[Co(MeTAA)] Metalloradical Catalytic Route to Ketenes via Carbonylation of Carbene Radicals

Article abstract of DOI:10.1002/ejic.201800101

An efficient synthetic strategy towards beta-lactams, amides, and esters involving “in situ” generation of ketenes and subsequent trapping with nucleophiles is presented. Carbonylation of carbene radical intermediates using the cheap and highly active cobalt(II) tetramethyltetraaza[14]annulene catalyst [Co(MeTAA)] provides a convenient one-pot synthetic protocol towards substituted ketenes. N-tosylhydrazones are used as carbene precursors, thereby bridging the gap between aldehydes and ketenes. Activation of these carbene precursors by the metalloradical cobalt(II) catalyst affords Co^III-carbene radicals, which subsequently react with carbon monoxide to form ketenes. In the presence of a nucleophile (imine, alcohol, or amine) in the reaction medium the ketene is immediately trapped, resulting in the desired products in a one-pot synthetic protocol. The β-lactams formed upon reaction with imines are produced in a highly trans-selective manner.

Full text of DOI:10.1002/ejic.201800101

A Journal of
Accepted Article
Title: [Co(MeTAA)] Metalloradical Catalytic Route to Ketenes via
Carbonylation of Carbene Radicals
Authors: Andrei Chirila, Kaj M. van Vliet,, Nanda D. Paul, and Bas de
Bruin
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: Eur. J. Inorg. Chem. 10.1002/ejic.201800101
Link to VoR: http://dx.doi.org/10.1002/ejic.201800101

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
[Co(MeTAA)] Metalloradical Catalytic Route to Ketenes via
Carbonylation of Carbene Radicals
Andrei Chirila,[a] Kaj M. van Vliet,[a] Nanda D. Paul,[a,b] and Bas de Bruin[a]*
Abstract: An efficient synthetic strategy towards beta-lactams,
amides and esters involving “in situ” generation of ketenes and
subsequent trapping with nucleophiles is presented. Carbonylation of
carbene radical intermediates using the cheap and highly active
cobalt(II) tetramethyltetraaza[14]annulene catalyst [Co(MeTAA)]
provides a convenient one-pot synthetic protocol towards substituted
ketenes. N-tosylhydrazones are used as carbene precursors, thereby
bridging the gap between aldehydes and ketenes. Activation of these
carbene precursors by the metalloradical cobalt(II) catalyst affords
Co(III)-carbene radicals, which subsequently react with carbon
monoxide to form ketenes. In the presence of a nucleophile (imine,
alcohol or amine) in the reaction medium the ketene is immediately
trapped, resulting in the desired products in a one-pot synthetic
protocol. The β-lactams formed upon reaction with imines are
produced in a highly trans-selective manner.
Introduction
However, most of these reactions are associated with low yields
and rather harsh reaction conditions (i.e. high temperatures and
high CO pressures).[8]
Ketenes are synthetically useful compounds with intriguing
electronic structures, a versatile reactivity and interesting physical
and chemical properties.[ 1 ] Ketenes are usually prepared in
stoichiometric reactions such as base promoted elimination of
acyl chlorides (first shown by Wedekind in 1901),[2] by metal
halide abstraction (exemplified by Staudinger in 1905),[3] via a
Wolff rearrangement from α-diazoketones,[ 4 ] or by (poorly
controlled) thermolysis (pyrolysis) pathways.[5] All these methods
have important limitations for practical multi-gram syntheses,
often involve unstable (intermediates and) starting materials, and
typically require specialized equipment.[1]
A more robust method for preparing ketenes[ 6 ] found in
several organic transformations is the carbonylation of metal-
carbenes[7,8] and “in situ” reaction of the thus produced highly
reactive ketenes with suitable reagents such as imines, amines or
alcohols.7 However, a significant drawback of these methods is
that the transformations usually rely on the use of Fischer-type
carbene complexes (thus leading to fast carbene-carbene
dimerization side reactions in many cases), and the reactions are
typically stoichiometric. Just a few examples have been reported
where the carbonylation is catalysed by a transition metal
complex.[7, 8] One of them uses noble metal catalysts in the form
of palladium complexes,[7c] while others use base metal catalysts.
The synthesis of reactive ketenes formed during the
carbonylation of metal carbenes leads to relevant products, such
as β-lactams formed by [2+2] ketene-imine cycloadditions
(Scheme 1). The β-lactam skeleton is of great importance for the
pharmaceutical industry,[9] as the β-lactam backbone is found in
many antibiotics such as penicillins, cephalosporins,
carbapenems and monobactams. The β-lactam structure strongly
interacts and even covalently binds to the active site of some
bacterial proteins, which eventually leads to their apoptosis.[10]
The [2+2] cycloaddition between ketenes and imines to yield β-
lactams is a reaction discovered by Staudinger, published only
two years after isolation of the first ketene.[1] The mechanism of
the Staudinger synthesis (Scheme 1) starts with nucleophilic
attack of the imine nitrogen to the central carbon of the ketene
moiety, forming an enol-type amide intermediate.[11] If the ketene
is substituted with electron donating substituents, the cyclization
step is fast and due to the default E-configuration of the imine and
subsequent conrotatory ring closure, the products formed are
generally the cis-isomers. If the ketene lacks strongly electron-
donating substituents, the cyclization step is slower, thus allowing
isomerization of the enol intermediate resulting in formation of
(thermodynamically favoured) trans-β-lactam isomers (Scheme
1).[11]
[a] Dr. A. Chirila, M.Sc. K.M. van Vliet, Dr. N.D. Paul, Dr. W.I. Dzik, and
Prof. Dr. B. de Bruin - Homogeneous, Supramolecular and Bio-Inspired
Catalysis (HomKat), van `t Hoff Institute for Molecular Sciences (HIMS),
University of Amsterdam, Science Park 904, 1098XH, Amsterdam, The
Netherlands, E-mail: B.deBruin@uva.nl
Homepage: http://www.homkat.nl/
[b] Dr. N.D. Paul - Present address: Indian Institute of Engineering Science
and Technology, Shibpur, Botanic Garden, Howrah 711 103, West
Bengal, India
Supporting information for this article is given via a link at the end of the
document.((Please delete this text if not appropriate))
Scheme 1. The Staudinger synthesis of -lactams via [2+2] cycloaddition
reactions.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
The use of cobalt(II) porphyrins [CoII(por)] as stable
metalloradicals with well-defined open-shell doublet d7-electronic
configurations have recently been reported in several
transformations involving catalytic carbene transfer reactions[12]
via stable CoIII-carbene radical intermediates.[ 13 ] A relevant
example is the efficient carbonylation of diazo compounds using
mild temperatures and CO pressures, previously reported in our
group.[14] The nucleophilic nature and radical-type character of the
carbene radical intermediates involved in these reactions is in
marked contrast with that of an electrophilic Fischer-type carbene,
thus preventing unwanted carbene dimerization reactions.
Therefore, the attack of the nucleophilic carbene radical at the
π-accepting carbon monoxide becomes possible, thus allowing
“in situ” generation of ketenes. These ketenes are subsequently
reacted with imines or amines to obtain amides or respectively
β-lactams (Scheme 2). Because the cobalt catalyst and the
carbene radical intermediates are not sensitive to the presence of
amines or imines, the reactions can be performed in a one-pot
catalytic reaction in which the “in situ” generated ketene is trapped
by the added nucleophile.
In this paper we report an easy one-pot cascade method to
form ketenes starting from tosylhydrazones (which are in turn
prepared easily from aldehydes.and tosylhydrazine). This
strategy expands the scope of previous studies by taking
advantage of the outstanding activity of [Co(MeTAA)] to activate
diazo compounds, thus forming carbene radicals intermediates
which react with carbon monoxide to afford ketenes. The “in situ”
catalytically formed ketenes are then trapped with imines, amines
or alcohols to afford beta-lactams, amides and esters as the final
organic products.
Figure 1. Structures of the cobalt(II) catalysts used in this study.
Due to the synthetic challenge in preparing substituted
cobalt(II) porphyrins, we decided to expand the catalytic scope of
this reaction by screening several cheap and easy to prepare
cobalt(II) complexes (Figure 1). Based on our current
understanding of the metalloradical mechanism, more electron-
rich low-spin cobalt(II) complexes facilitate the carbene reduction
and thus the formation of the key carbene radical intermediate.
Electron rich carbenes would be nucleophilic, therefore favouring
the CO carbonylation pathway. A lead candidate would be the
cobalt(II)tetramethyltetraaza[14]annulene complex [Co(MeTAA)]
(Figure 1), successfully used in our recent other studies.[15] It is
easily prepared from readily available staring materials,[16] and
presents a smaller macrocycle than corresponding porphyrins,
thereby inducing more electron density at the metal centre. The
complex proved to be an excellent catalyst for the activation of
diazo compounds and generation of reactive carbene radical
species.[15] As such, we argued that [Co(MeTAA)] should be a
potent catalyst for carbene carbonylation, with subsequent ketene
trapping.
Scheme 2. Proposed mechanism for catalytic ketene synthesis via cobalt(III)-
carbene radical carbonylation.
Results and Discussion
The cobalt(II) tetramethyltetraaza[14]annulene complex
[Co(MeTAA)] can be synthesized using a template condensation
between 2,4-pentandione and 1,2-diaminobenzene in the
presence of cobalt(II) acetate tetrahydrate (Scheme 3). However,
due the sensitivity of the complex towards oxygen and the low
yield of the direct cobalt-template synthesis, a stepwise route is
preferred in which the nickel(II) tetramethyltetraaza[14]annulene
[Ni(MeTAA)] is first prepared using the template synthesis around
nickel.[17] In this case the formed complex is not sensitive to
oxygen, and the complex can easily be obtained in high quantities
using bulk cheap starting materials. Demetallation using gaseous
hydrochloric acid, followed by deprotonation with trimethylamine
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
affords the free tetramethyltetraaza- [14]annulene ligand.
Complexation using cobalt acetate under reflux for a few minutes
produces the crystalline [Co(MeTAA)] complex, which
precipitates from solution in high yield. The thus formed
[Co(MeTAA)] complex was further used in catalysis screening.
9
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
PhMe
THF
-
7.5%
4%
4%
4%
4%
4%
4%
3%
3%
78
34
47
48
8
10
11
12
13
14
15
16[c]
17[c]
-
DCE
-
PhCl
-
MeCN
Dioxane
PhCF3
PhMe
PhMe
-
-
14
11
53
50
-
Pyridine
DMAP
1-Methyl
Imidazole
18[c]
19[d]
[Co(MeTAA)]
[Co(MeTAA)]
PhMe
PhMe
3%
3%
67
73
1-Methyl
Imidazole
Scheme 3. Synthesis of [Co(MeTAA)].
20[e]
21[f]
22[g]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
PhMe
PhMe
PhMe
-
-
-
3%
3%
3%
46
39
52
We decided to first focus on the reaction between the
benzaldehyde
tosylhydrazone
sodium
salt
and
N-methylbenzaldimine. In this reaction, phenylketene is
generated in situ, which together with the imine undergoes a [2+2]
cycloaddition resulting in formation of β-lactam 3 as the product
(Table 1). This reaction was used as a benckmark to find the
optimal catalytic conditions. A comparison of the activity of the
planar cobalt(II) complexes depicted in Figure 1 was investigated,
using 1.0 eq of tosylhydrazone, 1.2 eq. of imine, toluene as the
solvent, a CO pressure of 20 bar, and a reaction temperature of
60 °C as the initial reaction conditions.
[a] Reactions were carried out under 20 bar CO in a sealed autoclave with
1.0 eq. of sodium salt of tosylhydrazone and 1.2 eq. of imine for 16 h. Conc.
0.34 mmol tosylhydrazone/3.0 mL solvent. [b] Isolated yields. [c]
stoichiometry catalyst : additive = 1:1 [d] stoichiometry catalyst : additive =
1:2 [e] 10 bar CO [f] 2 bar CO [g] 40 °C
Entries 1-8 from Table 1 present the results of the initial
screening experiments, in which [Co(MeTAA)] affords the best
yield of 77% (entry 8), better than [Co(TPP)] yielding 68% product
(see entry 6). The other cobalt(II) complexes perform poorly,
affording a maximum yield of 16% (for [Co(Sal-tBu)], entry 5).
Control experiments using no catalyst, as well as the simple
[Co(OAc)2] precursor did not result in products formation, thus
showing that the reaction is indeed catalytic and requires a planar
low-spin cobalt(II) metalloradical catalyst (entries 1-2). Catalyst
loading screening (entries 7-9) indicates an optimum at around 3
mol% [Co(MeTAA)].
Table 1. Optimization of the reaction conditions for the synthesis of β-lactam
3
from
benzaldehyde
tosylhydrazone
sodium
salt
1
and
N-benzylidenemethanamine 2.
Solvent screening (entries 8-15) indicates that toluene is the
best solvent for this reaction, producing the product in a yield of
77%. Other solvents such as 1,2-dichloroethane or
chlorobenzene lead to lower yields of 47% (entry 11) and 48%
respectively (entry 12). Polar solvents (entries 13-15) are not
suitable for this reaction as the conversion drops significantly,
yielding the product in only 8% using acetonitrile and in 14% using
dioxane. In some of our previous reports,[15] coordinating additives
had a positive influence on the outcome of the reaction, increasing
the activity of the catalyst. However, in this study (Table 1, entries
16-19) the use of pyridine, 4-dimethylaminopyridine and
1-methylimidazole additives led to lower yields than in reactions
without these additives.
Entry
Catalyst
loading
Yield[b]
(%)
Catalyst
--
Solvent
PhMe
PhMe
Additive
[a]
1
-
-
<1
<5
Co(OAc)2*4H2
O
2
5%
3
4
5
6.
7
8
[Co(Sal-NMe2)]
[Co(Sal-CF3)]
[Co(Sal-tBu)]
[Co(TPP)]
PhMe
PhMe
PhMe
PhMe
PhMe
PhMe
-
-
-
-
-
-
5%
6
5%
10
16
68
71
77
5%
These additives might well hamper the [2+2] cycloaddition of
the ketene with the imine. Variation of the CO pressure (entries
20-21) indicates that β-lactams can be formed at 2 bar (39% yield),
but higher yields are obtained at higher CO pressures. At 10 bar
46% of the β-lactam 3 is obtained, but the highest yield (77%) is
3.5%
2%
[Co(MeTAA)]
[Co(MeTAA)]
3.5%
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
obtained when pressurising the autoclave to 20 bar. Temperature
plays also an important role, as it helps formation of the carbene
radical intermediate from the benzaldehyde tosylhydrazone
sodium salt. Lowering the reaction temperature to 40 °C is
possible, but leads to a drop in the product yield to 52% (Table 1,
entry 22). The reactions in Table 1 proved to be in all cases highly
trans-selective, producing trans β-lactam product 3 in an excellent
diastereoselectivity (>99%).
Based on all information gathered (Table 1), we decided to
continue with the following reaction conditions in the initial
screening experiments of most of the catalytic reactions
described below: 20 bar CO, 60 °C, 3 mol% catalyst and a 1.0:1.2
tosylhydrazone:imine molar ratio. For the more challenging
substrates, higher reaction temperatures were used.
5
-H
-H
-H
-H
60
77
69
6[c]
140
20
7
-Me
-Me
-H
-H
60
8[c]
140
46
33
9
-OMe
-OMe
-H
-H
60
10[c]
140
60
61
62
<5
To explore the versatility of the reaction, the substrate scope
was investigated (Table 2). First, the coupling of a variety of
para-substituted N-methylbenzaldimines to benzaldehyde
tosylhydrazone sodium salt was explored (Table 2, entries 1-5).
The non-substituted imine afforded the highest yield of 77% (entry
5), followed by electron-donating substituents (Me: 62% yield,
entry 1; MeO: 57% yield, entry 2). Introducing electron-
withdrawing substituents at the para-position of the
benzaldimines leads substantial lowering of the product yields (Cl:
52% yield, entry 3; NO2: 36% yield, entry 4).
11
-Cl
-Cl
-H
-H
60
12[c]
140
13
-NO2
-NO2
-H
-H
60
14[c]
140
<5
[a] Reactions were carried out under 20 bar CO in a sealed autoclave with
1.0 eq. of tosylhydrazone salt, 1.2 eq. of imine and 0.03 eq. [Co(MeTAA)].
Conc: 0.34 mmol tosylhydrazone salt/3.0 mL toluene. [b] Isolated yields. [c]
1.0 eq. tosylhydrazone salt, 2.0 eq. imine and 0.04 eq. [Co(MeTAA)]. Conc:
0.34 mmol tosylhydrazone salt/2.0 mL toluene.
Table 2. Substrate scope screening [Co(MeTAA)] catalysed trans-selective
β-lactam synthesis from substituted tosylhydrazone sodium salts and
substituted imines.
Next, the effect of para-substitution on the tosylhydrazone salt
was investigated (entries 6-14). These substrates proved to be
challenging, and thus we needed to increase the reaction
temperature to 140 °C. Apart from that, both electron donating
substituents (Me: 46%, entry 8; OMe: 60%, entry 10) and the
electron withdrawing chloro-substituent (62% yield, entry 12) are
tolerated. Unfortunately, a nitro substituted on the tosylhydrazone
proved to be too challenging, and did not produce any β-lactam.
Remarkably, reactions are possible in pressurized systems at
elevated temperatures of 140 °C showing the stability of the
catalytic system, in contrast to normal [2+2] Staudinger
cycloadditions that are usually performed bellow 0 °C. In all cases,
the reactions proved to be highly trans-selective, producing the
trans β-lactam products in excellent diastereoselectivities >99%.
The reactions in Table 1 and 2 afforded the trans products in
excellent diastereoselectivities (>99%), but as racemic mixtures.
Utilization of chiral cobalt(II) catalysts has no influence on this
reaction, because the ketene is dissociated from the metal centre
before reacting with the imine.[14] This was confirmed in reactions
using chiral catalyst [Co(Sal-tBu)] (Table 3, entry 6). Therefore,
we argued that a chiral co-catalyst could perhaps be used to
induce chirality in the [2+2] ketene-imine cycloaddition. Hence, we
decided to investigate a cascade one-pot reaction that uses the
cobalt(II) catalyst for ketene formation, and subsequently a chiral
co-catalyst that favours the β-lactam formation (Table 3).
Temp
(°C)
Yield[b]
(%)
Entry[a]
-R1
-H
-R2
Product
1
-Me
60
60
60
60
62
57
52
36
-
2
3
4
-H
-H
-H
OMe
-Cl
-NO2
Two chiral co-catalysts have been tested, the iron based
()-PPY* used successfully in enantioselective [2+2]
cycloadditions as reported by Fu and coworkers[18] (entries 2, 4),
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
and quinidine which is a well-known organocatalyst used in a
variety of reactions (entry 5).[19]
one-pot bi-catalytic system is possible, and the reactions are quite
tolerant to a variety of additives despite formation of reactive
intermediates such as cobalt(III) carbene radicals and ketenes in
these reactions.
Table 3. Cocatalyst screening in attempts to enable asymmetric β-lactam
cascade synthesis from benzaldehyde tosylhydrazone sodium salt 1 and
N-benzylidenemethanamine 2.
Table 4. Reaction optimizations for the synthesis of N-benzylbenzamide 5
from
benzaldehyde
tosylhydrazone
sodium
salt
1
and
N-benzylidenemethanamine 4.
Entry
[a]
Yield[b]
(%)
Yield[b]
Catalyst
Cocatalyst
ee [c]
Eq. of
additive
Entry[a]
Catalyst
Additive
(%)
<1
75
40
44
51
46
44
95
73
61
74
75
40
37
73
1
[Co(MeTAA)]
77
-
-
1
-
-
-
2
[Co(TPP)]
K3PO4
-
3
3
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(MeTAA)]
[Co(Sal-tBu)]
[Co(Sal-CF3)]
[Co(Sal-NMe2)]
-
2
3
53
68
<1%
4
K3PO4
Pyridine
2
(-)-PPY*
-
[Co(MeTAA)]
[Co(TPP)]
5
2
-
6
K3PO4 : pyridine
DMAP
1:1
2
7
4
62
<1%
8
DABCO
DABCO
DABCO
DABCO
DABCO
DABCO
DABCO
DABCO
2
(-)-PPY*
9
1
[Co(TPP)]
10
11
12
13
14
15
0.5
0.25
0.1
2
5.
58
16
<1%
<1%
Quinidine
[Co(TPP)]
2
2
6
-
[a] Reactions were carried out under 20 bar CO in a sealed autoclave with
1.0 eq. tosylhydrazone, 2.2 eq. aniline, 0.04 eq. catalyst and additive for 16h.
Conc. 0.4 mmol tosylhydrazone/3.0 mL toluene. [b] Isolated yields.
[Co(Sal-tBu)]
[a] Reactions were carried out under 20 bar CO in a sealed autoclave with
1.0 eq. tosylhydrazone, 1.2 eq. imine, 0.03 eq. catalyst and 0.03 eq.
cocatalyst for 16h. Conc. 0.34 mmol tosylhydrazone salt/3.0 mL solvent. [b]
Isolated yields. [c] enantiomeric excess determined by chiral HPLC analysis.
To show that not only β-lactams can be obtained as products
using the [Co(MeTAA)-catalyzed ketene formation protocol, we
decided to also investigate the possibility to trap the ketene
intermediates with amines and alcohols, in order to form amides
or esters in one-pot reactions. As a benchmark reaction,
benzaldehyde tosylhydrazone sodium salt 1 and aniline 4 were
used as substrates to form N-benzylbenzamide 5. Table 4 shows
the condition screening study, in which different bases are tested
together with [Co(MeTAA)] (entries 3-12). The best yield of 95%
was obtained when using 2.0 eq. of DABCO (entry 8).
Investigating different molar ratios of DABCO proved less
successful (entries 9-12), revealing that at least 2.0 equivalents
are needed to obtain the product in almost quantitative yield. The
bases 4-dimethylaminopyridine, pyridine and potassium
phosphate were also tested, but in these cases the desired
Although the cascade dual catalytic reactions were
compatible in both cases, affording the trans β-lactams in decent
yields (Table 3), no chirality transfer was observed when using
[Co(MeTAA)] or [Co(TPP)] in combination with quinidine or ()-
PPY*. A possible reason for the lack of chirality transfer is the high
reaction temperature of 60 °C used in these reactions, as chirality
transfer with organocatalysts or ()-PPY* typically requires low
temperatures (0 °C or below). Unfortunately we were unable to
reduce the reaction temperature to such required temperature for
efficient chirality transfer, as the ketene formation is too slow at
such a low temperature. However, the reactions do show that a
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
product was obtained in moderate yields (44%, 51% and 44%,
respectively; see Table 4, entries 4-7). The best conditions for
[Co(TPP)] have been screened in a previous study,[14] and the
result is presented in Table 4, entry 2 for comparison. The
afforded yield of 75% was obtained using 3.0 equivalents of
potassium phosphate, and is significantly lower than the 95% one
obtained using [Co(MeTAA)] and DABCO (entry 8). The salen
based cobalt(II) complexes (Figure 1) produced the desired
product in only moderate yields of 40%, 37% and 73% for [Co(Sal-
tBu)], [Co(Sal-CF3)] and [Co(Sal-NMe2)] respectively (entries 13-
15), further confirming that a more electron rich catalyst is better
suited for these radical-type transformations proceeding via
carbene radical intermediates.
13
14
15
16
-H
-H
-H
-H
46
70
43
15
iPr-OH
nBu-OH
sBu-OH
tBu-OH
[a] Reactions were carried out under 20 bar CO in a sealed autoclave with
1.0 eq. tosylhydrazone, 2.2 eq. aniline, 0.04 eq. catalyst and additive for 16h.
Conc. 0.4 mmol tosylhydrazone/3.0 mL toluene. [b] Isolated yields.
Using the abovementioned optimized reaction conditions, we
proceeded to further screen the substrate scope (Table 5). We
first investigated the reaction of para-substituted tosyhydrazone
salts, carbon monoxide and aniline catalysed by [Co(MeTAA)]
(Table 5, entries 1-4). The unsubstituted benzaldehyde
tosylhydrazone sodium salt afforded the desired amide product in
the highest yield (95%), followed by the Me-substituted hydrazone
(90%, entry 2), Cl-substituted hydrazone (82%, entry 4) and MeO-
substituted hydrazone (81% yield, entry 3). Next, reactions using
several para-substituted anilines were investigated (entries 5-8),
and as expected the results show that an electron-donating alkoxy
substituent on the nucleophile is beneficial (yield 90%, entry 5).
Electron withdrawing substituents such as an NO2 (50% yield
(entry 6) or chloro- (61%) and bromo-group (84%) lead to
somewhat lower yields (entries 7-8). Aliphatic and benzylic
amines are also suitable for this reaction (entries 9-11), with
benzylamine affording the amide product in 86% isolated yield.
The last entries in Table 5 show ester formation using differently
substituted alkyl alcohols (entries 12-16). Due to the lower
nucleophilic nature of alcohols as compared to amines, the yields
of these reactions are somewhat lower, with n-butanol affording
the ester product in 70% yield (entry 14), while tert-butanol reacts
poorly producing only 15% of the desired ester product (entry 16).
As can be expected, the more sterically congested and hence less
nucleophilic alcohol produced the ester product in a lower yield as
compared to the linear alcohols.
Table 5. Substrate scope screening in [Co(MeTAA)] catalysed amides and
ester synthesis from tosylhydrazone salts and corresponding nucleophiles.
Entry
Yield
-R
-H
NuH
Product
[a]
[b] (%)
1
2
95
90
-Me
3
4
5
6
-OMe
-Cl
81
82
90
50
-H
-H
Computational Studies
The mechanism of the [Co(MeTAA)]-catalyzed carbene
carbonylation reaction was also investigated computationally,
using DFT (see Supporting Information for details). In line with
similar systems that have previously been studied extensively, the
computations were performed at the BP86 and def2-TZVP level,
using Grimme’s version 3 dispersion corrections (Scheme 4). The
mechanism is similar to the one reported for related [Co(por)]-
catalyzed reactions.[14] The pathway starts with activation of the
diazo compound (which is generated in situ from the
tosylhydrazone salt in an uncatalyzed step) by the catalyst,
involving binding of the diazo compound via its carbon atom to
cobalt (B), followed by N2-loss (TS1) to form carbene radical
intermediate C (see Scheme 4 and Figure 2). This step has a
transition state barrier TS1 of +9.5 kcal mol-1 for R1 = -COOEt and
+7.7 kcal mol-1 when R1 = Ph. This step has a lower energy barrier
when compared to [Co(por)].[14]
7
-H
61
8
9
-H
-H
84
48
nBu-NH2
10
-H
86
11
12
-H
-H
79
55
nPr-OH
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
the catalytic cycle based on these DFT calculations. The free
ketene generated in this manner reacts subsequently with
nucleophiles or imines present in solution, leading to the final
ester, amide or β-lactam products.
20.0
TS1-EDA= 9.5
14.5
10.0
TS1-BzN2= 7.7
9.2
B’
4.9
TS3-EDA=12.8
0.0
1.3
1.5
B
0.0
TS3-BzN2=11.8
A
-3.0
C’
-10.0
-20.0
-30.0
-40.0
-50.0
-11.5
-14.8
C
F
E’
-33.5
-35.4
-3F3.’7
-3E8.4
R = -COOEt
R = -Ph
CoMeTAA-EDA-CO
CoMeTAA-BzN2-CO
Figure 2. Energy diagram comparison of the [Co(MeTAA)] catalyzed ketene
formation starting from ethyldiazoacetate and BzN2. All energies relative to A
(transition state barriers relative to the preceding intermediate also reported).
Conclusions
In summary, we have shown an efficient synthetic strategy of
producing diastereoselective β-lactams, amides and esters using
[Co(MeTAA)] as an effective metalloradical catalyst able to
activate tosyhydrazones to generate carbene radicals, which are
carbonylated to produce ketenes and subsequently trapped by
the applied nucleophile in one-pot reactions. The cheap and
highly active [Co(MeTAA)] catalyst proved to be compatible with
several different reaction conditions, including dual cascade
catalysis, tolerating many functional groups, coordinating
additives and high reaction temperatures. The protocol presented
in this paper provides a reliable method to produce ketenes in a
single pot starting from tosylhydrazone salts and carbon
monoxide. The reactions proceed via formation of carbene radical
intermediates which react with CO to form ketenes. The intrinsic
reactivity of ketenes requires in situ trapping of these reactive
intermediates with amines, alcohols or imines to obtain useful
organic building blocks. The substrate scope has been explored,
showing that this method can be applied to several different
substrates bearing a variety of substituents.
Scheme 4. DFT calculated pathway of the [Co(MeTAA)]-catalysed carbene
carbonylation (DFT-D3, BP86, def2-TZVP). All energies (also the transition
states) are in kcal mol-1 and relative to A + diazo compound; in black are
represented the energies for R1= -COOEt and in red the energies for R1= -Ph.
Carbene radical C can form a bridging carbene D via TS2
which has a barrier of +19.6 kcal mol -1 for R1 = -COOEt and +16.9
-1
kcal mol when R1 = Ph. Unlike what was observed for the
[Co(por)] system, species C is however more stable than species
D for the [Co(MeTAA)] system, thus leading to lower barriers of
the follow-up carbene carbonylation step (see also Figure S1 in
the supporting information). Carbonylation of carbene radical
intermediate C proceeds via CO addition to the carbene moiety
with a significantly lower TS3 transition state barriers (+12.8 kcal
mol -1 for R1 = -COOEt and +11.8 kcal mol -1 for R1 = Ph) than
-1
previously reported for [Co(por)] (+23.7 kcal mol for R1 = -
COOMe). For the [Co(MeTAA)] system, TS3 connects species C
with species E. This is somewhat different from what was
observed for the [Co(por)]-catalyzed reaction in which the ketene
spontaneously dissociates from the metal following the
carbonylation step. For the [Co(MeTAA)] system the ketene
adduct E is formed as an intermediate. However, ketene
dissociation (F) from ketene adduct E requires only a few kcal
mol-1. The computed TS3 energy barriers of the carbene
carbonylation steps and the TS1 barriers for carbene radical
formation are very similar, and the CO concentration (non-
standard conditions) could well have a significant influence on the
relative barriers of these steps (see Figure 2). As such, it is not
possible to predict what is/are the actual rate limiting step(s) of
Experimental Section
General Procedures. All manipulations, except the carbonylation
reactions, were performed under a nitrogen atmosphere using standard
Schlenk techniques. All solvents used for catalysis were dried over and
distilled from sodium (toluene, tetrahydrofuran, diethyl ether) or CaH2
(dichloromethane, hexane, ethyl acetate, methanol, acetonitrile). Unless
specified, all chemicals were purchased from commercial suppliers (Sigma
Aldrich, Acros or Strem) and used without further purification. NMR spectra
(1H, and 13C{1H}) were measured on a Bruker AV400, AV300, DRX 500 or
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
DRX 300 spectrometer or on a Varian Mercury 300 spectrometer,
referenced internally to residual solvent resonance of CHCl3 (δ = 7.26 ppm
for 1H, δ = 77.2 ppm for 13C) or DMSO (δ = 2.50 ppm for 1H, δ = 39.5 ppm
for 13C). Unless noted otherwise, the NMR spectra were measured in
CDCl3. Individual peaks are reported as: multiplicity (s = singlet, d =
doublet, t = triplet, q = quartet, m = multiplet), integration, coupling constant
in Hz. Mass spectra of the synthesized compounds were recorded on an
Agilent-5973 GC-MS spectrometer, and the corresponding HRMS data
were recorded on a JEOL AccuTOF 4G via direct injection probe using
either EI or ESI. Chirality was determined using a Shimadzu HPLC setup
equipped with an UV-Vis detector (SPD-10Avp).
Acknowledgements
We thank Ed Zuidinga for MS measurements, Jan Meine Ernsting
for NMR advice. Financial support from the Netherlands
Organization for Scientific Research (NWO-CW VICI project
016.122.613) and the University of Amsterdam (Research Priority
Area Sustainable Chemistry) is gratefully acknowledged. We
thank Prof. Helma Wennemers (ETH Zürich) for providing a
sample of quinidine and Dr. Wojciech I. Dzik for helpful
discussions.
Catalyst preparation. [Co(MeTAA)] has been synthesized according to
reported procedures.[20] [Co(TPP)] and [Co(salens)] have been purchased
from Sigma-Aldrich or Strem and used without further purification.
Keywords: Ketene • Cobalt • Catalysis • Carbonylation •
Metalloradical • Carbene Radical • N-Tosylhydrazone
General Procedure for Synthesis of the N-tosylhydrazone salts.[21] An
equimolar mixture of corresponding aldehyde and N-tosylhydrazide was
placed in a round bottom flask and dissolved in methanol (2 mL/mmol).
The reaction mixture was stirred overnight at room temperature. The white
precipitate was collected by filtration, and washed with cold methanol and
hexane to obtain the pure product. The formed N-tosylhydrazone was then
deprotonated in methanol using 1 equivalent of sodium methoxide. After
evaporation of methanol, the pure product was obtained as a white powder.
Typical Carbene Carbonylation Procedures. In a typical carbonylation
experiment a stainless steel autoclave (150 mL) equipped with inserts
suitable for five glass vials (4 mL) was employed. The vials were charged
with appropriate amounts of solvent, substrates (N-tosylhydrazone sodium
salts and nucleophiles or imines) and catalyst, together with Teflon stirring
bars. Before starting the catalytic reactions, the charged autoclave was
purged three times with CO and then pressurized to the desired pressure.
After catalysis, the autoclave was cooled to room temperature, and purged
with nitrogen so that any excess CO was removed. The reaction mixture
is then analysed and purified using typical organic work-up, preparative
TLC or flash column chromatography.
CoII-catalysed β-lactam synthesis using different N-tosylhydrazone
sodium salts and imines. Under a nitrogen atmosphere, [Co(MeTAA)] (3
mol%), and the desired N-tosylhydrazone sodium salt (1.0 eq, 0.34 mmol)
were added to a flame-dried 4 ml glass vial. Then the vial was sealed with
a stopper and evacuated. The desired imine (1.2 eq., 0.408 mmol) in 3.0
ml of toluene were added via syringe. A small needle was inserted at the
top of the vial. The vials were then inserted into an autoclave (150 mL)
equipped with inserts suitable for five such glass vials and pressurized to
20 bar of CO. The pressurized autoclave was then stirred at 60o C for 16h.
The autoclave was subsequently cooled to room temperature, and purged
with nitrogen so that any excess CO is removed. The resulting mixture was
concentrated and the residue was purified by preparative TLC or flash
silica gel chromatography.
CoII-Catalysed amide/ester synthesis using N-tosylhydrazone
sodium salt and different nucleophiles. Under a nitrogen atmosphere,
[Co(MeTAA)] (4 mol%), DABCO (2.0 eq., 0.8 mmol), and the desired N-
tosylhydrazone sodium salt (1.0 eq., 0.4 mmol) were added to a 4 ml
flame-dried glass vial. Then the vial was sealed with a stopper and
evacuated. Anhydrous toluene (3.0 mL) and the desired nucleophile
(alcohol or amine) (2.2 eq., 0.88 mmol) were added via syringe. A small
needle was inserted at the top of the vial. The vials were then inserted into
an autoclave (150 mL) equipped with inserts suitable for five such glass
vials and pressurized with 20 bar of CO. The pressurized autoclave was
then stirred at 60 oC for 16h. The autoclave was subsequently cooled to
room temperature, and purged with nitrogen so that any excess CO is
removed. The resulting mixture was concentrated and the residue was
purified by preparative TLC or flash silica gel chromatography.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
References
[1]
[2]
a) T. T. Tidwell, “Ketenes, 2nd edition“, 2006, John Wiley & Sons,
Hoboken, NJ; b) T. T. Tidwell, Angew. Chem. Int. Ed. 2005, 44, 5778. c)
D. H. Paull, A. Weatherwax, T. Lectka, Tetrahedron, 2009, 65, 6771.
M. S. Newman, A. Arkell, T. Fukunaga, J. Am. Chem. Soc. 1960, 82,
2498.
Ragaini, S. Cenini, Organometallics 2008, 27, 6143; i) M. P. Doyle,
Angew. Chem. Int. Ed. 2009, 48, 850; j) J. V. Ruppel, T. J. Gauthier, N.
L. Snyder, J. A. Perman, X. P. Zhang, Org. Lett. 2009, 11, 2273; k) S. F.
Zhu, X. Xu, J. A. Perman, X. P. Zhang, J. Am. Chem. Soc. 2010, 132,
12796.
[3]
[4]
[5]
H. Staudinger, Chem. Ber. 1905, 38, 1735.
[13] a) W. I. Dzik, X. Xu, X. P. Zhang, J. N. H. Reek, B. de. Bruin, J. Am.
Chem. Soc. 2010, 132, 10891. b) H. Lu, W. I. Dzik, X. Xu, L. Wojtas, B.
de. Bruin, X. P. Zhang, J. Am. Chem. Soc. 2011, 133, 8518. c) W. I. Dzik,
X. P. Zhang, B. de Bruin, Inorg. Chem., 2011, 50, 9896. d) W. I. Dzik, J.
N. H. Reek, B. de Bruin, Chem. Eur. J. 2008, 14, 7594.
Y. Chiang, A. J. Kresge, V. V. Popik, J. Am. Chem. Soc. 1999, 121, 5930.
a) H. W. Moore, D. S. Wilbur, J. Org. Chem. 1980, 45, 4483. b) G. Kresze,
W. Runge, E. Ruch, Justus Liebigs Ann. Chem. 1972, 756, 112.
a) K. H. Dötz, Angew. Chem., Int. Ed. Engl. 1975, 14, 644. b) K. H. Dötz,
P. Tomuschat, Chem. Soc. Rev. 1999, 28, 187.
[6]
[7]
[14] N.D. Paul, A. Chirila, H. Lu, X.P. Zhang, B. de Bruin, Chem. Eur. J., 2013,
19 (39), 12953-12958.
a) L. S. Hegedus, Tetrahedron 1997, 53, 4105. b) L. S. Hegedus, Topics
Organomet. Chem. 2004, 13, 157. c) Z. Zhang, Y. Liu, L. Ling, Y. Li, Y.
Dong, M. Gong, X. Zhao, Y. Zhang, J. Wang, J. Am. Chem. Soc. 2011,
133, 4330; d) Z. Zhang, Y. Zhang, J. Wang, ACS Catal. 2011, 1, 1621–
1630.
[15] a) B. G. Das, A. Chirila, M. Tromp, J. N. H. Reek, B. de Bruin, J. Am.
Chem. Soc. 2016, 138, 8968; b) A. Chirila, B.G. Das, N.D. Paul, B. de
Bruin, ChemCatChem, 2017, 9, 1413 – 1421.
[16] a) F. A. Cotton, J. Czuchajowska, Polyhedron 1991, 9, 2553; b) P.
Mountford, Chem. Soc. Rev. 1998, 27, 105.
[8]
a) A. Miyashita, H. Shitara, H. Nohira, J. Chem. Soc. Chem.Commun.
1985, 850. b) M. Huser, M. Youinou, J. A. Osborn, Angew. Chem., Int.
Ed. Engl. 1989, 28, 1386. c) J. Barletta, F. Karimi, H. Doi, B. Långström,
J. Label Compd. Radiopharm. 2006, 49, 801. d) N. Ungvári, E. Fördős,
J. Balogh, T. Kégl, L. Párkányi, F. Ungváry, Organometallics 2010, 29,
3837.
[17] a) J. H. Niewahner, K. A. Walters, A. Wagner, Journal of Chemical
Education 2007 84 (3), 477; b) J. R. Chipperfield, S. Woodward, Journal
of Chemical Education 1994 71 (1), 75.
[18] (a) J.C. Ruble, H.A. Latham, G.C. Fu, J. Am. Chem. Soc. 1997, 119,
1492 (b) B.L. Hodous, G.C. Fu, J. Am. Chem. Soc. 2002, 124, 1578 (c)
E.C. Lee, B.L. Hodous, E. Bergin, C. Shih ,G.C. Fu, J. Am. Chem. Soc.
2005, 127, 11586
[9]
a) P.M. Dewick, “Medicinal Natural Products, a Biosynthetic Approach,
3rd edition”, 2009, John Wiley & Sons, Inc; b) R.P. Elander, Appl.
Microbiol. Biotechnol. 2003, 61, 385.
[19] a) M. A. Calter, J. Org. Chem. 1996, 61, 8006-8007; b) Taggi, A. E.;
Hafez, A. M.; Wack, H; Young, B.; Ferraris, D; Lectka, T. J. Am. Chem.
Soc. 2002, 124, 6626.
[10]
D.J. Waxman, J.L. Strominger, Ann. Rev. Biochem. 1983, 52, 825.
[11] L. Jiao, X. Liang, J. X. Xu, J. Am. Chem. Soc. 2006, 128, 6060-6069
[12] a) L. Y. Huang, Y. Chen, G. Y. Gao, X. P. Zhang, J. Org. Chem. 2003,
68, 8179; b) A. Penoni, R. Wanke, S. Tollari, E. Gallo, D. Musella, F.
Ragaini, F. Demartin, S. Cenini, Eur. J. Inorg. Chem. 2003, 1452; c) Y.
Chen, X. P. Zhang, J. Org. Chem. 2004, 69, 2431; d) Y. Chen, X. P.
Zhang, J. Org. Chem. 2007, 72, 5931; e) Y. Chen, J. V. Ruppel, X. P.
Zhang, J. Am. Chem. Soc. 2007, 129, 12074; f) S. F. Zhu, J. V. Ruppel,
H. J. Lu, L. Wojtas, X. P. Zhang, J. Am. Chem. Soc. 2008, 130, 5042; g)
S. F. Zhu, J. A. Perman, X. P. Zhang, Angew. Chem. Int. Ed. 2008, 47,
8460; h) S. Fantauzzi, E. Gallo, E. Rose, N. Raoul, A. Caselli, S. Issa, F.
[20] a) J. Eilmes, Polyhedron 1991, 10, 1779; b) S. J. Dzugan, D. H. Busch,
Inorg. Chem. 1990, 29, 2528; c) J. Eilmes, Polyhedron 1985, 4, 943; d)
J. H. Niewahner, K. A. Walters, A. Wagner, Journal of Chemical
Education 2007 84 (3), 477; e) J. R. Chipperfield, S. Woodward, Journal
of Chemical Education 1994 71 (1), 75.
[21] a) J. R. Fulton, V. K. Aggarwal, J. de. Vicente, Eur. J. Org. Chem. 2005,
1479; b) D. Arunprasath, P. Muthupandi, G. Sekar, Org. Lett. 2015, 17,
5448.
This article is protected by copyright. All rights reserved.

10.1002/ejic.201800101
European Journal of Inorganic Chemistry
FULL PAPER
Entry for the Table of Contents
FULL PAPER
Ketenes by metalloradical Catalysis
A. Chirila, K.M. van Vliet, N.D. Paul,
B. de Bruin*
[Co(MeTAA)] Metalloradical Catalytic
Route to Ketenes via Carbonylation of
Carbene Radicals
Carbonylation of carbene radical intermediates generated using the highly active
cobalt(II) catalyst [Co(MeTAA)] provides a convenient one-pot synthetic protocol
towards substituted ketenes. Subsequent trapping with nucleophiles leads to beta-
lactams, amides and esters.
This article is protected by copyright. All rights reserved.