Dual Benzophenone/Copper-Photocatalyzed Giese-Type Alkylation of C(sp3)?H Bonds

Article abstract of DOI:10.1002/chem.201904111

Photocatalyzed Giese-type alkylations of C(sp³)?H bonds are very attractive reactions in the context of atom-economy in C?C bond formation. The main limitation of such reactions is that when using highly polymerizable olefin acceptors, such as unsubstituted acrylates, acrylonitrile, or methyl vinyl ketone, radical polymerization often becomes the dominant or exclusive reaction pathway. Herein, we report that the polymerization of such olefins is strongly limited or suppressed when combining the photocatalytic activity of benzophenone (BP) with a catalytic amount of Cu(OAc)₂. Under mild and operationally simple conditions, the Giese adducts resulting from the C(sp³)?H functionalization of amines, alcohols, ethers, and cycloalkanes could be synthesized. Preliminary mechanistic studies have revealed that the reaction does not proceed through a radical chain, but through a dual BP/Cu photocatalytic process, in which both Cu^II and low-valent Cu^I/0 species, generated in situ by reduction by the BP ketyl radical, may react with α-keto or α-cyano intermediate radicals, thus preventing polymerization.

Full text of DOI:10.1002/chem.201904111

A Journal of
Accepted Article
Title: Dual Benzophenone/Copper Photocatalyzed Giese-Type
Alkylation of C(sp3)−H Bonds
Authors: Baptiste Abadie, Damien Jardel, Gianluca Pozzi, Patrick
Toullec, and Jean-Marc Vincent
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Chem. Eur. J. 10.1002/chem.201904111
Link to VoR: http://dx.doi.org/10.1002/chem.201904111
Supported by

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
Dual Benzophenone/Copper Photocatalyzed Giese-Type
Alkylation of C(sp3)−H Bonds
Baptiste Abadie,[a] Damien Jardel,[a] Gianluca Pozzi,[b] Patrick Toullec,*[a] and
Jean-Marc Vincent*[a]
Abstract: Photocatalyzed Giese-type alkylations of C(sp3)−H bonds
are very attractive reactions in the context of atom-economy in C−C
bond formation. The main limitation of such reactions is that when
using highly polymerizable olefin acceptors, such as unsubstituted
acrylates, acrylonitrile or methyl vinyl ketone, radical polymerization
often becomes the dominant or exclusive reaction pathway. Herein,
we report that the polymerization of such olefins is strongly limited or
suppressed when combining the photocatalytic activity of
benzophenone with a catalytic amount of Cu(OAc)2. Using mild and
operationally simple conditions, the Giese adducts resulting from the
C(sp3)−H functionalization of amines, alcohols, ethers and
cycloalkanes could be synthesized. Preliminary mechanistic studies
revealed that the reaction does not proceed through radical chain, but
through a dual BP/Cu photocatalytic process in which both CuII and
low valent CuI/0 species generated in situ by reduction by the BP ketyl
radical may react with the α-keto or α-cyano intermediate radicals,
thus preventing polymerization.
transformation.
Interestingly,
these
complementary
methodologies enable alkylations of C−H bonds from a broad
scope of both donors and acceptors, with the noteworthy
exception of easily polymerizable olefin acceptors, such as
unsubstituted acrylates, acrylonitrile or methyl vinyl ketone, for
which radical polymerization is often the dominant reaction
pathway leading to polymers as the major or exclusive products.
In a recent work on site-selective α-C(sp3) −H alkylation of primary
amines, Rovis and coworkers overcame this limitation by
employing an excess of the HAT catalyst (quinuclidine) i.e. 200
mol%.[6c] It should be also noted that photocatalyzed Giese
reactions exploiting C−X bonds activation (X ≠ H) have also been
recently developed.[8] Overall, only
a
few examples of
photocatalyzed C−H/C−X bonds alkylations by unsubstituted
acrylates have been reported in these studies. With the objective
of avoiding polymerization issues, Hong and coworkers exploited
a dual photocatalysis/Ni catalysis approach combined to enone
photochemistry (Scheme 1, approach B).[9] While this
methodology proved to be very effective for the functionalization
of ethers, it has an obvious inherent limitation on the acceptor side,
being restricted to enones.
Introduction
The catalytic alkylation of C(sp3)−H bonds represents one of the
most attractive but challenging method to functionalize organic
compounds.[1] In this domain, the use of photocatalysts is
attracting considerable interest.[2] This is due to the ability of their
photoexcited triplet states to perform one-electron oxidations
and/or hydrogen atom abstractions to generate carbon-centered
radicals from C(sp3)−H bonds. Once generated, a particularly
useful way of functionalizing the alkyl radicals is by engaging them
in a Giese-type reaction (Scheme 1, approach A).[3,4] After fast
addition of such nucleophilic radicals to electron-deficient olefins,
the newly formed alkyl radicals could be trapped either by one-
electron reduction followed by protonation or by reverse hydrogen
atom transfer (RHAT). Catalytic systems employing photoredox
catalysts,[5] photoredox catalysts combined with HAT catalysts
(Indirect HAT process),[6] or HAT photocatalysts only (Direct HAT
process),[7] have been shown to be effective to perform such
[a]
[b]
B. Abadie, D. Jardel, Dr. Prof. P. Toullec, Dr. J.-M. Vincent
Institut des Sciences Moléculaires, CNRS UMR5255, Univ.
Bordeaux, 33405 Talence, France
E-mail: jean-marc.vincent@u-bordeaux.fr
patrick.toullec@u-bordeaux.fr
Dr. G. Pozzi
Scheme 1. Photocatalyzed Giese reaction of C(sp3)-H donors vs
CNR-Istituto di Scienze e Tecnologie Molecolari (ISTM), via Golgi
19, 20133 Milano, Italy
photosensitized radical polymerization of olefins.
Supporting information for this article is given via a link at the end of
the document.((Please delete this text if not appropriate))
In recent years our group has exploited the photoreactivity of
benzophenone (BP) to initiate the radical iodoperfluoroakylation
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
of alkenes, or the α-C(sp3)−H functionalization of ethers when
combined with NFSI as an oxidant.[10] We also merged the
benzophenone photoreactivity to that of copper(II) complexes to
develop photoreducible copper(II) precatalysts for application in
ethyl acrylate to afford the alkylated products 1, 4 and 5 in ~ 50%
isolated yields. Less selective reactions for the
Table 1. Optimization of the reaction conditions.[a]
copper-catalyzed
azide
alkyne
cycloaddition
or
trifluoromethylation of unactivated terminal alkenes.[11] Based on
these studies and previous literature examples, in particular the
seminal work of Inoue, Kamijo and coworkers on the aromatic
ketones photocatalyzed C(sp3)−H bonds functionalization,[12] we
hypothesized that combining the BP photoreactivity with that of
copper could be helpful to limit the polymerization in Giese type
reactions, as the alkyl radicals intermediates I and II (Scheme 1)
might react with CuII to generate masked radicals,[13] while low-
valent CuI/0 species photogenerated in situ might reduce II.[8c] We
report here that the combined use of catalytic amounts of BP and
Cu(OAc)2 under mild UVA irradiation allows to isolate the Giese
reaction products in moderate to good yields from a range of
donors, including trialkylamines, and highly polymerizable
acceptors such as ethyl acrylate, acrylonitrile and methyl vinyl
ketone.
deviation from « standard »
conditions
conversion
(%)[b]
yield
(%)[c]
entry
1
2
none
100
100
100
0
61
< 5
60[d]
0
without Cu(OAc)2
3
BP 20 mol%
4
without irradiation
5
without BP
15
0
6
without degassing
100
90
40
19
15
< 5
< 5
15
7
Cu(OTf)2 instead of Cu(OAc)2
CuCl2 instead of Cu(OAc)2
CuBr2 instead of Cu(OAc)2
Ni(OAc)2 instead of Cu(OAc)2
NiCl2.6H2O instead of Cu(OAc)2
Results and Discussion
8
83
9
100
100
58
The reaction between N-Boc-piperidine and ethyl acrylate (Table
1) served as a benchmark to test our working hypothesis. The
optimization experiments, run in NMR tubes in deuterated
solvents on 0.1 mmol of acrylate, were followed by 1H NMR.
Irradiation was conducted using a mild UVA source, i.e. a low
pressure mercury lamp type Thin Layer Chromatography set at
365 nm (6 W, emission from 320 to 390 nm, Figure S1).
Pleasingly, after optimization (Table 1) the Giese adduct 1 could
be obtained in satisfactory 61% yield in CD3CN (see table S1 for
solvents screening) in 1 h reaction using 50 mol% of BP and 2
mol% of Cu(OAc)2 (Table 1, entry 1). The acrylate was fully
10
11
[Cu(CH3CN)4]PF6 instead of
Cu(OAc)2
12
100
< 5
13
14
CuI instead of Cu(OAc)2
80
< 5
30
CuOAc instead of Cu(OAc)2
100
Co(OAc)2.4H2O instead of
Cu(OAc)2
15
100
14
consumed,
the
by-products
being
telomers/polymers.
Remarkably, in the absence of Cu(OAc)2 only telomers/polymers
were formed (Table 1, entry 2), highlighting its key role in limiting
the acrylate polymerization at low loading. Higher loadings (10
mol%, 100 mol%) led to slower and less selective reactions. The
BP loading could be reduced to 20 mol% when irradiation was
conducted with two lamps (Table 1, entry 3). Light and BP were
mandatory for the reaction (Table 1, entries 4-5), while without
degassing the reaction proceeded in 40% yield (Table 1, entry 6).
Cu(OAc)2 was found to be much superior to the other CuII, CuI,
NiII and CoII salts tested (Table 1, entries 7-15). Finally, BP was
by far the most reactive and selective photocatalyst among the
aromatic ketones tested, which included substituted
benzophenones, phenylglyoxylic acid, ethyl benzoylformate,
xanthone, thioxanthone and fluorenone (Table S1).
[a] Reaction conditions: amine (0.3 mmol), acrylate (0.1 mmol), BP, MnXn,
CD3CN (1 mL), degassing by Ar bubbling (10 min), irradiation for 1 h of the NMR
tube by a TLC lamp set at 365 nm (6W, emission: 320-390 nm) placed at ~ 0.5
cm of the tube. [b] Conversion of acrylate determined by 1H NMR using the BP
resonances as internal standard. [c] 1H NMR yields of 1 using the BP
resonances as internal standard. Due to the paramagnetism of the Co species
present at the end of the reaction, the yield was determined by isolating 1 by
column chromatography. [d] Irradiation with two lamps.
Giese adducts vs polymers were observed when employing the
methyl or benzyl acrylates, leading to isolated yields of 30% for 2
and 3. Interestingly, a new acrylate bearing a nonafluoro-tert-
butoxyethyl group could be employed to afford the fluorinated
piperidine 6 in 48% yield. Remarkably, engaging the Boc-
piperidine with methyl methacrylate or cyclohexenone in the
reaction led to polymerization only, along with the almost
complete consumption of the BP. This may reflect the lower
reactivity of the intermediate alpha keto tertiary radical towards
the reducing Cu(I/0) species generated in situ, compared to that
of the alpha keto secondary radical produced from ethyl acrylate
(see mechanistic study and discussion below). Acrylonitrile and
methyl vinyl ketone were also successfully engaged in the
With the best conditions in hands the scope of the reaction was
then evaluated (Scheme 2). First, it should be mentioned that in
the absence of Cu(OAc)2, all the reactions conducted to assess
the
substrate
scope
led
to
>
95%
olefin
telomerization/polymerization (For a representative example see
the 1H NMR spectra given in the Figure S2). In the presence of 2
mol% Cu(OAc)2, cyclic Boc-protected amines reacted well with
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
Scheme 2. Photocatalyzed Giese reaction of C(sp3)-H donors vs photosensitized radical polymerization of olefins. [a] 5 equiv of EtOH were used. [b]
Transesterification: 5 mol% TsOH, CHCl3, rt, 24 h.
reaction, leading to the N-Boc-piperidines 7 and 8 in 59% and
49% yields, respectively. The BP triplet state being a good one-
electron oxidizing agent of electron-rich amines,[14] trialkyl and
dialkyl aryl amines were also examined as donor partners. To our
delight, the tropane methyl group was alkylated smoothly by
acrylates, acrylonitrile and methyl vinyl ketone to afford 9-12 in
good yields. N,N-dimethylbenzylamine was selectively alkylated
on the methyl group to provide the trialkylamine 13 in 28%
isolated yield. As already observed with other photocatalytic
systems,[5e] the reaction between N,N-dimethylaniline and ethyl
acrylate led to the formation of an unseparable mixture of the
Giese adduct 14 and bicyclic product 14’ that were isolated in
26% overall yield. This highlights a competition between the
RHAT step and the intramolecular addition of the radical
intermediate on the phenyl
ring. The selectivity issue for the CH3 vs α-CH2 alkylation in
trialkylamines was evaluated employing N-methylpiperidine as
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
donor (Scheme 2). As anticipated, C−H functionalization occurred
on both methyl and methylene groups to afford a ~ 1/0.75 mixture
the α-hydroxyisopropyl radical.[16] As observed in previous studies,
prolonging the irradiation led to the fast appearance of metallic
copper NPs which, in the absence of stabilizing ligands, rapidly
aggregate to form a grey/black fine powder as evidenced by the
broad spectra recorded after 7, 10 and 15 min irradiation,
characteristic of the presence of light-diffusing particles (Figure
1A).
1
(estimated by H NMR) of amines 15 and 15’ that we could not
separate by silica gel column chromatography. Cyclic ethers,
including 3,3-dimethyloxetane, reacted smoothly with ethyl
acrylate to afford the Giese adducts 16-19 in good yields. The
linear ethyl tert-butyl ether (ETBE), a compound widely used as
gasoline additive, could be functionalized as-well to afford 20 in
satisfactory 45% yield.
Isopropanol, cyclohexanol and ethanol were functionalized
effectively with ethyl acrylate to afford the tertiary and secondary
alcohols 21-22 and 24 in good yields. The alcohols 22 and 24
could be easily transformed under acidic conditions into the γ-
valerolactone 23 and spirolactone 25, respectively.
Finally, cycloooctane was functionalized in a good 73% overall
yield to afford a separable mixture of the substituted cyclooctyl
derivatives 26 and 27. Interestingly, the diester 26 was obtained
in ~ 40% yield (average of 3 reactions), while the monoaddition
product 27 could still be isolated in ~ 30% yield. A recently
reported synthesis of this product through functionalization and
valence isomerization of a cubane derivative, followed by PtO2
catalyzed hydrogenation of the obtained cyclooctatetraene
intermediate, led to 27 in 23% overall yield.[15] With cyclohexane
the Giese adduct 28 could be isolated, albeit in low yield. Benzylic
C−H bonds, including those of toluene and benzyl alcohol were
found unreactive in this transformation. This was ascribed to the
very slow rate of addition of benzylic radicals on electron-poor
olefins. As a comparison, the rate of addition of the benzyl radical
1
1
to methyl acrylate (kobsd = 430 M− s− ) is about 5 orders of
magnitude lower than that of the α-hydroxyisopropyl radical (kobsd
1
1
= 3.5 x 107 M− s− ).[3c]
Mechanistic studies were then undertaken. The main mechanistic
issues deal with the role of copper and whether a radical chain
process could operate. The latter point was assessed by running
the reaction between N-Boc-piperidine and ethyl acrylate in the
conditions of entry 3 in table 1, but in the presence of 0.5 % v/v
D2O. NMR and mass spectrometry analysis (Figures S3-S4)
revealed that in the products ~ 85% of the methylene groups in α
position to the carbonyl was monodeuterated. This shows that a
chain reaction implying the direct H-atom abstraction of the
donors by intermediates II (Scheme 1) is a minor or inoperative
pathway, and that the hydrogen atom of the newly formed C−H
bond comes from an intermediate species with a labile proton,
possibly the one of the BP ketyl radical. Similar results were found
for the TBADT photocatalyzed addition of THF to acrylonitrile.[7h]
Concerning the role of copper, the fact that both Cu(OAc)2 and, to
a lower extent, CuOAc were effective in limiting the polymerization
(entries 1, 2 and 14 of Table 1) suggested that high valent and
low valent copper species might be involved in the process. To
get more insights, the reaction between isopropanol and ethyl
acrylate was followed by UV-vis spectroscopy and 1H NMR. In the
absence of ethyl acrylate the full reduction of CuII was observed
in ~ 5 min, as evidenced by the complete disappearance of the
band centered at 675 nm ascribed to a d-d electronic transition of
the d9 CuII ions, (Figure 1A). Such fast reduction agrees with
previous studies on the light-mediated reduction of transition
metal ions (AuIII, CoII, CuII, AgI…) by ketyl radicals, in particular
Figure 1. UV-vis monitoring of the Cu(OAc)2 photoreduction in the presence of
isopropanol (A), or isopropanol + ethyl acrylate (B) and (C). For A and B,
reaction run in a quartz UV cuvette (1 cm path length) closed with a rubber
septum: isopropanol (0.9 mmol), acrylate (0.3 mmol), BP (20 mol%), Cu(OAc)2
(2 mol%), CH3CN (3 mL), degassing by Ar bubbling (10 min), irradiation by a
TLC lamp set at 365 nm (6W, emission: 320-390 nm) placed at ~ 0.5 cm of the
UV cuvette. For C, same conditions as B except that the degassing is conducted
by freeze-pump-thaw cycles followed by blow-torch sealing of the cuvette.
In the presence of acrylate, the reduction proceeded very
effectively too, albeit at a slower rate, and the full CuII reduction
required ~ 20 min of irradiation. Interestingly, no metallic NPs
were generated upon prolonged irradiation, as revealed by the
electronic absorption spectra recorded over a 8 h period (Figure
1B), i.e. the reaction time necessary for the formation of 22 and
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
full consumption of the acrylate (checked by 1H NMR). During this
lapse of time, a very weak absorbance over the whole visible
spectrum was observed, the solution remaining colorless. 1H
NMR analysis of aliquots of the reaction mixture taken at various
times showed that at full CuII reduction, i.e. ~ 30 min reaction time,
neither product 22 nor polymeric by-products were formed, and
that benzophenone and acrylate were intact. It was only after ~
90 min of irradiation that 22 began to form, at which time copper
NPs were not detected (see UV-vis spectrum after 2 h of
irradiation in Figure 1B). Remarkably, it was when the acrylate
had been consumed and the Giese adduct formed that Cu0 NPs
accumulated, as shown by the appearance of the characteristic
plasmon resonance at 580 nm (see spectrum after 18 h irradiation
in Figure 1B) and the photograph of the UV cuvette taken 10 h
after the end of the reaction (Figure S5). It should be noted that
under the same reaction conditions, but in the absence of
Cu(OAc)2, full polymerization was observed in ~ 2.5 h. Also, a
similar behavior was observed by UV-vis spectroscopy for
reactions involving N-Boc-piperidine and tropane as donors, that
is a slowdown of the CuII reduction rate and of copper NPs
formation when the acrylate was present (Figures S6, S7).
Interestingly, the results gathered in Table 1 suggest that the
acetate counterion plays a key role in the reaction, since the use
of CuBr2, CuCl2, Cu(OTf)2, [Cu(CH3CN)4]PF6 or CuI led to acrylate
polymerization. Previous works have shown that the reducing
power of ketyl radicals is increased by the presence of a base, the
reduction process proceeding through proton-coupled electron
transfer (PCeT).[16a] In particular, PCeT assisted by a base has
been exploited to facilitate the reduction of low-valent MI to M0
metal species to generate metallic NPs.[16a] A possible role of the
much more basic acetate counterion (pKa CH3CO2H ~ 4.7)
[Cu(CH3CN)4]+
CH3COO
H
O
Ph
Ph
Scheme 3. Possible mechanism for CuI to Cu0 reduction in CH3CN assisted by
the acetate counterion.
We then wanted to assess whether the Cu0 species that led to the
NPs were formed by CuI disproportionation or through a light-
driven CuI reduction process. The model reaction between
isopropanol and ethyl acrylate was conducted under the usual
conditions, except that the CH3CN solution was deoxygenated by
freeze-pump-thaw cycles, followed by torch sealing of the UV
cuvette to ensure no possible entry of air. At full CuII reduction, i.e.
after ~ 30 min irradiation, the lamp was switched off and the
cuvette placed in the dark for 24 h. As shown in Figure 1C, no
reappearance of the CuII and Cu0 NPs characteristic bands at 675
nm and 580 nm, respectively, was noticed. However, the colorless
solution instantaneously recovered the original blue color when
opened to air (~ 70% and 90% recovery after 30 s and 15 min,
respectively, Figure 1C), confirming that the photogenerated low
valent copper species during the reaction are highly reactive
reducing species. It thus appears that CuI disproportionation did
not occur or was extremely slow compared to the reaction time
scale. Hence, the Cu0 NPs previously shown to accumulate under
irradiation at the end of the reaction were most probably formed
from a CuI to Cu0 light-driven reduction process. Based on
previous studies,[16,11] we propose that the photogenerated α-
22
Hc
Hb
Ha
red
hydroxyisopropyl radical (E1/2 = −0.61 V vs SCE in CH3CN)[17]
and/or BP keto radical (E1/2red = −0.25 V vs SCE in CH3CN)[17] are
likely involved in this process. First, both radicals should be
capable to reduce Cu(OAc)2 (E1/2red = −0.2 V and −0.4 V vs SCE
in acetic acid methanol)[18] to form CuOAc, HOAc, acetone and/or
BP. In acetonitrile, a ligand which strongly stabilizes CuI over CuII,
the well-known and rather stable tetrakis(acetonitrile)
[Cu(CH3CN)4]+ ion should form. Interestingly, its reduction
Figure 2. 1H NMR monitoring for 5 h of the photocatalyzed Giese-type reaction
between isopropanol and ethyl acrylate. Reaction run in a NMR tube closed with
a rubber septum: isopropanol (0.3 mmol), acrylate (0.1 mmol), BP (20 mol%),
Cu(OAc)2 (2 mol%), CD3CN (1 mL), degassing by Ar bubbling (10 min),
irradiation by a TLC lamp set at 365 nm (6W, emission: 320-390 nm) placed at
~ 0.5 cm of the tube.
red
potential to Cu0 (E1/2 = −0.12 V vs SCE in H2O/CH3CN)[19] is
accessible to both the α-hydroxyisopropyl radical and less
reducing BP ketyl radical. The formation of copper NPs at the end
of the reaction between cyclooctane and ethyl acrylate also
supports the fact that under the reaction conditions the reduction
of CuII to Cu0 can be mediated by the BP ketyl radical.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
compared to Br− (pKa HBr ~ −9), Cl− (pKa HCl ~ −8), I− (pKa HI ~
−10), OTf− (pKa HOTf ~ −14) or PF6− (pKa HPF6 ~ −20), could be
to favor the CuI to Cu0 reduction (Scheme 3). This hypothesis is
supported by the UV-vis follow up of the photoreduction of
[Cu(CH3CN)4]PF6 (Figure S8) in the presence of N-Boc-piperidine
as H-donor, which revealed that the appearance of the Cu0 NPs
is much slower compared to that observed with Cu(OAc)2 (Figure
S6A). Typically, an absorbance of ~ 1.2 at the maximum of the
Scheme 4. Proposed mechanism for the dual BP/CuX2 photocatalyzed Giese-type alkylation (In blue, side-reactions that become significant at low acceptor concentration).
plasmon resonance (580 nm) is reached after 30 min irradiation
with Cu(OAc)2 compared to an absorbance of only ~ 0.2 after 90
min of irradiation for [Cu(CH3CN)4]PF6. Further studies that will
specifically address the role of the copper counterion in tuning the
photoreduction efficiency will be conducted and reported in due
course.
of electron-rich trialkylamines, by a very fast charge transfer
followed by proton transfer.[14] Of the two radical species A and B
formed, radical A derived from the donor preferentially adds to the
acceptor to generate the radical C. Seminal works of Kochi and
coworkers have shown that alkyl radicals rapidly react with
Cu(OAc)2 to possibly generate metastable alkylcopper(III)
species.[13] More recently, such alkylcopper(III) have been
proposed as key intermediates in reactions exploiting dual
copper/photoredox catalysis processes.[22] Thus, in the early
stage of the reaction, i.e. when the CuII concentration is still high,
radical intermediate C might react with Cu(OAc)2 to generate an
alkylcopper species D that could be viewed as a masked radical.
This may be particularly important to limit the polymerization in
the early stage of the reaction, when the concentration of reducing
CuI/0 species is low. In the same time, the BP ketyl radical B
reduces CuII to CuI and, most probably CuI to Cu0 (Scheme 3).
Reduction of the radical C by the photogenerated low valent
copper species (CuI and/or Cu0) followed by protonation leads to
the desired Giese reaction product. We propose that the CuI/Cu0
manifold dominates the reduction of C for the following reasons:
a) in acetonitrile, Cu0 is a stronger reducing agent than CuI,[19] b)
the CuI to Cu0 photoreduction was shown to be an effective
process under the reaction conditions, c) the copper catalyst
resting state is not CuII, but most probably CuI and, d) copper NPs
accumulate rapidly once the acceptor has been consumed, as
expected if the redox active Cu0 species were not oxidized
anymore by the intermediate C.
Additional informations were provided by the follow up of the
1
reaction by H NMR (Figure 2). The spectra show that after 3 h
reaction the alkene conversion was ~ 70% and that ~ 50% of 22
was formed. Importantly, only a small amount of acetone was
formed, i.e. ~ 2.5% with respect to 22. This shows that the α-
hydroxyisopropyl radical is not the electron source necessary to
form 22, i.e. neither by direct reduction, nor through the
intermediacy of CuI or Cu0 species. On the other hand, ~ 90% of
the BP remained intact at that stage of the reaction, with
benzopinacol as by-product, strongly suggesting that the BP ketyl
radical is the main reducing agent in the process. This should hold
when the acrylate is still present in significant amount in solution
and can effectively trap the α-hydroxyisopropyl radical. Indeed,
the rate of addition of the latter to methyl acrylate (kobsd = 3.5 x 107
1
M− s−1)3c is ~ 4 orders of magnitude higher to that of the BP ketyl
1
1
radical (k = <9 x 103 M− s− ).[20] Accordingly, in the last 2 h of
reaction (Figure 2), when the acrylate concentration tends to zero,
an increase of the rate of formation of acetone (~ 10% with
respect to 22), benzopinacol (> 95%) and Cu0 NPs (Figure 1B)
was noticed, as anticipated if the α-hydroxyisopropyl radical then
competes with the BP ketyl radical as reducing species.
According to the present findings and to the results of previous
studies, the following general mechanism, depicted in Scheme 4,
is proposed. The reaction begins by direct abstraction of an
hydrogen atom of the donor from the BP triplet or,[21] in the case
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
70; Angew. Chem. Int. 1984, 96, 51-52; c) D. P. Curran, S. Hadida, S.-Y.
Kim, Z. Luo, J. Am. Chem. Soc. 1999, 121, 6607-6615; d) I. Ryu, S.
Uehara, H. Hirao, T. Fukuyama, Org. Lett. 2008, 10, 1005-1008.
a) Y. Miyake, K. Nakajima, Y. Nishibayashi, J. Am. Chem. Soc. 2012,
134, 3338-3341; b) P. Kohls, D. Jadhav, G. Pandey, O. Reiser, Org. Lett.
2012, 14, 672-675; c) L. Ruiz Espelt, E. M. Wiensch, T. P. Yoon, J. Org.
Chem. 2013, 78, 4107-4114; d) J. B. McManus, N. P. R. Onuska, D. A.
Nicewicz, J. Am. Chem. Soc. 2018, 140, 9056-9060; e) X. Dai, D. Cheng,
B. Guan, W. Mao, X. Xu, X. Li, J. Org. Chem. 2014, 79, 7212-7219; f) R.
A. Aycock, C. J. Pratt, N. T. Jui, ACS Catal. 2018, 8, 9115-9119.
a) K. A. Margrey, W. L. Czaplyski, D. A. Nicewicz, E. J. Alexanian, J. Am.
Chem. Soc. 2018, 140, 4213-4217; b) S. Rohe, A. O. Morris, T.
McCallum, L. Barriault, Angew. Chem. Int. Ed. Engl. 2018, 57, 15664-
15669; Angew. Chem. 2018, 130, 15890-15895; c) M. A. Ashley, H.
Yamauchi, J. C. K. Chu, S. Otsuka, H. Yorimitsu, T. Rovis, Angew. Chem.
Int. Ed. Engl. 2019, 58, 4002-4006; Angew. Chem. 2019, 131, 4042-4046
d) H.-B Yang, A. Feceu, D. B. C. Martin, ACS Catal. 2019, 9, 5708-5715;
e) J. L. Jeffrey, J. A. Terrett, D. W. C. MacMillan, Science 2015, 349,
1532-1536.
Conclusions
In summary, we have discovered that the polymerization
reactions that occur when performing BP photocatalyzed Giese-
type alkylations of C(sp3)−H bonds using unsubstituted acrylates,
acrylonitrile and methyl vinyl ketone as acceptors are strongly
inhibited in the presence of a catalytic amount of Cu(OAc)2.
Exploiting a dual photoredox/copper catalysis enables to expand
the scope of BP as photocatalyst, in particular in a process based
on the ability of its triplet state to effectively abstract hydrogen
atoms from C(sp3)−H bonds of a broad range of donors. The
methodology here reported is operationally simple and employs
readily available cheap reagents and a safe low intensity UVA
source present in most laboratories. It is complementary to
existing ones and of clear practical value. Mechanistic studies
revealed that reactions do not proceed through a radical chain,
but through a dual BP/Cu photocatalytic process in which both
CuII and low valent CuI/0 species generated in situ by reduction by
the BP ketyl radical may react with the α-keto or α-cyano
intermediate radicals, thus preventing polymerization.
[5]
[6]
[7]
a) X.-Z. Fan, J.-W. Rong, H.-L. Wu, Q. Zhou, H.-P. Deng, J. D. Tan, C.-
W. Xue, L.-Z. Wu, H.-R. Tao, J. Wu, Angew. Chem. Int. Ed. Engl. 2018,
57, 8514-8518; Angew. Chem. 2018, 130, 8650-8654; b) D. Dondi, A. M.
Cardarelli, M. Fagnoni, A. Albini, Tetrahedron 2006, 62, 5527-5535; c) S.
Dondi, S. Protti, A. Albini, S. Manas Carpio, M. Fagnoni, Green Chem.
2009, 11, 1653-1659; d) C. Manfrotto, M. Mella, M. Freccero, M. Fagnoni,
A. Albini, J. Org. Chem. 1999, 64, 5024-5028; e) D. Dondi, I. Caprioli, M.
Fagnoni, M. Mella, A. Albini, Tetrahedron 2003, 59, 947-957; f) R. Mosca,
M. Fagnoni, M. Mella, A. Albini, Tetrahedron 2001, 57, 10319-10328; g)
A. M. González-Cameno, M. Mella, M. Fagnoni, A. Albini, J. Org. Chem.
2000, 65, 297-303; h) D. Dondi, M. Fagnoni, A. Albini, Chem. Eur. J.
2006, 12, 4153-4163; i) D. Ravelli, M. Zoccolillo, M. Mella, M. Fagnoni,
Adv. Synth. Catal. 2014, 356, 2781-2786; j) T. Fukuyama, K. Yamada, T.
Nishikawa, D. Ravelli, M. Fagnoni, I. Ryu, Chem. Lett. 2018, 47, 207-209.
a) C. P. Seath, D. B. Vogt, Z. Xu, A. J. Boyington, N. T. Jui, J. Am. Chem.
Soc. 2018, 140, 15525-15534; b) A. ElMarrouni, C. B. Ritts, J. Balsells,
Chem. Sci. 2018, 9, 6639-6646; c) T. van Leeuwen, L. Buzzetti, L. A.
Perego, P. Melchiorre, Angew. Chem. Int. Ed. Engl. 2019, 58, 4953-
4957; Angew. Chem. 2019, 131, 5007-5011.
Acknowledgements
The CNRS, the University of Bordeaux and the Ministère de
l’Enseignement Supérieur de la Recherche et de l’Innovation
(Salary grant to BA) are gratefully acknowledged for their financial
support.
[8]
[9]
Keywords: C−H activation • photocatalysis • benzophenone •
copper • Giese reaction
[1]
Selected reviews on C(sp3)−H bond functionalization including
alkylations: a) L. Yang, H. Huang, Chem. Rev. 2015, 115, 3468-3517; b)
A. S. K. Hashmi, Acc. Chem. Res. 2014, 47, 864-876; c) T. W Lyons, M.
S. Sanford, Chem. Rev. 2010, 110, 1147-1169; d) J. F. Hartwig, M. A.
Larsen, ACS Cent. Sci. 2016, 2, 281-292; e) Z. Dong, Z. Ren, S. J.
Thompson, Y. Xu, G. Dong, Chem. Rev. 2017, 117, 9333-9403; f) Y. Qin,
L. Zhu, S. Luo, Chem. Rev. 2017, 117, 9433-9520; g) H. Yi, G. Zhang,
H. Wang, Z. Huang, J. Wang, A. K. Singh, A. Lei, Chem. Rev. 2017, 117,
9016-9085; h) J. He, M. Wasa, K. S. L. Chan, Q. Shao, J.-Q. Yu, Chem.
Rev. 2017, 117, 8754-8786; i) H. M. L. Davies, R. E. J. Beckwith, Chem.
Rev. 2003, 103, 2861-2903; j) T. G. Saint-Denis, R.-Y. Zhu, G. Chen, Q.-
F. Wu, J.-Q. Yu, Science 2018, 359, eaao4798; h) C.-G. Newton, S. G.
Wang, C. C. Oliveira, N. Cramer, Chem. Rev. 2017, 117, 8908-8976.
Selected reviews: a) J. Twilton, C. Le, P. Zhang, M. H. Shaw, R. W.
Evans, D. W. C. MacMillan, Nat. Rev. Chem. 2017, 1, 0052; b) J. Xie, H.
Jin, A. S. K. Hashmi, Chem. Soc. Rev. 2017, 46, 5193-5203; c) S. Protti,
M. Fagnoni, D. Ravelli, ChemCatChem 2015, 7, 1516-1523; d) M.
Fagnoni, D. Dondi, D. Ravelli, A. Albini, Chem. Rev. 2007, 107, 2725-
2756; e) K. Nakajima, Y. Miyake, Y. Nishibayashi, Acc. Chem. Res. 2016,
49, 1946-1956.
G. S. Lee, S. H. Hong, Chem. Sci. 2018, 9, 5810-5815.
[10] a) R. Beniazza, F. Molton, C. Duboc, C. Tron, N. D. McClenaghan, D.
Lastécouères, J.-M. Vincent, Chem. Commun. 2015, 51, 9571-9574; b)
R. Beniazza, M. Douarre, D. Lastécouères, J.-M. Vincent, Chem.
Commun. 2017, 53, 3547-3550; c) R. Beniazza, R. Atkinson, C. Absalon,
F. Castet, S. A. Denisov, N. D. McClenaghan, D. Lastécouères, J.-M.
Vincent, Adv. Synth. Catal. 2016, 358, 2949-2961; d) R. Beniazza, B.
Abadie, L. Remisse, D. Jardel, D. Lastécouères, J.-M. Vincent, Chem.
Commun. 2017, 53, 12708-12711.
[11] a) L. Harmand, S. Cadet, B. Kauffmann, L. Scarpantonio, P. Batat, G.
Jonusauskas, N. D. McClenaghan, D. Lastécouères, J.-M. Vincent,
Angew. Chem. Int. Ed. Engl. 2012, 51, 7137-7141; Angew. Chem. 2012,
124, 7249-7253; b) L. Harmand, R. Lambert, L. Scarpantonio, N. D.
McClenaghan, D. Lastécouères, J.-M. Vincent, Chem. Eur. J. 2013, 19,
16231-16239; c) R. Beniazza, R. Lambert, L. Harmand, F. Molton, C.
Duboc, S. Denisov, G. Jonusauskas, N. D. McClenaghan, D.
Lastécouères, J.-M. Vincent, Chem. Eur. J. 2014, 20, 13181-13187.
[12] a) Y. Amaoka, M. Nagatomo, M. Watanabe, K. Tao, S. Kamijo, M. Inoue,
Chem. Sci. 2014, 5, 4339-4345; b) S. Kamijo, T. Hoshikawa, M. Inoue,
Tetrahedron Lett. 2011, 52, 2885-2888; c) S. Kamijo, T. Hoshikawa, M.
Inoue, Org. Lett. 2011, 13, 5928-5931; d) T. Hoshikawa, S. Kamijo, M.
Inoue, Org. Biomol. Chem. 2013, 11, 164-169; e) T. Hoshikawa, S.
Yoshioka, S. Kamijo, M. Inoue, Synthesis 2013, 45, 874-887; f) T.
Hoshikawa, M. Inoue, Chem. Sci. 2013, 4, 3118-3123; g) S. Kamijo, M.
Hirota, K. Tao, M. Watanabe, T. Murafuji, Tetrahedron Lett. 2014, 55,
5551-5554; h) S. Kamijo, K. Tao, G. Takao, H. Murooka, T. Murafuji,
Tetrahedron Lett. 2015, 56, 1904-1907; i) S. Kamijo, G. Takao, K. Kamijo,
[2]
[3]
[4]
Selected reviews: a) B. Giese, Angew. Chem. Int. Ed. Engl. 1983, 22,
753-764; Angew. Chem. 1983, 95, 771-782; b) B. Giese Angew. Chem.
Int. Ed. Engl. 1985, 24, 553-565; Angew. Chem. 1985, 97, 555-567; c) H.
Fischer, L. Radom, Angew. Chem. Int. Ed. Engl. 2001, 40, 1340-1371;
Angew. Chem. 2001, 113, 1380-1414.
Selected examples: a) B. Giese, J. Dupuy, Angew. Chem. Int. Ed. Engl.
1983, 22, 622-623; Angew. Chem. 1983, 95, 633-634; b) B. Giese, J. A.
González-Gómez, T. Witzel, Angew. Chem. Int. Ed. Engl. 1984, 23, 69-
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
M. Hirota, K. Tao, T. Murafuji, Angew. Chem. Int. Ed. Engl. 2016, 55,
9695-9699; Angew. Chem. 2016, 128, 9847-9851; j) S. Kamijo, M.
Watanabe, K. Kamijo, K. Tao, T. Murafuji, Synthesis 2016, 48, 115-121;
k) S. Kamijo, G. Takao, K. Kamijo, T. Tsuno, K. Ishiguro, T. Murafuji, Org.
Lett. 2016, 18, 4912-4915; l) S. Kamijo, K. Kamijo, K. Maruoka, T.
Murafuji, Org. Lett. 2016, 18, 6516-6519; m) M. Kamijo, K. Kamijo, T.
Murafuji, J. Org. Chem. 2017, 82, 2664-2671.
[17] T. Lund, D. D. M. Wayner, M. Jonsson, A. G. Larsen, K. Daasbjerg, J.
Am. Chem. Soc. 2001, 123, 12590-12595.
[18] I. Toledo, M. Arancibia, C. Andrade, I. Crivelli, Polyhedron 1998, 17, 173-
178.
[19] J. A. Altermatt, S. E. Manahan, Anal. Chem. 1968, 40, 655-657.
[20] a) W. Schnabel In Lasers in Polymer Science and Technology:
Applications, Vol. 2 (Eds.: J.-P. Fouassier, J. F. Rabek), CRC Press:
Boca Raton, FL, 1991; b) J. Lalevée, X. Allonas, J.-P. Fouassier, J. Phys
Chem. A 2004, 108, 4326-7420.
[13] a) J. K. Kochi, A. Bemis, C. L. Jenkins, J. Am. Chem. Soc. 1968, 90,
4616-4625; b) C. L. Jenkins, J. K. Kochi, J. Am. Chem. Soc. 1972, 94,
843-855.
[21] Selected reviews and articles: a) J. C. Scaiano, J. Photochem. 1973, 2,
81-118; b) W. M. Nau, F. L. Cozens, J. C. Scaiano, J. Am. Chem. Soc.
1996, 118, 2275-2282; c) W. J. Leigh, E. C. Lathioor, M. J. St. Pierre, J.
Am. Chem. Soc. 1996, 118, 12339-2348; d) L. Giering, M. Berger, C.
Steel, J. Am. Chem. Soc. 1974, 96, 953-958; e) Y. M. A. Naguib, C. Steel,
S. G. Cohen, M. A. Young, J. Phys. Chem. 1987, 91, 3033-3036; f) S.
Abad, F. Bosca, L. R. Domingo, S. Gil, U. Pischel, M. A. Miranda, J. Am.
Chem. Soc. 2007, 129, 7407-7420.
[14] S. Inbar, H. Linschitz, S. G. Cohen, J. Am. Chem. Soc. 1981, 103, 1048-
1054.
[15] S. D. Houston, H. Xing, P. V. Bernhardt, T. J. Vanden Berg, J.
Tsanaktsidis, G. Paul Savage, C. M. Williams, Chem. Eur. J. 2019, 25,
2735-2739.
[16] a) J. C. Scaiano, K. G. Stamplecoskie, G. L. Hallett-Tapley, Chem.
Commun. 2012, 48, 4798-4808; b) M. L. Marin, K. L. McGilvray, J. C.
Scaiano, J. Am. Chem. Soc. 2008, 130, 16572-16584.
[22] Review: A. Hossain, A. Bhattacharyya, O. Reiser, Science 2019, 364,
eaav9713.
This article is protected by copyright. All rights reserved.

10.1002/chem.201904111
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 2:
FULL PAPER
Author(s), Corresponding Author(s)*
Page No. – Page No.
Title
Radical copper: The presence of 2 mol% of Cu(OAc)2 enables to strongly limit the
polymerization of highly polymerizable unsubstituted olefins engaged in the
benzophenone photocatalyzed radical Giese-type alkylations of C(sp3)−H bonds
(see scheme).
This article is protected by copyright. All rights reserved.