Catalytic α-Deracemization of Ketones Enabled by Photoredox Deprotonation and Enantioselective Protonation

Article abstract of DOI:10.1021/jacs.1c06637

This study reports the catalytic deracemization of ketones bearing stereocenters in the α-position in a single reaction via deprotonation, followed by enantioselective protonation. The principle of microscopic reversibility, which has previously rendered this strategy elusive, is overcome by a photoredox deprotonation through single electron transfer and subsequent hydrogen atom transfer (HAT). Specifically, the irradiation of racemic pyridylketones in the presence of a single photocatalyst and a tertiary amine provides nonracemic carbonyl compounds with up to 97% enantiomeric excess. The photocatalyst harvests the visible light, induces the redox process, and is responsible for the asymmetric induction, while the amine serves as a single electron donor, HAT reagent, and proton source. This conceptually simple light-driven strategy of coupling a photoredox deprotonation with a stereocontrolled protonation, in conjunction with an enrichment process, serves as a blueprint for other deracemizations of ubiquitous carbonyl compounds.

Full text of DOI:10.1021/jacs.1c06637

pubs.acs.org/JACS
Article
Catalytic α‑Deracemization of Ketones Enabled by Photoredox
Deprotonation and Enantioselective Protonation
Chenhao Zhang, Anthony Z. Gao, Xin Nie, Chen-Xi Ye, Sergei I. Ivlev, Shuming Chen,*
and Eric Meggers*
Cite This: J. Am. Chem. Soc. 2021, 143, 13393−13400
Read Online
sı
*
Metrics & More
Article Recommendations
Supporting Information
ACCESS
ABSTRACT: This study reports the catalytic deracemization of ketones
bearing stereocenters in the α-position in a single reaction via deprotonation,
followed by enantioselective protonation. The principle of microscopic
reversibility, which has previously rendered this strategy elusive, is overcome
by a photoredox deprotonation through single electron transfer and
subsequent hydrogen atom transfer (HAT). Specifically, the irradiation of
racemic pyridylketones in the presence of a single photocatalyst and a tertiary
amine provides nonracemic carbonyl compounds with up to 97% enantiomeric
excess. The photocatalyst harvests the visible light, induces the redox process,
and is responsible for the asymmetric induction, while the amine serves as a
single electron donor, HAT reagent, and proton source. This conceptually
simple light-driven strategy of coupling a photoredox deprotonation with a stereocontrolled protonation, in conjunction with an
enrichment process, serves as a blueprint for other deracemizations of ubiquitous carbonyl compounds.
which takes advantage of the intrinsic acidity of the α-C−H
bond. This transformation not only offers a straightforward
handle to introduce further functionality but often also
generates a stereocenter (Figure 1a). The α-deracemization
of carbonyl compounds is therefore an attractive strategy for
generating enantiomerically pure carbonyl compounds. Sur-
prisingly, α-deracemizations of carbonyl compounds are rare.
Tsunoda8,9 and Matsumoto10 reported the α-deracemization
of 2-alkylketones and 2-hydroxyketones, respectively, utilizing
host−guest chemistry with a chiral host, while Wu11 developed
an α-deracemization via crystallization-induced dynamic
resolution of corresponding chiral imines. In these cases,
equimolar amounts of a chiral reagent were needed.
No single-operation catalytic α-deracemizations of carbonyl
compounds in the prominent α-position have been reported.
At first glance this is surprising, as a deracemization can be
envisioned to proceed through an enolate intermediate upon
deprotonation, followed by an enantioselective protonation
(Figure 1b). In fact, the enantioselective protonation of
enolates or enolate equivalents such as silyl enol ether is well
established, including catalytic methods.12−18 However, the
principle of microscopic reversibility prohibits a one-pot
INTRODUCTION
■
Optically active compounds are in high demand for the
synthesis of chiral fine chemicals and pharmaceuticals. Despite
the progress that has been made in asymmetric organic
synthesis, the vast majority of standard organic reactions
provide racemic products if the generation of a new
stereocenter is involved. Deracemization reactions, in which
one enantiomer of a racemic mixture is converted to the other
enantiomer, thus providing a single stereoisomer without
otherwise modifying the overall chemical structure, are
therefore highly appealing.1,2 In contrast to dynamic kinetic
resolutions and related processes,3−5 such deracemizations, in
which substrate and product possess an identical chemical
structure, are surprisingly rare. Two obstacles must be
overcome in designing a successful deracemization reaction.
First, the deracemization process is thermodynamically
unfavorable due to a decrease in entropy.6 Second, the
deracemization must occur through an achiral intermediate
(e.g., an enolate), but if the forward and reverse reaction follow
the same mechanistic pathway, the principle of microscopic
reversibility will always result in the formation of the
racemate.7 Thus, new mechanistic concepts for the realization
of one-step deracemizations, ideally in a catalytic fashion, are
highly desirable.
Received: June 26, 2021
Published: August 14, 2021
Carbonyl groups are one of the most important and
abundant functional groups in organic chemistry. Carbonyl
compounds are often key intermediates for the synthesis of
more complex molecules. A prominent example is the C−H
functionalization of carbonyl compounds at the α-position,
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400
© 2021 American Chemical Society
13393

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
generate a chelator for binding to photoactive bis-cyclo-
metalated iridium or rhodium complexes for which we have
demonstrated versatile asymmetric photocatalysis over the
years.22,23 We envisioned that transition-metal coordination, in
combination with the withdrawing nature of the pyridine
moiety, would stabilize the ketyl intermediate in the proposed
single electron transfer/hydrogen atom transfer (SET/HAT)
sequence. Photolysis of rac-1 in the presence of the established
chiral iridium photocatalyst Λ-IrS24 (4 mol %) in acetone
together with DABCO (B1) or quinuclidine (B2) (each 3.0
equiv) as electron donors and HAT reagents25,26 provided rac-
1 unchanged as a complete racemate after 24 h at room
temperature (Table 1, entries 1 and 2). Replacing DABCO or
quinuclidine with Hunig’s base (B3) did not alter the outcome
̈
(entry 3). However, when we replaced the iridium photo-
catalyst Λ-IrS with the analogous bis-cyclometalated phenyl
benzothiazole rhodium catalyst Λ-RhS,27 ketone 1 showed an
enantiomeric excess (ee) of 26% (entry 4). Although this
relates to a ratio of R and S enantiomers of just 63:37, this
result encouraged us to further optimize the reaction
conditions. Substituting Hunig’s base with diisopropylbenzyl-
̈
amine (B4) saw the ee improve to 42% (entry 5). More
notable results were obtained with N-phenylpiperidine (B5,
81% ee) and N-phenylazepane (B6, 84% ee) (entries 6 and 7).
These results demonstrate that the efficiency of this light-
driven deracemization28,29 is strongly affected by the structure
of the amine. Nonetheless, despite significant screening efforts,
we were not able to further improve the deracemization by
modifying the amine. Fortuitously, we found that the related
rhodium catalyst Λ-RhInd,30 comprising of two cyclo-
metalated 6-tert-butyl-2-phenyl-2H-indazole ligands, in combi-
nation with the base B6, provided an improved ee of 87%
(entry 8). Combining Λ-RhInd with N-phenylpiperidine (B5)
afforded (R)-1 with 92% ee (entry 9). Having identified the
optimal photocatalyst/base pair, we proceeded to finetune the
reaction conditions. Other solvents or a reduced catalyst
loading provided less satisfactory results (entries 10−12).
However, when we decreased the amount of N-phenyl-
piperidine from 3.0 to 2.0 equiv, the ee rose to 94% (entry
13). Finally, optimal results were obtained upon addition of
CaSO4 as a drying agent.31 Blue light irradiation of rac-1 in the
presence of Λ-RhInd (4.0 mol %), N-phenylpiperidine (2.0
equiv), and CaSO4 (5% m/v) in acetone for 24 h provided the
deracemized ketone (R)-1 with 96% ee and 97% isolated yield
Figure 1. Catalytic deracemization of carbonyl compounds in α-
position: motivation, mechanistic plan, and realization.
reaction sequence of base-induced α-deprotonation of a
carbonyl compound to its enolate, followed by enantioselective
protonation, as the equilibrium would always result in the
formation of the entropically favored racemate. To circumvent
the restriction of microscopic reversibility but still exploit well-
established enolate chemistry, we envisioned that the enolate
formation could be split into two elementary steps, a
photoinduced electron transfer19 (+ electron) to generate a
ketyl intermediate, followed by a hydrogen atom transfer (−
hydrogen), thus resulting in a net deprotonation (see frame in
Figure 1b). In this scenario, deprotonation and protonation
would follow distinct pathways, and the enrichment of one
enantiomer would not violate the principle of microscopic
reversibility. Using photons as a driving force,20,21 the
conversion of a racemic carbonyl compound into its single
enantiomer may then be feasible. Herein, we report how
deracemization of ketones can be accomplished by combining
photoredox deprotonation with enantioselective protonation in
a single reaction (Figure 1c).
1
(entry 14). A H NMR yield of 99% also demonstrated that
under these reaction conditions, no photochemical side
reaction occurs. Control experiments verified that both the
amine and light are required for the reaction, while air must be
excluded (entries 15−17). However, catalytic amounts of
amine are sufficient, although the obtained ee of (R)-1
decreased somewhat (entry 18).
Substrate Scope. With optimal reaction conditions in
hand (Table 1, entry 14), we evaluated the substrate scope of
this ketone deracemization. First, we modified the phenyl
moiety of pyridylketone 1 (Scheme 1). A methyl group in the
para-, meta-, or ortho-position (2−4) as well as bulky tert-butyl
(5) or isobutyl (6) groups in the para-position were well-
tolerated and provided the deracemized ketones with 92−96%
ee (90−98% yield). The light-driven deracemization permits
electron-donating substituents in the phenyl moiety such as
para-methoxy (7) and 1,3-dioxole (8) as well as electron-
withdrawing substituents such as para-bromo (9), para- or
ortho-chloro (10, 11), 3-fluoro-4-phenyl (12), and 2,4-difluoro
RESULTS AND DISCUSSION
■
Reaction Optimization. We commenced our study with
the racemic ketone 1, which contains a tertiary stereogenic α-
carbon. A pyridyl moiety is connected to the carbonyl group to
13394
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
a
Table 1. Initial Experiments and Optimization
b
c
entry
catalyst
amine
solvent
conditions
yield (%)
ee (%)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
Λ-IrS (4.0)
Λ-IrS (4.0)
Λ-IrS (4.0)
B1 (3.0)
B2 (3.0)
B3 (3.0)
B3 (3.0)
B4 (3.0)
B5 (3.0)
B6 (3.0)
B6 (3.0)
B5 (3.0)
B5 (3.0)
B5 (3.0)
B5 (3.0)
B5 (2.0)
B5 (2.0)
none
acetone
acetone
acetone
acetone
acetone
acetone
acetone
acetone
acetone
MeCN
THF
acetone
acetone
acetone
acetone
acetone
acetone
acetone
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
photolysis
quant.
quant.
quant.
95
85
96
94
87
94
81
85
90
91
99 (97)
quant.
quant.
quant.
99
0
0
0
Λ-RhS (4.0)
Λ-RhS (4.0)
Λ-RhS (4.0)
Λ-RhS (4.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
Λ-RhInd (2.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
Λ-RhInd (4.0)
26
42
81
84
87
92
82
82
89
94
96
0
d
e
photolysis, CaSO4
photolysis
dark
photolysis, air
photolysis, CaSO4
B5 (3.0)
B5 (3.0)
B5 (0.1)
0
0
90
d
a
Conditions: 1a (0.05 mmol) and amine (0.1−3.0 equiv) in acetone (0.5 mL) with Rh catalyst (2.0−4.0 mol %) irradiated with blue LEDs (24 W)
b
for 24 h at room temperature under an atmosphere of N2, unless stated otherwise. Determined by 1H NMR of the crude products using 1,1,2,2-
tetrachloroethane as internal standard. Enantioselectivities determined by HPLC on chiral stationary phase. CaSO4 (vacuum-dried with heat
gun) with 5% m/v. Isolated yield shown in brackets.
c
d
e
(13) with 91−97% ee (87−95% yield). Only a para-CF3 group
provided a somewhat lower yield of 69% but with satisfactory
96% ee (14). A metal-coordinating thioether was also
compatible and afforded the deracemized ketone 15 with
97% ee (95% yield). Likewise, the double bond of an
isobutenyl group was tolerated and provided the deracemized
ketone 16 with 91% ee (94% yield). The phenyl moiety can
also be benzannulated to efficiently provide the deracemized
naphthalenes 17 (95% ee, 88% yield) and 18 (96% ee, 94%
yield). The thiophene 19 was obtained with 86% ee (93%
yield), while the dibenzofuran 20 was afforded with 95% ee
(93% yield).
Scheme 2 reveals that other parts of the structural scaffold
are also tolerant of modifications. Replacing the α-methyl
group with an ethyl (21) or isopropyl group (22) or having an
indane (23) functionality provided the deracemized ketones
with 86−89% ee (78−83% yield). An ether functionality in α-
position afforded the deracemized ketone 24 in 75% yield and
with only moderate 64% ee. However, α-amino groups are
compatible with the deracemization and afforded the α-
aminoketones 25−27 with 85−92% ee (82−95% yield).
Interestingly, even a fluorine can be incorporated at the
stereogenic carbon. The deracemized ketone 28 was isolated
with 72% ee (93% yield), while the related aliphatic ketone 29
was isolated with only 40% ee (85% yield). Furthermore,
examples 30−34 demonstrate that the pyridine moiety can be
functionalized. A methyl (32) or bromine (33) substituent in
ortho-position of the pyridine provided satisfactory enantiose-
lectivities of 90% and 94% ee, respectively, but required an
increased catalyst loading of 8.0 mol %, most likely due to a
less efficient catalyst coordination to the modified pyridine
moieties. The pyridine can also be replaced with a quinoline
moiety to afford the deracemized ketone 35 with 87% ee (92%
yield), but replacing the pyridine with a pyrazole or imidazole
moiety did not result in any deracemization (see Supporting
Information, Supplementary Figure 2). Finally, ketone 36
13395
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
a
Scheme 1. Substrate Scope with Pyridylketones Bearing an
Aryl Group at the Stereocenter
Scheme 2. Expanded Substrate Scope
a
a
The combination of catalyst and base varied for optimal results. Λ-
RhS (8.0 mol %) as catalyst and 3,5-(tBu)2PhCH2N(iPr)2 (3.0 equiv)
b
as base. Λ-RhS (4.0 mol %) as catalyst and N-phenylpiperidine as
c
base (3.0 equiv). Λ-RhInd (4.0 mol %) as catalyst and N-
d
phenylazepane as base (3.0 equiv). Λ-RhInd (4.0 mol %) as
catalyst, N-phenylpiperidine (3.0 equiv) as base, and CaSO4 (5% m/
a
e
Py = 2-pyridyl. Standard reaction conditions according to Table 1,
v). Λ-RhInd (8.0 mol %) as catalyst, N-phenylpiperidine (3.0 equiv)
b
f
entry 14. Deviation from standard reaction conditions: 3.0 equiv of
N-phenylpiperidine (B5) instead. The absolute configuration of
as base, and CaSO4 (5% m/v). Λ-RhInd (4.0 mol %) as catalyst and
c
N,N-diisopropylethylamine (3.0 equiv) as base.
compound 12 was determined by X-ray crystallography (CCDC
2081804) (see Supporting Information, Supplementary Figure 3,
Supplementary Table 1) and all other compounds were assigned by
analogy.
reveals that a stereocenter that is connected to two aliphatic
side chains provides only an inefficient deracemization.
Mechanistic Studies. The proposed mechanism for the
observed visible-light-driven deracemization of stereocenters in
α-position of carbonyl groups within pyridylketones is shown
in Figure 2. The catalytic cycle begins with the bidentate
coordination of the racemic pyridylketone to the enantiomeri-
cally pure rhodium catalyst to form complex 1I as a mixture of
two diastereomers, (R)-1I and (S)-1I. Photoexcitation to the
triplet state 3I, followed by SET from the tertiary amine,
furnishes the rhodium ketyl radical complex 2II.22,23 Sub-
sequent HAT from the α-position of the ketyl to the amine
radical cation generates rhodium enolate 1III and a protonated
amine, which are primed to undergo a diastereoselective
1
proton transfer to provide the rhodium coordinated ketone I
as a single stereoisomer. Dissociation of the ketone then leads
to a new catalytic cycle.
The independently synthesized and isolated catalyst/
substrate complex 1I displays a significantly increased
absorbance in the visible region compared to the initial
acetonitrile coordinated complex (see Supporting Information,
Supplementary Figure 4). It must therefore constitute the in
situ generated main light-absorbing species in this photo-
Figure 2. Proposed mechanism for the developed photoderacemiza-
tion.
13396
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 3. Supporting mechanistic experiments; bpy = 2,2′-bipyridine.
1
catalysis. Cyclic voltammetry of catalyst/substrate complex I
trace water interferes with the mechanism and provides
was used to determine a reduction potential Epc(1I/1I−) =
−0.90 V vs Ag/AgCl (see Supporting Information, Supple-
mentary Figure 5).32 Together with an estimated triplet energy
of 2.5 eV,33 this leads to E1/2(3I/1I−) = ca. 1.6 V and
support for an enolate protonation mechanism (1III → I).
1
A control experiment in which the photoderacemization was
conducted with α-deuterated tertiary amine and deuterated
acetone did not lead to any deuterium incorporation into the
deracemized ketone, which excludes an alternative reaction
pathway via α-deprotonation of the intermediate amine radical
cation as well as scenarios in which the solvent serves as the
hydrogen atom source (Figure 3c). The involvement of an α-
deprotonation of the intermediate amine radical cation,
otherwise a preferred reaction pathway for amine radical
cations,36 can be excluded based on another experiment in
which N-phenylpiperidine is replaced with triphenylamine,
which does not possess any aliphatic α-C-H bonds (Figure
3d). Triphenylamine can indeed serve as an alternative base,
providing the deracemized ketone with 85% ee (93% yield).
Additional experiments shine further light on the role of the
amine. Replacing the amine with a Hantzsch ester resulted in
efficient formation of the alcohol 37 (Figure 3e). This result
can be rationalized by reduction of the photoexcited catalyst/
3
demonstrates that the photoexcited complex I is capable of
oxidizing aromatic amines (ca. 0.7−1.0 V vs SCE),34 thereby
generating an amine radical cation and the Rh ketyl radical 2II.
The following experiments provide strong support for the
proposed reaction sequence. When the photoderacemization
was performed with racemic ketone rac-1d in which the chiral
center bears a deuterium atom (97% deuteration), the
deuterated product (R)-1d was obtained in 96% yield with
95% ee and a remaining deuteration level of 40% (Figure 3a).
The observed deuterium retention is a key experiment which
verifies that the α-hydrogen of the racemic substrate is
identical to the α-hydrogen of the deracemized product. This is
consistent with our proposal that the amine radical cation
serves both as the HAT agent35 and subsequently as the
1
1
protonating agent (2II → III → I). The somewhat reduced
deuteration level can be attributed to the interference of
residual water in the reaction mixture through direct
protonation of the enolate. When the deracemization of rac-
1 was performed in the presence of just 1 equiv of D2O under
otherwise standard conditions, deracemized (R)-1d was
isolated in 85% yield with 82% ee and a deuteration level of
30% (Figure 3b). This experiment demonstrates that even
3
substrate complex I with the Hantzsch ester to generate the
ketyl intermediate 2II,37 followed by protonation and hydrogen
donation from the Hantzsch ester radical cation to generate
1
the alcohol 37. Enolate III cannot be formed in this case
because although the Hantzsch ester can serve as an electron
donor, it lacks the ability to act as a hydrogen atom acceptor
after its initial oxidation. Finally, we probed the stereo-
13397
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
Figure 4. (a) Calculated free energy diagram of the Λ-RhInd-catalyzed conversion of (S)-1 into (R)-1 at the ωB97X-D/def2-TZVPP−SDD(Rh),
SMD (acetone)//B3LYP-D3BJ/def2-SVP−LANL2DZ(Rh) level of theory. (b) Calculated geometries and relative free energies (kcal/mol) of
intermediates. Interatomic distances are shown in Å.
controlled protonation step 1III → 1I. Starting from the
synthesized catalyst/substrate complex 38 from Δ-RhInd
(S)-1 to Λ-RhInd, with the displacement of two acetonitrile
ligands, leads to photoactive complex (S)-1I. Under visible
light, (S)-1I is excited to the triplet state (S)-3I with a 52.5
kcal/mol increase in free energy. The SET between (S)-3I and
N-phenylpiperidine is exergonic by 13.5 kcal/mol. The
resulting Rh ketyl complex (S)-2II undergoes HAT with the
amine radical cation with an activation free energy of 20.3
kcal/mol. Subsequent intersystem crossing generates closed-
shell Rh enolate 1III. Protonation of Λ-1III by the ammonium
species at the pro-R face gives rise to (R)-1I, the diastereomeric
counterpart of (S)-1I which is 2.9 kcal/mol higher in free
energy. As (R)-1I is less stable, ligand exchange readily occurs
to replace (R)-1 with another molecule of (S)-1 at its
coordination sites, commencing another iteration of the
photoinduced SET/HAT/protonation sequence. Over time,
this leads to the enrichment of (R)-1 in the reaction mixture.
In conjunction with the mechanistic control experiments
(deprotonation/asymmetric protonation sequence, Figure 3f),
our computational results show that two enantioselectivity
filters work in tandem in this reaction to achieve an efficient
overall deracemization with high ee values. First, the
stereoselectivity of the enolate protonation step leads to the
preferential formation of (R)-1I. Second, the favorability of
(S)-1I over (R)-1I ensures that product inhibition does not
occur and the enrichment process of (R)-1 does not arrest
1
(intermediate I in Figure 2), we used the strong base DBU
(1.1 equiv) to generate the enolate complex 39 (intermediate
1III in Figure 2) (Figure 3f). Without any purification, the
enolate 39 was protonated with the hydrochloride salt of N-
phenylpiperidine (5 equiv), followed by release of the
coordinated compound 1 from the rhodium center using
2,2′-bipyridine (10 equiv) as a competing ligand, to afford (S)-
1 in 96% yield and with 50% ee. This result verifies the ability
of enolate 1III to undergo a stereocontrolled protonation (1III
→ 1I in Figure 2). These experiments further support the triple
function of the amine as a single electron donor, hydrogen
atom acceptor, and proton source in sequential fashion within
the same catalytic cycle. It is noteworthy that although
stoichiometric and stepwise deprotonation/protonation allows
for a partial deracemization in this experiment, the rhodium-
catalyzed reaction in the presence of catalytic or stoichiometric
amounts of DBU only leads to racemization of (R)-1, thus
revealing that a base-catalyzed deracemization in the absence
of light is not feasible (Figure 3g, left reaction). However, (R)-
1, which was synthesized from rac-1 using Λ-RhInd, can be
readily converted by photoderacemization to (S)-1 by just
using the mirror-imaged catalyst Δ-RhInd (Figure 3g, right
reaction).
1
itself. The calculated structure of Rh enolate III provides
To provide further support to our proposed mechanism, we
performed density functional theory (DFT) calculations
(Figure 4a). The bidentate coordination of pyridylketone
insight on why the enolate protonation is stereoselective. The
π−π stacking between the enolate phenyl substituent and the
1
indazole ligand is particularly strong in III, with an Ar−Ar
13398
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
centroid distance of only 3.66 Å. This π−π stacking causes the
pro-R face of the enolate to be much more exposed and
accessible for protonation than the pro-S face. Calculated
geometries of (S)-1I and (R)-1I also reveal the structural
features underpinning the favorability of (S)-1I (Figure 4b). In
(S)-1I, stronger π−π stacking is observed between the α-
phenyl substituent of (S)-1 and the indazole ligand. The role of
π−π stacking in energetically differentiating the diastereomeric
(S)-1I and (R)-1I also explains the lower ee values when
ketones with aliphatic α-substituents are used. In addition,
(R)-1I suffers from a steric clash (H···H distance 2.14 Å)
between the α-methyl substituent of (R)-1 and the tert-butyl
group on the ligand. Both of these features explain why (S)-1I
is 2.9 kcal/mol more stable than (R)-1I.
bridge Crystallographic Data Centre, 12 Union Road,
Cambridge CB2 1EZ, UK; fax: +44 1223 336033.
AUTHOR INFORMATION
■
Corresponding Authors
Eric Meggers − Fachbereich Chemie, Philipps-Universität
Marburg, 35043 Marburg, Germany; orcid.org/0000-
0002-8851-7623; Email: meggers@chemie.uni-
marburg.de
Shuming Chen − Department of Chemistry and Biochemistry,
Oberlin College, Oberlin, Ohio 44074, United States;
Email: shuming.chen@oberlin.edu
Authors
Chenhao Zhang − Fachbereich Chemie, Philipps-Universität
Marburg, 35043 Marburg, Germany; orcid.org/0000-
0003-4336-4341
Anthony Z. Gao − Department of Chemistry and
Biochemistry, Oberlin College, Oberlin, Ohio 44074, United
States
Xin Nie − Fachbereich Chemie, Philipps-Universität Marburg,
35043 Marburg, Germany
Chen-Xi Ye − Fachbereich Chemie, Philipps-Universität
Marburg, 35043 Marburg, Germany; orcid.org/0000-
0003-4534-845X
CONCLUSIONS
■
We reported the first example of a photoderacemization of
carbonyl compounds bearing an α-stereogenic center, which
complements previous light-driven deracemizations reported
by Bach28 (allenes) and Knowles29 (N-aryl-imidazolidin-2-
ones). The visible light photolysis of racemic pyridylketones in
the presence of a single photocatalyst and a simple tertiary
amine provides nonracemic carbonyl compounds with up to
97% ee. Mechanistic experiments and DFT calculations
support a novel mechanism through a photoredox deprotona-
tion to an intermediate enolate, followed by a regular
protonation step. A combination of stereoselective protonation
and an enrichment process due to the different Rh-
coordination preference of the two enantiomers results in an
efficient overall deracemization. While catalytic one-step
deracemization through a simple deprotonation/reprotonation
sequence is prohibited by the principle of microscopic
reversibility, we demonstrate that this can be overcome by
splitting the deprotonation step into a photoinduced SET/
HAT sequence. A chiral rhodium catalyst serves a dual
function as the photoredox catalyst and chiral Lewis acid
catalyst, while a tertiary amine cocatalyst plays a triple role
sequentially as the single electron donor, hydrogen atom
acceptor, and proton source within one catalytic cycle.
Innumerable methods exist for introducing substituents into
the α-position of carbonyl compounds. Although the method
introduced herein is currently limited to pyridylketones (for
follow-up chemistry, see Supporting Information, Supplemen-
tary Figure 6), this photochemical deracemization will serve as
a blueprint for other deracemizations of ubiquitous carbonyl
compounds and thus expand the synthetic toolbox for the
synthesis of functionalized nonracemic carbonyl compounds.
Sergei I. Ivlev − Fachbereich Chemie, Philipps-Universität
Marburg, 35043 Marburg, Germany
Complete contact information is available at:
https://pubs.acs.org/10.1021/jacs.1c06637
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
E.M. gratefully acknowledges funding from the German
Research Foundation (ME 1805/17-1). C.Z. thanks the
China Scholarship Council for a stipend. S.C. is thankful to
Oberlin College for financial support. DFT calculations were
performed using SCIURus, the Oberlin College HPC cluster
(NSF MRI 1427949), and startup allocations awarded by
Extreme Science and Engineering Discovery Environment
(XSEDE TG-CHE200100).
REFERENCES
■
(1) Gruber, C. C.; Lavandera, I.; Faber, K.; Kroutil, W. From a
racemate to a single enantiomer: deracemization by stereoinversion.
Adv. Synth. Catal. 2006, 348, 1789.
ASSOCIATED CONTENT
(2) Rachwalski, M.; Vermue, N.; Rutjes, F. P. J. T. Recent advances
in enzymatic and chemical deracemisation of racemic compounds.
Chem. Soc. Rev. 2013, 42, 9268.
(3) Faber, K. Non-sequential processes for the transformation of a
racemate into a single stereoisomeric product: proposal for stereo-
chemical classification. Chem. - Eur. J. 2001, 7, 5004.
(4) Fu, G. C. Transition-metal catalysis of nucleophilic substitution
reactions: a radical alternative to SN1 and SN2 processes. ACS Cent.
Sci. 2017, 3, 692.
(5) Bhat, V.; Welin, E. R.; Guo, X.; Stoltz, B. M. Advances in
stereoconvergent catalysis from 2005 to 2015: transition-metal-
mediated stereoablative reactions, dynamic kinetic resolutions, and
dynamic kinetic asymmetric transformations. Chem. Rev. 2017, 117,
4528.
(6) Eliel, E.; Wilen, S. H.; Doyle, M. P. Basic Organic Stereochemistry;
Wiley-Interscience: New York, 2001.
■
sı
* Supporting Information
The Supporting Information is available free of charge at
https://pubs.acs.org/doi/10.1021/jacs.1c06637.
Supplementary Figures and Tables, experimental
procedures, compound characterization data, NMR
spectra, HPLC traces, crystallographic data, and details
of the computational study (PDF)
Accession Codes
CCDC 2081804 contains the supplementary crystallographic
data for this paper. These data can be obtained free of charge
via www.ccdc.cam.ac.uk/data_request/cif, or by emailing
data_request@ccdc.cam.ac.uk, or by contacting The Cam-
13399
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400

Journal of the American Chemical Society
pubs.acs.org/JACS
Article
(7) Blackmond, D. G. “If pigs could fly” chemistry: a tutorial on the
principle of microscopic reversibility. Angew. Chem., Int. Ed. 2009, 48,
2648.
(8) Tsunoda, T.; Kaku, H.; Nagaku, M.; Okuyama, E. Deracemiza-
tion of 2-alkylcyclohexanones utilizing host-guest molecular associa-
tion with optically active host compounds in basic suspension media.
Tetrahedron Lett. 1997, 38, 7759.
(9) Kaku, H.; Imai, T.; Kondo, R.; Mamba, S.; Watanabe, Y.; Inai,
M.; Nishii, T.; Horikawa, M.; Tsunoda, T. A method to prepare
optically active acyclic α-benzyl ketones by thermodynamically
controlled deracemization. Eur. J. Org. Chem. 2013, 2013, 8208.
(10) Matsumoto, K.; Otsuka, K.; Okamoto, T.; Mogi, H.
Deracemization of acyclic α-hydroxy ketone derivatives by dynamic
resolution using an optically active host compound. Synlett 2007,
2007, 729.
(28) (a) Hölzl-Hobmeier, A.; Bauer, A.; Silva, A. V.; Huber, S. M.;
Bannwarth, C.; Bach, T. Catalytic deracemization of chiral allenes by
sensitized excitation with visible light. Nature 2018, 564, 240.
(b) Plaza, M.; Jandl, C.; Bach, T. Photochemical deracemization of
allenes and subsequent chirality transfer. Angew. Chem., Int. Ed. 2020,
59, 12785. (c) Plaza, M.; Großkopf, J.; Breitenlechner, S.; Bannwarth,
C.; Bach, T. Photochemical deracemization of primary allene amides
by triplet energy transfer: a combined synthetic and theoretical study.
J. Am. Chem. Soc. 2021, 143, 11209.
(29) Shin, N. Y.; Ryss, J. M.; Zhang, X.; Miller, S. J.; Knowles, R. R.
Light-driven deracemization enabled by excited-state electron transfer.
Science 2019, 366, 364.
(30) Steinlandt, P. S.; Zuo, W.; Harms, K.; Meggers, E. Bis-
cyclometalated indazole chiral-at-rhodium catalyst for asymmetric
photoredox cyanoalkylations. Chem. - Eur. J. 2019, 25, 15333.
(31) Other drying agents provided inferior results. See Supporting
Information (Supplementary Figure 1) for more details.
(32) This favorable reduction potential of the ketone can be
attributed to the electron-withdrawing pyridyl group in combination
with a bidentate coordination to the rhodium Lewis acid. See
Supporting Information (Supplementary Figure 5) for more details.
(33) Huang, X.; Lin, J.; Shen, T.; Harms, K.; Marchini, M.; Ceroni,
P.; Meggers, E. Asymmetric [3 + 2] photocycloadditions of
cyclopropanes with alkenes or alkynes through visible-light excitation
of catalyst-bound substrates. Angew. Chem., Int. Ed. 2018, 57, 5454.
(34) Roth, H. G.; Romero, N. A.; Nicewicz, D. A. Experimental and
calculated electrochemical potentials of common organic molecules
for applications to single-electron redox chemistry. Synlett 2016, 27,
714.
(35) Capaldo, L.; Ravelli, D. Hydrogen atom transfer (HAT): a
versatile strategy for substrate activation in photocatalyzed organic
synthesis. Eur. J. Org. Chem. 2017, 2056.
(36) Hu, J.; Wang, J.; Nguyen, T. H.; Zheng, N. The chemistry of
amine radical cations produced by visible light photoredox catalysis.
Beilstein J. Org. Chem. 2013, 9, 1977.
̌
(11) Kosmrlj, J.; Weigel, L. O.; Evans, D. A.; Downey, C. W.; Wu, J.
Unfunctionalized, α-epimerizable nonracemic ketones and aldehydes
can be accessed by crystallization-induced dynamic resolution of
imines. J. Am. Chem. Soc. 2003, 125, 3208.
(12) Fehr, C.; Galindo, J. Catalytic enantioselective protonation of
enolates. Angew. Chem., Int. Ed. Engl. 1994, 33, 1888.
(13) Vedejs, E.; Lee, N.; Sakata, S. T. Deracemization via highly
enantioselective enolate protonation using a chiral aniline as the
“acid". J. Am. Chem. Soc. 1994, 116, 2175.
(14) Ishihara, K.; Nakamura, S.; Kaneeda, M.; Yamamoto, H. First
example of a highly enantioselective catalytic protonation of silyl enol
ethers using a novel Lewis acid-assisted Brønsted acid system. J. Am.
Chem. Soc. 1996, 118, 12854.
(15) Nakamura, S.; Kaneeda, M.; Ishihara, K.; Yamamoto, H.
Enantioselective protonation of silyl enol ethers and ketene disilyl
acetals with Lewis acid-assisted chiral Brønsted acids: reaction scope
and mechanistic insights. J. Am. Chem. Soc. 2000, 122, 8120.
(16) Uraguchi, D.; Kinoshita, N.; Ooi, T. Catalytic asymmetric
protonation of α-amino acid-derived ketene disilyl acetals using P-
spiro diaminodioxaphosphonium barfates as chiral proton. J. Am.
Chem. Soc. 2010, 132, 12240.
(17) Cheon, C. H.; Kanno, O.; Toste, F. D. Chiral Brønsted acid
from a cationic gold(I) complex: catalytic enantioselective proto-
nation of silyl enol ethers of ketones. J. Am. Chem. Soc. 2011, 133,
13248.
(37) Wang, P.-Z.; Chen, J.-R.; Xiao, W.-J. Hantzsch esters: an
emerging versatile class of reagents in photoredox catalyzed organic
synthesis. Org. Biomol. Chem. 2019, 17, 6936.
(18) Mandrelli, F.; Blond, A.; James, T.; Kim, H.; List, B.
Deracemizing α-branched carboxylic acids by catalytic asymmetric
protonation of bis-silyl ketene acetals with water or methanol. Angew.
Chem., Int. Ed. 2019, 58, 11479.
(19) Pandey, G. Photoinduced electron transfer (PET) in organic
synthesis. Top. Curr. Chem. 1993, 168, 175.
(20) McAtee, R. C.; McClain, E. J.; Stephenson, C. R. J. Illuminating
photoredox catalysis. Trends Chem. 2019, 1, 111.
(21) Prentice, C.; Morrisson, J.; Smith, A. D.; Zysman-Colman, E.
Recent developments in enantioselective photocatalysis. Beilstein J.
Org. Chem. 2020, 16, 2363.
(22) Zhang, L.; Meggers, E. Steering asymmetric Lewis acid catalysis
exclusively with octahedral metal-centered chirality. Acc. Chem. Res.
2017, 50, 320.
(23) Huang, X.; Meggers, E. Asymmetric photocatalysis with bis-
cyclometalated rhodium complexes. Acc. Chem. Res. 2019, 52, 833.
(24) Huo, H.; Shen, X.; Wang, C.; Zhang, L.; Röse, P.; Chen, L.-A.;
Harms, K.; Marsch, M.; Hilt, G.; Meggers, E. Asymmetric photoredox
transition-metal catalysis activated by visible light. Nature 2014, 515,
100.
́
(25) Jeffrey, J. L.; Petronijevic, F. R.; MacMillan, D. W. C. Selective
radical−radical cross-couplings: design of a formal β-Mannich
reaction. J. Am. Chem. Soc. 2015, 137, 8404.
(26) Jeffrey, J. L.; Terrett, J. A.; MacMillan, D. W. C. O−H
hydrogen bonding promotes H-atom transfer from α C−H bonds for
C-alkylation of alcohols. Science 2015, 349, 1532.
(27) Ma, J.; Shen, X.; Harms, K.; Meggers, E. Expanding the family
of bis-cyclometalated chiral-at-metal rhodium(III) catalysts with a
benzothiazole derivative. Dalton Trans. 2016, 45, 8320.
13400
https://doi.org/10.1021/jacs.1c06637
J. Am. Chem. Soc. 2021, 143, 13393−13400