Phenyl Silicates with Substituted Catecholate Ligands: Synthesis, Structural Studies and Reactivity

Article abstract of DOI:10.1002/chem.202100453

While the generation of aryl radicals by photoredox catalysis under reductive conditions is well documented, it has remained challenging under an oxidative pathway. Because of the easy photo-oxidation of alkyl bis-catecholato silicates, a general study of phenyl silicates bearing substituted catecholate ligands has been achieved. The newly synthesized phenyl silicates have been fully characterized, and their reactivity has been explored. It was found that, thanks to the substitution of the catecholate moiety, notably with the 4-cyanocatecholato ligand, the phenyl radical could be generated and trapped. Computational studies provided a rationale for these findings.

Full text of DOI:10.1002/chem.202100453

Accepted Article
Accepted Article
Title: Phenyl Silicates with Substituted Catecholate Ligands: Synthesis,
Structural Studies and Reactivity
Authors: Louis Gabriel Fensterbank, Etienne Levernier, Khaoula
Jaouadi, Heng-Rui Zhang, Vincent Corcé, Aurélie Bernard,
Geoffrey Gontard, Claire Troufflard, Laurence Grimaud,
Etienne Derat, and Cyril Ollivier
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.202100453
Link to VoR: https://doi.org/10.1002/chem.202100453
01/2020

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
Phenyl Silicates with Substituted Catecholate Ligands: Synthesis,
Structural Studies and Reactivity
Etienne Leverniera, Khaoula Jaouadia,b, Heng-Rui Zhanga, Vincent Corcéa, Aurélie Bernarda, Geoffrey
Gontarda, Claire Troufflard,a Laurence Grimaudb, Etienne Derata*, Cyril Olliviera*, Louis Fensterbanka*
Dedicated to Professor Ilhyong Ryu in celebration of his 70th birthday.
Abstract: While the generation of aryl radicals by photoredox
catalysis is well-documented under reductive conditions, it has
remained challenging under an oxidative pathway. Because of the
easy photooxidation of alkyl bis-catecholato silicates, a general
study on phenyl silicates bearing substituted catecholate ligands has
been achieved. The newly synthesized phenyl silicates have been
fully characterized and their reactivity has been explored. It was
found that thanks to the substitution of the catecholate moiety and
notably with the 4-cyanocatecholato ligand, the phenyl radical could
be generated and trapped. Computational studies provided a
rationale for these findings.
photooxidation of alkylsilicates bearing the hexafluorocumyl
alcohol dianion ligand (Martin’s ligand), a single example of
oxidation of the corresponding phenylsillicate gave 10% of
Giese-type adduct from the phenyl radical. This underlines that
the generation of aryl radicals under photooxidative conditions is
highly challenging.
Introduction
Alkyl bis-catecholatosilicates have recently elicited intense
interest in photooxidative catalysis as valuable alkyl radical
precursors in radical addition reactions,[1] radical-polar crossover
2
]
reactions[
as well as dual photoredox-nickel cross-
couplings.[1a,3] In contrast, aryl silicates counterparts have been
much less utilized. Their main use has been disclosed by
DeShong in palladium catalyzed cross-coupling reactions with
arylhalides[ 4 ] (Scheme 1, eq. 1) and by Hashmi for cross-
coupling reactions with aryldiazonium salts by gold(I) catalysis[5]
(Scheme 1, eq. 2).
Scheme 1. Use of aryl silicates for sp2-sp2 cross-coupling reactions
Second, a new class of phenyl silicate derivatives could
open new perspectives in terms of reactivity either in radical
addition reactions or in dual catalysis. This study aimed at
determining the influence of the substitution of the catechol
ligands in their ability to promote the generation of aryl radicals.
We therefore undertook the synthesis of various phenyl silicates
and studied their structural features and reactivity.
The photooxidation of arylsilicates would be of interest for
two reasons. First, the generation of aryl radicals by photoredox
catalysis is well-documented under photoreductive conditions
from a variety of precursors such as aryl diazoniums,[ 6 ]
iodoniums,[7,8] sulfoniums,[7,9] arylhalides,[10] and also benzoyl
hypohalites[11] (Scheme 2). In contrast and to the best of our
knowledge, only one full study was published recently by the
Yoshimi group under photooxidative conditions with aryl
carboxylates.[ 12 ] Up to 150 mol % of biphenyl (BP)/1,4-
dicyanonaphthalene (DCN) as photocatalytic mixture had to be
used under UV irradiation to provide moderate yields (~50%) of
aryl radical adducts. Of note also, at the occasion of a very
13
]
recent study by Morofuji and Kano[
dedicated to the
[a]
E. Levernier, K. Jaouadi, H. R. Zhang, Dr. V. Corcé, A. Bernard, G.
Gontard, C. Troufflard, Dr. E. Derat, Dr. C. Ollivier, Prof. L.
Fensterbank
Sorbonne Université, CNRS, Institut Parisien de Chimie Moléculaire
4 Place Jussieu, CC 229, F-75252 Paris Cedex 05 (France)
E-mail: etienne.derat@sorbonne-universite.fr,
cyril.ollivier@sorbonne-universite.fr, louis.fensterbank@sorbonne-
universite.fr
[b]
K. Jaouadi, Dr. L. Grimaud
Laboratoire de biomolécules (LBM), Département de chimie,
Sorbonne Université, École normale supérieure, PSL University,
CNRS, 75005 Paris (France)
Supporting information for this article is given via a link at the end of
the document.
Scheme 2. Generation of aryl radicals through photocatalytic conditions
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
Results and Discussion
We first focused on the reactivity of the simplest term, the
phenyl bis-catecholatosilicate 1a. It was synthesized based on a
previously reported protocol[1a,14] using catechol (2 equiv), 18-C-6
(1 equiv), MeOK (1 equiv) and phenyltrimethoxysilane (1 equiv).
Silicate 1a was obtained in a satisfying 87% yield (reaction time:
2 h at room temperature, solvent of crystallization: acetone/Et2O).
Its half-wave oxidation potential in DMF was measured by cyclic
voltammetry and the observed value (E1/2 = + 0.89 V vs.
ox
SCE[1a]) was compared with the reduction potential of the
photocatalyst [Ir(dF(CF3)ppy)2(bpy)]PF6 (3) in its excited state
(Ered(Ir(III)*/Ir(II) = + 1.32 V vs SCE[15]). These data suggested
that silicate 1a could be oxidized by this photocatalyst under
irradiation. Nevertheless, all attempts to generate a phenyl
radical with photocatalyst 3 and to trap it with allylsulfone 2 met
limited success. Allylation product 4 was observed in only 5 %
yield. This finding appeared puzzling since efficient
phosphorescence quenching of 3 with 1a was observed. Indeed,
the quenching rate (kq) constant was determined by Stern-
Volmer analysis and found to be kq = 5.7 x 108 mol-1 L s-1 (in
comparison to kq = 7.9 x 109 mol-1 L s-1 for benzyl silicate[1a]). In
order to check whether this low yield was due to the instability of
1a, 29Si NMR experiments were carried out before and after
reaction with acceptor 2. After 24 hours of blue LED irradiation
(λmax = 450 nm) in DMF-d7, only 1a as silicon derivative could be
detected in the mixture and no obvious degradation was also
evidenced by 1H NMR (Figure 1).
Figure 1. 29Si NMR monitoring of the photocatalytic reaction of 1a with 2a in
DMF-d7.
We first tested the putative substitution effect exerted by the
catechols
on
the
highly
reactive
cyclohexyl
bis-
catecholatosilicate substrate. Notably, we wished to check that
the formation of the cyclohexyl radical was still possible. Based
on the previously described procedure using cyclohexyl-
trimethoxysilane, MeOK, 18-C-6 and catechols with diverse
substitution patterns (electron donating or withdrawing groups),
cyclohexylsilicates 6a-6d were obtained after crystallization in
acetone/Et2O (Scheme 4).[1a,16]
Additionally, when an equimolar mixture of 1a and 3 was
irradiated under blue LED in deuterated THF as solvent, the
1
1
formation of C6H5D 5 could be observed by H NMR ( H 7.29
2
2
ppm in THF-d8) and quantified by H NMR ( H 7.35 ppm in
THF-d8) to 21% NMR yield (see Scheme 3 and the supporting
information for more details).
Scheme 3. Photooxidation of phenyl silicate 1a and phenyl radical trapping
Scheme 4. Synthesis of cyclohexyl silicates 6 bearing differently substituted
catechol moieties
All together, these findings suggest an inefficient oxidation
process of 1a. Considering the favorable redox potentials and
the efficient luminescence quenching (see above), this low
reactivity of 1a aroused our curiosity and led us to study the role
of substituents on the catechol moiety and their potential effect
on the reactivity of the corresponding phenyl silicates.
Upon treatment with photocatalyst 3 and allylsulfone 2 under
blue LED irradiation, silicates 6b (Eox1/2 = + 0.91 V vs SCE[17]
)
and 6d (Eox1/2 = + 1.01 V), bearing electron withdrawing groups,
afforded the corresponding radical allylation product 7, but in
slightly lower yields (77% and 60% yields respectively) than
plain silicate 6a (Eox1/2 = + 0.69 V, 88% of 7) (Table 1).
Interestingly, an electron donating group such as a methoxy on
the catechols (6c) drastically decreased the yield of this reaction
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
and most of the starting material was recovered. In view of the
lower oxidation potential of 6c (Eox1/2 = + 0.63 V) this result
appears contradictory but was corroborated by DFT calculations
(vide infra).
Table 1. Generation of cyclohexyl radical from cyclohexyl silicates 6 probed by
allylation reaction
Cyclohexyl silicate 6 Yield of 7 (in %)a
6a
6b
6c
6d
88
77
14
60
a
1H NMR yield of 7 using 1,3,5 trimethoxybenzene as NMR standard
Encouraged by these results showing that the catechol
substitution can modulate the reactivity of the corresponding
silicates but does not prevent the photooxidation process, we
prepared
a library of phenyl silicates with mono- and
polysubstituted catechols.
Using the same procedure as above with the appropriate
catechol, most of the silicates 1 were efficiently prepared (yield >
80%) in crystalline form suitable for X-ray diffraction analysis
(Scheme 5). Thus, silicates 1b, 1d and 1h bearing electron
donating (MeO) and electron withdrawing (F, CN) groups
respectively were analyzed by XRD (Figure 2).[18]
All three crystal structures exhibit silicates in which the
silicon center adopts a square pyramidal geometry, quite similar
to that of 1a. Indeed, the electronic features of the catechol do
not seem to have much impact on the environment of the silicon
atom. 1b and 1h show a marked interaction between silicate and
potassium with K-O distances ranging from 2.7 to 3.0 Å, similar
to previously published structures of potassium hypervalent
silicates.[1a] Only 1d features weaker interactions. While the
asymmetric units of 1b and 1h contain only one discrete silicate,
silicate 1d shows a statistical disorder of the fluorine atoms that
suggests the presence of cis and trans isomers but in
undetermined proportions (see SI for more details).
Scheme 5. Preparation of phenyl silicates 1 bearing substituted catechols
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
T = 393 K (DMSO-d6)
T = 373 K (DMSO-d6)
T = 353 K (DMSO-d6)
T = 333 K (DMSO-d6)
T = 303 K (DMSO-d6)
T = 273 K (Acetone-d6)
T = 253 K (Acetone-d6)
Figure 3. 13C NMR of silicate 1h at various temperatures
A coalescence of the peaks was observed above 333 K
probably due to Berry pseudo-rotation[ 20 ] suggesting that an
equilibrium exists between the two isomers. DFT calculations
also indicated that the energetic gap between the cis and trans
form for 1b is 3.3 kcal/mol in favor of the cis form. For 1h, the
trans form is slightly favored by 0.6 kcal/mol.
After studying the structural features of these new silicon
derivatives, we examined other key properties. First, UV-vis
absorption spectra were recorded. All silicates 1 exhibited a
unique absorption band spanning from 276 nm for 1d to 303 nm
for 1g. Importantly, whatever the substitution pattern, no
noticeable absorption was observed in the wavelength range of
blue LED, previously used for the photooxidation of alkyl
silicates (from 400 to 520 nm) (see Figure 4 and the supporting
information for more details and UV data).
Figure 2. XRD analyses of silicates 1b, 1d and 1h. Only cis isomer of 1d
depicted and hydrogens omitted for clarity.
The presence of two isomers was also confirmed by 13C
NMR. All resonances of the non-symmetrical catechol moieties
in 13C NMR were doubled with identical integration for both
species in the case of 1d, but also for the other non-symmetrical
phenylsilicates 1b, 1c, 1h. It has to be mentioned that this type
of 1:1 mixture of two isomers was also observed in the
cyclohexyl series (silicates 6b-d).
Intrigued by this finding, we wondered if an equilibrium exists
between both cis and trans isomers in solution. To answer this
question, 13C NMR of 1h at various temperatures was conducted
(in acetone-d6 at low temperature (below 273 K) and DMSO-d6
at higher temperature (from 303 K), Figure 3).[19]
1d
1h
1g
1.00
0.75
0.50
0.25
0.00
Cl
CN
Cl
Cl
K, 18 C 6
Cl
F
O
O
O
O
O
O
Si
Si
Si
O
K, 18 C 6
O
O
K, 18 C 6
O
O
O
Cl
Cl
1d
F
Cl
1g
1h
CN
Cl
300
400
500
)
600
Wavelength(nm
Figure 4. Absorption spectra of silicates 1b, 1d and 1h
Second, the oxidation potentials of silicates 1a-i were
measured in DMF as displayed in Table 2.
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
Table 2. Half-wave oxidation potentials of silicates 1a-i
Silicate
1a
E ox1/2 (vs SCE in DMF)
+ 0.89 V
+ 0.81 V
+ 1.09 V
+ 1.05 V
+ 1.13 V
+ 1.46 V
+ 1.37 V
+ 1.33 V
+1.62 V
1b
Silicate 1
Yield of 4 (in %)1
1c
1d
1a
1b
1c
1d
1e
1f
5%
9%
1e
1f
8%
1g
1h
6%
1i
6%
15%
23%
35%
7%
1g
1h
1i
Interestingly, catechol modifications resulted in important
1/2
variations of the oxidation potentials, from Eox = + 0.81 V for
1/2
1/2
1b to Eox = + 1.62 V for 1i with Eox = + 0.89 V for the
unsubstituted phenylsilicate 1a. Thus, a donating group on the
catechols as for 1b was found to logically decrease the oxidation
potential while electron withdrawing groups significantly
1
1H NMR yield of 4 using 1,3,5 trimethoxybenzene as NMR standard
Encouraged by these results, we optimized the reaction of
the most promising silicate 1h with acceptor 2 (Table 4). The
influence of different counterions (ammoniums and potassium
with and without 18-C-6, entries 6, 11 and 12) was studied but
no significant improvement was observed. The temperature
(room temperature or 100°C, entries 5 and 6) did not affect the
reaction efficiency. Switching to some organic photocatalysts
resulted in a decreased yield (entries 1, 2 and 3).[22,23] Even the
very oxidizing Fukuzumi's acridinium (Ered(PC*/PC•-) = + 2.06 V
vs SCE[24]) did not afford 4 in a better yield. These results are
consistent with what was previously observed with
alkylsilicates.[25] Different solvents (DMF, DMSO, MeCN, EtOH),
reaction times and concentrations were also screened but,
despite all our efforts, the yield remained modest. Notably,
prolonged reaction times did not result in significant
improvements (entry 6 vs entries 9 and 10). Hence, the best
conditions were found to be in DMSO with 3 at room
temperature for 24 h (40%, entry 4). Some control experiments
were also carried out. The reaction in the dark or photocatalyst
free conditions did not afford any product (entries 7 and 8).
Therefore, the presence of a photocatalyst under irradiation is
mandatory to generate the phenyl radical.
1/2
increased the values above + 1.0 V (Eox = + 1.05 V for 1d).[21]
Higher oxidation potentials could even be reached by using the
per-bromocatechol or by adding two cyano groups (Eox1/2 (1f) = +
1.46 V and Eox1/2 (1f) = + 1.62 V). Even if the oxidation of 1f and
1i by the photoactivated
3
[Ir(dF(CF3)ppy)2-(bpy)]PF6
(Ered(Ir(III)*/Ir(II) = + 1.32 V) appeared difficult, all the other
silicates 1 could potentially lead to the generation of the phenyl
radical. Thus, we proceeded this study with the photocatalytic
allylation reaction from silicates 1a-i with allylsulfone 2 in the
presence of 2 mol% of 3. Results are summarized in Table 3.
Although a yield < 10% was obtained from silicates 1a-e and
1i, silicate 1h bearing a cyano group on the catechol afforded 4
in 35% yield. Per-brominated and per-chlorinated catechols 1f
and 1g also provided the desired product 4 in slightly better
yields than silicate 1a (15% and 23% respectively). Nevertheless,
these yields are still lower than the one obtained with silicate 1h.
The higher oxidation potentials of 1f and 1g in comparison to 1h
might explain this result. Also, as previously observed in Table 1
with cyclohexyl silicates, the methoxy substitution was found to
be detrimental resulting in a decreased yield. Thus, the
modification of the catechol moiety proved to modulate the
reactivity of the phenyl silicates 1 in the photocatalytic allylation
type reaction.
Table 4. Optimization of the allylation reaction
Table 3. Photocatalytic allylation reaction of phenylsilicates1
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
Yield
of 41
1,2
1,15
1,1
Stern-Volmer plot of 1h
Entry Solvent
T
Photocatalyst
t
Y+
Fukuzumi's
acridinium
Pyrylium salt2
K+/18-
C-6
1
2
DMF
DMF
DMF
DMSO
DMF
DMF
DMF
DMF
DMF
DMF
rt
24 h
24 h
24 h
24 h
24 h
24 h
24 h
24 h
68 h
6%
7%
K+/18-
C-6
rt
K+/
y = 0.0003x + 1
R² = 0.97
3
rt
4CzIPN
7%
18-C-6
1,05
1
K+/
4
rt
3
3
3
-
40%
38%
35%
0%
18-C-6
K+/
5
100°C
18-C-6
K+/
6
rt
rt
rt
rt
rt
0,95
18-C-6
K+/
0
100
200
300
400
500
600
7
Concentration of 1h (μM)
18-C-6
K+/
8
3
3
3
0%3
35%
41%4
18-C-6
K+/
Figure 5. Stern-Volmer plot - quench of silicate 1h and 3
9
18-C-6
K+/
10
68 h
24 h
18-C-6
K+
To better understand the difference of reactivity between
alkylsilicates and phenyl silicates, we resorted to computational
studies. The fragmentation of the intermediate hypercoordinated
radical resulting from the oxidation by the photocatalyst was
assessed for both types of silicates. Energy barriers for the
cleavage of the silicon-carbon bond and the free energy of the
reaction are summarized in Table 5, in the presence or absence
of DMF as coordinating ligand on the silicon center. In all cases,
it was found that the presence of DMF reduces the barrier of
radical expulsion. With DMF, the generation of a primary alkyl
radical is rather easy with an energy barrier of 17.45 kcal/mol,
which explains why various transformations have been observed
with this type of substrates.[1,3] For cyclohexyl silicates, we
observe a consistent and slightly lower barrier of 13.23 kcal/mol
to expel the cyclohexyl radical from I•Cy. Adding cyano groups
helps to reduce this barrier (I•Cy•CNII•Cy•CN) to 10.51
kcal/mol. Interestingly, methoxy groups on I•Cy•OMe resulted in
a much higher calculated barrier (37.73 kcal/mol) that is
corroborated by the poor reactivity of 6c in an allylation reaction
(see Table 1). Calculations also featured a distinct scenario for
this silicate. While in the other cases, upon departure of the
radical, the DMF strongly interacts with the silicon center of the
generated bis-catecholato spirosilane to give a square pyramid
hypercoordinated silicon species, II•Cy•OMe fragments
differently and the resulting methoxy substituted bis-catecholato
spirosilane adopts a tetrahedral shape with no coordination of
the DMF. The electron donation of the methoxy groups
presumably disfavors the formation of the hypercoordinated
species.
11
12
13
14
DMF
DMF
CH3CN
EtOH
rt
rt
rt
rt
3
3
3
3
30%
35%
0%
24 h Et3NH+
24 h
24 h
TBA+
TBA+
20%
15
DMSO
rt
3
24 h
TBA+
K+/
35%
16
DMF
rt
35
24 h
40%
18-C-6
1 [Silicate] = 0.1 mol.L-1, 3 (2 mol%), 1H NMR yield of 4 using 1,3,5 trimethoxybenzene
as NMR standard; 2 2,4,6-tri(p-tolyl)pyrylium tetrafluoroborate salt; 3 reaction in the
dark; 4 [Silicate] = 0.2 mol.L-1; 5 3 (10 mol%).
Inspired by the pioneering work of Nishigaichi showing that
silicates bearing catechol or 2,3-hydroxynaphthalene ligands
can be photoactivated by direct irradiation,[26] the photo allylation
reaction of silicate 1h was also tested at 300 nm with and
without 3 but limited success was met since only 10% and 6% of
product 4 were isolated respectively. So, even if this silicate
absorbs UV-B light, the use of 3 as photocatalyst under blue
LED irradiation remains more efficient. This was further
corroborated by the fact that the phosphorescence quenching of
3 with silicate 1h was also observed and thanks to a Stern-
Volmer plot, a quenching constant of 1.31 x 108 mol-1 L s-1 was
obtained (compared to 5.7 x 108 mol-1 L s-1 for silicate 1a). Thus,
silicate 1h does not quench the iridium photocatalyst more
efficiently than silicate 1a (Figure 5).
D-abstraction from deuterated THF-d8 was also probed and
2
led to the formation of C6H5D in 51% yield (as quantified by H
NMR) from 1h, in comparison to 21% yield obtained with
unsubstituted silicate 1a (see above) confirming the effect of the
catecholate substitution.[27]
Table 5. Barrier and Gibbs free energy of the reaction (in kcal/mol)
corresponding to radical formation for various oxidized silicates, as calculated
at the wB97M-D3BJ/def2-SV(P) level. Values in parentheses correspond to
calculations done without DMF.
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
Importantly, the calculations also supported that
phenylsilicates 1 are less prone to expel a radical upon oxidation.
Without any substituent on the catechol moiety, the energy
barrier was calculated to be 27.46 kcal/mol, which is 10 kcal/mol
more than for the extrusion of a primary radical and 14 kcal/mol
more than for the cyclohexyl one, and the generated
intermediate radical is rather unstable (with a Gibbs free energy
of 22 kcal/mol).[28] When the C-Si bond starts to elongate to
reach II•Ph, the silicon species maintains a square-planar
pyramid geometry around the silicon center through an
interaction with one molecule of solvent (DMF) and with the
phenyl radical being -stacked on top of one of the catechol
moieties, which brings stablization. Thus, for II•Ph and analogs
the expelled phenyl radical interacts through - interactions
with one of the catechol moieties and interaction distances
between centroids are respectively 3.21 Å, 3.23 Å and 3.12 Å for
II•Ph, II•Ph•CN and II•Ph•OMe (Figures 6 and 7). This marks a
sharp contrast with the expelled alkyl radicals that do not interact
at all with the catechol ligands and are less bound to the silicon
entity. Finally, the situation proved to be different with the
experimentally more reactive silicate 1h since the energy barrier
to generate the radical was about 19 kcal/mol (I•Ph•CN), thus far
more accessible. Nevertheless, the radical is still rather unstable
Free
Barrier
(kcal/mol)
energy of
reaction
(kcal/mol)
Silicate
1.86
(13.78)
17.45
(28.88)
7.77
2.03
13.23
10.51
37.73
I•Cy
with
a
highly endergonic process of 19 kcal/mol
(I•Ph•CNII•Ph•CN). Finally, the high barrier for I•Ph•OMe (>40
kcal/mol) is consistent with what was found for I•Cy•OMe. These
results prompted us to investigate the electronic structure of
these radicals more in detail.
I•Cy•CN
I•Cy•OMe
I•Ph•CN
I•Ph
17.45
19.05
(25.57)
18.99
(34.89)
22.61
(25.98)
27.46
(36.49)
29.16
32.53
(28.03)
(43.00)
I•Ph•OMe
Figure 6. Spin density and frontier orbitals diagram for the oxidized aryl
silicate II•Ph
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
To our surprise, these radicals are not classical and do not
follow the Aufbau principle. Indeed, for most of the radicals, the
orbital encompassing the unpaired electron corresponds to the
highest occupied molecular orbital (HOMO). Here, the orbital
better suited to describe the calculated spin density is not the
HOMO but a lower lying molecular orbital. This phenomenon is
known as SOMO-HOMO inversion and has been previously
described for various systems. Their common feature is to
combine a stable radical moiety with a donor unit, presenting
numerous -conjugated orbitals.[29] Our system constitutes an
original example since it is originating from a hypercoordinated
species.
The frontier orbitals, as well as the spin density, are depicted
in Figures 6 and 7 for radicals II•Ph and II•Ph•CN originating
from 1a and 1h. It can be observed that, in the case of the
radical species II•Ph•CN, the SOMO is lower in energy (-9.10
eV) than for the non-substituted catechol radical species (-8.63
eV). Overall, these findings could rationalize why the CN
substituted silicate 1h behaves differently from the unsubstituted
species 1a. First, the barrier for radical extrusion is weaker
(27.46 kcal/mol for I•Ph vs 19 kcal/mol for I•Ph•CN) allowing an
easier generation of the phenyl radical. In the case of II•Ph•CN,
the extra stability of the radical (-9.10 eV vs. -8.63 eV) would
result in an optimized radical reactivity with presumably less side
reactions.
In this study, we have synthesized and fully characterized by
NMR, UV-vis absorption, X-ray diffraction and cyclic voltammetry
a series of phenyl bis-catecholatosilicate derivatives 1 featuring
electronic modulation on the catechol ligand. While they share a
lot of common features with their alkylsilicate congeners, a
notable difference with these compounds resides in their higher
oxidation potentials that generally lay above 1.0 V vs SCE,
except for the unsubstituted phenylsilicate 1a or the methoxy
substituted one 1b, and bode for a less favorable oxidation. The
study of their photooxidation gave average results from a
synthetic point of view but revealed for the first time an important
effect of the catechol substitution. Counter-intuitively, phenyl
silicates with the most donor substituents on catechol, and
therefore with the lowest oxidation potentials, did not give the
best results. It is the 4-cyanocatecholato ligand (silicate 1h)
which provided the best results and led to about 40% of
allylation of the phenyl radical. Although modest, these yields
should be understood in a context where the generation of aryl
radicals by oxidative photoredox catalysis is hardly described.
DFT calculations have allowed to rationalize the different
observations and have highlighted an intriguing SOMO-HOMO
inversion that has never been described for hypervalent species.
These results also emphasize that silicates should be
considered as first-rate precursors to challenging radicals.
Experimental Section
General procedure for conjugate addition:
To a dried Schlenk flask were added the appropriate silicate (1.0
equiv), 3 (2 mol %) and the allyl sulfone (4 equiv). The Schlenk
flask was sealed with a rubber septum and evacuated / purged
with vacuum / argon three times. Then degassed solvent (0.1 M)
was introduced and the reaction mixture was irradiated with blue
LED (477 nm) at room temperature for 24 h under an argon
atmosphere. The reaction mixture was diluted with diethyl ether,
washed with water (2 times), dried over MgSO4 and evaporated
1
under reduced pressure. The crude residue was analyzed by H
NMR (with 1equiv of 1,3,5-trimethoxybenzene as NMR
standard).
Acknowledgements
We thank Sorbonne Université, PSL University, CNRS, ANR-17-
CE07-0018 HyperSilight (PhD grant to EL and KJ), the Chinese
Scholarship Council (PhD grant to HZ) for financial support and
Alexandre Millanvois for the graphical abstract.
Keywords: aryl radical, oxidation, photoredox catalysis, silicate
Figure 7. Spin density and frontier orbitals diagram for the oxidized
phenylsilicate II•Ph•CN
[1]
a) V. Corcé, L.-M. Chamoreau, E. Derat, J.-P. Goddard, C. Ollivier, L.
Fensterbank, Angew. Chem. Int. Ed. 2015, 54, 11414-11418; Angew.
Chem. 2015, 127, 11576-11580; b) A. Cartier, E. Levernier, V. Corcé, T.
Fukuyama, A.-L. Dhimane, C. Ollivier, I. Ryu, L. Fensterbank, Angew.
Chem. Int. Ed. 2019, 58, 1789-1793; Angew. Chem. Int. Ed. 2019, 131,
1803-1807; c) A. Cartier, E. Levernier, A.-L. Dhimane, T. Fukuyama, C.
Conclusions
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
Ed. 51, 12303–12306; c) E. H. Discekici, N. J. Treat, S. O. Poelma, K.
M. Mattson, Z. M. Hudson, Y. D. Luo, C. J. Hawker, J. R. de Alaniz,
Chem. Commun. 2015, 51, 11705−11708, d) I. Ghosh, B. Kꢀnig,
Angew. Chem. Int. Ed. 2016, 55, 7676−7679; Angew. Chem. 2016, 128,
7806−7810; e) M. Majek, U. Faltermeier, B. Dick, R. Perez-Ruiz, Jacobi
von Wangelin, A. Chem. Eur. Joc. 2015, 21, 15496−15501.
Ollivier, I. Ryu, L. Fensterbank, Adv. Synth. Catal. 2020, 362, 2254-
2259; d) N. R. Patel, C. B. Kelly, A. P. Siegenfeld, G. A. Molander, ACS
Catal. 2017, 7, 1766-1770; e) S. T. J. Cullen, G. F. Friestad, Org. Lett.
2019, 21, 8290-8294.
[2]
a) S. B. Lang, R. J. Wiles, C. B. Kelly, G. A. Molander, Angew.Chem.Int.
Ed. 2017, 56, 15073–15077; Angew. Chem. 2017,129,15269–15273;
b) J. P. Phelan, S. B. Lang, J. S. Compton, C. B. Kelly, R. Dykstra, O.
Gutierrez, G. A. Molander, J. Am. Chem. Soc. 2018, 140, 8037-8047; c)
T. Guo, L. Zhang, X. Liu, Y. Fang, X. Jin, Y. Yang, Y. Li, B. Chen, M.
Ouyang, Adv. Synth. Catal. 2018, 360, 4459-4463; d) L. R. E. Pantaine,
J. A. Milligan, J. K. Matsui, C. B. Kelly, G. A. Molander, Org. Lett. 2019,
21, 2317–2321
[11] L. Candish, M. Freitag, T. Gensch, F. Glorius, Chem. Sci. 2017, 8,
3618-3622.
[12] S. Kubosaki, H. Takeuchi, Y. Iwata, Y. Tanaka, K. Osaka, M. Yamawaki,
T. Morita, Y. Yoshimi, J. Org. Chem. 2020, 85, 5362-5369.
[13] T. Morofuji, Y. Matsui, M. Ohno, G. Irakashi, N. Kano, Chem. Eur. J.
2021, 10.1002/chem.202005300.
[14] a) C. L. Frye, J. Am. Chem. Soc. 1964, 86, 3170-3171; b) G. Cerveau,
C. Chuit, R. J. P. Corriu, L. Gerbier, C. Reye, J. L. Aubagnac, B. El
Amrani, Int. J. Mass. Spectron Ion Phys. 1988, 82, 259-271; c) V.
Corcé, C. Lévêque, C. Ollivier, L. Fensterbank, Science of Synthesis:
Photocatalysis in Organic Synthesis, 2019, 427.
[3]
a) C. Lévêque, L. Chennenberg, V. Corcé, J.-P. Goddard, C. Ollivier, L.
Fensterbank, Org. Chem. Front. 2016, 3, 462-465; b) E. Levernier, V.
Corcé, L.-M. Rakotoarison, A. Smith, M. Zhang, S. Ognier, M.
Tatoulian, C. Ollivier, L. Fensterbank, Org. Chem. Front. 2019, 6, 1378-
1382; c) M. Jouffroy, D. N. Primer, G. A. Molander, J. Am. Chem. Soc.
2016, 138, 475-478; d) N. R. Patel, C. B. Kelly, M. Jouffroy and G. A.
Molander, Org. Lett. 2016, 18, 764-767; e) M. Jouffroy, C. B. Kelly, G.
A. Molander, Org. Lett. 2016, 18, 876-879; f) J. P. Phelan, S. B. Lang, J.
Sim, S. Berritt, A. K. Peat, K. Billings, L. Fan, G. A. Molander, J. Am.
Chem. Soc. 2019, 141, 3723-3732; g) J. K. Matsui, S. B. Lang, D. R.
Heitz, G. A . Molander, ACS Catal. 2017, 7, 2563-2575; h) K. Lin, R. J.
Wiles, C. B. Kelly, G. H. M. Davies, G. A. Molander, ACS. Catal. 2017,
7, 5129-5133; i) B. A. Vara, X. Li, S. Berritt. C. R. Walters, E. J.
Petersson, G. A. Molander, Chem. Sci. 2018, 9, 336-344; j) S. O. Badir,
J. Sim, K. Billings, A. Csakai, X. Zhang, W. Dong, G. A. Molander, Org.
Lett. 2020, 22, 1046-1051; k) K. D. Raynor, G. D. May, U. K.
Bandarage, M. J. Boyd, J. Org. Chem. 2018, 83, 1551-1557.
a) W. M. Seganish, P. DeShong, J. Org. Chem. 2004, 69, 1137-1143;
b) W. M. Seganish, P. DeShong, Org. Lett. 2004, 5, 4379-4381.
S. Witzel, K. Sekine, M. Rudolf, S. K. Hashmi, Chem. Commun. 2018,
54, 13802-13804.
[15] D. Hanss, J. C. Freys, G. Bernardinelli, O. S. Wenger, Eur. J. Inorg.
Chem. 2009, 2009, 4850-4859.
[16] a) C. L. Frye, J. Am. Chem. Soc. 1964, 86, 3170-3171; b) K. Lin, C. B.
Kelly, M. Jouffroy, G. A. Molander, Org. Synth. 2017, 94, 16-33.
[17} All given potentials are referenced to SCE.
[18] Deposition Numbers 2009048 (for 1b), 2009049 (for 1h), and 2009050
(for 1d) contain the supplementary crystallographic data for this paper.
These data are provided free of charge by the joint Cambridge
Crystallographic Data Centre and Fachinformationszentrum Karlsruhe
Access Structures service www.ccdc.cam.ac.uk/structures.
[19] R. J. Abraham, D. S. Ribeiro, J. Chem. Soc.
2001, 2, 302-307.
Perkin.
Trans.
[20] a) A. R. Bassindale, M. Sohail, P. G. Taylor, A. A. Korlyukov, D. E.
Arkhipov, Chem. Commun. 2010, 46, 3274-3276, b) D. Kost, I.
Kalikhman, Acc. Chem. Res. 2009, 42, 303-314.
[4]
[5]
[6]
[21] A. G. Larsen, A. H. Holm, M. Roberson, K. Daasbjerg, J. Am. Chem.
Soc. 2001, 123, 1723-1729.
For a general revue on radical precursor in reduction see: I. Ghosh, L.
Marzo, A. Das, R. Shaikh, B. König, Acc. Chem. Res. 2016, 49, 1566-
1577; For a revue on diazoniums see: F. Mo, G. Dong, Y. Zhang, J.
Wang, Org. Biomol. Chem. 2013, 11, 1582-1593; a) T. Hering, D. P.
Hari, B. König, J. Org. Chem. 2012, 77, 10347-10352; b) Z. Xia, O.
Khaled, V. Mouriès-Mansuy, C. Ollivier, L. Fensterbank, J. Org. Chem,
2016, 81, 7182-7190; b) D. P. Hari, T. Hering, B. Kꢀnig, Org. Lett.
2012,14, 5334−5337; d) P. Maity, D. Kundu,B. C. Ranu, Eur. J. Org.
Chem. 2015, 2015, 1727−1734; e) J. Zhang, J. Chen, X. Zhang, X. Lei,
J. Org. Chem. 2014, 79, 10682−10688.
[22] C. Lévêque, L. Chenneberg, V. Corcé, C. Ollivier, L. Fensterbank,
Chem. Commun. 2016, 52, 9877-9880.
[23] J. Zhou, P.S. Mariano, Photochem. Photobiol. Sci. 2008, 7, 3936404.
[24] K. Ohkubo, K. Mizushima, R. Iwata, K. Souma, N. Suzuki, S. Fukuzumi,
Chem. Commun. 2010, 46, 601-603.
[25] L. Chenneberg, C. Lévêque, V. Corcé, A. Baralle, J. P. Goddard, C.
Ollivier, L. Fensterbank, Synlett 2016, 27, 731-735.
[26] a) D. Matsuoka, Y. Nishigaichi, Chem. Lett. 2015, 44, 163-165; b) Y.
Nishigaichi, A. Suzuki, T. Saito, A. Takuwa, Tetrahedron Lett. 2005, 46,
5149- 5151; c) Y. Nishigaichi, A. Suzuki, A. Takuwa, Tetrahedron Lett.
2007, 48, 211-214.
[7]
[8]
L. Fensterbank, J.-P. Goddard, M. Malacria, C. Ollivier, Chimia 2012,
66, 425-432.
[27] T. A. Halgren, J. L. Firkins, T. A. Fujimoto, H. H. Suzukawa, J. D.
Roberts, Proc. Natl. Acad. SCI. USA. 1971, 68, 3216-3218.
[28] These high barrier values are reminiscent of the values found in ref. 13.
[29] In metalloporphyrins, a) B. L. Westcott; N. E. Gruhn; L. J. Michelsen, D.
L. Lichtenberger J. Am. Chem. Soc. 2000, 122, 8083-8084; In
metalladithiolenes, b) T. Kusamoto, S. Kume, H. Nishihara, J. Am.
Chem. Soc. 2008, 130, 13844–13845; In distonic radical anions, c) G.
Gryn'ova, D. L. Marshall, S. J. Blanksby, M. L. Coote, Nature Chem.
2013, 5, 474–481, d) G. Gryn'ova, M. L. Coote, J. Am. Chem. Soc.
2013, 135, 15392–15403; in DNA base pairs, e) A. Kumar, M. D.
Sevilla, J. Phys. Chem. B 2017, 122, 98–105.
a) E. A. Merritt, B. Olofsson, Angew. Chem. Int. Ed, 2009, 48, 9052-
9070; b) A. Baralle, L. Fensterbank, J. P. Goddard, C. Ollivier, Chem.
Eur. J. 2013, 19, 10809-10813.
[9]
a) J. Li, J. Chen, R. Sang, W.-S. Ham, M. B. Plutschack, F. Berger, S.
Chabbra, A. Schnegg, C. Genicot, T. Ritter, Nat. Chem. 2020, 12, 56-
62; b)M. H. Aukland, M.Šiaučiulis, A. West, G. J. P. Perry, D. J. Procter,
Nat. Catal 2020, 3, 163-169; c) S. Donck, A. Baroudi, L. Fensterbank,
J.-P. Goddard, C. Ollivier, Adv. Synth. Catal. 2013, 355, 1477–1482.
For recent a review: d) Á. Péter, G. J. P. Perry, D. J. Procter, Adv.
Synth. Catal. 2020, 362, 2135-2142.
[10] a) J. D. Nguyen, E. M. D’amato, M. R. Narayanam, C. R. J. Stephenson,
Nature Chem. 2012, 4, 854-859; b) H. Kim, C. Lee, Angew. Chem. Int.
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
This article is protected by copyright. All rights reserved.

10.1002/chem.202100453
Chemistry - A European Journal
FULL PAPER
Entry for the Table of Contents (Please choose one layout)
Layout 1:
FULL PAPER
Etienne Levernier, Khaoula Jaouadi,
Heng-Rui Zhang, Vincent Corcé,
Aurélie Bernard, Geoffrey Gontard,
Claire Troufflard, Laurence Grimaud,
Etienne Derat*, Cyril Ollivier*, Louis
Fensterbank*
Page No. – Page No.
Phenyl Silicates with Substituted
Catecholate Ligands: Synthesis,
Structural Studies and Reactivity
- general study on phenyl silicates
- modification of the catechol ligands
- formation of a phenyl radical
While the generation of aryl radicals by photoredox catalysis is well-documented
under reductive conditions, it has remained challenging under an oxidative
pathway. Because of the easy photooxidation of alkyl bis-catecholato silicates, a
general study on phenyl silicates bearing substituted catecholate ligands has been
achieved. The newly synthesized phenyl silicates have been fully characterized and
their reactivity has been explored. It was found that thanks to the substitution of the
catecholate moiety and notably with the 4-cyanocatecholato ligand, the phenyl
radical could be generated and trapped. Computational studies provided a rationale
for these findings.
This article is protected by copyright. All rights reserved.