Mechanism of the cobalt-catalyzed carbonylation of ethyl diazoacetate

Article abstract of DOI:10.1016/j.ica.2009.09.026

Carbonyl cobalt complexes serve as catalysts or catalyst precursors for the facile and selective transformation of primary diazoalkanes into the corresponding ketene. The mechanism of this carbonylation reaction has been elucidated in the case of ethyl diazoacetate as model diazoalkane using octacarbonyl dicobalt as the catalyst precursor. Dinuclear cobalt complexes having ethoxycarbonylcarbene ligand(s) in bridging position(s) have been identified as active intermediary of the catalytic cycles and their relevant chemical properties have been explored. Key step of the carbonylation is the formation of the highly reactive ethoxycarbonylketene by intramolecular coupling of a carbonyl ligand with the ethoxycarbonylcarbene ligand. DFT calculations reveal that the ketene formation takes place via a rapid coupling of the carbene ligand with one terminal CO followed by coordination of an external carbon monoxide and by a facile intramolecular rearrangement and ketene elimination. The ethoxycarbonylketene can be in situ trapped by OH, NH, or CH acid compounds or by N-substituted imines. In the presence of ethanol diethyl malonate is the only product of the catalytic carbonylation of ethyl diazoacetate. On the bases of the kinetics of the composing steps of the catalytic cycles, localization of the rate-determining step(s) under various reaction conditions has been made.

Full text of DOI:10.1016/j.ica.2009.09.026

Inorganica Chimica Acta 363 (2010) 2016–2028
Contents lists available at ScienceDirect
Inorganica Chimica Acta
journal homepage: www.elsevier.com/locate/ica
Mechanism of the cobalt-catalyzed carbonylation of ethyl diazoacetate
*
}
Neszta Ungvári, Eszter Fördos, Tamás Kégl, Ferenc Ungváry
Department of Organic Chemistry, University of Pannonia, 8201 Veszprém, Egyetem u. 10, Hungary
a r t i c l e i n f o
a b s t r a c t
Article history:
Carbonyl cobalt complexes serve as catalysts or catalyst precursors for the facile and selective transfor-
mation of primary diazoalkanes into the corresponding ketene. The mechanism of this carbonylation
reaction has been elucidated in the case of ethyl diazoacetate as model diazoalkane using octacarbonyl
dicobalt as the catalyst precursor. Dinuclear cobalt complexes having ethoxycarbonylcarbene ligand(s)
in bridging position(s) have been identified as active intermediary of the catalytic cycles and their rele-
vant chemical properties have been explored. Key step of the carbonylation is the formation of the highly
reactive ethoxycarbonylketene by intramolecular coupling of a carbonyl ligand with the ethoxycarbonyl-
carbene ligand. DFT calculations reveal that the ketene formation takes place via a rapid coupling of the
carbene ligand with one terminal CO followed by coordination of an external carbon monoxide and by a
facile intramolecular rearrangement and ketene elimination. The ethoxycarbonylketene can be in situ
trapped by OH, NH, or CH acid compounds or by N-substituted imines. In the presence of ethanol diethyl
malonate is the only product of the catalytic carbonylation of ethyl diazoacetate. On the bases of the
kinetics of the composing steps of the catalytic cycles, localization of the rate-determining step(s) under
various reaction conditions has been made.
Received 27 February 2009
Received in revised form 28 August 2009
Accepted 11 September 2009
Available online 16 September 2009
Dedicated to Professor László Markó on the
occasion of his 80th birthday.
Keywords:
Cobalt
Carbene and carbonyl ligands
C–C coupling
Ketene
Density functional calculations
Ó 2009 Elsevier B.V. All rights reserved.
1. Introduction
In the absence of external carbon monoxide the coordinated carbon
monoxide ligands of octacarbonyl dicobalt are consumed in the (tri-
Octacarbonyl dicobalt was found to be an effective dediazotation
reagent for (trimethylsilyl)diazomethane [1] or ethyl diazoacetate
[2]. In the dediazotation reaction of (trimethylsilyl)diazomethane
with Co2(CO)8 (trimethylsilyl)ketene is formed selectively either un-
der catalytic (reaction 1) or under stoichiometric (reaction 2) reac-
tion conditions. In the presence of 10 mol% Co2(CO)8 solutions of
(trimethylsilyl)diazomethane in n-heptane are selectively carbony-
lated at 10 °C under an atmosphere of carbon monoxide to (trimeth-
ylsilyl)ketene in up to 80% isolated yield (reaction 1).
methylsilyl)ketene formation (reaction 2).
Ar; 1 bar
8Me3SiCHN2 þ Co2ðCOÞ8
!
8Me3SiCH@C@O þ 8N2 þ 2Co
10 ꢀC; 30 min
ð2Þ
No intermediary cobalt complexes were detected by infrared spec-
troscopy in reactions 1 and 2. Ethyl diazoacetate behaves differently
in the dediazotation reaction. At ambient conditions ethoxycarbon-
ylcarbene-bridged dicobalt carbonyl complexes 1 and 2 were iso-
lated from the reaction mixtures in good yields (reactions 3 and 4).
Under an atmosphere of carbon monoxide the ethoxycarbonylcar-
Me SiCHN þ CO 10 mol% Co
8 Me SiCH@C@O þ N
ð1Þ
2ðCOÞ
ƒƒƒƒƒƒƒƒƒ!
3
2
3
2
10 ꢀC; 1 bar; 4 h
EtO C
O
C
2
O
CH
O
C
C
5-25 °C, 1 bar
- CO, - N2
Co
Co
Co (CO) + EtO CCHN
2
2
8
2
82% isolated yield
ð3Þ
C
O
C
O
C
C
O
O
1
* Corresponding author. Tel.: +36 88 624 156; fax: +36 88 624 469.
E-mail address: ungvary@almos.vein.hu (F. Ungváry).
0020-1693/$ - see front matter Ó 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.ica.2009.09.026

N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
2017
EtO C
CO Et
2
2
CH
CH
O
O
C
C
5-25 °C, 1 bar
- CO, - N2
1 + EtO CCHN
Co
Co
2
2
73% isolated yield
ð4Þ
C
O
C
O
C
C
O
O
2
bene ligand(s) of these complexes couple(s) with CO to give the
highly reactive ethoxycarbonylketene, which can not be isolated
but can be scavenged in situ by OH, NH, or CH acid compounds
[2,3] or by N-substituted imines in a [2 + 2] cycloaddition reaction
[4]. In the presence of ethanol both complexes 1 and 2 convert
quantitatively to Co2(CO)8 and diethyl malonate (reactions 5 and 6).
EtO CCHN
N
+ CO
2
2
2
Co (CO) (CHCO Et)
Co (CO) (CHCO Et)
25 ꢀC; 50 bar
2
7
2
2
6
2
2
1 þ 2CO þ EtOH
Co ðCOÞ þ EtO CCH CO Et
ð5Þ
ð6Þ
ƒƒƒƒƒƒƒ!
2
2
2
2
8
60 min
25 ꢀC; 50 bar
2 þ 4CO þ 2EtOH
Co ðCOÞ þ 2EtO CCH CO Et
ƒƒƒƒƒƒƒ!
2
2
2
2
8
60 min
Experiments at atmospheric pressure of carbon monoxide have
shown that complex 2 converts first rapidly to 1 and diethyl malon-
ate, and in a much slower reaction complex 1 converts further to
octacarbonyl dicobalt and to a second mol of diethyl malonate
(Eq. (7)).
O=C=CHCO Et
2
2CO
EtOH
EtO CCH CO Et
2
2
2
10 ꢀC; 1 bar
10 ꢀC; 1 bar
2 þ 2CO þ EtOHꢁEtO
1
Co ðCOÞ þ EtO CCH CO Et
ƒƒƒƒƒƒƒ! ƒƒƒƒƒƒ!
2
2
2
2
8
2CCH2CO2Et
þ2COþEtOH
Scheme 2.
ð7Þ
In both steps the formation of the highly reactive eth-
oxycarbonylketene was assumed by coupling of the ethoxycarbon-
ylcarbene ligand with one of the coordinated carbon monoxide [2].
The combination of reactions 3–6 led to an effective one-pot proce-
dure for the preparation of diethyl malonate by catalytic carbonyl-
ation of ethyl diazoacetate in the presence of a few mol% of
octacarbonyl dicobalt [2]. Since diethyl malonate may be the prod-
uct from both complex 1 and 2 two different catalytic cycles can be
imagined depicted in Schemes 1 and 2.
Triphenylphosphane- or Ph2PCH2PPh2-substituted derivatives
of complex 1 were found to show similar catalytic activity in ethyl
diazoacetate carbonylation as Co2(CO)8, or complexes 1 or 2 [5–7].
In order to obtain a more detailed mechanism of the catalytic
dediazotation and carbonylation of ethyl diazoacetate we investi-
gated the kinetics of the catalytic reaction and the kinetics of the
reactions in Schemes 1 and 2 at 10 °C in methylene chloride
solutions.
2. Results
EtO CCHN
N + CO
2
2
2
2.1. Experiments using Co2(CO)8 as the catalyst precursor
Co2ðCOÞ8
10 ꢀC; CH2Cl2
EtO CCHN þ EtOH þ CO
N þ EtO CCH CO Et
ð8Þ
ƒƒƒƒƒ!
2
2
2
2
2
2
The effect of the concentrations of Co2(CO)8, ethyl diazoacetate,
ethanol, and carbon monoxide on the rate of diethyl malonate for-
mation in the catalytic carbonylation of ethyl diazoacetate (Eq. (8))
was studied in experiments performed between 1 and 100 bar
pressure of carbon monoxide with various initial concentrations
at 10 °C in CH2Cl2 solutions using 3 h reaction time (Table 1). The
data in Table 1 show that under atmospheric pressure of carbon
monoxide ([CO] = 0.00507–0.00515 mol/dm3) the rate of diethyl
malonate formation is practically linear with the initial concentra-
tions of Co2(CO)8 and that of ethyl diazoacetate, and independent
of the ethanol concentrations (entries 1–15). The carbon monoxide
concentration has a strong positive effect on the rate in the range
between 0.005 and 0.03 mol/dm3 (compare entries 1–15 with that
of entries 16 and 17). On the other hand the carbon monoxide con-
centration has a little negative effect on the rate in the range be-
tween 0.44 and 1.12 mol/dm3 (entries 29–33).
Co (CO)
Co (CO) (CHCO Et)
2
8
2
7
2
O=C=CHCO Et
2
2CO
EtOH
EtO CCH CO Et
2
2
2
Scheme 1.

2018
N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
Table 1
The observed concentration of diethyl malonate [DEM]3h and the observed rate of DEM formation (rDEM = [DEM]3h/3 ꢂ 3600 in the catalytic carbonylation reaction of ethyl
diazoacetate (EDA) at 10 °C under CO in 3 h reaction time applying Co2(CO)8 as the catalyst precursor in the presence of ethanol using various initial concentrations of the
reactants.
102 ꢂ [CO]0
[Co2(CO)8]0
(mol/dm3)
[EDA]0
[Ethanol]0
(mol/dm3)
[DEM]3h
105 ꢂ rDEM
[EDA]0/[CO]0
a
b
Entry
(mol/dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3 s)
1
2
3
4
5
6
7
8
0.512
0.509
0.512
0.512
0.508
0.512
0.507
0.510
0.508
0.515
0.512
0.508
0.507
0.510
0.510
1.02
0.005
0.005
0.005
0.005
0.0125
0.025
0.050
0.050
0.050
0.050
0.100
0.050
0.050
0.050
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.005
0.005
0.005
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.025
0.005
0.005
0.010
0.020
0.050
0.100
0.250
0.250
0.125
0.125
0.250
0.250
0.250
0.250
0.500
0.500
0.500
0.500
0.500
0.500
0.500
0.500
0.500
0.010
0.020
0.100
0.500
0.250
0.125
0.050
0.500
0.500
0.500
0.500
0.05
0.05
0.05
0.05
0.05
0.25
0.25
0.25
0.25
0.125
0.25
0.25
0.50
0.25
0.25
0.50
0.50
0.50
0.50
0.50
0.50
0.50
0.05
0.05
0.05
0.50
0.25
0.25
0.25
0.50
0.50
0.50
0.50
0.02
0.01
0.00135
0.00256
0.00820
0.01651
0.0138
0.0276
0.0299
0.0275
0.0552
0.0529
0.0989
0.0522
0.1150
0.1173
0.0582
0.088
0.0125
0.0237
0.0759
0.1529
0.128
0.256
0.277
0.255
0.511
0.490
0.916
0.511
1.065
1.086
0.493
0.820
1.20
1.24
1.30
1.56
1.70
0.0185
0.0417
0.077
1.92
0.889
0.482
0.204
1.96
1.95
3.93
9.77
19.53
49.21
48.83
24.65
24.51
49.21
48.54
48.83
49.21
98.62
98.04
98.04
49.02
17.86
6.88
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
2.80
7.27
0.130
0.134
0.140
0.169
14.76
22.24
29.73
37.22
37.22
37.22
37.22
37.22
37.22
37.22
44.71
59.69
74.66
112.11
112.11
112.11
3.39
2.25
1.68
0.184
0.0020
0.0045
0.0083
0.207
0.087
0.052
0.022
0.212
0.171
0.134
0.027
0.054
0.269
1.340
0.67
0.336
0.134
1.12
1.58
1.24
1.03
0.062
0.017
0.84
0.672
0.446
0.0446
0.0089
0.111
0.0067
0.0019
0.01
a
Calculated from the partial pressure of CO and the known solubility of CO in CH2Cl2 expressed in mol fraction: xCO (10 °C, 1 bar of P(CO) = 0.000473 converted to
[CO] = 0.007489 mol/dm3 (see Refs. [9,10]).
b
Determined by quantitative infrared spectroscopy.
The composition of the cobalt complexes in the reaction mix-
ture checked by infrared spectroscopy after 3 h of reaction time
shows the presence of three different cobalt complexes: Co2(CO)8,
Co2(CO)7(CHCO2Et) (1), and Co2(CO)6(CHCO2Et)2 (2) which can be
The effect of the concentrations of complex 1, ethyl diazoace-
tate, and ethanol on the rate of diethyl malonate formation in
the catalytic carbonylation of ethyl diazoacetate (Eq. (9)) was stud-
ied in experiments performed under atmospheric pressure of car-
bon monoxide with various initial concentrations at 10 °C in
CH2Cl2 solutions using various reaction times (Table 2). The data
in Table 2 show that the rate of diethyl malonate formation is first
order with respect to complex 1, less than first order in ethyl diazo-
acetate, and independent of the ethanol concentrations. The com-
position of the cobalt complexes in the reaction mixture checked
by infrared spectroscopy shows the presence of mainly complex
1, and less of complex 2, which can be recognized in the mixtures
recognized in the mixtures by their characteristic m(CO) bands at
2022, 2075 and 2080 cmꢁ1, respectively. The composition depends
basically on the applied CO pressure and on the [EDA]/[CO] con-
centration ratio in the experiment. Thus, at 1 bar CO pressure using
[EDA]/[CO] = 1.95 (entry 1) the cobalt composition is in form of
>16% complex 2, 80% complex 1, and <4% Co2(CO)8. If the [EDA]/
[CO] ratio is 10-fold higher, 19.53 (entry 4), the cobalt composition
is in form of <64% complex 2, 16% complex 1, and 0% Co2(CO)8.
Interestingly in this experiment and even more pronounced in
those performed at higher [EDA]/[CO] ratios (entries 5, 6, 10, and
by their characteristic m
(CO) bands at 2075 and 2080 cmꢁ1, respec-
tively. Higher concentration of ethyl diazoacetate influences not
only the rate of reaction 9, but results in a little higher ratio of 2/
1. This latter finding arouses the suspicion that the effect of ethyl
diazoacetate on the rate works by the formation of the much more
active complex 2. Results of experiments using Co2(CO)6(CH-
CO2Et)2 (2) as the catalyst precursor (see Section 2.3) support this
suspicion.
11), in addition of the
m(CO) bands of the known complexes a
shoulder appears at 2060 cmꢁ1. At 50 bar CO pressure using
[EDA]/[CO] = 1.34 (entry 25) the cobalt composition is in form of
16% complex 2, 56% complex 1, and 28% Co2(CO)8. If the [EDA]/
[CO] ratio is 50-fold lower, 0.027 (entry 22), >95% of the cobalt is
in form of Co2(CO)8. The amounts of complex 1 and complex 2 in
this case could not be determined from the infrared spectra.
2.3. Experiments using Co2(CO)6(CHCO2Et)2 (2) as the catalyst
precursor
2.2. Experiments using Co2(CO)7(CHCO2Et) (1) as the catalyst
precursor
2ðCOÞ6ðCHCO2EtÞ2
10 ꢀC; CH2Cl2
EtO2CCHN2 þ EtOH þ CO Co
ð2Þ N2 þ EtO2CCH2CO2Et
2ðCOÞ7ðCHCO2
10 ꢀC; CH2Cl2
EtO CCHN þ EtOH þ CO Co
EtÞ ð1Þ N þ EtO CCH CO Et
ƒƒƒƒƒƒƒƒƒƒƒ!
ƒƒƒƒƒƒƒƒƒƒƒ!
2
2
2
2
2
2
ð10Þ
ð9Þ

N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
2019
Table 2
The observed concentration of diethyl malonate [DEM]t, the observed rate of DEM formation ðr1DEM ¼ ½DEMꢃt =t ꢂ 60Þ, and the calculated observed rate constant
kD1EM ¼ rD1EM=½Co2ðCOÞ7ðCHCO2EtÞꢃ0 ꢂ ½EDAꢃ0 in the Co2(CO)7(CHCO2Et) (1)-catalyzed carbonylation reaction of ethyl diazoacetate (EDA) at 10 °C under atmospheric pressure of
CO in the presence of ethanol using various initial concentrations of the reactants.
[DEM]tb
[1]t
102 ꢂ ½COꢃ0
ðmol=dm3Þ
105 ꢂ r1DEM
105 ꢂ k1DEM
a
Entry
[1]0
[EDA]0
[Ethanol]0
(mol/dm3)
t Time
(Min)
(mol/dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3)
ðmol=dm3 sÞ
ðdm3=mol sÞ
1
2
3
4
5
0.518
0.500
0.521
0.500
0.500
0.005
0.010
0.010
0.010
0.010
0.01
0.01
0.01
0.02
0.04
0.043
0.02
0.04
0.02
0.02
252
180
216
150
180
0.0018
0.0025
0.0029
0.0032
0.0044
0.0037
0.0083
0.0073
0.0069
0.0060
0.0118
0.0232
0.0224
0.0342
0.0407
235
232
224
171
102
a
Calculated from the partial pressure of CO and the known solubility of CO in CH2Cl2 (see Refs. [9,10]).
Determined by quantitative infrared spectroscopy.
b
k1k2½Co2ðCOÞ7ꢃ½EtO2CCHN2ꢃ
The effect of the concentrations of complex 2, ethyl diazoace-
r ¼ k2½CO2ðCOÞ7ꢃ½EtO2CCHN2ꢃ ¼
ð12Þ
ð13Þ
tate, and ethanol on the rate of diethyl malonate formation in
the catalytic carbonylation of ethyl diazoacetate (Eq. (10)) was
studied in experiments performed between 1 and 3.4 bar pressure
of carbon monoxide with various initial concentrations at 10 °C in
CH2Cl2 solutions (Table 3). The rates in Table 3 are one order of
magnitude higher than those in Table 2. The data in Table 3 show
that the rate of diethyl malonate formation is first order with re-
spect to complex 2, and independent of the concentrations of ethyl
diazoacetate, and that of ethanol. During the catalytic reaction 7–
30% of complex 2 converts to complex 1.
k
ꢁ1
½COꢃ þ k2½EtO2CCHN2ꢃ
½Co2ðCOÞ8ꢃ
k
½COꢃ
1
ꢁ1
¼
þ
r
k1k2 ½EtO2CCHN2ꢃ k1
A plot of [Co2(CO)8]/r against [CO]/[EtO2CCHN2] using the experi-
mental initial concentrations and the observed initial rates allowed
the calculation of k1 (10 °C, n-heptane) = (1.22 0.06) ꢂ 10ꢁ3 sꢁ1
and the ratio of kꢁ1/k2 (10 °C, n-heptane) = 1.34 0.07 from the
intercept and the slope, respectively.
In order to have all kinetic information of the steps of Schemes 1
and 2 measured in CH2Cl2 solution we repeated the kinetic inves-
tigation of reaction 11 in CH2Cl2 as well. The results showed the
same kinetic patterns. A plot of [Co2(CO)8]/rin against [CO]/
[EtO2CCHN2] using the experimental initial concentrations and
the observed initial rates (rin) in Table 4 allowed the calculation
of k1 (10 °C) = (0.42 0.03) ꢂ 10ꢁ3 sꢁ1 and the ratio of kꢁ1/k2
(10 °C) = 7.6 0.2 from the intercept and the slope, respectively
(see Fig. 1). Thus, in CH2Cl2 the dissociation of a CO ligand from
Co2(CO)8 is 0.344-times slower than in n-heptane. The 5.7-times
higher ratio of kꢁ1/k2 in CH2Cl2 indicate a strong solvent effect
for the ratio of the competing reactions as well.
2.4. The reaction of Co2(CO)8 with EtO2CCHN2
10 ꢀC
Co ðCOÞ þ EtO CCHN þ
N þ CO þ Co ðCOÞ ðCHCO EtÞ
2
1
ƒƒƒ!
2
2
2
2
2
8
7
CH2Cl2
ð11Þ
The kinetics of the reaction of Co2(CO)8 with ethyl diazoacetate
(Eq. (11)) has been studied earlier in n-heptane solution by mea-
suring the initial rate of gas evolution and by measuring the initial
rate of the decrease of the Co2(CO)8 concentration at various initial
concentrations of Co2(CO)8, EtO2CCHN2 and CO at 10 °C [8]. The ob-
served kinetics was found to be in accord with a reversible disso-
2.5. The reaction of Co2(CO)7(CHCO2Et) (1) with CO and EtOH
ciation of carbon monoxide from Co2(CO)8, followed by
reaction with ethyl diazoacetate (Scheme 3).
Assuming steady-state concentration for the Co2(CO)7 interme-
diate, the rate could be described by Eq. (12) or after rearrange-
ment by Eq. (13).
a
10 ꢀC
Co ðCOÞ ðCHCO EtÞþEtOH þ 2CO
EtO CCH CO Et þ Co ðCOÞ
ƒƒƒ!
2
2
2
2
2
2
7
8
CH2Cl2
1
ð14Þ
Table 3
The observed concentration of diethyl malonate [DEM]t, the observed rate of DEM formation ðr2DEM ¼ ½DEMꢃt =t ꢂ 60Þ, and the calculated rate constant
kD2EM ¼ rD2EM=½Co2ðCOÞ6ðCHCO2EtÞ2ꢃ0 ꢂ ½COꢃ0 in the Co2(CO)6(CHCO2Et)2 (2)-catalyzed carbonylation reaction of ethyl diazoacetate (EDA) at 10 °C under CO in the presence of
ethanol using various initial concentrations of the reactants.
102 ꢂ [2]0
[EDA]0
[Ethanol]0
(mol/dm3)
t Time
(Min)
102 ꢂ [1]t
102 ꢂ ½COꢃ0
ðmol=dm3Þ
102 ꢂ ½DEMꢃt
ðmol=dm3Þ
105 ꢂ r2DEM
105 ꢂ k2DEM
a
b
Entry
(mol/dm3)
(mol/dm3)
(mol/dm3)
ðmol=dm3 sÞ
ðdm3=mol sÞ
1
2
3
4
5
6
7
8
0.499
0.503
0.504
0.504
0.504
0.510
0.510
0.510
0.510
0.510
1.478
2.370
2.355
2.335
0.50
0.50
0.50
0.50
0.50
0.50
0.50
1.00
1.00
1.00
0.50
0.50
0.50
0.50
0.010
0.025
0.050
0.050
0.050
0.100
0.100
0.010
0.010
0.010
0.050
0.050
0.100
0.500
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.05
0.40
0.02
0.02
0.05
0.25
60
60
60
90
120
90
125
30
60
60
15
15
30
75
0.164
0.166
0.152
0.230
0.305
0.234
0.322
0.152
0.299
0.306
0.087
0.106
0.204
0.469
0.128
0.113
0.089
0.101
0.103
0.051
0.049
0.167
0.250
0.265
0.082
0.104
0.151
0.025
0.0456
0.0460
0.0422
0.0426
0.0423
0.0433
0.0447
0.0843
0.0831
0.0850
0.0966
0.1181
0.1130
0.1042
1830
1829
1675
1690
1679
1697
1752
1651
1629
1665
1308
997
9
10
11
12
13
14
960
893
a
Calculated from the partial pressure of CO and the known solubility of CO in CH2Cl2 (see Refs. [9,10]).
Determined by quantitative infrared spectroscopy.
b

2020
N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
k1
(s)
Co2(CO)8
Co2(CO)
7 + CO
k-1
k2
+
Co2(CO)7 + EtO2CCHN2
+
Co2(CO)7(CHCO2Et)
N2
+
10
Scheme 3.
r
+
+
In the presence of ethanol the ethoxycarbonylcarbene-bridged
complex Co2(CO)7(CHCO2Et) (1) reacts with CO according to Eq.
(14) to give diethyl malonate and Co2(CO)8. The rate of the reaction
was followed by measuring the increase of the diethyl malonate
concentration versus reaction time using quantitative infrared
spectroscopy. The data in Table 5 show that the rate of diethyl
malonate formation is first order with respect to both the complex
and the CO concentration, and zero order with respect to the
ethanol concentration. Accordingly the rate constant of the reac-
tion was calculated as k1DEM ð10 ꢀCÞ ¼ rD1EM=½1ꢃ½COꢃ ¼ ð40 ꢄ 5Þꢂ
10ꢁ5 dm3=mol s. The observed first order in complex 1 and carbon
monoxide concentration suggests that the rate-determining step
involves 1 mol of complex 1 and 1 mol of external carbon monox-
ide. The rate-determining reaction might involve the synchronous
coordination of an external carbon monoxide and the coupling of
the ethoxycarbonylcarbene ligand with one of the coordinated
CO ligands. The subsequent formation of Co2(CO)8 and eth-
oxycarbonylketene involves the fast coordination of a second mol
of external carbon monoxide. The independency of the rate from
the ethanol concentration supports the assumption that the final
diethyl malonate formation is also a fast reaction between ethanol
and the intermediate ethoxycarbonylketene. It is remarkable that
under atmospheric pressure reaction 14 takes place twice as fast
in the absence of carbon monoxide (entry 4 in Table 5). Source of
carbon monoxide in this case may be the partial decomposition
of the carbonyl cobalt complexes in the reaction mixture.
+
k-1
k2
5
_
+
(10 °C)
-3
= 7.6 0.2
+
+
+
+
k1 (10 °C) = (0.42 0.03)x10 s-1
+_
+
1.5
1.0
2.0
0.5
2.5
[CO]0/[EtO2CCHN2]0
Fig. 1. Plot of the observed initial gas evolution rates (rgas) in the reaction:
Co2(CO)8 + EtO2CCHN2 ? Co2(CO)7(CHCO2Et) + N2 + CO in CH2Cl2 solution at 10 °C
at various initial concentrations according to Eq. (13).
reaction by measuring the initial rate of gas evolution ðr1gasÞ and by
measuring the initial rate of the decrease of the Co2(CO)7(CHCO2Et)
(1) concentration ðr1ꢁ½1ꢃÞ at various initial concentrations of complex
1, EtO2CCHN2 and CO at 10 °C in CH2Cl2 solution. The decrease of
the Co2(CO)7(CHCO2Et) concentration was followed by the de-
crease of intensity of the well separated bridging m(C@O) band of
complex 1 at 1853 cmꢁ1. In experiments started under an atmo-
sphere of argon we found a linear dependence of the observed
rates on the initial concentrations of complex 1 (see Fig. 2).
Raising the initial concentrations of EtO2CCHN2 has a positive
effect on the observed rates. On the other hand the presence of
CO in the gas atmosphere has a negative effect on the rate of the
dediazotation reaction. This kinetic behavior is in accordance with
a reversible dissociation of CO from complex 1, followed by a reac-
tion with EtO2CCHN2 (Scheme 4).
2.6. The reaction of Co2(CO)7(CHCO2Et) (1) with EtO2CCHN2
Assuming steady-state concentration for the Co2(CO)6(CH-
CO2Et) intermediate, the rate can be described by Eq. (16) or after
rearrangement by Eq. (17).
10 ꢀC
Co2ðCOÞ ðCHCO2EtÞ þEtO2CCHN2
N þ CO
2
ƒƒƒ!
7
CH2Cl2
1
r1ꢁ½1ꢃ ¼ k1½Co2ðCOÞ ðCHCO2EtÞꢃ½EtO2CCHN2ꢃ
þ Co2ðCOÞ6ðCHCO2EtÞ2
ð15Þ
2
6
2
k1k1½Co2ðCOÞ ðCHCO2EtÞꢃ½EtO2CCHN2ꢃ
1
2
7
Complex 1 also is an effective dediazotation reagent for ethyl
diazoacetate. Products of the dediazotation reaction are N2, CO,
and complex 2 according to reaction 15. It takes only a few minutes
to achieve a complete conversion if the CO formed in the reaction is
continuously swept away by argon. We studied the kinetics of the
¼
ð16Þ
ð17Þ
k1ꢁ1½COꢃ þ k1½EtO2CCHN2ꢃ
2
½Co2ðCOÞ7ðCHCO2EtÞꢃ
k1ꢁ1
½COꢃ
1
k11
¼
þ
r1ꢁ½1ꢃ
k11k12
½EtO2CCHN2ꢃ
Table 4
The effect of concentrations of Co2(CO)8 and ethyl diazoacetate on the initial rate (rin) of the reaction: Co2(CO)8 + N2CHCO2Et ? Co2(CO)7(CHCO2Et) + CO + N2 under atmospheric
pressure (740–750 mmHg) of carbon monoxide in CH2Cl2 solution at 10 °C.
[Co2(CO)8]0 (mol/dm3)
[EDA]0 (mol/dm3)
[CO]0 (mol/dm3)
105 ꢂ rin (mol/dm3 s)
[Co2(CO)8]0/rin (s)
[CO]0/[N2CHCO2Et]0
a
0.01
0.01
0.01
0.02
0.04
0.01
0.01
0.01
0.01
0.01
0.01
0.01
0.04
0.02
0.01
0.01
0.00506
0.00506
0.00506
0.00507
0.00510
0.00506
0.00508
0.00509
0.00508
0.00498
0.00510
0.00510
0.634
0.418
0.295
0.579
1.275
0.174
0.1225
0.1226
0.0952
0.0809
0.0691
0.0623
1577
2392
3396
3454
3137
5744
8165
8156
10504
12361
14461
16051
0.126
0.253
0.506
0.507
0.510
1.013
1.527
1.528
2.034
2.488
3.075
3.496
0.01
0.005
0.00333
0.00333
0.0025
0.0020
0.00166
0.00125
a
Calculated from the partial pressure of CO and the known solubility of CO in CH2Cl2 (see Refs. [9,10]).

N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
2021
150
100
50
1
75
5
r
10 x
gas
+
[1]0
(mol/dm3 s)
-
10 3
x
(s)
r1
gas
+
+
3
6
50
25
+
k1
k1
-1
+
_
+
(10 °C)
=
410 140
r
2
4
2
r
+
+
2
1
+
+
-3
5)x10 s-1
k11 (10 °C) = (15
_
+
+
+
+
+
+
+
0.02
0.01
1
2
4
5
3
[1]0 (mol/dm3)
[CO]0/[EtO2CCHN2]0
Fig. 2. Plot of the observed initial gas evolution rates ðr1gasÞ and initial disappearance
rates of complex 1 ðrꢁ1 ½1ꢃÞ in the reaction: Co2(CO)7(CHCO2Et) + EtO2CCHN2 ? Co2-
(CO)6(CHCO2Et)2 + N2 + CO in CH2Cl2 solution started under an atmosphere of argon
at 10 °C at [EtO2CCHN2]0 = 0.010 mol/dm3 at various initial concentrations of
complex 1.
Fig. 3. Plot of the observed initial gas evolution rates ðr1gasÞ and initial disappearance
rates of complex 1 ðrꢁ1 ½1ꢃÞ in the reaction: Co2(CO)7(CHCO2Et) + EtO2CCHN2 ? Co2-
(CO)6(CHCO2Et)2 + N2 + CO in CH2Cl2 solution at 10 °C at various initial concentra-
tions according to Eq. (17).
Accordingly the rate constant of the reaction was calculated as
k2DEM ð10 ꢀCÞ ¼ r2DEM=½2ꢃ½COꢃ ¼ ð2640 ꢄ 220Þ ꢂ 10ꢁ5 dm3=mol s, which
is 67.5-times faster than the analogous reaction with complex 1.
The observed first order in complex 2 and carbon monoxide con-
centration suggests that the rate-determining step involves one
mol of complex 2 and 1 mol of carbon monoxide, which might be
the coordination of an external carbon monoxide synchronously
to the coupling of the ethoxycarbonylcarbene ligand with one of
the coordinated CO ligands. The subsequent fast formation of com-
plex 1 and free ethoxycarbonylketene involves the coordination of
a second mol of carbon monoxide. The independency of the rate
from the ethanol concentration supports the assumption that the
final diethyl malonate formation is also a fast reaction between
ethanol and the intermediate ethoxycarbonylketene. These
assumptions were verified by the fact that reaction of complex 2
with 13CO in the presence of excess ethanol (reaction 19) results
in diethyl malonate with natural isotopic distribution of 13C [11].
k1
1
Co2(CO)7(CHCO2Et)
Co2(CO)6(CHCO2Et) + CO
k1
-1
k1
2
Co2(CO)6(CHCO2Et) + EtO2CCHN2
Co2(CO)6(CHCO2Et)2 + N2
Scheme 4.
A plot of ½1ꢃ=r1ꢁ½1ꢃ against [CO]/[EtO2CCHN2] using the experi-
mental initial concentrations and the observed initial rates allowed
the calculation of k11 ð10 ꢀCÞ ¼ ð15 ꢄ 5Þ ꢂ 10ꢁ3 sꢁ1 and the ratio of
k1ꢁ1=k12 ð10 ꢀCÞ ¼ 410 ꢄ 140 from the intercept and the slope,
respectively (see Fig. 3). A comparison of k11 with k1 shows that
the formation of the coordinatively unsaturated intermediate
Co2(CO)6(CHCO2Et) from complex 1 is more than 35-times faster
than that of Co2(CO)7 from Co2(CO)8. Despite of this difference
the overall rate of the reaction of ethyl diazoacetate with Co2(CO)8
(k1k2/kꢁ1 = 5.53 ꢂ 10ꢁ5 sꢁ1) is only 1.52-times faster than that with
complex 1 ðk11k21=k1ꢁ1 ¼ 3:65 ꢂ 10ꢁ5 sꢁ1Þ.
EtOH
Co2ðCOÞ ðCHCO2EtÞ þx13CO
Co ðCOÞ7ꢁxð13COÞ ðCHCO2EtÞ
ƒƒƒƒƒ!
2
6
2
x
10 ꢀC; 1 bar
2
þ EtO2CCH2CO2Et þ ðx ꢁ 2ÞCO
ð19Þ
If a solution of complex 2 and ethanol was kept under the same con-
ditions but under an atmosphere of argon instead of carbon monox-
ide, 20-times slower formation of diethyl malonate could be
observed parallel to the decrease of the concentration of complex
2 (see entries 4 and 5, respectively). The much slower formation
of diethyl malonate suggests that the coupling of the ethoxycarbon-
ylcarbene ligand with one of the coordinated CO ligands might oc-
cur in a small extent even in the absence of the beneficial external
carbon monoxide, and the ethoxycarbonylketene product is suc-
cessfully trapped by the ethanol present. Source of carbon monox-
ide in this case may be the partial decomposition of the carbonyl
cobalt complexes in the reaction mixture.
2.7. The reaction of Co2(CO)6(CHCO2Et)2 (2) with CO and EtOH
10 ꢀC
Co ðCOÞ ðCHCO EtÞ þEtOH þ 2CO
EtO CCH CO Et
2 2 2
ƒƒ!
2
2
6
2
CH2Cl2
2
þ Co2ðCOÞ7ðCHCO2EtÞ
ð18Þ
1
In the presence of ethanol the ethoxycarbonylcarbene-bridged
complex Co2(CO)6(CHCO2Et)2 (2) reacts with CO according to Eq.
(18) to give diethyl malonate and Co2(CO)7(CHCO2Et) (1) in the
first step. The rate of the reaction was followed by measuring the
increase of the diethyl malonate concentration versus reaction
time using quantitative infrared spectroscopy. The data in Table
6 show that the rate of diethyl malonate formation is first order
with respect to both the complex and the CO concentration, and
zero order with respect to the ethanol concentration. In a few cases
the rates of the reaction were calculated based on the amount of
complex 1 formed in the reaction (see for example entry 2), and
based on the amount of CO absorbed in the reaction (see for
example entry 3) in good agreement with the rates calculated on
the basis of the amount of diethyl malonate formed in the reaction.
The experimentally determined rate constants of the assumed
steps in the cobalt–carbonyl catalyzed carbonylation of ethyl
diazoacetate at 10 °C in CH2Cl2 solution have been compiled in
Scheme 5.
2.8. Computational studies on the ketene formation step
In order to get deeper insights into the mechanism of the ketene
formation from the bridged carbene complex 1 the ketene forma-

2022
N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
Table 5
The effect of initial concentrations of Co2(CO)7(CHCO2Et) (1), CO, and ethanol on the initial rate of diethyl malonate formation ðr1 Þ in the reaction: Co2(CO)7(CHCO2Et) + 2-
ꢀ
ꢁ
DEM
CO + EtOH ? Co2 (CO)8 + EtO2CCH2CO2Et at 10 °C in CH2Cl2 solution using various initial concentrations of the reactants, and the calculated rate constant k1DEM ¼ rD1EM=½1ꢃ0½COꢃ
.
102 ꢂ [1]0
102 ꢂ [CO]a
102 ꢂ [EtOH]0
t Time
(Min)
102 ꢂ [DEM]tb
105 ꢂ r1DEM
105 ꢂ k1DEM
Entry
(mol/dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3)
ðmol=dm3 s)
ðdm3=mol sÞ
1
2
3
4
5
6
7
8
0.50
1.00
1.00
1.00
1.00
2.00
2.00
4.90
1.00
1.00
1.00
0.514
0.513
0.511
0.000c
3.53
3.53
3.53
3.53
7.27
4.27
2.00
4.00
2.00
4.00
4.00
8.00
4.00
4.00
4.00
4.00
1618
475
240
300d,b
90
45
45
22
45
0.0860
0.0653
0.0326
0.000845
0.0772
0.0801
0.0702
0.0983
0.0828
0.0680
0.0708
0.00089
0.0023
0.0022
0.00469
0.0143
0.0297
0.0260
0.0744
0.0307
0.0515
0.1311
34.6
44.7
44.3
40.5
42.1
36.8
43.0
42.2
34.9
34.7
9
10
11
14.76
37.23
22
9
a
b
c
Calculated from the partial pressure of CO and the known solubility of CO in CH2Cl2 (see Refs. [9,10]).
Determined by quantitative infrared spectroscopy.
Reaction started under argon atmosphere.
d
After 300 min. reaction time the concentration of complex 1 is 0.00837 mol/dm3.
Table 6
The effect of initial concentrations of Co2(CO)6(CHCO2Et)2 (2), CO, and ethanol on the initial rate of diethyl malonate (DEM) formation ðrD2EM
Þ
in the reaction:
Co2(CO)6(CHCO2Et)2 + 2CO + EtOH ? Co2(CO)7(CHCO2Et) + EtO2CCH2CO2Et at 10 °C in CH2Cl2 solution using various initial concentrations of the reactants, and the calculated
rate constant ðk2DEM ¼ rD2EM=½2ꢃ0½COꢃÞ.
105 ꢂ r2DEM
105 ꢂ k2DEM
Entry
102 ꢂ [CO]a
102 ꢂ [2]0
102 ꢂ [EtOH]0
t Reaction
time (Min)
102 ꢂ [DEM]t
(mol/dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3)
ðmol=dm3 sÞ
ðdm3=mol sÞ
1
2
3
4
5
6
7
8
0.513
0.56
0.58
2.00
2.00
30
0.126
0.053
0.070
2437
2776b
2998c
Argond
240
0.0037
0.0102e
0.193
0.317
0.321
0.617
0.317
0.308
0.700
0.427
0.858
0.933
1.485
2.222
2.305
2.247
2.234
2.202
2.143
3.521
3.521
7.265
0.50
0.50
0.50
1.00
0.50
0.50
1.265
0.50
1.00
0.50
2.00
1.00
2.00
2.00
4.00
8.00
4.00
2.00
2.00
2.00
10
4
4
3
4
4
2
5
4
3
0.116
0.076
0.077
0.111
0.076
0.074
0.084
0.128
0.206
0.168
2604
2855
2785
2746
2838
2797
2556
2425
2438
2569
9
10
11
12
13
14
15
a
b
c
Calculated from the partial pressure of CO and the known solubility of CO in CH2Cl2 (see Refs. [9,10]).
Rate calculated based on the amount of complex 1 formed in the reaction.
Rate calculated based on the amount of CO absorption in the reaction.
Reaction started under an atmosphere of argon.
d
e
Rate calculated based on the amount of complex 2 consumed in the reaction.
k1 = (42 + 3) x 10-5 s-1
k11 = (1500 + 500) x 10-5 s-1
-
-
(Co2(CO)8 + N2CHCO2Et reaction)
(1 + N2CHCO2Et reaction)
k1-1/k12 = 410 + 140
k-1/k2 = 7.6 + 0.2
-
-
EtO CCHN
N
2
EtO CCHN
N
2
2
2
2
2
k21
k2
k1k2/k-1 = 5.53x10-5 s-1
k11k12/k1-1 = 3.65x10-5 s-1
k11
k1
-CO
-CO
+CO
k1-1
Co (CO) (CHCO Et)
Co (CO) (CHCO Et)
2 6 2 2
Co (CO)
Co (CO)
Co (CO) (CHCO Et)
2 7 2
2
6
2
2
8
2
7
+CO
k-1
1
2
A
B
kD1EM
kD2EM
O=C=CHCO Et
O=C=CHCO Et
CO
CO
2
2
k1DEM = (40 + 5) x 10-5 dm3/mol s
kD2EM = (2640 + 220) x 10-5 dm3/mol s
-
fast
+ EtOH
+ EtOH
fast
(1+CO+EtOH reaction)
(2+CO+EtOH reaction)
EtO CCH CO Et
EtO CCH CO Et
2
2
2
2
2
2
Scheme 5.

N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
2023
tion and the ketene elimination elementary steps were studied by
means of density functional calculations using the BP86 functional
which gives usually a reasonable prediction for carbonyl vibra-
tional frequencies [12]. More detailed computational investiga-
tions on the cobalt-catalyzed ketene formation reaction
comprising all elementary steps are underway.
that of the syn-periplanar isomer. The energy difference of both
conformers was recomputed at the BP86 level of theory and was
found to be 2.5 kcal/mol. The relative stability of anti-periplanar
and syn-periplanar conformations of the species occurring in this
computational study has also been checked and the former one
was proved to be slightly more stable in all cases.
Isomers of 1 were described thoroughly in one of our previous
papers [13]. It was reported that the arrangement of the H–C–
C@O is anti-periplanar as its energy is by 2.1 kcal/mol lower than
The structure of 1 computed at the BP86 level reveals symmetry
very close to Cs. It was shown previously [13] that this species tend
to release one CO with a relatively small free energy of dissocia-
Fig. 4. Computed structure of 1 and its coordinative unsaturated derivative 3 formed by dissociation of one CO. Bond distances are given in Å.
Fig. 5. Computed structures of the transition state for the intramolecular ketene formation (4TS), the coordinative unsaturated ketene complex (5), the transition state for the
coordination of an additional CO (6TS) and the coordinative saturated ketene complex (7). Bond distances are given in Å.

2024
N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
tion. The predicted
D
G for the reaction 1 ? 3 + CO is 9.3 kcal/mol,
ceeds in a fast reaction as the free energy of activation is only
8.9 kcal/mol whereas the reaction free energy is 8.5 kcal/mol, thus
endothermic. The transition state 4TS related to the formation of
slightly higher than that obtained at the B3LYP level (6.9 kcal/mol).
The geometry parameters computed with the two functionals are
in close agreement except the distance between the carbonyl oxy-
gen of the ethoxycarbonyl ligand and the interacting cobalt, which
is somewhat higher at the BP86 level suggesting a slightly weaker
computed interaction. The structure of complexes 1 and 3 are de-
the ketene complex has a single imaginary frequency of 185i cmꢁ1
.
The ketene complex 5 is an interesting intermediate example be-
tween the (l1 2) and (l2 2) coordination modes as the CH moi-
–g
–g
ety is bound stronger to Co1, however the strength of the
interaction to Co2 is not negligible, suggested by the 2.207 Å C–
Co bond length (see Fig. 5). The coordinative unsaturated complex
5 takes up one external carbon monoxide resulting in the saturated
species 7 via transition state 6TS, which has a characteristic imag-
inary frequency of 122i cmꢁ1. Unlike 5, complex 7 can be charac-
picted in Fig. 4. Their characteristic m(CO) bands are computed to
be at 2054 and 2041 cmꢁ1, respectively.
The CH moiety belonging to the bridging carbene group is found
to couple with one of the axial terminal carbonyl ligands resulting
in ketene complex 5, which has a computed characteristic
m(CO)
band at 2045 cmꢁ1. The intramolecular carbene–CO coupling pro-
terized as a true (l2 2) ketene complex with a computed
–g
Fig. 6. Computed structures of the transition state for the interconversion of the coordinative saturated ketene complex (8TS), the coordinative saturated (l1 2) ketene
–g
complex (9), the transition state for the dissociation of ethoxycarbonylketene (10TS) and the reformed ketene complex (11) bounding ketene via its carbonyl oxygen. Bond
distances are given in Å.
Fig. 7. Proposed free energy profile of the ketene formation and dissociation pathway.

N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
2025
Table 7
In order to shed some light on the coordination of eth-
oxycarbonylketene ligand in the more stable saturated com-
plexes 7, 9, and 11, the nature of interaction is studied within
the framework of Charge Decomposition Analysis (CDA), devel-
oped by Dapprich and Frenking [15]. For all calculations
O@C@CHCO2Et was considered as the donor fragment, whereas
Co2(CO)7 was considered as the acceptor fragment. The results
are given in Table 7.
Charge Decomposition Analysis (CDA) of complexes 7, 9, and 11. Interaction energy
values are given in kcal/mol.
7
9
11
Donation (d)
Back-donation (b)
Repulsive polarization (r)
Residual term (
Interaction energy
0.515e
0.099e
ꢁ0.242e
0.091e
ꢁ81.64
0.590e
0.278e
ꢁ0.180e
ꢁ0.037e
ꢁ38.64
0.645e
0.200e
ꢁ0.159e
ꢁ0.020e
ꢁ39.14
D)
The small residual term for complexes 9, and 11 indicate that
the ligand is bound to the transition metal in a classical donor–
characteristic m
(CO) band at 2052 cmꢁ1. The coordination of CO is
acceptor manner. The slightly higher
D value, however, for
complex 7, as well as the low amount of back-donation and the
unusually high interaction energy suggest that 7 has some of cova-
lent metallacycle character as well. On the other hand all the CDA
descriptors of complex 9 are remarkably similar to those found for
Pt–olefin complexes [16].
almost isoenergetic as the free energy of the reaction is 0.9 kcal/
mol.
It is predicted that the coordination mode of 7 is not the most
stable one as it is prone to form 9 in a very facile rearrangement
via a free energy barrier of 2.6 kcal/mol. The transition state related
to the interconversion is designated as 8TS, having a single imagi-
nary frequency of 76i cmꢁ1, is featured by a simultaneous migra-
tion of the ketene ligand from Co1 to Co2 and the decrease of the
Co1–Co2–C angle of one of the terminal CO ligand leading to a ke-
tene complex possessing to bridging carbonyl ligands (see Fig. 6).
The donation and back-donation values of each canonical orbi-
tal of a given complex may also provide valuable qualitative infor-
mation about the donor–acceptor character. Interestingly, in case
of complex 7 HOMO is mostly responsible for both donation and
the slight back donation (see Fig. 8a). The electron donation in
the (l1
–g
2) complex 9 is mainly described by the HOMOꢁ1 orbital
The (l1 2) complex 9 is by 2.5 kcal/mol more stable in term of
–g
free energy than 7 with characteristic m
(CO) bands at 2031 cmꢁ1
(Fig. 8b) whereas the back-donation is described by HOMOꢁ2 and
the results suggest that both cobalt atoms are involved in it
(Fig. 8c). The unusually high donation in complex 11 may be ratio-
nalized by the overlap of a filled orbital of the ethoxycarbonylke-
tene moiety with mainly the unfilled dz2 orbital of Co2 (Fig. 8d).
For the back-donation the HOMOꢁ1 orbital is mostly responsible
which just like the HOMOꢁ2 of 9 has a contribution on both cobalt
atoms (Fig. 8e).
and 1863 cmꢁ1 related to the harmonic vibrations of terminal
and bridging carbonyls, respectively.
The coordinated ketene ligand dissociates from 9 via a small
free energy barrier of 6.6 kcal/mol suggesting that ketene com-
plexes formed by the coupling of the bridging carbene and one
terminal CO ligands cannot be observed experimentally by IR
spectroscopy; they provide readily free ketene instead. The ke-
tene dissociation step is exothermic in term of free energy by
ꢁ5.5 kcal/mol. The transition state 10TS has a single imaginary
frequency of 73i cmꢁ1. Intrinsic reaction coordinate (IRC) calcula-
tion revealed that 10TS indeed connects 9 and the dissociated
ketene along with the dibridged Co2(CO)7 complex. Similar
homoleptic dirhodium heptacarbonyl complex is described re-
cently [14], but to the best of our knowledge dibridged Co2(CO)7
has not been described in the literature. More theoretical work is
underway related to the structures and interconversions of this
species.
3. Discussion
In the catalytic carbonylation experiments under atmospheric
pressure of carbon monoxide using Co2(CO)8 as the catalyst precur-
sor the prevailing complex in the reaction mixture is complex 2 if
the applied [EDA]/[CO] ratio in the experiment was >10 (see for
example entry 4 in Table 1 and the joining text in the results sec-
tion). It seems to be obvious that under these reaction conditions
the formation of the diethyl malonate product occurs mainly
according to the catalytic cycle: B. In order to check this assump-
tion we compared the observed rates of DEM formation with those
calculated for the various steps using the experimentally deter-
mined rate constants compiled in Scheme 5. From the experiments
performed under atmospheric CO pressure we chose entry 2 of Ta-
ble 1, where we have measured both the rate of DEM formation
[rDEM (obs)] and the composition of the cobalt complexes in the
reaction mixture after 1, 2, and 3 h reaction time. The results of
the comparison can be seen in Table 8. First we calculated the rate
of complex 1 formation from the precursor Co2(CO)8 and ethyl
diazoacetate (EDA) as r1 (calcd) = 0.0716 ꢂ 10ꢁ5 mol/dm3 s (entry
The entire proposed pathway of the ketene formation and elim-
ination steps are summarized in Fig. 7.
On the Co2(CO)7(O@C@CHCO2Et) potential energy hypersurface
yet another ketene adduct 11 has been found with an unusual
coordination mode. In this complex, which is only by 0.7 kcal/
mol higher in free energy than 9, the ketene ligand is mainly bound
by the carbonyl oxygen revealing a Co–O bond distance of 2.039 Å.
On the other hand the coordination of the ketene carbon is also
noticeable. Although the Co–C bond distance is relatively large,
the C–C–O bond angle is 146°, significantly deviating from 180°,
indicating a week coordination of carbon towards Co2.
Fig. 8. Canonical molecular orbitals attributed to electron donation and back-donation in the coordinative saturated ketene complexes.

2026
N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
Table 8
The observed concentrations of diethyl malonate (DEM), Co2(CO)8, Co2(CO)7(CHCO2Et) (1), and Co2(CO)6(CHCO2Et)2 (2), the observed rate of DEM formation (rDEM = [DEM]t/
t ꢂ 3600), the calculated rates of DEM formation rD1EM ¼ kD1EM=½1ꢃt ½COꢃ0 and rD2EM ¼ kD2EM=½2ꢃt ½COꢃ0, the calculated rates of complex 1 formation r1 = [Co2(CO)8]t/{(k
[CO]0/
ꢁ1
k1k2[EDA]t) + 1/k1} and the calculated rates of complex 2 formation r2 ¼ ½1ꢃt =fðkꢁ1 1½COꢃ0=k11k12½EDAꢃt Þ þ 1=k11g in the catalytic carbonylation reaction of ethyl diazoacetate
[EDA]0 = 0.02 mol/dm3 at 10 °C under atmospheric pressure of CO, [CO]0 = 0.00509 mol/dm3, applying [Co2(CO)8]0 = 0.005 mol/dm3 as the catalyst precursor in the presence of
[Ethanol]0 = 0.05 mol/dm3 using different reaction times (t). The experimentally determined rate constants k1DEM; k2DEM; kꢁ1=k1k2; k1; kꢁ1 1=k11k21, and k11 were taken from Scheme 5.
105 ꢂ r1DEM
ðcalcdÞ
105 ꢂ r2DEM
ðcalcdÞ
Entry Reaction [EDA]t/
[DEM]t
(mol/
dm3)
[Co2(CO)8]t [1]t (mol/
[2]t(mol/
105 ꢂ rDEM
(obs)
105 ꢂ r1
105 ꢂ r2
time t
[CO]0
(mol/dm3) dm3)
dm3)
(calcd)
(calcd)
(h)
(mol/dm3 s)
(mol/dm3 s)
(mol/dm3 s)
ðmol=dm3 sÞ
ðmol=dm3 sÞ
1
2
3
4
0
1
2
3
3.93
2.55
2.28
2.11
0
0.005
0.0003
0
0
0
0
0.0716
0.00094
0.00178
0.00256
0.0033
0.0034
0.0033
0.0014
0.0016
0.0017
0.0261
0.0247
0.0237
0.0007
0.0007
0.0007
0.0188
0.0215
0.0228
0.0306
0.0282
0.0253
Table 9
The observed and calculated concentrations of diethyl malonate (DEM), Co2(CO)7(CHCO2Et) (1), and Co2(CO)6(CHCO2Et)2 (2) in the carbonyl cobalt-catalyzed carbonylation
reaction of ethyl diazoacetate (EDA) in CH2Cl2 solution in the presence of ethanol at 10 °C and 3 h reaction time using different initial concentrations. For the calculations the
experimentally determined rate constants k1DEM; kD2EM; kꢁ1=k1k2; k1; kꢁ1 1=k11k21, and k11 were taken from Scheme 5.
Entrya [EDA]0
[CO]0
[EDA]0/ [Co2(CO)8]0
[CO]0
[DEM]3h
[1]3h obsb
[2]3h obsc
[1]3h calcd
[2]3h calcd
1
2
d
e
½DEMꢃ3h
½DEMꢃ3h
(mol/dm3) (mol/
dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3)
(mol/dm3)
ðmol=dm3Þ
ðmol=dm3Þ
25
26
32
33
34
0.500
0.250
0.500
0.050
0.010
0.3722
0.3722
1.1211
1.1211
1.1211
1.343
0.672
0.446
0.0446
0.0089
0.025
0.025
0.025
0.005
0.005
0.207
0.087
0.111
0.0067
0.0019
0.0140
0.0040
0.0137
0.0074
0.0153
0.000321
0.0000644
0.2386
0.0649
0.0883
0.000186
0.0000075
0.0121
0.0037
a
b
c
The numbers refer to entries in Table 1.
Determined by quantitative infrared spectroscopy at 1853 cmꢁ1
.
Determined by quantitative infrared spectroscopy at 2080 cmꢁ1
.
1
d
e
Calculated as ½DEMꢃ3h ¼ ðr1DEMÞ1:5 h ꢂ 3 ꢂ 3600.
2
Calculated as ½DEMꢃ3h ¼ ðr2DEMÞ1:5 h ꢂ 3 ꢂ 3600.
1 in Table 8), which is 2.88-times faster than the average rate of
DEM formation. Therefore this step could not be the bottle-neck
of the catalytic reaction. Then we calculated the rate of complex
2 formation from the precursor complex 1 and ethyl diazoacetate
as r2 (calcd) = 0.0306 ꢂ 10ꢁ5 (entry 2 in Table 8), 0.0282 ꢂ 10ꢁ5
(entry 3 in Table 8), and 0.0253 ꢂ 10ꢁ5 mol/dm3 s (entry 4 in Table
8) at 1, 2, and 3 h reaction time, respectively. Here again these rates
are higher than the observed rate of DEM formation, hence the
complex 2 formation is not rate determining either. The calculated
rates of DEM formation ½r1DEM ðcalcdÞꢃ are much lower than the ob-
served rates of DEM formation, but ½r2DEM ðcalcdÞꢃ are close to those.
This means that these reactions are the bottle-necks of the cata-
lytic reaction. Their ratio in the product formation is in the average
1:30. This means that the bulk of DEM is formed in the catalytic cy-
cle B. The catalytic cycle A plays a minor role under the chosen
experimental conditions. Especially the calculations from the con-
centrations in the third hour reaction time (entry 4 in Table 8)
show that the sum of r1DEM ðcalcdÞ ¼ 0:0007 ꢂ 10ꢁ5 mol=dm3 s
and r2DEM ðcalcdÞ ¼ 0:0228 ꢂ 10ꢁ5 mol=dm3 s is practically the
same as the observed rate of DEM formation rDEM (ob-
s) = 0.0237 ꢂ 10ꢁ5 mol/dm3 s. This result is in accord with the
observation that in the low pressure range ([CO] = 0.005–
0.030 mol/dm3) the catalytic reaction is first order with respect
of carbon monoxide.
complex 2. For the latter we used Eq. (17) as r1 ¼ ½1ꢃt=ðkꢁ1 1½COꢃ0=
k11k12½EDAꢃtÞ þ 1=k11g. From the half-time concentration of complex
2 we calculated the rate of DEM formation in cycle B and its concen-
2
tration ð½DEMꢃ3hÞ. The results of the calculations are compiled in
Table 9.
The date in Table 9 show that the diethyl malonate (DEM)
formed in the experiments origins mainly in cycle B if the ratio
of [EDA]0/[CO]0 is higher than 0.1. On the other hand mainly cycle
A is the source of DEM if the [EDA]0/[CO]0 is below 0.1. In the for-
mer case the carbon monoxide concentration has a positive effect
on the rate of DEM formation because the reactions of the com-
plexes 1 and (mainly) 2 are the rate determining. In the latter case
the carbon monoxide concentration has a negative effect on the
rate of DEM formation because the formations of the complexes
(mainly) 1 and 2 are the rate determining.
4. Conclusion
The presented results have proved that the catalytic carbonyl-
ation of ethyl diazoacetate in CH2Cl2 solution at 10 °C in the pres-
ence of ethanol using Co2(CO)8 as the catalyst precursor proceeds
parallel in two different cycles A and B (see Scheme 5). Both cata-
lytic cycles produce the intermediary ethoxycarbonylketene,
which is scavenged in a fast reaction by ethanol to form diethyl
malonate as the final organic product.
In order to explore the correlations in the high pressure range
([CO] = 0.030–1.121 mol/dm3) we performed calculations using
the experimental initial concentrations of entries 25, 26, 32, 33,
and 34 of Table 1. From the initial concentrations of Co2(CO)8, EDA,
and CO in these experiment we calculated the rate and the concen-
tration of complex 1 formation using Eq. (13) as r = [Co2(CO)8]0/
{(kꢁ1[CO]0/k1k2[EDA]0) + 1/k1} at the beginning and at half-time
(1.5 h) of the reaction. From the half-time concentration of complex
1 we calculated the rate of diethyl malonate (DEM) formation in cy-
½Coꢃ-cat:
EtO CCHN þ CO
N
2
ƒƒƒƒƒƒƒƒ!
2
2
rate-determining
þ O@C@CHCO Et EtOH EtO CCH CO Et
ƒƒ!
2
2
2
2
fast
In the first cycle Co2(CO)7 and Co2(CO)7(CHCO2Et) (1), whereas in
the second one Co2(CO)6(CHCO2Et) and Co2(CO)6(CHCO2Et)2 (2)
are the working repeating species. Source of Co2(CO)7 in the first cy-
cle is the catalyst precursor Co2(CO)8. Complex Co2(CO)7(CHCO2Et)
1
cle A and its concentration ð½DEMꢃ3hÞ and the rate of formation of

N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
2027
(1) is the source for Co2(CO)6(CHCO2Et) in the second cycle. The for-
5.4. Computational details
mation of the latter connects cycle A with cycle B. From the compo-
sition of the catalytic reaction mixtures and the observed kinetics of
the individual steps in the cycles we conclude that using sufficient
high concentration ratio of [N2CHCO2Et]/[CO] ꢅ 0.1, the rate-deter-
mining steps at low (1 bar) or high (50 bar) pressure of carbon mon-
oxide are the reactions of Co2(CO)7(CHCO2Et) (1) and
Co2(CO)6(CHCO2Et)2 (2) with carbon monoxide. Complex 2 is 66-
times more effective than complex 1 in this reaction. At low concen-
tration ratio [N2CHCO2Et]/[CO] < 0.1, the reactions of the catalyst
precursors Co2(CO)8 and complex 1 with ethyl diazoacetate become
rate determining, in which reactions the carbon monoxide concen-
tration has a negative effect. If the low concentration ratio of
[N2CHCO2Et]/[CO] is combined with a high CO concentration (for
example at 150 bar pressure), more diethyl malonate is produced
in catalytic cycle A than in catalytic cycle B.
Full geometry optimizations have been performed at the den-
sity functional level of theory without any symmetry constraints
using the GAUSSIAN 03 suite of programs [21]. For all the calculations
the gradient-corrected exchange functional developed by Becke
[22] was utilized in combination with a correlation functional
developed by Perdew [23] and denoted as BP86. For cobalt the va-
lence triple-f SDD basis set following the (8s7p6d1f) ? [6s5p3d1f]
contraction pattern is utilized with the corresponding relativistic
effective core potential [24]. For the other atoms the 6-31G(d,p)
basis set [25] was used. The density fitting basis sets were gener-
ated automatically from the AO primitives by the GAUSSIAN 03 pro-
gram. The stationary points were characterized by frequency
calculations in order to verify that they have zero imaginary fre-
quencies for equilibrium geometries and one imaginary frequency
for transition states. Thermochemistry corrections were taken
from frequency calculations at 298.15 K and 1 atm. Intrinsic reac-
tion coordinate (IRC) analyses [26] were carried out throughout
the reaction pathways to confirm that the stationary points are
smoothly connected to each other. For charge decomposition anal-
yses (CDA) the AOMIX software was used [27].
Computational studies on the ketene formation step have
shown that the coupling of the bridging carbene ligand with a
terminal CO ligand takes place via a low barrier resulting in a
coordinative unsaturated ketene complex, which is converted to
a (l2
g
–g –
2) ketene complex by an uptake of external CO. The (l2
2) complex quickly transforms to a (l1 2) species which
–g
releases ethoxycarbonylketene via a small barrier resulting in the
dibridged heptacarbonyl-dicobalt which reforms complex 1 after
reacting with ethyl diazoacetate.
Acknowledgements
The authors thank the Hungarian Academy of Sciences and the
Hungarian Scientific Research Fund for financial support under
Grant Nos. OTKA NK 71906 and F046959 and also thank the Super-
computer Center of the National Information Infrastructure Devel-
opment (NIIF) Program as well as Flextra Lab KFT. The support of
Bolyai Grant of the Hungarian Academy of Sciences for T.K. is also
acknowledged.
5. Experimental
5.1. General comments
Handling of the carbonyl cobalt complexes was carried out in an
atmosphere of dry (P4O10) and deoxygenated (BTS contact, room
temp.) argon or carbon monoxide utilizing standard Schlenk tech-
niques [17]. Methylene chloride were dried and distilled under an
atmosphere of argon or carbon monoxide according to standard
procedures [18]. IR spectra were recorded on Thermo Nicolet Ava-
tar 330 and Bruker Tensor 27 FTIR spectrometers using 0.00265,
0.00765, 0.0218 or 0.05097 cm CaF2 solution cells, calibrated by
the interference method [19]. Octacarbonyl dicobalt [20] and the
References
[1] N. Ungvári, T. Kégl, F. Ungváry, J. Mol. Catal. A: Chem. 219 (2004) 7.
[2] R. Tuba, F. Ungváry, J. Mol. Catal. A: Chem. 203 (2003) 59.
[3] N. Ungvári, F. Ungváry, in: L. Kollár (Ed.), Modern Carbonylation Methods,
Wiley-VCH, Weinheim, 2008, p. 199 (Chapter 8).
l2-ethoxycarbonylcarbene complexes 1 [2] and 2 [2] were pre-
}
[4] E. Fördos, R. Tuba, L. Párkányi, T. Kégl, F. Ungváry, Eur. J. Org. Chem. (2009)
1994.
pared by literature procedures. Carbon monoxide was obtained
from Messer Hungary.
}
[5] R. Tuba, E. Fördos, F. Ungváry, J. Mol. Catal. A.: Chem. 236 (2005) 113.
[6] E. Fördos, N. Ungvári, L. Párkányi, G. Szalontai, T. Kégl, F. Ungváry, Inorg. Chim.
}
Acta 361 (2008) 1832.
}
[7] N. Ungvári, E. Fördos, T. Kégl, F. Ungváry, Inorg. Chim. Acta 362 (2009) 1333.
5.2. Measurements at atmospheric pressure
}
[8] R. Tuba, E. Fördos, F. Ungváry, Inorg. Chim. Acta 358 (2005) 4081.
[9] A. Jonasson, O. Persson, P. Rasmussen, J. Chem. Eng. Data 45 (2000) 642.
[10] M.P. Magee, H.-Q. Li, O. Morgan, W.H. Hersh, Dalton Trans. (2003) 387.
[11] E. Fördos, N. Ungvári, T. Kégl, F. Ungváry, Eur. J. Inorg. Chem. (2006) 1875.
[12] J.P. Kenny, R.B. King, H.F. Schaefer III, Inorg. Chem. 40 (2001) 900.
[13] T. Kégl, F. Ungváry, J. Organomet. Chem. 692 (2007) 1825.
[14] X. Feng, C. Xie, Z. Liu, Y. Xie, R.B. King, H.F. Schaefer III, Dalton Trans. (2009)
2599.
The reactions of Co2(CO)8, complex 1 and complex 2 with ethyl
diazoacetate and ethanol at atmospheric pressure was performed
in a thermostatted glass reactor connected to a gas burette under
an atmosphere of carbon monoxide or argon. The gas volume
change was followed by reading the gas burette in appropriate
time intervals (usually 8–15 readings for calculation of the initial
rates). The concentration of diethyl malonate, Co2(CO)8, complex
1 and complex 2 in samples of the reaction mixture was calculated
from the IR spectrum using the molar absorbances of diethyl mal-
onate eM (CH2Cl2, 1749 cmꢁ1) = 579 cm2/mmol and eM (CH2Cl2,
1732 cmꢁ1) = 666 cm2/mmol, Co2(CO)8 eM (CH2Cl2, 2022 cmꢁ1) =
}
[15] S. Dapprich, G. Frenking, J. Phys. Chem. 99 (1995) 9352.
[16] A. Bedekovits, L. Kollár, T. Kégl, Inorg. Chim. Acta (submitted for publication).
[17] D.F. Shriver, The Manipulation of Air-Sensitive Compounds, Krieger, Malabar,
FL, 1982.
[18] W.L.F. Armarego, D.D. Perrin, Purification of Laboratory Chemicals, 4th ed.,
Butterworth-Heinemann, Oxford, 1996.
[19] H.H. Willard, L.L. Merritt Jr., J.A. Dean, F.A. Settle Jr., Instrumental Methods of
Analysis, 6th ed., Wadsworth, Belmont, CA, 1981.
697 cm2/mmol, complex
1
eM (CH2Cl2, 1853 cmꢁ1) = 953 cm2/
[20] P. Szabó, L. Markó, G. Bor, Chem. Techn. (Leipzig) 13 (1961) 549.
[21] GAUSSIAN 03, Revision B.05, M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria,
M.A. Robb, J.R. Cheeseman, J.A. Montgomery Jr., T. Vreven, K.N. Kudin, J.C.
Burant, J.M. Millam, S.S. Iyengar, J. Tomasi, V. Barone, B. Mennucci, M. Cossi, G.
Scalmani, N. Rega, G.A. Petersson, H. Nakatsuji, M. Hada, M. Ehara, K. Toyota, R.
Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, M.
Klene, X. Li, J.E. Knox, H.P. Hratchian, J.B. Cross, C. Adamo, J. Jaramillo, R.
Gomperts, R.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W.
Ochterski, P.Y. Ayala, K. Morokuma, G.A. Voth, P. Salvador, J.J. Dannenberg, V.G.
Zakrzewski, S. Dapprich, A.D. Daniels, M.C. Strain, O. Farkas, D.K. Malick, A.D.
Rabuck, K. Raghavachari, J.B. Foresman, J.V. Ortiz, Q. Cui, A.G. Baboul, S.
Clifford, J. Cioslowski, B.B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A.
mmol, complex 2 eM (CH2Cl2, 2080 cmꢁ1) = 3768 cm2/mmol.
5.3. Measurements at higher than atmospheric pressure
Reactions above the atmospheric pressure using CO pressures
were performed in a thermostatted stainless steel autoclave of
20 cm3 capacity. The concentration of the components in the reac-
tion mixture was determined from samples by quantitative infra-
red spectroscopy (see above).

2028
N. Ungvári et al. / Inorganica Chimica Acta 363 (2010) 2016–2028
Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M. W. Wong,
C. Gonzalez, J.A. Pople, Gaussian Inc., Pittsburgh, PA, 2003.
[22] A.D. Becke, Phys. Rev. A 38 (1988) 3098.
[25] W.J. Hehre, R. Ditchfield, J.A. Pople, J. Chem. Phys. 56 (1972) 2257.
[26] C. Gonzalez, H.B. Schlegel, J. Chem. Phys. 90 (1989) 2154.
[27] (a) S.I. Gorelsky, AOMIX: Program for Molecular Orbital Analysis, http://www.sg-
chem.net/, University of Ottawa, 2009;
[23] J.P. Perdew, Phys. Rev. B 33 (1986) 8822.
[24] M. Dolg, H. Stoll, H. Preuss, R.M. Pitzer, J. Phys. Chem. 97(1993) 5852.
(b) S.I. Gorelsky, A.B.P. Lever, J. Organomet. Chem. 635 (2001) 187.