Synthesis of Rhodium Complexes with Chiral Diene Ligands via Diastereoselective Coordination and Their Application in the Asymmetric Insertion of Diazo Compounds into E?H Bonds

Article abstract of DOI:10.1002/anie.202105179

A new method for the synthesis of chiral diene rhodium catalysts is introduced. The readily available racemic tetrafluorobenzobarrelene complexes [(R₂-TFB)RhCl]₂ were separated into two enantiomers via selective coordination of one of them with the auxiliary S-salicyl-oxazoline ligand. One of the resulting chiral complexes with an exceptionally bulky diene ligand [(R,R-iPr₂-TFB)RhCl]₂ was an efficient catalyst for the asymmetric insertion of diazoesters into B?H and Si?H bonds giving the functionalized organoboranes and silanes with high yields (79–97 %) and enantiomeric purity (87–98 % ee). The stereoselectivity of separation via auxiliary ligand and that of the catalytic reaction was predicted by DFT calculations.

Full text of DOI:10.1002/anie.202105179

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Synthesis of Rhodium Complexes with Chiral Diene Ligands
via Diastereoselective Coordination and Their Application in the
Asymmetric Insertion of Diazo Compounds into E-H Bonds
Authors: Nikita M. Ankudinov, Denis A. Chusov, Yulia V. Nelyubina,
and Dmitry Perekalin
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.202105179
Link to VoR: https://doi.org/10.1002/anie.202105179

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
Synthesis of Rhodium Complexes with Chiral Diene Ligands via
Diastereoselective Coordination and Their Application in the
Asymmetric Insertion of Diazo Compounds into E-H Bonds
Nikita M. Ankudinov, Denis A. Chusov, Yulia V. Nelyubina, and Dmitry S. Perekalin*
[a]
N. M. Ankudinov, D. A. Chusov, Y. V. Nelyubina, D. S. Perekalin
A.N. Nesmeyanov Institute of Organoelement Compounds, Russian Academy of Sciences
28 Vavilova str., Moscow, Russia
E-mail: dsp@ineos.ac.ru
Supporting information for this article is given via a link at the end of the document
Abstract: A new method for the synthesis of chiral diene rhodium
catalysts is introduced. The readily available racemic tetrafluoro-
benzobarrelene complexes [(R2-TFB)RhCl]2 were separated into two
enantiomers via selective coordination of one of them with the
auxiliary S-salicyl-oxazoline ligand. One of the resulting chiral
complexes with an exceptionally bulky diene ligand [(R,R-iPr2-
TFB)RhCl]2 proved to be an efficient catalyst for the asymmetric
insertion of diazoesters into B−H and Si−H bonds giving the
functionalized organoboranes and silanes with high yields (79−97%)
and enantiomeric purity (87−98% ee). The stereoselectivity of
separation via auxiliary ligand and that of the catalytic reaction was
predicted by DFT calculations.
The application of this method allowed us to synthesize a series
of the new chiral rhodium complexes with unprecedentedly bulky
tetrafluorobenzobarrelene ligands. The resulting complexes
appeared to be efficient catalysts for the asymmetric insertion of
diazo compounds into B−H and Si−H bonds, producing chiral
functionalized organoboranes and silanes with 87-98%
enantiomeric excess. The origin of enantioselectivity and the
concerted mechanism of this reaction were investigated by DFT
calculations.
Introduction
The development of enantioselective catalysis largely relies on a
diversity of the available chiral ligands. Chiral dienes (Scheme 1)
have distinct electronic and steric properties which make them
effective ligands for many reactions.[1,2] In particular, rhodium
complexes with chiral dienes are among the best catalysts for
asymmetric addition of arylboronic acids to the enones, imines,
and other polar double bonds.[3–5] Although these catalysts have
shown a remarkable performance, their accessibility is still rather
limited. Only several chiral dienes are readily available via
transformations of natural terpenes.[6,7] More typically chiral
dienes are obtained via multi-step synthesis involving catalytic
asymmetric reactions, such as hydrosilylation[8] or aldol
condensation.[9] Moreover, expensive preparative chiral
chromatography is necessary for the separation of some of the
most effective ligands.[10–12]
Following an interesting alternative approach, racemic diene
ligands can be initially coordinated with rhodium and then the
resulting complexes can be separated by crystallization of
diastereomeric adducts with R-binaphthyl-diamine.[13–16] However,
it is difficult to predict if this method can be applied for any
particular diene because of the unpredictable nature of the
crystallization process itself. This motivated us to propose and
develop a new approach to the chiral diene complexes based on
their diastereoselective coordination with the auxiliary
salicyloxazoline ligands. The approach has two important
differences from the crystallization method: firstly, only one of two
enantiomers coordinates with an auxiliary ligand, and secondly,
the selectivity of such coordination can be efficiently predicted
before the synthesis by DFT calculations.
Scheme 1. Some examples of the previously reported chiral diene ligands and
our approach to new diene rhodium complexes.
Results and Discussion
Synthesis of catalysts
We chose the tetrafluorobenzobarrelenes (TFB; Scheme 2) as
illustrative ligands for our investigation because they strongly bind
with transition metals and have interesting electron-acceptor
properties.[17] Rhodium complexes with such ligands have been
used as catalysts for a variety of transformations.[10,18–25] However,
the chiral R2-TFB ligands (R = Me, Bn, Ph, Fc) have been
obtained in 4 steps involving cross-coupling of vinyl triflates and
expensive separation of enantiomers by preparative HPLC.[10,26,27]
1
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
We synthesized a series of racemic R2-TFB ligands (1a−d, R =
Me, iPr, tBu) in one step by the Diels-Alder reaction of 1,4-
substituted benzenes with tetrafluoro-dehydrobenzene C6F4,
which was generated from C6F5Br by BuLi similarly to the
literature procedure (Scheme 2).[28] In contrast to the previously
used methods,[10] this approach allows for introduction of bulky
substituents R, including secondary and tertiary groups, which
can improve catalyst selectivity. The structure of the free
tetrafluorobenzobarrelene ligand was established by the X-ray
diffractions for the first time for the particular member tBu2-TFB
(1d).
Interestingly, the homo-chiral dimer R,R-2d had a bent structure,
that is similar to the parent complex [(R,R-Me2-TFB)RhCl]2 (R,R-
2a),[26] while analogous bending was absent in the hetero-chiral
dimer (see SI), possibly because of the steric repulsion of tBu
substituents. The type of dimeric structure, however, had little
influence on the chemical behavior of the complexes, because the
dimers can dissociate into the monomeric species [(R2-TFB)RhCl]
and react as such. For example, the addition of pyridine converted
2d into the monomeric adduct (tBu2-TFB)RhCl(Py) (3). The X-ray
structure of this adduct revealed a notable deviation from the
expected square-planar coordination environment of rhodium
because of the steric repulsion between two tBu substituents on
the one side and chloride and pyridine ligands on the other side.
Figure 1. The X-ray structures of the homo-chiral dimeric complex [(tBu2-
TFB)RhCl]2 (R,R-2d) (top) and the adduct (tBu2-TFB)RhCl(Py) (3) (bottom).
Atoms are shown as 50% thermal ellipsoids; independent molecules and
hydrogen atoms are omitted for clarity. The green dotted line on the structure of
3 corresponds to N-Rh-Cl plane, the red line shows the positions of Cl and Py
ligands expected for the ideal square-planar geometry. Selected distances (Å)
for R,R-2d: Rh−Cl 2.382−2.417, C=C 1.395−1.416; for 3: Rh−Cl 2.3620(14),
Rh−N 2.096(4), C=C 1.373(7) and 1.405(7), cf. to C=C 1.327(2) and 1.335(2)
for 1d.
Scheme 2. Synthesis of the tetrafluorobenzobarrelene ligands and their
rhodium complexes.
The reactions of dienes 1a−d with the commercially available
rhodium precursor [(C2H4)2RhCl]2 led to the clean replacement of
the ethylene ligands and formation of the corresponding racemic
complexes [(R2-TFB)RhCl]2 (2a−d) in good to quantitative yields.
Noteworthy, the most hindered ligand tBu2-TFB (1d) reacted so
slow that heating for several days at 40 °C was necessary to
reach the high yields. Less hindered complexes [(Me2-TFB)RhCl]2
(2a) and [(iPr2-TFB)RhCl]2 (2c) were also obtained by heating the
ligands directly with RhCl3×nH2O in aqueous isopropanol at
80 °C.[29] However, in the case of the bulkier ligand 1d similar
reaction led to the deposition of metallic rhodium, apparently
because coordination of the diene was slower than the metal
reduction.
The steric effects of the TFB ligands were quantitatively assessed
using the steric maps produced by SambVca 2.1 tool[31] for the
calculated structures (R2-TFB)Rh(acac) (Figure 2). The results
clearly showed that the new ligands iPr2-TFB (1c) and tBu2-TFB
(1d) are more bulky (judging by the total buried volume Vburied) and
more asymmetric (judging by the difference between Vburied of the
most and the least occupied quadrants) than the previously used
ligands Me2-TFB[26] and Ph2-TFB.[20] To the best of our knowledge,
The NMR spectra of some of the racemic complexes [(R2-
TFB)RhCl]2 (2a−d) in CDCl3 unexpectedly displayed several sets
of signals apparently due to the formation of dimeric or oligomeric
associates with different chirality of diene ligands.[30] In
accordance with this hypothesis, only one set of signals was
observed for the enantiomerically pure complexes (vide infra). We
managed to establish the structures of both homo- and hetero-
chiral of dimers for the complex [(tBu2-TFB)RhCl]2 (2d) by the X-
ray diffraction analysis (Figure 1).
the tBu2-TFB ligand has the largest total buried volume (Vburied
58%) of all known chiral dienes (compare for example to Vburied
48% for simple 1,5-cyclooctadiene and Vburied = 52% for bulky 2,5-
diphenyl-norbornadiene).[2]
In order to separate the racemic diene complexes, it was
necessary to find a chiral auxiliary ligand L, which would
coordinate selectively with one of the enantiomers of [(R2-
TFB)RhCl]2. To do that we calculated the difference in the
thermodynamic stability of the corresponding diastereomeric
=
=
2
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
adducts (R,R-tBu2-TFB)Rh(L) and (S,S-tBu2-TFB)Rh(L) by DFT
methods for various ligands L including L = S-proline (0.03 kcal
mol−1), N-isopropyl-S-proline (2.77 kcal mol−1), diphenyl-S-
prolinol (1.05 kcal mol−1), diphenylphosphanyl-R-BINOL (0.29
kcal mol−1) and some other (see SI for details). As the result of
this screening, we chose isopropyl-salicyloxazoline (S-Salox) as
one of the most convenient and effective auxiliary ligands (Table
1). S-Salox can be easily obtained in one step[32] from the
commercially available ortho-hydroxybenzonitrile and S-valinol,
which is derived from the naturally chiral amino acid valine. Salox
ligand has been successfully used previously by the group of Prof.
Meggers for chromatographic separation of the diastereomers of
the octahedral coordination compounds with metal-centered
chirality.[33–36]
Table 1. The calculated stabilities of the diastereomeric complexes
(R,R-R2-TFB)Rh(S-Salox) (4) and (S,S-R2-TFB)Rh(S-Salox) (5).
R1,R2
ΔGf° of 4 vs 5
Predicted de (%)
of 4 [b]
Experimental ee
(%) of R,R-2 [c]
[a]
(kcal mol−1
)
Me2 (a)
1.54
2.57
3.90
6.58
86
67
Me, iPr (b)
iPr2 (c)
97
92
>99
>99
>99
>99
tBu2 (d)
[a] Calculated difference between ΔGf° of 4 and 5. [b] Calculated ratio between
the diastereomers 4 and 5 assuming the equilibrium with the constant K =
exp(−ΔG°/RT). [c] Determined by chiral GS of the diene ligand after the
separation (see below).
Experimentally, the target complexes 4a−d can be synthesized by
two possible approaches, differing by whether the starting
rhodium complex is chiral or racemic (Scheme 3). The first
approach is the stereoselective coordination of the chiral rhodium
Salox complex with the racemic tetrafluorobarrelene ligands 1a−d.
An alternative way is the selective coordination of the racemic
diene rhodium complexes 2a−d with the chiral Salox anion.
Figure 2. Steric maps of the model complexes (R2-TFB)Rh(acac), R = Me (top
left), Ph (bottom left), iPr (top right), tBu (bottom right). Vburied indicates the
fraction of space occupied by the ligand in the coordination sphere of the metal.
NW, NE, SW, SE indicate the fraction of space occupied by ligand in the
corresponding quadrants.
DFT calculations at M06L/TZVP level[37] predicted that complexes
(R,R-R2-TFB)Rh(S-Salox) (4) with the matching combination of
R,R-diene and S-Salox ligands are significantly more stable than
the corresponding diastereomers (S,S-R2-TFB)Rh(S-Salox) (5)
with the alternative combination of S,S-diene and S-Salox (Table
1). The main reason for the different stability is the steric repulsion
between R and iPr substituents, which distorts the structure in the
second case. Accordingly, the energy difference between the
diastereomers increases with the size of the substituents R in the
diene ligand from 1.5 kcal mol−1 for R = Me to 6.6 kcal mol−1 for R
= tBu. The calculations also predicted that S-Salox auxiliary ligand
is able to discriminate the enantiomers of the rhodium complexes
with other C2-symmetrical dienes. In particular, the difference in
thermodynamic stability of the corresponding diastereomers (R-
diene)Rh(S-Salox) and (S-diene)Rh(S-Salox) is sufficiently large
for separation of complexes with such dienes as 2,5-Ph2-bicyclo-
2,2,2-octadiene (2.98 kcal mol−1), 1,5-Ph2-cyclooctadiene (4.77
kcal mol−1), 5,11-Ph2-dibenzo-cyclooctatetraene (4.97 kcal mol−1),
as well as some other ligands (see SI for details).
Scheme 3. Two alternative pathways to the chiral complexes
(S,S-R2-TFB)Rh(S-Salox) (4a−d).
Following the first approach, we prepared the rhodium precursors
L2Rh(S-Salox) (L = CO (6), C2H4 (7)) from S-Salox anion and the
corresponding chlorides [L2RhCl]2 (Scheme 4). The carbonyl
complex 6 was rather inert and reacted with an excess of the
diene Me2-TFB (1a) only in the presence of Me3N−O to give a
mixture of diastereomeric complexes 4a and 5a in 6:1 ratio. The
replacement of ethylene in the complex 7 proceeded more readily
but gave the same mixture with the same ratio. The ratio between
diastereomers
was
apparently
determined
by
their
thermodynamic stability rather than by the kinetics of ligand
exchange, because it remained unchanged when the reaction of
7 with Me2-TFB was carried out at lower or higher temperatures
(0 or 45 °C).
3
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
Scheme 4. Coordination of Me2-TFB ligand with Rh-Salox precursors.
Treatment of the resulting mixture of 4a and 5a with HCl produced
the chloride complex R,R-2a in almost quantitative yields and with
67% ee in good accordance with the observed 4a:5a ratio as well
as with the computational predictions (Table 1). Interestingly, the
reaction of the racemic diene Me2-TFB with the rhodium complex
containing less hindered methyl-salicyloxazoline ligand
(C2H4)2Rh(S-Me-Salox)
gave
the
mixture
of
similar
diastereomeric complexes (R,R-Me2-TFB)Rh(S-Me-Salox) and
(S,S-Me2-TFB)Rh(S-Me-Salox) with lower 4:1 ratio because of
the weaker steric interactions (see SI for details). The reactions of
the ethylene complex 7 with the more bulky tetrafluorobarrelene
ligands iPr2-TFB (1c) or tBu2-TFB (1d) were expected to have
much higher stereoselectivity, but unfortunately they were very
slow at 45 °C and led mainly to decomposition products at higher
temperatures.
The second approach appeared to be more fruitful. Thus, the
reaction of the racemic complex [(iPr2-TFB)RhCl]2 (2c) with 1
equivalent of Na[S-Salox] (Rh:Salox ratio = 2:1) selectively
produced only one diastereomeric complex (R,R-iPr2-TFB)Rh(S-
Salox) (4c) while the second enantiomer [(S,S-iPr2-TFB)RhCl]2
(S,S-2c) remained intact (Scheme 5). These complexes can be
easily separated by flash chromatography in the presence of
morpholine, which prevents acid-catalyzed decomposition of 4c
and forms well-separable adduct with S,S-2c.[38] The structure of
4c was confirmed by X-ray diffraction (Figure 3). Both
intermediate complexes can be converted into the respective
chiral chlorides [(R,R-iPr2-TFB)RhCl]2 (R,R-2c) and [(S,S-iPr2-
TFB)RhCl]2 (S,S-2c) by treatment with HCl during the work-up
procedure; the auxilary S-Salox ligand can be also regenerated.
For preparative purposes it was more convenient to carry out the
reaction with a slightly lower amount of Na[S-Salox] (0.9 equiv.).
This way the first enantiomeric complex R,R-2c was directly
obtained with >99% enantiomeric purity in 35% yield (50% is
theoretically possible). The second enantiomer S,S-2c was
initially obtained in 37% yield and with 84% ee, which was then
enhanced to >99% ee by single recrystallization. It is important to
emphasize that unlike most known methods[1,2] our approach
provides easy access to both enantiomers of the chiral diene
complexes. The same procedure allowed us to obtain the more
bulky chiral complexes [(R,R-tBu2-TFB)RhCl]2 (R,R-2d) and
[(S,S-tBu2-TFB)RhCl]2 (S,S-2d) with >99% ee. On the other hand,
the less bulky complexes [(R,R-Me2-TFB)RhCl]2 (S,S-2a) and
[(R,R-Me-iPr-TFB)RhCl]2 (S,S-2b) were obtained with a lower
enantiomeric excess of 67% and 92% in accordance with the
computational prediction (Table 1).
Scheme 5. Separation of enantiomers of [(R2-TFB)RhCl]2 (2a−d) via
diastereoselective coordination with S-Salox. Enantiomeric purity of the
complexes was determined by chiral GS of the corresponding diene ligands
after the separation.
4
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
enantiomeric excess was almost independent of the substituents
in the aromatic ring but correlated with the type ester group. In
particular, tert-butyl esters of diazocompounds gave products
9a−h with 97-98% ee, while ethyl esters gave similar products
9i−p with 91-94% ee. The X-ray diffraction analysis of the
compound 9m confirmed that the catalyst R,R-2c provided R-
enantiomer of the product (Figure 4), which correlated with the
calculated mechanism of the reaction (vide infra). The opposite
enantiomer of the catalyst S,S-2c expectedly produced the
second enantiomer of the product S-9j with almost the same
enantiomeric purity 92% ee. Similar selectivity and
stereochemistry have been previously observed for the closely
related [R,R-(diaryl-bicyclo-[2.2.2]-octadiene)RhCl]2 catalyst.[42]
Table 2. Catalytic insertion of diazo compounds into B−H bonds.
Figure 3. Single-crystal X-ray structure of the complex 4c. Atoms are shown as
50% thermal ellipsoids, hydrogen atoms are omitted for clarity. Selected
distances (Å): Rh−O 2.031(5), Rh−N 2.048(7), C=C 1.390(11) and 1.406(13).
For the demonstration purposes, we also used the developed
method for separation of the racemic rhodium complex 2e with
the known Ph2-TFB ligand.[10] As the result, both enantiomers
R,R-2e and S,S-2e were obtained in good yields and with 96%
and 88% ee, respectively, without application of the preparative
chiral HPLC, that have been used previously in the literature.
#
Ar
R
9
yield (%)[a]
84
ee (%)[b]
98
1
Ph
tBu
tBu
tBu
tBu
tBu
tBu
tBu
tBu
Et
9a
9b
9c
9d
9e
9f
2
4-MeC6H4
4-FC6H4
91
97
3
98
97
Catalytic studies
With new chiral complexes in hand, we decided to demonstrate
their potential for application as catalysts and had chosen the
insertion of diazo compounds into B−H bonds as a showcase
reaction. This reaction is interesting and useful because it
provides the functionalized boranes that are not accessible via
traditional hydroboration. The chiral boranes of this kind can be
transformed into various derivatives, such as amines or alcohols,
with the retention of stereochemical information.[39–41] Despite
these advantages, the asymmetric insertion of diazo compounds
into B-H bonds has been described only recently in several
reports using closely related diene Rh(I) catalysts,[42] Rh(II)
carboxylates,[41] Ru(II) metallacycles,[43] bisoxazoline Cu(I)
complexes,[44,45] or engineered cytochrome enzymes.[46,47]
We started the investigation with the reaction of tert-butyl ester of
phenyldiazoacetate (8a) with the readily available borane adduct
of N-methylpyrrolidine[48] BH3·N(C4H8)Me (Table 2). The tBu-
substituted catalyst R,R-2d was tried first, assuming that bulky
substituents should provide the highest enantioselectivity.
Surprisingly, it reacted very slowly with diazocompound 8a even
at 50 °C, apparently due to the steric overload. Although such
hindrance may be excessive for catalytic reactions it may be also
helpful for isolation and investigation of reactive intermediates.[49]
Fortunately, the iPr-substituted catalyst R,R-2c (1 mol%)
promoted the reaction of 8a with BH3·N(C4H8)Me in DCE already
at room temperature giving the desired organoborane product 9a
in 84% yield and with 98% ee. Investigation of the substrate scope
demonstrated that this reaction is compatible with both electron-
donating (Me, Ph, OMe) and electron-withdrawing (NO2, Ac, F,
CF3) substituents in the arene ring. The process was a bit slower
in the case of sterically hindered substrates such as 2-
methylphenyl or 1-napthyl diazoacetates (Table 2, entries 7, 8),
which required the use of 1.5 mol% of the catalyst. The
4
4-ClC6H4
4-MeOC6H4
3,4-(MeO)2C6H3
2-MeC6H4
1-naphtyl
4-NO2C6H4
3-NO2C6H4
4-PhC6H4
4-AcC6H4
4-BrC6H4
4-CF3C6H4
3,5-F2C6H3
2-fluorenyl
3-NO2C6H4
89
98
5
85
97
6
84
98
7[c]
8[c]
9
9g
9h
9i
88
98
83
89
86
91
10
11
12
13
14
15
16
17[d]
Et
9j
98
92
Et
9k
9l
93
92
Et
89
92
Et
9m
9n
9o
9p
S-9j
87
94
Et
94
91
Et
97
93
Et
95
93
Et
99
92
[a] Isolated yield; [b] Determined by chiral HPLC; [c] 1.5 mol% of R,R-2c
catalyst; [d] 1 mol% of S,S-2c catalyst.
Next, we briefly investigated the catalytic activity of the complex
R,R-2c for analogous insertion of diazocompounds 8 into Si−H
bond of triethylsilane (Table 3).[50] This reaction proceeded
somewhat slower than the one with boranes but nevertheless
gave the corresponding organosilane products 10a−e in good
yields and with high enantiomeric purity. Again the selectivity was
5
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
influenced by the ester group and rose from 87% ee for the methyl
ester 10a to 97% ee for the tert-butyl ester 10e. This similarity with
B−H insertion suggested that the stereoselectivity was mainly
determined by the chirality of a certain carbene intermediate,
while the nature of the RnE−H reagent determined only the speed
of the reaction. Accordingly, the more hindered and less
nucleophilic triethylsilane reacted slower than amino borane.
strongly improved the stereoselectivity of the insertion giving the
carbazole derivative 11c with 72% ee. tBuOH itself did not react
with 8a because of the steric hindrance. At the same time, the
less bulky 2-naphthol did react with 8a in the presence of R,R-2c
but gave only the racemic insertion product 2-(2-naphthyloxy)-
phenylacetate.
Figure 4. Single-crystal X-ray structure of the borane insertion product 9m
which confirms the stereochemistry of the catalytic reaction. Atoms are shown
as 50% thermal ellipsoids, except hydrogens shown as spheres.
Scheme 6. Catalytic insertion of diazo compounds into N−H bonds.
The effect of tBuOH and the differences between the insertion of
diazo compounds into B−H, Si−H, N−H, and O−H bonds can be
explained by the unified mechanism proposed in accordance with
generally accepted concepts[51–53] (Scheme 7). First the dimeric
chloride complex R,R-2c dissociates into monomeric species and
reacts with diazoacetate to give the rhodium-carbene
intermediate IM1. Such carbenes have electrophilic character[54]
and therefore the reaction of the carbene IM1 with B−H and Si−H
compounds proceeds as the attack of nucleophilic hydride on the
carbon atom followed by the concerted asynchronous formation
of the C−B or C−Si bond.[55] On the other hand, the reaction with
N-H compounds proceeds as the attack of a nucleophilic nitrogen
atom on the carbon atom of IM1 giving the zwitterion intermediate
IM2.[52] Intra- or intermolecular proton transfer in IM2 can lead to
the opposite enantiomer of the insertion product as it was
observed experimentally. In addition, if the proton transfer is
hindered, the intermediate IM2 can undergo dissociation with the
formation of enol-type species and the loss of stereochemical
information. Apparently, this is why the stereoselectivity of the
reaction was improved by the addition of an external proton
source such as tBuOH. Further improvement may be possible by
finding the more efficient proton transfer reagents (see SI for
additional results).[56,57]
We were interested to see whether the DFT calculations can
support the proposed mechanism and correctly predict the
observed enantioselectivity of the reaction. According to
calculations at M06L/TZVP level with solvent corrections for DCE
(Scheme 8) the presumed active species [(R,R-iPr2-TFB)RhCl]
can reversibly react with diazoester 8a to give one of two possible
adducts R-I or S-I.[58] Further removal of N2 from these
intermediates leads to the alternative carbene complexes R-II or
S-II, which have similar stability but the opposite configuration of
the rhodium-bound carbenoid center. The activation barrier for the
formation of R-II is 1.8 kcal mol−1 lower than that for S-II. This
energy difference corresponds to the ratio of R-II:S-II ≈ 95:5,
which qualitatively correlates with the experimental formation of
Table 3. Catalytic insertion of diazo compounds into Si−H bonds.
#
Ar
R
10
yield (%)[a]
ee (%)[b]
87
1
Ph
Me
Et
10a
10b
10c
10d
10e
S-10b
83
82
82
78
89
82
2
4-PhC6H4
3-NO2C6H4
4-AcC6H4
3,4-(MeO)2C6H3
4-PhC6H4
91
3
Et
96
4
Et
93
5
tBu
Et
97
6[c]
88
[a] Isolated yield; [b] Determined by chiral HPLC; [c] 1 mol% of S,S-2c catalyst.
This hypothesis seemed to be correct for the insertion of diazo
compounds into nonpolar E−H bonds such as B−H and Si−H.
However, the catalytic reactions of diazoester 8a with N−H
compounds were markedly different (Scheme 6). In particular, the
insertion of 8a into N−H bond of benzamide proceeded cleanly in
DCE in the presence of 1 mol% of the catalyst R,R-2c and gave
the desired product 11a in good yield 70% but without any
enantiomeric excess. The analogous reactions with phthalimide
and carbazole were less clean and gave the target products 11b
and 11c in lower yields but with higher stereoselectivity 33 and
42% ee, respectively. Noteworthy, the enantioselectivity was
reversed: in contrast to R-boranes 9 and R-silanes 10, S-
enantiomer of the amide 11b was the major product. It is also
interesting to note that the addition of tBuOH as co-solvent (10%)
6
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
the products R-9 with >90% ee. The simplified explanation for the
difference in activation barriers is the unfavorable steric repulsion
in the transition state S-TS1 between iPr substituent of the diene
ligand and the CO2tBu group of the diazo ester (see SI for detailed
analysis of non-covalent interactions). The formation of carbene
complexes R-II and S-II is essentially irreversible because of the
nitrogen evolution and therefore it determines the
stereoselectivity of the whole process.
Scheme 7. Proposed mechanism for the catalytic insertion reactions.
For steric reasons the addition of the borane BH3·N(C4H8)Me to
the carbene complexes R-II and S-II can proceed only from one
side of the Rh=C bond, which is not blocked by the diene ligand.
In accordance with the proposed model, the transfer of a hydride
from BH3 to the carbene carbon occurs first and has a reasonably
low activation barrier of 18.5 kcal mol−1 (R-TS2) or 22.7 kcal mol−1
(S-TS2).[59] Subsequent formation of C−B bond has essentially no
barrier according to the potential surface scan and therefore no
racemization can occur during this step. Thus the catalyst R,R-2c
was correctly predicted to provide R-enantiomer of the product 9a.
In contrast to the borane, the reaction of benzamide with the
carbene complex R-II proceeds in a stepwise manner (Scheme
9). The initial nucleophilic attack of the nitrogen[60] via transition
state R-TS3 with 18.6 kcal mol−1 barrier leads to the zwitterion
intermediate R-IV, which is stabilized by the hydrogen bond
between NH and COOtBu groups. Subsequent proton transfer
from nitrogen to oxygen via R-TS4 gives complex R-V with weakly
coordinated enol form of the product. The transfer of proton to the
carbon atom of R-V can occur only from the side that is opposite
to the rhodium atom and it would lead to the S-enantiomer of the
product 11a. However, the dissociation of R-V apparently
proceeds faster than the proton transfer. It regenerates the active
species [(R,R-iPr2-TFB)RhCl] and gives the racemic form of the
product 11a in accordance with the proposed mechanism on
Scheme 7 and experimental observations.
Scheme 8. Free energy diagram of the insertion of diazoester 8a into B−H bond
catalyzed by the active species [(R,R-iPr2-TFB)RhCl] (at M06L/TZVP level).
7
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
Federation using the equipment of the Center for molecular
composition studies of INEOS RAS.
Keywords: rhodium • asymmetric catalysis • diene ligands •
DFT calculations • carbene insertion
References
[1]
[2]
C. Defieber, H. Grutzmacher, E. M. Carreira, Angew. Chem. Int. Ed. 2008,
47, 4482–4502; Angew. Chem. 2008, 120, 4558–4579.
M. Hirano, N. Komine, E. Arata, T. Gridneva, A. Hatori, N. Kaizawa, K.
Kamakura, A. Kuramochi, S. Kurita, S. Machida, H. Okada, A. Sawasaki,
T. Uchino, Tetrahedron Lett. 2019, 60, 150924.
[3]
M. M. Heravi, M. Dehghani, V. Zadsirjan, Tetrahedron Asymmetry 2016,
27, 513–588.
[4]
[5]
T. Hayashi, K. Yamasaki, Chem. Rev. 2003, 103, 2829–2844.
T. Yasukawa, H. Miyamura, S. Kobayashi, Acc. Chem. Res. 2020, 53,
2950–2963.
[6]
[7]
K. Okamoto, T. Hayashi, V. H. Rawal, Org. Lett. 2008, 10, 4387–4389.
C. Fischer, C. Defieber, T. Suzuki, E. M. Carreira, J. Am. Chem. Soc.
2004, 126, 1628–1629.
Scheme 9. Free energy diagram of the insertion of the carbene complex R-II
into N−H bond of benzamide (at M06L/TZVP level).
[8]
[9]
T. Hayashi, K. Ueyama, N. Tokunaga, K. Yoshida, J. Am. Chem. Soc.
2003, 125, 11508–11509.
S. Abele, R. Inauen, D. Spielvogel, C. Moessner, J. Org. Chem. 2012,
77, 4765–4773.
Conclusion
[10] T. Nishimura, H. Kumamoto, M. Nagaosa, T. Hayashi, Chem. Commun.
2009, 5713.
In summary, we developed a new approach to the rhodium
complexes with chiral diene ligands. The method is based on the
diastereoselective coordination of one enantiomer from the
racemic mixture of the complexes with a chiral auxiliary ligand,
which allows for their further separation. Although somewhat
similar methods have been employed previously,[35,61–63] the
potential of this approach is clearly undeveloped and it can be
applied to the complexes of metals other than rhodium and
ligands other than dienes. The DFT calculations are especially
helpful for this method because they allow one to screen a variety
of auxiliary ligands and estimate the selectivity of coordination
with sufficient accuracy before the actual synthesis.
[11] Y. Otomaru, N. Tokunaga, R. Shintani, T. Hayashi, Org. Lett. 2005, 7,
307–310.
[12] Y. Otomaru, A. Kina, R. Shintani, T. Hayashi, Tetrahedron: Asymmetry
2005, 16, 1673–1679.
[13] A. Kina, K. Ueyama, T. Hayashi, Org. Lett. 2005, 7, 5889–5892.
[14] F. Läng, F. Breher, D. Stein, H. Grützmacher, Organometallics 2005, 24,
2997–3007.
[15] M.-C. Melcher, T. Ivšić, C. Olagnon, C. Tenten, A. Lützen, D. Strand,
Chem. Eur. J. 2018, 24, 2344–2348.
[16] M.-C. Melcher, B. Rolim Alves da Silva, T. Ivšić, D. Strand, ACS Omega
2018, 3, 3622–3630.
[17] M. A. Esteruelas, L. A. Oro, Coord. Chem. Rev. 1999, 193–195, 557–
618.
New chiral rhodium complexes with exceptionally bulky
tetrafluoro-benzobarrelene ligands were synthesized and used as
catalysts for asymmetric insertion of diazo esters into B−H and
Si−H bonds. A number of boranes and silanes with various
functional groups were produced in high yields (79−97%) and
enantiomeric purity (87−98% ee). Analogous catalytic insertion of
diazo esters into N−H bonds gave the products with opposite
chiral configuration thus providing an experimental evidence for
the unified mechanism of such transformations. The DFT
calculations suggested that enantioselectivity was determined by
the structure of the intermediate carbene complex and more
specifically by the steric interactions between the ester group and
the substituents in the diene ligand. Hopefully, the new rhodium
catalysts will find application in other asymmetric transformations.
[18] T. Nishimura, H. Makino, M. Nagaosa, T. Hayashi, J. Am. Chem. Soc.
2010, 132, 12865–12867.
[19] T. Nishimura, A. Noishiki, T. Hayashi, Chem. Commun. 2012, 48, 973–
975.
[20] R. Takechi, T. Nishimura, Chem. Commun. 2015, 51, 8528–8531.
[21] Y. Pan, X. Lu, H. Qiu, T. Hayashi, Y. Huang, Org. Lett. 2020, 22, 8413–
8418.
[22] Y. Huang, T. Hayashi, J. Am. Chem. Soc. 2016, 138, 12340–12343.
[23] Y. Huang, T. Hayashi, J. Am. Chem. Soc. 2015, 137, 7556–7559.
[24] K. M.-H. Lim, T. Hayashi, J. Am. Chem. Soc. 2015, 137, 3201–3204.
[25] M. Umeda, K. Sakamoto, T. Nagai, M. Nagamoto, Y. Ebe, T. Nishimura,
Chem. Commun. 2019, 55, 11876–11879.
[26] T. Nishimura, Y. Ichikawa, T. Hayashi, N. Onishi, M. Shiotsuki, T.
Masuda, Organometallics 2009, 28, 4890–4893.
[27] Separation by crystallization of menthol ethers was also used, but
apparently has been abandoned: T. Nishimura, M. Nagaosa, T. Hayashi,
Chem. Lett. 2008, 37, 860–861.
[28] S. C. Cohen, M. L. N. Reddy, D. M. Roe, A. J. Tomlinson, A. G. Massey,
J. Organomet. Chem. 1968, 14, 241–251.
[29] D. S. Perekalin, A. V Kolos, Mendeleev Commun. 2021, 31, 1–7.
[30] Similar situation has been observed previously for rhodium complexes
with diferent chiral ligands: M. Mayr, C. J. R. Bataille, S. Gosiewska, J.
A. Raskatov, J. M. Brown, Tetrahedron Asymmetry 2008, 19, 1328–1332.
[31] L. Falivene, Z. Cao, A. Petta, L. Serra, A. Poater, R. Oliva, V. Scarano,
L. Cavallo, Nat. Chem. 2019, 11, 872–879.
Acknowledgements
We thank Alexey A. Tsygankov and Dr. Dmitry V. Muratov for their
help with analytical measurements. This work was supported by
the Russian Science Foundation (grant no. 17-73-20144). X-ray
diffraction data were collected with financial support from the
Ministry of Science and Higher Education of the Russian
[32] D. Franco, M. Gómez, F. Jiménez, G. Muller, M. Rocamora, M. A.
Maestro, J. Mahía, Organometallics 2004, 23, 3197–3209.
8
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
[33] L. Gong, M. Wenzel, E. Meggers, Acc. Chem. Res. 2013, 46, 2635–2644.
[34] J. Ma, X. Zhang, X. Huang, S. Luo, E. Meggers, Nat. Protoc. 2018, 13,
605–632.
[35] E. Meggers, Chem. Eur. J. 2010, 16, 752–758.
[36] E. Meggers, Eur. J. Inorg. Chem. 2011, 2011, 2911–2926.
[37] D. G. Gusev, Organometallics 2013, 32, 4239–4243.
[38] Noteworthy, the chromatographic separation of the diastereomers 4a
and 5a was not successful.
[39] V. K. Aggarwal, G. Y. Fang, A. T. Schmidt, J. Am. Chem. Soc. 2005, 127,
1642–1643.
[40] J. R. Fulton, V. K. Aggarwal, J. de Vicente, Eur. J. Org. Chem. 2005,
2005, 1479–1492.
[41] Y. Pang, Q. He, Z.-Q. Li, J.-M. Yang, J.-H. Yu, S.-F. Zhu, Q.-L. Zhou, J.
Am. Chem. Soc. 2018, 140, 10663–10668.
[42] D. Chen, X. Zhang, W.-Y. Qi, B. Xu, M.-H. Xu, J. Am. Chem. Soc. 2015,
137, 5268–5271.
[43] N. Otog, S. Chanthamath, I. Fujisawa, S. Iwasa, Eur. J. Org. Chem. 2021,
2021, 1564–1567.
[44] Q. Cheng, S. Zhu, Y. Zhang, X. Xie, Q. Zhou, J. Am. Chem. Soc. 2013,
135, 14094–14097.
[45] S. Hyde, J. Veliks, B. Liégault, D. Grassi, M. Taillefer, V. Gouverneur,
Angew. Chem. Int. Ed. 2016, 55, 3785–3789; Angew. Chem. 2016, 128,
3849–3853.
[46] S. B. J. Kan, X. Huang, Y. Gumulya, K. Chen, F. H. Arnold, Nature 2017,
552, 132–136.
[47] K. Chen, X. Huang, S.-Q. Zhang, A. Z. Zhou, S. B. J. Kan, X. Hong, F. H.
Arnold, Synlett 2019, 30, 378–382.
[48] This type of borane adduct was chosen to achieve better separation of
enantiomers of the products by analytical HPLC. Similar results were
obtained with more common and commercially available BH3·NEt3.
[49] B. Dutta, B. F. E. Curchod, P. Campomanes, E. Solari, R. Scopelliti, U.
Rothlisberger, K. Severin, Chem. Eur. J. 2010, 16, 8400–9.
[50] For the closely related reaction see: D. Chen, D.-X. Zhu, M.-H. Xu, J. Am.
Chem. Soc. 2016, 138, 1498–1501.
[51] H. M. L. Davies, D. Morton, Chem. Soc. Rev. 2011, 40, 1857.
[52] D. Gillingham, N. Fei, Chem. Soc. Rev. 2013, 42, 4918–4931.
[53] H. Keipour, V. Carreras, T. Ollevier, Org. Biomol. Chem. 2017, 15, 5441–
5456.
[54] C. Werlé, R. Goddard, P. Philipps, C. Farès, A. Fürstner, J. Am. Chem.
Soc. 2016, 138, 3797–3805.
[55] X. Li, D. P. Curran, J. Am. Chem. Soc. 2013, 135, 12076–12081.
[56] B. Xu, S. Zhu, X. Xie, J. Shen, Q. Zhou, Angew. Chem. Int. Ed. 2011, 50,
11483–11486; Angew. Chem. 2011, 123, 11685–11688.
[57] Y.-Y. Ren, S.-F. Zhu, Q.-L. Zhou, Org. Biomol. Chem. 2018, 16, 3087–
3094.
[58] The formation of the adducts R-I and S-I from [(S,S-iPr2-TFB)RhCl] and
diazoester 8a has favorable enthalpy but unfavorable entropy.
[59] The difference between activation barriers of R-TS2 and S-TS2 also
favors the formation of R-enantiomer of the product. However, we
assume that it is less relevant because the formation of carbene
intermediates R-II and S-II is irreversible and their ratio is fixed by the
difference of R-TS1 and S-TS1.
[60] Alternative attack of carbene by the hydrogen atom of benzamide has at
least 47 kcal mol−1 activation barrier according to the potential energy
scan.
[61] Z. Zhou, Y. Tan, T. Yamahira, S. Ivlev, X. Xie, R. Riedel, M. Hemming,
M. Kimura, E. Meggers, Chem 2020, 6, 2024–2034.
[62] M. A. Bau, S. Wiesler, S. L. Younas, J. Streuff, Chem. Eur. J. 2019, 25,
10531–10545.
[63] T. Punniyamurthy, M. Mayr, A. S. Z. Dorofeev, C. J. R. Bataille, S.
Gosiewska, B. Nguyen, A. R. Cowley, J. M. Brown, Chem. Commun.
2008, 5092–5094.
9
This article is protected by copyright. All rights reserved.

10.1002/anie.202105179
Angewandte Chemie International Edition
RESEARCH ARTICLE
Entry for the Table of Contents
New rhodium complexes with bulky racemic dienes were synthesized and separated into pure enantiomers by diastereoselective
coordination with the chiral auxiliary ligand. The optimal structure of the auxiliary ligand and the selectivity of coordination were predicted
by DFT calculations. The resulting chiral complexes catalyze the asymmetric insertion of diazoesters into B−H and Si−H bonds.
10
This article is protected by copyright. All rights reserved.