Deoxygenation of Epoxides with Carbon Monoxide

Article abstract of DOI:10.1002/chem.202002651

The use of carbon monoxide as a direct reducing agent for the deoxygenation of terminal and internal epoxides to the respective olefins is presented. This reaction is homogeneously catalyzed by a carbonyl pincer-iridium(I) complex in combination with a Lewis acid co-catalyst to achieve a pre-activation of the epoxide substrate, as well as the elimination of CO₂ from a γ-2-iridabutyrolactone intermediate. Especially terminal alkyl epoxides react smoothly and without significant isomerization to the internal olefins under CO atmosphere in benzene or toluene at 80–120 °C. Detailed investigations reveal a substrate-dependent change in the mechanism for the epoxide C?O bond activation between an oxidative addition under retention of the configuration and an S_N2 reaction that leads to an inversion of the configuration.

Full text of DOI:10.1002/chem.202002651

Accepted Article
Accepted Article
Title: Deoxygenation of Epoxides with Carbon Monoxide
Authors: Theo Maulbetsch, Eva Jürgens, and Doris Kunz
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: Chem. Eur. J. 10.1002/chem.202002651
Link to VoR: https://doi.org/10.1002/chem.202002651
01/2020

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
Deoxygenation of Epoxides with Carbon Monoxide
Theo Maulbetsch, Eva Jürgens and Doris Kunz*[a]
Abstract: The use of carbon monoxide as a direct reducing agent for
the deoxygenation of terminal and internal epoxides to the respective
olefins is presented. This reaction is homogeneously catalyzed by a
carbonyl pincer-iridium(I) complex in combination with a Lewis acid
co-catalyst to achieve a pre-activation of the epoxide substrate as well
as the elimination of CO2 from a γ-2-iridabutyrolactone intermediate.
Especially terminal alkyl epoxides react smoothly and without
significant isomerization to the internal olefins under CO atmosphere
in benzene or toluene at 80-120 °C. Detailed investigations reveal a
substrate-dependent change in the mechanism for the epoxide C-O
bond activation between an oxidative addition under retention of the
configuration and an SN2 reaction that leads to an inversion of the
configuration.
are variations – in addition to electrochemical methods[20]
– in
which hydrogen is used as a deoxygenation agent to obtain water
as a byproduct. Methylrheniumtrioxide (Re(CH3)O3) is a suitable
catalyst for this reaction that is carried out at 150 °C. However,
the hydrogenation of the product olefins can be an unwanted side
reaction.[21] Silver and gold nanoparticles are heterogeneous
catalysts in a process that uses hydrogen directly[22] or generates
it in situ from alcohols.[23] The currently most efficient system uses
reactive hydrogen species that are produced in an in situ water
gas shift reaction from carbon monoxide by a hydrotalcite-
supported gold nanoparticle catalyst.[24,25] However, above 50 °C
the formation of free hydrogen is observed, which increases the
CO consumption and requires additional security measures.
While aryl oxiranes already react at room temperature, alkyl
oxiranes require temperatures of 110 °C and an organic solvent.
Introduction
Wittig-Reaction (XR2 = PPh3)
Apart from the water-gas shift reaction itself as well as
reductions using hydrogen produced by this reaction in situ[1], the
use of carbon monoxide as a direct deoxygenation agent is very
rare in homogenous catalysis and hard to distinguish from the
former one depending on the system[1-4]. The deoxygenation of
epoxides (Scheme 1) to alkenes is an important reaction in
organic chemistry and some non-catalytic as well as catalytic
reactions (homogeneous and heterogeneous) are known. One
early report on the deoxygenation of epoxides was in 1955 by
Wittig and Haag,[5] who used triphenylphosphine as a deoxy-
genation reagent at 180 °C to deoxygenate α,β-epoxy esters that
were obtained from the Darzens reaction. They recognized that
the temperature can be reduced when adding hydroquinone.
Vedejs and Fuchs developed this reaction further by reacting cis
or trans epoxides to the betaines with lithium diphenylphosphide
and methyl iodide, which subsequently led to formation of the
olefin with inversion of the configuration at room temperature.[6,7]
Later, rhenium and molybdenum catalyzed variations were deve-
loped.[8,9] In these cases, the stoichiometric amounts of triorgano-
phosphine oxide as a side product is problematic to get recycled
back to the phosphine or phosphide. Other stoichiometric or over-
stoichiometric deoxygenation reagents for epoxides[10,11] are
Co2(CO)8,[12] Fe(CO)5,[13] SmI2,[14] In/InCl,[15] lithium amide bases
with silylboranes[16] or diazomalonate[17]. The latter two are
catalyzed by copper-nanoparticles and copper complexes
respectively. Not yet fully understood is the deoxygenation of
arene oxides and their oxepine tautomers by [Rh(CO)2Cl]2 or
[Rh(C2H4)2Cl]2 in CHCl3.[18,19] Environmentally much more benign
- OPPh3
Deoxygenation
Johnson-Corey-
R1
O
"
- "
O
Chaykovsky
O
R2X
R1
+
(XR2 = SMe2)
- SMe2
R1
R
R
H
R
2
this work: CO + [ ]
O
C
2
Scheme 1. Deoxygenation of epoxides as part of a two-step Wittig-reaction.
In the following, we will report on the efficient and stereoselec-
tive deoxygenation of terminal and internal alkyl oxiranes as well
as aryl oxiranes with carbon monoxide as a direct deoxygenation
agent catalyzed by the nucleophilic carbonyl pincer-complexes
1[26] and 2[27] (Figure 1). Using CO directly is advantageous as the
formation of hydrogen, which can cause
further side reactions, is avoided per se. In
combination with the Johnson-Corey-Chay-
kovsky reaction of aldehydes with sulfur ylides
to epoxides, this sequence equals to the Wittig
reaction, but with formation of CO2 instead of
triphenylphosphine oxide (Scheme 1).
Figure 1. Structure of the pincer complexes 1 and 2 bearing the bimca ligand.
Results and Discussion
In 2015 we reported on the nucleophilic isomerization of
terminal epoxides to methyl ketones using the electron rich
carbonyl pincer-rhodium catalyst 1 in combination with LiNTf2 as
a Lewis acid co-catalyst.[28] As one of the observed side reactions
was a CO insertion into the ring-opened intermediate, we
envisaged the possibility of a catalytic reaction to obtain β-
lactones, as it is known for Co2(CO)8.[29,30] However, we were
[a]
MSc. T. Maulbetsch, Dr. E. Jürgens, Prof. Dr. D. Kunz
Institut für Anorganische Chemie
University of Tübingen
Auf der Morgenstelle 18, D-72076 Tübingen, Germany
E-mail: Doris.Kunz@uni-tuebingen.de
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
surprised, when we found propylene as the only organic product
when reacting propylene oxide, 5 mol% of 1 and 20 mol% of
LiNTf2 under 15 bar of CO in C6D6 at 80 °C (Scheme 2). The
formation of CO2 was recognized in the 13C NMR spectrum.
independent (entries 8-10). However, at low pressure, the amount
of CO (1.2 equivalents) comes close to the required stoichiometric
amounts in our setup (Wilmad Pressure NMR tubes) and thus the
insufficient amount of CO dissolved in benzene explains the lower
yields. At 15 bar the yield seems to decrease slightly (entry 10).
Therefore, we chose 10 bar as an optimal pressure in all further
experiments. To shorten the reaction time, we also increased the
temperature and found full conversion at 100 °C after 8 h and at
120 °C after 2 h (entry 11-12). Even at room temperature some
product formed, but at a very slow reaction rate (entry 13). With
1-hexene oxide, we recognized a slow isomerization of 1-hexene
to internal hexenes. The degree of isomerization increases, when
the reaction mixture is kept at the reaction conditions after full
conversion of the epoxide. This isomerization also requires the
presence of the Lewis acid as a co-catalyst as it was confirmed
by independent measurements (Supporting Information).
5 mol% [1/2],
20 mol% LiNTf2
5 mol% [1/2],
20 mol% LiNTf2
15 bar CO
+
O
O
+
CO2
CH3
CH3
CH3
80 °C, C6D6
80 °C, C6D6
3a
4a
Scheme 2. Deoxygenation of epoxides (right) with CO catalyzed by the electron
rich pincer complexes 1 and 2. The formation of β-lactones is not observed.
This prompted us to investigate this unusual reactivity
further.[31] While the rhodium catalyst 1 seemed to be unstable at
the elevated temperature, we were pleased to find that the analo-
gous iridium complex 2 was not only more stable, but also
considerably more active (Table 1, entry 5). Other (commercially
available) rhodium, iridium, cobalt or iron complexes were much
less active under the identical conditions (entries 2-4 and 7-9).
With the pincer complex [Ir(bimcaC5)CO], in which the carbene
moieties are connected via a 1,5-pentadiyl chain (bimcaC5),
mainly the epoxide isomerization product, methyl butyl ketone,
was detected (entry 6).
Table 2. Optimizing the reaction conditions for the catalytic deoxygenation of
1,2-epoxyhexane with COa
Entry [2]
LiNTf2 CO
Solvent Temp. Time Conv.b Yieldb Isomerb
[mol%] [mol%] [bar]
[°C] [h]
[%]
[%]
[%]
1
5
--
10
10
10
10
10
10
10
2.0
5.9
15
10
10
C6D6
C6D6
C6D6
C6D6
C6D6
Tol-d8
80
80
80
80
80
80
24
24
24
24
24
24
24
24
24
24
8
0
0
Table 1. Catalyst screening for the catalytic deoxygenation of 1,2-epoxyhexane
with COa
2
--
1
30
6.0
15
30
30
30
30
30
30
30
30
0
0
3
32
81
100
92
7
32
70
97
91
3
4
2.5
5
Entry Catalyst
Time
Yield
(alkene) b
Side product b
5
2
6
5
1
2
3
4
[Rh(bimcaMe)CO] (1)
24 h
24 h
24 h
50 %
0 %
--
--
--
--
7
5
THF-d8 80
--
4
4
3
[RhCl(PPh3)3]
8
5
C6D6
C6D6
C6D6
Tol-d8
Tol-d8
80
83
94
94
100
79
90
91
93
[Ir(CO)Cl(PPh3)2]
[Ir(acac)(CO)2]
0 %
9
5
80
24 h
90 h
31 %
68 %
10
11
12
5
80
5
6
[Ir(bimcaMe)CO] (2)
24 h
24 h
98 %
8 %
2 % int. olefin
5
100
[Ir(bimcaC5)(CO)]
87 % methyl butyll
ketone
5
120 0.5
2
96
100
92
82
7
8
9
[Co(Cp)(CO)2]
[Co2(CO)8]
[Fe(CO)5]
24 h
24 h
24 h
0 %
0 %
0 %
--
--
--
13
5
30
10
C6D6
rt
168
6
6
[a] reaction carried out in a pressure NMR tube (Wilmad) at 0.2 M epoxide
concentration. [b] from 1H NMR calibrated to 1,3,5-trimethoxybenzene as
internal standard.
[a] reaction carried out in a pressure NMR tube (Wilmad) at 0.2 M epoxide
concentration. [b] from 1H NMR calibrated to 1,3,5-trimethoxybenzene as
internal standard.
We screened a broad range of substrates and found that
terminal epoxides react most readily (Figure 2). The reaction
under the optimized conditions works very well for aliphatic,
terminal epoxides (3a-3e) including benzyl epoxide (3k) (no
isomerization to methyl styrene is observed), and 1,1-
disubstituted epoxides (3f), whereas internal epoxides (3g-3j) are
much harder to deoxygenate as well as terminal epoxides bearing
functional groups (3l-3y). In case of styrene oxide (3n) the
formation of 28% of benzyl aldehyde indicates a Lewis acid
Therefore, we optimized the reaction conditions for catalyst 2
(Table 2). Without catalyst or without co-catalyst, the reaction did
not proceed (entries 1 and 2). THF as a solvent is not favorable,
presumably as it reduces the Lewis acidity of the LiNTf2 co-
catalyst (entry 7). Afterwards we probed the pressure depen-
dence of carbon monoxide. The reaction rate is roughly pressure-
This article is protected by copyright. All rights reserved.

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
5 mol% [2], 30 mol% LiNTf2,
catalyzed epoxide isomerization as side reaction. The use of LiBr
(solubilized with 4 eq of tetrahydrofuran) instead of LiNTf2, which
circumvented this reaction in the nucleophilic epoxide isomeriza-
tion,[32] did not have any beneficial effect. Although the amount of
the aldehyde by-product could be reduced pronouncedly using
LiBr or LiI, the reaction did not go to completion due to deacti-
vation of the catalyst by formation of [Ir(bimca)(CO)X2] (X = Br, I)
(see Supporting Information). While electron rich styrene oxides
(3l, 3m) lead to an increase of side reactions, electron poor
styrene oxides 3o-3t react smoothly at 100 °C and good to excel-
lent yields were obtained with few or negligible formation of the
aldehyde. Moreover, also the sterically hindered o-CF3 substi-
tuted styrene oxide 3t can be deoxygenated to 4t in very high yield,
albeit at a lower reaction rate. Most striking is the stereochemistry
when 1,2-disubstituted epoxides are reacted: deoxygenation of
cis epoxide cis-3j led to full retention of the configuration in olefin
cis-4j. Slow subsequent isomerization of the olefin into the more
stable trans-butene, however, can occur towards the end of the
reaction. Respectively, trans-butene oxide (trans-3j) is converted
into trans-butene (trans-4j), albeit at a slower reaction rate. The
opposite selectivity, inversion of the configuration, is observed
with the doubly ester functionalized substrates cis-3z and trans-
3z. Although the reaction is extremely slow at 80 °C in benzene
(10 d) it proceeds with moderate yields and low isomerization. At
120 °C the epoxy succinate cis-3z reacted to diethyl fumarate
(trans-4z) in 32 % yield and the trans epoxide trans-3z to diethyl
maleate (cis-4z) in 25 % already after 24 h and with still good
selectivity. The selectivity of the deoxygenation can be explained
with a substrate-dependent change in the epoxide activation me-
chanism (vide infra) from oxidative addition in the case of alkyl
epoxides to an SN2 mechanism for the C-O bond activation in the
case of ethyl carboxylates (3z) (vide infra).
O
10 bar CO
+
CO2
C6D6, 80 °C, 24 h
R
R
3
4
15
4a
4f
4b
4c
4d
4e
64 %
69 %
*
65 %
*
97 %
71 %
*
)
a
(84%
*
c
c
4g
4h
4i
cis-4j
trans-4j
70 %
*
1 %
*
0 %
1 %
7 %
2 %
*
*
b
d
e
(17 %
*
)
(38 %
*
)
(52 %
*
)
4l
:
R = p-MeO 4 % (23 % aldehyde)
4m: R = p-Me 26 % (31 % aldehyde)
4n: R = H
4o
: R = p-F
35 % (28 % aldehyde)
62 % (33 % aldehyde)
f
4k
89 % (56 % )
4p: R = p-Cl 82 %g (11 % aldehyde)
4q: R = p-Br 85 %g (10 % aldehyde)
R
:
R = m-CF3 98 %g (1 % aldehyde)
4r
4l-s
4s: R = p-CF3 99 %g (0 % aldehyde)
(61 %f)
4u
6 %
*
4t
:
R = o-CF3 91 %h (0 % aldehyde)
O
O
H
N
O
Ts
Et
O
Et
O
4v 6 %
4w 52 %
4x 31 %
4y 37 %
O
O
O
trans-4z 42 %i
cis-4z 34 %j
O
O
-3z
-3z
from cis
from trans
O
Et
O
O
(32 %k)
(25 %l)
Et
Et
4za
99 %
from terminal epoxide
To demonstrate the applicability of our reaction to natural
product derived substrates, we obtained 4za in 99 % yield from
deoxygenation of 3za, which can be easily obtained from
citronellal in a Johnson-Corey-Chaykovsky reaction with dime-
thylsulfonium ylide (Scheme 1). For practical reasons, we also
found that the reaction can be run at 1 bar of CO on a preparative
scale, to avoid the necessity of using special autoclaves. On a 1
mmol scale (at cost of a longer reaction time of 72 h) we isolated
the products 4k (56%) and 4s (61%) by transfer of the volatiles in
vacuo and subsequent removal of benzene (except of 0.5 eq
remaining). Product loss (24 % (4k) and 17 %(4s)) was found in
the benzene fraction as the scale was too small for a fractional
distillation.
Figure 2. Chemo- and stereoselective deoxygenation of various epoxides.
Reactions were carried out in a medium-wall NMR tube with pressure valve
(Wilmad) at 0.2 M epoxide concentration. Yields were determined by 1H NMR
spectroscopy calibrated to 1,3,5-trimethoxybenzene as internal standard after
release of the overpressure. * indicates the 1H NMR yield of the dissolved
amount of the gaseous reaction product, measured under CO pressure. [a]
120 °C, 24 h; [b] 120 °C, 96 h; [c] 224 h; [d] 120 °C, 72 h, 11 % trans-4j; [e]
120 °C, 72 h, 2 % cis-4j; [f] isolated yield, 1 mmol scale, 1 bar CO; 72 h; [g]
100 °C, 24 h; [h] 100 °C, 48 h; [i] 240 h, 9 % cis-3z; [j] 240 h; 8 % trans-3z; [k]
120 °C, 48 h, 7% cis-3z; [l] 120 °C, 48 h, 16 % trans-3z.
Mechanistic investigations.
First, we tested whether a second CO ligand coordinates to 2
to form an 18 VE complex 2+CO. However, under 10 bar of CO
no change of the 1H NMR signals of 2 was observed. Moreover,
in the 13C NMR spectrum we see two separate CO signals, one
for the iridium complex 2 and one for free CO. This confirms that
no fast exchange of the CO ligand on the NMR timescale occurs.
In the IR spectrum (toluene) at atmospheric pressure of CO,
neither a shift nor an additional band was observed, and under
CO atmosphere only complex 2 crystallized out. Therefore, we
This article is protected by copyright. All rights reserved.

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
conclude that the 16 VE complex 2 is the catalytic active species.
Also for rhodium complex 1 we had not observed any exchange
of the CO ligand by 13CO under 3 bar, neither under UV light nor
by heating at 60-70 °C.[26]
platina-γ-lactone (1644 cm-1 (NaCl)), the only structurally
characterized 2-metalla-γ-lactone in literature.[34] and to an acyclic
iridaester (1638 (benzene) and 1650 cm-1 (KBr)).[35] When
monitoring the catalytic reaction NMR-spectroscopically (with
Lewis acid) at 60 °C over a period of 23 h, this intermediate can
also be detected. Its concentration reaches quickly a maximum of
5 mol% and slowly declines during the course of the reaction (see
Supporting Information). This shows, that the epoxide activation
step A and the decarboxylation step E have about roughly the
same reaction rate under the applied conditions for substrate 3a.
tBu
tBu
Li+
O
X–
[LiX]
O
+
CO2
N
+IIr
R
A
N
N
N
R
R
B
B'
N
CO
E
ox. Add.
SN2 (Inversion)
CO
2
(Retention)
- LiX
The reaction of cis- and trans-1,2-dialkyl-substituted epoxide
(cis-3j, trans-3j) led to retention of the configuration in the
respective products (cis-4j and trans-4j) and to inversion in the
case of 1,2-diethylcarboxylate-substituted epoxides (3z). As an
epoxide opening that follows an SN2 mechanism would lead to
inversion of the configuration in a possible intermediate Int-1b,
we conclude that the activation step B proceeds via oxidative
addition under C-O bond cleavage to the iridaoxetane[10,11,36] Int-
1a with the substrates cis-3j and trans-3j. This is followed by CO
induced migration of the alkoxide from Ir to a CO ligand[35,37,38] to
form 5j. Under the applied conditions, Int-1b could also be formed
from Int-1a by Lewis acid ring opening followed by CO
coordination and lactone formation by nucleophilic addition of the
alkoxide to CO[39,40] to form 5j without change of the configuration.
Attempts to detect 5j from 1,2-dialkylsubstituted epoxides 3j failed,
as these substrates do not react without presence of the Lewis
acid. During the catalytic conditions, no intermediate was
observed which counts for the oxidative addition B to be the rate-
determining step with these substrates. In case of the ester
functionalized epoxides 3z the intermediate 5z was obtained
much easier without Lewis acid compared to propylene oxide (3a)
as a substrate. The 3JHH coupling between the methine protons of
the metallalactone moiety of trans-5z (obtained from cis-3z) of
6.8 Hz and that of cis-5z (obtained from trans-3z) of 3.2 Hz,
confirm the inversion of the configuration. Proof for the configu-
ration as well as for the formation of the 2-irida-γ-lactone was
obtained from the X-ray crystal structure analyses (Figure 3). To
our knowledge, this is the second example of a structurally
characterized 2-metalla-γ-lactone.
tBu
tBu
tBu
tBu
tBu
tBu
X
X
N
N
N
CO
N
N
N
CO
N
N
CO N
N
N
N
N
+III Ir
+III Ir
+IIIIr
N
O
N
CO
O
Li
O
R
OLi
R
R
- LiX
LiX
6
Int-1a
Int-1b
CO
tBu
tBu
D
C
LiX
CO
N
LiX
CO N
N
N
+IIIIr
N
O
O
R
5
Scheme 3. Proposed mechanism of the deoxygenation of epoxides with CO
catalysed by iridium complex 2. The Lewis acid co-catalyst is necessary to
activate the epoxide (A) as well as to induce the decarboxylation (D, E). The
involvement of Int-1a depends on the substrate: 1,2-dialkyl epoxides react to
Int-1a in an oxidative addition of the epoxide C-O bond, while cis- and trans-
1,2-diethylcarboxyl epoxide get activated through an SN2 mechanism.
The first step of the catalytic cycle (Scheme 3) is the cleavage
of the C-O bond of the epoxide (B). This step requires pre-
activation of the epoxide by the Lewis acid (step A), as it is already
known from the nucleophilic epoxide isomerization with rhodium
catalyst 1 and its congeners.[28,32] Although the complete catalytic
cycle does not operate in absence of the Lewis acid (Table 2,
entry 1), the C-O bond cleavage of propylene oxide still occurs at
80 °C, albeit extremely slowly (B’). We took advantage of this
observation and reacted 2 with propylene oxide (3a) and 10 bar
of CO at 80 °C for 10 days, and were able to isolate intermediate
5a. Due to the presence of the CO atmosphere, further reaction
to acetone or a hydridoalkyl complex[33] is blocked. The identity of
5a, which contains a hitherto unprecedented 2-irida-γ-lactone
moiety, was proven by spectroscopic methods. The chiral center
of the metallalactone moiety reduces the symmetry of the
complex so that 8 signals for the aromatic H atoms and two for
the N-methyl groups are observed. The two signals of the
diastereomeric methylene protons of the lactone ring are detected
at 1.71 and 1.80 ppm and the methine signal at 3.92 ppm. From
the 13C NMR spectrum further evidence for the formation of the
CO complex 5a was obtained (172.6 ppm (IrCO2R), 182.8 (CO),
78.8 (O-CH(CH3)-), and 27.6 (Ir-CH2) ppm). The IR spectrum
(ATR) confirms this with characteristic bands at 2014 (CO) and
1630 cm-1 (IrCO2R). The latter value is similar to the one of a 2-
Figure 3. Molecular structure of the isolated intermediates cis-5z from (trans-
3z) (left) and trans-5z (from cis-3z) (right) from reactions of 2 and 3z without
added Lewis acid. Atoms are shown with anisotropic atomic displacement
parameters at the 50% probability level for the lactone moiety and the rest in
wire-frame style for clarity reasons. Hydrogen atoms (except for the iridacycle)
as well as co-crystallized benzene molecules are omitted for clarity.
This article is protected by copyright. All rights reserved.

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
An inversion of the configuration was also observed by Dowd
and Kang, using stoichiometric amounts (referring to CO) of
Co2(CO)8 in the reaction with 3z.[12] They suggested an analogous
intermediate. Their reaction occurred already at room tempe-
rature, albeit in neat epoxide. SN2 reactions with Rh(I) and Ir(I)
complexes are typically observed in the oxidative addition of alkyl
iodide in the Monsanto or Cativa acetic acid process,[41] while me-
tallaoxetane formation from oxiranes with Rh(I) and Ir(I) was in-
vestigated by Milstein and co-workers.[33,42] In our case, it seems
energetically favored for iridium complex 2 to open the electron
poor epoxide 3z in an SN2 fashion and the more electron rich
internal epoxide 3j by oxidative addition of the C-O bond.
with a stronger Ir-alkyl bond. It is known that migratory insertion
to form acyl ligands occurs slowly in iridium complexes and is e.g.,
the rate limiting step in the Cativa process.[44,45]
The accumulation of intermediate 5 in absence of a Lewis acid
also means that the subsequent decarboxylation step E requires
the presence of a Lewis acid as well. To confirm this, we heated
intermediate 5a in C6D6 (without CO atmosphere) up to 80 °C and
found no CO2 elimination, while addition of the Lewis acid led to
slow formation of the signals of propene (4a) already at room
temperature along with the signals of 2. When stoichiometric
Scheme 4. Literature known formation of β-lactones (left) and their C-O
activation with iridium(I) complexes (bottom). The deoxygenation of epoxides
(right, this work) proceeds via a new isomeric iridalactone which does not
reductively eliminate β-lactones.
amounts of LiNTf2 were added, we observed a slight shift in the Conclusion
1H NMR signals, which is most pronounced for the N-CH3 and the
adjacent imidazole signals (see Supporting Information Figure
S1). Therefore, we propose formation of Lewis acid adduct 6 (step
D) to be mandatory for the CO2 elimination (step E). In the case
of the isolated intermediates cis-5z and trans-5z the CO2
elimination step required even heating to 60 °C and thus is the
rate limiting step using these substrates.
In summary we presented a new homogeneous catalyzed
deoxygenation of epoxides that uses CO directly as a traceless
and environmentally benign deoxygenation agent. Especially
terminal alkyl epoxides react smoothly and without significant
isomerization to the internal olefins. Internal epoxides react under
either retention or inversion of the configuration, depending on
their substituents.[46] This can be explained by two different modes
of the epoxide opening. Either by an oxidative addition of the
epoxy-CO bond which leads to retention of the configuration or by
an SN2-pathway under inversion of the configuration. Various
iridalactones 5 were isolated and in some cases structurally
characterized. Under stoichiometric conditions, the coordination
of the Lewis acid to 5 forming 6 is observed, from which the olefin
is released. Thus, the role of the Lewis acid is not only pre-
activation of the epoxide, but also inducing the CO2 elimination to
produce the product olefin.
As β-lactones are known to easily eliminate CO2 under
elevated temperatures or Lewis acidic conditions, the reductive
elimination of β-lactones from intermediate 6 and subsequent CO2
elimination could also be a possible way to obtain the olefin.
However, as we observed olefin formation from intermediate 5a
in the presence of a Lewis acid already at room temperature
without detecting any signals of a β-lactone, we are convinced
that the olefins are formed by direct CO2 elimination from inter-
mediate 6. In addition, literature known syntheses of β-lactones
often proceed under elevated temperatures as well. Moreover, we
could not observe the reverse reaction, the oxidative addition at
2, neither with nor without the presence of the Lewis acid and only
very slow deoxygenation at 80 °C of the β-lactones to the olefin
(after several days). In contrast, Milstein and co-workers have
shown that electron rich 16 VE iridium(I) complexes oxidatively
add β-propriolactones readily at low temperatures by Calkyl-O bond
cleavage, thus forming 4-irida-γ-lactones (Scheme 4, bottom).[43]
C-Cacyl bond activation, which would form 2-irida-γ-lactones (like
in 5), had not been observed.
Experimental Section
General Information. Unless otherwise stated, all reactions were carried
out under an argon atmosphere in dried and degassed solvents using
Schlenk technique. Toluene, pentane, 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. Carbon
monoxide was purchased from Westfalen with a purity of 99.97 %. High
pressure NMR scale experiments were performed in heavy or medium wall
pressure valve NMR tubes (Wilmad). 1H NMR spectra of catalytic
experiments were recorded with an increased delay time d1 of 60 s to
insure reliable integration values. See Supporting Information for the
numbering scheme of the compounds.
To answer the question about the particularity of our systems
1 and 2 in comparison with Co2CO8 catalyzed CO/epoxide
reactions that produce polymers or β-lactones is the fact that no
migratory insertion step of the alkyl group to the carbonyl ligand
is involved after epoxide opening as the molecular structure of
intermediate 5 revealed. Fast CO migratory insertion after
epoxide opening and C-O bond formation to obtain a 2-metalla-
oxolan-3-one (Scheme 4, left) is usually considered the key step
in the formation of β-lactones.[11,30] In contrast, a nucleophilic
attack of the alkoxide O-atom at the CO ligand forms the 2-irida-
γ-lactone in our case (Scheme 4, right). This can be explained
Synthesis of the catalyst [Ir(bimcaMe)(CO)] (2). Benzyl potassium (58.6
mg, 450 µmol) and [Ir(acac)(CO)2] (52.1 mg, 150 µmol) were added to a
suspension of HbimcaMe•2HI (104.3 mg, 150 µmol) in 12 mL of toluene at
This article is protected by copyright. All rights reserved.

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
room temperature and stirred for 24 h. The resulting yellow suspension
was filtered, and the filtrate dried in vacuo to obtain the desired product as
a yellow solid (96 mg, 91%). 1H NMR (400 MHz, C6D6) = 1.54 (s, 18H, H-
11), 3.81 (s, 6H, H-14), 6.14 (d, J = 2.2 Hz, 2H, H-2), 7.31 (d, J = 2.2 Hz,
2H, H-4), 7.64 (d, J = 1.6 Hz, 2H, H-4’), 8.48 (d, J = 1.6 Hz, 2H, H-5’). The
NMR data in thf-d8 is identical to that of a sample obtained with a Li-base[27]
however, using benzyl potassium gives a much cleaner reaction.
(C28), 60.5 (C23), 79.8 (C18), 110.7, 110.8, 115.6, 116.2 (C2+C4+C5+C7),
116.4, 117.9 (C5’, C10’), 124.2, 124.8 (C1+C8 or C4a+C5a), 125.8, 127.0
(C4’+C9’), 127.9, 128.0 C1+C8 or C4a+C5a), 134.1, 134.9 (C1a+C8a),
139.0 (C3+C6), 144.3, 144.7 (C2’, C7’), 168.8 (C16), 170.4 (C21), 179.3
(CO), 180.3 (C26). ESI+ (MeCN): m/z 875.28 [M]+, 848.31 [M-CO+H]+. IR
(ATR, cm-1): 2034 (s, CO), 1747 (m, ester), 1691 (m, ester), 1645 (m,
lactone).
Trans-5z: 1H NMR (C6D6, 400 MHz): δ = -0.17 (t, J = 7.2 Hz, 3H, H-29),
0.76 (t, J = 7.2 Hz, 3H, H-24), 1.47 (s, 9H, H-11/13), 1.48 (s, 9H, H-11/13),
2.36 (dq, J = 10.6, 7.1 Hz, 1H, H-28), 3.20 (dq, J = 10.5, 7.3 Hz, 1H, H-28),
3.58 (dq, J = 10.4, 7.1 Hz, 1H, H-23), 3.60 (d, J = 3.2 Hz, 1H, H-19), 3.76
(dq, J = 10.7, 7.1 Hz, 1H, H-23), 4.04 (s, 3H, H-14/15), 4.05 (s, 3H, H-
14/15), 5.46 (d, J = 3.2 Hz, 1H, H-18), 6.12 (d, J = 2.0 Hz, 1H, H-4’/9’),
6.16 (d, J = 1.9 Hz, 1H, H-4’/9’), 7.14 (d, J = 2.1 Hz, 1H, H-5’/10’), 7.37 (d,
J = 2.1 Hz, 1H, H-5’/10’), 7.41 (d, J = 1.1 Hz, 1H, H-2/7), 7.43 (d, J = 1.1
Hz, 1H, H-2/7), 8.36 (d, J = 1.3 Hz, 1H, H-4/5), 8.38 (d, J = 1.4 Hz, 1H, H-
4/5). 13C NMR (C6D6,101 MHz): δ = 12.8 (C29), 14.5 (C24), 32.8
(C11+C13), 35.3 (C10+C12), 38.7 (C19), 41.4 (C14+C15), 59.6 (C28),
60.8 (C23), 79.6 (C18), 110.6, 110.7 (C2, C7), 115.7, 115.9 (C4, C5),
116.4, 117.5 (C5’, C10’), 124.7, 124.9 (C4’, C9’), 125.9, 126.4, 127.9,
128.0 (C1, C8, C4a, C5a), 134.6, 134.8 (C1a, C8a), 139.0 (C3+C6), 142.8,
144.5 (C2’, C7’), 168.7 (C16), 173.0 (C21), 181.5, 181.7 (CO, C26). ESI+
(MeCN): m/z 875.28 [M]+, 848.31 [M-CO+H]+. IR (ATR, cm-1): 2035 (s, CO),
1738 (m, ester), 1683 (m, ester), 1645 (m, lactone).
General procedure for the catalytic deoxygenation. 2 (3.3 mg, 5.0
µmol), lithium bis(trifluoromethylsulfonyl)imide (8.6 mg, 30 µmol) and a
certain amount of 1,3,5-trimethoxybenzene as internal standard were
dissolved in 0.5 mL of benzene-d6 or toluene-d8 in a pressure NMR tube.
Then 100 µmol of epoxide were added and the NMR-ampule was
pressurized with 10 bar CO, and heated in an oil bath at 80 °C, if not
otherwise noted. The yield was determined via 1H NMR.
General procedure for the 1 mmol-scale catalytic deoxygenation. 2
(33 mg, 50 µmol), lithium bis(trifluoromethylsulfonyl)imide (86 mg, 0.30
mmol) were dissolved in 5 mL of benzene in a 100 mL Schlenk flask. Then
1.0 mmol of epoxide was added, the argon atmosphere exchanged for CO
at ambient pressure and the reaction mixture heated in an oil bath at 80 °C
for 3 days. The product was then vacuum transferred from the reaction into
a trapped cooled with liquid nitrogen and solvent then distilled off. Due to
the lack of fractional distillation the product still contains 0.5 equivalents of
benzene and losses of about 15% of product in the benzene fraction.
X-ray structure analysis. Data collection was carried out on a
Bruker APEX Duo CCD with an Incoatec IµS Microsource with a
Quazar MX mirror using Mo Kα radiation (λ = 0.71073 Å) and a
graphite monochromator. Corrections for absorption effects were
applied using SADABS.[47] All structures were solved by direct
methods using SHELXS and refined using SHELXL.[48] CCDC
Synthesis of 5a. Synthesis of 5a. 2 (6.6 mg 10 µmol) and epoxypropane
(1.4 µL, 20 µmol) were dissolved in 0.5 mL of C6D6 in a pressure NMR
tube and pressurized with 10 bar CO. The reaction mixture was heated to
80 °C for 10 d. The solvent was evaporated and the residue extracted with
DCM. After concentration do dryness the residue was washed with
pentane to obtain 5a as a pale-yellow solid (Yield: NMR: 89 %; isolated:
1951759
(trans-5z),
1951760
(cis-5z),
1951761
1.6 mg, 23 %). 1H NMR (C6D6, 500 MHz): δ = 1.27 (d, 3JHH = 6.0 Hz, 3H,
([Ir(bimca)(CO)Br2]) and 1951762 ([Ir(bimca)(CO)I2]) 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.
2,3
H-20), 1.50 (s, 9H, H-11), 1.51 (s, 9H, H-13), 1.71 (dd,
J
= 11.0, 5.8
HH
Hz, 1H, H-19), 1.80 (dd, 2,3JHH = 11.0 Hz, 11.0 Hz, 1H, H-19), 3.77 (s, 3H,
H-14), 3.89 (s, 3H, H-15), 3.92 (ddq, JHH = 11.0, 5.8 Hz, 6.0 Hz, 1H, H-
3
18), 5.91 (d, 3JHH = 2.2 Hz, 1H, H-4’ or 9’), 5.93 (d, 3JHH = 2.2 Hz, 1H, H-4’
or 9’), 7.17 (d, 3JHH = 2.2 Hz, 1H, H-5’ or 10’), 7.20 (d, 3JHH = 2.2 Hz, 1H,
H-5’ or 10’), 7.42 (d, 4JHH = 1.5 Hz, 1H, H-2 or 7), 7.43 (d, 4JHH = 1.5 Hz,
1H, H-2 or 7), 8.37 (d, 4JHH = 1.5 Hz, 1H, H-4 or 5), 8.38 (d, 4JHH = 1.5 Hz,
1H, H-4 or 5). 13C NMR (C6D6, 125 MHz): δ = 23.7 (C20), 27.6 (C19), 32.8
(C11 or C13), 35.3 (C10 or C12), 40.6 (C15), 41.1 (C14), 78.8 (C18), 110.7,
110.8 (C2, C7), 115.8, 116.0 (C4, C5), 117.1, 117.1 (C5’, C10’), 124.7
(C4’), 125.1 (C1+C8), 125.3 (C9’), 128.0 (C4a+5a), 134.6, 135.9 (C1a,
C8a), 138.8, 138.9 (C3, C6), 147.5 (C7’), 148.6 (C2’), 172.6 (C16), 182.8
(CO). ESI+ (MeCN): m/z 718.3 [M-CO+H]+. Anal. Calcd. for C33H38IrN5O3:
C, 53.21; H, 5.14; N, 9.40. Found: C, 53.27; H, 5.24; N, 9.52. IR (ATR, cm-
1): 2014 (m, CO), 1630 (w, lactone).
Acknowledgements
Theo Maulbetsch and Eva Jürgens thank the MWK-BW for a
fellowship (Landesgraduiertenförderung). We are grateful to
Yingying Tian for providing some of the epoxide starting material
and Ronja Jordan for a sample of [Ir(bimcaC5)(CO)], to Cäcilia
Maichle-Mössmer for help with the X-ray structure analyses, to
Klaus Eichele and Kristina Strohmaier for measuring all NMR
samples under CO pressure and to Dominik Brzecki for operating
the 600 NMR spectrometer. We also acknowledge help of Fotios
Fotakis with catalytic experiments of the styrene oxide substrates.
Synthesis of the Intermediates cis- and trans-5z. 2 (9.9 mg, 15 µmol)
and 2.8 mg (15 µmol) of either cis- or trans-diethyl epoxy succinate (3z)
were dissolved in 0.5 mL of C6D6 in a pressure NMR tube and pressurized
with 10 bar CO. The reaction mixture was heated to 80 °C for 1 d. Single
crystals suitable for X-ray diffraction were obtained by evaporation of the
solvent at room temperature.
Cis-5z: 1H NMR (Tol-d8, 600 MHz): δ = -0.11 (t, J = 7.2 Hz, 3H, H-29), 0.87
(t, J = 7.0 Hz, 3H, H-24), 1.46 (s, 9H, H-11), 1.49 (s, 9H, H-13), 2.43 (dq,
J = 10.5, 7.2 Hz, 1H, H-28), 3.15 (dq, J = 10.5, 7.2 Hz, 1H, H-28), 3.44 (d,
J = 6.8 Hz, 1H, H-19), 3.78 (s, 3H, H-14/15), 3.85 (dq, J = 10.9, 7.0 Hz,
1H, H-23), 4.07 (dq, J = 10.9, 7.0 Hz, 1H, H-23), 4.09 (s, 3H, H-14/15),
4.25 (d, J = 6.8 Hz, 1H, H-18), 6.33 (s, 2H, H-5’ and 10’), 7.42 (d, J = 1.7
Hz, 1H, H-4’ or 9’), 7.47 (s, 2H, H-4/5 or 2/7), 7.49 (d, J = 1.7 Hz, 1H, H-4’
or 9’), 8.32 (d, J = 1.1 Hz, 1H, H-4/5 or 2/7), 8.32 (d, J = 1.1 Hz, 1H, H-4/5
or 2/7). 13C NMR (Tol-d8, 151 MHz): δ = 13.0 (C29), 14.6 (C24), 32.7, 32.8
(C11, C13), 35.3 (C10+C12), 39.4 (C19), 41.2, 41.7 (C14, C15), 59.1
Keywords: Deoxygenation • Epoxides • Homogeneous
Catalysis • Iridium • Pincer Ligands
References
[1]
A. Ambrosi, S. E. Denmark, Angew. Chem., Int. Ed. 2016, 55, 12164–
12189.
[2]
D. Chusov, B. List, Angew. Chem., Int. Ed. 2014, 53, 5199–5201.
This article is protected by copyright. All rights reserved.

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
[3]
A. A. Tsygankov, M. Makarova, D. Chusov, Mendeleev Commun. 2018,
[32] a) Y. Tian, E. Jürgens, D. Kunz, Chem. Commun. 2018, 54, 11340–
11343; b) Y. Tian, E. Jürgens, K. Mill, R. Jordan, T. Maulbetsch, D. Kunz,
ChemCatChem 2019, 11, 4028–4035.
28, 113–122.
[4]
A. P. Moskovets, D. L. Usanov, O. I. Afanasyev, V. A. Fastovskiy, A. P.
Molotkov, K. M. Muratov, G. L. Denisov, S. S. Zlotskii, A. F. Smol’Yakov, D. A.
[33] D. Milstein, J. C. Calabrese, J. Am. Chem. Soc. 1982, 104, 3773–3774.
[34] J. Wu and R. P. Sharp, Organometallics 2008, 27, 4810–4816.
[35] K. A. Bernard, J. D. Atwood, Organometallics 1989, 8, 795–800.
[36] D. P. Klein, J. C. Hayes, R. G Bergman, J. Am. Chem. Soc. 1988, 110,
3704–3706.
Loginov, et al., Org. Biomol. Chem. 2017, 15, 6384–6387.
[5]
[6]
[7]
[8]
G. Wittig, W. Haag, Chem. Ber. 1955, 88, 1654–1666.
E. Vedejs, P. L. Fuchs, J. Am. Chem. Soc. 1971, 93, 4070–4072.
E. Vedejs, P. L. Fuchs, J. Am. Chem. Soc. 1973, 95, 822–825.
a) K. P. Gable, E. C. Brown, Synlett 2003, 2243–2245; b) K. P. Gable,
[37] While CO insertion is preferred to proceed into Ir-O over the Ir-C bond, it
is different for Ru oxetanes: J. F. Hartwig, R. G. Bergman, R. A. Andersen, J.
Am. Chem. Soc. 1990, 112, 3234–3236.
E. C. Brown, J. Am. Chem. Soc. 2003, 125, 11018–11026.
[9] a) T. Nakagiri, M. Murai, K. Takai, Org. Lett. 2015, 17, 3346–3349; b)
S. Asako, T. Sakae, M. Murai, K. Takai, Adv. Synth. Catal. 2016, 358, 3966–
3970.
[38] Group 10 complexes (d8) prefer CO insertion into the M-OR over the M-R
bond: a) H. E. Bryndza,, Organometallics 1985, 4, 1686–1687; b) R. A.
Michelin, M. Napoli, R. Ros, J. Organomet. Chem. 1979, 175, 239–255; DFT:
c) S. A. Macgregor, G. W. Neave, Organometallics 2003, 22, 4547–4556.
[39] 2-Rhodaoxetanes can be activated by protonation and insert acetonitrile:
a) B. De Bruin, M. J. Boerakker, R. de Gelder, J. M. M. Smits, A. W. Gal,
Angew. Chem. Int. Ed. 1999, 38, 219–222; b) B. De Bruin, M. J. Boerakker, R.
A. W. Verhagen, R. de Gelder, J. M. M. Smits, A. W. Gal, Chem. Eur. J. 2000,
6, 298–312.
[10] K. A. Jørgensen, B. Schioett, Chem. Rev. 1990, 90, 1483–1506.
[11] A. Dauth, J. A. Love, Chem. Rev. 2011, 111, 2010–2047.
[12] P. Dowd, K. Kang, J. Chem. Soc. Chem. Commun. 1974, 384–385.
[13] H. Alper, D. Des Roches, Tetrahedron Lett. 1977, 18, 4155–4158.
[14] J. M. Concellón, E. Bardales, Org. Lett. 2002, 4, 189–191.
[15] M. Mahesh, J. A. Murphy, H. P. Wessel, J. Org. Chem. 2005, 70, 4118–
4123.
[16] B. Xiao, Z. Niu, Y.-G. Wang, W. Jia, J. Shang, L. Zhang, D. Wang, Y.
Fu, J. Zeng, W. He, et al., J. Am. Chem. Soc. 2015, 137, 3791–3794.
[17] J. Yu, Y. Zhou, Z. Lin, R. Tong, Org. Lett. 2016, 18, 4734–4737.
[18] H. C. Volger, H. Hogeveen, C. F. Roobeek, Recl. des Trav. Chim. des
Pays-Bas 2010, 92, 1223–1231. All reactions in references 18-20 were
performed without the presences of CO gas and in chloroform, which is
proposed to act as the deoxygenation agent as phosgene was detected.
Although carbonyl containing catalysts (15 or 20 mol% of [Rh(CO)2Cl]2) were
used and produced some CO2, full conversion can only be explained with the
presence of chloroform.
[40] Observed for Mn and Re carbonyl complexes: J. C. Selover, G. D.
Vaughn, C. E. Strouse, J. A. Gladysz, J. Am. Chem. Soc. 1986, 108, 1455–
1462.
[41] P. M. Maitlis, A. Haynes, G. J. Sunley, M. J. Howard, J. Chem. Soc. -
Dalton Trans. 1996, 2187–2196.
[42] M. J. Calhorda, A. M. Galvao, C. Unaleroglu, A. A. Zlota, F. Frolow, D.
Milstein, Organometallics 1993, 12, 3316–3325.
[43] A. A. Zlota, F. Frolow, D. Milstein, Organometallics 1990, 9, 1300–
1302.
[44] D. Forster, J. Chem. Soc. Dalton Trans. 1979, 1639.
[45] A. Haynes, P. M. Maitlis, G. E. Morris, G. J. Sunley, H. Adams, P. W.
Badger, C. M. Bowers, D. B. Cook, P. I. P. Elliott, T. Ghaffar, et al., J. Am.
Chem. Soc. 2004, 126, 2847–2861.
[19] a) R. Roulet, J. Wenger, M. Hardy, P. Vogel, Tetrahedron Lett. 1974,
15, 1479–1482; b) R. W. Ashworth, C. A. Berchtold, Tetrahedron Lett. 1977,
18, 339–342.
[20] J.-M. Huang, Z.-Q. Lin, D.-S. Chen, Org. Lett. 2012, 14, 22–25.
[21] J. E. Ziegler, M. J. Zdilla, A. J. Evans, M. M. Abu-Omar, Inorg. Chem.
2009, 48, 9998–10000.
[46] During the course of publishing this work, deoxygenation of internal
alkyl epoxides with a Mn/Al – system that proceeds under inversion of the
configuration was presented. L. R. Lamb, A. K. Hubbell, S. N. MacMillan, G.
W. Coates, J. Am. Chem. Soc. 2020, 142, 8029-8035.
[22] A. Noujima, T. Mitsudome, T. Mizugaki, K. Jitsukawa, K. Kaneda,
Angew. Chem., Int. Ed. 2011, 50, 2986–2989.
[47] G. M. Sheldrick, SADABS 2012/1; University of Göttingen, Göttingen,
Germany, 2012.
[23] T. Mitsudome, A. Noujima, Y. Mikami, T. Mizugaki, K. Jitsukawa, K.
Kaneda, Angew. Chem., Int. Ed. 2010, 49, 5545–5548.
[24] T. Mitsudome, A. Noujima, Y. Mikami, T. Mizugaki, K. Jitsukawa, K.
Kaneda, Chem. - Eur. J. 2010, 16, 11818–11821.
[48] a) G. M. Sheldrick, Acta Cryst. 2008, A64, 112-122; b) G. M. Sheldrick,
Acta Cryst. 2015, C71, 3-8; c) C. B. Hübschle, G. M. Sheldrick, B. Dittrich, J.
Appl. Cryst. 2011, 44, 1281−1284.
[25] K. Kaneda, H. Matsuda, Method for Producing Alkene by Deoxidation
of Epoxy Compound, 2010, JP2011236165A.
[26] M. Moser, B. Wucher, D. Kunz, F. Rominger, Organometallics 2007,
26, 1024–1030.
[27] A. Seyboldt, B. Wucher, M. Alles, F. Rominger, C. Maichle-Mössmer,
D. Kunz, J. Organomet. Chem. 2015, 775, 202–208.
[28] E. Jürgens, B. Wucher, F. Rominger, K. W. Törnroos, D. Kunz, Chem.
Commun. 2015, 51, 1897–1900.
[29] J. T. Lee, P. J. Thomas, H. Alper, J. Org. Chem. 2001, 66, 5424–5426.
[30] T. L. Church, Y. D. Y. L. Getzler, C. M. Byrne, G. W. Coates, Chem.
Commun. 2007, 657–674.
[31] D. Kunz, E. Jürgens, T. Maulbetsch, DE 102018003492 A1, German
patent and trademark office (DMPA), 30.04.2018.
This article is protected by copyright. All rights reserved.

10.1002/chem.202002651
Chemistry - A European Journal
RESEARCH ARTICLE
Entry for the Table of Contents (Please choose one layout)
Layout 1:
RESEARCH ARTICLE
Theo Maulbetsch, Eva Jürgens, Doris
Kunz*
Page No. – Page No.
Iridium Catalyzed Deoxygenation of
Epoxides with Carbon Monoxide
Deoxygenate it! epoxides: Carbon monoxide is the direct and environmentally benign
deoxygenation agent for epoxides. This reaction is catalyzed by iridium pincer complex
2 and co-catalyst LiNTf2 via an unprecedented γ-2-iridabutyrolactone intermediate.
LiNTf2 is not only necessary for the epoxide activation, but also for the decarboxylation
of the iridalactone.
This article is protected by copyright. All rights reserved.