ARTICLE
DOI: 10.1038/s41467-018-05375-2

OPEN

Enantioselective radical conjugate additions driven
by a photoactive intramolecular iminium-ion-based
EDA complex
1234567890():,;

Zhong-Yan Cao1, Tamal Ghosh1 & Paolo Melchiorre

1,2

The photochemical activity of electron donor–acceptor (EDA) complexes provides a way to
generate radicals under mild conditions. This strategy has found application in chemical
synthesis and recently in enantioselective catalysis. Reported methods classically relied on the
formation of intermolecular EDA complexes, generated upon aggregation of two suitable
reagents. Herein, we further expand the synthetic utility of this strategy demonstrating that an
intramolecular EDA complex can trigger a photochemical catalytic enantioselective radical
process. This approach enables radical conjugate additions to β-substituted cyclic enones to
form quaternary carbon stereocenters with high stereocontrol using visible light irradiation.
Crucial for success is the use of an amine catalyst, adorned with a carbazole moiety, which
generates, upon condensation with enones, chiral iminium ions that show a broad absorption
band in the visible region. This optical property originates from an intramolecular charge
transfer π–π interaction between the electron-rich carbazole nucleus and the electron-deficient
iminium double bond.

1 ICIQ,

Institute of Chemical Research of Catalonia - the Barcelona Institute of Science and Technology, Avinguda Països Catalans 16, 43007 Tarragona,
Spain. 2 ICREA, Catalan Institution for Research and Advanced Studies, Passeig Lluís Companys 23, 08010 Barcelona, Spain. Correspondence and requests
for materials should be addressed to P.M. (email: pmelchiorre@iciq.es)
NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

1

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

T

he charge-transfer theory was formulated in 1952 by
Robert Mulliken to rationalize the appearance of strong
color on bringing together two colorless organic compounds1. This theory explains how the association of an electronrich substrate (a donor D) and an electron-accepting molecule
(an acceptor A) can induce the formation of a molecular aggregation in the ground state. This new intermediate, termed electron donor–acceptor (EDA) complex (Fig. 1a)2, has physical
properties that differ from those of the separated substrates. For
example, its formation is generally accompanied by the appearance of a new absorption band, the charge-transfer band (hνCT),
which is associated with a transfer of a single electron (SET) from
the donor to the acceptor. Often, the energy of this transition lies
in the visible frequency range.
The photo-physics of EDA complexes have been extensively
studied since the 1950s3–5, while their use in chemical synthesis
was exploited in a more intensive way only 20 years later6–9.
Recently, the resurgence of visible light-driven processes has
motivated chemists to reinvestigate the potential of EDA complex
activation for promoting photochemical processes10–24. For
example, our laboratories applied this strategy for enantioselective
catalysis25–28. Specifically, we exploited the ability of a chiral
catalyst to interact with a weakly polarized substrate (e.g. aldehydes and β-ketoesters) and transform it into an electron-rich
chiral organocatalytic intermediate (such as enamines25–27 or
enolates28). This enabled the aggregation with an electron-poor
reagent to form a colored EDA complex, whose photoactivity
provided access to reactive open-shell intermediates (Fig. 1b). At
the same time, the chiral organocatalytic intermediate ensured
effective stereochemical control over the ensuing radical bondforming process. In general, all the synthetic methods reported so
far6–28 relied on the excitation of intermolecular EDA complexes
formed upon aggregation of two substrates/intermediates.
Here, we further expand the synthetic potential of this activation strategy demonstrating the possibility to form an

a

intramolecular EDA complex and using its photochemical activity
to drive a stereoselective radical reaction (Fig. 1c). Specifically, we
used a chiral catalyst to generate an electron-poor intermediate
upon condensation with a weakly polarized substrate. To induce
the formation of an intramolecular EDA aggregation29–32, the
catalyst was adorned with an electron-rich moiety. The resulting
intramolecular charge transfer π–π interaction elicited a broad
absorption band in the visible region. Irradiation with visible light
initiated a stereoselective radical process. To our knowledge, this
study offers the first demonstration of the potential of photonabsorbing intramolecular EDA complexes in synthetic
applications.
Results
Background. The present study was motivated by our interest in
developing enantioselective radical conjugate additions33–36 via
iminium ion activation37. This organocatalytic activation
mechanism has found many applications for facilitating conjugate
additions in ionic domains. In contrast, it was difficult to develop
a stereoselective iminium ion trap of open-shell species. Recently,
our laboratories reported a strategy, based on the combined
action of a chiral organic catalyst 3, bearing a redox-active carbazole moiety, and a photoredox38 catalyst (PC), that enabled
enantioselective radical conjugate additions to β-substituted cyclic
enones 1 (Fig. 2a)39,40. During these studies, we isolated stable
tetrafluoroborate salts of the chiral iminium ion A-1, generated
upon condensation of a catalyst of type 3 (bearing a 3,6-di-tertbutyl-carbazole moiety) and an aliphatic enone (R′ = Me in 1,
Fig. 2b), which were characterized by X-ray single-crystal analysis. Surprisingly, these crystals showed an intense bright-yellow
coloration (see UV–Vis spectrum in Fig. 2c). Typically, aliphatic
iminium ions of type A can only absorb in the UV region41,42
(below 350 nm, see for example, the absorption spectrum of
iminium ion A-2 in Fig. 2c).

Intermolecular EDA complex between two substrates
Ground state
D

A

+

Donor

KEDA

Excited state
SET

[D,A]

Acceptor

Coloured
EDA complex

Colourless

+

–

Product

[D,A]
hνCT

Radical ion pair

b Intermolecular EDA between a catalytic intermediate and a substrate
Donor

Catalyst

c

S

+

Cat

A

e– rich
intermediate

Substrate

Product

[D,A]

D

Coloured
EDA complex

Asymmetric
process

This work: photochemical activity of an intramolecular EDA complex

D
D

Cat

+

A

Substrate

e– poor
Intermediate

e-rich unit
Catalyst

Visible light

S

Product
Asymmetric process

Intramolecular EDA complex

Fig. 1 The EDA complex activation strategy for photochemical synthetic applications. a The principles of EDA complex activation and synthetic use enabled
by light. b In situ generated catalytic intermediates can form an intermolecular EDA complex with a substrate. c A catalyst, adorned with an electron-rich
unit, reacts with a substrate to generate an electron-poor intermediate prone to intramolecular EDA complex formation. KEDA association constant of EDA
complex formation, D electron-rich molecule, A electron-poor molecule, SET single-electron transfer
2

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

a

O

O
Photocatalyst (PC)
light
H

R
R′

N

R

Radical
precursor

NH2

R′
2
Quaternary carbon

1
Cyclic enones

Chiral amine 3

b
BF4
H
Z-configured
C=N bond

97

Å

N

3.0

e– rich

H
+

+

N

N

PF6

H

A-1
Me
e– poor

Iminium ion

Me
Me

Me
A-2

Iminium ion
Intramolecular EDA complex

CCDC 1437991

Absorbance (a.u.)

c

A-2

1

Catalyst 3b

0.8

A-1

0.6
0.4
0.2
0
300

350

400
Wavelength (nm)

450

500

Fig. 2 Background and initial insights. a Original strategy for the enantioselective iminium ion trapping of radicals based on the combination of a carbazole
amine catalyst 3 and an exogenous photoredox catalyst (PC). b X-ray crystal structure of the carbazole-based iminium ion A-1. c UV–Vis spectra of the
carbazole catalyst 3b (structure in Table 1) and iminium ions A-1 and A-2 [1.0 mM] in CH3CN

The X-ray structure of A-1 provided a rationalization for its
optical properties. The broad absorption band in the visible
region is induced by a stabilizing intramolecular charge-transfer
π–π interaction between the electron-deficient iminium ion
double bond and the electron-rich carbazole nucleus. This type
of intramolecular EDA29–32 association generally results in a
considerable redshift of the absorption bands. The intramolecular
EDA complex also accounted for the interatomic separation
between the carbazole nitrogen and the sp2 α-carbon of the
iminium ion (3.10 Å), which was significantly shorter than the
van der Waals distance.
Given our interest in the photochemical activity and the
synthetic utility of intermolecular EDA complexes generated from
chiral organocatalytic intermediates25–28,43, we wondered if the
visible-light-absorbing properties of the iminium ion-based EDA
complex A could be harnessed to design an enantioselective lighttriggered process44 without the need for an external photoredox
catalyst.

Design plan. Figure 3 details our strategy for exploiting the
photoactivity of the intramolecular iminium-ion-based EDA
complex A, which is transiently generated in the ground state
upon condensation of the carbazole aminocatalyst 3 with cyclic
enone 1.

Visible-light excitation of A would trigger an intracomplex SET
from the electron-rich carbazole (the donor) to the iminium
double bond (the acceptor), furnishing the chiral radical
intermediate B. Using the photoexcitation of EDA complexes in
chemical synthesis is not simple, mainly because of the
unproductive, fast back-electron transfer (BET) which restores
the ground-state EDA complex, making further reactivity
difficult. Therefore, a central element of our approach was to
identify a suitable mechanism to interrupt the BET process45
between B and A. We envisioned that the long-lived carbazole
radical cation in B, which is known to be a persistent species46,
could act as an effective oxidant. Specifically, we hoped that an
SET would occur from an easily oxidazible electron-rich alkyl
trimethylsilane 4 to B, furnishing the neutral 5π-electron βenaminyl radical intermediate C and thus precluding the BET.
This SET manifold would also generate the silyl radical cation D.
The latter intermediate can readily undergo rapid desilylation in
the presence of weak nucleophiles, including water47,48, to form a
carbon radical, which can then enter a radical conjugated
addition manifold39. Importantly, after the stereoselective iminium ion radical trap, the electron-rich carbazole moiety would
be positioned at a strategic position to promote a rapid
intramolecular SET reduction of the unstable α-iminyl radical
cation E49. This regenerates the carbazole radical cation within F,
which would be able to oxidize a new molecule of the organic

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

3

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

Oxidant
N

+

H

N

N
N

+

HN

HN

Visible light

R

SET

3

TMS
C

4

1
– H2O

SET

BET

R′

R′

R′
TMS
D

R

B

A (colored)
O

+

Solv
R

Solv-TMS

Iminium ion A

+

R′
1

H

N
NH2

O

3

N
N

+

Electron relay
chain mechanism

R
R′

E

R
R′
2
H2O

+

N

N
N

N
SET
G

R
R′

R
R

R

TMS

F

R′

4

Fig. 3 Mechanistic proposal. Central to this study is the ability of the carbazole radical cation within B, generated upon visible-light excitation of the
intramolecular EDA complex A, to drive the generation of radicals acting as an SET oxidant. SET single-electron transfer, BET back-electron transfer, Solv
solvent, TMS trimethylsilyl

silane 4. By regenerating the radical, this SET event would be the
propagation step of a radical chain pathway. Hydrolysis of the
neutral imine intermediate G would turn over catalyst 3 while
providing product 2 bearing a quaternary carbon stereocenter.
Optimization of the model reaction. To validate our photochemical plan, we selected β-methyl cyclohexenone 1a as the
model substrate, while using the carbazole catalysts 3 to promote
the formation of the chiral iminium ion A (Table 1). We performed the experiments in CH3CN at 35 °C under irradiation by
a single high-power (HP) LED (λmax = 420 nm) and using an
irradiance of 15 mW/cm2, which was controlled by an external
power supply (full details of the illumination set-up are reported
in Supplementary Figure 1). Water (2 equiv) was added to
facilitate the generation of radicals upon desilylation of intermediate of type D. As radical precursors, we selected a pool of
organic silanes 4a–c with different oxidation potentials (Eox (4·+/
4), as measured by cyclic voltammetry versus Ag/Ag+ in
CH3CN). In consonance with the mechanistic requirement that
radicals originate upon SET oxidation from the carbazole radical
cation in B, only substrates 4 with comparable or lower oxidation
potential than the catalyst carbazole unit should be suitable for
reaction. This was the exact reactivity trend observed in the
photochemical radical conjugate addition promoted by catalyst
4

3a (entries 1–3). Given the oxidizing power of the carbazole
radical cation in 3a (Eox (3a·+/3a) = + 1.15 V), it is no surprise
that only the trimethylsilyl carbazole 4a (Eox (4a·+/4a) = + 0.95
V) provided the desired conjugate addition product 2 bearing a
quaternary stereocenter (entry 1). Substrates 4b and 4c remained
completely unreacted. Substrate 4d, having a dimethyl(phenyl)
silyl group, provided better results than the trimethylsilyl (TMS)
analog (speculatively because of the enhanced β-silyl stabilization
of the intermediate radical cation of type D50,51, compare entries
1 and 4), and was selected for further studies.
We then evaluated a family of chiral catalysts 3b–e bearing
different substitution patterns, which provided the redox-active
carbazole unit with a range of electronic properties. As previously
reported by our group39,40, both enantiomers of catalysts 3 are
accessible in good yields from readily available carbazoles through
a five-step sequence (see Supplementary Note 1 in the Supporting
Information for details on the synthesis of catalyst 3e). In
congruence with the mechanistic proposal, we found a correlation
between the catalytic activity and the oxidation ability of the
carbazole unit. The reaction was effectively promoted by catalysts
bearing more electron-withdrawing groups (3d and 3e), which
magnified the oxidizing power of the carbazole radical cation
(entries 7 and 8). In contrast, the presence of an electrondonating group in 3c completely inhibited the process (entry 6).
The best results in terms of reactivity and stereocontrol were

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

Table 1 Optimization studies

Entry
1
2
3
4
5
6
7
8

Eox (3·+/3) (V)a
+1.15
+1.15
+1.15
+1.15
+1.05
+0.96
+1.16
+1.09

Catalyst 3
3a
3a
3a
3a
3b
3c
3d
3e

[%] Yield of 2b
58
0
0
82
16
0
79
74

4
4a
4b
4c
4d
4d
4d
4d
4d

[%] ee of 2
79
–
–
80
86
–
81
88

Reactions performed at 35 °C on a 0.1 mmol scale using 1.5 equiv. of 4 under illumination by a single high-power (HP) LED (λmax = 420 nm)
aE of the redox-active carbazole unit within catalysts 3a–e, as measured by cyclic voltammetry vs. Ag/Ag+ in CH CN
ox
3
bYield of isolated 2

a
O
Catalyst 3e (20 mol%)
single LED (420 nm)

1a
+

Ph

4d

Ph

5
(1.5 equiv)

Conditions as in
entry 8, Table 1

2a

Ph

N

+

Traces

+
Me

Ph

7: 19% yield

6: 37% yield

b

O

H
Catalyst 3e (20 mol%)
single LED (420 nm)

1a
+
4d
8
(5 equiv)

O
+

Me

Me

N

Conditions as in
entry 8, Table 1

H
2a: 45% yield
85% ee

c

Amine catalyst 3

O

7: 33% yield
50% ee
O

Salicylic acid
(twice the amount of 3)
+

N

Me
1a

4d

SiMe2Ph

Single LED (420 nm)
H2O (2.0 equiv)
CH3CN, 35 °C, 24 h

(S )

Me
N
2a

Catalyst (R,R)-3d, (15 mol%); (S)-2a: 44% yield, 80% ee
Catalyst (S,S)-3b, (5 mol%); no reaction
Catalyst (R,R)-3d (15 mol%) + (S,S)-3b (5 mol%); (S)-2a: 40% yield, 42% ee

Fig. 4 Mechanistic investigations. a Trapping the radical intermediate generated from 4d. b The presence of an hydrogen donor 8 favors the formation of 3methyl cyclohexanone 7. c Entrainment experiment supporting a radical chain propagation mechanism
NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

5

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

Table 2 Substrate scope for the enantioselective trap of carbazole-derived radicals

Reactions performed over 48 h at 35 °C on a 0.1 mmol scale using 1.5 equiv. of 4 under illumination by a single high-power (HP) LED (λmax = 420 nm)
aYield and ee measured on the isolated 2 (average of two runs per substrate)
bReaction time: 72 h
cBenzoic acid (40 mol%) was used

6

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

Table 3 Substrate scope for the enantioselective trap of α-silyl amine-derived radicals

9
Entry
1
2
3
4
5
6
7
8
9

n
1
1
1
1
1
1
1
1
2

R1
Me
Me
Me
Me
Me
Me
Et
Et
Me

R2
H
Cl
H
H
H
H
H
H
H

R3
H
H
Cl
F
MeO
H
H
H
Cl

R4
H
H
H
H
H
Br
H
Cl
H

10
10a
10b
10c
10d
10e
10f
10g
10h
10i

[%] Yielda
82
62
79
81
71
83
57
79
79

[%] ee
80
63
84
79
83
80
79
81
93

Reactions performed over 48 h at 35 °C on a 0.1 mmol scale using 1.5 equiv. of 9 under illumination by a single HP LED (λmax = 420 nm)
aYield and ee measured on the isolated 10 (average of two runs per substrate)

provided by catalyst 3e, which was adorned with substituents
that can enhance both its steric shielding and oxidizing ability
(entry 8).
Mechanistic investigation. Catalyst 3e was selected for further
investigations to better elucidate the mechanism. No product
formation was detected in the absence of light or when replacing
catalyst 3e with cyclohexylamine, which lacked the redox-active
carbazole core. Optical absorption studies (detailed in Supplementary Figure 4) indicated that only the in situ generated iminium ion A could absorb at 420 nm, the operative λ of the system.
When performing the model reaction in the presence of ethene1,1-diyldibenzene 5, we observed the formation of adduct 6,
arising from the trap of the radical photogenerated from substrate
4d (Fig. 4a). We also detected 3-methyl cyclohexanone 7, a
compound which likely arose from the β-enaminyl radical
intermediate C (Fig. 3) upon hydrogen abstraction. The last
observation suggests a possible pathway to turn over the catalyst 3
involved in the light-triggered radical generation manifold.
We then performed an additional experiment, where the model
reaction was carried out in the presence of an excess of 1,4cyclohexdiene 8, a hydrogen donor that afforded 7 in a better
yield and with moderate enantioselectivity (Fig. 4b). Control
experiments established that catalyst 3e, substrate 4d, and light
were all necessary for the formation of product 7. Overall, these
observations are congruent with the proposed mechanism for
radical formation, where the photogenerated intermediate B
oxidizes substrate 4 to furnish the radical along with intermediate
C.
Finally, we conducted an experiment using two carbazole
catalysts with an opposite absolute configuration (Fig. 4c). We
mixed 15 mol% of (R,R)-3d, a catalyst that can efficiently
promote the model reaction to give product (S)-2a in 80% ee,
with 5 mol% of catalyst (S,S)-3b, which is completely inactive in
these conditions. This catalyst mixture afforded the product (S)2a with a greatly reduced enantioselectivity (42% ee) with respect
to using 3d alone. This result implies that, while only (R,R)-3d
can be involved in the radical generation pattern, the subsequent
iminium ion trap is governed by both catalysts, thus causing an

erosion of stereocontrol. This entrainment experiment (the nonactive catalyst 3b at initiation is instead apt at propagation)
supports the mechanism proposed in Fig. 3, where the EDAcomplex-based photochemical radical generation serves to initiate
a radical conjugate addition sustained by a chain propagation
mechanism. Our attempts to determine the quantum yield of the
process, which would be useful to support the feasibility of a
chain manifold, were frustrated by the impossibility of carefully
establishing the photon flux of the high-power violet LED needed
for the photochemical processes, a crucial requirement for
reliably determining the quantum yield.

Substrate scope and synthetic applications. Adopting the optimized conditions described in Table 1, entry 8, we demonstrated
the generality of the radical conjugate addition by evaluating a
variety of cyclic enones 1 and α-carbazole silyl reagents 4
(Table 2). The resulting chiral products 2 contain a quaternary
carbon center bearing a carbazole moiety, a structure found in
optoelectronic materials, synthetic dyes, conducting polymers52,
and naturally occurring alkaloids53.
A wide range of substituents at the carbazole core were welltolerated, regardless of their electronic properties and position
(products 2a–k). Heterocycle rings, including furyl and thienyl
moieties, could be included in the products (entries 7 and 8). As
for the enones, a wide range of carbocycles and β-olefin
substituents could be used. Ring systems that incorporate βalkyl groups other than methyl reacted with a reduced
enantioselectivity (entries 12–14), while the radical conjugate
addition performed well for a diverse range of ring sizes,
including cyclopentenyl and cycloheptenyl architecture (entries
15 and 16). Interestingly, linear enones reacted smoothly to
provide the corresponding compounds 2q-r, albeit with moderated or no stereoselectivity (entries 17–18). Crystals from
compound 2a were suitable for X-ray crystallographic analysis,
which established the absolute configuration of the quaternary
stereocenter.
The general applicability of a chemical strategy is crucial for
evaluating its usefulness. We wondered if the photochemical

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

7

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

a

O

(R,R)-3e (20 mol%)
PhHN

+

1a

HP single LED (420 nm)
Irradiance 40 ± 2 mW/cm2

CO2H

Me

PhCO2H (40 mol%)
CH3CN, 35 °C, 48 h

11 (2.0 equiv)

H
N

Ph

12
61%, 47% ee

OBn

O

O

EtO2C

(R,R)-3d (20 mol%)

CO2Et

HP single LED (405 nm)
Irradiance 50 ± 2 mW/cm2

+
Me

Me

N
H

Me

13 (3.0 equiv)

b

PhCO2H (80 mol%)
CH3CN, 35 °C, 4 d

Me
OBn
14
42%, 43% ee
F

F
O

(R,R)-3e (20 mol%)

+

LED (460 nm)
PhCO2H (40 mol%)

N

CH3CN, 15 °C, 3 d

i) NH2NH2, KOH
diethyleneglycol
130 °C, 2 h
N
ii) 200 °C, 4 h
(R)-15
57% yield, 50% ee

SiMe2Ph
Ref 58: 7 steps, 1.2% total yield of racemic 15

Over 2 steps

c

Cl

Cl

OH

(R,R)-3e (20 mol%)

O

LED (420 nm)
PhCO2H (40 mol%)
+

N

L-Selectride
(2 equiv)

CH3CN, 35 °C, 5 d

–78 °C to rt,
THF, 2 h

Me
SiMe2Ph

Me
N
Trans-16
41% yield, 11:1 dr, 71% ee
Over 2 steps

Fig. 5 Expansion of the scope and synthetic utility. a Using different radical precursors than silanes. b, c Enantioselective synthesis of biologically relevant
compounds

activity of the iminium-ion-based EDA complex could be
expanded to successfully generate radicals from α-silyl amines
935,54–56, organic silanes with a low oxidation potential (Eox < 1.0
V, see Supplementary Figure 12 for details). As detailed in
Table 3, the resulting photochemical radical conjugate addition to
cyclic enones efficiently afforded an array of adducts 10 with a
quaternary stereocenter. In contrast, both α-indole and α-pyrrole
silyl reagents remained unreacted under the optimal conditions
(see Supplementary Figure 2).
Preliminary investigations indicated that our protocol is
compatible with different radical precursors other than silanes.
For example, N-phenylglycine 1157 (Eox = + 0.42 V vs. SCE in
CH3CN) and the (benzyloxy)methyl-substituted dihydropyridine
13 (Eox = + 1.08 V vs. Ag/Ag+ in CH3CN) provided the
corresponding adducts 12 and 14, respectively, with promising
enantioselectivity (Fig. 5a). We also applied our method for the
streamlined preparation of biologically relevant compounds.
Figure 5b details a straightforward enantioselective two-step
synthesis of compound 15, which is potentially useful to prevent
ophthalmic disease58. Last, we could stereoselectively prepare
trans-alcohol 16, the epimer of a compound useful for the
treatment of influenza (Fig. 5c)59.
In summary, we have demonstrated that an intramolecular
EDA complex can be used to trigger a photochemical asymmetric
radical process. Crucial for implementing the process was a chiral
carbazole catalyst that do not contain any photosensitive unit, but
rather it guides the photochemical formation of radicals by
inducing the transient formation of visible-light-absorbing
intramolecular iminium-ion-based EDA complexes. The method,
8

which avoids the need for external photoredox catalysts, offers the
first example of synthetic applications triggered by the photochemical activity of an intramolecular EDA complex.
Methods
General. For the exemplary reaction setup, see Supplementary Fig. 1. For more
information on unreactive substrates, see Supplementary Fig. 2. For UV–Vis
absorption studies, see Supplementary Figs. 3, 4. For cyclic voltammetry studies, see
Supplementary Figs. 5–13. For the NMR spectra of compounds in this article, see
Supplementary Figs. 14–47. For the product analysis with HPLC on chiral stationary phase, see Supplementary Figs. 48–80. For crystallographic information, see
Supplementary Fig. 81 and the Supplementary Tables 1, 2, 3. For the synthesis of
catalyst 3e and substrates 2, 4, and 9, see Supplementary Note 1. For the preparation of the iminium ion A-2, see Supplementary Note 2. For details of
mechanistic studies, see Supplementary Note 3. For details of synthetic applications, see Supplementary Note 4. For general information, general experimental
procedure, and analytic data of compounds synthesized, see Supplementary
Methods. For additional references pertinent to the Supplementary Information,
see Supplementary References.
Catalytic enantioselective radical addition reaction to enones. Exemplary, a 15
mL Schlenk tube was charged with the chiral carbazole-derived primary amine
catalyst (R,R)-3e (0.04 mmol, 20 mol%), acid (0.08 mmol, 40 mol%, salicylic acid
for substrate 4 and benzoic acid for substrates 9), the organic silane 4 or 9 (0.15
mmol, 150 mol%), enone 1 (0.1 mmol, 100 mol%), H2O (0.2 mmol, 200 mol%), and
200 μL of CH3CN. The mixture was placed under an atmosphere of argon, cooled
with liquid nitrogen, and degassed via vacuum evacuation (5 min), backfilled with
argon, and warmed to room temperature. The freeze-pump–thaw cycle was
repeated three times, and then the Schlenk tube was placed into a support fitted
with a 420 nm high-power single LED (λ = 420 nm). The irradiance was regulated
at 15 ± 2 mW/cm2, as controlled by an external power supply and measured using a
photodiode light detector at the start and the end of each reaction; the temperature
was kept at 35 °C with a chiller connected to the irradiation plate (the setup is
detailed in Supplementary Fig. 1). This setup secured a reliable irradiation and

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

temperature while keeping a distance of 1 cm between the reaction vessel and the
light source. Stirring was maintained for the indicated time (generally 48 h), and
then the irradiation was stopped. The reaction volatiles were removed in vacuum
and the residue was purified by column chromatography to give the products 2 or
10 in the stated yield and enantiomeric purity. The reported yield and ee are
average of two runs per substrate.
Data availability. The X-ray crystallographic coordinate for structures reported in
this study have been deposited at the Cambridge Crystallographic Data Centre
(CCDC) under deposition numbers CCDC-1819014 (2a). These data can be
obtained free of charge from The Cambridge Crystallographic Data Centre via
www.ccdc.cam.ac.uk/data_request/cif. The authors declare that all other data
supporting the findings of this study are available within the article and Supplementary Information files, and also are available from the corresponding author
upon reasonable request.

Received: 1 May 2018 Accepted: 3 July 2018

References
1.
2.
3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.
17.
18.

Mulliken, R. S. Molecular compounds and their spectra. II. J. Am. Chem. Soc.
74, 811–824 (1952).
Foster, R. Electron donor-acceptor complexes. J. Phys. Chem. 84, 2135–2141
(1980).
Rosokha, S. V. & Kochi, J. K. Fresh look at electron-transfer mechanisms via
the donor/acceptor bindings in the critical encounter complexes. Acc. Chem.
Res. 41, 641–653 (2008).
Hilinski, E. F., Masnovi, J. M., Amatore, C., Kochi, J. K. & Rentzepis, P. M.
Charge-transfer excitation of electron donor-acceptor complexes. Direct
observation of ion pairs by time-resolved (picosecond) spectroscopy. J. Am.
Chem. Soc. 105, 6167–6168 (1983).
Singh, J. O., Anunziata, J. D. & Silber, J. J. n–π Electron donor–acceptor
complexes. II. Aliphatic amines dinitrobenzenes. Can. J. Chem. 63, 903–907
(1985).
Rathore, R. & Kochi, J. K. Donor/acceptor organizations and the electrontransfer paradigm for organic reactivity. Adv. Phys. Org. Chem. 35, 193–318
(2000).
Fox, M. A., Younathan, J. & Fryxell, G. E. Photoinitiation of the SRN1 reaction
by excitation of charge-transfer complexes. J. Org. Chem. 48, 3109–3112
(1983).
Sankararaman, S., Haney, W. A. & Kochi, J. K. Annihilation of aromatic cation
radicals by ion-pair and radical pair collapse. Unusual solvent and salt effects
in the competition for aromatic substitution. J. Am. Chem. Soc. 109,
7824–7838 (1987).
Gotoh, T., Padias, A. B. & Hall, J. H. K. An electron donor-acceptor complex
and thermal triplex as intermediates in the cycloaddition reaction of Nvinylcarbazole with dimethyl 2,2-dicyanoethylene-1,1-dicarboxylate. J. Am.
Chem. Soc. 113, 1308–1312 (1991).
Lima, C. G. S., Lima, T., Duarte, M., Jurberg, I. D. & Paixão, M. W. Organic
synthesis enabled by light-irradiation of EDA complexes: theoretical
background and synthetic applications. ACS Catal. 6, 1389–1407 (2016).
Tobisu, M., Furukawa, T. & Chatani, N. Visible light-mediated direct arylation
of arenes and heteroarenes using diaryliodonium salts in the presence and
absence of a photocatalyst. Chem. Lett. 42, 1203–1205 (2013).
Pham, P. V., Nagib, D. A. & MacMillan, D. W. C. Photoredox catalysis: a mild,
operationally simple approach to the synthesis of α-trifluoromethyl carbonyl
compounds. Angew. Chem. Int. Ed. 50, 6119–6122 (2011).
Nappi, M., Bergonzini, G. & Melchiorre, P. Metal-free photochemical
aromatic perfluoroalkylation of α-cyano arylacetates. Angew. Chem. Int. Ed.
53, 4921–4925 (2014).
Kandukuri, S. R. et al. X-ray characterization of an electron donor-acceptor
complex that drives the photochemical alkylation of indoles. Angew. Chem.
Int. Ed. 54, 1485–1489 (2015).
Davies, J., Booth, S. G., Essafi, S., Dryfe, R. A. W. & Leonori, D. Visible-lightmediated generation of nitrogen-centered radicals: metal-free hydroimination
and iminohydroxylation cyclization reactions. Angew. Chem. Int. Ed. 54,
14017–14021 (2015).
Cheng, Y., Yuan, X., Ma, J. & Yu, S. Direct aromatic C–H trifluoromethylation
via an electron-donor–acceptor complex. Chem. Eur. J. 21, 8355–8359 (2015).
Spell, M. L. et al. A visible-light-promoted O-glycosylation with a
thioglycoside donor. Angew. Chem. Int. Ed. 55, 6515–6519 (2016).
Quint, V. et al. Metal-free, visible light-photocatalyzed synthesis of benzo[b]
phosphole oxides: synthetic and mechanistic investigations. J. Am. Chem. Soc.
138, 7436–7441 (2016).

19. Böhm, A. & Bach, T. Radical reactions induced by visible light in
dichloromethane solutions of Hünig’s base: synthetic applications and
mechanistic observations. Chem. Eur. J. 22, 15921–15928 (2016).
20. Beatty, J. W. et al. Photochemical perfluoroalkylation with pyridine N-oxides:
mechanistic insights and performance on a kilogram scale. Chem. 1, 456–472
(2016).
21. Marzo, L., Wang, S. & König, B. Visible-light-mediated radical arylation of
anilines with acceptor-substituted (hetero)aryl halides. Org. Lett. 19,
5976–5979 (2017).
22. Liu, Y.-Y. et al. Visible-light-driven aza-ortho-quinone methide generation for
the synthesis of indoles in a multicomponent reaction. Angew. Chem. Int. Ed.
56, 9527–9532 (2017).
23. Zhang, J., Li, Y., Xu, R. & Chen, Y. Donor–acceptor complex enables alkoxyl
radical generation for metal-free C(sp3)–C(sp3) cleavage and allylation/
alkenylation. Angew. Chem. Int. Ed. 56, 12619–12623 (2017).
24. Liu, B., Lim, C.-H. & Miyake, G. M. Visible-light-promoted C−S crosscoupling via intermolecular charge transfer. J. Am. Chem. Soc. 139,
13616–13619 (2017).
25. Arceo, E., Jurberg, I. D., Álvarez-Fernández, A. & Melchiorre, P.
Photochemical activity of a key donor–acceptor complex can drive
stereoselective catalytic α-alkylation of aldehydes. Nat. Chem. 5, 750–756
(2013).
26. Arceo, E., Bahamonde, A., Bergonzini, G. & Melchiorre, P. Enantioselective
direct α-alkylation of cyclic ketones by means of photo-organocatalysis. Chem.
Sci. 5, 2438–2442 (2014).
27. Bahamonde, A. & Melchiorre, P. Mechanism of the stereoselective αalkylation of aldehydes driven by the photochemical activity of enamines. J.
Am. Chem. Soc. 138, 8019–8030 (2016).
28. Woźniak, Ł., Murphy, J. J. & Melchiorre, P. Photo-organocatalytic
enantioselective perfluoroalkylation of β-ketoesters. J. Am. Chem. Soc. 137,
5678–5681 (2015).
29. Cram, D. J. & Day, A. C. Macro rings. XXXI. Quinone derived from [2.2]
paracyclophane, an intramolecular-molecular complex. J. Org. Chem. 31,
1227–1232 (1966).
30. Yoshida, M., Tatemitsu, H., Sakata, Y. & Misumi, S. Layered compounds.
XXXVIII. Synthesis of an intramolecular charge transfer complex containing a
7,7,8,8-tetracyanoquinodimethane unit. Tetrahedron Lett. 17, 3821–3822
(1976).
31. Slama-Schwok, A., Blanchard-Desce, M. & Lehn, J.-M. Intramolecular charge
transfer in donor-acceptor molecules. J. Phys. Chem. 94, 3894–3902 (1990).
32. Iwanaga, T., Ogawa, M., Yamauchi, T. & Toyota, S. Intramolecular chargetransfer interaction of donor-acceptor-donor arrays based on anthracene
bisimide. J. Org. Chem. 81, 4076–4080 (2016).
33. Sibi, M. P., Ji, J., Wu, J. H., Gürtler, S. & Porter, N. Chiral Lewis acid catalysis
in radical reactions: enantioselective conjugate radical additions. J. Am. Chem.
Soc. 118, 9200–9201 (1996).
34. Sibi, M. P., Ji, J., Sausker, J. B. & Jasperse, C. P. Free radical-mediated
intermolecular conjugate additions. Effect of the Lewis acid, chiral auxiliary,
and additives on diastereoselectivity. J. Am. Chem. Soc. 121, 7517–7526
(1999).
35. Espelt, R. L., McPherson, I. S., Wiensch, E. M. & Yoon, T. P. Enantioselective
conjugate additions of α-amino radicals via cooperative photoredox and Lewis
acid catalysis. J. Am. Chem. Soc. 137, 2452–2455 (2015).
36. Huo, H., Harms, K. & Meggers, E. Catalytic, enantioselective addition of alkyl
radicals to alkenes via visible-light-activated photoredox catalysis with a chiral
rhodium complex. J. Am. Chem. Soc. 138, 6936–6939 (2016).
37. Lelais, G. & MacMillan, D. W. C. Modern strategies in organic catalysis: the
advent and development of iminium activation. Aldrichim. Acta 39, 79–87
(2006).
38. Shaw, M. H., Twilton, J. & MacMillan, D. W. C. Photoredox catalysis in
organic chemistry. J. Org. Chem. 81, 6898–6926 (2016).
39. Murphy, J. J., Bastida, D., Paria, S., Fagnoni, M. & Melchiorre, P. Asymmetric
catalytic formation of quaternary carbons by iminium ion trapping of radicals.
Nature 532, 218–222 (2016).
40. Bahamonde, A., Murphy, J. J., Savarese, M., Brémond, É., Cavalli, A. &
Melchiorre, P. Studies on the enantioselective iminium ion trapping of radicals
triggered by an electron-relay mechanism. J. Am. Chem. Soc. 139, 4559–4567
(2017).
41. Cai, X., Chang, V., Chen, C., Kim, H.-J. & Mariano, P. S. A potentially general
method to control relative stereochemistry in enone–olefin 2+2photocycloaddition reactions by using eniminium salt surrogates. Tetrahedron
Lett. 41, 9445–9449 (2000).
42. Hörmann, F. M., Chung, T. S., Rodriguez, E., Jakob, M. & Bach, T. Evidence
for triplet sensitization in the visible light-induced [2+2] photocycloaddition
of eniminium ions. Angew. Chem. Int. Ed. 57, 827–831 (2018).
43. Silvi., M. & Melchiorre, P. Enhancing the potential of enantioselective
organocatalysis with light. Nature 554, 41–49 (2018).

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications

9

ARTICLE

NATURE COMMUNICATIONS | DOI: 10.1038/s41467-018-05375-2

44. Brimioulle, R., Lenhart, D., Maturi, M. M. & Bach, T. Enantioselective catalysis
of photochemical reactions. Angew. Chem. Int. Ed. 54, 3872–3890 (2015).
45. Szostak, M., Fazakerley, N. J., Parmar, D. & Procter, D. J. Cross-coupling
reactions using samarium(II) iodide. Chem. Rev. 114, 5959–6039 (2014).
46. Prudhomme, D. R., Wang, Z. & Rizzo, C. J. An improved photosensitizer for
the photoinduced electron-transfer deoxygenation of benzoates and m(trifluoromethyl)benzoates. J. Org. Chem. 62, 8257–8260 (1997).
47. Yoshida, J., Kataoka, K., Horcajada, R. & Nagaki, A. Modern strategies in
electroorganic synthesis. Chem. Rev. 108, 2265–2299 (2008).
48. Dockery, K. P. et al. Nucleophile-assisted cleavage of benzyltrialkylsilane
cation radicals. J. Am. Chem. Soc. 119, 1876–1883 (1997).
49. Jakobsen, H. J., Lawesson, S. O., Marshall, J. T. B., Schroll, G. & Williams, D. H.
Mass spectrometry. XII. Mass spectra of enamines. J. Chem. Soc. B 940–946 (1966).
50. Yoshida, J., Maekawa, T., Murata, T., Matsunaga, S. & Isoe, S. The origin of βsilicon effect in electron-transfer reactions of silicon-substituted heteroatom
compounds. Electrochemical and theoretical studies. J. Am. Chem. Soc. 112,
1962–1970 (1990).
51. Remeur, C., Kelly, C. B., Patel, N. R. & Molander, G. A. Aminomethylation of
aryl halides using α‑silylamines enabled by Ni/photoredox dual catalysis. ACS
Catal. 7, 6065–6069 (2017).
52. Blouin, N. & Leclerc, M. Poly(2,7-carbazole)s: structure-property
relationships. Acc. Chem. Res. 41, 1110–1119 (2008).
53. Roy, J., Jana, A. K. & Mal, D. Recent trends in the synthesis of carbazoles: an
update. Tetrahedron 68, 6099–6121 (2012).
54. Yoon, U. C. & Mariano, P. S. Mechanistic and synthetic aspects of amineenone single electron transfer photochemistry. Acc. Chem. Res. 25, 233–240
(1992).
55. Zhang, X.-M. & Mariano, P. S. Mechanistic details for SET-promoted
photoadditions of amines to conjugated enones arising from studies of
aniline-cyclohexenone photoreactions. J. Org. Chem. 56, 1655–1660 (1991).
56. Hasegawa, E., Xu, W., Mariano, P. S., Yoon, U. C. & Kim, J. U. Electrontransfer-induced photoadditions of the silyl amine, Et2NCH2TMS to α,βunsaturated cyclohexenones. Dual reaction pathways based on ion pairselective cation-radical chemistry. J. Am. Chem. Soc. 110, 8099–8111 (1988).
57. Millet, A., Lefebvre, Q. & Rueping, M. Visible-light photoredox-catalyzed
Giese reaction: decarboxylative addition of amino acid derived α-amino
radicals to electron-deficient olefins. Chem. Eur. J. 22, 13464–13468 (2016).
58. Quach, T., Berman, J. & Garvey, D. S. Opsin-binding Ligands, Compositions
and Methods of Use. F. Hoffmann-La Roche AG, Grenzacher-strasse, Basel,
WO 2013/081642 A1 (2013).
59. Chen, L. et al. Compounds for the Treatment and Prevention of Influenza.
Bikam pharmaceuticals, INC., Thorndike Street, Cambridge, WO 2011/
095576 A1 (2011).

10

Acknowledgements
Financial support was provided by the ICIQ, Agencia Estatal de Investigación (AEI,
CTQ2016-75520-P and SEV-2013-0319), and the European Research Council (ERC
681840-CATA-LUX). Z.-Y. Cao thanks the EU for a Horizon 2020 Marie SkłodowskaCurie Fellowship (grant 702405).

Author contributions
P.M. conceived the idea and supervised the whole project. Z.-Y. C. and T.G. designed and
carried out the experiments. All authors discussed the results, contributed to writing the
manuscript, commented on the manuscript, and approved the final version of the
manuscript for submission.

Additional information
Supplementary Information accompanies this paper at https://doi.org/10.1038/s41467018-05375-2.
Competing interests: The authors declare no competing interests.
Reprints and permission information is available online at http://npg.nature.com/
reprintsandpermissions/
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.

Open Access This article is licensed under a Creative Commons
Attribution 4.0 International License, which permits use, sharing,
adaptation, distribution and reproduction in any medium or format, as long as you give
appropriate credit to the original author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made. The images or other third party
material in this article are included in the article’s Creative Commons license, unless
indicated otherwise in a credit line to the material. If material is not included in the
article’s Creative Commons license and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this license, visit http://creativecommons.org/
licenses/by/4.0/.
© The Author(s) 2018

NATURE COMMUNICATIONS | (2018)9:3274 | DOI: 10.1038/s41467-018-05375-2 | www.nature.com/naturecommunications