa). to the reaction, only 9% of 3aa was obtained.
2
6
A substantial gain of yield was observed (56% of 3aa) by switch-
ing PPh AuCl to (p-CF Ph) PAuCl [Au-CF ] (entry 4). Finally, Mechanistic investigations. This preliminary set of findings drove
3
3
3
3
after substantial optimization it was found that the combination us to delineate a plausible mechanism for the further development
of [Au-CF ] (5mol%), ([Ir-F]) (1mol%), 1,10-phenanthroline of the reaction. We first considered the addition of a radical inter-
3
(
10mol%), K CO (2.5 equiv.) in degassed MeCN at room tem- mediate stemming from the photocatalytic cycle to produce the
2
3
perature overnight under a blue light-emitting diode (LED) light corresponding intermediate of type B through A (Fig. 1). Alkynyl
gave the best result, since a 71% isolated yield of 3aa was obtained radicals remain an elusive species, but they have been noted spo-
4
3,44
(
entry 5). The reaction can work without a photocatalyst (entry 8) radically in the literature to be generated from alkynyl iodides
.
and the use of a more reductive photocatalyst such as fac-Ir(ppy)
Nevertheless, by using alkynyl iodide 2b as a probe (as it bears a
3
was not beneficial (entry 2). Control experiments regarding the fluorine label), this hypothesis was rapidly discarded. First, as the
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

NATure CHeMisTry
Articles
Table 1 | Defining the key parameters of the alkynylative cyclization
OH
O
[Au] (5 mol %), [PC] (1 mol%),
O
I
2.5 equiv. K2CO3, additive
+
+
+
Ph
Ph
Ph
MeCN, blue LED,
r.t., overnight
I
1
a
2a
1.5 equiv.
3aa
Ph
4a
5a
1
equiv.
Entry
[Au]a
[PC]a
Additive, 10ꢄmol%a
3aa yield (%)c
4a yield (%)
1b
2b
3b
4
5
PPh AuCl
Ru(bpy) (PF )
-
15
21
22
33
22
15
25
19
29
29
8
3
3
6
2
PPh AuCl
fac-Ir(ppy)3
-
8
3
PPh AuCl
[ꢃr-F]
[ꢃr-F]
[ꢃr-F]
[ꢃr-F]
[ꢃr-F]
-
-
26
3
[Au-CF3]
[Au-CF3]
[Au-CF3]
[Au-CF3]
[Au-CF3]
[Au-CF3]
-
-
56
72 (71)f
phen.
6
phen. (1 equiv.)
quinuclidine
phen.
25
65
26
11
7
8
9
-
-
1
1
1
0
1d
2e
[ꢃr-F]
[ꢃr-F]
[ꢃr-F]
phen.
-
[Au-CF3]
[Au-CF3]
PAuCl; [ꢃr-F] = Ir[dF(CF
phen.
-
27
16
phen.
-
a[Au-CF
] = (p-CF
Ph)
)ppy]
(dtbbpy)PF
; phen. = 1,10-phenanthroline. Only 1ꢀequiv. of K
b
CO
was used. Yields were determined by H NMR using 1,3,5-trimethoxybenzene as
c
1
3
3
3
3
2
6
d
2
3
e
f
2 3
the internal standard, the yield in parentheses is the isolated yield. No K CO . No light. Isolated yield.
reductive potential of 1-fluoro-4-(iodoethynyl)benzene at the when the concentration of 6 was increased, which suggested that
3
ground state (E (2b)=−1.47V versus saturated calomel electrode 6 acts as a quencher of [Ir-F] (Fig. 2e). In parallel, we recorded
1/2
(
SCE) is substantially higher than the reduction potential of the the luminescence lifetime of the same solutions and observed that
excited state of the [Ir-F] catalyst (E* =−0.89V versus SCE), pho- it decreased from 2.4µs in the absence of 6 (in agreement with pre-
1/2
46,47
toreductive formation of an alkynyl radical seemed unlikely. This vious reports ) to 290ns in the presence of 745µM of 6, thereby
3
was corroborated by fluorescence quenching studies that showed no confirming the quenching of [Ir-F] by 6 (see Supplementary
3
3
9
−1 −1
9
−1 −1
quenching of excited [Ir-F] ( T ), denoted [Ir-F], by 2b, therefore Fig. 3). Values of 2.9×10 moll s and 4.1×10 moll s were
1
precluding a photocatalysed electron transfer event. Another path- extracted for the bimolecular quenching rate constant (k ) from
q
way for the formation of the gold(ꢀꢀꢀ) intermediate of type B was the luminescence intensity (I /I) and lifetime (τ /τ) Stern–Volmer
0
0
investigated due to the fact that alkynyl iodide 2b and the gold com- plots respectively (see inset of Fig. 2e and Supplementary Fig. 4).
plex [Au-CF ] convert into a new unstable gold species under blue These k values are almost in the range of the encounter rate under
3
q
LED irradiation, but the catalytic role of the latter remains elusive the control of molecular diffusion. This observation suggests that
(
see Supplementary Section V.5).
no major molecular reorganization occurs during the reaction
9
,14,16
3
Literature reports
indicate that a vinylgold(ꢀꢀꢀ) intermedi- between [Ir-F] and 6, which would be reasonably in line with a
2
8
ate of type D of Fig. 1, which would undergo reductive elimination Dexter-type energy transfer according to equation (1) (ref. ), while
3
to provide benzofurans 3, is presumably involved. To investigate considering that the triplet level of [Ir-F] is considerably higher
3
48
this, vinylgold(ꢀ) 6 of type E was prepared in 82% yield by an inde- than that of 6 (ref. ).
45
pendent route as shown in Fig. 2a. An X-ray diffraction analysis
of suitable crystals of 6 confirmed its structure (CCDC 1850902,
Fig. 2b) and provided useful structural data for further modelling
studies. No conversion of 6 when reacted with one equivalent of
3
3
½Ir � F þ 6 ! ½Ir � F þ 6
ð1Þ
We also recorded the transient absorption spectra of [Ir-F]
alkynyliodide 2b was observed after 3h at room temperature (25°C) solutions containing various concentrations of 6 (Fig. 2f). The
3
in the absence of irradiation with a blue LED. However, blue LED differential spectrum of [Ir-F] exhibited a maximum at around
4
9
irradiation changed the scenario. A low conversion (<10%) was 480–500nm, which is in agreement with the literature . The addi-
3
observed after 2h at room temperature, while overnight irradiation tion of an equimolar amount of 6 yields a decrease in the [Ir-F] sig-
3
resulted in the formation of 33% of benzofuran 3ab (accompanied nal, consistent with the [Ir-F] luminescence quenching observed
3
by 20% of 4a and 35% of protodeauration product 7 as determined in Fig. 2e. In the presence of a ninefold excess of 6, the [Ir-F] signal
1
by H NMR (Fig. 2c)). The addition of 10mol% of [Ir-F] markedly almost disappears and is replaced by a broad differential absorp-
3
altered the outcome and yielded 3ab almost quantitatively (Fig. 2d). tion tentatively attributed to 6 (new contributions below 450nm
Therefore, experimental conditions to trigger the key C–C bond and above 550nm). The formation of 3ab is also enhanced by the
formation were successfully found. Although benzofuran forma- presence of 10mol% benzophenone in the reaction medium, which
tion could be achieved under direct irradiation conditions without presumably also acts as a sensitizer (Fig. 2d). Another important
photocatalyst [Ir-F] (Fig. 2c,d), which brings some rationalization point to check was the formation of 6 in the reaction conditions.
for the finding in entries 8 and 9 of Table 1, benzofuran formation This was achieved by exposing 1a to a stoichiometric amount of
seems to be greatly enhanced in its presence. This was confirmed [Au-CF ] in CD CN overnight. After this reaction time in the
3
3
by measuring the steady-state luminescence spectra of mixtures of dark, the formation of 6 was observed by NMR in 26% yield (See
Ir-F] and 6. We observed a drop in the [Ir-F] luminescence signal Supplementary Section V. 4).
[
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

Articles
NATure CHeMisTry
a
b
OH
O
[Au-CF3]
+
(1) AgOTs, THF
2) Et3N
(
1a
AuP(p-CF3Ph)3
, 82%
6
c
O
Blue LED
CD3CN,
Overnight
O
O
O
+
2b
+
+
AuP(p-CF3Ph)3
I
H
6
3ab, 33%
4a, 20%
7, 35%
F
d
g
O
3
ab
0 min, 26%
.5 h, 85%
1
0 mol% [Ir-F]
3
2
+
2b
Blue LED
CD3CN
AuP(p-CF3Ph)3
3.5 h, 95%
6
Photosensitization
Direct
irradiation
CD3CN,
2.5 h
1
0 mol%
benzophenone
3ab
λ = 365 nm = 35%
3[Ir–F] + 6
3
ab
CD3CN, 2.5 h
λ = 400 nm = 42%
λ = 365 nm, 13%
λ = 400 nm, 43%
h
e
f
2
.5
.0
.5
.0
.5
.0
0
6
4
2
0
50
2
1
1
0
0
[6]
40
30
0
5
00 µM
0
200 400 600
6] (µM)
4.5 µM
[
2
1
0
0
0
745 µM
3[Ir–F]
450
500
550
600
650
700
450
500
550
600
650
700
Wavelength (nm)
Wavelength (nm)
45
Fig. 2 | A vinylgold(i) species as a plausible intermediate. a, Preparation of 6 according to Hashmi’s method . b, X-ray diffraction structure of complex
3
6
6
. c, Alkynylation from 2b and 6 under blue LED irradiation. d, Alkynylation of 6 by 2b by direct irradiation and sensitization. e, Quenching of [ꢃr-F] by
3
monitored by steady-state fluorimetry. f, Differential absorption spectra of [ꢃr-F] (500ꢀµM in N -saturated acetonitrile) in the presence of increasing
amounts of 6, showing the quenching of the absorption of [ꢃr-F] at 500ꢀnm and the appearance of new contributions assigned to 6 below 450ꢀnm and above
50ꢀnm. Spin density (difference between the α and β densities) isosurface (isovalue 0.0006 a.u.). g, [ꢃr-F] in the vicinity of 6. h, Isolated [ꢃr-F] complex.
2
3
3
3
3
5
All these studies converge to imply an excited state of 6, whose an excited electronic state, which may further react with 2a. Note
3
50
3
formation is promoted by the long-lived triplet state [Ir-F] (ref. ). that the same calculations performed on [Ir-F] approached by 2a
This finding is in line with a recent study on excited-state organo- show that no transfer is occurring to 2a (see Supplementary Fig. 7).
51
metallic catalysis by McCusker and MacMillan , who reported This finding is consistent with the quenching studies carried out on
an energy-transfer-mediated reductive elimination of an excited [Ir-F] in the presence of 6 or 2a. Note that [Au-CF ] also does not
3
arylnickel(ꢀꢀ) intermediate, in the same vein as other related reports quench the fluorescence of [Ir-F]. To determine which electronic
52,53
54–56
on copper
and nickel intermediates . Further support for states of 6 are accessible via this energy transfer, time-dependent
these conclusions was provided by calculations. The spin densi- density functional theory calculations were carried out. Results
3
3
ties of [Ir-F] in isolation or in the vicinity of 6 were compared show that only 6 was accessible within the blue LED energy range
(
Fig. 2g,h and Supplementary Sections VIII. 2–3 for calculation (470nm, see Supplementary Section VIII. 3 for details).
3
details). As shown in Fig. 2g, part of the [Ir-F] spin density is trans-
ferred to the approaching furan moiety of 6, intimating that energy Modelling studies. Following these findings, we conducted a
3
transfer is taking place. This would lead to the formation of 6 in detailed theoretical study of the reaction of 6 with 2a. The ground
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

NATure CHeMisTry
Articles
O
O
O
Au
AuP(p-CF3Ph)3
3
P(p-CF Ph)3
I
36
Au
I
P(p-CF3Ph)3
0
TS4
(
3 3
p-CF Ph)
(
3 3
p-CF Ph)
–
6.1
3II
3IV
TS3
16.4
–
I
–18.5
2a
O
O
–
25.1
Au
I
O
P(p-CF3Ph)3
Au
I
P(p-CF3Ph)3
Au P(p-CF3Ph)3
3II
I
TS5
O
3III
–33.3
O
–33.3
(p-CF3Ph)3P
Au
I
Au
I
TS6
P(p-CF3Ph)3
3IV
–53.9
–
60.0
O
–65.9
I
Au
p-CF3Ph)3P
O
(
3T1
1S0
3V
33
aa
3T1
1S0
O
Au
I
P(p-CF3Ph)3
–114.3
1IV
O
Dissociative
complex
3
aa
+
(p-CF3Ph)3PAuI
3
3
3
Fig. 3 | Potential surface energy of the reaction of 6 with 2a. The reaction of 6 with 2aa proceeds via several steps to lead to intermediate ꢃV. Final
3
3
3
formation of 3aa follows either the dissociative decay of ꢃV or occurs after rearrangement and dissociation of ꢃV into 3aa. Optimized geometries of
intermediates ꢃꢃ and ꢃV are given in the inset (top right). Salient features of complex ꢃꢃ are the I–C–C angle of 115° and a C–Au bond of 2.12ꢀÅ. Gibbs free
energies (CH CN) are given relative to the starting products and are in kcalꢀmol .
3
3
3
−
1
3
state reactivity of 6 was also checked to help rationalize the role taking into account solvent effects by SMD procedure. If not
of blue LED in the mechanism efficiency. For the sake of clar- stated otherwise, reported energies are Gibbs free energies in
1
3
ity, we will now refer to 6 either as 6 or as 6 to unambiguously CH CN DGCH CN (ΔG
).
3
3
MeCN
1
3
point to the S ground state or the T excited state of 6, respec-
On the singlet potential energy surface, the reaction pathway
0
1
tively. All calculations presented below were obtained at the PBE0/ was quite straightforward to determine. However the barrier to
1
SDD(Au), 6–311G*(I), 6–31G** (other atoms) level of theory, formation of the oxidative addition adduct, gold(ꢀꢀꢀ) complex I is
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

Articles
NATure CHeMisTry
OH
lower energies than the starting products. This would be consis-
tent with the great efficiency of the observed reaction of Fig. 2d.
The aforementioned data all support the plausible catalytic path-
way of Fig. 4.
1
Ph
Reductive
elimination
O
+
–
base
HX
In addition to the use of the photosensitizer [Ir-F], two other
factors seem to optimize this process. First, the substitution of
Ph PAuCl by [Au-CF ] seemed to be highly beneficial (Table 1, entry
I
Ph
L
O
L–Au –X
Ph
X = Cl or I
AuI
III*
Ir–F]
3
3
[
2
5
3
3 versus 4), perhaps due to the higher electrophilicity of [Au-CF ] .
3
R
Second, the adjunction of 1,10-phenanthroline gave a significant
yield increase regardless of the pathway followed (compare entry
4 versus 5 and entry 8 versus 9 in Table 1). The reason for this is
not clearly established and several hypotheses have been proposed.
For instance, some halogen bonding between phenanthroline and
Gold
catalysis
Energy
transfer
Photo
sensitization
O
*
Ph
AuI
[Ir–F]III
L
I
57
R
O
*
the alkynyl iodides 2 (which are known halogen-bonding donors)
Ph
might be at work and explain the increased reactivity of the system.
Oxidative
addition
AuI
I
R
Scope of the reaction. We then explored the scope of this trans-
formation. For this, a series of substituted aryl iodoalkynes 2 were
reacted with 2-(phenylethynyl)phenols 1 under optimized reaction
conditions (Fig. 5a). Aryl iodoethynyl bearing electron-withdraw-
L
2
Fig. 4 | Mechanism proposal. Two intertwined catalytic cycles are
responsible for the alkynylative cyclization of o-alkylnylphenols with
iodoalkynes. The connecting points involve the interaction of [ꢃr-F]
ing groups in the para position, such as F and CF , produced the cor-
3
3
responding benzofurans 3ab and 3ac in good yields. Meta F or ortho
Cl-substituted aryl iodoalkynes reacted smoothly with 2-(phenyl-
ethynyl)phenol 1a, providing 3ad and 3ae in 78% and 63% yields,
respectively. The presence of the electron-donating groups (–Me,
3
with 6 to promote the formation of 6, which is prone to undergoing
oxidative addition.
–
t-Bu, –OMe) at the para position, or no substitution, furnished the
−
1
prohibitive: almost 28kcalmol (see Supplementary Section VIII. alkynylbenzofurans in slightly lower (3af, 3ag and 3ba) to moderate
4
) and renders the overall pathway unlikely. On the other hand, the (3ah) yields. No reaction was observed with an iodoalkyne bearing
reaction on the triplet potential energy surface seemed to be more a 4-nitro arylgroup, instead the protodeauration cyclization product
sinuous, but also more favourable thermodynamically (see Fig. 3). 7 was obtained. In addition, alkylalkynes bearing n-pentyl, cyclo-
3
The approach of 2a to 6 (along the Au–C (bearing the iodine) bond hexyl, and 3-nitrile butyl groups could be incorporated into the
reaction coordinate) leads to the formation of an intermediate com- benzofuran scaffold (3aj–3al) from the corresponding alkyl iodo-
3
−1
plex II, located 18.5kcalmol below the reagents. Interestingly, the alkynes, albeit in significantly lower yields (25–42%).
geometry adopted by 2a in this complex is bent (I–C–C angle of
The effect of substitution on both aromatic rings of o-alkynyl-
3
1
15°) and reminiscent of that of modelled 2a (I–C–C angle of 129° phenols 1 was then investigated in reactions with 1-(iodoethynyl)-
versus 180° for 2a in its ground state, see Supplementary Fig. 15). 4-(trifluoromethyl)benzene 2c (Fig. 5b). Arylalkynes with no
3
This feature suggests that the 6 complex may transfer energy to 2a substitution or bearing a CH F group or F atom at the para or meta
3
when these reactants approach each other. Checking the spin den- position to the alkyne gave good yields of benzofurans (3bc, 3ac,
sity along the Au–C (bearing the iodine) bond reaction coordinate 3ec and 3fc). Similarly, precursors with an electron-withdrawing
effectively confirms that a transfer is occurring at a relatively long ester group on the phenol moiety at the para or meta position of the
3
distance (from 3.6Å, see Supplementary Fig. 16). Therefore, 6 could alkyne delivered corresponding ethynylbenzofurans in good yields
act as a relay for transferring energy to 2a, which would provide (3cc and 3dc).
access to a reactive bent structure of the acetylenic compound. From
3
complex II and by approaching the iodine atom to gold, intermedi- Postfunctionalization and other systems. The benzofurans 3
3
ate III is localized on which Au(ꢀ) is oxidized to Au(ꢀꢀꢀ) and organic are valuable scaffolds for further elaboration, notably through the
precursors lie in the trans position. This step occurs with a low acti- potential reactivity of the alkyne moiety. For instance, the prod-
−
1
vation energy of 2.2kcalmol via TS3. By reducing the C–Au–C uct 3aa can be hydrogenated by formic acid under palladium(0)
3
angle on III (that is bringing together the two C atoms involved in catalysis. Depending on the reaction conditions Z-alkene 8 or the
the forthcoming new C–C bond), it was possible to localize a transi- E-isomer 9 can be selectively obtained. Triazole 10 could also be
−1
tion structure TS4, requiring a formation barrier of 18.8kcalmol . formed in thermal conditions through a Huisgen type of reaction
3
TS4 connects to the Au(ꢀꢀꢀ) complex intermediate IV, on which between 3aa and sodium azide (Fig. 5c, see Supplementary Section
the formation of the key C–C bond between 6 and 2a is observed VI for details).
but with the iodine still interacting with the slightly elongated triple
Finally, the sensitization protocol is not restricted to substrates
bond (d(I–C)=2.29Å). Finally, two pathway variants can be envis- 1 and 2. Other organogold intermediates such as vinylgold 11 and
3
3
aged from IV. First, the IV complex may further rearrange to lead arylgold 13 can undergo the oxidative addition (reductive elimina-
3
−1
3
to a V complex (−27kcalmol below IV) via an inexpensive TS5 tion sequence to provide respectively 12 and 14 as shown in Fig. 5d).
−
1
3
transition structure (+0.001kcalmol compared with IV). Then One pot reactions are also possible using a vinyliodide electrophile
3
3
V easily dissociates to lead to 3aa+ (p-CF Ph) PAuI (barrier TS6 (15) or an o-alkynyl tosylaniline (17) as nucleophilic precursor. In
3
3
−
1
3
of 6kcalmol ). An electronic decay of 3aa can then be envisaged the latter case, a new route to 2,3-disubstituted indoles is available.
to lead to the final coupling product 3aa+ (p-CF Ph) PAuI. Another
3
3
3
possible pathway would imply a direct electronic decay of IV from Conclusions
S toT leading to a complex that was revealed to be dissociative. It This study describes a dual catalysis transformation, involving elec-
0
1
would therefore directly provide the final coupling products.
trophilic gold catalysis and iridium photosensitization to allow a
3
2
The proposed reaction pathway using 6 as a starting product C(sp) –C(sp) cross-coupling reaction, useful for the alkynylation of
seems to be extremely favourable from a thermodynamic point of benzofurans. A thorough luminescence study, supported by density
view, as all intermediate complexes and transition structures have functional theory calculations, revealed the mechanistic pathway.
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

NATure CHeMisTry
Articles
CF3
–
F
F
PF6
Cl
OH
O
[
Au-CF3], [Ir-F]
R1
t-Bu
N
Au
a
I
K2CO3, 1,10-phenanthroline
N
N
F
F
+
+
R
F3C
P
CF3
Ir
MeCN, blue LED,
Room temperature, overnight
N
R1
t-Bu
1
2
3
R
CF3
1
equiv.
1.5 equiv.
CF3
[Ir-F]
[Au-CF3]
a
O
O
O
O
O
O
Cl
F
3
aa, 71%
3ba, 67%
3ab, 70%
F
3ac, 75%
CF3
O
3ad, 78%
3ae, 63%
O
O
O
O
O
N
3af, 63%
3ag, 44%
t-Bu
3ah, 35%
OMe
3aj, 42%
3ak, 41%
3al, 25%
O
b
F
O
O
O
O
O
O
F
O
O
3bc, 72%
CF3
3cc, 65%
CF3
3dc, 83%
CF3
3ec, 51%
CF3
3fc, 62%
CF3
c
O
2
equiv. HCO2H,
mol% Pd2(dba)3,
mol% dppb
3 equiv. aq. HCO2H (25%),
1 mol% Pd2(dba)3,
2 mol% dppb
O
O
1
4 equiv. NaN3,
DMF, 180 °C
2
3aa
3aa
H
3aa
H
Dioxane, 80 °C
H
Dioxane, 80 °C
N
N
8, 85%
9, 86%
N
10, 62%
H
H
d
O
n-C6H13
O
O
N
O
O
O
n-C6H13
Conditionsb
Conditionsa
Conditionsb
+
2b
2b
1a
+
6
+
15
N
35%
Ph3PAu
O
I
16
12, 60% (78%)c
15
11
F
Ts
Ts
N
N
H
Conditionsb
Conditionsa
+
2b
AuPPh3
+
F
13
14, 82%
17
18, 50%
F
Fig. 5 | Scope of the alkynylation process. a, Variation of the iodoalkynes 2. b, Variation of the o-alkynylphenols 1. c, Post-functionalization of benzofuran
3
aa by stereoselective hydrogenation and [3ꢀ+ꢀ2] cycloaddition. d, Other working systems involving other nucleophilic and electrophilic partners.
a
Experimental details: 5ꢀmol% [Au-CF ], 1ꢀmol% [ꢃr-F], K CO (2.5ꢀequiv.), 1,10-phenanthroline (10ꢀmol%) MeCN, blue LED, room temperature, overnight;
3
2
3
b
c
1
ꢀmol% [ꢃr-F], 1,10-phenanthroline (10ꢀmol%) MeCN, blue LED, room temperature, overnight. NMR yield.
Blue LED-excited [Ir-F] interacts with a vinylgold(ꢀ) intermedi- other words, the triplet excited state of the vinylgold(ꢀ) intermedi-
ate, stemming from a gold(ꢀ)-promoted 5-endo-dig O-cyclization ate and the alkynyl iodide partner readily engage in an oxidative
via energy transfer, to trigger oxidative addition at gold(ꢀ). In addition–trans/cis isomerization sequence that forges the desired
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

Articles
NATure CHeMisTry
2
18. Um, J., Yun, H. & Shin, S. Cross-coupling of Meyer–Schuster intermediates
C(sp) –C(sp) bond, followed by reductive elimination that deliv-
under dual gold–photoredox catalysis. Org. Lett. 18, 484–487 (2016).
ers the benzofuran product. The process described in this Article
enables the oxidative addition–reductive elimination pathway to be
used with organogold intermediates and, because other nucleophilic
1
9. Huang, L., Rudolph, M., Rominger, F. & Hashmi, A. S. K. Photosensitizer-free
visible light mediated gold catalyzed 1,2-difunctionalization of alkynes.
Angew. Chem. Int. Ed. 55, 4808–4813 (2016).
and electrophilic partners are competent, it opens new avenues in 20. Tlahuext-Aca, A., Hopkinson, M. N., Garza-Sanchez, R. A. & Glorius, F.
Alkyne difunctionalization by dual gold/photoredox catalysis. Chem. Eur. J
the field of excited-state gold catalysis.
2
2, 5909–5913 (2016).
2
1. Alcaide, B., Almendros, P., Busto, E. & Luna, A. Domino Meyer–Schuster/
arylation reaction of alkynols or alkynyl hydroperoxides with diazonium
salts promoted by visible light under dual gold and ruthenium catalysis.
Adv. Synth. Catal. 358, 1526–1533 (2016).
2. Cornilleau, T., Hermange, P. & Fouquet, E. Gold-catalysed cross-coupling
between aryldiazonium salts and arylboronic acids: probing the
usefulness of photoredox conditions. Chem. Commun. 52,
Methods
General procedure for alkynylative cyclization of o-alkynylphenols with
iodoalkynes. A Schlenk tube was equipped with a magnetic stirring bar and
charged with the photocatalyst [Ir-F](1mol%) before the addition of the gold(ꢀ)
2
3 2 3
complex [Au-CF ] (5mol%), K CO (2.5equiv), 1,10-phenanthroline (10mol%),
the appropriate iodoalkyne 2 (0.15mmol) and o-alkynylphenol derivative 1
(
0.1mmol), with MeCN (2ml). ꢁe mixture was degassed using three freeze
1
0040–10043 (2016).
pump–thaw cycles and purged with Ar, then irradiated for 16h (unless otherwise
stated) with blue LED light (see Supplementary Section VI for set-up). ꢁe stirring
2
2
2
3. Gauchot, V. & Lee, A.-L. Dual gold photoredox C(sp2)-C(sp2) cross
couplings – development and mechanistic studies. Chem. Commun. 52,
2
speed was ≥1,200r.p.m. ꢁe reaction was quenched with Et O (3ml) and a 2M
1
0163–10166 (2016).
HCl solution (3ml) and the solution was then extracted by Et
2
O (3×5ml). ꢁe
4. Gauchot, V., Sutherland, D. R. & Lee, A.-L. Dual gold and photoredox
catalysed C–H activation of arenes for aryl–aryl cross couplings. Chem. Sci 8,
combined organic layer was dried over MgSO , ꢂltered and concentrated under
4
reduced pressure to give the crude product. ꢁe residue was puriꢂed by ꢃash
2
885–2889 (2017).
chromatography (FC) on silica gel to aꢄord 3.
5. Tlahuext-Aca, A., Hopkinson, M. N., Daniliuc, C. G. & Glorius, F. Oxidative
addition to gold(I) by photoredox catalysis: straightforward access to
diverse (C,N)- cyclometalated gold(III) complexes. Chem. Eur. J 22,
11587–11592 (2016).
Data availability
Crystallographic data for the structures reported in this Article have been
deposited at the Cambridge Crystallographic Data Centre under deposition
numbers 1850903 (3aa) and 1850902 (6). Copies of the data can be obtained free
of charge via https://www.ccdc.cam.ac.uk/structures/. All other data supporting
the findings of this study are available within the Article and the Supplementary
Information, or from the corresponding authors on reasonable request.
26. Zhang, Q., Zhang, Z.-Q., Fu, Y. & Yu, H.-Z. Mechanism of the visible
light-mediated gold-catalyzed oxyarylation reaction of alkenes. ACS Catal 6,
798–808 (2016).
27. Zhou, Q.-Q., Zou, Y.-Q., Lu, L.-Q. & Xiao, W.-J. Visible-light-induced organic
photochemical reactions through energy-transfer pathways. Angew. Chem. Int.
Ed. 58, 1586–1604 (2019).
2
8. Strieth-Kalthoꢄ, F., James, J. M., Teders, M., Pitzer, L. & Glorius, F. Energy
transfer catalysis mediated by visible light: principles, applications, directions.
Chem. Soc. Rev. 47, 7190–7202 (2018).
Received: 5 September 2018; Accepted: 18 June 2019;
Published: xx xx xxxx
2
9. de Haro, T. & Nevado, C. Gold-Catatalyzed Ethynylation of Arenes. J. Am.
Chem. Soc. 132, 1512–1513 (2010).
ꢂeferences
1.
2.
3.
Dorel, R. & Echavarren, A. M. Gold(I)-catalyzed activation of alkynes for the
construction of molecular complexity. Chem. Rev. 115, 9028–9072 (2015).
Fensterbank, L. & Malacria, M. Molecular complexity from polyunsaturated
substrates: the gold catalysis approach. Acc. Chem. Res. 47, 953–965 (2014).
Harris, R. J. & Widenhoefer, R. A. Gold carbenes, gold-stabilized
carbocations, and cationic intermediates relevant to gold-catalysed enyne
cycloaddition. Chem. Soc. Rev. 45, 4533–4551 (2016).
30. Hopkinson, M. N., Ross, J. E., Giuꢄredi, G. T., Gee, A. D. & Gouverneur, V.
Gold-catalyzed cascade cyclization−oxidative alkynylation of allenoates.
Org. Lett. 12, 4904–4907 (2010).
31. Li, Y., Brand, J. P. & Waser, J. Gold-catalyzed regioselective synthesis of
2- and 3-alkynyl furans. Angew. Chem. Int. Ed. 52, 6743–6747 (2013).
32. Xia, Z., Khaled, O., Mouriès-Mansuy, V., Ollivier, C. & Fensterbank, L. Dual
photoredox/gold catalysis arylative cyclization of o-alkynylphenols with
aryldiazonium salts: a ꢃexible synthesis of benzofurans. J. Org. Chem. 81,
7182–7190 (2016).
4
5
6
.
.
.
Wang, W., Hammond, G. B. & Xu, B. Ligand eꢄects and ligand design in
homogeneous gold(I) catalysis. J. Am. Chem. Soc. 134, 5697–5705 (2012).
Liu, L.-P. & Hammond, G. Recent advances in the isolation and reactivity of
organogold complexes. Chem. Soc. Rev. 41, 3129–3139 (2012).
33. Joost, M., Amgoune, A. & Bourissou, D. Reactivity of gold complexes
towards elementary organometallic reactions. Angew. Chem. Int. Ed. 54,
15022–15045 (2015).
Hashmi, A. S. K. et al. Scope and limitations of palladium-catalyzed
cross-coupling reactions with organogold compounds. Adv. Synth. Catal. 352,
34. Levin, M. D. & Toste, F. D. Gold-catalyzed allylation of aryl boronic acids:
accessing cross-coupling reactivity with gold. Angew. Chem. Int. Ed. 53,
6211–6215 (2014).
1307–1314 (2010).
7.
Garcia-Dominguez, P. & Nevado, C. Au−Pd bimetallic catalysis: the
importance of anionic ligands in catalyst speciation. J. Am. Chem. Soc. 138,
35. Asomoza-Solis, E. O., Rojas-Ocampo, J., Toscano, R. A. & Porcel, S.
Arenediazonium salts as electrophiles for the oxidative addition of gold(I).
Chem. Commun. 52, 7295–7298 (2016).
3266–3269 (2016).
8
9
1
.
Zheng, Z., Wang, Z., Wang, Y. & Zhang, L. Au-catalysed oxidative cyclisation.
Chem. Soc. Rev. 45, 4448–4458 (2016).
36. Huang, L., Rominger, F., Rudolph, M. & Hashmi, A. S. K. A general access
to organogold(III) complexes by oxidative addition of diazonium salts.
Chem. Commun. 52, 6435–6438 (2016).
.
Hopkinson, M. N., Tlahuext-Aca, A. & Glorius, F. Merging visible light
photoredox and gold catalysis. Acc. Chem. Res. 49, 2261–2272 (2016).
0. Sahoo, B., Hopkinson, M. N. & Glorius, F. Combining gold and photoredox
catalysis: visible light-mediated oxy- and aminoarylation of alkenes. J. Am.
Chem. Soc. 135, 5505–5508 (2013).
37. Winston, M. S., Wolf, W. J. & Toste, F. D. Photoinitiated oxidative addition of
CF I to gold(I) and facile aryl-CF reductive elimination. J. Am. Chem. Soc.
3
3
136, 7777–7782 (2014).
11. Tlahuext-Aca, A., Hopkinson, M. N., Sahoo, B. & Glorius, F. Dual gold/
photoredox-catalyzed C(sp)–H arylation of terminal alkynes with diazonium
salts. Chem. Sci 7, 89–93 (2016).
12. Shu, X.-Z., Zhang, M., Frei, H. & Toste, F. D. Dual visible light photoredox
and gold-catalyzed arylative ring expansion. J. Am. Chem. Soc. 136,
38. Serra, J., Parella, T. & Ribas, X. Au(III)-aryl intermediates in oxidant-free
C-N and C-O cross-coupling catalysis. Chem. Sci. 8, 946–952 (2017).
39. Joost, M. et al. Oxidative addition of carbon–carbon bonds to gold.
Angew. Chem. Int. Ed. 54, 14512–14516 (2015).
40. Joost, M. et al. Facile oxidative addition of aryl iodides to gold(I) by
ligand design: bending turns on reactivity. J. Am. Chem. Soc. 136,
14654–14657 (2014).
41. Zeineddine, A. et al. Rational development of catalytic Au(I)/Au(III)
arylation involving mild oxidative addition of aryl halides. Nat. Commun. 8,
565 (2017).
42. Harper, M. J. et al. Oxidative addition, transmetalation, and reductive
elimination at a 2,2-bipyridyl-ligated gold center. J. Am. Chem. Soc. 140,
4440–4445 (2018).
5844–5847 (2014).
1
3. Kim, S., Rojas-Martin, J. & Toste, F. D. Visible light-mediated gold-catalysed
carbon(sp2)–carbon(sp) cross-coupling. Chem. Sci. 7, 85–88 (2016).
4. Levin, M. D., Kim, S. & Toste, F. D. Photoredox catalysis unlocks single-
electron elementary steps in transition metal catalyzed cross-coupling. ACS
Cent. Sci. 2, 293–301 (2016).
1
1
1
1
5. Skubi, K. L., Blum, T. R. & Yoon, T. P. Dual catalysis strategies in
photochemical synthesis. Chem. Rev. 116, 10035–10074 (2016).
6. Twilton, J., Le, C., Zhang, P., Evans, R. W. & MacMillan, D. W. C. ꢁe merger
of transition metal and photocatalysis. Nat. Rev. Chem 1, 0052 (2017).
7. Patil, D. V., Yun, H. & Shin, S. Catalytic cross-coupling of vinyl golds with
diazonium salts under photoredox and thermal conditions. Adv. Synth. Catal.
43. Martelli, G., Spagnolo, P. & Tiecco, M. Homolytic aromatic substitution by
phenylethynyl radicals. J. Chem. Soc. B 0, 1413–1418 (1970).
44. Xie, J. et al. A highly eꢅcient gold–catalyzed photoredox α–C(sp3)–H
alkynylation of tertiary aliphatic amines with sunlight. Angew. Chem. Int. Ed.
54, 6046–6050 (2015).
3
57, 2622–2628 (2015).
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry

NATure CHeMisTry
Articles
4
5. Hashmi, A. S. K., Ramamurthi, T. D. & Rominger, F. On the
57. Dumele, O., Wu, D., Trapp, N., Goroꢄ, N. & Diederich, F. Halogen
trapping of vinylgold intermediates. Adv. Synth. Catal. 352,
bonding of (iodoethynyl)benzene derivatives in solution. Org. Lett. 16,
971–975 (2010).
4722–4725 (2014).
46. Lowry, M. S. et al. Single-layer electroluminescent devices and photoinduced
hydrogen production from an ionic iridium(III) complex. Chem. Mater. 17,
Acknowledgements
5712–5719 (2005).
We thank Sorbonne Université, CNRS and ANR HyperSiLight for funding and the
Chinese Scholarship Council (for PhD grants to Z.X. and F.Z.). We are grateful to O.
Khaled for HRMS. This work was granted access to the high performance computing
HPC) resources of the HPCaVe Centre at Sorbonne Université and the authors wish to
acknowledge support from the ICMG Chemistry Nanobio Platform-PCECIC, Grenoble,
for calculations facilities. J. Forté is acknowledged for the X-ray diffraction analyses.
4
4
4
5
7. Teders, M. et al. ꢁe energy-transfer-enabled biocompatible disulꢂde–ene
reaction. Nat. Chem. 10, 981–988 (2018).
8. Porter, G. & Wilkinson, F. Energy transfer from the triplet state. Proc. R. Soc.
A 264, 1–18 (1961).
(
9. Kaga, A. et al. Degenerative xanthate transfer to oleꢂns under visible-light
photocatalysis. Beilstein J. Org. Chem. 14, 3047–3058 (2018).
0. Lu, Z. & Yoon, T. P. Visible light photocatalysis of [2+2] styrene
cycloadditions by energy transfer. Angew. Chem. Int. Ed. 51,
Author contributions
Z.X., V.C. and F.Z. performed the synthetic experiments and undertook all the
physicochemical analyses. C.P. conducted the MS analyses. A.E., L.J. and T.L.S.
performed and analysed the luminescence and transient absorption experiments. Y.G.
and H.D. carried out computational studies and V.M-M., C.O. and L.F. designed the
experiments, collated the data and prepared the manuscript.
10329–10332 (2012).
5
1. Welin, E. R., Le, C., Arias-Rotondo, D. M., McCusker, J. K. & MacMillan, D.
W. C. Photosensitized energy transfer-mediated organometallic
catalysis through electronically excited nickel(II). Science 355,
380–385 (2017).
5
5
2. Creutz, S. E., Lotito, K. J., Fu, G. C. & Peters, J. C. Photoinduced Ullmann
C–N coupling: demonstrating the viability of a radical pathway. Science 338,
Competing interests
647–651 (2012).
The authors declare no competing interests.
3. Yoo, W.-J., Tsukamoto, T. & Kobayashi, S. Visible light-mediated Ullmann-
type C–N coupling reactions of carbazole derivatives and aryl iodides.
Org. Lett. 17, 3640–3642 (2015).
Additional information
Supplementary information is available for this paper at https://doi.org/10.1038/
s41557-019-0295-9.
54. Hwang, S. J. et al. Trap-free halogen photoelimination from mononuclear
Ni(III) complexes. J. Am. Chem. Soc. 137, 6472–6475 (2015).
55. Shields, B. J. & Doyle, A. G. Direct C(sp )−H cross-coupling enabled
3
Reprints and permissions information is available at www.nature.com/reprints.
Correspondence and requests for materials should be addressed to C.O., V.M.-M or L.F.
by catalytic generation of chlorine radicals. J. Am. Chem. Soc. 138,
12719–12722 (2016).
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in
5
6. Heitz, D. R., Tellis, J. C. & Molander, G. A. Photochemical nickel-catalyzed
C–H arylation: synthetic scope and mechanistic investigations. J. Am. Chem.
Soc. 138, 12715–12718 (2016).
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2019
NATꢁꢂE CHEMꢃSTꢂY | www.nature.com/naturechemistry