Dehalogenation reaction photocatalyzed by homoleptic copper(i) complexes associated with strongly reductive sacrificial donors

Article abstract of DOI:10.1039/d1cy01209f

In order to perform challenging reduction reactions with light, at low cost and low toxicity, we aim at using for the first time a reductive quenching cycle with a simple, strongly colored homoleptic copper(i) complex [Cu(dipp)₂]⁺(dipp = 2,9-diisopropyl-1,10-phenanthroline). Complexes of this family being weak photo-oxidants, we specifically designed and synthesized powerful, recyclable sacrificial electron donors D. We demonstrate that, during irradiation with LED light in the presence of D, the strong reductant [Cu(dipp)₂]⁰is efficiently photo-generated. Further, we present the first photochemical reaction using photo-generated [Cu(dipp)₂]⁰and evidence that the kinetics of the overall reaction are strongly affected by the oxidation potential of the sacrificial donorE(D⁺/D). Adapting the thermodynamics of sacrificial donors D and [Cu(dipp)₂]⁺has thus allowed us to unlock a brand new concept, giving access to cheap, non-toxic solar light-generated very strong reductive power.

Full text of DOI:10.1039/d1cy01209f

Catalysis
Science &
Technology
View Article Online
View Journal
COMMUNICATION
Dehalogenation reaction photocatalyzed by
homoleptic copper(I) complexes associated with
Cite this: DOI: 10.1039/d1cy01209f
strongly reductive sacrificial donors†
Received 6th July 2021,
Accepted 13th July 2021
Lea Gimeno, Clémence Queffélec, * Kevin Mall Haidaraly,
DOI: 10.1039/d1cy01209f
Errol Blart and Yann Pellegrin
*
rsc.li/catalysis
I
+
In order to perform challenging reduction reactions with light, at
low cost and low toxicity, we aim at using for the first time a
formula [Cu (NN) ] (PSCu) and heteroleptic ones of the
2
I
+
general formula [Cu (PP)(NN)] (PSCu′) where NN is a sterically
burdened chelating diimine ligand and PP is a bulky
bisphosphine ligand. Both sorts of complexes are abundantly
reductive quenching cycle with
homoleptic copper(I) complex
a
simple, strongly colored
+
6
[Cu(dipp)
2
]
(dipp
=
7
2
,9-diisopropyl-1,10-phenanthroline). Complexes of this family
used in the frame of oxidative quenching-based processes
being weak photo-oxidants, we specifically designed and
synthesized powerful, recyclable sacrificial electron donors D. We
because they exhibit very strong photoreductive power (e.g.
+
−1.43 V vs. SCE for [Cu(dap) ] (ref. 8 and 9) and −1.44 for
2
+
10
demonstrate that, during irradiation with LED light in the
Cu(Xantphos)(dmp) )
and can react with a variety of
0
presence of D, the strong reductant [Cu(dipp)
2
]
is efficiently
electron acceptors, once they are in their excited state.
Conversely, Cu(I) complexes are less used in the frame of
photo-generated. Further, we present the first photochemical
0
7
reaction using photo-generated [Cu(dipp)
2
]
and evidence that
reductive quenching-based processes. Several examples of
the kinetics of the overall reaction are strongly affected by the
photochemical cycles involving the reductive quenching of the
excited state of heteroleptic Cu(I) complexes can be found in
+
oxidation potential of the sacrificial donor E(D /D). Adapting the
+
11–15
thermodynamics of sacrificial donors D and [Cu(dipp)
2
]
has thus
the literature.
In those examples, RQ was advantageously
I
− 0
allowed us to unlock a brand new concept, giving access to
cheap, non-toxic solar light-generated very strong reductive
power.
used to generate the reduced state [Cu (PP)(NN˙ )] in the
presence of an electron donor (e.g. triethylamine). This reduced
species is a very strong reductant (from ca. −1.6 V to −1.8 V vs.
13,16,17
SCE)
and is better than or comparable to many
ruthenium and iridium PS under the same conditions (e.g. −1.3
Introduction
+
2
V vs. SCE for [Ru(bpy) ] (ref. 18) and −1.4 V vs. SCE for
3
+
19
Organic photochemistry is a very active research field where
light is used to drive challenging reactions (involving, in
particular, reactive radicals) under mild, green and
[
3 2
Ir(dFCF ppy) (bpy)] ).
RQ with homoleptic Cu(I) complexes (PSCu) would also
present many advantages for the following reasons: they
feature very negative reduction potentials like the
sustainable conditions, with
a
high tolerance towards
1
,2
functional groups.
Most of the time, the addition of a
heteroleptic parent complexes, making them equally potent
photosensitizer PS in the reaction mixture is necessary. While
heavy metal (e.g. ruthenium or iridium) based complexes are
frequently used as a PS, copper(I) complexes are playing a
more and more important role in modern photochemistry as
cheaper, less noxious and efficient photosensitizers. In the
context of organic photochemistry, there are mainly two sorts
of Cu(I) complexes: homoleptic complexes of the general
+
reductants after RQ (e.g. −1.6 V vs. SCE for [Cu(dipp) ] where
2
dipp stands for 2,9-diisopropyl-1,10-phenanthroline). Second,
they exhibit a broad absorption band in the visible region,
where usual efficient PSCu′ are less efficient light collectors
3
–5
(
Fig. S1†) although significant progress has been made to
20–23
+
red-shift their absorption spectrum.
Third, [Cu(NN)2]
complexes are easy to prepare and more stable in the long
+
24,25
term than the [Cu(PP)(NN)] ones.
Finally, the excited
2
6
state of PSCu is sufficiently long-lived to ensure reactivity
with a donor D.
CEISAM, UMR 6230, Université de Nantes, CNRS, 2 chemin de la Houssinière,
Nantes F-44000, France. E-mail: clemence.queffelec@univ-nantes.fr,
yann.pellegrin@univ-nantes.fr
However, reports of RQ processes involving PSCu are
extremely rare compared to PSCu′ because the excited state of
homoleptic copper(I) complexes is a weak photo-oxidant,
†
Electronic supplementary information (ESI) available. See DOI: 10.1039/
d1cy01209f
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.

View Article Online
Communication
Catalysis Science & Technology
2
7–29
−
making them unable to react with usual electron donors.
E(C1*/C1 ) is equal to 0.46 V vs. SCE in THF, and the relevant
+
Efforts have been made to increase the photo-oxidizing
strength of PSCu (quantitatively described by E(PS* /PSCu ))
) to less
negative values. This strategy proved to be efficient to
implement RQ but the reductive power of PSCu was
target D should exhibit an irreversible oxidation step at E(D /
D) = 0.16 V vs. SCE at least.
Benzimidazoline donors, epitomized by famous BIH
(Scheme 1), fulfill almost all required parameters. However,
E(BIH /BIH) is too positive to allow RQ with common
homoleptic complexes. The potential of benzimidazoline
donors can, however, be efficiently tuned by chemical
engineering. Jin-Pei Chen et al. successfully demonstrated
that E(BIH /BIH) could be substantially lowered when
electron donating groups are tethered either to cycles A or B
(Scheme 1). Aiming at implementing RQ involving C1 and a
−
Cu
by shifting the reduction potential E(PSCu/PSCu
−
+
−
3
4
sacrificed. Another more recent strategy is based on
increasing the energy of the excited state. Improvements of
35
E(PS* /PS
−) are reported without altering E(PSCu/PSCu
Cu
−). A
Cu
+
different, complementary approach is to shift the oxidation
potential of the electron donor D to less positive values. This
3
0,31
method was successfully used by McMillin et al.
Several
ferrocene derivatives featuring very low oxidation potentials
benzimidazoline donor, we thus prepared new BI
H and
1
+
were used as a D in the presence of [Cu(dipp) ] or
BI H donors with multiple electron-rich methoxy groups
2
2
+
[
Cu(dpp)
2
]
(where dpp stands for 2,9-diphenyl-1,10-
grafted on cycles A and B of the BIH molecular skeleton
(Scheme 1).
phenanthroline) as a PSCu. A minimum experimental driving
Ã
force of 300 meV for RQ to occur between PS and D was
In this contribution, we focus on the design of new donors
D based on a benzimidazoline molecular frame such that RQ
with benchmark C1 is thermodynamically possible. We
implement for the first time reductive quenching of the
excited state of a strongly colored homoleptic copper(I)
complex, by using specifically designed benzimidazoline
donors and mild light sources. A very powerful and cheap
Cu
determined on this occasion. However, the ferrocenium/
ferrocene couples are reversible and
recombination takes place between D and PSCu
a
fast charge
, preventing
+
−
the accumulation of the latter and to manage a reduction
+
reaction with a substrate. Trials involving [Cu(dpp)2] with
irreversible
electron
donors
(dithionite
and
−
0
dihydroxyquinone) were unsuccessful owing to their
reductant [Cu(dipp)(dipp˙ )]
is photo-generated and
oxidation potential being too positive with respect to E(PS* /
accumulated in the photolyzed medium, and this
unprecedented process is used to achieve a challenging
reduction reaction in the field of organic photochemistry.
Efficient sacrificial donors D should be colorless (in order
not to interfere with light collection by C1) and exhibit an
Cu
3
2
PSCu ). It thus appears relevant to develop electron donors
−
+
D such that D is irreversibly oxidized and E(D /D) is shifted
at a very low anodic potential. RQ can be basically described
Ã
as a photo-induced electron transfer process from D to PS
.
Cu
+
oxidation at a sufficiently low potential E(D /D) to allow
photo-induced electron transfer from D to C1*, and the
The thermodynamics of the reaction thus depend on both
species. In order to refine our objectives, we focus on
oxidation must be irreversible to avoid counter-productive
charge recombination between D and C1 . Indeed, McMillin
et al. successfully evidenced RQ between C1* and ferrocene
implementing RQ with
a
simple homoleptic complex
(hereafter named C1). The latter was chosen
because it features a very negative reduction potential (ca.
+
−
+
2
[Cu(dipp) ]
3
0,31
1
7,33
donors
but the reversibility of the ferrocenium/ferrocene
−
1.6 V vs. SCE)
and one of the most intense absorption
−
couples prevented C1 from accumulating and using its
strong reductive power in further reactions. However, this
work permitted to determine that a minimum driving force
of 300 meV is necessary to drive RQ. Knowing that for C1, the
band in the visible region, for a homoleptic copper(I) complex
ε455nm = ca. 6700 M cm , Fig. S1†). Additionally, C1 is
−
1
−1
(
stable, with a long excited state lifetime (200 ns in degassed
1
7
THF). Knowing that for C1, the photo-oxidation potential
Scheme 1 Illustration of the strategy developed in this contribution: to implement for the first time reductive quenching of the excited state of
homoleptic copper(I) complexes by using benzimidazoline donors.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Catalysis Science & Technology
Communication
−
photo-oxidation potential E(C1*/C1 ) is equal to 0.46 V vs.
Importantly, no degradation was monitored for at least 3
hours, namely the duration of the photolysis trials.
+
SCE in THF, and the relevant target D should exhibit E(D /D)
=
0.16 V vs. SCE at least. Benzimidazoline donors, epitomized
1
Cyclic voltammetry experiments were conducted for BI H
by famous BIH (Scheme 1), fulfill all required parameters but
their oxidation potential must be adapted to the low photo-
and BI H in dry THF (cyclic voltammograms are shown in
2
Fig. S4 and S5†). The forward scans feature in both cases an
3
5
+
oxidation power of C1. Jin-Pei Chen et al. successfully
intense oxidation wave associated with BI
m
H → BI
m
H . A less
+
demonstrated that E(BIH /BIH) could be substantially
intense cathodic wave appears on the reverse scan, indicating
+
lowered when electron donating groups are tethered either to
cycles A or B (Scheme 1). Aiming at implementing RQ
involving C1 and a benzimidazoline donor, we thus prepared
new BI H and BI H donors with multiple electron rich-
the partial reversibility of the BI H /BI H couple. In the
m
m
presence of one equivalent of triethylamine (TEA), the
cathodic waves disappear (Fig. S4 and S5†). Indeed, oxidation
+
of BI H into BI H˙ is then followed by deprotonation of the
1
2
m
m
methoxy groups grafted on cycles A and B of the BIH
m
latter into BI ˙, a strongly reductive species (Fig. S6†).
molecular skeleton (Scheme 1). BI H was synthesized using
the procedure developed by Jin-Pei Cheng et al., using
dithionite as a reductant agent and para-anisaldehyde. The
Assigning a reliable potential value for an irreversible process
is not straightforward. Following Nicewicz and Vullev's
1
3
5
3
7,38
+
lead,
m m
we choose to quantify E(BI H /BI H) under our
same strategy was envisioned in the case of BI H. However,
conditions as the half-peak potential Ep/2 in the absence of
2
in order to save unnecessary synthesis steps, we started with
TEA (Table 1), considering that the CE character of BI H
m
4
,5-dimethoxy-1,2-dinitrobenzene (1′) (easily obtained on a
oxidation in basic medium distorts the oxidation waves. Edge
potentials Ee (where the faradaic current starts to grow,
indicating the beginning of the donors' electroactivity) are
3
6
gram scale instead of 4,5-dimethoxynitroaniline), under the
same above experimental conditions, with dithionite and
benzaldehyde (2′). Rewardingly, after 7 days of reaction, the
desired benzimidazole derivative (3′) was isolated with 90%
yield. Both precursors 3 and 3′ could be obtained on a gram
scale starting with 5 g of 1 or 1′. Then, the imidazole nitrogen
atoms were methylated in the presence of methyl iodide and
target compounds (BI
colorless powders, by reduction of [BI
NaBH , respectively, (Scheme 2 and detailed procedures in
1
also given. As expected, the values of Ep/2 for both BI H and,
in particular, BI H are significantly less positive than the case
2
of benchmark BIH due to the electron donating effect of the
methoxy groups (Table 1; Fig. S4 and S5†). BI H notably
2
features a remarkably low Ep/2 value of 0.13 V vs. SCE, which
is strongly shifted towards less positive potentials by ca. 300
1 2
H and BI H) were finally obtained as
+
−
+ −
1
,I ] or [BI
2
,I ] with
mV compared
benzimidazoline donors.
to
that
of
previously
published
3
5
4
the ESI†).
We next monitored the evolution of the UV–Vis spectrum
of C1 in THF, in the presence of benzimidazoline donors and
triethylamine, under irradiation with a blue LED. For both
1 2
The synthesis of BI H and BI H thus proceeded smoothly:
it is reproducible, scalable, does not require to isolate
unstable species nor drastic conditions for the reduction of
the nitro groups, and the overall yield of the synthesis (72%
over 3 steps for BI H and 55% for BI H) is satisfying.
1 2
BI H and BI H, the solution darkened and a fine powder
precipitated. Light scattering prevents suitable UV-vis spectra
from being recorded and the same experiment is thus
performed in acetonitrile where photolyzed solutions
remained transparent. Under those conditions, a broad band
at 575 nm quickly appears (Fig. S8 and S9†) which is in good
2
1
Both BI
1
H and BI
2
H are stable as powders (shelf stability:
1
several months). H NMR studies of BI
1
H and BI
2
H in
aerated deuterated THF (Fig. S2 and S3†) revealed mild
degradation of BI H after 20 hours (ca. 10%) under ambient
conditions. A fine precipitate was observed, collected and
−
m
agreement with the spectrum of C1 obtained by reductive
spectro-electrochemistry (Fig. S10†). We thus assigned the
+
+
identified as BI1 (counter-anion undetermined) and BI2
monitored spectral changes to the photo-induced formation
−
(counter-anion undetermined), probably resulting from the
of C1 , proving that C1* is reductively quenched by BI
1
H and
O
2
-promoted oxidation of corresponding benzimidazolines.
2 1
BI H despite the rather weak estimated driving force for BI H
1 2
Scheme 2 Synthesis of BI H and BI H sacrificial donors.
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.

View Article Online
Communication
Catalysis Science & Technology
Table 1 Electrochemical data and deduced driving forces for electron
transfer in the frame of RQ. Potentials are given vs. SCE. Cyclic
voltammetry has been conducted in THF with 0.1 M tetrabutylammonium
hexafluorophosphate
involving the BI H/C1 couple in a demanding reductive
m
process. We selected the photo-assisted dehalogenation of
1
3,39
aromatic compounds which is known to be challenging,
paving the way towards fruitful aryl radical chemistry.
1
5,39,40
+
+
a
Ep/2(BI
m
H /BI
m
H)
E (BI
e m
H /BI
m
H)
ΔGRQ (eV)
We chose to start with 4-bromo-benzophenone Bzp-Br as a
substrate (Ep/2 = −1.62 V vs. SCE, Fig. S12†) and investigate
the photogeneration of plain benzophenone Bzp-H in
degassed THF and in the presence of C1, TEA and
benzimidazoline donors with visible light, under the same
BI
BI
1
H
H
0.38
0.13
0.2
0.0
−0.08
−0.33
2
a
−
+
Calculated using the equation ΔGRQ = −F·[E(C1*/C1 ) – Ep/2(BI
m
H /
−
−
BI
m
H)] with E(C1*/C1 ) = 0.46 V vs. SCE. E(C1*/C1 ) estimated with
−
−
13
the Rehm and Weller equation E(C1*/C1 ) = E(C1/C1 ) + E00, where
E(C1/C1 ) = −1.65 V vs. SCE (Fig. S7†) and E00 = 2.11 eV.
conditions described by Michelet et al. (blue LED, 450 nm).
−
17
The photolyzed mixture is regularly monitored by gas
chromatography. Knowing that copper(I) complexes are
strong photo-reductants, we first confirmed that C1 was
unable to reduce Bzp-Br via the well-known oxidative
quenching process (Table 2, entry 1). On the other hand,
(
Table 1). Stern–Volmer analysis (Fig. S11†) allowed the
7
estimation of the rate constants for RQ at kRQ = 4.4 × 10 and
1
8
−1
−1
.4 × 10
s
M
when the quencher is BI H and BI H,
when a mixture of C1, TEA, BI H and Bzp-Br was irradiated
1
2
1
respectively. RQ is faster with BI H than with BI H thanks to
(entry 2) the solution darkened and a fine white powder
precipitated. The conversion gradually increases with time to
reach a final value of 95% after 3 hours of irradiation. The
precipitate and the filtrate were separated by centrifugation.
2
1
the improved driving force for the electron transfer (Table 1).
Nevertheless, the kinetics are sluggish compared to that of
the case of a heteroleptic copper(I) complex in the presence
2
5
1
of BIH. In the latter situation, however, the Gibbs free
energy for the electron transfer is much larger than in our
case, and it is known that RQ with copper(I) complexes is
particularly demanding from the thermodynamic point of
view because of the large reorganization energy which is
H NMR of the filtrate confirms that all Bzp-Br has been
consumed and the signals of Bzp-H are clearly observed,
+
−
1
1
while the precipitate is identified as [BI , Br ] by H NMR
(Fig. S13†). When the same experiment is conducted with
BI H instead of BI H (entry 3), the kinetics of the reaction are
2
1
−
32
expected in the transformation from C1* to C1 .
spectacularly improved: GC monitoring reveals a conversion
of 99% after only 30 minutes of light soaking (for comparison
Achieving reductive quenching of C1* with an irreversible
electron donor not only allows the generation of the strong
reductant C1 but also permits the accumulation of the latter
with BI H after 30 min, the conversion is 42%). Analysis
results of the filtrate reveals the spectral features of Bzp-H
1
−
in the photolyzed medium, opening a new avenue towards
the photo-assisted reduction of challenging substrates. This
encouraged us to implement a photochemical reaction
(Fig. S14†). This improvement is likely due to the larger
driving force of the reductive quenching of C1* by BI H.
2
Importantly, when light is masked, the conversion stops and
Table 2 Results of the photochemical experiments. Reaction conditions: C1 = 5 mol%, [TEA] = 0.95 M in THF, 0.1 mmol Bzp-Br, and 1 or 2 eq. of D. All
−
counter-anions for the complexes are PF
6
a
Entry
Catalyst
Donor (D)
None
Light
Time (h)
Conversion (yield)
+
1
2
3
4
5
6
7
8
9
[Cu(dipp)
[Cu(dipp)
[Cu(dipp)
[Cu(dipp)
None
2
2
2
2
] , C1
Blue LED
Blue LED
Blue LED
None
3
3
0.5
3
3
3
3
3
3
3
3
3
3
0.5
3
0%
+
] , C1
BI
1
BI
2
BI
2
BI
2
BI
2
H
H
H
H
H
95% (89%)
99% (93%)
0%
0%
0%
+
] , C1
+
] , C1
Blue LED
Blue LED
Blue LED
Blue LED
Blue LED
Green LED
Green LED
Simulated solar light
Blue LED
Blue LED
Blue LED
3
Cu(CH CN)
4
PF
6
+
[Cu(dipp)
[Cu(dipp)
[Cu(Xantphos)(dmp)
[Cu(dipp)
[Cu(Xantphos)(dmp)
2
] , C1
HE
DmFc
16%
0%
+
2
] , C1
+
2
] , C2
BI
2
BI
2
BI
2
BI
2
BI
2
BI
2
BI
2
H
H
H
H
H
H
H
85% (80%)
75% (73%)b
23% (n.d.)
84% (80%)
+
10
11
12
13
14
15
2
] , C1
+
2
] , C2
+
[Cu(dipp)
Ru(bpy)
2
] , C1
c
d
3
PF
6
73% (55%)
+
c
[Cu(dipp)
[Cu(dipp)
2
] , C1
99% (88%)
82% (66%)
+
e
2
] , C1
a
b
c
d
NMR yield with p-anisaldehyde as the internal standard. Not determined. Reaction performed in acetonitrile. Yield obtained after
e
purification by preparative TLC. Reaction performed with one equivalent of D.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Catalysis Science & Technology
Communication
restarts when light is restored, ensuring that the reaction is
not photo-initiated. Control experiments confirmed that no
reaction occurred in the absence of light or C1, and that
successful when the photocatalyst is C1 (entry 9), it is
significantly impeded when C1 is replaced by C2 (entry 11).
C1 still features significant absorbance at this wavelength
+
−1
−1
plain Cu was (photo)catalytically inactive (entries 4, 5 and 6,
(ε5
≈ 230 L mol cm ) compared to C2 (ε
≈ 10 L
50nm
550nm
−
1
−1
respectively). In the absence of BI H, no reaction occurred
mol
cm , Fig. S1†), rationalizing the different
2
proving that TEA (entry 1) cannot play the role of a sacrificial
donor. A traditional donor Hantzsch ester (HE) (entry 7)
showed very little efficiency compared to BI H and BI H
performances of the two photosensitizers. Finally, under
simulated solar light (entry 12), the reaction performed with
C1 was successful. Rewardingly, C1 outperforms the classical
1
2
2
+
because its oxidation potential is too positive to allow RQ
[Ru(bpy)
3
]
complex used as a PS under the same conditions
with C1*. Decamethylferrocene (dmFc) is so far the only
reported donor able to reduce C1* because E(dmFc /dmFc) is
(entries 13 and 14) proving the relevance of using homoleptic
copper complexes to drive challenging reduction reactions.
Furthermore, we tested the reaction with only one equivalent
+
3
0,31
sufficiently low.
Yet, attempts with DmFc as a D did not
+
work (entry 8) because dmFc /dmFc is a reversible couple,
and fast recombination between C1 and dmFc prevents
accumulation of C1 . Significantly, when C1 is replaced by
the heteroleptic complex [Cu(xantphos)(dmp)] (C2, Fig. S1†),
of BI H. The reaction is still very efficient (entry 15) but
2
−
+
slower than the case when two equivalents were used (entry
−
+
−
1
3). At the end of each run, we could simply recover [BI , Br ]
+
+
−
and [BI , Br ] by filtration of the photolyzed solution. In the
2
the reaction is slowed down because the heteroleptic complex
case where only one equivalent of BI H was used (entry 15),
2
−
1
−1
+
−
is less sensitive to visible light (ε450nm ≈ 100 L mol cm ,
Fig. S1†) (entry 9). This is more blatant when the
dehalogenation reaction is driven with a longer wavelength
light source (green LED at 550 nm). While the reaction is still
70 μmol of [BI
initial loading of the donor can be restored. We could
regenerate BI H and BI H from the corresponding collected
benzimidazolium by treatment with NaBH under mild
2
, Br ] was collected showing that 70% of the
1
2
4
2
Table 3 Scope of the photochemical experiments. Reaction conditions: C1 = 5 mol%, [TEA] = 0.95 M in THF, 0.2 mmol BI H, 0.1 mmol R–X
Entry
1
R–X
Time
3 h
Conversion (yield)
a
73% (72%)
a
2
30 min
99% (96%)
b
3
1 h
99% (89%)
c
4
5
3 h
3 h
5% (n.d.)
c
20% (n.d.)
c
6
7
3 h
3 h
50% (n.d.)
b
55% (46%)
c
8
3 h
3% (n.d.)
a
b
c
NMR yield with p-anisaldehyde as the internal standard. GC yield with hexadecane as the internal standard. Not determined.
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.

View Article Online
Communication
Catalysis Science & Technology
Scheme 3 Supposed mechanism for the photo-induced dehalogenation reaction of R–X.
conditions, and re-use the compounds in new photochemical
Based on the gathered experimental facts and on previous
studies on a similar reaction, we propose the following
mechanism (Scheme 3): photogenerated C1* is efficiently
4
1
13
reactions. No alteration of the photoconversion efficiencies
was noted while using recycled BI H or BI H. The fact that
1
2
−
the sacrificial donor can be easily regenerated by a cheap
chemical is a strong advantage of this photochemical cycle
and demonstrates the overall renewability of this BIH-based
photochemical reaction. It is worth noting that the ketone
functional group of Bzp-Br was untouched during the
photolysis, proving the selectivity of the photochemical
process. At the end of each run, no elemental copper was
observed in the photolyzed mixtures.
reduced into C1 either by BI
1
H or BI
2
H. Bzp-Br is then reduced
−
−
m
by C1 into Bzp˙ and Br thus regenerating C1. Meanwhile, BI -
H˙ is deprotonated by TEA yielding BI ˙, which is also a strong
reductant (Ep/2 (BI
eventually reduced into a carbanion Bzp either by C1 (path A)
+
m
+
2
2
/BI ˙) = −1.7 V vs. SCE, Fig. S15†). Bzp˙ is then
−
−
or BI ˙ (path B), affording Bzp-H after protonation. We note that
m
−
BI
m
˙ could compete with C1 to also reduce Bzp-Br. However, the
fact that the reaction is considerably less efficient when
2+
Under the optimized conditions and to gain more
information about the substrates that could be activated under
[Ru(bpy)3] is used as a photosensitizer instead of C1 points to
−
the fact that C1 is very active in the reduction of the substrates,
+
these conditions, we then studied the scope and limitation of
being a better reductant than [Ru(bpy)
3
] by more than 300 mV.
+
2
the light-mediated, Cu(dipp) -catalyzed reduction of a series of
Further work is ongoing to decipher the full mechanism
associated with this completely new photochemical cycle.
In summary, we synthesized in a few efficient steps new
benzimidazoline donors BI H and BI H exhibiting very strong
aryl halides possessing various substituents on the backbone.
Reduction of various aryl iodides was first evaluated and found
to proceed very rapidly in all cases (less than 1 h; Table 3,
entries 2 and 3). Reduction of 4-iodobiphenyl was slower
1
2
reductive power. We used them as sacrificial donors in a
photochemical cycle and for the first time we were able to
monitor the productive reductive quenching of the excited
state of a simple and cheap colored homoleptic copper(I)
(Table 3, entry 1) likely because it is less activated than
2-iodoacetanilide or 4-iodomethylbenzoate. Corresponding
deiodinated arenes were obtained in good to excellent yields
regardless of the substituents on the molecular skeleton of the
aryl iodide. The reduction of aryl bromides is more challenging
and debromination reactions proceeded more slowly as
expected, because the C–Br bond cleavage kinetics are slower
complex C1. Using only mild light sources, the strong
−
reductant C1 was photo-generated in the presence of BI H
1
2
and BI H, performing a more efficient challenging photo-
assisted dehalogenation reaction than that of classical
photosensitizers. Importantly, the photo-oxidation of the
designed benzimidazoline donors is well controlled and they
can be easily regenerated. A preliminary scope study revealed
that the photochemical process is compatible with many
substrates. This work paves the way towards virtually waste-
free demanding reduction reactions, driven only by solar
energy, with a cheap and essentially non noxious, highly
colored homoleptic copper(I) complex.
42
upon aryl bromide reduction. While the reaction yield was
very satisfactory in the case of Bzp-Br (Table 2), the conversions
and yields dropped for 4-bromomethylbenzoate (Table 3, entries
6
and 7). Conversions are particularly poor for 4-bromo
biphenyl and 2-bromo acetanilide (entries 4 and 5) because the
13
substrates were not sufficiently activated. Finally, almost no
conversion was monitored for 4-chlorobenzophenone (Table 3,
entry 8), even when the irradiation time was prolonged to 12 h.
This is likely due to the very challenging C–Cl cleavage kinetic
42
rate upon reduction.
Importantly, no conversion was
monitored for all substrates in the absence of C1, except in the Funding sources
case of 4-iodomethylbenzoate. A conversion of 45% was
observed after 3 h of irradiation, which is significantly less
efficient and more sluggish than in the presence of C1.
The authors wish to thank the ANR (PERCO program no.
ANR-16-CE07-0012-01) for financial support.
Catal. Sci. Technol.
This journal is © The Royal Society of Chemistry 2021

View Article Online
Catalysis Science & Technology
Communication
Author contributions
19 D. Hanss, J. C. Freys, G. Bernardinelli and O. S. Wenger, Eur.
J. Inorg. Chem., 2009, 2009, 4850–4859.
The manuscript was written through contributions of all
authors. All authors have given approval to the final version
of the manuscript.
2
0 B. J. McCullough, B. J. Neyhouse, B. R. Schrage, D. T. Reed,
A. J. Osinski, C. J. Ziegler and T. A. White, Inorg. Chem.,
2
018, 57, 2865–2875.
2
1 M. Sandroni, M. Kayanuma, M. Rebarz, H. Akdas-Kilig, Y.
Pellegrin, E. Blart, H. L. Bozec, C. Daniel and F. Odobel,
Dalton Trans., 2013, 42, 14628–14638.
2 H. Takeda, Y. Monma, H. Sugiyama, H. Uekusa and O.
Ishitani, Front. Chem., 2019, 7, 418.
Conflicts of interest
There are no conflicts to declare.
2
References
2
3 S.-P. Luo, N.-Y. Chen, Y.-Y. Sun, L.-M. Xia, Z.-C. Wu, H. Junge,
M. Beller and Q.-A. Wu, Dyes Pigm., 2016, 134, 580–585.
1
N. A. Romero and D. A. Nicewicz, Chem. Rev., 2016, 116,
0075–10166.
1
24 A. J. J. Lennox, S. Fischer, M. Jurrat, S.-P. Luo, N. Rockstroh,
H. Junge, R. Ludwig and M. Beller, Chem. – Eur. J., 2016, 22,
1233–1238.
2
M. H. Shaw, J. Twilton and D. W. C. MacMillan, J. Org.
Chem., 2016, 81, 6898–6926.
3
4
S. Engl and O. Reiser, ACS Catal., 2020, 10, 9899–9906.
A. C. Hernandez-Perez and S. K. Collins, Acc. Chem. Res.,
25 Y. Sakaguchi, A. Call, M. Cibian, K. Yamauchi and K. Sakai,
Chem. Commun., 2019, 55, 8552–8555.
2
016, 49, 1557–1565.
26 M. S. Lazorski and F. N. Castellano, Polyhedron, 2014, 82, 57–70.
27 S. Paria and O. Reiser, ChemCatChem, 2014, 6, 2477–2483.
28 C. Sandoval-Pauker, G. Molina-Aguirre and B. Pinter,
Polyhedron, 2021, 199, 115105.
29 N. Armaroli, G. Accorsi, F. Cardinali and A. Listorti, in
Photochemistry and Photophysics of Coordination Compounds
I, ed. V. Balzani and S. Campagna, Springer, Berlin,
Heidelberg, 2007, pp. 69–115.
5
R. Giereth, M. Obermeier, L. Forschner, M. Karnahl, M.
Schwalbe and S. Tschierlei, ChemPhotoChem, 2021, 5,
6
44–653.
A. Hossain, A. Bhattacharyya and O. Reiser, Science,
019, 364, eaav9713.
6
7
8
9
2
B. M. Hockin, C. Li, N. Robertson and E. Zysman-Colman,
Catal. Sci. Technol., 2019, 9, 889–915.
J.-M. Kern and J.-P. Sauvage, J. Chem. Soc., Chem. Commun.,
30 K. L. Cunningham and D. R. McMillin, Inorg. Chem.,
1998, 37, 4114–4119.
31 K. L. Cunningham, C. R. Hecker and D. R. McMillin, Inorg.
Chim. Acta, 1996, 242, 143–147.
1
987, 546–548.
M. Pirtsch, S. Paria, T. Matsuno, H. Isobe and O. Reiser,
Chem. – Eur. J., 2012, 18, 7336–7340.
1
1
0 Y. Zhang, M. Heberle, M. Wächtler, M. Karnahl and B.
Dietzek, RSC Adv., 2016, 6, 105801–105805.
1 B. Wang, D. P. Shelar, X.-Z. Han, T.-T. Li, X. Guan, W. Lu, K.
Liu, Y. Chen, W.-F. Fu and C.-M. Che, Chem. – Eur. J.,
32 A. K. I. Gushurst, D. R. McMillin, C. O. Dietrich-Buchecker
and J. P. Sauvage, Inorg. Chem., 1989, 28, 4070–4072.
33 M. K. Eggleston, D. R. McMillin, K. S. Koenig and A. J.
Pallenberg, Inorg. Chem., 1997, 36, 172–176.
2
015, 21, 1184–1190.
34 A. Rosas-Hernández, C. Steinlechner, H. Junge and M.
Beller, Green Chem., 2017, 19, 2356–2360.
1
1
1
2 A. Call, C. Casadevall, F. Acuña-Parés, A. Casitas and J.
Lloret-Fillol, Chem. Sci., 2017, 8, 4739–4749.
3 B. Michelet, C. Deldaele, S. Kajouj, C. Moucheron and G.
Evano, Org. Lett., 2017, 19, 3576–3579.
4 A. Nitelet, D. Thevenet, B. Schiavi, C. Hardouin, J. Fournier,
R. Tamion, X. Pannecoucke, P. Jubault and T. Poisson,
Chem. – Eur. J., 2019, 25, 3262–3266.
35 X.-Q. Zhu, M.-T. Zhang, A. Yu, C.-H. Wang and J.-P. Cheng,
J. Am. Chem. Soc., 2008, 130, 2501–2516.
36 T. J. Collins, R. D. Powell, C. Slebodnick and E. S. Uffelman,
J. Am. Chem. Soc., 1991, 113, 8419–8425.
37 H. G. Roth, N. A. Romero and D. A. Nicewicz, Synlett,
2016, 27, 714–723.
1
1
1
1
5 H. Baguia, C. Deldaele, E. Romero, B. Michelet and G.
Evano, Synthesis, 2018, 50, 3022–3030.
6 O. Green, B. A. Gandhi and J. N. Burstyn, Inorg. Chem.,
38 E. M. Espinoza, J. A. Clark, J. Soliman, J. B. Derr, M. Morales
and V. I. Vullev, J. Electrochem. Soc., 2019, 166, H3175.
39 J.-H. Shon, D. Kim, M. D. Rathnayake, S. Sittel, J. Weaver
and T. S. Teets, Chem. Sci., 2021, 12, 4069–4078.
40 R. S. Shaikh, S. J. S. Düsel and B. König, ACS Catal., 2016, 6,
8410–8414.
41 S. Rohrbach, R. S. Shah, T. Tuttle and J. A. Murphy, Angew.
Chem., Int. Ed., 2019, 58, 11454–11458.
42 C. Costentin, M. Robert and J.-M. Savéant, J. Am. Chem. Soc.,
2004, 126, 16051–16057.
2
009, 48, 5704–5714.
7 S. Garakyaraghi, P. D. Crapps, C. E. McCusker and F. N.
Castellano, Inorg. Chem., 2016, 55, 10628–10636.
8 S. Campagna, F. Puntoriero, F. Nastasi, G. Bergamini and V.
Balzani, in Photochemistry and Photophysics of Coordination
Compounds I, ed. V. Balzani and S. Campagna, Springer,
Berlin, Heidelberg, 2007, pp. 117–214.
This journal is © The Royal Society of Chemistry 2021
Catal. Sci. Technol.