Synthesis of Sterically Hindered Primary Amines by Concurrent Tandem Photoredox Catalysis

Article abstract of DOI:10.1021/jacs.9b10871

Primary amines are an important structural motif in active pharmaceutical ingredients (APIs) and intermediates thereof, as well as members of ligand libraries for either biological or catalytic applications. Many chemical methodologies exist for amine synthesis, but the direct synthesis of primary amines with a fully substituted α carbon center is an underdeveloped area. We report a method which utilizes photoredox catalysis to couple readily available O-benzoyl oximes with cyanoarenes to synthesize primary amines with fully substituted α-carbons. We also demonstrate that this method enables the synthesis of amines with α-trifluoromethyl functionality. Based on experimental and computational results, we propose a mechanism where the photocatalyst engages in concurrent tandem catalysis by reacting with the oxime as a triplet sensitizer in the first catalytic cycle and a reductant toward the cyanoarene in the second catalytic cycle to achieve the synthesis of hindered primary amines via heterocoupling of radicals from readily available oximes.

Full text of DOI:10.1021/jacs.9b10871

Article
pubs.acs.org/JACS
Cite This: J. Am. Chem. Soc. XXXX, XXX, XXX−XXX
Synthesis of Sterically Hindered Primary Amines by Concurrent
Tandem Photoredox Catalysis
Michael C. Nicastri,† Dan Lehnherr,*,‡ Yu-hong Lam,§ Daniel A. DiRocco,‡
and Tomislav Rovis*,†
†Department of Chemistry, Columbia University, New York, New York 10027, United States
‡Process Research and Development, Merck & Co., Inc., Rahway, New Jersey 07065, United States
§Computational and Structural Chemistry, Merck & Co., Inc., Rahway, New Jersey 07065, United States
S
* Supporting Information
ABSTRACT: Primary amines are an important structural
motif in active pharmaceutical ingredients (APIs) and
intermediates thereof, as well as members of ligand libraries
for either biological or catalytic applications. Many chemical
methodologies exist for amine synthesis, but the direct synthesis of primary amines with a fully substituted α carbon center is an
underdeveloped area. We report a method which utilizes photoredox catalysis to couple readily available O-benzoyl oximes with
cyanoarenes to synthesize primary amines with fully substituted α-carbons. We also demonstrate that this method enables the
synthesis of amines with α-trifluoromethyl functionality. Based on experimental and computational results, we propose a
mechanism where the photocatalyst engages in concurrent tandem catalysis by reacting with the oxime as a triplet sensitizer in
the first catalytic cycle and a reductant toward the cyanoarene in the second catalytic cycle to achieve the synthesis of hindered
primary amines via heterocoupling of radicals from readily available oximes.
The synthesis of amines is an area of research rich in
methodologies which generate products via C−N bond
formation, addition to CN double bonds, and, more
recently, α-C−H functionalization of amines.3 When the
synthesis of a primary amine is required, specifically one with a
fully substituted α-carbon, the number of applicable method-
ologies decreases dramatically. These methodologies can be
sorted into 2e− approaches and 1 e− approaches to amine
synthesis.
INTRODUCTION
■
Primary amines are constituent members of ligand libraries as
either final ligands or valuable synthetic intermediates.1 When
the retrosynthesis of a target molecule is designed, amines
represent key points of disconnection, thereby having a
significant impact on the choice of synthetic route. The ability
to utilize amines in the synthesis of a diverse library of final
ligands is directly affected by access to a wide array of primary
amines. General motif 1 is readily found in a number of
patented pharmaceutical agents, a small selection of which is
illustrated in Figure 1.2
The 2e− approach, pioneered by Ellman and co-workers, can
be effected via addition of nucleophiles to tert-butyl
sulfinimides (Scheme 1A).4 When this methodology is applied
to the synthesis of primary amines with fully substituted α-
carbons, the synthesis of ketone-derived sulfinimides often
requires the use of strong Lewis acids and heat. The synthesis
of amines from sulfinimides then requires an organometallic
reagent, followed by deprotection with strong acid. We
postulated that an alternative synthetic methodology based
on a single-electron disconnection could bypass the need to
utilize harsh organometallic reagents and enable the use of
starting materials which are bench stable and easily
synthesized.
The rapid adoption of photoredox catalysis has resulted in a
multitude of methodologies which effect α-amino functional-
ization (Scheme 1B).5 However, the vast majority of these
methodologies require full substitution of the nitrogen with
alkylations, arylation, or activating groups, a strategy that
Figure 1. Pharmaceutical agents containing primary amines with fully
substituted α-carbons.
Received: October 15, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
starting materials, such as oximes and iminium salts, that could
be utilized to generate α-amino radicals under mild conditions.
Herein (Scheme 1C), we report a photoredox-catalyzed
method to synthesize primary amines starting from bench-
stable, easily formed O-benzoyl oximes8 or iminium chloride
salts and a cyanoarene.9 This methodology enables rapid
access to fully substituted primary amines through easily
synthesized bench-stable precursors without the requirement
for subsequent deprotection.
Scheme 1. Approaches to the Synthesis of Sterically
Hindered Amines: (A) Prior Art of Conventional 2e−
Synthesis of Primary Amines; (B) Prior Art of Photoredox-
Catalyzed Syntheses of Primary Amines Compared to This
Work (C)
RESULTS AND DISCUSSION
■
We utilized a high-throughput experimentation (HTE)10
approach to identify conditions which would produce our
desired product from oxime 3a and cyanopyridine 4a via
photoredox catalysis (Table 1). This approach enabled rapid
identification of a number of conditions that provide the
desired product 5, with the best results being in DMSO with
diisopropylamine (DIPA) as the terminal reductant and Ir-
based photocatalysts, particularly Ir[dF(Me)ppy]2dtbbpyPF6
(PC1), which furnishes product in 71% yield.
Interestingly, the tris-phenylpyridine derivative, Ir(ppy)3
(PC4), affords only trace product, despite being the most
reducing iridium-based photocatalyst in the screen.11 The
organic photocatalyst 4CZIPN (PC8), which has redox
properties similar to those of PC3, also gives only trace
product when DMSO is used as the solvent.12 Reactions using
acetonitrile or acetone as the solvent deliver only trace
amounts of the desired product, while 1,2-dichloroethane
proved unsuccessful, except for a moderate 19% yield when
PC2 was utilized as a photocatalyst. Analysis of the reaction
profile revealed multiple pyridine-based byproducts, but no
homocoupling of the reduced 4-cyanopyridine was observed
(i.e., formation of 4,4′-dipyridyl).13 Using 4-cyanopyridine in
excess dramatically improves the reaction profile, eliminating
most pyridine-based byproducts (Supporting Information,
Table S2).
Using the oxime coupling partner as the limiting reagent and
4a in excess (2.2 equiv), we chose to fine-tune our optimized
reaction conditions. Photocatalyst loading could be decreased
to as low as 2 mol% while maintaining high yields of product
5a. Increasing the limiting substrate (oxime) concentration
from 0.1 to 0.2 M did not affect the yield at a 0.3 mmol scale;
under these conditions, however, the reaction became
heterogeneous. We anticipated that a homogeneous reaction
would be more advantageous for reaction scale-up, enabling
improved light penetration into the reaction mixture, and
therefore we elected to conduct our reactions at 0.1 M. We
hypothesized that the overall transformation would generate
benzoic acid during the course of the reaction via N−O bond
scission of the oxime and DIPA oxidation. We elected to utilize
1 equiv of benzoic acid to prevent a potential induction phase
for the reduction of 4-cyanopyridine, should this process
require a proton source.14 Empirically, we observed that, at
high concentrations of DIPA, reactions initially form a biphasic
mixture. We observed that the inclusion of benzoic acid
improved the solubility of DIPA, generating a homogeneous
solution after 30 s of mixing.
requires subsequent deprotection to yield a primary amine.
Finally, many of these methodologies do not contain examples
for synthesizing amine-bearing fully substituted carbons.
Two notable exceptions to the above statement have
recently appeared. Dixon reported an interesting strategy
toward α-amino functionalization of primary amines utilizing
quinones as in situ oxidants to generate imine intermediates,
with in situ alkylation by both two- and one-electron methods.6
Concurrently with our work, Gilmore reported three examples
of hindered amine syntheses via photoredox catalysis, utilizing
ketones in the presence of excess ammonia gas (1 atm NH3),
strong acid, and excess Hantzsch ester at elevated temper-
ature.7 Our efforts focused on developing benchtop-stable
Our initial HTE experiments revealed that DIPA was the
optimal stoichiometric reductant in DMSO when the oxime
was used as the excess coupling partner. Diisopropylethylamine
(DIPEA) is commonly used in photoredox chemistry as a
stoichiometric reductant.15 While DIPEA is easier to oxidize
than DIPA, oxidation generates a secondary α-amino radical
B
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Table 1. Reaction Discovery via High-Throughput
Experimentation
Table 2. Optimized Reaction and Control Experiments
a
b
entry
deviation from standard reaction conditions
yield (%)
1
2
3
4
5
6
none
no DIPA
no photocatalyst
DIPEA instead of DIPA
Hantzsch Ester instead of DIPA
(PMP)3N amine instead of DIPA
86
0
0
27
26
0
a
3a (0.3 mmol), 4a (2.2 equiv), benzoic acid (1.0 equiv), PC1 (2.0
mol%), amine base (3.6 equiv), DMSO (0.1 M), blue LEDs (455 nm,
b
5W per reaction), 40 °C. Assay yield determined by a UPLC
calibrated to an authentic sample of purified product.
(27%). Additionally, adducts of DIPEA and 4-cyanopyridine
were observed via UPLC-MS analysis. Reactions with
Hanstzch ester give yields similar to those obtained using
DIPEA as the terminal reductant. The use of (PMP)3N instead
of DIPA does not afford any product 5a. From this result we
inferred the importance of DIPA as both an electron source
and either a proton or hydrogen atom source for the reaction.
While the solvent can function as a hydrogen atom donor in
many radical reactions, we note that the estimated bond
dissociation energy of the α-amino C−H bonds in DIPA
should be significantly lower than those of the C−H bonds
present in DMSO. DFT calculations predict that hydrogen
atom abstraction from DMSO by the α-amino radical of DIPA
is unfavorable, based on the Gibbs free energy for the reaction
being +17.4 kcal/mol.16
Encouraged by the success of the reaction and postulating
that it proceeds via a benzylic α-amino radical, we wondered if
this radical intermediate could also be attained via the
reduction of iminium salts. The reduction potential of benzylic
ketimines in the presence of strong acid was reported to be ca.
−0.7 V vs SCE.7 Based on these results, the photocatalytic
manifold could reduce benzylic iminium salt 6a, the analog of
the oxime 3a, for coupling with cyanopyridine toward
generating product 5a (Scheme 2).
a
Assay yields determined by ultraperformance liquid chromatography
(UPLC) analysis of crude reactions (10 μmol scale) relative to an
internal standard.
Scheme 2. Reductive Coupling via Iminium Chloride Salt
which we observed to compete for coupling with 4-cyano-
pyridine. Interestingly, DIPA was never observed to cross-
couple with 4-cyanopyridine. To test the role of the reductant
in the reaction under the conditions utilizing pyridine as the
excess coupling partner, we compared DIPA, DIPEA, Hantzsch
ester, and tris-p-methoxyphenylamine ((PMP)3N) in the
reductive coupling reaction (Table 2). We observed that, in
larger scale reactions (0.3 mmol), DIPA gives product 5a in
86% yield. As expected, removing DIPA as a reductant
produces no product in the reaction. The use of the more
oxidizable tertiary amine DIPEA results in a much lower yield
As a model reaction, we subjected iminium salt 6a to a
photocatalytic reaction with our optimal photocatalyst, PC1,
and DIPA as stoichiometric reductant. With the concern that
the α-amino radical may dimerize more readily under these
conditions, we utilized the iminium salt as the excess reagent
(2.2 equiv). Gratifyingly, our first reaction under these
C
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a b
,
Table 3. Oxime and Iminium Chloride Scope for Primary Amine Synthesis
a
Method A: 0.3 mmol of oxime 3 at 0.1 M in DMSO containing 4-cyanopyridine (2.2 equiv), i-Pr2NH (3.6 equiv), benzoic acid (1.0 equiv), PC1
b
(2 mol%) illuminated by blue LEDs (455 nm, 5W per reaction) at 40 °C for 2.5 h. Method B: 4-cyanopyridine (0.3 mmol) at 0.1 M in DMSO
containing iminium salt 6 (2.2 equiv), i-Pr2NH (3.6 equiv), PC1 (2 mol%) illuminated by blue LEDs (455 nm, 5W per reaction) at 40 °C for 2.5 h.
c
Yields from Method A and Method B are denoted by values next to the letters A and B and unless noted otherwise are assay yields determined by
d
a UPLC calibrated to an authentic sample of purified product. Value in brackets are isolated yields. Reaction was run on a 3.6 mmol scale.
e
f
Reaction duration of 3.5 h. Reaction duration of 16 h.
conditions furnished the desired product in 81% yield. The
reaction profile showed clean formation of product without the
formation of significant amounts of side products.17 Therefore,
we elected to move forward with these conditions to elaborate
a complementary scope using either oximes (Method A) or
iminium chloride salts (Method B) for coupling with
cyanoarenes (Table 3).
With optimized catalytic conditions in hand, we chose to
explore the scope of the oxime coupling partner in the reaction
(Method A, Table 3). We began by testing our reaction at a 0.3
mmol scale, utilizing our model oxime 3a (1 equiv) and excess
4-cyanopyridine 4a (2.2 equiv). The reaction was irradiated for
2.5 h in a temperature-controlled (40 °C) TAK 120
photoreactor with 455 nm light (5 W per vial), enabling up
to 10 reactions to be run in parallel at a time.18 These
optimized conditions provided primary amine 5a in 86% yield.
Next, we tested the scope of oximes compatible in the
reductive coupling.
First, benzylic oximes which contain para-substituents of
varying electronic properties were tested using the optimized
reaction conditions. Gratifyingly, electronically varied products
5a−5f are obtained in excellent yield, regardless of whether the
substituent on the arene is electron-withdrawing or electron-
donating. Compound 5g, containing an ortho-methyl sub-
stituent, is formed in 32% yield, requiring extended reaction
times of 16 h for complete consumption of the oxime (vide
infra). Products that incorporate larger arenes (5h−5k) or
longer alkyl groups (5l) were successfully synthesized.
Compound 5n, containing a cyclopropyl substituent, was
synthesized in 56% yield without producing the related ring-
opened product.19 Synthesis of compound 5o provides a
desirable ester functionality in 32% yield. The analogous 2-
pyridyl-derived oxime reacts to form only trace amounts of
primary amine product.20 Other heterocyclic oximes tested did
not yield product.21
When expanding our scope, we recognized that primary
amines containing an α-trifluoromethyl group represent a
unique architecture which is not easily accessed with
traditional synthetic methodologies. Based on our initial
mechanistic hypothesis, we thought this functional group
should be well tolerated in the reaction. Therefore, we tested
the reactivity of O-benzoyl oximes derived from readily
available trifluoromethyl ketones. Compound 5p, a primary
amine with an α-trifluoromethyl group, was synthesized in 65%
yield. We note that aryl bromides are also tolerated in the
reaction, with compound 5q forming in 29% yield. Compound
5r containing a pyrazole is assembled in 57% yield, indicating
the oxime reaction tolerates heterocycles more efficiently when
they are included through a biaryl linkage to the benzylic
oxime (i.e., separating the α-amino radical from the hetero-
cycle via a phenylene linker is beneficial for high yield of the
coupling product). Trifluoromethyl derivatives of acetophe-
D
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
a b c
, ,
Table 4. Cyanoarene Scope for Primary Amine Synthesis
a
Method A: 0.3 mmol of oxime 3 at 0.1 M in DMSO containing 4-cyanopyridine (2.2 equiv), i-Pr2NH (3.6 equiv), benzoic acid (1.0 equiv), PC1
b
(2 mol%) illuminated by blue LEDs (455 nm, 5 W per reaction) at 40 °C for 2.5 h. Method B: 4-cyanopyridine (0.22 mmol) at 0.1 M in DMSO
containing iminium salt (2.2 equiv), i-Pr2NH (3.6 equiv), PC1 (2 mol%) illuminated by blue LEDs (455 nm, 5 W per reaction) at 40 °C for 2.5 h.
c
Yields from Method A and Method B are denoted by values next to the letters A and B and unless noted otherwise are assay yields determined by
a UPLC calibrated to an authentic sample of purified product.
none are widely commercially available and can easily provide
bench-stable α-trifluoromethyl oxime derivatives suitable for
generation of products 5p−5r. We believe this methodology
represents a uniquely rapid method of synthesizing α-
trifluoromethyl primary amines, substructures of high interest
to medicinal chemists as a result of their improved resistance
toward oxidative metabolic degradation and improved
membrane permeability arising from the increase in lip-
ophilicity imparted by the incorporation of fluorine atoms.22
To complement oxime reactivity (Method A), we explored
the coupling reaction of iminium chlorides (Method B). When
using electronically varied iminium salts (6a−6f), the yield is
clearly affected, in contrast to reactions with the analogous
oximes (3a−3f). Iminium salts containing electron-donating
substituents afford higher yields in the coupling reaction.
Specifically, reaction efficiency correlates directly with the
electron-rich nature of the aromatic ring, with electron-
donating p-methoxy compound 6f yielding primary amine
product 5f in 94% yield. The strongly electron-deficient
trifluoromethyl group significantly diminishes product for-
mation, resulting in a yield of 14% for compound 5b. In
contrast, the corresponding oxime chemistry is less sensitive
toward the impact of electronic substituents, converting p-
trifluoromethyl oxime (3b) into 5b in 91% yield (vide supra).
Ortho-methyl substitution improves the coupling reaction
when using iminium salts, delivering primary amine 5g in
excellent yield. In contrast, oximes with an ortho-methyl
substituent do not efficiently convert into the desired product
5g, requiring 16 h to consume the starting oxime. In both
cases, ortho-methyl substitution causes the arene ring to rotate
out of the plane of either the iminium or oxime, resulting in a
decrease in the extent of conjugation between the π-system of
the phenyl ring and the imine CN group. In the case of the
iminium substrate, the energy of the LUMO is raised,
analogous to the placement of an electron-donating group at
the para-position. Additionally, the resulting benzylic α-amino
radical is postulated to be more reactive due to decreased
resonance delocalization. In the case of the oxime substrate,
additional effects are present, related to the energy of its triplet
excited state, that may lead to decreased reactivity (vide infra).
Further, Method B produces amines with sterically demanding
groups (e.g., t-Bu) adjacent to the new C−C bond (5m) in
good yield.
Photochemical reactions are sometimes limited in their
scalability due to light penetration. To test this, we increased
E
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
the scale of our reaction of 3a and 4a by 12-fold. To
accommodate the larger reaction vessel volume, we utilized a
PennOC m1 photoreactor which can accommodate a 40 mL
glass vial.23 Gratifyingly, on this scale, 3.6 mmol of 3a gives
product 5a in 80% yield. At temperatures slightly above room
temperature, ca. 35−40 °C, we observed the reaction to be
homogeneous at a concentration of 0.1 M in oxime 3a,
allowing for greater light penetration.
Next, the cyanoarene coupling partner scope was tested with
both oxime 3j and iminium salt 6j. Heterocycles are tolerated
in the cyanoarene substrate scope in the form of 2-biaryl-4-
cyanopyridines (5s−5x) (Table 4). More electron-rich hetero-
cycles couple more efficiently with oxime (3j) coupling
partners. As a comparison, product 5y is formed in 89%
yield when the pyridine contains a 2-phenyl substituent, 5z is
produced in 97% yield with 4-methoxyphenyl, and 5aa is
generated in only 25% yield when the pyridine contains a 2-
carbomethoxyphenyl. For the analogous iminium salt (6j),
product formation follows a trend similar to that of the
respective oxime reaction (5s−5aa).
Interestingly, PC4, the most reducing iridium-based photo-
catalyst (IrII, Ered = 2.20 V vs SCE), is not the most efficient
catalyst for oxime consumption.11 PC1 has a maximum
reduction potential in the reduced state (IrII) of −1.43 V vs
SCE.25 We utilized DFT calculations to determine the single-
electron reduction potential of O-benzoyl oximes and found
that the model compound 3a has a reduction potential of
−2.03 V vs SCE (Table 5). This reduction potential indicates
that reduction of the oxime by PC1 would be a significantly
endergonic process (13.8 kcal mol−1), thus making direct
reduction unlikely.
Table 5. Triplet Energy Sensitization of O-Benzoyl Oximes
by Photoredox Catalysts
Functional groups in the 2-position of the cyanopyridine
coupling partner have a strong influence on reactivity with
both oxime (3j) and iminium (6j) coupling partners. For
reactions with 3j, products containing electron-donating
substituents (5ac−5ah) increased the efficiency of the
coupling reaction. Coupling of 3j with the respective
cyanoarene produced no product for compounds (5ai−5ak).
Coupling of iminium salt 6j with electron-rich cyanoarenes
produced results similar to those obtained for the reaction with
oxime 3j. Notably, iminium 6j couples effectively with
cyanoarenes containing electron-withdrawing groups at the 2-
position (5ai, 5aj). This complementary reactivity between
oximes and iminiums allows this catalytic manifold to access a
greater substrate scope than the oxime reaction alone.
Substituents at the 3-position are also tolerated, as witnessed
with 5am and 5ap.
Coupling of oximes and iminium salts with 2-cyanopyridine
derivatives occurs with similar reactivity to their 4-cyano-
pyridine analogs (5aq, 5ar). The reaction does not occur with
non-heterocyclic cyanoarenes (5at, 5au). The reaction was
also tested with other heterocyclic cyanoarenes, where both 3j
and 6j couple to form azaindole 5av. More electron-poor
heterocycles can be synthesized using iminium salt 6j to afford
products 5aw and 5ax.
oxime
E(T1−S0)
Ered (V) E*ox (V)
remaining
b
c
entry photocatalyst [kcal·mol−1
]
vs SCE
vs SCE
(%)
1
2
3
4
5
PC1
PC3
PC4
PC6
PC7
55.8
61.8
58.1
46
−1.42
−1.37
−2.20
−1.33
−1.06
0.97
1.21
0.31
0.77
0.83
0
51
0
98
98
43.6
a
Triplet−singlet energy gap (E[T1−S0]) refers to triplet (T1) to
singlet (S0) energy gap calculated using M06-2X/6-31+G(d,p)
b
SMD=DMSO. Catalyst triplet-state energy values are experimental
values.25,30,34a 3a (0.3 mmol), 4a (2.2 equiv), benzoic acid (1.0
c
equiv), PC1 (2.0 mol%), DMSO (0.1 M), blue LEDs (455 nm, 5 W
per reaction), 40 °C. No stoichiometric reductant included.
Based on these results, we considered that N−O bond
cleavage could occur via an alternative, nonreductive path-
way.24,26−28 To test this, we revisited our control reaction
where the terminal reductant DIPA was excluded (Table 5).
The reaction was irradiated in a temperature-controlled (40
°C) TAK 120 photoreactor with 455 nm light at a power of 5
W per vial at a reaction scale of 0.3 mmol. Under these
conditions, PC1 caused decomposition of 100% of the starting
oxime 3a and the formation of a new byproduct which was
identified as azine dimer 7 (Scheme 3). We envisioned that 7
could form via the dimerization of two iminyl radicals
generated from N−O bond scission. However, it appears
unlikely that the catalyst would reduce the oxime with high
efficiency under these conditions, because no stoichiometric
reductant is available to reduce the photocatalyst to turn over
the catalyst and complete the catalytic cycle. Furthermore, we
observe no consumption of the 4-cyanopyridine coupling
partner in reactions where DIPA is excluded. The reduction
potential of the heterocycle is dramatically more electro-
Initially, we hypothesized that O-benzoyl oximes could be
reductively deprotected in situ to generate low concentrations
of the corresponding N−H imine in solution. Previously, Zard
and co-workers described single-electron reductive methods
for iminyl radical generation from O-benzoyl oximes.24 They
proposed, using SmI2, single-electron reduction of the benzoyl
moiety, which then undergoes beta-scission to release an
iminyl radical and benzoate. While our HTE screen revealed
reaction conditions which promoted the desired reductive
coupling, our data did not support direct reduction of the O-
benzoyl oxime. Specifically, the data reveals that the magnitude
of the reduction potential of a catalyst does not correlate with
oxime consumption (cf. Table 1). Interestingly, iridium-based
photocatalysts catalyzed decomposition of the oxime across a
wide range of catalyst reduction potentials. Ruthenium-based
and organic photoredox catalysts which possess reduction
potentials similar to those of the successful iridium-based
photocatalysts do not catalyze oxime consumption.
F
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 3. Evidence of Iminyl Radical Formation
Table 6. DFT Calculated Triplet Energy Values and
Reduction Potentials
oxime
geometry
GZ relative to GE
E(T1−S0)
E°redvs
SCE [V]
R
[kcal·mol−1
]
[kcal·mol−1
]
4-CF3
Z
E
+1.6
51.9
53.5
−1.74
−1.75
4-Cl
Z
E
+1.8
+6.6
+2.3
+1.7
−0.2
+1.6
52.2
54.0
−2.01
−2.10
4-F
Z
E
52.0
54.6
−1.81
−1.99
positive than the reduction potential of the oxime. Therefore, it
is unlikely that the O-benzoyl oxime would be preferentially
reduced in comparison to the 4-cyanopyridine. The presence
of an iminyl radical intermediate is further supported by the
observation of off-pathway byproducts in the syntheses of
general compound 5 that are unique to Method A (e.g., Table
3).29
4-H
Z
E
52.7
54.2
−2.03
−2.04
4-Me
2-Me
4-MeO
Z
E
57.8
54.5
−2.03
−2.00
Photoredox catalysts can also act as triplet sensitizers as an
alternative to engaging in redox chemistry.31 It has been
observed previously that oximes which are promoted into the
triplet state can undergo N−O bond homolysis to generate an
N-centered radical.14,32,33 Triplet energy sensitization has been
invoked in multiple reactions which utilize photoredox
catalysts to induce excited-state reactivity in an electron-
neutral process.
Z
E
75.8
57.9
−2.03
−2.02
Z
E
57.2
54.6
−2.04
−2.01
a
E(T1−S0) refers to triplet (T1) to singlet (S0) energy gap calculated
using M06-2X/6-31+G(d,p) SMD=DMSO.
We next tested this hypothesis by repeating the experiment
with multiple photocatalysts, where the oxime starting material,
4-cyanopyridine, and benzoic acid were mixed in DMSO in the
absence of DIPA. We chose to screen five catalysts which differ
in both triplet energies and reduction potentials.25,34 We then
compared the amount of O-benzoyl oxime consumption in
each reaction (0.3 mmol scale) and found again that only
iridium-based photocatalysts catalyze the decomposition of
oxime 3a (Table 5). The reduction potential of Ru(bpy)3PF6,
PC6, in the reduced state (RuI) is −1.33 V vs SCE, similar to
the reduced state iridium-based (IrII) PC1 and PC3. However,
PC6 does not catalyze oxime decomposition, while both PC1
and PC3 do under these reaction conditions. If an active redox
cycle is occurring at the catalyst, some 4-cyanopyridine (−1.60
V vs SCE) or 4-cyanopyridinium (−0.69 V vs SCE) should
have been reduced in an off-pathway reaction.35 In all cases,
minimal consumption of 4-cyanopyridine starting material is
observed. For each of the five photocatalysts screened on scale,
a triplet (T1)-to-singlet (S0) energy gap [E(T1−S0)],
commonly referred to as triplet energies, has been reported
in the literature. We next utilized DFT to calculate the triplet-
to-singlet energy gap36 for our model O-benzoyl oxime
substrate (3a). Oximes Z-3a and E-3a have a predicted
triplet−singlet energy gap of 52.7 and 54.2 kcal/mol,
respectively (Table 6). Iridium-based photocatalysts PC1,
PC3, and PC4 all possess a triplet-singlet energy gap >55 kcal/
mol, and all these photocatalysts can consume oxime 3 in the
absence of a reductant. In contrast, both PC6 and PC7 have
triplet-to-singlet energy gaps below 50 kcal/mol (46.0 and 43.6
kcal/mol, respectively) and are unable to consume oxime 3a in
the absence of a reductant. Based on the dichotomy related to
triplet-to-singlet energies and oxime consumption in these
experiments, we infer that triplet sensitization is responsible for
the decomposition of the N−O bond in O-benzoyl oximes.
A Stern−Volmer analysis of the emission of PC1* (0.01 M
in DMSO) using various concentrations of oxime 3a (0−15
mM) reveals 3a can quench the excited state of PC1 (see
Supporting Information for details), consistent with the
proposed mechanism. While the Stern−Volmer relationship
supports quenching of the excited state of PC1, it cannot be
used alone to determine if this proceeds via an energy-transfer
or electron-transfer mechanism. To discern between the two
potential mechanisms, we photolyzed (450 nm) solutions of
oxime 3a in the presence of various photocatalysts (2 mol%) in
the absence of a terminal reductant (DIPA). If oxime
consumption is observed in quantities significantly greater
than the amount of photocatalyst present, this would rule out
an electron-transfer process and support an energy-transfer
mechanism. Indeed, photocatalysts PC1−PC3 successfully
consumed oxime (see Table 5) well beyond the amount of
catalyst present in solution. This data supports an energy-
transfer process to convert oxime via N−O bond scission. Due
to the lack of spectral overlap between the UV−vis emission
spectrum of photocatalyst PC1 and the UV−vis absorption
spectrum of oxime 3a (see Supporting Information for details),
̈
we rule out a Forster energy-transfer process in favor of a
Dexter energy-transfer process.
In most photoredox reactions, the concentration of radicals
is often estimated to be approximately the concentration of
photocatalyst. While reductive couplings are commonly
achieved with stoichiometric metal reductants, intermediates
which require single-electron reduction are in much lower
concentrations relative to the species responsible for the single-
G
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Scheme 4. Mechanistic Insight into the Synthesis of Sterically Hindered Primary Amines: (A) Proposed Mechanism for the
Coupling of O-benzoyl Oximes and 4-Cyanopyridine (Method A); (B) Proposed Mechanism for the Coupling of Iminium
Chloride Salts and 4-Cyanopyridine
electron reduction. The relatively low concentration of reduced
photocatalyst present in the reaction makes the successive
reduction of radical intermediates to anions increasingly
unlikely, thus strengthening the alternate pathway involving
triplet sensitization.
concentration in IV when compared to DIPA, this pathway is
unlikely. The resultant imine IX possesses a high reduction
potential (ca. −2.4 V vs SCE). However, protonation occurs to
generate iminium salt X, which is much more easily reduced
(ca. −0.8 V vs SCE).35
We hypothesize that this reductive coupling proceeds
through a catalytic mechanism (Scheme 4) where the iridium
engages in concurrent tandem catalysis,37 acting simulta-
neously as both a photoredox catalyst and a triplet sensitizer.38
First, photocatalyst I absorbs a photon to generate its
corresponding excited singlet state (II), and subsequent
intersystem crossing provides excited-state triplet III. This
excited state III can be quenched reductively to generate a
ground state IrII (IV). IV can then reduce 4-cyanopyridine via
a proton-coupled electron-transfer (PCET) mechanism14 to
generate persistent radical V and a ground state I.35 Excitation
of I can again generate excited state III via photon absorption
and intersystem crossing, where relaxation occurs via Dexter
energy transfer39 between III and O-benzoyl oxime 3a. Triplet-
state oxime VI can undergo N−O bond homolysis, resulting in
iminyl radical VII. DFT calculations reveal that a hydrogen
atom transfer event between iminyl radical VII and DIPA is
exothermic (Scheme 4, inset), making this a likely pathway for
the formation of imine IX. Alternatively, iminyl radical VII
could be reduced by IV; however, due to the large disparity in
Iminium X can be reduced via two likely reaction
intermediates, illustrated in Scheme 4. In path A, iminium X
can be reduced via the α-amino radical of DIPA, VIII.40
Radical VIII is generated either by the reductive quenching of
excited state photocatalyst III, followed by deprotonation, or
via hydrogen atom abstraction by iminyl radical VII on DIPA.
DFT calculations reveal that electron transfer between α-
amino radical of DIPA (VIII) and iminium X is substantially
exothermic, with a ΔGrxn of −7.4 kcal/mol (Scheme 4, inset),
generating α-amino radical XII and hindered iminium XI. In
path B (Scheme 4) iminium X is reduced by the reduced-state
photocatalyst IV, regenerating catalyst I and forming radical
XII.
Penultimate intermediate XIII can be formed via a
heterocoupling between persistent radical V and radical XII.
Intermediate XIII can formally lose both a proton and a
cyanide to give the reaction product 5a.
We hypothesize that the analogous photocatalyzed reaction
from iminium salts (e.g., 6) occurs via a similar mechanistic
pathway to that of the oximes, including the reduction of 4a via
H
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
(2) (a) Janssen Pharmaceutica NV. Farnesyl protein transferase
inhibiting (imidazole-5-yl) methyl-2-quinolinone derivatives, their
preparation, pharmaceutical compositions containing them and their
preparation, and medicaments comprising them. Patent WO 9721701
A1, June 19, 1997. (b) Zhang, Y.; Hodous, B. L.; Kim, J. L.; Wilson,
K.; Wilson, D. Compositions useful for treating disorders related to
kit. Patent WO 2015057873 A1, October 15, 2014. (c) Kotian, P. L.;
Babu, Y. S.; Kumar, V. S.; Zhang, W.; Vogeti, L. Preparation of
pyrrolidinecarboxamide derivatives as human plasma kallikrein
inhibitors. Patent WO 2017059178 A1, April 6, 2017. (d) Adams,
C. M.; Babu, C.; Capparelli, M. P.; Ding, J.; Ehara, T.; Jendza, K.; Ji,
N.; Karki, R. G.; Kawanami, T.; Mainolfi, N.; Powers, J. J.; Serrano-
Wu, M. H.; Xue, L.; Zhang, C. 2-(1h-indol-4-ylmethyl)-3h-imidazo-
[4,5-b]pyridine-6-carbonitrile derivatives as complement factor b
inhibitors useful for the treatment of ophthalmic diseases. Patent WO
2014143638 A1, September 18, 2014.
PCET from the photoexcited catalyst (vide supra). The
iminium reaction differs from the oxime reaction by removing
the requirement to generate the iminium in situ. Excited-state
photocatalyst III can oxidize DIPA, generating an ammonium
radical cation. The α-C−H bonds of an ammonium radical
cation become markedly more acidic than those in the parent
amine.3a A second unit of DIPA or 4a could deprotonate the
ammonium radical cation, resulting in radical VIII. The
iminium 6a can be reduced by both pathways considered for
the oxime reaction, where either reduced photocatalyst IV or
radical VIII can donate an electron to 6a, leading to α-amino
radical XII. Subsequent heteroradical coupling of XII and V
and formal loss of both a proton and cyanide would afford
product 5a.
(3) Selected reviews on amine synthesis: (a) Beatty, J. W.;
Stephenson, C. R. J. Amine Functionalization via Oxidative
Photoredox Catalysis: Methodology Development and Complex
Molecule Synthesis. Acc. Chem. Res. 2015, 48, 1474−1484. (b) Enders,
D.; Reinhold, U. Asymmetric Synthesis of Amines by Nucleophilic
1,2-Addition of Organometallic Reagents to the CN-Double Bond.
Tetrahedron: Asymmetry 1997, 8, 1895−1946. (c) Lu, Z.; Ma, S.
Metal-Catalyzed Enantioselective Allylation in Asymmetric Synthesis.
Angew. Chem., Int. Ed. 2008, 47, 258−297. (d) Legnani, L.; Bhawal, B.
N.; Morandi, B. Recent Developments in the Direct Synthesis of
Unprotected Primary Amines. Synthesis 2017, 49, 776−789. (e) Ell-
man, J. A. Applications of Tert-Butanesulfinamide in the Asymmetric
Synthesis of Amines. Pure Appl. Chem. 2003, 75, 39−46.
(4) (a) Liu, G.; Cogan, D. A.; Ellman, J. A. Catalytic Asymmetric
Synthesis of tert-Butanesulfinamide. Application to the Asymmetric
Synthesis of Amines. J. Am. Chem. Soc. 1997, 119, 9913−9914.
(b) Robak, M. T.; Herbage, M. A.; Ellman, J. A. Synthesis and
Applications of tert-Butanesulfinamide. Chem. Rev. 2010, 110, 3600−
3740. (c) McMahon, J. P.; Ellman, J. A. Highly Stereoselective
Addition of Organometallic Reagents to N-tert-Butanesulfinyl Imines
Derived from 3- and 4-Substituted Cyclohexanones. Org. Lett. 2004,
6, 1645−1647. (d) Cogan, D. A.; Liu, G.; Ellman, J. Asymmetric
synthesis of chiral amines by highly diastereoselective 1,2-additions of
organometallic reagents to N-tert-butanesulfinyl imines. Tetrahedron
1999, 55, 8883−8904. (e) Cogan, D. A.; Ellman, J. A. Asymmetric
Synthesis of α,α-Dibranched Amines by the Trimethylaluminum-
Mediated 1,2-Addition of Organolithiums to tert-Butanesulfinyl
Ketimines. J. Am. Chem. Soc. 1999, 121, 268−269. (f) Liu, G.;
Cogan, D. A.; Owens, T. D.; Tang, T. P.; Ellman, J. A. Synthesis of
Enantiomerically Pure N-tert-Butanesulfinyl Imines (tert-Butanesulfi-
nimines) by the Direct Condensation of tert-Butanesulfinamide with
Aldehydes and Ketones. J. Org. Chem. 1999, 64, 1278−1284.
(5) Select examples of photoredox-catalyzed functionalization of
amines: (a) Douglas, J. J.; Cole, K. P.; Stephenson, C. R. J.
Photoredox Catalysis in a Complex Pharmaceutical Setting: Toward
the Preparation of JAK2 Inhibitor LY2784544. J. Org. Chem. 2014, 79,
11631−11643. (b) Joe, C. L.; Doyle, A. G. Direct Acylation of
C(sp3)-H Bonds Enabled by Nickel and Photoredox Catalysis. Angew.
Chem., Int. Ed. 2016, 55, 4040−4043. (c) Ahneman, D. T.; Doyle, A.
G. C-H functionalization of amines with aryl halides by nickel-
photoredox catalysis. Chem. Sci. 2016, 7, 7002−7006. (d) Zuo, Z.;
Ahneman, D. T.; Chu, L.; Terrett, J. A.; Doyle, A. G.; MacMillan, D.
W. C. Merging Photoredox with Nickel Catalysis: Coupling of α-
Carboxyl Sp3-Carbons with Aryl Halides. Science 2014, 345, 437−
440. (e) Patel, N. R.; Kelly, C. B.; Siegenfeld, A. P.; Molander, G. A.
Mild, Redox-Neutral Alkylation of Imines Enabled by an Organic
Photocatalyst. ACS Catal. 2017, 7, 1766−1770. (f) Alam, R.;
Molander, G. A. Photoredox-Catalyzed Direct Reductive Amination
of Aldehydes without an External Hydrogen/Hydride Source. Org.
Lett. 2018, 20, 2680−2684. (g) Musacchio, A. J.; Lainhart, B. C.;
Zhang, X.; Naguib, S. G.; Sherwood, T. C.; Knowles, R. R. Catalytic
Intermolecular Hydroaminations of Unactivated Olefins with
Secondary Alkyl Amines. Science 2017, 355, 727−730. (h) Trow-
CONCLUSION
■
In summary, we have developed photoredox-catalyzed
methodology for the direct synthesis of primary amines
without the use of harsh organometallic reagents. We have
demonstrated that two separate bench-stable starting materials,
benzylic O-benzoyl ketoximes and benzylic iminium chloride
salts, can be coupled efficiently with 4-cyano-heteroarenes and
tolerate a great diversity of functional groups. Further, we have
identified a reaction where the photoredox catalyst is able to
efficiently catalyze a net reductive coupling by acting as both a
reductant and a triplet sensitizer, resulting in a reaction which
is facilitated by concurrent tandem catalysis. This method
enables scientists to synthesize libraries of hindered primary
amines via parallel photoredox, including potential high-
throughput experimentation for medicinal chemists.41,42
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.9b10871.
■
S
Synthetic procedures, spectral data, compound charac-
terization, and Cartesian coordinates of computational
structures (PDF)
AUTHOR INFORMATION
■
Corresponding Authors
*dan.lehnherr@merck.com
*tr2504@columbia.edu
ORCID
Michael C. Nicastri: 0000-0002-6283-5543
Dan Lehnherr: 0000-0001-8392-1208
Yu-hong Lam: 0000-0002-4946-1487
Tomislav Rovis: 0000-0001-6287-8669
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was funded by the Fellowship for Disruptive
Chemistry at Merck & Co., Inc., Kenilworth, NJ, USA.
REFERENCES
■
(1) Campos, K. R.; Coleman, P. J.; Alvarez, J. C.; Dreher, S. D.;
Garbaccio, R. M.; Terrett, N. K.; Tillyer, R. D.; Truppo, M. D.;
Parmee, E. R. The Importance of Synthetic Chemistry in the
Pharmaceutical Industry. Science 2019, 363, No. eaat0805.
I
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
bridge, A.; Reich, D.; Gaunt, M. J. Multicomponent synthesis of
tertiary alkylamines by photocatalytic olefin-hydroaminoalkylation.
phore. J. Am. Chem. Soc. 2014, 136, 5257−5260. (k) Cuthbertson, J.
D.; MacMillan, D. W. C. The direct arylation of allylic sp3 C−H
bonds via organic and photoredox catalysis. Nature 2015, 519, 74−77.
(l) Yan, C.; Li, L.; Liu, Y.; Wang, Q. Direct and Oxidant-Free
Electron-Deficient Arylation of N-Acyl-Protected Tetrahydroisoqui-
nolines. Org. Lett. 2016, 18, 4686−4689. (m) Betori, R. C.; Scheidt,
K. A. Reductive Arylation of Arylidene Malonates Using Photoredox
Catalysis. ACS Catal. 2019, 9, 10350−10357.
(10) (a) Shevlin, M. Practical High-Throughput Experimentation for
Chemists. ACS Med. Chem. Lett. 2017, 8, 601−607. (b) Krska, S. W.;
DiRocco, D. A.; Dreher, S. D.; Shevlin, M. The Evolution of Chemical
High-Throughput Experimentation to Address Challenging Problems
in Pharmaceutical Synthesis. Acc. Chem. Res. 2017, 50, 2976−2985.
(c) DiRocco, D. A.; Dykstra, K.; Krska, S.; Vachal, P.; Conway, D. V.;
Tudge, M. Late-Stage Functionalization of Biologically Active
Heterocycles Through Photoredox Catalysis. Angew. Chem., Int. Ed.
2014, 53, 4802−4806.
́
Nature 2018, 561, 522−527. (i) Floden, N. J.; Trowbridge, A.;
Willcox, D.; Walton, S. M.; Kim, Y.; Gaunt, M. J. Streamlined
Synthesis of C(sp3)-Rich N-Heterospirocycles Enabled by Visible-
Light-Mediated Photocatalysis. J. Am. Chem. Soc. 2019, 141, 8426−
̈
8430. (j) Hari, D. P.; Konig, B. Eosin y Catalyzed Visible Light
Oxidative C-C and C-P Bond Formation. Org. Lett. 2011, 13, 3852−
3855. (k) McManus, J. B.; Onuska, N. P. R.; Nicewicz, D. A.
Generation and Alkylation of α-Carbamyl Radicals via Organic
Photoredox Catalysis. J. Am. Chem. Soc. 2018, 140, 9056−9060.
(l) Ye, J.; Kalvet, I.; Schoenebeck, F.; Rovis, T. Direct α-alkylation of
primary aliphatic amines enabled by CO2 and electrostatics. Nat.
Chem. 2018, 10, 1037−1041. (m) Ashley, M. A.; Yamauchi, C.; Chu,
J. C. K.; Otsuka, S.; Yorimitsu, H.; Rovis, T. Photoredox-Catalyzed
Site-Selective α-C(sp3)−H Alkylation of Primary Amine Derivatives.
Angew. Chem., Int. Ed. 2019, 58, 4002−4006. (n) Thullen, S. M.;
Rovis, T. A Mild Hydroaminoalkylation of Conjugated Dienes Using
a Unified Cobalt and Photoredox Catalytic System. J. Am. Chem. Soc.
2017, 139, 15504−15508.
(11) The literature value for the reduction potential of Ir(ppy)3
(PC4) is (IrII) Ered = −2.20 V vs SCE. See: Koike, T.; Akita, M.
Visible-Light Radical Reaction Designed by Ru- and Ir-Based
Photoredox Catalysis. Inorg. Chem. Front. 2014, 1, 562−576.
(12) The photocatalyst 4CZIPN (PC8) has a reduction potential of
(6) Vasu, D.; Fuentes de Arriba, A. L.; Leitch, J. A.; de Gombert, A.;
Dixon, D. J. Primary α-tertiary amine synthesis via α-C−H
functionalization. Chem. Sci. 2019, 10, 3401−3407.
(7) Rong, J.; Seeberger, P. H.; Gilmore, K. Chemoselective
Photoredox Synthesis of Unprotected Primary Amines Using
Ammonia. Org. Lett. 2018, 20, 4081−4085.
(8) For selected examples of N−O bond scission of oximes in
photoredox catalysis, see: (a) Morcillo, S. P.; Dauncey, E. M.; Kim, J.
H.; Douglas, J. J.; Sheikh, N. S.; Leonori, D. Photoinduced Remote
Functionalization of Amides and Amines Using Electrophilic Nitrogen
Radicals. Angew. Chem., Int. Ed. 2018, 57, 12945−12949. (b) Davies,
J.; Morcillo, S. P.; Douglas, J. J.; Leonori, D. Hydroxylamine
Derivatives as Nitrogen-Radical Precursors. Chem. - Eur. J. 2018,
24, 12154−12163. (c) Dauncey, E. M.; Morcillo, S. P.; Douglas, J. J.;
Sheikh, N. S.; Leonori, D. Photoinduced Remote Functionalisations
by Iminyl Radical Promoted C−C and C−H Bond Cleavage
Cascades. Angew. Chem., Int. Ed. 2018, 57, 744−748.
(9) For selected examples of ipso-substitution of cyanoarenes in the
context of radical chemistry, see: (a) Vittimberga, B. M.; Minisci, F.;
Morrocchi, S. The Photoinitiated Electron Transfer-Substitution
Reaction of Diphenyl Ketyl with Protonated 4-Cyanopyridine. J.
Am. Chem. Soc. 1975, 97, 4397−4398. (b) Furihata, T.; Sugimori, A.
Competitive photo-ethoxylation and -hydroxyethylation of pyridine-2-
carbonitrile in acidic ethanolic solutions. J. Chem. Soc., Chem.
Commun. 1975, 241b−242. (c) Borg, R. M.; Arnold, D. R.;
Cameron, T. S. Radical ions in photochemistry. 15. The photo-
substitution reaction between dicyanobenzenes and alkyl olefins. Can.
J. Chem. 1984, 62, 1785−1802. (d) Zeng, X.; Cai, J.; Gu, Y. A Novel
Hydroalkyl-decyanation of 4-Pyridinecarbonitrile: A Facile Selective
Synthesis of 4-Pyridinemethanols. Tetrahedron Lett. 1995, 36, 7275−
7276. (e) Bernardi, R.; Caronna, T.; Morrocchi, S.; Ursini, M.
Photochemically induced reactions of selected ketones with 4-
cyanopyridine in neutral and acidic medium. J. Heterocycl. Chem.
1996, 33, 1137−1142. (f) Bernardi, R.; Caronna, T.; dal Pio Luogo,
D.; Morrocchi, S.; Poggi, G.; Vittimberga, B. M. A study of solvent
effect on photochemically induced reactions between pyridinedicar-
bonitriles and alkenes: an easy approach to the synthesis of cyclopenta
[b or c] pyridines. J. Chem. Soc., Perkin Trans. 1 1996, 1593−1600.
(g) McNally, A.; Prier, C. K.; MacMillan, D. W. C. Discovery of an α-
Amino C−H Arylation Reaction Using the Strategy of Accelerated
Serendipity. Science 2011, 334, 1114−1117. (h) Pirnot, M. T.; Rankic,
D. A.; Martin, D. B. C.; MacMillan, D. W. C. Photoredox Activation
for the Direct β-Arylation of Ketones and Aldehydes. Science 2013,
339, 1593−1596. (i) Qvortrup, K.; Rankic, D. A.; MacMillan, D. W.
C. A General Strategy for Organocatalytic Activation of C−H Bonds
via Photoredox Catalysis: Direct Arylation of Benzylic Ethers. J. Am.
Chem. Soc. 2014, 136, 626−629. (j) Zuo, Z.; MacMillan, D. W. C.
Decarboxylative Arylation of α-Amino Acids via Photoredox
Catalysis: A One-Step Conversion of Biomass to Drug Pharmaco-
E
red(P/P−) = − 1.21V vs SCE. See: Luo, J.; Zhang, J. Donor-Acceptor
Fluorophores for Visible-Light-Promoted Organic Synthesis: Photo-
redox/Ni Dual Catalytic C(Sp3)-C(Sp2) Cross-Coupling. ACS Catal.
2016, 6, 873−877.
(13) Products of Minisci-type heteroarene alkylation were observed
via UPLC-MS analysis of crude reaction mixtures. See Supporting
Information, Figure S1, for more information.
(14) Reduction of 4-cyanopyridine is postulated to proceed via a
proton-coupled electron-transfer (PCET) event.
(15) For selected examples of using DIPEA as terminal reductant in
photochemistry, see: (a) Narayanam, J. M. R.; Tucker, J. W.;
Stephenson, C. R. J. Electron-Transfer Photoredox Catalysis:
Development of a Tin-Free Reductive Dehalogenation Reaction. J.
Am. Chem. Soc. 2009, 131, 8756−8757. (b) Wallentin, C.-J.; Nguyen,
J. D.; Finkbeiner, P.; Stephenson, C. R. J. Visible Light-Mediated
Atom Transfer Radical Addition via Oxidative and Reductive
Quenching of Photocatalysts. J. Am. Chem. Soc. 2012, 134, 8875−
8884. (c) Devery, J. J.; Nguyen, J. D.; Dai, C.; Stephenson, C. R. J.
Light-Mediated Reductive Debromination of Unactivated Alkyl and
Aryl Bromides. ACS Catal. 2016, 6, 5962−5967. (d) Karkas, M. D.;
Porco, J. A.; Stephenson, C. R. J. Photochemical Approaches to
Complex Chemotypes: Applications in Natural Product Synthesis.
Chem. Rev. 2016, 116, 9683−9747.
(16) DFT calculations using M06-2X/6-31+G(d,p) SMD=DMSO
•
for the reaction of Me2SO + iPr-HN-(C•)Me2 → Me-SO-CH2 + (i-
Pr)2NH give a ΔGrxn of 17.4 kcal/mol, supporting that this process is
energetically unfavorable.
(17) Watson, W. On Byproducts and Side Products. Org. Process Res.
Dev. 2012, 16, 1877.
(18) The Photoreactor TAK120. http://wppr.photoreactor.de/
tak120 (accessed August 28, 2019).
(19) We utilized DFT calculations to estimate the stability of the
cyclopropylmethyl radical intermediate. We observed that the ring-
closed intermediate is favored by ca. 11 kcal/mol (Supporting
Information, Figure S93) over its analogous ring-opened structure, in
agreement with our experimental results.
(20) Product was identified via LC/MS analysis. The yield was
calculated to be <10%.
(21) Benzylic oximes were prepared containing furan, pyridine, and
pyrazine as their aromatic components. Furan- and pyrazine-derived
oximes produced no detectable product. The starting oximes were
completely consumed in the reaction.
(22) (a) Purser, S.; Moore, P. R.; Swallow, S.; Gouverneur, V.
Fluorine in medicinal chemistry. Chem. Soc. Rev. 2008, 37, 320−330.
(b) Shah, P.; Westwell, A. D. The role of fluorine in medicinal
chemistry. J. Enzyme Inhib. Med. Chem. 2007, 22, 527−540. (c) Gillis,
E. P.; Eastman, K. J.; Hill, M. D.; Donnelly, D. J.; Meanwell, N. A.
J
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Applications of Fluorine in Medicinal Chemistry. J. Med. Chem. 2015,
58, 8315−8359.
1999, 1978−1980. (h) Boivin, J.; Fouquet, E.; Zard, S. Z. A new and
synthetically useful source of iminyl radicals. Tetrahedron Lett. 1991,
32, 4299−4302. (i) Yu, X. Y.; Chen, J. R.; Wang, P. Z.; Yang, M. N.;
Liang, D.; Xiao, W. J. A Visible-Light-Driven Iminyl Radical-Mediated
C-C Single Bond Cleavage/Radical Addition Cascade of Oxime
Esters. Angew. Chem., Int. Ed. 2018, 57, 738−743. (j) Li, L.; Chen, H.;
Mei, M.; Zhou, L. Visible-light promoted γ-cyanoalkyl radical
generation: three-component cyanopropylation/etherification of
unactivated alkenes. Chem. Commun. 2017, 53, 11544−11547.
(k) Shen, X.; Zhao, J. J.; Yu, S. Photoredox-Catalyzed Intermolecular
Remote C−H and C−C Vinylation via Iminyl Radicals. Org. Lett.
2018, 20, 5523−5527. (l) Zhao, B.; Wang, M.; Shi, Z. Single-
Electron-Transfer-Induced C(sp3)−N Couplings via C−C Bond
Cleavage of Cycloketoxime Esters. J. Org. Chem. 2019, 84, 10145−
10159. (m) Jian, Y.; Chen, M.; Yang, C.; Xia, W. Minisci-Type C−H
Cyanoalkylation of Heteroarenes Through N−O/C−C Bonds
Cleavage. Eur. J. Org. Chem. 2019, DOI: 10.1002/ejoc.201900406.
(n) Fan, X.; Lei, T.; Liu, Z.; Yang, X.-L.; Cheng, Y.-Y.; Liang, G.;
Chen, B.; Tung, C.-H.; Wu, L.-Z. Benzyl C-O and C-N Bond
Construction via C-C Bond Dissociation of Oxime Ester under
Visible Light Irradiation. Eur. J. Org. Chem. 2019, DOI: 10.1002/
ejoc.201901060.
(29) An impurity was observed via UPLC-MS under optimized
reaction conditions (e.g., Table 3, Method A) consistent with
formation of a diarylated product derived from an iminyl radical
performing ipso-substitution on 4-cyanopyridine and subsequent
generation of an α-amino radical from that arylated imine, ultimately
leading to a second ipso-substitution on 4-cyanopyridine (see
Supporting Information, Figure S1). The same analysis of reactions
using iminium chloride salts (e.g., Table 3, Method B) did not show
the same diarylated reaction byproduct via UPLC-MS. Product 5a
(produced via Method A) was subjected to reaction conditions for
Method A instead of oxime 3a, and no diarylated impurity was
observed via UPLC-MS.
(23) Le, C.; Wismer, M. K.; Shi, Z.-C.; Zhang, R.; Conway, D. V.; Li,
G.; Vachal, P.; Davies, I. W.; MacMillan, D. W. C. A General Small
Scale Reactor To Enable Standardization and Acceleration of
Photocatalytic Reactions. ACS Cent. Sci. 2017, 3, 647−653.
(24) Callier-Dublanchet, A.-C.; Quiclet-Sire, B.; Zard, S. Z. A new
source of nitrogen centered radicals. Tetrahedron Lett. 1995, 36,
8791−8794.
(25) Ladouceur, S.; Fortin, D.; Zysman-Colman, E. Enhanced
Luminescent Iridium(III) Complexes Bearing Aryltriazole Cyclo-
metallated Ligands. Inorg. Chem. 2011, 50, 11514−11526.
(26) For cleavage of oximes using radical initiators, see: (a) El Kaim,
L.; Meyer, C. An Unprecedented Radical Reaction of Benzotriazole
Derivatives. A New Efficient Method for the Generation of Iminyl
Radicals. J. Org. Chem. 1996, 61, 1556−1557. (b) Boivin, J.; Callier-
Dublanchet, A.-C.; Quiclet-Sire, B.; Schiano, A.-M.; Zard, S. Z. Iminyl,
amidyl, and carbamyl radicals from O-benzoyl oximes and O-benzoyl
hydroxamic acid derivatives. Tetrahedron 1995, 51, 6517−6528.
(c) Callier, A.-C.; Quiclet-Sire, B.; Zard, S. Z. Amidyl and carbamyl
radicals by stannane mediated cleavage of O-benzoyl hydroxamic acid
derivatives. Tetrahedron Lett. 1994, 35, 6109−6112. (d) Boivin, J.;
Schiano, A.-M.; Zard, S. Z. Iminyl radicals by stannane mediated
cleavage of oxime esters. Tetrahedron Lett. 1994, 35, 249−252.
(e) Boivin, J.; Fouquet, E.; Zard, S. Z. Iminyl radicals: part II. ring
opening of cyclobutyl- and cyclopentyliminyl radicals. Tetrahedron
1994, 50, 1757−1768. (f) Boivin, J.; Fouquet, E.; Zard, S. Z. Ring
opening induced by iminyl radicals derived from cyclobutanones: new
aspects of tin hydride cleavage of S-phenyl sulfenylimines. J. Am.
Chem. Soc. 1991, 113, 1055−1057. (g) Boivin, J.; Fouquet, E.; Zard,
S. Z. Iminyl radicals: Part I. generation and intramolecular capture by
an olefin. Tetrahedron 1994, 50, 1745−1756. (h) Boivin, J.; Fouquet,
E.; Zard, S. Z. Multiple radical additions starting with iminyl radicals
derived from sulphenylimines. Tetrahedron Lett. 1990, 31, 3545−
3548.
(30) Srivastava, V.; Singh, P. P. Eosin y Catalysed Photoredox
Synthesis: A Review. RSC Adv. 2017, 7, 31377−31392.
(31) (a) Daub, M. E.; Jung, H.; Lee, B. J.; Won, J.; Baik, M.-H.;
Yoon, T. P. Enantioselective [2 + 2] Cycloadditions of Cinnamate
Esters: Generalizing Lewis Acid Catalysis of Triplet Energy Transfer.
J. Am. Chem. Soc. 2019, 141, 9543−9547. (b) Teders, M.; Henkel, C.;
(27) For cleavage of oximes using heat, see: (a) Cai, Y.; Jalan, A.;
Kubosumi, A. R.; Castle, S. L. Microwave-Promoted Tin-Free Iminyl
Radical Cyclization with TEMPO Trapping: A Practical Synthesis of
2-Acylpyrroles. Org. Lett. 2015, 17, 488−491. (b) Portela-Cubillo, F.;
Scott, J. S.; Walton, J. C. Microwave-assisted preparations of
dihydropyrroles from alkenone O-phenyl oximes. Chem. Commun.
2007, 4041−4043. (c) Portela-Cubillo, F.; Scott, J. S.; Walton, J. C.
Microwave-Assisted Syntheses of N-Heterocycles Using Alkenone-,
Alkynone- and Aryl-carbonyl O-Phenyl Oximes: Formal Synthesis of
Neocryptolepine. J. Org. Chem. 2008, 73, 5558−5565. (d) Blake, J. A.;
Pratt, D. A.; Lin, S.; Walton, J. C.; Mulder, P.; Ingold, K. U.
Thermolyses of O-Phenyl Oxime Ethers. A New Source of Iminyl
Radicals and a New Source of Aryloxyl Radicals. J. Org. Chem. 2004,
69, 3112−3120. (e) Lin, X.; Stien, D.; Weinreb, S. M. A New Method
for the Generation and Cyclization of Iminyl Radicals via the Hudson
Reaction. Org. Lett. 1999, 1, 637−640.
(28) For photocleavage of oximes, see: (a) Wang, P. Z.; He, B. Q.;
Cheng, Y.; Chen, J. R.; Xiao, W. J. Radical C−C Bond Cleavage/
Addition Cascade of Benzyl Cycloketone Oxime Ethers Enabled by
Photogenerated Cyclic Iminyl Radicals. Org. Lett. 2019, 21, 6924−
6929. (b) Faulkner, A.; Race, N. J.; Scott, J. S.; Bower, J. F. Copper
catalyzed Heck-like cyclizations of oxime esters. Chem. Sci. 2014, 5,
2416−2421. (c) Bingham, M.; Moutrille, C.; Zard, S. Z. A Simple
Route to Functionalized Δ1-Pyrrolenines. Heterocycles 2014, 88, 953−
960. (d) Portela-Cubillo, F.; Lymer, J.; Scanlan, E. M.; Scott, J. S.;
Walton, J. C. Dioxime oxalates; new iminyl radical precursors for
syntheses of N-heterocycles. Tetrahedron 2008, 64, 11908−11916.
(e) Alonso, R.; Campos, P. J.; Rodríguez, M. A.; Sampedro, D.
Photocyclization of Iminyl Radicals: Theoretical Study and Photo-
chemical Aspects. J. Org. Chem. 2008, 73, 2234−2239. (f) Alonso, R.;
Campos, P. J.; García, B.; Rodríguez, M. A. New Light-Induced Iminyl
Radical Cyclization Reactions of Acyloximes to Isoquinolines. Org.
Lett. 2006, 8, 3521−3523. (g) Gagosz, F.; Zard, S. Z. Generation and
Capture of Iminyl Radicals from Ketoxime Xanthates. Synlett 1999,
́
́
̈
Anhauser, L.; Strieth-Kalthoff, F.; Gomez-Suarez, A.; Kleinmans, R.;
Kahnt, A.; Rentmeister, A.; Guldi, D.; Glorius, F. The Energy-
Transfer-Enabled Biocompatible Disulfide−Ene Reaction. Nat. Chem.
2018, 10, 981−988.
(32) Patra, T.; Mukherjee, S.; Ma, J.; Strieth-Kalthoff, F.; Glorius, F.
Visible-light photosensitized aryl and alkyl decarboxylative carbon-
heteroatom and carbon-carbon bond formations. Angew. Chem., Int.
Ed. 2019, 58, 10514−10520.
(33) Soni, V. K.; Lee, S.; Kang, J.; Moon, Y. K.; Hwang, H. S.; You,
Y.; Cho, E. J. Reactivity Tuning for Radical−Radical Cross-Coupling
via Selective Photocatalytic Energy Transfer: Access to Amine
Building Blocks. ACS Catal. 2019, 9, 10454−10463.
(34) (a) Strieth-Kalthoff, F.; James, M. J.; Teders, M.; Pitzer, L.;
Glorius, F. Energy transfer catalysis mediated by visible light:
principles, applications, directions. Chem. Soc. Rev. 2018, 47, 7190−
7202. (b) Srivastava, V.; Singh, P. P. Eosin Y Catalysed Photoredox
Synthesis: A Review. RSC Adv. 2017, 7, 31377−31392.
(35) Reduction potentials can be found in our related paper:
Lehnherr, D.; Lam, Y.-h.; Nicastri, M. C; Liu, J.; Newman, J. A.;
Regalado, E. L.; DiRocco, D. A.; Rovis, T.; et al. Electrochemical
Synthesis of Hindered Primary and Secondary Amines via Proton-
Coupled Electron Transfer. J. Am. Chem. Soc. 2019, DOI: 10.1021/
jacs.9b10870.
(36) For literature precedent for using DFT to calculate T1-to-S0
energy gaps and successfully correlating these with experimental
observations in photocatalyzed organic transformations, see:
(a) Zhao, J.; Brosmer, J. L.; Tang, Q.; Yang, Z.; Houk, K. N.;
Diaconescu, P. L.; Kwon, O. Intramolecular Crossed [2 + 2]
Photocycloaddition through Visible Light-Induced Energy Transfer. J.
K
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX

Journal of the American Chemical Society
Article
Am. Chem. Soc. 2017, 139, 9807−9810. (b) James, M. J.; Schwarz, J.
L.; Strieth-Kalthoff, F.; Wibbeling, B.; Glorius, F. Dearomative
Cascade Photocatalysis: Divergent Synthesis through Catalyst
Selective Energy Transfer. J. Am. Chem. Soc. 2018, 140, 8624−
8628. (c) Blum, T. R.; Miller, Z. D.; Bates, D. M.; Guzei, I. A.; Yoon,
T. P. Enantioselective photochemistry through Lewis acid−catalyzed
triplet energy transfer. Science 2016, 354, 1391−1395. (d) Chang, X.-
P.; Zheng, Y.; Cui, G.; Fang, W.-H.; Thiel, W. Photocycloaddition
reaction of atropisomeric maleimides: mechanism and selectivity.
Phys. Chem. Chem. Phys. 2016, 18, 24713−24721.
(37) For reviews on tandem catalysis and definition of concurrent
tandem catalysis, see: (a) Wasilke, J.-C.; Obrey, S. J.; Baker, R. T.;
Bazan, G. C. Concurrent Tandem Catalysis. Chem. Rev. 2005, 105,
1001−1020. (b) Fogg, D. E.; dos Santos, E. N. Tandem catalysis: a
taxonomy and illustrative review. Coord. Chem. Rev. 2004, 248, 2365−
2379.
(38) Concurrent tandem catalysis is a synthetic transformation that
involves the cooperative action of two or more catalytic cycles in a
single reactor. In the example from this Article, the catalyst operating
in both catalytic cycles is the same, acting in different roles as a redox
catalyst and as an energy transfer catalyst.
̈
(39) Forster energy transfer is ruled out due to a lack of spectral
overlap between the UV−vis emission spectra of photocatalyst PC1
and the oxime UV−vis absorption spectra.
(40) (a) Wayner, D. D. M.; Dannenberg, J. J.; Griller, D. Oxidation
potentials of α-aminoalkyl radicals: bond dissociation energies for
related radical cations. Chem. Phys. Lett. 1986, 131, 189−191.
(b) Ismaili, H.; Pitre, S. P.; Scaiano, J. C. Active participation of
amine-derived radicals in photoredox catalysis as exemplified by a
reductive cyclization. Catal. Sci. Technol. 2013, 3, 935−937.
(41) For a discussion of direct-to-biology and parallel library
synthesis, see: Santanilla, A. B.; Regalado, E. R.; Pereira, T.; Shevlin,
M.; Bateman, K.; Campeau, L. C.; Schneeweis, J.; Berritt, S.; Shi, Z.-
C.; Nantermet, P.; Liu, Y.; Helmy, R.; Welch, C. J.; Vachal, P.; Davies,
I. W.; Cernak, T.; Dreher, S. D. Nanomole-scale high-throughput
chemistry for the synthesis of complex molecules. Science 2015, 347,
49−53.
(42) For a related method to the work herein, in which hindered
amines are synthesized from iminium chlorides and cyanoheteroar-
enes using electrochemistry, see: ref 35.
L
DOI: 10.1021/jacs.9b10871
J. Am. Chem. Soc. XXXX, XXX, XXX−XXX