Nucleophilic Isomerization of Epoxides by Pincer-Rhodium Catalysts: Activity Increase and Mechanistic Insights

Article abstract of DOI:10.1002/cctc.201900594

Herein, we present the efficient isomerization of epoxides into methyl ketones with a novel pincer-rhodium complex under very mild conditions. The catalyst system has an excellent functional group tolerance and a wide array of epoxides was tested. The corresponding methyl ketones were obtained in very high yields with excellent chemo- and regioselectivity. In addition, we investigated mechanistic details like the isomerization of the catalyst, and we obtained evidence that the catalytic cycle follows a β-hydride elimination-reductive elimination pathway after the nucleophilic ring opening of the epoxide.

Full text of DOI:10.1002/cctc.201900594

Full Papers
DOI: 10.1002/cctc.201900594
1
2
3
4
5
6
7
8
9
Nucleophilic Isomerization of Epoxides by Pincer-Rhodium
Catalysts: Activity Increase and Mechanistic Insights
Yingying Tian,[a] Eva Jürgens,[a] Katharina Mill,[a] Ronja Jordan,[a] Theo Maulbetsch,[a] and
Doris Kunz*[a]
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Herein, we present the efficient isomerization of epoxides into
methyl ketones with a novel pincer-rhodium complex under
very mild conditions. The catalyst system has an excellent
functional group tolerance and a wide array of epoxides was
tested. The corresponding methyl ketones were obtained in
very high yields with excellent chemo- and regioselectivity. In
addition, we investigated mechanistic details like the isomer-
ization of the catalyst, and we obtained evidence that the
catalytic cycle follows a β-hydride elimination-reductive elimi-
nation pathway after the nucleophilic ring opening of the
epoxide.
Introduction
tion of [Co2(CO)8] in methanol that is required as a solvent. In
2015 we have reported on a selective ring opening using the
nucleophilic Rh-catalyst B (5 mol%) along with 20 mol% of the
In contrast to the common Lewis acid catalyzed isomerization
of terminal epoxides (Meinwald reaction),[1] which leads to the
formation of aldehydes via generation of the more stable
carbenium intermediate,[2] the isomerization by nucleophilic
catalysts is rare and results in the formation of methyl ketones
(Scheme 1).[3] This pathway is very attractive because, in
combination with the Corey-Chaykovsky reaction, it provides an
oxidation-free route for the synthesis of methyl ketones from
aldehydes or, together with the epoxidation of olefins, an
alternative to the Wacker oxidation. Since the discovery of this
opposite regioselectivity in 1962 by Eisenmann,[3a] it was
evident that the catalysis is not only achieved by a nucleophilic
catalyst on its own, but typically requires a Lewis acid co-
catalyst [LA+] which pre-activates the epoxide by coordination,
but does not lead to ring opening under carbocation formation.
In case of the results of Eisenmann, the catalytic system A
(Figure 1) most likely consists of the [Co(CO)4]– nucleophile and
the Lewis acid [Co(CH3OH)6]2+ formed in situ by disproportiona-
[3h]
°
Lewis acid co-catalyst LiNTf2 at 60 C. A more active system
was reported by Coates using [Al(salen)][Co(CO)4].[3i] While both
systems were highly selective with terminal alkyl epoxides, the
isomerization of phenyl oxirane led to a mixture of the
aldehyde and the ketone in almost equal amounts. This
prompted us to develop the more nucleophilic CO-free rhodium
catalyst D, in which the highly reactive metal center is
intramolecularly stabilized by coordination of both N-homoallyl
m
substituents of the so-called bimcaH
(1,8-bis(imidazolin-2-
° °
ylidene)-3,6-di(tert-butyl)carbazolide) carbene-pincer-ligand.[3j]
Due to its higher nucleophilicity, a mixture of the weak Lewis
acids LiBr and LiCl (2:1), which are present from the synthesis
of D in THF, is sufficient to act as a co-catalyst in benzene. With
Scheme 1. Lewis acid- (left) and nucleophile- (right) catalysed Meinwald
reaction of terminal epoxides: formation of aldehydes versus methyl
ketones.
[a] Y. Tian, E. Jürgens, K. Mill, R. Jordan, T. Maulbetsch, Prof. Dr. D. Kunz
Institut für Anorganische Chemie
Eberhard Karls Universität Tübingen
Auf der Morgenstelle 18, 72076 Tübingen (Germany)
E-mail: Doris.Kunz@uni-tuebingen.de
Supporting information for this article is available on the WWW under
https://doi.org/10.1002/cctc.201900594
Figure 1. Rhodium pincer complexes and suitable Lewis acid co-catalysts for
the nucleophilic Meinwald rearrangement.
This manuscript is part of the Special Issue dedicated to the Women of
Catalysis.
ChemCatChem 2019, 11, 1–9
1
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
this system it was not only possible to achieve a selectivity of
methyl ketone vs. aldehyde of 40:1 and 95% yield in the
isomerization of phenyl oxirane after 2 h at room temperature
(5 mol% D), but also to isomerize the highly sensitive p-
methoxyphenyl oxirane to the corresponding ketone with a
ratio of 10:1 and in 57% yield.
1
2
3
4
5
6
The functional group tolerance of catalyst D is very high,
but sterically demanding substrates such as tert-butyloxirane,
À NH2, or À OH containing substrates, some ortho-substituted
aryl oxiranes or internal epoxides required very long reaction
times and/or elevated temperatures, which in some cases still
led to low yields. Therefore, we were interested in developing
our catalyst system further along with gaining more insight into
the reaction mechanism.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
As the 18 e– rhodium complex D requires the dissociation of
one olefin moiety to react as a nucleophile, we envisaged the
16 e– complex 4 being an interesting target to fulfill our
purpose: The complex would still be stabilized intramolecularly
by one olefin moiety during catalysis but in addition exhibit a
higher nucleophilicity and thus catalytic activity. However, an
unsymmetrically N-substituted bimca ligand or related systems
have hitherto not yet been reported. In the following, we will
report on the synthesis of the unsymmetrical bimca ligand
bimcaMe,Homo and its Rh complex 4 as well as the catalytic
activity of 4 in the regioselective epoxide isomerization. In
addition, more insight into the reaction mechanism and
deactivation pathways will be provided.
Scheme 2. Synthesis of the unsymmetrical rhodium catalyst 4.
The preparation of 4LiX was achieved by deprotonation of 3
with LiHMDS and subsequent addition of [Rh(μ-Cl)(COD)]2.
However, the generated COD and HMDS during the reaction
cannot be removed in vacuo without decomposition of the
complex, possibly due to the presence of the lithium salts. To
obtain the pure complex 4, we transmetalated the ligand from
the potassium complex [K(bimcaMe,Homo)] generated in situ by
deprotonation of 3 with KHMDS. The potassium halides formed
could be readily removed by filtration and complex 4 isolated
by removal of all volatiles in vacuo.
Results and Discussion
The structure of complex 4 could be identified by NMR
experiments. Bonding of the terminal double bond to the
1
The synthesis of an unsymmetrically substituted bimca ligand
requires the selective monoalkylation of bisimidazol 1. Based on
the observation that the monoalkylated product 2 was formed
as a byproduct when reacting bisimidazol 1 with an excess
methyl iodide in THF,[4] we investigated the monoalkylation
further. In acetonitrile the reaction of equimolar amounts of
bisimidazol 1 and methyl iodide or Meerwein’s salt (Me3OBF4)
leads to a 1:1 mixture of the dialkylated product and the
starting material 1. However, if the reaction is carried out with
even a twofold excess of methyl iodide in THF at room
temperature the desired product 2 is formed due to its
precipitation under these conditions, which prevents further
alkylation (Scheme 2). After workup by filtration, the imidazo-
lium salt 2 was obtained in 90% yield as a yellow solid. The
reduced symmetry leads to four signals for the carbazole
moiety and six signals for the imidazolium moieties. The
presence of only one NH signal with the same integral as the
other aromatic peaks excludes a potential 1:1 mixture of 1 and
the dimethylated product. The introduction of the N-homoallyl
substituent was carried out with a twofold excess of 4-bromo-1-
rhodium centre is proven by the JRhC coupling constants of
13.2 Hz (C16) and 13.7 Hz (C17) in the 13C NMR spectrum and
the 4 signals for the diastereotopic hydrogen atoms at the two
methylene groups of the homoallyl substituent. The cis/trans
assignment of the olefinic signals at 3.50 ppm (3J=7.7 Hz, H-
17cis) and 2.94 ppm (3J=11.5 Hz, H-17trans) is based on the vicinal
coupling constants, which are smaller than those typically
observed in non-coordinated olefins (e.g. in compound 3: 3JHH
=
17.2 Hz (trans) and 10.3 Hz (cis)), and can be explained with the
lower s-character due to a rehybridization upon coordination of
the metal ion. The relative assignment of axial and equatorial
methylene signals is based on the Karplus equation. Finally,
information on the relative conformation of the metallacycle
were obtained with an NOE experiment, which confirmed the
aforementioned assignments. Out of several minimum confor-
mations of similar energy obtained by DFT-calculations, the
NOE data is in accordance with the structure that is second
lowest in energy (Δ=1.5 kJ/mol; Figure 2 and Supporting
Information).
With the new rhodium complex 4 in hand, we were able to
investigate the isomerization of phenyl oxirane (5a) as the
model substrate. To our satisfaction, an initial test using 5 mol%
of complex 4 together with 10 mol% LiBr (pre-activated with
5 μL THF-d8) in C6D6 revealed full conversion at room tempera-
ture in a short time and demonstrated that the catalyst was
more active than the previous catalyst D (Table 1, entry 1).
°
butene at 110 C in dimethylformamide and the desired
bisimidazolium salt 3 was isolated after workup in 95% yield.
The 10 signals for the aromatic imidazolium and carbazole
protons can be clearly assigned with the help of 2D NMR
experiments including an NOE experiment to assign the peaks
to the relative sides of the carbazole backbone.
ChemCatChem 2019, 11, 1–9
www.chemcatchem.org
2
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
Using complex 4 as the catalyst, various functionalized
epoxides were converted successfully into the desired methyl
ketones (Table 2). Phenyl oxirane (5a) and aryl oxirane 5b
bearing the strong electron-withdrawing trifluoromethyl sub-
stituent were transformed regioselectively in excellent yield. To
our delight, epoxides 5c and 5d possessing electron-donating
groups (methyl or methoxy) were converted into 6c and 6d
with still very good regioselectivities of 50:1 and 22:1 favoring
the methyl ketones (6c, 6d) over the aldehydes (7c, 7d) in
good to moderate yields. Even the ortho-substituted methyl
ketone 6e was obtained from 5e in high yield and with
excellent regioselectivity. Besides aryl ketones, epoxides bearing
NH or OH groups like sulfonamide (5f) and hydroxyl groups
(5g) were rearranged to the desired methyl ketones with 1 mol
% catalyst loading in 83% (6f) and 90% (6g) yield, albeit at
longer reaction time (24 h). The high nucleophilicity of complex
4 is also compatible with an ester group. The isomerization of
1
2
3
4
5
6
7
8
9
Figure 2. Minimum conformation of complex 4 based on DFT-D3 calcu-
lations (Turbomole: BP86/def2-TZVP) and observed NOE cross peaks
(indicated with arrows). For clarity reasons, part of the carbazole backbone is
depicted in wireframe style.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Table 1. Optimization of the reaction conditions for the isomerization of
phenyl oxirane in presence of rhodium catalyst 4.[a]
°
epoxide 5h afforded 6h in moderate yield after 24 h at 60 C
and 5 mol% catalyst loading. As the extended reaction times
may derive from competing coordination of LiBr at the Lewis
basic centers of functional groups, we have increased the
amount of LiBr to 50 mol% and achieved full conversion already
after 1 h reaction time at moderate to excellent yield (entries 6–
8) at room temperature. In the case of 5g, the moderate yield
resulted from unknown side products which will be studied
further. Furthermore, the sterically demanding 2-(tert-butyl)
oxirane (5i) can be also converted to the desired product 6i,
albeit with higher catalyst loading (5 mol%) and reaction
temperature.
Entry Catalyst [mol
%]
Conc.5a [mol/
L]
t
Yield Ratio
[%][b] (6a:7a)
1
2
3
5
3
1
1
1
1
1
1
–
0.1
0.1
0.1
0.1
0.1
0.2
0.4
0.8
0.8
40 min
70 min
120 min >99 >99:1
24 h
24 h
60 min
50 min
40 min
60 min
84
89
>99:1
>99:1
4[c]
5[d]
6
7
6
99
95
92
<5
5:1
10:1
>99:1
>99:1
>99:1
–
7
8
9
The internal epoxides 5j–l react considerably faster com-
[a] Carried out in J. Young NMR tubes and with 10 mol% LiBr and 5 μL THF-
d8. [b] Yield (1H NMR) of 6a calibrated to 1,3,5-trimethoxybenzene (internal
standard). [c] Without LiBr. [d] THF-d8 as the solvent.
°
pared with rhodium catalyst D. At 80 C, 7-oxabicyclo[4.1.0]
heptane (5j) was isomerized into cyclohexanone in almost
quantitative yield. Furthermore, cis-2,3-epoxybutane (5k) was
also compatible with the catalyst system and yielded 99%
butan-2-one. In contrast, the conversion of trans-2,3-epoxybu-
tane (5l) under otherwise identical conditions yielded only 7%
of butan-2-one. This outcome is of considerable interest as it is
contrary to the results from Coates and coworkers, who
obtained less than 40% conversion with cis-2-hexene oxide but
98% yield with trans-2-hexene oxide applying [(salcy)Al(THF)2]+
[Co(CO)4]À as the catalyst.[5]
Encouraged by this result, we tried to reduce the catalyst
loading: 1 mol% of the catalyst was sufficient for a fast and
quantitative rearrangement without influencing the excellent
regioselectivity (Table 1, entries 2 and 3). To test whether
lithium bromide was still necessary as a co-catalyst we applied
the isolated catalyst 4 without adding any Lewis acids. The
catalytic activity was reduced dramatically under these con-
ditions (Table 1, entry 4). This indicates that a Lewis acid is
essential for the pre-activation of the substrates. Low con-
version was also obtained using THF-d8 as the solvent which
most likely stems from competitive binding between THF and
the substrate to the Lewis acid co-catalyst (Table 1, entry 5).
Afterward, the concentration of phenyl oxirane was studied.
Epoxide concentrations of 0.2 mol/L (Table 1, entry 6) shortened
the reaction time substantially and the reaction was 10 times
faster compared to catalyst D. Further increase of the concen-
tration led to a reduced reaction time whereas the yields
decreased slightly (Table 1, entries 7 and 8). In addition, the
need for catalyst 4 was also tested. Without rhodium catalyst 4
the yield remained below 5% (Table 1, entry 9). Thus, the
optimized reaction conditions (1 mol% catalyst 4, 0.2 M epox-
ide, C6D6, room temperature) were applied to investigate the
generality of this protocol.
Mechanistic considerations. When we tried to isolate
catalyst 4 by chromatography at silica gel under argon, we
isolated the isomerized complex 8 (Figure 3, top) as indicated
by the characteristic substitution pattern in the NMR spectra.
The signal at 1.17 ppm with an integral of 3H shows a 3J (6.3 Hz)
and a 4J coupling to the olefinic signals at 4.56 (H-15) and
4.09 ppm (H-16). The cis-conformation of the double bond is
3
not only confirmed by the J coupling constant of 6.3 Hz, but
also by the NOE between methyl signal H-17 and the signal H-
14b of the methylene group. In addition, an NOE cross-peak is
observed between the signals of both olefinic protons and the
N-methyl signal. The assignment of the peaks of the imidazole
moieties and carbazole backbone are based on the observed
NOE cross peaks (Figure 3, bottom).
Complex 8 is of importance as it is also formed under the
conditions applied for the catalysis. While the in situ prepared
ChemCatChem 2019, 11, 1–9
www.chemcatchem.org
3
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
Table 2. Substrate scope in the isomerization of epoxide catalysed by rhodium complex 4.[a]
1
2
3
4
Entry
Substrate
Product
t
Yield [%][b]
Ratio
(6:7)
5
6
6a
5a
7
8
9
1
2
3
1 h
1 h
1 h
99
98
77
>99:1
>99:1
50:1
6b
6c
6d
5b
5c
5d
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
4
5
1 h
42
92
22:1
6e
5e
5f
24 h
33:1
6f
24 h
1 h
83
94[e,g]
6
7
8
–
–
–
6g
6h
24 h
1 h
90
5g
5h
42[e,g]
24 h
1 h
67[c,e]
80[e,g]
6i
6j
5i
5j
9[d–f]
24 h
24 h
99
99
–
–
10[d–f]
6k
6k
5k
5l
11[d–f]
24 h
24 h
99
7
–
–
12[d–f]
[a] Carried out in J. Young NMR tubes and with 10 mol% LiBr and 5 μL THF-d8, 0.2 M concentration of epoxides. [b] Yield (1H NMR) of 6a calibrated to 1,3,5-
trimethoxybenzene (internal standard). [c] At 60 C. [d] 20 mol% LiNTf2 instead of LiBr. [e] 5 mol% catalyst 4. [f] At 80 C. [g] 50 mol% LiBr.
°
°
complex 4LiX is stable in THF-d8, isomerization takes place within
one hour at room temperature when 2 eq of LiBr are added to
a solution of the isolated complex 4 in benzene-d6 without the
presence of any amount of epoxide. Obviously, the Lewis
acidity of the lithium cation in benzene is enhanced to such an
extent that the isomerization of the double bond can occur.
To prove that complex 8 is an active catalyst itself, we have
carried out the Meinwald reaction of phenyl oxirane with the
isolated complex 8 under the general conditions (1 mol% 8,
10 mol% LiBr, 5 μL THF-d8, room temperature). The reaction rate
is comparable to that when starting from complex 4. We
assume that the isomerization of complex 4 to 8 as well as the
Meinwald reaction of phenyl oxirane catalyzed by complex 4 or
8 occurs at comparable reaction rates. Isomerizations of the
double bond can also be observed using catalyst D. However,
due to the two N-homoallyl moieties, mixtures are obtained,
and the NMR spectra become very complex.
ChemCatChem 2019, 11, 1–9
www.chemcatchem.org
4
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Scheme 4. D/H exchange during the isomerization of 1-D-phenyl oxirane
with the N-homoallyl substituted catalysts D and 4 indicate a hydridoÀ Rh(III)
intermediate Int-1 in the catalytic cycle.
Figure 3. Top: Isomerization of the terminal double bond of complex 4 to an
internal cis-double bond in complex 8 by Lewis acids. Bottom: Minimum
conformation of the isomerized complex 8 based on DFT calculations
(Turbomole: BP86/def2-TZVP) and observed NOE cross peaks (indicated with
arrows). For clarity reasons, part of the carbazole moiety is depicted in
wireframe style.
As the hydride intermediate (Int-1) was not observed during
the catalysis with complex B (L=CO), we envisaged that the N-
homoallyl moiety of catalyst D or 4 could insert into the RhÀ H
intermediate Int-1 to give the rhodium alkyl intermediate Int-2,
which could undergo a D/H exchange in case deuterated
phenyl oxirane was used. To prove this hypothesis, we carried
out catalytic experiments with 1-D-phenyl oxirane as substrate
with both N-homoallyl-substituted catalysts D and 4 under
catalytic as well as stoichiometric conditions (Scheme 4). In all
cases, we found an exchange of the deuterium by hydrogen in
the methylphenyl ketone of 5–7%. (deuteration degree of the
phenyl oxirane: 98%).
Based on these observations, we favor pathway II for the
catalytic cycle of the nucleophilic Meinwald reaction with our
rhodium catalysts. This pathway is also supported by the
observations by Milstein, who isolated hydrido-2-oxoalkyl
complexes when reacting RhCl(PMe3)3 with neat methyl or
phenyl oxirane.[3g] In this case, the reductive elimination to
acetone or acetophenone is the rate limiting step.[6]
Catalyst deactivation pathways. In addition, we observed
two deactivation products of the catalyst. We recognized the
formation of red crystals after finishing a VT NMR experiment of
catalyst 4 with 1.0 equiv. of phenyl oxirane at room temper-
ature. The X-ray structure analysis reveals the formation of a
bromidoÀ Rh(III) complex, in which the N-homoallyl substituent
is coordinated in a η3-allyl coordination mode to the oxidized
metal center (Figure 4). The allylÀ Rh bonds measure 2.137
(RhÀ C17), 2.114 (RhÀ C16) and 2.220 Å (RhÀ C15). As the isomer-
ization of phenyl oxide 5a is very fast, it is likely that the
Scheme 3. Initially proposed catalytic cycle (pathway I and pathway II) for
complex B (L=CO) and evidence for pathway II following the observed H/D
exchange in the case of the homoallyl complexes D and 4 (vide infra).
Earlier, we had proposed a catalytic cycle for catalyst B, in
which the oxidative addition product OA was confirmed by
NMR spectroscopy (Scheme 3).[3h] From this intermediate two
possible pathways can lead to the product: a concerted 1,2-H
shift under reductive elimination of the catalyst (c) or β-hydride
elimination (d) to form the Rh-hydride intermediate (Int-1),
which subsequently undergoes reductive elimination (e) to
release the product and the catalyst.
formation of complex
9
occurs after the catalysis. CÀ H
activation at the allylic position to form an allylhydrido complex
and subsequent substitution of the hydrido by a bromido
ligand can explain its formation.
From the testing of cis-1,2-(diethoxycarboxyl)oxirane (5 mol
%
4 (2.3 mg), 20 mol% LiNTf2 (4.6 mg), cis-1,2-(diethoxy
carboxyl)oxirane (15.1 mg) and 1,3,5-trimethoxybenzene
ChemCatChem 2019, 11, 1–9
www.chemcatchem.org
5
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
and the carbazole nitrogen atoms with a mean distance of
2.74 Å. The allyl rhodium bonds measure 2.143 (RhÀ C17), 2.124
(RhÀ C16) and 2.243 Å (RhÀ C15) (mean).
1
2
3
4
5
To avoid erroneous catalytic results in the NMR experiments,
catalytic tests with substrates of low reactivity should be carried
out either with an inert or without an internal standard.
6
7
8
9
Conclusions
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
Herein, we have presented the most active and selective
catalyst system for the nucleophilic Meinwald reaction of
terminal epoxides so far. The reactivity enhancement of catalyst
4 was achieved by providing only one coordinating N-homoallyl
substituent at the ligand scaffold. The isolated complex 4 can
undergo isomerization of the double bond in the presence of
weak Lewis acids and benzene as solvent to yield the N-
methallyl complex 8, which is itself a comparably active catalyst
in the Meinwald reaction. D/H exchange experiments provide
strong evidence for a β-hydride elimination/reductive elimina-
tion pathway via a hydridoÀ Rh intermediate for the Rh-
catalyzed nucleophilic Meinwald reaction.
Figure 4. Molecular structure of the catalyst deactivation product 9. Atoms
are shown with anisotropic atomic displacement parameters at the 50%
probability level. Hydrogen atoms as well as 3.5 co-crystallized benzene
molecules are omitted for clarity.
Experimental Section
General information. Unless otherwise noted, all reactions were
carried out under an argon atmosphere in dried and degassed
solvents using Schlenk technique. Toluene, pentane, dichloro-
methane and tetrahydrofuran were purchased from Sigma Aldrich
and dried using an MBraun SPS-800 solvent purification system. All
lithium salts used were obtained from commercial suppliers, dried
in vacuum and used without further purification. Chemicals from
commercial suppliers were degassed through freeze-pump-thaw
cycles prior to use. Rhodium complex D was synthesized according
to the literature procedure.[3j] 1H and 13C NMR spectra were
recorded using a Bruker ARX 250 or AVANCE II+400 spectrometer.
Chemical shifts δ (ppm) are given relative to the solvent’s residual
proton and carbon signal respectively: THF-d8: 3.58 ppm (1H NMR)
and 67.57 ppm (13C{H} NMR); C6D6: 7.16 ppm (1H NMR) and
128.39 ppm (13C{H} NMR); DMSO-d6: 2.50 ppm (1H NMR) and
39.51 ppm (13C{H} NMR). Coupling constants (J) are expressed in Hz.
Signals were assigned as s (singlet), d (doublet), t (triplet), q
(quartet), quint (quintet), m (multiplet) and variations thereof; (br)
refers to a broad signal. The assignment of peaks is based on 2D
NMR correlation and NOE spectra.
m
Synthesis of HbimcaMeH ·HI·HBr (3). a) Synthesis of 2: To a
° °
Figure 5. Molecular structure of the side product 10. Atoms are shown with
anisotropic atomic displacement parameters at the 50% probability level.
Hydrogen atoms as well as the two NTf2À counter ions and two co-
crystallized benzene molecules are omitted for clarity.
suspension of 1 (0.74 g, 1.8 mmol) in 10 mL THF was added MeI
(0.51 g, 3.6 mmol) dropwise. The reaction was stirred at room
temperature for 24 h to form a pale precipitate. The mixture was
filtered and dissolved in dichloromethane. After removal of solvent,
product 2 was obtained as a light yellow solid (0.89 g, 1.6 mmol,
yield: 90%). 1H NMR (250.13 MHz, DMSO-d6) δ=11.24 (s (br), 1H,
4
NH), 9.69 (s (br), 1H, H-7’), 8.57 (d, JHH =1.7 Hz, 1H, H-5), 8.42 (d,
°
°
(4.3 mg), 24 h at 80 C and 24 h at 100 C) we obtained red
single crystals in the NMR tube after several days at room
temperature. The X-ray structure analysis reveals that under the
elevated temperature the internal standard 1,3,5-trimethoxy-
benzene has reacted with the N-homoallyl moiety of complex 4
in a formal dehydrogenation and CÀ C bond formation to the Rh
(III)(η3-allyl) complex 10 (Figure 5). In the solid state the 16eÀ
Rh-d6 complex forms a Lewis-pair dimer between the rhodium
3
4JHH =1.7, 1H, H-4), 8.21 (s (br), 1H, H-2’), 8.19 (ps t, JHH =1.8 Hz, 1H,
3
3
H-10’), 7.98 (ps t, JHH =1.7 Hz, 1H, H-9’), 7.73 (ps t, JHH =1.3 Hz, 1H,
4
4
H-5’), 7.66 (d, JHH =1.7 Hz, 1H, H-7), 7.50 (d, JHH =1.7 Hz, 1H, H-2),
3
7.22 (ps t, JHH =1.0 Hz, 1H, H-4’), 3.99 (s, 3H, H-18), 1.45 (s, 18H, H-
11 and H-13). 13C{H} NMR (62.90 MHz, DMSO-d6) δ=143.6 and 143.2
(C3 and C6), 138.0 (C7’), 137.2 (C2’), 132.7 (C8a), 132.1 (C1a), 129.1
(C4’), 125.7 (C4a), 125.0 (C5a), 123.7 (2x) (C10’ and C9’), 121.7 (C1),
121.0 (C7), 120.2 (C5’), 120.0 (C2), 119.1 (C5), 118.9 (C8), 116.7 (C4),
ChemCatChem 2019, 11, 1–9
www.chemcatchem.org
6
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
36.0 (C18), 34.78 and 34.70 (C10 and C12), 31.65 and 31.62 (C11
and C13). C27H32N5I (553.50): calcd C 58.59, H 5.83, N 12.65; found C
filtered to remove potassium halides. The collected filtrate was
dried under vacuum and the residue was washed with pentane (3×
5 mL). Pure catalyst 4 was obtained after removal of solvent as a
1
2
3
4
5
6
7
8
9
°
58.38, H 6.29, N 12.30. m.p.: 287 C (dec.).
yellow solid (43.6 mg, 75.0 μmol, yield: 52%). The NMR data (THF-
b) Synthesis of 3: To a solution of the monomethylated product 2
m
d8) correspond to the results obtained from using [Li(bimcaMe,H )].
° °
(0.89 g, 1.6 mmol) in 10 mL DMF was added 4-bromobutene
C31H36N5Rh (581.56): calcd C 64.02, H 6.24, N 12.04; found C 61.70, H
6.02, N 11.36. Possibly contains residual KBr: C31H36N5Rh·0.2 KBr:
calcd C 61.51, H 5.99, N 11.57.
°
(0.43 g, 3.2 mmol) in one portion. The reaction was stirred at 110 C
for 21 h. The solvent was removed via oil pump and a good
precipitate was achieved after adding a bit of ethanol and diethyl
ether. The mixture was filtered, and the residue was resolved in
dichloromethane. After removal of the solvent, the product 3 was
obtained as a light yellow solid (1.0 g, 1.5 mmol, yield: 95%). 1H
NMR (250.13 MHz, DMSO-d6) δ=11.63 (s (br), 1H, NH), 10.06 (s (br),
1H, H-2’), 10.01 (s (br), 1H, H-7’), 8.65 (s (br), 2H, H-4 and H-5), 8.32
(s (br), 1H, H-5’), 8.29 (s (br), 1H, H-10’), 8.17 (s (br), 1H, H-4’), 8.06 (s
Synthesis of the isomerized rhodium complex 8. To a J. Young
NMR tube containing LiBr (6.9 mg, 80 μmol) and 20 μL of THF-d8,
was added catalyst 4 (23.2 mg, 40.0 μmol) with C6D6 (0.8 mL).
Catalyst 8 was generated at room temperature after 1 h in
quantitative yield as determined by NMR spectroscopy. Another
method for achieving complex 8 was to run a chromatography on
silica gel with complex 4. 1H NMR (400 MHz, THF-d8) δ=8.19 (d,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
4
4
(br), 1H, H-9’), 7.76 (d, JHH =1.3 Hz) and 7.75 (d, JHH =1.4 Hz) (2H,
H-2 and H-7), 5.92 (ddt, 3JHH =17.2 Hz, 3JHH =10.3 Hz, 3JHH =6.8 Hz,
1H, H-16), 5.23 (dd, JHH =17.2 Hz, JHH =1.0 Hz, 1H, H-17trans), 5.16 (d
3JHH =2.2 Hz, 1H, H-10’), 8.14 and 8.13 (each d, each JHH =1.6 Hz,
4
3
2
each 1H, H-5 and H-4), 8.02 (d, 3JHH =2.2 Hz, 1H, H-5’), 7.81 (d, 4JHH
=
4 3
3
3
(br), JHH =10.3 Hz, 1H, H-17cis), 4.44 (t, JHH =7.2 Hz, 2H, H-14), 4.03
1.6 Hz, 1H, H-7), 7.73 (d, JHH =1.6 Hz, 1H, H-2), 7.27 (d, JHH =2.0 Hz,
3
3
3
(s, 3H, H-18), 2.78 (dt, JHH =6.8 Hz, JHH =7.2 Hz, 2H, H-15), 1.47 (s,
18H, H-11 and H-13). 13C{H} NMR (62.90 MHz, DMSO-d6) δ=143.7
(C3 and C6), 137.9 (C7’), 137.3 (C2’), 133.7 (C16), 132.3 and 132.2
(C1a and C8a), 125.5 and 125.4 (C4a and C5a), 123.9 (C9’), 123.3
(C5’), 123.2 (C10’), 122.9 (C4’), 120.92 and 120.89 (C2 and C7),
119.42 and 119.37 (C4 and C5), 119.0 (C1 and C8), 118.3 (C17), 48.4
(C14), 36.1 (C18), 34.9 (C10 and C12), 33.2 (C15), 31.6 (C11 and C13).
1H, H-4’), 7.17 (d, 3JHH =2.0 Hz, 1H, H-9’), 4.56 (br ps t, 1H, JHH
=
2 3
7.2 Hz, H-15), 4.45 (dd, JHH =11.7 Hz, JHH =6.8 Hz, 1H, H-14a), 4.09
(d ps quint, 3JHH =6.2 Hz, 4JHH =2.3 Hz, 1H, H-16), 3.97 (br d, 2J=
11.7 Hz, 1H, H-14b), 3.71 (s, 3H, H-18), 1.55 and 1.54 (each s, 18H, H-
3
4
11 and H-13), 1.17 (dd, JHH =6.3 Hz, JHH =0.6 Hz, 3H, H-17). 13C{H}
1
NMR (101 MHz, THF-d8) δ=187.9 (d, JRhC =42.4 Hz, C7’), 187.2 (d,
1JRhC =44.3 Hz, C2’), 139.5 and 139.2 (C3 and C6), 137.5 (C1a), 136.3
(C8a), 127.9 and 127.6 (C4a and C5a), 126.4 (C1), 126.2 (C8), 124.6
(C9’), 118.3 (C4’), 116.5 (C10’), 114.8 (C5’), 114.1 (C5), 113.8 (C4),
°
m.p.: 263 C.
Preparation of catalysts 4LiX. a) In situ generation of [Li(bim-
caMe,Homo)]: Lithium bis(trimethylsilyl)amide (7.3 mg, 44 μmol) was
added to the suspension of 3 (10.0 mg, 14.5 μmol) in 0.5 mL of
THF-d8 at room temperature and a light yellow solution with blue
fluorescence was formed. After 10 min, the quantitative formation
of [Li(bimcaMe,Homo)] was confirmed by 1H NMR spectroscopy. 1H
NMR (400 MHz, THF-d8) δ=7.99 (s (br), 2H, H-4 and H-5), 7.73 (s (br),
2H, H-5’ and H10’), 7.39 (s (br), 2H, H-2 and H-7), 7.21 and 7.15
(each s (br), each 1H, H-4’ and H-9’), 5.80–6.05 (m, 1H, H-16), 5.15 (d,
1
1
111.4 (C7), 109.7 (C2), 57.3 (d, JRhC =14.1 Hz, C16), 53.8 (d, JRhC
=
12.7 Hz, C15), 50.0 (C14), 37.6 (C18), 35.72 and 35.65 (C10 and C12),
1
32.93 and 32.90 (C11 and C13), 20.9 (C17). H NMR (400 MHz, C6D6)
δ=8.46 (s, 2H), 7.67 (s, 1H), 7.59 (s, 1H), 7.49 (s, 1H), 7.23 (s, 1H),
6.36 (s, 1H), 6.14 (s, 1H), 4.48–4.40 (m, 1H, H-15), 4.31 (ps quint,
2
3
3JHH =6.0 Hz, 1H, H-16), 4.09 (dd, JHH =11.2 Hz, JHH =6.8 Hz, 1H, H-
2
14a), 3.66 (br d, JHH =11.7 Hz, 1H, H-14b), 3.12 (s, 3H, H-18), 1.55 (br
s, 18H, H-11 and H-13), 1.37 (br d, 3JHH =6.0 Hz, 3H, H-17).
3
3JHH =16.5 Hz, 1H, H-17trans), 5.04 (d, JHH =10.5 Hz, 1H, H-17cis), 4.26–
DFT calculations. Performed based on density functional theory at
the BP86/def2-SVP and/or BP86/def2-TZVP[7] level implemented in
Turbomole.[8] The RI-approximation[9] and the Grimme dispersion
correction D3-BJ[10] were used all over. Several structures were
optimized differing in the conformation of the rings formed by the
coordination of the double bond. Minimum structures were verified
at the BP86/def2-SVP level by calculating the Hessian matrix and
ensuring that it has no imaginary frequency.
4.35 (m, 2H, H-14), 3.93 (s (br), 3H, H-18), 2.70–2.80 (m, 2H, H-15),
1.50 (s, 18H, H-11 and H-13).
b) In situ generation of 4LiX: [Rh(μ-Cl)(COD)]2 (0.5 eq) was added to
the previous prepared solution of [Li(bimcaMe,Homo)] (1.0 eq) at room
temperature. The solution was stirred for 10 min. Catalyst 4LiX was
obtained as an orange solution in quantitative yield as determined
1
3
by NMR spectroscopy. H NMR (400 MHz, THF-d8) δ=8.17 (d, JHH
=
3
2.2 Hz, 1H, H-10’), 8.10 (d, JHH =2.3 Hz, 1H, H-5’), 8.08–0.09 (m, 2H,
X-ray structure analysis. CCDC 1907047 (9) and 1907048 (10)
contain the supplementary crystallographic data. These data can be
obtained free of charge from the Cambridge Crystallographic Data
Centre at www.ccdc.cam.ac.uk/data_request/cif.
4
4
H-4 and H-5), 7.73 (d, JHH =1.5 Hz, 1H, H-2), 7.70 (d, JHH =1.5 Hz,
1H, H-7), 7.17 (d, 3JHH =2.3 Hz, 1H, H-4’), 7.16 (d, 3JHH =2.2 Hz, 1H, H-
9’), 4.65–4.58 (m, 1H, H-16), 4.28 (ps td, 2/3 HH =12.4 Hz, 3JHH =2.2 Hz,
J
1H, H-14ax), 4.09 (d ps t, 2JHH =13.0 Hz, 3JHH =3.6 Hz, 1H, H-14eq), 3.92
(s, 3H, H-18), 3.50 (d, 3J=7.7 Hz, 1H, H-17cis), 2.94 (d, 3J=11.5 Hz, 1H,
2
3
H-17trans), 2.67 (br dd, JHH =15.1 Hz, JHH =12.4 Hz, 1H, H-15ax), 2.33–
2.25 (m, 1H, H-15eq) (overlap with COD signal), 1.53 and 1.52 (each
s, 18H, H-11 and H13). 13C{H} NMR (101 MHz, THF-d8) δ=184.8 (d,
Acknowledgements
1JRhC =43.3 Hz, C7’), 179.0 (d, JRhC =45.7 Hz, C2’), 139.2 (C3), 139.1
1
Yingying Tian thanks the China Scholarship Council (CSC) for a
predoctoral fellowship and Eva Jürgens the MWK-BW for funding
(Landesgraduiertenförderung). We are grateful to Prof. Karl W.
Törnroos for assistance in the X-ray structure analyses and thank
Mario R. Rapp for helpful discussions as well as Nina F. Liska for
help with the substrate synthesis.
(C6), 137.1 (C1a), 136.8 (C8a), 127.3 and 127.2 (C4a and C5a), 126.6
(C9’), 126.0 (C8), 125.7 (C1), 123.9 (C4’), 116.1 (C10’), 116.5 (C5’),
1
113.6 (C5), 113.4 (C4), 111.3 (C7), 110.2 (C2), 51.3 (d, JRhC =13.2 Hz,
1
C16), 47.5 (C14), 38.9 (C18), 37.0 (d, JRhC =13.7 Hz, C17), 35.6 (C10
and C12), 32.8 (C11 and C13), 31.9 (C15).
Preparation of catalyst 4. Imidazolium salt 3 (100 mg, 145 μmol),
potassium bis(trimethylsilyl)amide (86.9 mg, 436 μmol) and [Rh(μ-
Cl)(COD)]2 (35.8 mg, 72.6 μmol) were added to a dry flask at room
°
temperature. The flask was cooled down to À 60 C for 30 min. Cold
Conflict of Interest
THF was injected into the flask and the reaction was stirred for
°
another 30 min at À 60 C. After completion, the reaction was
The authors declare no conflict of interest.
ChemCatChem 2019, 11, 1–9
www.chemcatchem.org
7
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
[5] J. R. Lamb, M. Mulzer, A. M. LaPointe, G. W. Coates, J. Am. Chem. Soc.
2015, 137, 15049–15054.
[6] D. Milstein, Acc. Chem. Res. 1984, 17, 221–226.
Keywords: Rhodium · Homogenous catalysis · Epoxides ·
Isomerization · Carbene ligands
1
2
[7] a) A. Becke, Phys. Rev. A 1988, 38, 3098–3100; b) J. P. Perdew, Phys. Rev.
B 1986, 33, 8822–8824; c) A. Schäfer, H. Horn, R. Ahlrichs, J. Chem. Phys.
1992, 97, 2571–2577; d) A. Schäfer, C. Huber, R. Ahlrichs, J. Chem. Phys.
1994, 100, 5829–5835; e) F. Weigend, R. Ahlrichs, Phys. Chem. Chem.
Phys. 2005, 7, 3297–3305; f) F. Weigend, Phys. Chem. Chem. Phys. 2006,
8, 1057–1065.
[8] a) TURBOMOLE V6.3.1 2011, a development of University of Karlsruhe
and Forschungszentrum Karlsruhe GmbH, 1989–2007, TURBOMOLE
GmbH, since 2007; available from http://www.turbomole.com; b) O.
Treutler, R. Ahlrichs, J. Chem. Phys. 1995, 102, 346–354; c) M. von Arnim,
R. Ahlrichs, J. Comp. Chem. 1998, 19, 1746–1757; d) C. van Wüllen, J.
Comp. Chem. 2011, 32, 1195–1201; e) P. Deglmann, F. Furche, R.
Ahlrichs, Chem. Phys. Lett. 2002, 362, 511–518; f) P. Deglmann, F. Furche,
J. Chem. Phys. 2002, 117, 9535–9538; g) R. Ahlrichs, M. Bär, M. Häser, H.
Horn, C. Kölmel, Chem. Phys. Lett. 1989, 162, 165–169; h) M. K.
Armbruster, F. Weigend, C. van Wüllen, W. Klopper, Phys. Chem. Chem.
Phys. 2008, 10, 1748–1756; i) D. Peng, N. Middendorf, F. Weigend, M.
Reiher, J. Chem. Phys. 2013, 138, 184105.
[9] a) K. Eichkorn, O. Treutler, H. Öhm, M. Häser, R. Ahlrichs, Chem. Phys.
Lett. 1995, 240, 283–290; b) K. Eichkorn, O. Treutler, H. Öhm, M. Häser, R.
Ahlrichs, Chem. Phys. Lett. 1995, 242, 652–660; c) K. Eichkorn, F.
Weigend, O. Treutler, R. Ahlrichs, Theor. Chem. Acc. 1997, 97, 119–124;
d) P. Deglmann, K. May, F. Furche, R. Ahlrichs, Chem. Phys. Lett. 2004,
384, 103–107; e) F. Weigend, Phys. Chem. Chem. Phys. 2002, 4, 4285–
4291; f) M. Sierka, A. Hogekamp, R. Ahlrichs, J. Chem. Phys. 2003, 118,
9136–9148.
3
4
5
6
7
8
9
[1] J. Meinwald, S. S. Labana, M. S. Chadha, J. Am. Chem. Soc. 1963, 85, 582–
585.
[2] Ni2+: a) A. Miyashita, T. Shimada, A. Sugawara, H. Nohiya, Chem. Lett.
1986, 15, 1323–1326; In3+; b) B. C. Ranu, U. Jana, J. Org. Chem. 1998, 63,
8212–8216; Fe3+
; c) K. Suda, K. Baba, S.-I. Nakajima, T. Takanami,
Tetrahedron Lett. 1999, 40, 7243–7246; Bi3+; d) A. M. Anderson, J. M.
Blazek, P. Garg, B. J. Payne, R. S. Mohan, Tetrahedron Lett. 2000, 41,
1527–1530; Er3+; e) A. Procopio, R. Dalpozzo, A. De Nino, M. Nardi, G.
Sindona, A. Tagarelli, Synlett 2004, 2004, 2633–2635; Sn; f) M. Banerjee,
U. K. Roy, P. Sinha, S. Roy, J. Organomet. Chem. 2005, 690, 1422–1428;
Cu2+; g) M. W. C. Robinson, K. S. Pillinger, I. Mabbett, D. A. Timms, A. E.
Graham, Tetrahedron 2010, 66, 8377–8382; [ReBr(CO)3]; h) R. Umeda, M.
Muraki, Y. Nakamura, T. Tanaka, K. Kamiguchi, Y. Nishiyama, Tetrahedron
Lett. 2017, 58, 2393–2395; Ru2+; i) C.-L. Chang, M. P. Kumar, R.-S. Liu, J.
Org. Chem. 2004, 69, 2793–2796 (although Lewis-acid catalyzed, methyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
ketones from α-keto oxiranes possibly due to chelation effects; Pd2+
;
j) D. J. Vyas, E. Larionov, C. Besnard, L. Guénée, C. Mazet, J. Am. Chem.
Soc. 2013, 135, 6177–6183; k) N. Humbert, D. J. Vyas, C. Besnard, C.
Mazet, Chem. Commun. 2014, 50, 10592–10595; l) S. Kulasegaram, R. J.
Kulawiec, J. Org. Chem. 1997, 62, 6547–6561; m) S. Kulasegaram, R. J.
Kulawiec, Tetrahedron 1998, 54, 1361–1374; Ni in a reaction sequence;
n) A. N. Desnoyer, J. L. Geng, M. W. Drover, B. O. Patrick, J. A. Love,
Chem. Eur. J. 2017, 23, 11509–11512.
[3] a) J. L. Eisenmann, J. Org. Chem. 1962, 27, 2706; b) B. Rickborn, R. M.
Gerkin, J. Am. Chem. Soc. 1968, 90, 4193–4194; c) B. Rickborn, R. M.
Gerkin, J. Am. Chem. Soc. 1971, 93, 1693–1700; d) Z.-W. An, R. D’Aloisio,
C. Venturello, Synthesis 1992, 1992, 1229–1231; e) S. Kulasegaram, R. J.
Kulawiec, J. Org. Chem. 1994, 59, 7195–7196; f) J. Prandi, J. L. Namy, G.
Menoret, H. B. Kagan, J. Organomet. Chem. 1985, 285, 449–460; g) D.
Milstein, J. Am. Chem. Soc. 1982, 104, 5227–5228; h) E. Jürgens, B.
Wucher, F. Rominger, K. W. Törnroos, D. Kunz, Chem. Commun. 2015, 51,
1897–1900; i) J. R. Lamb, Y. Jung, G. W. Coates, Org. Chem. Front. 2015,
2, 346–349; j) Y. Tian, E. Jürgens, D. Kunz, Chem. Commun. 2018, 54,
11340–1134.
[10] a) S. Grimme, J. Antony, S. Ehrlich, H. Krieg, J. Chem. Phys. 2010, 132,
154104; b) S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem. 2011, 32,
1456–1465.
Manuscript received: April 1, 2019
Revised manuscript received: May 20, 2019
Version of record online: ■■■, ■■■■
[4] M. Moser, Dissertation, Heidelberg University, Heidelberg, 2007.
ChemCatChem 2019, 11, 1–9
www.chemcatchem.org
8
© 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
��
These are not the final page numbers!

Full Papers
FULL PAPERS
1
2
3
4
5
Finding the path: The efficient iso-
merization of a wide range of
Y. Tian, E. Jürgens, K. Mill, R. Jordan, T.
Maulbetsch, Prof. Dr. D. Kunz*
epoxides into methyl ketones is
achieved using a novel pincer-
rhodium complex under very mild
conditions with excellent chemo-
and regioselectivity. Investigations
on the mechanism reveal the con-
comitant isomerization of the
catalyst and provide evidence for a
β-hydride elimination-reductive elim-
ination step in the catalytic cycle.
6
7
8
9
1 – 9
Nucleophilic Isomerization of
Epoxides by Pincer-Rhodium
Catalysts: Activity Increase and
Mechanistic Insights
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57