A Bifunctional Photoaminocatalyst for the Alkylation of Aldehydes: Design, Analysis, and Mechanistic Studies

Article abstract of DOI:10.1021/acscatal.8b01331

A bifunctional photoaminocatalyst based on imidazolidinone and thioxanthone is presented. The preparation of these catalysts proceeds in a two-step synthesis that allows an easy tuning of the steric properties. The photophysical and electrochemical data of the imidazolidinone photocatalysts have been determined, indicating that the catalysts can work under visible light conditions. To corroborate the experimental observations, ground state geometry optimization and energy transition studies of thioxanthone and bifunctional catalyst 4c were optimized by time-dependent density functional theory (TD DFT) calculations. The alkylation of aldehydes with this photoaminocatalyst works with high enantioselectivities and yields due to the stereoelectronic properties of the catalyst. A rational mechanistic cycle based on different mechanistic experiments, TD DFT calculations, and laser flash photolysis is presented.

Full text of DOI:10.1021/acscatal.8b01331

Subscriber access provided by UNIV OF DURHAM
Article
A Bifunctional Photo-Aminocatalyst for the Alkylation of
Aldehydes: Design, Analysis and Mechanistic Studies
Thomas Rigotti, Antonio Casado-Sanchez, Silvia Cabrera, Raul Perez-
Ruiz, Marta Liras, Víctor A. de la Peña O'Shea, and José Alemán
ACS Catal., Just Accepted Manuscript • DOI: 10.1021/acscatal.8b01331 • Publication Date (Web): 15 May 2018
Downloaded from http://pubs.acs.org on May 15, 2018
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
A Bifunctional PhotoꢀAminocatalyst for the
Alkylation of Aldehydes: Design, Analysis and
Mechanistic Studies
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Thomas Rigotti,a Antonio Casado-Sánchez,a Silvia Cabrera,b,c Raúl Pérez-Ruiz,d Marta Liras,d
Víctor A. de la Peña O'Shea,d and José Alemána,c*
a
Organic Chemistry Department, Módulo 1, Universidad Autónoma de Madrid, 28049 Madrid
(Spain)
b Inorganic Chemistry Department, Módulo 7, Universidad Autónoma de Madrid, 28049 Madrid
(Spain)
c
Institute for Advanced Research in Chemical Sciences (IAdChem), Universidad Autónoma de
Madrid, 28049 Madrid (Spain).
d Photoactivated Processes Unit, IMDEA Energy, Av. Ramón de la Sagra 3, 28935, Móstoles
(Madrid, Spain)
ACS Paragon Plus Environment
1

ACS Catalysis
Page 2 of 42
1
2
3
4
5
6
7
8
9
ABSTRACT. Abifunctional photoaminocatalyst based on imidazolidinone and thioxanthone is
presented. The preparation of these catalysts proceeds in a twoꢀstep synthesis which allows an
easy tuning of the steric properties. The photophysical and electrochemical data of the
imidazolidinone photocatalysts have been determined, indicating that the catalyst can work under
visible light conditions. In order to corroborate the experimental observations, ground state
geometry optimization and energy transition studies of thioxanthone and the bifunctional catalyst
4c were optimized by TD DFT calculations. The alkylation of aldehydes with this aminoꢀ
photocatalyst works with high enantioselectivities and yields due to the stereoelectronic
properties of the catalyst. A rational mechanistic cycle based on different mechanistic
experiments, TD DFT calculations, and laser flash photolysis is presented.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
KEYWORDS: Bifunctional photocatalyst • Organocatalysis • Pyrrolidines • Photochemistry •
Mechanism
INTRODUCTION
Photocatalysis has emerged as one of the best strategies for the development of sustainable and
novel processes for the synthesis of organic compounds.1 In spite of the enormous growth in this
area, the vast majority of the methodologies developed provide achiral or racemic compounds
using either metallic complexes or organic molecules as a photocatalyst. The main reason for this
is the difficulty of combining a chiral catalyst to induce a sterochemical control, with the high
reactivity and low activation barriers of radical intermediates.2 In spite of this difficulty,
asymmetric photocatalytic transformations have been successfully accomplished by the use of a
dualꢀcatalyst approach using a combination of a photocatalyst and a chiral organocatalyst.3 The
ACS Paragon Plus Environment
2

Page 3 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
use of only one catalyst that combines chirality and photoredox properties is not only more
convenient but also more challenging and consequently, is less explored. The pioneering work
carried out by Bach’s group developed the photoꢀasymmetric [2+2] cycloaddition by employing
a chiral bifunctional catalyst, containing a chromophore unit, that induced a stereocontrol by
hydrogen bond interactions.4 A similar strategy was applied by Xiao’s group who developed a
bifunctional metalꢀphotocatalyst for an enantioselective aerobic oxidation in which the
photosensitizer is linked to a chiral bisoxazoline metal complex.5 More recently, Meggers and
coworkers have developed new chiral iridium and rhodium complexes that simultaneously act as
photocatalysts and as chiral Lewis acids to provoke the asymmetric induction.6 However, to the
best of our knowledge, a bifunctional photoꢀaminocatalyst which contains a photocatalyst and
aminocatalyst has not been developed yet.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
During the last few years, the asymmetric αꢀalkylation of aldehydes and ketones7 has become
an attractive topic because of the synthetic utility of the products obtained. In 2008, MacMillan
2+
developed a pioneering and remarkable example in which the photoredox catalyst Ru(bpy)3
and an imidazolidinone organocatalyst were employed in a cooperative manner3a to furnish the α
ꢀalkylation of aldehydes using different radical precursors (left, Figure 1). Since then diverse
dual catalytic systems based on other metal complexes8 and semiconductors9 as photocatalysts
have also been reported. Alternatively, the αꢀalkylation of aldehydes using metalꢀfree
photocatalysts was developed by Zeitler’s10 and Ferroud’s11 groups using the dyes Eosin Y and
Rose Bengal, respectively. The combination of organic photocatalyst (Eosin Y or Rose Bengal)
and MacMillan’s imidazolidinone catalyst made the alkylation of aldehydes with alkyl halides
possible, but the enantioselectivity achieved was in some cases lower compared to that of metalꢀ
based dual catalytic systems. These findings emphasize that the αꢀalkylation of aldehydes is not a
ACS Paragon Plus Environment
3

ACS Catalysis
Page 4 of 42
1
2
3
4
5
6
7
8
9
trivial process and that there are difficulties in finding the appropriate conditions for the
compatibility of two different catalytic systems. In addition, the development of a general photoꢀ
organocatalyst for the αꢀalkylation of aldehydes (instead of using an external photocatalyst for
each photoactivatable reagent) will be a significant advance in the field of visible light
photoredox catalysis. The design of a unique metalꢀfree catalyst – called a bifunctional photoꢀ
aminocatalyst – capable of activating both, the photoactivatable reagent and the substrate
simultaneously, would be highly desirable (right, Figure 1). In the field of organocatalysis, the
dualꢀactivation of reagents by a bifunctional organocatalyst is an area of great importance, which
has led to the development of novel transformations.12 Most of the reported bifunctional catalysts
are combinations of a Lewis base with a Lewis acid moiety. However, the incorporation of a
photoorganocatalyst in a bifunctional catalyst is much scarcer, and more particularly, the
combination of a photoorganocatalyst and an aminocatalyst has not yet been developed.7
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Comparison of the two catalytic systems, dual catalytic and bifunctional photoꢀ
aminocatalyst systems.
For the design of the new photoꢀaminocatalysts, different factors need to be taken into account
(Figure 2). Firstly, the election of the chiral scaffold that will incorporate the two catalytic
moieties is crucial for the success of its application. 4ꢀImidazolidinone has been chosen for two
ACS Paragon Plus Environment
4

Page 5 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
reasons: i) The catalysts usually employed for the alkylation of aldehydes in most of the dual
catalytic systems are MacMillan’s imidazolidinone catalysts, which are chiral 4ꢀimidazolidinone
derivatives. ii) These catalysts possess two groups (R1 and R2) that could allow the incorporation
of two different units (topꢀFigure 2). The R1 group would permit the easy modulation of the
steric hindrance of the aminocatalyst, whereas the second (R2) would also incorporate the
photocatalytic unit. Moreover, the synthesis of chiral 4ꢀimidazolidinones is simple, short, and
assembles the R1 and R2 groups from two separate reactants which allows an easy fineꢀtuning of
the properties of the catalyst. Regarding the photocatalytic unit, thioxanthones have been widely
studied as photoinitiators in polymerization processes,13 and as efficient photocatalysts in
energyꢀ and electronꢀtransfer proceses.14 In addition, Bach’s group has reported that
thioxanthones can also be used as visible lightꢀphotocatalysts in asymmetric transformations.15
We envisaged that thioxanthones would be good candidates as the photocatalytic unit for the
synthesis of the bifunctional photoꢀaminocatalyst. In particular, 2ꢀsubstituted thioxanthones were
chosen as building blocks for the synthesis of the photoꢀaminocatalyst.16 The second
consideration is related to the geometry of the enamine in the aminocatalyst which has been
studied in depth by MacMillan and others.3 Therefore, in our case, because the incorporation of
the aromatic photocatalytic unit (R2) is required, and in order to control the geometry in the
enamine, the R1 group must be of a different relative size compared to R2 (BottomꢀFigure 1).
Herein, we present a bifunctional photoꢀaminocatalyst by a combination of both an aminoꢀ
organocatalyst (imidazolidinone) and a photoꢀorganocatalyst (thioxanthone) in the same
molecule. In addition, the photophysical and electrochemical data of photo aminocatalysts, their
application in the alkylation of aldehydes and a rational catalytic cycle based on different
mechanistic studies, TD DFT calculations, and laser flash photolysis, are also presented.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
5

ACS Catalysis
Page 6 of 42
1
2
3
4
5
6
7
8
DESIGN OF THE PHOTO-AMINOCATALYST
• Structure
Easy Incoportation
Into the Catalytic Unit
("Visible light" dye)
O
O
N
N
9
H
COR4
R2
R2
R1
Photochemistry
R1
N
+
NH2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
H
Control
O
Easy Steric Modulation
(large number of optically
pure amino-acids)
Steric Control
S
• Enamine-geometry
O
O
N
• Relative Size of R1 and R2 controls
the geometry
N
R2
R2
R1
R1
N
N
R3
favored when R1 bulkier than R2
R3
favored when R2 bulkier than R1
Figure 2. Design of the photoꢀaminocatalysts and initial considerations.
RESULTS AND DISCUSSION
Synthesis of the Bifunctional Photocatalysts and Characterization. The
photoaminocatalysts 4 were prepared in two steps, starting from commercially available
aminoacids (Scheme 1). Firstly, the aminoacids 1a-c were easily transformed into Nꢀmethyl
aminoamides 2a-c in high yields. The aminoacetylation of the corresponding
carboxythioxanthone17 3 with amides 2a-c led to the desired photoꢀaminocatalyst 4a-d with
moderate to good trans/cis ratios. The absolute configuration of the major diastereoisomer of
catalysts 4b, and 4c was determined by singleꢀcrystal Xꢀray analysis18 (see bottom Scheme 1). In
the structures both substituents (R1 and thioxanthone) are in trans geometry with the
thioxanthone moiety in a pseudoꢀequatorial disposition.
ACS Paragon Plus Environment
6

Page 7 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of photoꢀaminocatalysts 4 and Xꢀray crystal structure of 4b and 4c.
Next, we studied the photophysical and electrochemical properties of the bifunctional
catalysts 4 to understand their suitability for a photocatalytic process (Table 1). For comparison
the same study was also carried out using thioxanthone under identical experimental conditions.
The UVꢀVis absorption spectrum of thioxanthone presents a maximum absorption maximum at
λmax = 381 nm with a significant part of the absorption band in the visible region. The different
structures and stereochemistry of the aminoꢀthioxanthone catalysts 4 showed little effect on the
maximum absorption (λmax = 380 ± 1 nm). Similar findings were described by Neumann et al.
who reported that the 2ꢀsubstitution in the aromatic ring of the thioxanthone did not produce any
significant effect on the absorption maximum, or on the molar absorption coefficients, regardless
of the nature of the substituent.16 Moreover, strong emissions at λmax = 413ꢀ420 nm were
observed for all catalysts upon photoexcitation of the absorption maximum at room temperature.
The excited singlet state energy of the photoaminocatalyst was estimated at the intersection of
the normalized absorption and emission spectra. As we expected, all the S1 energy values are
ACS Paragon Plus Environment
7

ACS Catalysis
Page 8 of 42
1
2
3
4
5
6
very close to that of thioxanthone, which indicates that the imidazolidinone fragment has little
effect on the photophysical properties of the thioxanthone moiety.
7
8
9
The groundꢀstate redox potentials of the catalysts were determined by cyclic voltammetry (CV)
in acetonitrile as solvent (Table 1). Aminoꢀthioxanthone catalysts 4 show two irreversible
oxidation peaks: the most positive corresponds to the oxidation of the thioxanthone moiety and
the less cathodic to the amine moiety (see S.I. for further details). This assignment was
confirmed by comparing the redox values of the bifunctional catalysts 4 with the oxidation peaks
of thioxanthone and imidazolidinone catalysts 4f (1.40 V vs. SCE in CH3CN). In addition, one
reversible reduction potential was also determined, which corresponds to the thioxanthone
reduction. Taking into account the redox potentials and the singlet state energy, the oxidation and
reduction excited state potentials of catalysts 4 were estimated. Both, oxidation and reduction
potentials in the excited state of the bifunctional catalysts 4 are similar to those calculated for
thioxanthone which indicates the preservation of the electrochemical properties of the
thioxanthone moiety on the bifunctional catalyst properties, which confirms the adequate design
of the bifunctional catalysts 4.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
8

Page 9 of 42
ACS Catalysis
1
2
3
4
5
6
Table 1. Photophysical and electrochemical data of thioxanthone and catalysts 4a-d, and
structures of parent imidazolidinone catalyst 4e and 4f.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Eox
E*
Ered
E*
red
ox
Emission
Absorption
S1 Energy
(eV)c
Compound
(λmax
,
(λmax, nm)a
(V vs
(V vs
(V vs
(V vs
nm)b
SCE) d
SCE)e
SCE)d
SCE)e
Thioxanthone
381
381
409
3.13
3.14
+1.82
+1.32
ꢀ1.31
ꢀ1.31
ꢀ1.48
ꢀ1.45
+1.65
+1.69
412
416
413
420
414
4a
+1.83
+1.44
381
381
379
380
3.12
3.13
3.14
3.13
ꢀ1.32
ꢀ1.33
ꢀ1.32
ꢀ1.29
ꢀ1.50
ꢀ1.50
ꢀ1.35
ꢀ1.44
+1.62
+1.63
+1.79
+1.69
4b
4c
+1.80
+1.36
+1.80
+1.08
epi-4c
4d
+1.82
+1.49
+1.84
+1.50
epi-4d
379
413
3.15
ꢀ1.31
ꢀ1.41
+1.74
+1.84
a UVꢀVis spectra recorded in 0.025 mM CH2Cl2 solutions at rt. b Fluorescence recorded in 0.25
mM CH2Cl2 solutions at rt. c Determined from the intersection of the normalized absorbance and
ACS Paragon Plus Environment
9

ACS Catalysis
Page 10 of 42
1
2
3
4
5
6
7
8
9
d
emission spectra (converted into eV). Conditions: 50 mV·sꢀ1 scan rate; 1 mM solution of
compound in argonꢀsaturated CH3CN solution; 0.1 M solution of Bu4NPF6; glassy carbon disk
(2.8 mm diameter) as working electrode; platinum wire as counter electrode; Ag/AgCl as
d
reference electrode. Excited state redox potentials of the thioxanthone moiety estimated using
the equation: E*ox = Eox ꢀ E0ꢀ0 or E*red = Ered + E0ꢀ0, where E0ꢀ0 is the singlet state energy.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In order to corroborate the experimental photophysical observations, the ground state geometry
optimization of thioxanthone and the bifunctional catalyst 4c, as well as their energy transition
studies, were optimized by TD DFT calculations (see Figure 3 and S.I.). Geometry optimization
of the thioxanthone ground state (S0) revealed a planar structure with C2v symmetry similar to
that previously described in the literature (right, Figure 3).19,20 Other possible conformations such
as nonꢀplanar structures with Cs symmetry described by Rubio Pons et al.21 were not considered
because they are less stable conformations.22 The absorption spectra of thioxanthone exhibited a
first band located at 381 nm that, by taking our calculations into account, can be assigned to the
πꢀπ∗ transitions from HOMO to LUMO orbitals (see S.I. for further details). In addition, an nꢀπ∗
optically forbidden transition (S2), energetically closeꢀlying to S1 was also observed, which is in
agreement with Ishijima’s studies.23 On the other hand, Mundt et al.19 observed a change in the
energy of these excited states with a lower nꢀπ∗ transition in vacuum conditions. These variations
are due to the solvent effect, which stabilizes the πꢀπ∗ excited state of the thioxanthone. In
addition, the optimization of the triplet state leads to an energy of 2.73 eV that is close to the
experimental value of 2.84 eV.24 As described above (see Table 1), the UVꢀvis of 4c depicts a
similar profile to thioxanthone (see left, Figure 3). The TDꢀDFT calculations showed that S1
transition takes place between the HOMO, which is delocalized over the whole molecule with a
contribution from the amine moiety, and the LUMO localized predominantly on the
thioxanthone (Figure 3). We found that the triplet state of 4c exhibited an energy value of 2.73
eV, which is identical to that of free thioxanthone.
ACS Paragon Plus Environment
10

Page 11 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3. Comparison between the experimental (black line) and the calculated vertical (red
bars) excitation by TDꢀDFT B3LYP (6ꢀ311++G**) for catalyst 4c (left), using DMF as solvent,
and thioxanthone (right) using DCM as solvent. Inset: Molecular orbitals involved in the S1
transition.
The αꢀalkylation of aldehydes with diethyl bromomalonate was chosen as a model
reaction3a to evaluate the catalytic activity of the bifunctional photocatalysts 4 (see below, Table
2). Moreover, using the usual formulation for the determination of the free energies of a
photoinduced electron transfer (PET) process (∆GPET = 23.06[Eox ꢀ Ered] ꢀ E*(S1 or T1)), we can
estimate the thermodynamic driving force of the PET between the diethyl bromomalonate as the
reactant vide infra (Ered = ꢀ 1.0 V vs. SCE in CH3CN) and catalysts 4 (see Eox and E* values in
Table 1). All the estimated ∆GPET values are negative (ranging from ꢀ9.0 to ꢀ9.9 Kcal/mol) which
indicates that the process is thermodynamically favorable for all the catalysts 4.
ACS Paragon Plus Environment
11

ACS Catalysis
Page 12 of 42
1
2
3
4
Optimization Studies and Scope of the Reaction
5
6
7
8
Initially, the αꢀalkylation of hydrocinnamaldehyde (5a) was carried out using MacMillan’s
imidazolidinone catalyst 4e (20 mol%) without any external photosensitizer (entry 1) under
visible light irradiation (23 W commercial fluorescent bulb) using DMF as solvent. The reaction
furnished the αꢀalkylated product 7a with a 41% conversion and a 91% enantiomeric excess. The
catalyst 4f was then employed in the same reaction conditions (entry 2) obtaining a negligible
conversion (<5%), indicating that this catalyst did not work under these reaction conditions.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2+
Then, catalysts 4e and 4f has been employed in combination with Ru(bpy)3 as an external
photocatalyst (entries 3 and 4) obtaining, respectively, an 85% and a 59% conversion and
enantiomeric excess values of 91% and 97%. We then tested our bifunctional photocatalyst 4a-d
(entries 5ꢀ15). To our delight, the bifunctional photocatalyst 4a allowed the synthesis of 7a with
complete conversion and a moderate enantiomeric excess (64% ee, entry 5). The use of catalyst
4b with a slightly bulky substituent, such as the benzyl group (R1= Bn), provided comparable
enantioselectivity (57%, entry 6), whereas the catalyst with the bulkier tert-butyl moiety (4c)
achieved an outstanding >99% ee and full conversion (entry 7). By contrast, the epimer
photocatalyst epiꢀ4c showed very low reactivity and enantioselectivity (entry 8). Furthermore,
catalysts 4d and epi-4d with a quaternary stereocenter showed lower enantiomeric excesses and
conversions than catalyst 4c (entries 9ꢀ10). Using the best catalyst 4c, other solvents such as
DMSO or CH2Cl2 were also studied, obtaining in all cases identical enantioselectivities, but
lower conversions (entries 11ꢀ12). Conversely, an apolar solvent such as toluene afforded the
product with a low conversion (entry 13). Moreover, the reduction of the catalyst loading
involved a strong conversion decrease (entry 14) and the use of blue LEDs (λmax= 450 nm) as an
irradiation source afforded aldehyde 7a in lower enantioselectivity (entry 15). We then carried
ACS Paragon Plus Environment
12

Page 13 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
out the reaction in the presence of catalyst 4f and thioxanthone (dual catalytic system) and the
results were compared to catalyst 4c (entries 7 and 16). As it can be observed, only a very slight
better result was obtained in the case of the bifunctional catalytic system (4c), The latter result
forced us to investigate the mechanistic differences between the dual catalytic system and the
bifunctional catalyst (see below).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 2. Screening of different bifunctional photoꢀorganocatalysts 4 in the αꢀalkylation of
aldehyde 5a.a
Catalyst
(x mol%)
ee
(%)c
91
Ent.
Solvent Conv.b (%)
1
2
4e (20)
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMF
DMSO
CH2Cl2
41
<5
85
4f (20)
n.d.
91
3
4e (20) + Ru(bpy)32+ (1)
4f (20) + Ru(bpy)32+ (1)
4a (20)
4
59
97
5
100
77
64
6
4b (20)
57
7
4c (20)
100
11
>99
8d
8
epi-4c (20)
4d (20)
9
71
37
10
11
12
epi-4d (20)
4c (20)
30
29
60
>99
>99
4c (20)
14
ACS Paragon Plus Environment
13

ACS Catalysis
Page 14 of 42
1
2
3
4
5
6
7
8
9
13
14
4c (20)
4c (10)
Toluene
5
96
91
70
97
DMF
DMF
DMF
35
15e
4c (20)
100
97
16
4f (20) + thioxanthone (20)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
5a (0.3 mmol), 6a (0.15 mmol), 2,6ꢀlutidine (0.3 mmol) and catalyst 4 in 0.3 mL of the
o
b
indicated solvent were stirred at 15 C for 18 h under visible light irradiation (CFLꢀ23 W).
Conversion of 7a determined by 1H NMR. c Determined by SFC on chiral stationary phase after
d
e
derivatization (see S.I.). Opposite enantiomer obtained. Reaction carried out under blue LED
(385 mW) irradiation. Conv. Conversion. n.d. Not determined
Once the optimal conditions had been determined (entry 8, Table 2), we studied the scope of
the αꢀalkylation of aldehyde 5a using a variety of alkylating reagents (Table 3). Bromomalonates
6a and 6b (a tertiary bromide) were reduced by catalyst 4c to give aldehydes 7a and 7b with an
excellent yield and enantioselectivies. Later, the electron reduction of bromoꢀnitrile species
(bromoacetonitrile, Ered= ꢀ1.23 V vs SCE in CH3CN, see S.I. for the cyclic voltammetry) by the
photoꢀorganocatalyst 4c was explored, and the negative value of the free energy (∆GPET= ꢀ2.3
Kcal/mol) guarantees the thermodynamic driving force of the process. The reaction with nitriles
6c and 6d provided the aldehydes 7c and 7d respectively, in high yields and excellent
enantioselectivities. It is worth mentioning that these results are better than those previously
described in the literature.7 αꢀBromoketone derivatives 6e-g (e.g. bromoacetophenone, Ered= ꢀ
0.89 V vs SCE) were also reduced using our catalytic system, yielding the ketones 7e, 7f and 7g,
respectively, with an excellent yield and high levels of enantiocontrol (84ꢀ95% ee). Furthermore,
the reaction of benzyl derivative 6h with aldehyde 5a was successfully catalyzed by the
bifunctional thioxanthone 4c to afford 7h in 97% ee.
ACS Paragon Plus Environment
14

Page 15 of 42
ACS Catalysis
1
2
3
4
Table 3. Scope of the αꢀalkylation of 5a with bromoderivatives 6.a
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
5a (0.24 mmol), 6 (0.12 mmol), 2,6ꢀlutidine (0.24 mmol) and
catalyst 4 in 0.24 mL of DMF were stirred at 15 oC under visible light
irradiation (CFLꢀ23 W). b Based on a 30% conversion by 1 H NMR. c 5
d
e
eq. of 5a were employed. DMSO used as solvent. Overall yield of
f
both diastereoisomers (dr 2:1). Reaction performed on a 1 mmol
scale.
Next, the applicability of the photocatalyst was evaluated using different aldehydes (5) with
bromomalonate 6a as bromoꢀderivative (Table 4). Primary and secondary βꢀalkylꢀaldehydes 5i
and 5j exhibited good reactivity, affording the corresponding aldehydes 7i and 7j with excellent
enantioselectivies (96% and 97% ee, respectively). To the best of our knowledge, these
enantioselectivies are the highest reported in the literature for those substrates.7 The bifunctional
catalytic system also allowed the successful alkylation of functionalized aldehydes 1k and 1l.
ACS Paragon Plus Environment
15

ACS Catalysis
Page 16 of 42
1
2
3
4
5
6
7
8
9
Accordingly, the alkylatedꢀaldehydes 7l and 7k, containing an NꢀBocꢀpiperidine and a double
bond respectively, were isolated in good yields and high enantioselectivities (96% and 99% ee,
respectively). It must be highlighted that the level of enantioselectivity achieved for the αꢀ
alkylation of aldehydes using the bifunctional catalyst 4c are not only comparable, but also better
than those reported in the literature using dual catalytic systems. These better enantioselectivities
are due to the more sterically hindered nature of the bifunctional photocatalyst 4c compared to
MacMillan’s catalyst 4e (compare entries 3, 4, 7 and 16 of Table 2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Table 4. Scope of the reaction of different aldehydes (5) with diethyl bromomalonate (6a).a
O
Br
O
Catalyst 4c (20 mol%)
R
R
+
EtO2C
CO2Et
DMF, 15 ºC,
2,6-lutidine
23 W CFL
EtO2C
CO2Et
5
6a
7
O
O
O
( )4
EtO2C
Ph
EtO2C
CO2Et
CO2Et
EtO2C
CO2Et
7a, 95% Yield
7i, 81% Yield
7j, 55% Yieldb
>99% ee
96% ee
97% ee
Boc
N
Et
O
O
( )3
EtO2C
CO2Et
EtO2C
CO2Et
7k, 62% Yieldb
7l, 73% Yieldb
96% ee
99% ee
a
5 (0.24 mmol), 6a (0.12 mmol), 2,6ꢀlutidine (0.24
mmol) and catalyst 4 in 0.24 mL of DMF were stirred at
o
b
15 C under visible light irradiation (CFLꢀ23 W). 5 eq.
of the corresponding aldehyde 5 were used.
ACS Paragon Plus Environment
16

Page 17 of 42
ACS Catalysis
1
2
3
4
5
6
7
Mechanistic Considerations
8
9
Originally Proposed Photo-Catalytic Cycle: According to the original mechanism proposed by
MacMillan and others,3,7ꢀ11 we postulate the mechanism outlined in Scheme 2 for the
photoalkylation of aldehydes using the bifunctional catalyst 4c. The reaction starts with the
condensation of catalyst 4c with aldehyde 5 to give the first enamine intermediate A. The
thioxanthone moiety of intermediate A can absorb light to reach the excited intermediate A*
(step 1). It is wellꢀknown that thioxanthones can promote a single electron transfer process
(SET).14 Therefore, this excited intermediate A* can reduce the bromoꢀalkane derivative 6 (step
2) through a SET reduction, to give intermediate B and the alkyl radical. Then, the alkyl radical
is added to the nucleophilic enamine B to obtain the αꢀamino radical C that can be
intramolecularly oxidized by the thioxanthone radicalꢀcation. Finally, the resultant iminium D is
hydrolyzed to give the final αꢀalkylated aldehyde 7. In order to confirm this mechanistic
proposal, different experiments involved in the process, and related to the different steps, were
carried out.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
17

ACS Catalysis
Page 18 of 42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Proposed mechanism for the αꢀalkylation of aldehydes employing catalyst 4c in
accordance with the original mechanism proposed by MacMillan.3a
Initiation Process (Step 2): Three different mechanisms have been proposed in the literature
for the photochemical initiation process of this reaction, depending on the organocatalyst
employed, the presence of an external photosensitizer, and/or the type of bromoꢀderivative
(Scheme 3). The first was proposed by MacMillan3 and coworkers for the dual catalytic system,
using iridium(III) or ruthenium(II) complexes and MacMillan’s imidazolidinone catalyst (topꢀ
left, Scheme 3). In this case, the reducing excited state of the metal photocatalyst is able to
reduce the bromoalkane thought a SET process, giving the radical intermediate I (topꢀleft,
Scheme 3). The second mechanism was described by Melchiorre et al.25 In the latter case, a
colored EDA complex intermediate between the enamine and a benzylbromide is formed and can
be photoexcited to give a SET from the enamine to the bromobenzyl derivative obtaining the
ACS Paragon Plus Environment
18

Page 19 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
radical intermediate II (bottomꢀleft, Scheme 3). The third and most recent mechanism proposed
is related to the direct photoexcitation of the enamine,26 generated upon condensation of an
aldehyde and JørgensenꢀHayashi’s catalyst, which is able to reach an excited state and to
undergo a SET with bromo malonates to form the radical intermediate I (RightꢀScheme 3). Once
the radical intermediate I or II was formed by some of these mechanisms,27 the enamine can trap
the intermediate (I or II) as described in step 3 of Scheme 2.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Three different pathways for the initiation process of the αꢀalkylation of aldehydes.
In order to distinguish which of the initiation mechanisms from the three different pathways
was applicable, and to elucidate the role of the thioxanthone in our catalytic system, we studied
the αꢀalkylation of hydrocinnamaldehyde using three different catalysts 4e, 4f and 4g (Scheme
4). We chose diethyl bromomalonate as the reagent in order to discard the EDA complex type
mechanism because this requires an electron deficient bromobenzyl (equation b, Scheme 3).
Catalyst 4f was chosen because it has an aromatic substituent with a similar steric hindrance
around the secondary amine active center compared to that of our bifunctional catalyst (4c), but
ACS Paragon Plus Environment
19

ACS Catalysis
Page 20 of 42
1
2
3
4
5
6
7
8
9
without the thioxanthone moiety. The results obtained using 4f will allow us to distinguish the
role of the thioxanthone moiety in the bifunctional catalyst 4c. In addition, the imidazolidinone
catalyst 4e and JørgensenꢀHayashi’s catalyst (4g) were also studied for comparison with the
mechanisms reported in the literature.3,27 Firstly, the reaction was carried out in the presence of
catalyst 4e-g, without any external photocatalyst (Scheme 4). As Melchiorre reported,25 the
JørgensenꢀHayashi catalyst 4g and catalyst 4e afforded the alkylated product 7a, with 100% and
41% conversion respectively, due to the photoexcitation of the intermediate enamine (equation c,
Scheme 3). By contrast, catalyst 4f did not produce any alkylated product while a 41%
conversion was obtained with the MacMillan’s catalyst 4e. This last result seems to indicate the
possible photoꢀabsorption of the enamine generated upon the condensation of aldehyde 5e and
catalyst 4e. On the other hand, the addition of the ruthenium(II) complex as an external
photocatalyst in combination with aminocatalyst 4e or 4f increased the formation of the αꢀ
alkylated product 7a up to 85% and 59%, respectively. These results would indicate that the
enamine formed upon condensation of catalyst 4f and aldehyde 5a is not able to generate the
alkyl radical (through a SET between the photoexcited enamine and the bromo derivative) and
that the reaction needs an external photocatalyst to proceed.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
20

Page 21 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 4. αꢀAlkylation of hydrocinnamaldehyde using catalyst 4e, 4f and 4g in the absence and
presence of external photocatalyst.
To obtain more information and to understand the differences in terms of the reactivity in the
reaction conducted with catalyst 4e, 4f and 4g, three different enamines (4e-En, 4f-En and 4g-
En, Figure 4)28 were synthesized by condensation of phenylacetaldehyde with the different
aminocatalysts.27 CV and UVꢀVis spectra of the enamines were performed and the excited redox
potentials of the enamines were calculated (see S.I. and Figure 4). According to the excited
reduction potential of the three enamines (E*red from ꢀ 2.38 to ꢀ2.74 V vs SCE in CH3CN), all
should be able to reduce the diethyl bromomalonate (Ered = ꢀ 1.0 V vs. SCE in CH3CN) without
the need of an external photocatalyst. The UVꢀVis spectra of the enamines were also compared.
Enamine 4g-En derived from the JørgensenꢀHayashi´s catalyst has a significant absorption
around 400 nm that guarantees the visible light absorption of the enamine, which can act as
photoinitiator and reduce the bromomalonate 6a via direct photooxidation without the addition of
an external photocatalyst. However, this cannot be the reason for the different behavior between
ACS Paragon Plus Environment
21

ACS Catalysis
Page 22 of 42
1
2
3
4
5
6
7
8
9
the enamines 4e-En and 4f-En since they have almost identical UVꢀVis absorption spectra but a
different reactivity (41% vs <5% conversion, respectively). An additional factor must to be
involved in the null conversion obtained in the case of catalyst 4f (with a phenyl substituent). We
think that in the case of catalyst 4f the steric hindrance of the phenyl group is responsible for the
lack of any initiation process needed to generate the alkyl radical from the bromo derivative
through the direct photoexcitation of the enamine. The addition of an external photocatalyst in
combination with 4f overcomes the null contribution of the direct photoexcitation of the
enamine, enabling the initiation step of the reaction and giving good conversion (59%, Scheme
4). Therefore, taking into account that catalyst 4f and the bifunctional catalyst 4c possess a
similar steric hindrance around the secondary amine active center, we assumed that the
thioxanthone group is responsible for any initiation step process in the case of the bifunctional
catalyst 4c.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
F3C
CF3
CF3
O
O
N
N
tBu
tBu
Ph
Me
N
N
N
OTMS
CF3
4g-En
4e-En
4f-En
E*red = - 2.38 V
E*red = - 2.73 V
E*red = - 2.74 V
ACS Paragon Plus Environment
22

Page 23 of 42
ACS Catalysis
1
2
3
4
Figure 4. Spectroscopic and CV studies of enamines 4e-En, 4f-En and 4g-En (V vs SCE in
5
6
CH3CN).
7
8
9
Catalyst steric effects in the initiation process: As described previously, the reaction under
catalyst 4f led to a null conversion whereas the alkylation took place with 100% conversion
under catalyst 4c (Scheme 4). Since we thought that the steric hindrance around the active center
of the catalyst is the main responsible for the different reactivities observed, some additional
experiments were performed. It is well known that the SET process proceeds between two
spatially close molecules (enamine and bromomalonate derivative), and therefore, the steric
factor around the enamine must play an important role in the initiation process. With these ideas
in mind we synthetized a variety of imidazolidinone catalysts 4e-i and performed the alkylation
under the same reaction conditions (see Scheme 5).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Firstly, we carried out the reaction using each catalyst 4, without the addition of an
external photocatalyst, with the aim of identifying the effect of the steric hindrance of the
intermediate enamine on the initiation process. Regarding the catalysts in transꢀconfiguration,
the conversions obtained using the different catalysts transꢀ4 show that only the catalyst with the
smallest substituent (methyl, transꢀ4e) is able to promote the initiation of the reaction by direct
photoexcitation of the enamine. The transꢀdisposition of the substituents shielded both faces of
the enamine (Scheme 5), and as a consequence the bulkiest catalysts (trans-4f, trans-4h and
trans-4i) inhibit the SET process from the enamine to the bromoderivative 6a. In these cases, the
2+
addition of Ru(bpy)3 as an external photocatalyst, responsible for the photoꢀinitiation step
processes, overcame this inhibition and allowed the reaction to progress with all the catalysts
except in the case of the bulkiest catalyst trans-4i and cis-4i, indicating that the corresponding
enamines trans-4i-En and cis-4i-En were not formed. Furthermore, the enamine ratio (versus the
ACS Paragon Plus Environment
23

ACS Catalysis
Page 24 of 42
1
2
3
4
1
secondary amine catalyst) was determined by H NMR (in CD3CN and DMFꢀd7), showing that
5
6
7
8
4e-En and 4f-En are present in comparable quantities under reaction conditions (see S.I.). These
results indicate that a different amount of the enamine intermediate cannot be the responsible of
the different activity observed for the two catalysts 4e and 4f, in the absence of an external
photosensitizer. In addition, quenching studies of the two enamines (4e-En and 4f-En) in the
presence of diethyl bromomalonate were conducted. These latter studies showed that the Sternꢀ
Volmer constant value in the cases of 4e-En is the double than in the case of 4f-En (see S.I.).
Moreover, a similar study was performed in DMF (under the same reaction conditions,
concentration and solvent), obtaining a negligible quenching of 4f-En. By contrast, the
quenching phenomena of 4e-En was observed, suggesting that a different behavior of the two
catalysts under reaction conditions is taking place.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Conversely, it is well known that enamines formed from imidazolidinones with substituents in
cisꢀorientation have one of the faces unblocked (bottom, Scheme 5). Therefore, phenyl and
cyclopropyl catalysts cisꢀ4f and cisꢀ4h were able to carry out the αꢀalkylation of aldehydes even
in the absence of an external photocatalyst because the enamine and the bromo derivative can
come into closer proximity compared to the case of the transꢀcatalysts. Although the reactivity of
catalysts cisꢀ4 was higher than the corresponding transꢀ4, their enantiomeric discrimination was
poor and for this reason these cisꢀcatalysts are not usually employed in photoꢀaminocatalysis.3,7
To conclude this section, the initiation step under the bifunctional catalyst 4c proceeds towards
the thioxanthone moiety and the electronic transfer between the enamine and the bromo
derivative can be discarded.
ACS Paragon Plus Environment
24

Page 25 of 42
ACS Catalysis
1
2
3
4
5
6
O
CO2Et
CO2Et
Ph
Cat. (20 mol%)
DMF, 2,6-lutidine
O
EtO2C
CO2Et
H
H
N2, rt
CFL 23W
18 h
Br
Ph
5a
6a
7a
7
8
9
O
O
O
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
N
N
N
trans
tBu
Ph
tBu
Me tBu
tBu
tBu
N
H
N
N
N
H
H
H
trans-4h
trans-4e
trans-4i
trans-4f
Conv.
(without
photocat)
0%
0%
41%
<5%
Conv.
(with [Ru])a
85%
59%
82%
0%
O
N
O
N
O
N
tBu
cis
tBu
tBu
tBu
Ph
N
H
N
H
N
H
cis-4h
cis-4i
cis-4f
Conv.
(without
photocat)
25%
18%
0%
Conv.
(with [Ru])a
47%
80%
3%
a 1 mol% of Ru(bpy)32+ was added
Scheme 5. Effect of the size and disposition of the catalyst substituents in the αꢀalkylation of
aldehydes.
Propagation Process and Quantum Yields: In the αꢀalkylation of aldehydes, Yoon29 and
Melchiorre27 have shown that a chain propagation mechanism is involved in the reaction.
According to these studies, the accepted mechanism of the alkylation of aldehyde is outlined in
Scheme 6. The radical I is trapped by enamine II to form the αꢀaminoꢀradical III that can be
oxidized by the bromo derivative 6. Next, the iminium ion IV generated is hydrolyzed to recover
catalyst 4. Therefore, the generation of additional radical species to react with II, via the
initiation process, is unnecessary since propagation of the process can take place (6 to I). Yoon
ACS Paragon Plus Environment
25

ACS Catalysis
Page 26 of 42
1
2
3
4
5
6
7
8
9
and Melchiorre have demonstrated the propagation processes by quantum yield measurements
(Φ). In Melchiorre’s photocatalytic system, the JørgensenꢀHayashi’s catalyst presented a Φ =
18,26 whereas Yoon reported that the MacMillan’s catalyst had a Φ = 22 in this photocatalytic
system.27
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 6. Accepted radicalꢀchain propagation process for the αꢀalkylation of aldehydes.27, 29
In order to study the propagation process of our photocatalytic system, we measured the
quantum yield of the overall reaction with the bifunctional catalyst 4c. Using our irradiation
system, the quantum yield of catalysts 4g, and 4e or 4f with thioxanthone as the external
photocatalyst was also determined (Figure 5). The quantum yields were measured by the
ferrioxalate actinometer methodology, following the procedure described by the IUPAC.30
A
quantum yield of 9.2 was obtained in the reaction catalysed by JørgensenꢀHayashi’s catalyst (4g)
under our irradiation system (Figure 5). This datum is lower than that reported by Melchiorre due
to the different reaction and irradiation conditions, but is in agreement with a selfꢀpropagation
mechanism (Φ > 1). For catalysts 4e and 4f, quantum yields of 5.6 and 4.4 respectively, were
ACS Paragon Plus Environment
26

Page 27 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
measured using thioxanthone as external photocatalyst. By contrast, a quantum yield of 2.4 was
obtained using our bifunctional catalyst 4c. We think that the trend of these values indicates the
level of accessibility of the αꢀamino radical III by the bromo derivative 6. Indeed, the greatest
quantum yield value (Φ = 9.2) was observed for the less hindered intermediate III obtained
employing catalyst 4g, which has one of the faces of the pyrrolidine ring plane completely
unblocked. On the other hand, the methyl substituent in catalyst 4e is partially hindering one of
the two faces, observing a quantum yield value of 5.6. The presence of a bulkier substituent like
phenyl or thioxanthone decreased this value to 4.4 and 2.4, respectively, because of the
additional steric hindrance present on one of the faces of the imidazolidinone ring plane, which is
blocking the approach of the bromo derivative 6 to the αꢀamino radical III. This reduction in the
overall quantum yield, in the case of our bifunctional catalyst 4c, can also be attributed to the
contribution of the internal redox process (the oxidation of the αꢀaminoꢀradical and the reduction
of the radical cation thioxanthone; see intermediate C, step 4 in Scheme 2). Therefore, an
additional SET between the thioxanthone group of catalyst 4c and the bromoderivative to initiate
the catalytic cycle is necessary (step 2, in Scheme 2) which leads to a further decrease in the
quantum yield value observed.31
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACS Paragon Plus Environment
27

ACS Catalysis
Page 28 of 42
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Overall quantum yield of the αꢀalkylation of hydrocinnamaldehyde with
diethylbromomalonate using different catalytic systems.
Laser flash photolysis experiments: In order to gain further insight into the mechanism, laser
flash photolysis experiments (LFP) were performed to investigate the excited state of
bifunctional catalyst 4c involved in the process, and its comparison with thioxanthone. The
transient spectrum of thioxanthone in deoxygenated DMF shows an absorption maximum at 630
nm, which corresponds to the characteristic tripletꢀtriplet band of thioxanthone.32 Next, we
focused our attention on the kinetic decays of both compounds at 630 nm (Figure 6). The decays
of catalyst 4c and thioxanthone were markedly different, with triplet lifetime values (τ) much
higher for thioxanthone (1.879 ꢁs) than for the bifunctional photocatalyst 4c (0.682 ꢁs).
Therefore, the triplet quenching rate constant (kq) was determined to be 9.3 x 105 sꢀ1. These
results can be explained based on the photochemical studies of thioxanthones as photoinitiatiors
in the polimerization of olefins.13 In these studies, the thioxanthone triplet excited state is
quenched by amines given the corresponding diꢀradicals by an electroꢀtransfer mechanism
ACS Paragon Plus Environment
28

Page 29 of 42
ACS Catalysis
1
2
3
4
5
6
7
8
9
(Scheme 7). This intermediate could undergo decomposition via proton transfer from the aminoꢀ
radical to the ketone radical or backꢀelectron transfer (BET) to the ground state. We have
experimentally confirmed that the degradation of 4c was negligible, confirming the rapid BET
process. Similarly, a decrease in the fluorescence emission of the bifunctional catalyst compared
to that of thioxanthone was also observed due to deactivation by the amino moiety (see S.I.).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6. Normalized triplet decay traces of thioxanthone and catalyst 4c at 630 nm (λexc = 385
nm).
Selective irradiation
to the thioxanthone moiety
hν
O
O
O
S
N
O
S
N
*
λ
exc = 385 nm
tBu
N
tBu
N
H
H
4c
O
Electron transfer
fast BET
O
S
N
tBu
N
H
Scheme 7. Deactivation mechanism of the bifunctional catalyst 4c.
ACS Paragon Plus Environment
29