Halide-Free and Bifunctional One-Component Catalysts for the Coupling of Carbon Dioxide and Epoxides

Article abstract of DOI:10.1021/acs.inorgchem.9b00262

In this paper, we first report a new class of halide-free and bifunctional one-component catalysts for the coupling of CO₂ with epoxides. The catalysts do not need halide-based additives or tethered salts attached to the ligand when used for this coupling reaction. As the halide-free and bifunctional one-component catalysts, we chose nonionic and monomeric tetracarbonylchromium(0), tetracarbonylmolybdenum(0), and tetracarbonyltungsten(0) complexes chelated by modified ethylenediamines, namely N,N-dimethylethylenediamine, N,N′-dimethylethylenediamine, N,N,N′-trimethylethylenediamine, and N,N,N′,N′-tetramethylethylenediamine. A simple mixture of M(CO)₆ (M = Cr, Mo, and W) with the modified ethylenediamines shows only one-third of the activity achieved with the tetracarbonyl metal complexes precoordinated to the corresponding modified ethylenediamines. Increasing the number of methyl substituents on the nitrogen atoms of the ethylenediamine derivatives as well as the chromium metal center in the metal carbonyl complex significantly enhanced the catalytic activity. Thus, among the 12 catalysts tested, tetracarbonyl(tetramethylethylenediamine)chromium(0) exhibited the best catalytic activity under the same reaction conditions. Various terminal and internal epoxides were easily converted into the corresponding cyclic carbonates using this chromium system. Calculations based on density functional theory were also carried out to elucidate the mechanism of the coupling reaction.

Full text of DOI:10.1021/acs.inorgchem.9b00262

Article
pubs.acs.org/IC
Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
Halide-Free and Bifunctional One-Component Catalysts for the
Coupling of Carbon Dioxide and Epoxides
Yoseph Kim,†,# Seol Ryu,‡,# Woolee Cho,† Min Kim,† Myung Hwan Park,§ and Youngjo Kim*,†
†Department of Chemistry and BK21+ Program Research Team, Chungbuk National University, Cheongju, Chungbuk 28644,
Republic of Korea
‡Department of Chemistry, Chosun University, Gwangju 61452, Republic of Korea
§Department of Chemistry Education, Chungbuk National University, Cheongju, Chungbuk 28644, Republic of Korea
S
* Supporting Information
ABSTRACT: In this paper, we first report a new class of
halide-free and bifunctional one-component catalysts for the
coupling of CO2 with epoxides. The catalysts do not need
halide-based additives or tethered salts attached to the ligand
when used for this coupling reaction. As the halide-free and
bifunctional one-component catalysts, we chose nonionic and
monomeric tetracarbonylchromium(0), tetracarbonylmolyb-
denum(0), and tetracarbonyltungsten(0) complexes chelated
by modified ethylenediamines, namely N,N-dimethylethyle-
nediamine, N,N′-dimethylethylenediamine, N,N,N′-trimethy-
lethylenediamine, and N,N,N′,N′-tetramethylethylenediamine.
A simple mixture of M(CO)6 (M = Cr, Mo, and W) with the modified ethylenediamines shows only one-third of the activity
achieved with the tetracarbonyl metal complexes precoordinated to the corresponding modified ethylenediamines. Increasing
the number of methyl substituents on the nitrogen atoms of the ethylenediamine derivatives as well as the chromium metal
center in the metal carbonyl complex significantly enhanced the catalytic activity. Thus, among the 12 catalysts tested,
tetracarbonyl(tetramethylethylenediamine)chromium(0) exhibited the best catalytic activity under the same reaction
conditions. Various terminal and internal epoxides were easily converted into the corresponding cyclic carbonates using this
chromium system. Calculations based on density functional theory were also carried out to elucidate the mechanism of the
coupling reaction.
To date, numerous catalytic systems have been reported in the
literature for the synthesis of cyclic carbonates.5−14 As shown in
Chart 1, catalytic systems can be roughly classified into three
INTRODUCTION
■
The rapid increase in the carbon dioxide (CO2) concentration in
the atmosphere is of great concern globally, and this issue must
be addressed. Extensive studies, including CO2 emission
reduction, carbon capture and sequestration, and carbon capture
and utilization, are actively underway to reduce atmospheric
CO2 levels.1 From a synthetic chemist’s point of view, CO2 is an
attractive C1 source because of its safety, abundance, and
renewability.2 However, a considerable amount of energy is
required for the chemical activation of highly oxidized and
thermodynamically stable CO2. Nonetheless, various chemical
substances, such as urea, methanol, cyclic carbonates,
polycarbonates, and sodium 2-hydroxybenzoate, have been
commercially produced from CO2.3 Among the processes for
synthesizing these compounds, the coupling of CO2 and
epoxides to make cyclic carbonates is one of the most active
research fields of CO2 conversion due to its atom economy (no
side products) and broad applicability. Since aliphatic cyclic
carbonates were first synthesized in the 1950s,4 they have been
used in many applications,3 including as aprotic solvents,
electrolytes for lithium-ion batteries, monomers for polymer-
izations, and pharmaceutical intermediates.
Chart 1. Schematic Representation of Catalytic Systems for
the Synthesis of Cyclic Carbonates
categories. The most effective catalytic systems are binary
systems consisting of a metal-based catalyst in conjunction with
a halide additive.5,6 The metal complexes chelated by well-
defined ligands can act as Lewis acids to activate the epoxides.
Lewis basic halides serve as nucleophiles to open the epoxide
ring in the catalytic cycle.7 Unlike binary systems, the
Received: January 27, 2019
© XXXX American Chemical Society
A
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Chart 2. Logical Approach to Conceptionally Novel Halide-Free and Bifunctional One-Component Catalysts
Scheme 1. Synthetic Routes for LnM(CO)4 (n = 1−4; M = Cr, Mo, W)
bifunctional one-component systems depicted in Chart 1b could
definitely reduce the translational entropy cost by combining the
Lewis acid and Lewis base in a single molecule, and
consequently, the close proximity of the two functional groups
could maximize the synergistic effect. Various kinds of
bifunctional one-component catalytic systems equipped with
both Lewis acid metallic regions and nucleophilic fragments in a
single molecule for the synthesis of cyclic carbonates have been
reported in the literature.8−12 Recently, oxo-bridged dimeric
aluminum(salen) complexes as catalysts in the absence of any
additives or tethered salts have been reported for the synthesis of
cyclic carbonates.13,14 Unexpectedly, the μ-oxo group in the oxo-
bridged dimeric aluminum(salen) complexes was easily
converted into a μ-carbonato group in the first step of the
coupling reaction, and epoxide insertion subsequently occurred.
Because no halide-based additive was present and two aluminum
centers in a single molecule were involved, this system belongs
to the halide-free dimeric systems shown in Chart 1c. However,
only two papers on aluminum-based dimers have been reported
to date.13,14
Apart from the three systems shown in Chart 1, another
promising catalytic system is halide-free and bifunctional one-
component catalysts. As shown in Chart 2a, we hypothesized
that catalysts of this new concept could be achieved if the ligand
molecule has both a binding site and a Lewis basic site, and at the
same time, the Lewis acidic metal center remains coordinatively
unsaturated to allow epoxide coordination. Thus, we chose
LnM(CO)4 (M = Cr, Mo, and W; Ln = bidentate diamine
ligands) as model catalysts for the halide-free and bifunctional
one-component systems. The LnM(CO)4 compounds were
selected as model catalysts for the following reasons. (1)
Bidentate diamine ligands could bind to the metal center and
still have a free Lewis basic site because in solution, bidentate
diamines of these metal complexes are hemilabile, and this
lability results in a vacant site for epoxide coordination as
depicted in Chart 2b. The kinetics of the reaction between
(N,N,N′,N′-tetramethylethylenediamine)M(CO)4 (M = Cr
and W) and phosphite were reported in the literature.15,16 (2)
The Lewis basicities17 of diamine ligands Ln and chromium
complexes18−20 for cyclic carbonate synthesis are well-known.
Thus, we chose LnM(CO)4 as a halide-free and bifunctional one-
component system. (3) The presence of strong-field CO ligands
permits the generation of LnM(CO)4 as the sole product
without contamination by any side products from the reaction
between commercially available M(CO)6 and bidentate ligand
L.21−27 Unlike other chromium-based catalysts with oxidation
states of +3 to +5,18−20 the oxidation state of LnM(CO)4 is zero,
which means that LnM(CO)4 is diamagnetic and its purity could
be determined by NMR experiments.
In this context, we report here group VI metal (0)
tetracarbonyls containing four different kinds of diamines as a
new class of halide-free and bifunctional one-component
monomeric catalysts for the coupling of CO2 with terminal or
internal epoxides. The detailed mechanism for this reaction was
also confirmed by density function theory calculations.
RESULTS AND DISCUSSION
■
Synthesis and Characterization of Halide-Free and
Bifunctional One-Component Catalysts. As outlined in
Scheme 1, 12 target products, group VI metal tetracarbonyls
chelated by bidentate ligands, namely, N,N-dimethylethylenedi-
amine (L1), N,N’-dimethylethylenediamine (L2), N,N,N′-
B
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
trimethylethylenediamine (L3), and N,N,N′,N′-tetramethyle-
thylenediamine (L4), as halide-free and bifunctional one-
component catalysts, were prepared in one step directly from
commercially available M(CO)6 (M = Cr, Mo, and W) and
bidentate ligands L1−L4. To synthesize LnM(CO)4, a mixture of
polar solvents di-n-butyl ether and THF (20:3 v/v) was applied.
The synthesis of the Cr and Mo compounds required a relatively
short reaction time of 2 h, while the W compounds need an
extended reaction time of 7 h. After appropriate workup, high
purity products LnCr(CO)4, LnMo(CO)4, and LnW(CO)4 were
obtained as yellow solids in yields of 49−64%, 78−92%, and
83−90%, respectively. As the metal center becomes heavier and
the number of methyl groups attached to the N atom increases,
the yield of the final product increases.
Table 1. Catalyst Screening for the Coupling Reaction of
Epichlorohydrin and CO2
a
b
c
entry
catalyst
Cr(CO)6
yield [%]
TON
TOF [h−1
]
1
2
3
4
5
6
7
8
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Mo(CO)6
W(CO)6
All LnM(CO)4 compounds were freely soluble in polar
organic solvents such as THF, acetone, MeOH, and DMSO but
insoluble in toluene and nonpolar solvents such as hexane.
These compounds were relatively stable in air and even in polar
solvents for a few days. The solubility of LnM(OC)4 compounds
increases as the increased number of methyl substituents on the
nitrogen atoms increases. All compounds were fully charac-
L1
L2
0
L3
19.4
23.4
27.8
28.1
28.4
27.4
19.0
8.8
47.8
28.6
25.6
65.9
46.9
28.6
87.2
56.3
47.6
194
234
278
281
284
374
190
88
478
286
256
659
469
286
872
563
476
55.4
66.9
79.4
80.3
81.1
107
54.3
25.1
137
81.7
73.1
188
134
81.7
249
161
136
L4
Cr(CO)6 + L4
Mo(CO)6 + L4
W(CO)6 + L4
L1Cr(CO)4
L1Mo(CO)4
L1W(CO)4
L2Cr(CO)4
L2Mo(CO)4
L2W(CO)4
L3Cr(CO)4
L3Mo(CO)4
L3W(CO)4
L4Cr(CO)4
L4Mo(CO)4
L4W(CO)4
9
1
terized by H and 13C NMR and IR spectroscopy (see the
10
11
12
13
14
15
16
17
18
19
20
21
22
Supporting Information). Newly synthesized L2Cr(CO)4,
L3Cr(CO)4, and L3W(CO)4 species were also characterized
by elemental analysis.
1
In comparison with free ligands L1−L4, all signals in the H
and 13C NMR spectra are shifted downfield, which is a result of
complexation with the metal. The magnitude of the downfield
shifts in the 1H NMR spectra roughly decreased in the order Cr
to Mo to W under the fixed ligand system. In addition, in the 1H
NMR spectra, the magnitude of the downfield shift is greater for
N−H resonances than for N−CH3. The N−H and N−CH3
resonances shifted downfield by approximately 1.55 and 0.53
ppm, respectively, relative to those of the corresponding free
ligand.
a
b
Yield of pure isolated cyclic carbonate. TON (turnover number) =
c
(mol of epichlorohydrin consumed)/(mol of catalyst). TOF
(Turnover frequency) = (mol of epichlorohydrin consumed)/(mol
of catalyst × time (h)).
Coupling of CO2 with Epoxides. To determine if
LnM(CO)4 could act as halide-free and bifunctional one-
component catalysts in the coupling reaction, the coupling of
CO2 with epichlorohydrin (EH) was performed using a fixed
[EH]/[catalyst] molar ratio of 1000 and a CO2 pressure of 10
bar at 100 °C for 3.5 h. The results are summarized in Table 1.
As expected, M(CO)6 (M = Cr, Mo, and W) did not show any
catalytic activity (Table 1, entries 1−3). Our recent report17 on
L1−L4 as organocatalysts by themselves for the synthesis of
cyclic carbonates showed that the greater number of methyl
substituents on the N atoms of the ethylenediamine ligand
provided the higher catalytic activity (entries 4−7). Under the
same conditions, M(CO)6 in conjunction with L4 showed
activities similar to that achieved with L4 alone (entries 8−10).
These results indicate that simple mixtures of M(CO)6 with L4
did not show any cooperative effect in their activity. Then, we
explored tetracarbonyl group VI metal complexes chelated by
L1−L4. Surprisingly, all complexes could act as halide-free and
bifunctional one-component catalysts (entries 11−22). Inter-
estingly, the catalytic activities of the metal complexes with more
methyl substituents are greater than those of complexes with
fewer methyl groups (entries 11, 14, 17, and 20; entries 12, 15,
18, and 21; entries 13, 16, 19, and 22). In addition, the activity
decreased going down the group from Cr to Mo to W with the
same ligand (entries 11−13, 14−16, 17−19, and 20−22).
Among the 12 tetracarbonyl complexes, L4Cr(CO)4 showed
the highest TON (872) and TOF (249.1) (entry 20). Many
kinds of chromium-based complexes have been reported as
catalysts in conjunction with a cocatalyst for the synthesis of
cyclic carbonates.18−20 Surprisingly, bidentate and tridentate
ligands chelated to chromium have never been used in this
reaction. Although the metal sources have been extended to
molybdenum and tungsten, to the best of our knowledge, group
VI complexes in zero oxidation states chelated by bidentate
ligands have never been used as catalysts for the synthesis of
cyclic carbonate.
The influences of the concentration of L4Cr(CO)4 catalyst,
CO2 pressure, and temperature on the coupling reaction
between CO2 and epichlorohydrin were investigated, and the
results are summarized in Table 2. Using a fixed catalyst
concentration and temperature, the catalytic activity was found
to be proportional to the CO2 pressure (Table 2, entries 1−4;
entries 5−8; entries 9−12; entries 19−22). The highest TOF
(2010) was obtained using 0.01 mol % catalyst loading and 30
bar of CO2 pressure at 100 °C in 1 h (entry 4). As expected,
higher catalyst concentrations resulted in higher activities
(entries 1, 5, 9, and 19; entries 2, 6, 10, and 20; entries 3, 7,
11, and 21; entries 4, 8, 12, and 22). The TOF increased rapidly
from 42 to 541 as the reaction temperature was gradually
increased from 25 to 125 °C (entries 14−18).
To demonstrate the important discovery that L4Cr(CO)4
acting as a halide-free and bifunctional one-component catalyst
C
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
is not restricted to epichlorihydrin, the synthesis of a variety of
cyclic carbonates from the corresponding terminal and internal
epoxides as substrates was carried out (Figure 1). Substrates
include seven terminal epoxides, propylene oxide (1a), 1,2-
epoxybutane (1b), 1,2-epoxyhexane (1c), epichlorohydrin
(1d), glycidyl methyl ether (1e), 1,2-epoxy-3-phenoxypropane
(1f), and tert-butyl glycidyl ether (1g), and four internal
epoxides, 1,2-epoxy-2-methylpropane (1h), 2,3-epoxybutane
(1i), cis-cyclohexene oxide (1j), and cis-cyclopentene oxide
(1k). The reactions of epoxides 1a−1c with simple alkanes (R =
Me, Et, and n-Bu) showed a substantial dependence on chain
length. Interestingly, epoxides with short chains gave higher
conversions. Among the tested substrates, 1d was the most
active epoxide in the coupling reaction, and it showed a TON of
938 and TOF of 312.7 within only 3 h. 1e with a −CH2OMe
substituent and 1f with a −CH2OPh substituent showed
excellent TONs of 893 and 971, respectively; however, 1g
with a sterically hindered −CH2OBut group exhibited a
relatively slow conversion. Interestingly, oxo-bridged dimeric
aluminum(salen) complexes, a typical halide-free dimeric
system (shown in Chart 1c), could not catalyze the reaction of
internal epoxides for the synthesis of the corresponding cyclic
carbonates even under harsh conditions of 2.5 mol % catalyst
loading and 50 bar of CO2 pressure at 50 °C for 60 h.13 Our
system L4Cr(CO)4 could catalyze the reaction of internal
epoxides 1h−1k using 0.1 mol % catalyst loading and 10 bar of
CO2 pressure at 120 °C in 48 h. In addition, halide-free
bifunctional one-component catalyst L4Cr(CO)4 showed serval
times better higher TOF-based catalytic activities than halide-
free dimeric system in the synthesis of terminal epoxides.13 As
shown in Figures S67−S70, cis-hexahydrobenzo[d][1,3]dioxol-
2-one (2j)28 and cis-tetrahydro-4H-cyclopenta[d][1,3]dioxol-2-
one (2k)29 with the retained original configuration of the
corresponding epoxides were obtained. This means that the
cyclic carbonate obtained in the coupling reaction is made by the
nucleophilic ring-opening (SN2) of epoxide, followed by the SN2
ring-closing pathway, rather than the SN1 ring-closing pathway.
Table 2. Screening of Coupling Reaction Conditions Using
L4Cr(CO)4
CO2
L4Cr(CO)4 pressure
time yield
TOF
c
a
b
Entry
[mol %]
[bar]
T [°C] [h]
[%]
TON
[h−1
]
1
2
3
4
5
6
7
8
0.01
0.01
0.01
0.01
0.05
0.05
0.05
0.05
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.1
0.5
0.5
0.5
1
5
10
30
1
100
100
100
100
100
100
100
100
100
100
100
100
100
100
125
75
1
1
1
1
1
1
1
1
1
1
1
3
4
1
1
1
1
1
1
1
1
1
3.2
320
880
1390
2010
82
228
336
498
66
167
248
749
990
324
541
156
84
320
880
1390
2010
82
228
336
498
66
167
248
250
248
324
541
156
84
8.8
13.9
20.1
4.1
11.4
16.8
24.9
6.6
16.7
24.8
74.9
99
32.4
54.1
15.6
8.4
5
10
30
1
9
10
11
12
13
14
15
16
17
18
19
20
21
22
5
10
10
10
30
30
30
30
30
1
50
25
100
100
100
100
4.2
8.3
24.1
35.8
46.7
42
17
48
72
42
17
48
72
5
10
30
0.5
93
93
a
b
Yield of pure isolated cyclic carbonate. TON (turnover number) =
c
(mol of epichlorohydrin consumed)/(mol of catalyst). TOF
(turnover frequency) = (mol of epichlorohydrin consumed)/(mol
of L4Cr(CO)4 × time (h)).
Figure 1. Scope of epoxides using L4Cr(CO)4.
D
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 2. Variable temperature NMR spectra for L1W(CO)4.
Figure 3. Plausible mechanism for coupling the reaction of CO2 with propylene oxide using L4Cr(CO)4. For clarity, the four carbonyl groups attached
to the chromium center in L4Cr(CO)4 are omitted.
E
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 4. Computed energy profile of cyclic propylene carbonate formation with catalyst L4Cr(CO)4 at 373.15 K and 10 bar. For clarity, the four
carbonyl groups attached to the chromium center in L4Cr(CO)4 are omitted.
Figure 5. Two possible mechanistic pathways from intermediate D.
Mechanism for the Coupling Reaction of CO2 with
Epoxides. Figure 2 shows the chemical shifts for −NMe2 and
−NH2 arm protons in the variable temperature NMR spectra of
L1W(CO)4 in DMSO-d6. Interestingly, −NMe2 protons of
L1W(CO)4 appear as a two singlets (or a doublet) at room
temperature. Similar observation for 5-membered ring com-
pounds with W−NMe2 bond was reported in the literature.30−32
This peak splitting is probably due to the coordination of ring
nitrogen of L1 with tungsten (I = 1/2). A clear upfield shift in the
−NH2 arm of L1W(CO)4 as a broad singlet from 5.34 ppm at 25
°C to 5.13 ppm at 100 °C was observed. As the temperature goes
up, the methyl protons of the −NMe2 arm shifted downfield and
new two singlets (or a doublet) appear at the downfield region.
The upfield shift of −NH2 protons and the peak splitting of
NMe2 protons in complex L1W(CO)4 indicated that the −NH2
arm of the bound diamine in L1W(CO)4 dissociates reversibly at
high temperature and other −NMe2 arm is firmly attached to
metal center even at high temperature of 100 °C. This result
supports the existence of hemilability for our catalytic systems
(see Figures S71−S75 in the Supporting Information).
Figures 3 and 4 summarize the computational mechanistic
investigations of the coupling reaction of carbon dioxide and
F
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 6. Comparison of the energy barriers in the reactions catalyzed by L4M(CO)4. For clarity, the four carbonyl groups attached to the metal center
in L4M(CO)4 are omitted.
propylene oxide using our halide-free and bifunctional one-
component catalyst L4Cr(CO)4, and the proposed mechanism
takes into account all our experimental observations and
theoretical calculation studies.
intermediate B with a solution-phase binding free energy of only
1.65 kcal/mol. Then, the sterically less hindered carbon of the
propylene oxide is attacked by the basic −NMe2 group in a
nucleophilic fashion to give eight-membered metallacycle C via
transition state B−TS. This step features the highest activation
barrier (36.18 kcal/mol) in the catalytic cycle from A to C
although there is another comparable, but lower barrier in the
later part. Intermediate C can attack CO2 in a nucleophilic
manner leading to five-coordinated intermediate E with a vacant
site via D. As shown in Figure 5, it is also possible that
intermediate D may produce 10-membered metallacyclic
intermediate E′ with six-coordinated Cr center through other
pathway. However, this mechanistic pathway was removed from
consideration due to its high activation energy of 54.70 kcal/mol
to give the final products (see Supporting Information).
Intriguingly, the activation energy from E to E−TS is only
4.57 kcal/mol, which implies the formation of the C−O bond
and cleavage of the C−N bond are facile. It is also interesting
that the energy barrier of 35.36 kcal/mol from C to E−TS in the
second half of the cycle is comparable to the highest one, 36.18
kcal/mol from A to B−TS. We found that the most important
TDI and TDTS in Figure 4 are A (0.60) and B-TS (0.80) with
the energetic span ΔG‡ = 36.18 kcal/mol. The TDI and TDTS
in alternative mechanistic pathway are E′ (0.64) and E′−TS
(1.00) with ΔG‡ = 54.70 kcal/mol, where the numbers 0.64 and
In order to help understand catalytic cycle in our work, we
adopted the energetic span model which systematically
expresses TOF in terms of the energies of all states in catalytic
cycle.33,34 The model quantitatively identifies TOF-determining
intermediate (TDI) and TOF-determining transition state
(TDTS) in a single catalytic cycle based on the degree of
TOF control of each state, which measures how much influential
the state is on the TOF value. The TDTS − TDI energy
difference ΔG‡, called the energetic span, can serve as the
apparent activation energy, expressing approximately TOF ≈
‡
kBT e−ΔG /RT. Thus, the ΔG‡ values characterizing catalytic
h
cycles allow us to compare the plausibility of proposed
mechanisms and the efficiency of various catalysts.
Six-coordinated precatalyst L4Cr(CO)4, denoted as A0, was
activated to hemilabile species A with a vacant site, resulting
from one −NMe2 group detaching from the chromium center of
A0. The actual catalytic species A was first optimized, and the
Gibbs free energies, ΔG(sol), of all other structures are shown
relative to that of A. The catalytic cycle starts with the
occupation of the vacant site by propylene oxide to afford
G
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
Figure 7. Schematic computed energy profiles of cyclic propylene carbonate formation with catalyst LnCr(CO)4 (n = 1−4) at 373.15 K and 10 bar.
Figure 8. Schematic computed energy profiles of cyclic propylene carbonate formation with catalyst L4Cr(CO)4 at 373.15 K and 1, 5, 10, and 30 bar.
1.00 in parentheses represent the degree of TOF control of the
state. Therefore, the barrier of 36.18 kcal/mol from A to C can
be considered an apparent rate-determining step of the overall
catalytic cycle. On the other hand, we also checked the
possibility that the reaction proceeds by intermolecular attack of
free diamine (such as entry 8 in Table 1, Cr(CO)6 + L4) on the
H
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
metal-bound epoxide. However, we found the barrier of 43.0
kcal/mol for such attack, and so such catalysts proved to be less
efficient than L4Cr(CO)4 (see Supporting Information).
As shown in Table 1, the activities decrease going down the
group from Cr to Mo to W with the same ligand. For DFT
calculations, we consistently used ligand L4, which showed the
highest activity among the four bidentate ligands. Figure 6 shows
the computed energy profiles of cyclic propylene carbonate
formation using L4M(CO)4 (M = Cr, Mo, and W). It is
noteworthy that the rate-determining step for L4Cr(CO)4 is
primarily in the first half of catalytic cycle with a comparable
barrier in the second half; however, for the two other metals,
L4Mo(CO)4 and L4W(CO)4, the second half steps are solely
rate-determining. Although the overall shapes of profiles look
similar, the first half cycle tends to be stabilized while the other
half is destabilized as the central metal becomes heavier. This
trend shifts the TOF-determining spotlight to the second half,
making C (1.00) and E−TS (1.00) be TDI and TDTS,
respectively, for the heavier catalysts. Note that the degree of
TOF control in this single set of TDI and TDTS is 1.00, which
implies that TDI and TDTS are unambiguously defined in the
case of heavier metal catalysts. According to the energetic span
model shown in Figure 6, the apparent activation energies are
found to be ΔG‡ = 36.18 kcal/mol for L4Cr(CO)4, 42.66 kcal/
mol for L4Mo(CO)4, and 50.11 kcal/mol for L4W(CO)4, which
matches with experimental trends.
According to Table 1, the catalytic activities of the metal
complexes with more methyl substituents are higher than those
of complexes with fewer methyl groups. Unlike L4Cr(CO)4,
catalysts L1Cr(CO)4−L3Cr(CO)4 can have two or four ΔG‡
values depending on which nitrogen atom is detached from the
chromium center. Figure 7 illustrates that the average apparent
activation energies ΔG‡ obtained from the energetic span model
are 36.18 kcal/mol for L4Cr(CO)4, 40.89 kcal/mol for
L3Cr(CO)4, 44.93 kcal/mol for L2Cr(CO)4, and 45.23 kcal/
mol for L1Cr(CO)4 (see Tables S3 and S4 in the Supporting
Information). In other words, the coupling reaction of carbon
dioxide and propylene oxide experiences 3.71−9.05 kcal/mol
higher hindrance with L1Cr(CO)4−L3Cr(CO)4 than with
L4Cr(CO)4. Therefore, our calculations suggest that L4Cr-
(CO)4 should exhibit a considerably higher activity than the
other catalysts when taking the Arrhenius factor into account. In
fact, L4Cr(CO)4 (Table 1, entry 20) showed 1.3−3.2 fold higher
activity than the three other catalysts (Table 1, entries 11, 14,
and 17).
Finally, it is also noteworthy to mention pressure effects as the
catalytic activities in experiment were found to be proportional
to the pressure of CO2. Unfortunately, the energetic span model
we adopted for activity analysis deals with only the temperature
and the energetics of the intermediates and transition states, and
the concentration or pressure effects are not taken into account
in calculating TOF. However, we can consider changes in the
pressure under which the whole reaction proceeds, not limiting
the pressure effects just to that of CO2. As illustrated in Figure 8,
the two barriers in Gibbs energy profiles tend to decrease as the
pressure increases, from 1 to 5, 10, and 30 bar. The entropy
change for each step is negative except for the formation of final
products. As the pressure increases, on the other hand, the
translational entropy generally decreases while the number of
reactants is greater than that of products (see Table S5 in the
Supporting Information). It results in less negative entropy
change and less positive Gibbs energy change, meaning less
hindered catalytic reaction cycles. The decreases in energy
barriers are 1.2, 1.7, and 2.5 kcal/mol as the pressure increases
from 1 bar to 5, 10, and 30 bar, respectively. Note also that the
whole reaction Gibbs energy decreases from −2.03 to −3.22, −
3.73, and −4.55, as the pressure increases from 1 to 5, 10, and 30
bar.
CONCLUSION
■
In conclusion, we hypothesized and synthesized a new class of
halide-free and bifunctional one-component catalysts for the
synthesis of cyclic carbonates from CO2 and epoxides. The
catalysts were rationally and systematically designed, and we
chose well-known classical group VI metal(0) tetracarbonyl
complexes with bidentate diamine ligands as halide-free and
bifunctional one-component catalysts. These kind of catalysts
were selected because of the facile generation of a vacant site on
the metal center for epoxide coordination due to the hemilability
of bidentate diamines on group VI(0) carbonyls. In addition, in
situ generated and neutral amine-centered nucleophiles, which
are less likely to cause reactor corrosion than halide-based
nucleophiles, can open the epoxide ring, and the translational
entropy penalty was reduced by using an intramolecular reaction
with nucleophile and epoxide in close proximity. Furthermore,
the nucleophilicity of the diamine ligand can be easily controlled
by the introduction of electron-donating alkyl groups into the N
atoms. The optimized catalyst was highly effective in forming
cyclic carbonates with a wide range of different substrates. These
new catalysts can be readily synthesized and obtained in pure
form with almost no contamination. Thorough evaluation of the
catalytic activity using DFT theoretical mechanistic studies also
supported the utility of these novel halide-free and bifunctional
one-component catalysts.
ASSOCIATED CONTENT
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.inorg-
chem.9b00262.
■
S
Experimental procedures, analytical data, computational
details, and Cartesian coordinates of all computed
1
structures and copies of the H and 13C NMR spectra
and IR spectra for all products (PDF)
AUTHOR INFORMATION
Corresponding Author
*(Y.K.) E-mail: ykim@chungbuk.ac.kr.
■
ORCID
Seol Ryu: 0000-0003-0160-5004
Min Kim: 0000-0002-1710-2682
Myung Hwan Park: 0000-0002-8858-5204
Youngjo Kim: 0000-0001-8571-0623
Author Contributions
#Y.K. and S.R. contributed equally to this work.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work was financially supported by National Research
Foundation of Korea (NRF) grants funded by the Korea
government (No. 2017M1A2A2049159 and No.
2018R1A2B6004037) and the National Institute of Super-
computing and Network/Korea Institute of Science and
I
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry
Article
(19) Darensbourg, D. J.; Chung, W.-C. Availability of Other Aliphatic
Polycarbonates Derived from Geometric Isomers of Butene Oxide and
Carbon Dioxide Coupling Reactions. Macromolecules 2014, 47, 4943−
4948.
(20) Castro-Osma, J. A.; Lamb, K. J.; North, M. Cr(salophen)
Complex Catalyzed Cyclic Carbonate Synthesis at Ambient Temper-
ature and Pressure. ACS Catal. 2016, 6, 5012−5025.
Technology Information with supercomputing resources
including technical support (KSC-2017-C2-0019). Y.K. ex-
presses thanks for the financial support from the research year of
Chungbuk National University in 2018.
REFERENCES
■
́
́
(21) Herve, A.-C.; Yaouanc, J.-J.; Clement, J.-C.; des Abbayes, H.;
Toupet, L. Hemilability of the Primary Amine−Metal Bond in
Polyamine-(Group 6) Metal Carbonyl Complexes. J. Organomet.
Chem. 2002, 664, 214−222.
(1) Cokoja, M.; Bruckmeier, C.; Rieger, B.; Herrmann, W. A.; Ku
̈
hn,
F. E. Transformation of Carbon Dioxide with Homogeneous
Transition-Metal Catalysts: a Molecular Solution to a Global
Challenge? Angew. Chem., Int. Ed. 2011, 50, 8510−8537.
(2) Aresta, M.; Dibenedetto, A.; Angelini, A. Catalysis for the
Valorization of Exhaust Carbon: from CO2 to Chemicals, Materials, and
Fuels. Technological Use of CO2. Chem. Rev. 2014, 114, 1709−1742.
(3) Maeda, C.; Miyazaki, Y.; Ema, T. Recent Progress in Catalytic
Conversions of Carbon Dioxide. Catal. Sci. Technol. 2014, 4, 1482−
1497.
(22) Shiu, K.-B.; Wang, S.-L.; Liao, F.-L. Organotransition-metal
Complexes of Multidentate Ligands: XVI. On the Nature of the Sigma-
Donicity of the Saturated Nitrogen Ligands; Chelate-assisted
Weakening of the α-N−H Bond: Synthesis, and Spectral and Structural
Study of [Mo(N−N)(CO)4] (N−N = Saturated Nitrogen Bidentate
Ligands). J. Organomet. Chem. 1991, 420, 207−215.
(23) Powell, J.; Lough, A.; Raso, M. Synthesis, Structure and Relative
Hydrolytic Stability of cis-[PtCl2L] and cis-[W(CO)4L] Where L is the
Bidentate Triaminophosphine Ligand [MeNCH2CH2N(Me)PN(Me)-
CH2]2. J. Chem. Soc., Dalton Trans. 1994, 1571−1576.
(24) Connor, J. A.; Lloyd, J. P. Chromium(0) Complexes of Sterically
Hindered vic-Diamines. J. Chem. Soc. A 1970, 3237−3242.
̃
(25) Kromer, L.; Coelho, A. C.; Bento, I.; Marques, A. R.; Romao, C.
(4) Peppel, W. Preparation and Properties of the Alkylene Carbonates.
Ind. Eng. Chem. 1958, 50, 767−770.
(5) Martín, C.; Fiorani, G.; Kleij, A. W. Recent Advances in the
Catalytic Preparation of Cyclic Organic Carbonates. ACS Catal. 2015,
5, 1353−1370.
(6) Comerford, J. W.; Ingram, I. D. V.; North, M.; Wu, X. Sustainable
Metal-Based Catalysts for the Synthesis of Cyclic Carbonates
Containing Five-Membered Rings. Green Chem. 2015, 17, 1966−1987.
(7) North, M.; Pasquale, R. Mechanism of Cyclic Carbonate Synthesis
from Epoxides and CO2. Angew. Chem., Int. Ed. 2009, 48, 2946−2948.
(8) Miao, C.-X.; Wang, J.-Q.; Wu, Y.; Du, Y.; He, L.-N. Bifunctional
Metal-Salen Complexes as Efficient Catalysts for the Fixation of CO2
with Epoxides under Solvent-Free Conditions. ChemSusChem 2008, 1,
236−241.
C. The Effect of Specific Modifications of the Amine Ligands on the
Solubility, Stability, CO Release to Myoglobin and Whole Blood, Cell
Toxicity and Haemolytic Index of [Mo(CO)4(NR3)2] Complexes. J.
Organomet. Chem. 2014, 760, 89−100.
(26) King, R. B.; Fronzaglia, A. Organometallic Chemistry of the
Transition Metals. XV. New Olefinic and Acetylenic Derivatives of
Tungsten. Inorg. Chem. 1966, 5, 1837−1846.
(27) Szymanska-Buzar, T. Photochemical Synthesis of the Tungsten-
(II) Complex [WCl2(CO)3(bipy)]. J. Organomet. Chem. 1989, 375,
85−89.
́
(9) Melendez, J.; North, M.; Villuendas, P. One-Component Catalysts
for Cyclic Carbonate Synthesis. Chem. Commun. 2009, 2577−2579.
(10) Tian, D.; Liu, B.; Gan, Q.; Li, H.; Darensbourg, D. J. Formation of
Cyclic Carbonates from Carbon Dioxide and Epoxides Coupling
Reactions Efficiently Catalyzed by Robust, Recyclable One-Compo-
nent Aluminum-Salen Complexes. ACS Catal. 2012, 2, 2029−2035.
(11) Luo, R.; Zhou, X.; Zhang, W.; Liang, Z.; Jiang, J.; Ji, H. New Bi-
functional Zinc Catalysts Based on Robust and Easy-to-Handle N-
Chelating Ligands for the Synthesis of Cyclic Carbonates from
Epoxides and CO2 under Mild Conditions. Green Chem. 2014, 16,
4179−4189.
(28) Tezuka, K.; Komatsu, K.; Haba, O. The Anionic Ring-Opening
Polymerization of Five-Membered Cyclic Carbonates Fused to the
Cyclohexene Ring. Polym. J. 2013, 45, 1183−1187.
(29) Buttner, H.; Steinbauer, J.; Werner, T. Synthesis of Cyclic
̈
Carbonates from Epoxides and Carbon Dioxide by Using Bifunctional
One-component Phosphorus-based Organocatalysts. ChemSusChem
2015, 8, 2655−2669.
(30) Okazaki, M.; Suzuki, E.; Miyajima, N.; Tobita, H.; Ogino, H.
Facile Isomerization of a Tungsten Silyl Complex to a Base-Stabilized
Silyene Complex via 1,2-Migration of an Aryl Group. Organometallics
2003, 22, 4633−4635.
(12) Ren, W.-M.; Liu, Y.; Lu, X.-B. Bifunctional Aluminum Catalyst
for CO2 Fixation: Regioselective Ring Opening of Three-Membered
Heterocyclic Compounds. J. Org. Chem. 2014, 79, 9771−9777.
(13) Castro-Osma, J. A.; North, M.; Wu, X. Development of a Halide-
Free Aluminium-Based Catalyst for the Synthesis of Cyclic Carbonates
from Epoxides and Carbon Dioxide. Chem. - Eur. J. 2014, 20, 15005−
15008.
(31) Buffin, B. P.; Richmond, T. G. Coordination Chemistry of
Chelating Nitrogen Ligands with Tungsten Carbonyl Nitriles.
Polyhedron 1990, 9, 2887−2893.
(32) Dreisch, K.; Andersson, C.; Stalhandske, C. Synthesis of
MO2Cl2(N,N,N′,N′-tetramethylethylenediamine) (M = Mo and W)
and Crystal structure of WO2Cl2(N,N,N′,N′-tetramethylethylenedi-
amine)−an Unprecedented Coordination Geometry in the WO2Cl2
Core. Polyhedron 1992, 11, 2143−2150.
(33) Kozuch, S.; Shaik, S. How to Conceptualize Catalytic Cycles?
The Energetic Span Model. Acc. Chem. Res. 2011, 44, 101−110.
(34) Kozuch, S.; Shaik, S. A Combined Kinetic−Quantum
Mechanical Model for Assessment of Catalytic Cycle: Application to
Cross-Coupling and Heck Reactions. J. Am. Chem. Soc. 2006, 128,
3355−3365.
(14) Castro-Osma, J. A.; North, M.; Offermans, W. K.; Leitner, W.;
Muller, T. E. Unprecedented Carbonato Intermediates in Cyclic
̈
Carbonate Synthesis Catalysed by Bimetallic Aluminium(Salen)
Complexes. ChemSusChem 2016, 9, 791−794.
(15) Faber, G. L.; Walsh, T. D.; Dobson, G. R. Octahedral Metal
Carbonyls. IX.1 Kinetics of the Reaction of N,N,N’,N’-Tetramethyle-
thylenediaminechromium Tetracarbonyl with Triethyl Phosphite. J.
Am. Chem. Soc. 1968, 90, 4178−4179.
(16) Dobson, G. R.; Moradi-Araghi, A. Octahedral Metal Carbonyls.
XLV. *Kinetics and Mechanism of Ligand Exchange in Tungsten
Carbonyl Complexes Containing Chelating Ligands Bonding Through
Nitrogen. Inorg. Chim. Acta 1978, 31, 263−269.
(17) Cho, W.; Shin, M. S.; Hwang, S.; Kim, H.; Kim, M.; Kim, J. G.;
Kim, Y. Tertiary Amines: A New Class of Highly Efficient Organo-
catalysts for CO2 Fixations. J. Ind. Eng. Chem. 2016, 44, 210−215.
(18) Ramidi, P.; Sullivan, S. Z.; Gartia, Y.; Munshi, P.; Griffin, W. O.;
Darsey, J. A.; Biswas, A.; Shaikh, A. U.; Ghosh, A. Catalytic Cyclic
Carbonate Synthesis Using Epoxide and Carbon Dioxide: Combined
Catalytic Effect of Both Cation and Anion of an Ionic CrV(O) Amido
Macrocyclic Complex. Ind. Eng. Chem. Res. 2011, 50, 7800−7807.
J
DOI: 10.1021/acs.inorgchem.9b00262
Inorg. Chem. XXXX, XXX, XXX−XXX