Electrophile, Substrate Functionality, and Catalyst Effects in the Synthesis of α-Mono and Di-Substituted Benzylamines via Visible-Light Photoredox Catalysis in Flow

Article abstract of DOI:10.1002/cctc.201800754

We report herein the facile and one-pot synthesis of α-mono- and di-substituted benzylamines from cheap and readily available α-amino acids, via photocatalytic decarboxylative arylation in flow. This enables to access intermediates and building blocks that are difficult to obtain via other synthetic routes, but are key for the manufacture of pharmaceuticals, agrochemicals, and fine chemicals. The optimal decarboxylative conditions were identified through a high-throughput evaluation of catalysts, organic or inorganic bases, ligands, and reaction parameters (i. e., contact time, temperatures, and photoelectron power). The reaction turned out to be electronically controlled as the yields increased with increasing electron-density on the aryl moiety. The results were correlated with the redox properties of the photocatalysts, deriving catalyst structure-performance relationships which can facilitate the future identification of even better materials. In addition, compared to traditional batch chemistry, the use of a flow protocol led to quicker reactions (30 min instead of 12–72 h) and ensured more predictable reaction scale-ups.

Full text of DOI:10.1002/cctc.201800754

Accepted Article
Title: Electrophile, Substrate Functionality, and Catalyst Effects in
the Synthesis of α-Mono and Di-Substituted Benzylamines via
Visible-Light Photoredox Catalysis in Flow
Authors: Gianvito Vilé
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.201800754
Link to VoR: http://dx.doi.org/10.1002/cctc.201800754
A Journal of
www.chemcatchem.org

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
Electrophile, Substrate Functionality, and Catalyst Effects in the
Synthesis of -Mono and Di-Substituted Benzylamines via
Visible-Light Photoredox Catalysis in Flow
Gianvito Vilé,*[a] Sylvia Richard,[a] Arnaud Lhuillery,[a] and Georg Rueedi[a]
Abstract: We report herein the facile and one-pot synthesis of α-
mono- and di-substituted benzylamines from cheap and readily
available α-amino acids, via photocatalytic decarboxylative arylation
in flow. This enables to access intermediates and building blocks that
are difficult to obtain via other synthetic routes, but are key for the
manufacture of pharmaceuticals, agrochemicals, and fine chemicals.
The optimal decarboxylative conditions were identified through a high-
throughput evaluation of catalysts, organic or inorganic bases, ligands,
and reaction parameters (i.e., contact time, temperatures, and
photoelectron power). The reaction turned out to be electronically
controlled as the yields increased with increasing electron-density on
the aryl moiety. The results were thus correlated with the redox
properties of the photocatalysts, deriving catalyst structure-
performance relationships. In addition, compared to traditional batch
chemistry, the use of a flow protocol led to quicker reactions (30 min
instead of 12-72 h) and ensured more predictable reaction scale-ups.
transmetallation in the presence of ZnCl2, and the
subsequent Negishi coupling in the presence of Pd(OAc)2
(
Figure 1).12 One of the major concerns with this long, three-
step synthesis is the use of organolithium reagents which are
corrosive, flammable, and in certain cases pyrophoric.
The development of more viable catalytic methodologies for
the manufacture of these important pharmacophores has
long remained an important challenge in chemistry, until
MacMillan13,14 and Molander15,16 independently developed
batch photocatalytic methods to construct C(sp2)-C(sp3)
bonds under relatively-mild decarboxylative conditions, using
Ir[dF(CF3)ppy]2(dtbbpy)PF6
(dF(CF3)ppy
=
2-(2,4-
difluorophenyl)-5-(trifluoromethyl)pyridine; dtbbpy = 4,4′-di-
tert-butyl-2,2′-dipyridyl). These reactions are typically
performed at room temperature and under inert atmosphere,
upon irradiation of the reaction mixture with a 34 W blue LED
for more than 12 h (and up to three days). The scalability of
this method is evidently limited by the long reaction time, the
need for a high amount of expensive transition-metal
catalysts (ca. 10 mg of catalyst per 100 mg of product),
transport phenomena limitations within the batch reactor,
and the attenuation effect due to the transfer of photons
(Bourgeur-Lambert-Beer law).17,18 To overcome these
issues, photoredox catalysis can be combined with
continuous-flow microreactor technology, ensuring a uniform
irradiation within the narrow reaction channels, more
predictable reaction scale-ups, decreased safety hazards,
less waste generation, improved reproducibility, and the
possibility of a machine-assisted 24/7 working regime.19-22
Introduction
Visible-light photoredox catalysis has re-emerged as a
powerful tool for the synthesis of pharmaceutical,
agrochemical, and fine chemical intermediates.1-5 This class
of catalysis focuses on the use of transition metals (i.e.,
ruthenium, iridium, or nickel polypyridyl complexes)6-10
which, upon excitation with visible light, activate common
functional groups present in organic compounds. Much of the
renewed interest in this catalytic methodology is related to
the possibility of accessing compounds that are difficult to
obtain via other chemical routes. For example, -mono- and
di-substituted benzylamines are central components of a
wide range of pharmaceutically-relevant compounds.11 The
method employed for the synthesis of these privileged
scaffolds often involves the deprotonation of
a tert-
butyloxycarbonyl (Boc) protected pyrrolidine in the presence
of an organolithium reagent (typically, n-butyllithium), the
[a]
Dr. G. Vilé, A. Lhuillery, Dr. S. Richard, Dr. G. Rueedi
Idorsia Pharmaceuticals Ltd
Chemistry Technologies and Lead Discovery
Department of Drug Discovery Chemistry
Hegenheimermattweg 91, CH-4123 Allschwil, Switzerland.
*Corresponding author. E-mail: gianvito.vile@idorsia.com
Supporting information for this article is given via a link at
the end of the document
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
A first example of flow photocatalysis was
reported by Alcázar and co-workers23 who
developed an approach to couple Boc-protected
prolines with aryl bromides. The reaction
conditions, however, were neither optimized nor
generalized to the synthesis of a broad variety of
Figure 1. Summary of the strategies for the decarboxylative arylation of
-amino acids.
Boc-protected benzylamines. Besides, no insights into
catalyst structure-properties-performance relations were
identified to rationalize the effect of the ligand with the
electrophile and substrate functionalities. Determining the
catalyst, ligand, base, and reaction parameters which are
deemed as ‘optimal’ for a given transformation is however of
key importance in order to maximize the reaction rate and
minimize the content of transition-metal catalyst. And this is
particularly relevant considering the large number of
photoredox catalysts (nine), ligands (four), and bases (eight)
that are commercially available for this reaction (Figure 2).
Taking inspiration from the pioneering developments of
MacMillan13,14 and Alcázar,23 we rationalize for the first time
catalyst, ligand, and base effects and we demonstrate that
by controlling the redox
properties of the iridium-
based
photocatalysts, it is possible
to develop mild and
homogeneous
sustainable photocatalytic
decarboxylative processes
in continuous-flow mode.
Notable features of our
protocol are its broad
substrate scope, including
both
-mono and di-
substituted benzylamines,
its operational simplicity, the
low catalyst loading, short
residence time, and usage
of similar equivalents of
reagents.
Results and
Discussion
Catalytic materials. The
continuous-flow
decarboxylative arylation of

-amino acids has been
investigated using
Figure 2. Typical chemical space for the decarboxylative arylation of
-amino acids via
photoredox catalysis in flow. The figure shows the variety of potential catalysts, ligands, and
bases that are commercially available and could be employed in the reaction.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FTUabLleL1.PAPER
Compositional characterization of the iridium-based homogeneous catalysts employed in this study.
Ir[1]
C[2]
H[2]
F[2]
N[2]
P[2]
(wt.%)
Code
Name
(wt.%)
(wt.%)
(wt.%)
(wt.%)
(wt.%)
cat1
[Ir(dtbbpy)(ppy)2]PF6
20
17
23
20
20
26
20
19
20
53
45
52
58
62
56
54
50
54
4
3
2
5
5
3
5
4
5
12
27
15
12
6
6
5
6
5
5
6
6
6
6
3
3
0
0
0
0
3
3
3
cat2
cat3
cat4
cat5
cat6
cat7
cat8
cat9
Ir[dF(CF3)ppy]2(dtbbpy)PF6
Ir(dFppy)3
Ir[dF(tBu)ppy]3
Ir[p-F(tBu)ppy]3
Ir(p-F-ppy)3
8
Ir(dmppy)2(dtbbpy)PF6
Ir[dF(CH3)ppy]2(dtbbpy)PF6
12
19
12
Ir[dF(CF3)ppy]2(bpy)PF6
[1]
[2]
Elemental analysis.
Nominal values. dtbbpy = 4,4′-di-tert-butyl-2,2′-dipyridyl, ppy = 2-phenylpyridine, dF = difluoro,
tBu = tert-butyl, dm = dimethyl, p-F = para-fluoro, bpy = 2,2'-bipyridine.
commercial homogeneous catalysts where the ligand
coordinated to the active metal (i.e., Ir) has been
systematically varied introducing fluorine and other lipophilic
moieties on the ligand backbone. The use of commercial
catalysts, in particular, is justified by the need to develop a
solid photocatalytic process which makes use of catalysts
available ‘on shelf’. The catalysts in use have been
characterized in depth to determine compositional and
structural properties of the materials. Chemical analyses by
inductively coupled plasma optical emission spectroscopy
(ICP-OES) corroborate the high purity of the samples (Table
1H, 13C, and 19F nuclear magnetic resonance (NMR)
spectroscopy confirms the molecular structures of the
materials.
Reaction condition optimization. In order to identify the
optimal reaction conditions, we have considered the
synthesis
of
tert-butyl
2-(4-
from
(methoxycarbonyl)phenyl)pyrrolidine-1-carboxylate
methyl 4-bromobenzoate and (tert-butoxycarbonyl)proline as
a model reaction (Figure 3). To this aim, we have employed
one of the commercial kits recently released by
HepatoChem.24 This kit contains several vials with pre-
weighted combinations of Ir-based homogeneous catalysts
and organic or inorganic bases (Figure 3a), and is an
efficient and fast tool for the screening of catalysts, ligands,
and bases. The reactions, in particular, were performed in
1
), which are based on iridium and contain no trace of any
other transition metal. The loading of iridium is consistent
with the nominal value in the commercial samples. Besides,
Figure 3. Screening kit and batch photocatalytic setup used for
the identification of suitable reaction conditions (a). The system
was irradiated with visible light
( = 430 nm) at room
temperature for 16 h. Decarboxylative batch arylation of 4-
bromobenzoate and (tert-butoxycarbonyl)proline to tert-butyl
2-(4-(methoxycarbonyl)phenyl)pyrrolidine-1-carboxylate (b).
The experimental conditions are reported in the experimental
section. The green ticks refer to combination of catalyst and
base leading to >70% conversion of the starting material and
>80% selectivity to the desired product.
Figure 4. Schematic view of the continuous-flow microreactor
used in this work. The reactor consists of peristaltic pumps, a
power supplier for visible light irradiation, LEDs, and a coil tube
of 10 mL where the reaction takes place.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
Figure 5. Influence of residence time (a), temperature (b), and photoelectron power (c) on the conversion and product selectivity during
decarboxylative arylation of 4-bromobenzoate and (tert-butoxycarbonyl)proline to produce tert-butyl 2-(4-
(methoxycarbonyl)phenyl)pyrrolidine-1-carboxylate. The reaction scheme is indicated in Figure 3b. Conditions: T = 60°C and  = 100% (a),
 = 30 min and  = 100% (a), and T = 60°C and  = 30 min (c).
DMA as solvent, since this was reported to be the most
effective solvent for photoredox Ir/Ni dual catalysis.23 An
aliquot of the stock solution containing the reagents was
introduced in each vial; the kit was then irradiated with visible
the better photon transfer within its small tubes. Besides, the
reaction kinetics is controlled efficiently in flow mode,
circumventing competing side reactions to happen.13,14
As it turned out, the temperature has a lower impact on both
the conversion and product selectivity, since only minor
changes are detected in the product feed by varying this
parameter between 30 and 60°C (Figure 5b). Nevertheless,
the relatively mild reaction temperatures, compared to those
associated with the use of heterogeneous catalysts in flow
photochemistry,18 enable to favorably convert highly-
functionalized organic building blocks in a selective manner,
inhibiting the decomposition of these highly-sensitive fine
chemicals. As a result, this appears as one of the major
advantages for using transition-metal homogeneous
photocatalysts. Figure 5c reports the influence of the
photoelectron power on the conversion and product
light (
 = 430 nm) at room temperature for 16 h. The results
(Figure 3b) demonstrate that it is possible to reach >70%
conversion of the starting material and >80% selectivity to
the desired product in a variety of cases (cat1 or cat2 with
Cs2CO3, CsF or DBU, and cat8 or cat9 with Cs2CO3 or CsF).
Specifically, the combination of cat1 and DBU provides the
highest conversion level (100%). This is in line with the work
of Alcázar et al.23 and differs from the results published by
MacMillan et al.,13 where the suboptimal combination of cat2
and CsF was proposed. The latter would provide an
incomplete conversion (ca. 70-75%).
These conditions have been extrapolated in flow mode,
using the system shown in Figure 4. It consists of peristaltic
pumps to feed the reagents, a visible-light power supplier
selectivity; the conversion increases from 64% at

= 50% to
= 100%, and the product selectivity passes from
= 50% to 67% at = 100%. Based on these results,
= 30-50 min, T = 60°C, and = 100% as
68% at
38% at


(which can be tuned between 75 W at
power and 150 W at = 100% photoelectron power), visible-
light emitting diodes (LEDs), and a coil tube of 10 mL internal
volume, which is irradiated at = 430 nm, and where the

= 50% photoelectron


we have selected


optimal reaction conditions for the substrate scope
evaluation.

reaction takes place. Focusing on the use of cat1 and DBU
as catalyst and non-nucleophilic base, respectively, Figure
Role of the electrophile and substrate functionality on
the catalytic activity. The catalytic system and
experimental conditions identified above have been applied
5
shows the influence of residence time
, temperature T,
and photoelectron power , on the conversion of (tert-

butoxycarbonyl)proline and selectivity to the benzylamine
product. The conversion increases with increasing residence
to the synthesis of several -mono and di-substituted
benzylamines, starting from aryl halides containing a variety
time (Figure 5a), from 39% at

= 5 min to ca. 70% at

= 30
of substituents. As shown in Figure 6, we have prepared tert-
min. Above = 30 min, the conversion remains almost

butyl
2-(4-(methoxycarbonyl)phenyl)pyrrolidine-1-
6a 39% yield), tert-butyl 2-(4-
constant. The product selectivity, on the other hand, first
increases with the residence time, reaching a maximum (60-
70% selectivity) at a residence time of 15-30 min; and then
stays constant. Compared to the batch protocols reported in
the literature,13,14 where the reactions take several hours
(and up to three days) to complete, it is obvious that the use
of a continuous microreactor improves the throughput due to
carboxylate
(
,
chlorophenyl)pyrrolidine-1-carboxylate (6b, 44% yield), tert-
butyl 2-(1H-indol-5-yl)pyrrolidine-1-carboxylate (6c, 26%
yield), tert-butyl 2-(4-(trifluoromethyl)phenyl)pyrrolidine-1-
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
Figure 6. Decarboxylative arylation of -amino acids in the presence of functionalized aryl bromides.
carboxylate
(
6d
,
26%
yield),
tert-butyl
2-(4-
butyl
(3R)-2-(4-(tert-butyl)phenyl)-3-hydroxypyrrolidine-1-
phenoxyphenyl)pyrrolidine-1-carboxylate (6e, 62% yield),
carboxylate (6r, 19% yield), tert-butyl 2-(4-(tert-butyl)phenyl)-
4-methylenepyrrolidine-1-carboxylate (6s, 18% yield), tert-
tert-butyl
2-(3-(1,2,4-oxadiazol-5-yl)phenyl)pyrrolidine-1-
6f 93% yield), tert-butyl 2-(2-
carboxylate
(
,
butyl
azabicyclo[3.1.0]hexane-3-carboxylate (6t, 32% yield), tert-
butyl 2-(4-(tert-butyl)phenyl)azetidine-1-carboxylate 6u
53% yield), tert-butyl 2-(4-(tert-butyl)phenyl)piperidine-1-
carboxylate 6v 35% yield), tert-butyl 2-(4-(tert-
butyl)phenyl)azepane-1-carboxylate (6w, 41% yield), tert-
butyl 1-(4-(tert-butyl)phenyl)-2-azaspiro[3.4]octane-2-
(1R,5R)-2-(4-(tert-butyl)phenyl)-3-
isopropylphenyl)pyrrolidine-1-carboxylate (6g, 3% yield),
tert-butyl 2-(4-tert-butylphenyl)pyrrolidine-1-carboxylate (6h
76% yield), tert-butyl 2-(3-propionyl-phenyl)pyrrolidine-1-
carboxylate 6i 79% yield), tert-butyl 2-(3-
,
(
,
(
,
(
,
isopropylphenyl)pyrolidine-1-carboxylate (6j, 53% yield),
tert-butyl 2-(4-cyclopropylphenyl)pyrrolidine-1-carboxylate
(6k, 66% yield), tert-butyl 2-(4-acetylphenyl)pyrrolidine-1-
carboxylate (6l, 68% yield), tert-butyl 2-(4'-(tert-butyl)-[1,1'-
carboxylate (6x, 31% yield).
Surprisingly, when the aryl halide is functionalized with
hydroxy and alkoxy groups in meta position and an open-
chain amino acid is used, the conditions describing above
(involving the use of cat1) are ineffective. This result does
not change by varying the temperature, residence time, or
photoelectron power. Thus, we have employed again one of
the kits released by HepatoChem,24 investigating the
influence of the catalyst, ligand, and base in the reaction. We
have discovered that the latter proceeds smoothly only in the
presence of cat2 and K3PO4 as a base (Figure S2). These
biphenyl]-4-yl)pyrrolidine-1-carboxylate
(6m, 74% yield),
tert-butyl 2-(2,2-difluorobenzo[d][1,3]dioxol-5-yl)pyrrolidine-
1-carboxylate (6n, 55% yield), tert-butyl 2-(4-(pyridin-2-
yl)phenyl)pyrrolidine-1-carboxylate (6o, 77% yield), tert-butyl
2-(4-(pyridin-3-yl)phenyl)pyrrolidine-1-carboxylate (6p, 75%
yield).
The conditions described above can eventually be applied to
other aminoacids. For example, method 1 has been utilized
to prepare tert-butyl (1R,5R)-3-(4-(tert-butyl)phenyl)-2-
azabicyclo[3.1.0]hexane-2-carboxylate (6q, 48% yield), tert-
conditions (method 2, Figure 7) have been utilized for the
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
Figure 7. Decarboxylative arylation of -amino acids in the presence of aryl bromides with m-hydroxyl and m-alkoxy (a) and nitriles (b).
flow
synthesis
of
tert-butyl
(1-(3-
within a few minutes (30-50 min), which represents a
significant improvement compared to the batch methods.8,9
hydroxyphenyl)ethyl)carbamate (7a, 21% yield), tert-butyl (1-
(3-methoxyphenyl)ethyl)carbamate (7b, 10% yield), tert-
butyl (1-(3-ethoxyphe-nyl)ethyl)carbamate (7c, 32% yield),
and tert-butyl (1-(3-(neopentyloxy)phenyl)ethyl)carbamate
Molecular and mechanistic understanding. The substrate
scope study suggests that the photocatalytic activity is
dictated by the specific type of iridium photocatalyst and base
used. Thus, depending on the reagents, a particular
photocatalyst (and a specific base) is recommended. The
(
7d, 29% yield). Similarly, when the benzylic moiety is
functionalized with a nitrile, the conditions reported in
methods 1 and are unsuccessful. Retrosynthetically, the
most efficient approach to prepare -mono- and di-
2

photocatalysts are all similar in structure: cat1, cat2, cat7,
substituted benzylamines is by starting from nitriles rather
cat8, and cat9 are constituted by a complex carrying a 1+
than aryl halides, as reported elsewhere.9 Using cat5 and
net positive charge, which is balanced by a negative counter-
–
CsF (method 3
,
Figure 7), it is possible to form the tert-butyl
ion, specifically PF6
(
Figure 2). It has been reported that this
(2-(4-cyanophenyl)propan-2-yl)carbamate adduct with 89%
ionic space charge creates a high electric field which
enhances the electronic charge injection into the transition
metal complex.25-30 The latter contains two cyclometalating
ligands (ppy: 2-phenylpyridine) to coordinate the iridium
metal center and increase the ligand field splitting energy.
The other ligand (dtbbpy: 4,4’-di-tert-butyl-2,2’-dipyridyl)
ensures redox reversibility and decreases self-quenching
yield (7f). The reaction has been generalized to the synthesis
of tert-butyl (2-(2-cyanophenyl)propan-2-yl)carbamate (7e
51% yield), tert-butyl 2-(4-cyanophenyl)pyrrolidine-1-
carboxylate 7g 56% yield), and tert-butyl (1-(4-
,
(
,
cyanophenyl)cyclopentyl)carbamate (7h, 38% yield). The
method can be applied as well to the arylation of
heterocycles. For example, the synthesis of tert-butyl 2-
(pyridin-4-yl)pyrrolidine-1-carboxylate from isonicotinonitrile
and (tert-butoxycarbonyl)proline results in 58% yield.
Similarly, the production of tert-butyl (2-(pyridin-4-yl)propan-
2-yl)carbamate gives 41% yield. In all cases, the use of a
flow microreactor results in the completion of the reaction
(deactivation) of the catalyst (Figure 2). Instead, cat3, cat4,
cat5, and cat6 are constituted by three ppy cyclometalating
ligands coordinating the iridium metal (Figure 2). This core
structure, made of Ir, ppy, and eventually dtbbpy, is
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
complemented by the presence of several -F, -CH3, -tBu, and
-CF3 substituents (in different ratio and speciation). It is the
tailoring of this iridium complex that appears to be a powerful
tool to control the activity and selectivity of the homogeneous
photocatalysts.
To rationalize the need for three different catalytic methods,
we have considered the reaction mechanisms proposed in
the iridium-catalyzed arylation process. It has been
suggested13,14 that the initial irradiation of the heteroleptic IrIII
photocatalyst (8a
,
Figure 8) produces the photoexcited *IrIII
8b. Deprotonation of the

-amino acid 8c in the presence of
an appropriate base and subsequent oxidation by the *IrIII
complex generate an instable carboxyl radical which
undergoes rapid decarboxylation to form an
-amino radical
(
8d) and the *IrII species (8e). At this stage, MacMillan et al.
propose that the oxidative addition of the Ni0 (8f) with an aryl
halide (8g) produces the NiII intermediate (8h),13,14 which
would intercept the
-amino radical 8d. Subsequent
reductive elimination leads to the required C-C bond
formation delivering the α-amino arylation product 8i and
releasing the NiI intermediate 8j. Finally, the reduction of the
nickel complex and oxidation of the iridium species close the
two catalytic cycles simultaneously. This hypothesis has
been lately questioned by Oderinde et al.,31,32 who reported
experimental and theoretical evidences showing that a
NiI/NiII pathway is highly possible in Ir/Ni-dual catalysis.
For electron-deficient cyanoarenes, on the other hand, it has
been observed that upon formation of the photoexcited *IrIII
Figure 8. Proposed mechanisms for the decarboxylative arylation of
-amino acids in the presence of aryl bromides (top) and nitriles
(bottom). The mechanisms have been adapted from refs. 13,14,32.
species
(8j) under photocatalytic conditions, a facile
reduction of cyanobenzene species 8m takes place,
generating the corresponding aryl radical anion (8n), and
producing a photoexcited *IrIV species (8o). At this stage,
oxidation of the -
amino acid derivative (8p) by IrIV produces
the corresponding carboxyl radical species (8q) which
intercepts the aryl radical anion 8n, thereby forming the
product 8r after elimination of the cyanide.
reduction step. As exemplified by the redox potential listed in
Table 2, the addition of substituents to the cyclometalating
ligand has a strong effect on the ligand-field stabilization
energy of the iridium(III) complexes. Strong electron-
withdrawing substituents pull the electron density away from
the metal center, thereby stabilizing the metal d-orbitals and
The two mechanisms indicate that the reaction follows
different paths in the presence of aryl bromides (methods 1
rendering the catalyst prone to reduction. Notably, cat1
,
and
2
) and nitriles (method 3). In the first case, the iridium is
used in method 1, where iridium(III) is activated and reduced
reduced to *IrII, requiring a catalyst that facilitates this
to iridium(II), is the material with the lowest E1/2 (-1.51 V)
III/II
Table 2. Redox properties of the iridium catalysts investigated.9,20-25
III/II
*III/II
IV/*III
E1/2
(V vs SCE)
E1/2
E1/2
Code
Complex
(V vs SCE)
(V vs SCE)
cat1
cat2
cat3
cat4
cat5
cat6
cat7
cat8
cat9
[Ir(dtbbpy)(ppy)2]+
Ir[dF(CF3)ppy]2(dtbbpy)+
Ir(dFppy)3
-1.51
-1.37
n.a.[1]
n.a.
+0.66
-0.96
+0.89
n.a.
-1.21
-1.44
-1.48
-1.67
-1.60
-0.94
-0.92
n.d.
Ir[dF(tBu)-ppy]3
n.a.
Ir[p-F(tBu)-ppy]3
n.a.
n.a.
Ir(p-F-ppy)3
n.a.
n.a.
Ir(dmppy)2(dtbbpy)+
Ir[dF(CH3)ppy]2(dtbbpy)+
Ir[dF(CF3)ppy]2(bpy)+
-1.50
-1.44
-1.26
+0.77
+0.97
+1.70
[1] Not available.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
*III/II
and lowest E1/2
(+0.66 V). The other catalysts are less
improved reproducibility, and automatized operation. The
efficient, requiring higher E1/2III/II and E1/2*III/II energies. This is
in line with the experimental results in the previous section.
A similar analysis can be derived based on the electronic
properties induced by the substituents on the brominated
arenes. With substituents having acceptor (-CO2H, -
CO2R, -CHO, -COR, -CF3, -CCl3), weak donor (-CH3, -R),
reactions
were
optimized
using
high-throughput
methodologies and generalized to a variety of starting
materials, demonstrating a wide functional group tolerance.
The reactivity differences among the substrates were
correlated with the steps of the reaction mechanism and with
the redox properties of the catalysts, deriving key catalyst
structure-performance relationships that can drive the design
of better and improved catalysts. We expect that the flow
photocatalytic methods described herein will find widespread
applications in academic and industrial labs for the synthesis
of pharmaceutical and agrochemical intermediates.
and acceptor and weak donor (-F, -Cl, -Br, -I)
III/II
properties, a photocatalyst with low E1/2 properties (cat1
)
and a non-nucleophilic base (e.g., DBU, pKa,DMSO = 12) is
required. With open amino acids and strong donors (-OH,
-OR in meta position), a milder catalyst (cat2) and a stronger
base (K3PO4 or Cs2CO3, pKa,DMSO = 12.6) are recommended.
For reactions involving nitriles, instead, an oxidizing
photocatalyst is needed to complete the catalytic cycle, since
the iridium is being oxidized from IrIII to *IrIV (Figure 8). The
highest photocatalytic activity in terms of oxidation potential
Experimental Section
General Procedure. Liquid chromatography-mass spectrometry (LC-MS)
analyses were performed with an analytical Agilent G4220A pump coupled
with Thermo MSQ Plus mass spectrometer (ionization: ESI+), Dionex
DAD-3000RS, evaporative light scattering detector (ELSD) Sedex 90,
using the Water (2.1 mm×50 mm, 2.5 µm) column from Agilent
Technologies. 1H nuclear magnetic resonance spectra were recorded at
room temperature on a Bruker NMR 500 MHz spectrometer equipped with
a DCH cryoprobe. Chemical shift () values are reported in parts per million
(ppm) downfield using the residual solvent signals as internal reference.
The multiplicity is described as singlet (s), doublet (d), and multiplet (m).
is obtained with cat5, which exhibits the lowest transition
IV/*III
energy for oxidation, E1/2
(-1.67 V), in line with the
experimental results. In all cases, the position of the leaving
group on the aromatic moiety of the brominated arene can
play a key role in the reaction: while para- and meta-positions
are tolerated, substituents in ortho-position could result in an
activity drop, likely due to steric hindrance effects. This is well
visible in Figure 6, where the yield of purified 6g is much
lower than that of 6j
.
It is important to highlight that the compounds depicted in
Figure 6 and Figure 7 are all formed as racemic mixtures.
This does not change in the presence of chiral ligands for the
Photocatalyst, base, and ligand screening (batch mode). In order to
investigate the performance of the different catalysts and bases, the
HepatoChem photochemical kits were used. This enables to conveniently
screen multiple reaction conditions simultaneously using pre-weighed
catalysts and reagents. Prior to the batch reaction, DMA was degassed in
N2. The aryl derivative (0.2 mmol) and the carboxylic acid (0.3 mmol) were
then dissolved in DMA (2 mL). A fraction of this stock solution (0.1 mL)
was transferred to each reaction vial containing the catalyst and the
reagents. The kit was then irradiated under visible light conditions ( = 430
nm) at 25°C for 16 h. Afterwards, an aliquot of the reaction product was
analyzed by liquid chromatography. Conversions were calculated with
LC/MS analyses and based on the following formula: Area products / (Area
starting material still present + Area products).
Ni-catalyzed cross coupling cycle.25 As shown in Figure S3
,
in fact, racemic products are formed in the presence of chiral
ligands as well. This is correlated with the mechanism in
Figure 8. After decarboxylation of 8c to 8d and CO2
extrusion, the stereochemical information in fragment 8d is
lost, resulting in a racemic mixture of products.
Scale-up synthesis. To finally demonstrate the scalability of
the flow method developed, the synthesis of compound 6h
was scaled up, producing 3.7 g of product within 8 h, with a
1
purified yield of 77% (100% pure by LC-MS and H-NMR).
Photocatalytic reactions (flow mode). Prior to the reaction, DMA was
degassed in N2. Typically, the aryl derivative (0.2 mmol), the carboxylic
acid (0.3 mmol), NiCl2.glyme (0.02 mmol), 4,4'-Di-tert-butyl-2,2'-dipyridyl
(0.03 mmol), and cat1 or cat2 (0.002 mmol) were dissolved in anhydrous
DMA (2 mL). Afterwards, DBU or K3PO4 (0.3 mmol) was added. For
decarboxylative arylations involving nitriles, the nitrile aryl substrate (0.2
mmol), the carboxylic acid (0.6 mmol), cesium fluoride (0.6 mmol), and
cat5 (0.004 mmol) were dissolved in anhydrous DMA (10 mL) and
deionized water (0.2 mL). In all cases, the resulting solution was stirred at
room temperature until complete dissolution of the solids, and degassed
with N2 for 20 min. The reaction was conducted on the reactor shown in
Figure 3. The solution, in particular, was injected in the photoreactor (10
mL internal volume) at T = 30-60°C, P = 1 bar, and F(mixture) = 0.2-2
mL/min, corresponding to a residence time,  of 5-50 min. Isolation of the
product and recovery of the catalyst are more complicated in
This further confirms the benefits of using a continuous flow
microreactor to have predictable reaction scale-ups and the
possibility of a machine-assisted 24/7 working regime.
Conclusions
We have developed one-pot continuous processes to
construct photocatalytically a wide variety of
substituted benzylamines from cheap and readily available
-amino acids. Compared to traditional batch chemistry, the
-mono and di-

use of a flow photoreactor enables to overcome typical
photochemistry issues, such as non-uniform irradiation,
leading to quicker reactions (30 min instead of 12-72 h),
more predictable scale-ups, decreased safety hazards,
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
[12] G. Barker, J. L. McGrath, A. Klapars, D. Stead, G. Zhou, K. R.
Campos, P. O’Brien, J. Org. Chem. 2011, 76, 5936.
[13] Z. Zuo, D. T. Ahneman, L. Chu, J. A. Terrett, A. G. Doyle, D.
W. C. MacMillan, Science 2014, 345, 437.
[14] Z. Zuo, D. W. C. MacMillan, J. Am. Chem. Soc. 2014, 136,
5257.
[15] J. C. Tellis, D. N. Primer, G. A. Molander, Science 2014, 345,
433.
[16] M. Jouffroy, D. N. Primer, G. A. Molander, J. Am. Chem.
Soc. 2016, 138, 475.
[17] T. van Gerven, G. Mul, J. Moulijn, A. Stankiewicz, Chem.
Eng. Process. 2007, 46, 781.
[18] K. Han, Y.-C. Lin, C.-M. Yang, R. Jong, G. Mul, B. Mei,
ChemSusChem 2007, 10, 4510.
homogeneous than in heterogeneous catalysts. Thus, upon conclusion of
the reaction, the collected mixture, was diluted with saturated aqueous
NaHCO3 solution (5 mL) and water (15 mL). Ethyl acetate (15 mL) was
added and the two layers formed were separated. The aqueous layer was
extracted three times with ethyl acetate. The combined organic layers were
washed with saturated aqueous NaCl solution, dried over MgSO4, filtered,
and concentrated in vacuum. The residue was purified on a Combiflash
(12 g SiO2 column, 30 mL/min, 16 mL fractions, product added on 2 g of
isolute, Heptane to Heptane + 30% EtOAc). The fractions containing the
products were concentrated in vacuum and dried under high vacuum to
yield the desired compounds. A detailed description of the conditions
applied in method 1, 2, and 3 is reported in the Supporting Information.
[19] G. Jas, A. Kirschning, Chem. Eur. J. 2003, 9, 5708.
[20] M. Bu, C. Cai, F. Gallou, B. H. Lipshutz, Green Chem. 2018
20, 1233.
,
Acknowledgements
[21] C. Wiles, P. Watts, Eur. J. Org. Chem. 2008, 14, 1655.
[22] F. Levesque, P. H. Seeberger, Org. Lett. 2011, 13, 5008.
[23] I. Abdiaj, J. Alcázar, Bioorganic Med. Chem. 2017, 25, 6190.
[24] The kits are commercialized by HepatoChem Inc. (USA)
with the name EvoluChem™. See www.hepatochem.com/
[25] Z. Zuo, H. Cong, W. Li, J. Choi, G. C. Fu, D. W. C. MacMillan,
J. Am. Chem. Soc. 2016, 138, 1832.
The authors sincerely thank Dr. Thomas Weller for the continuous
support.
Keywords: photochemistry • homogeneous catalysis •
arylations • iridium • solar-driven transformations
[26] J. D. Slinker, A. A. Gorodetsky, M. S. Lowry, J. Wang, S.
Parker, R. Rohl, S. Bernhard, G. G. Malliaras, J. Am. Chem.
Soc. 2004, 126, 2763.
[1] J. Xuan, W. J. Xiao, Angew. Chem. Int. Ed. 2012, 51, 6828.
[2] C. K. Prier, D. A. Rankic, D. W. C. MacMillan, Chem. Rev.
2013, 113, 5322.
[3] S. Paria, O. Reiser, ChemCatChem 2014, 6, 2477.
[4] D. M. Schultz, T. P. Yoon, Science 2014, 343, 1239176.
[5] M. N. Hopkinson, B. Sahoo, J.-L. Li, F. Glorius, Chem. Eur. J.
2014, 20, 3874.
[6] K. Zeitler, Angew. Chem., Int. Ed. 2009, 48, 9785.
[7] J. M. R. Narayanam, C. R. J. Stephenson, Chem. Soc. Rev.
2011, 40, 102
[8] D. Rackl, P. Kreitmeier, O. Raiser, Green Chem. 2016, 18,
214.
[9] H. Huo, X. Shen. C. Wang, L. Zhang, P. Rose, L.-A. Chen, K.
Harms, M. Marsch, G. Hilt, E. Meggers, Nature 2014, 515,
100.
[10] A. J. Perkowski, D. A. Nicewitz, J. Am. Chem. Soc. 2013, 135,
10334.
[11] T. Yan, B. L. Feringa, K. Barta, ACS Catal. 2016, 6, 381.
[27] M. S. Lowry, J. I. Goldsmith, J. D. Slinker, R. Rohl, R. A.
Pascal, G. G. Malliaras, S. Bernhard, Chem. Mater. 2005
17, 5712.
,
[28] H.-C. Su, F.-C. Fang, T.-Y. Hwu, H.-H. Hsieh, H.-F. Chen, G.-
H. Lee, S.-M. Peng, K.-T. Wong, C.-C. Wu, Adv. Funct.
Mater. 2007, 17, 1019.
[29] S. Ladouceur, D. Fortin, E. Z. Colman, Inorg. Chem. 2010
49, 5625.
,
[30] A. Joshi-Pangu, F. Lévesque, H. G. Roth, S. F. Oliver, L.-C.
Campeau, D. Nicewicz, D. A. DiRocco, J. Org. Chem. 2016
81, 7244.
,
[31] M. S. Oderinde, A. Varela-Alvarez, B. Aquila, D. Q. Robbins,
J. W. Johannes, J. Org. Chem. 2015, 80, 7642.
[32] M. S. Oderinde, M. Frenette, D. W. Robbins, B. Aquila, J.
W. Johannes, J. Am. Chem. Soc. 2016, 138, 1760.
This article is protected by copyright. All rights reserved.

10.1002/cctc.201800754
ChemCatChem
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Gianvito Vilé,* Sylvia Richard, Arnaud
Lhuillery, and Georg Rueedi
Page No. – Page No.
Electrophile, Substrate Functionality,
and Catalyst Effects in the Synthesis
of -Mono and Di-Substituted
Benzylamines via Visible-Light
Photoredox Catalysis in Flow
We report three optimized catalytic methodologies to perform the decarboxylative arylation of -amino acids in continuous-flow mode.
The methods provide efficient yields in a wide substrate scope and are rationalized molecularly, depending on the redox properties of
the photocatalysts, the choice of the electrophiles, and the substrate functionalities.
This article is protected by copyright. All rights reserved.