[Zinc(II)(Pyridine-Containing Ligand)] Complexes as Single-Component Efficient Catalyst for Chemical Fixation of CO2 with Epoxides

Article abstract of DOI:10.1002/ejoc.202100409

The reaction between epoxides and CO₂ to yield cyclic carbonates is efficiently promoted under solvent-free and relatively mild reaction conditions (0.5 mol % catalyst loading, 0.8 MPa, 125 °C) by zinc(II) complexes of pyridine containing macrocyclic ligands (Pc?L pyridinophanes). The zinc complexes have been fully characterized, including X-ray structural determination. The [Zn(II)X(Pc?L)]X complexes showed good solubility in several polar solvents, including cyclic carbonates. The scope of the reaction under solvent-free conditions has been studied and good to quantitative conversions with excellent selectivities have been obtained, starting from terminal epoxides. When solvent-free conditions were not possible (solid epoxides or low solubility of the catalyst in the oxirane) the use of cyclic carbonates as solvents has been successfully investigated. The remarkable stability of the catalytic system has been demonstrated by a series of consecutive runs.

Full text of DOI:10.1002/ejoc.202100409

European Journal of Organic Chemistry
Accepted Article
Accepted Article
Title: [Zinc(II)(Pyridine-Containing Ligand)] Complexes as Single-
Component Efficient Catalyst for Chemical Fixation of CO2 with
Epoxides
Authors: Matteo Cavalleri, Nicola Panza, Armando di Biase, Giorgio
Tseberlidis, Silvia Rizzato, Giorgio Abbiati, and Alessandro
Caselli
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Eur. J. Org. Chem. 10.1002/ejoc.202100409
Link to VoR: https://doi.org/10.1002/ejoc.202100409
01/2020

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
[Zinc(II)(Pyridine-Containing Ligand)] Complexes as Single-
Component Efficient Catalyst for Chemical Fixation of CO2 with
Epoxides
Matteo Cavalleri,[a] Nicola Panza,[a] Armando di Biase,[a] Giorgio Tseberlidis,[a,b] Silvia Rizzato,[a]
Giorgio Abbiati,[c] and Alessandro Caselli*[a]
Dedicated to Professor Franco Cozzi on the occasion of his 70th birthday. “Senatores boni viri, senatus mala bestia“
[a]
Mr. M. Cavalleri, Mr. N. Panza, Mr. A. di Biase, prof S. Rizzato and prof. A. Caselli
Department of Chemistry
Università degli Studi di Milano and CNR-SCITEC
via Golgi 19, 20133 Milano, Italy
E-mail: alessandro.caselli@unimi.it
[b]
[c]
Dr. G. Tseberlidis
Department of Materials Science and Solar Energy Research Center (MIB-SOLAR)
University of Milano-Bicocca
via Cozzi 55, 20125 Milano, Italy
Prof. G. Abbiati
Dipartimento di Scienze Farmaceutiche, Sezione di Chimica Generale e Organica “A. Marchesini”
Università degli Studi di Milano
via Venezian 21, 20133 Milano, Italy
Supporting information for this article is given via a link at the end of the document.((Please delete this text if not appropriate))
Abstract: The reaction between epoxides and CO2 to yield cyclic
carbonates is efficiently promoted under solvent-free and mild
reaction conditions (0.5 mol% catalyst loading, 0.8 MPa, 125 °C) by
zinc(II) complexes of pyridine containing macrocyclic ligands (Pc-L
pyridinophanes). The zinc complexes have been fully characterised,
including X-ray structural determination. The [Zn(II)X(Pc-L)]X
complexes showed good solubility in several polar solvents, including
cyclic carbonates. The scope of the reaction under solvent-free
conditions has been studied and good to quantitative conversions with
excellent selectivities have been obtained starting from terminal
epoxides. When solvent-free conditions were not possible (solid
epoxides or low solubility of the catalyst in the oxirane) the use of
cyclic carbonates as solvents has been investigated with success.
The remarkable stability of the catalytic system has been
demonstrated by a series of consecutive runs.
stable and difficult to activate and use. Therefore, efficient CO2
activation appears to be still a highly coveted target.
The synthesis of cyclic carbonates from carbon dioxide and
epoxides is a 100% atom economical reaction and an attractive
pathway for CO2 utilisation;[3] it is a topic of great relevance in this
field especially because cyclic carbonates have
a great
applicability in industrial processes.[4] Cyclic carbonates are
mainly employed as aprotic polar solvents, electrolytes in
secondary batteries, monomers for polycarbonate based
polymers and intermediates of fine chemicals.[5]
In the last twenty years, to overcome the limitations inherent to
homogeneous catalysts, such as stability and recovery issues,
heterogeneous catalysts such as modified zeolites,[6] metal
oxides,[7]
MOFs,[8]
supported
catalysts,[9]
mesoporous
materials,[10] silica,[11] cellulose[12] and poly-ionic liquids (PILs)[13]
have been investigated. Heterogeneous catalytic systems seem
to be an interesting option to develop a reusable and easily
separable catalyst. However, heterogeneous catalysts are
inherently more stable compared to homogeneous ones. This
implies that all these catalysts suffer from low activity, thus
requiring higher temperatures and pressures to display
comparable results, so limiting a wide applicability.[14]
Introduction
Carbon dioxide is the principal greenhouse gas, largely
recognized as responsible for global warming and related climate
changes, but it represents also an abundant C1 source. It is
naturally present in the atmosphere and it is produced on large
scale by many human activities, sensitively enhancing its natural
greenhouse effect. Limiting CO2 emissions can only stem the
problem but to solve it a circular economy based on carbon
dioxide should be pursued.[1] This means developing new and
efficient carbon capture and utilization technologies to sequester
the large quantities of CO2 produced in industrial processes and
to obtain high valuable products with low energy consumption.[2]
However, carbon dioxide is kinetically and thermodynamically
According to the literature, the synthesis of cyclic carbonates from
CO2 and epoxide consists in three main steps: (i) the activation of
the epoxide and with the subsequent ring-opening by an external
nucleophile; (ii) the CO2 insertion into the oxygen anion
intermediate; and (iii) the ring-closing reaction.[15] Because CO2 is
a thermodynamically stable molecule, for an efficient coupling
reaction to give cyclic carbonates the use of co-catalysts is
needed in the ring opening of the epoxide and in reducing the
activation energy of the CO2 conversion.[16]
1
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
Considering environmental compatibility and the high-efficiency
catalytic conversion of CO2, there is the strong need to develop
green catalysts capable of high conversions and selectivities,
especially under solvent-free conditions.[17]
the structure in solution, with two signals for each couple of
equivalent methylene groups.
The development of new catalysts for a so crucial reaction, able
to work at mild process conditions with good performance, that
are also reusable and easily separable, is therefore of high
industrial interest.
We have recently reported that iron(III) complexes of pyridine
based 12 membered tetraaza macrocyclic ligands (Pc-L) are
efficient catalyst for the epoxidation of alkenes with H2O2 as the
terminal oxidant and we have also disclosed that under certain
conditions they are also suitable catalysts for the epoxide
activation towards ring-opening with a nucleophile.[18]
Scheme 2. Synthetic route used to obtain zinc complexes 2a-c.
As a continuation of our studies on the catalytic activity of
pyridinophane type macrocyclic ligands,[18-19] we envisioned that
Zn(II) complexes would be perfect candidates to promote a clean
conversion of epoxides and CO2 in cyclic carbonates under mild
reaction conditions and without the addition of any co-catalyst
(Scheme 1). Herein we disclose our findings.
It should be noted that by changing from chlorine to bromine, the
nature of the counteranion affected the NMR spectra in DMSO-d6
only to a minor extent, whilst this is not the case when the bulkier
iodine is used, indicating that a slight different geometry of the
donor atoms in the macrocyclic skeleton is imposed by steric
crowding. The MS (ESI) analyses clearly showed that in all
complexes only one halogen was coordinated to the metal centre,
confirming the saline nature of the complexes (see Supporting
Information, page S3).
For all complexes, 2a, 2b and 2c, single crystals suitable for X-
ray crystallography were grown. Although the refinement left no
doubt about the validity of the structural model of complex 2c, it
was not possible to obtain satisfactory R values and as a
consequence the full crystal structure determination for 2c is not
reported here. Therefore, structural details and geometrical
parameters will be given for complexes 2a and 2b (Figures 1 and
2, respectively). The overall solid state structures of compounds
2a, 2b and 2c are different. However, they share some features
like the same molecular geometry for the cationic complex
[ZnII(X)(Pc-L)]+ (with X = Cl, Br, I). Four nitrogen atoms of the
macrocycle and a monodentate X- ligand form a distorted square
pyramidal environment around the Zn(II) ion, while an additional
X- anion balances the positive charge in the second coordination
sphere. The macrocycle adopts a cis-folded (+++) conformation
with the metal atom that is displaced from the N4 macrocyclic
cavity.[21] This arrangement and all the structural parameters are
comparable to those of similar complexes which have been
already reported in the literature.[22] For both complexes 2a and
2b, the shortest Zn-N bond length is that with the N3 donor atom
(2.053 Å for complex 2a and 2.050 Å for complex 2b). The Zn-X
bond distance varies consistently with the X- anion size (2.2139
Å for Zn-Cl and 2.3516 Å for Zn-Br).
To better describe the geometry of the [ZnII(X)(Pc-L)]+ five-
coordinated system, the τ-value was calculated. τ = (β-α)/60,
where β and α (°) are the largest angles in the system, those that
define the basal plane of a square-pyramid. The τ-value was first
introduced by Addison and colleagues[23] in order to have a
quantitative esteem of the deviation from ideal square-pyramidal
(τ = 0) or trigonal-bipyramidal geometry (τ = 1). In the current
study, β and α refer to N2-Zn-N4 and X-Zn-N1 bond angles
respectively and N3 represents the axial ligand. Considering that
both complexes under investigation lie in the continuum between
the two ideal geometries, it appears that while complex 2b tends
to approach closely the square-pyramidal (τ = 0.26), complex 2a
is slightly more distorted away from it (τ = 0.35).
Scheme 1. Cycloaddition of CO2 and epoxides catalysed by [ZnII(X)(Pc-L)]X
complexes reported in this work.
Results and Discussion
Preparation of the zinc complexes. We recently reported the
straightforward synthesis of macrocycle 1[20] in good overall yield
by treatment of N-tritosyl-diethylenetriamine with pyridine-2,6-
diylbis(methylene) dimethanesulfonate, followed by the hydrolysis
in strong mineral acids of the tosyl protecting groups.[19a] Zinc
complexes 2a-c were obtained by slowly adding the metal salt to
a stirred dichloromethane (DCM) solution of the macrocyclic
ligand (Scheme 2). A small defect of the zinc salt was used to
avoid the presence of free metal salt in the final product, being the
excess of ligand easily removable by simple washing of the
crystalline metal complex with DCM. The choice of the different
halogen anions was made in order to assess their different
reactivity in the nucleophilic ring opening of the epoxide in the
catalytic tests.
The metal complexes showed a good solubility in different polar
media. They were fully characterized by NMR spectroscopy, MS
spectrometry and elemental analyses. All experimental evidences
1
are in perfect agreement with the proposed structures. H NMR
data in DMSO-d6 are consistent with an apparent Cs symmetry of
2
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
absence of TBAX salts as co-catalysts and results are
summarised in Table 1.
Table 1. Screening of the [Zn(II)(X)(Pc-L)]X, 2a-c, catalysed cycloaddition of
CO2 to styrene epoxide[a]
Entry
Catalyst
TBAX
Selectivity[b]
(%)
Conversion[b]
(%)
TOF[c]
(h-1)
Figure 1. Molecular structure of the complex [Zn(II)(Cl)(Pc-L)]Cl·1.5H2O, 2a
(50% probability thermal ellipsoids). Selected bond lengths (Å): ): Zn-N1
2.063(1), Zn-N2 2.214(2), Zn-N3 2.053(1), Zn-N4 2.220(1), Zn-Cl1 2.2139(5).
Water molecules have been omitted for clarity. Full crystal structure data of 2a
are reported in the Supporting Information.
1
2
3
4
5
6
7
8
9
2a
-
TBACl
TBACl
-
95
90
11
99
97
12
2a
2b
-
92
78
19.5
>12
>12
25
TBAB
TBAB
-
>99
91
>99
>99
>99
96
2b
2c
-
>99
93
TBAI
TBAI
-
12
98
95
12
2c
97
97
24
[a] Reaction conditions: neat, 250 L of 3a (2.19 mmol), [Zn] = 1 mol%, TBAX
= 2 mol%, at 125 °C; P(CO2) = 0.8 MPa; reaction time = 4 h. [b] Conversions
and selectivities determined by 1H NMR using 1,3,5-trimethylbenzene as
internal standard. [c] Turnover frequency (mol3a(converted)·molcat-1·reaction time-1).
As it can be seen from data reported in the table and according to
literature,[25] TBAX alone can efficiently promote the cycloaddition
reaction, with only slight differences depending on the nature of
the halide. Under the selected reaction conditions, TBAB is the
most active catalyst (>99% of conversion, TOF >12 h-1, entry 5,
Table 1), whilst TBACl is the most selective (99% selectivity, entry
2 Table 1). It should be pointed out, however, that at lower
catalytic loadings, quaternary ammonium salts are less efficient
and with 1 mol% of TBAB the observed conversion was only 60%
(Table S1). Moreover, two problems arise with the use of these
reagents. First, they are very hygroscopic, so the reactions need
strict anhydrous conditions. Second, the presence of organic
counterparts make trickier the purification step, which requires
demanding column chromatography. Although complex 2a alone
is a competent catalyst, it demonstrated to be less active than
TBACl (compare entries 3 and 2, Table 1) and even when
employed in combination with TBACl, only moderate beneficial
effect has been observed, mainly in terms of conversion (compare
entry 1 and 3, Table 1).[26] Better results were obtained with
complex 2c, where very similar conversion and selectivity were
observed (entries 8 and 9, Table 1). Conversely, the use of the
binary catalytic system seemed to decrease slightly the selectivity
(entry 7, Table 1). Better results were observed with complex 2b,
whose presence in combination with TBAB gave the best
selectivity and conversion (entry 4, Table 1). Noteworthy, complex
2b alone gave the best TOF observed (25 h-1), with quantitative
conversion and full selectivity (entry 6, Table 1). It should be
pointed out that in the absence of the ammonium salt the styrene
carbonate 4a was easily isolated in almost quantitative yield
Figure 2. Molecular structure of the complex [Zn(II)(Br)(Pc-L)]Br·H2O, 2b (50%
probability thermal ellipsoids). Selected bond lengths (Å): Zn-N1 2.058(2), Zn-
N2 2.264(3), Zn-N3 2.050(3), Zn-N4 2.181(3), Zn-Br1 2.3516(5). The water
molecule has been omitted for clarity. Full crystal structure data of 2b are
reported in the Supporting Information.
Optimization of the zinc-catalysed cyclic carbonate
synthesis. At the outset, we decided to compare the catalytic
activity of our zinc complexes 2a-c in the cycloaddition reaction of
CO2 with styrene oxide 3a with that of quaternary ammonium
halides, which are well known to exhibit good activities in carbon
dioxide fixation with epoxide.[15] Very efficient binary catalytic
systems composed by Zn complexes and ammonium salts have
been reported and a detailed DFT study of the mechanism of the
cycloaddition reaction of CO2 to epoxides by the
Zn(salphen)/TBAX (TBAX = tetra-n-butylammonium halide) has
been reported by Bo and co-workers.[24] The presence of both
Lewis acid and nucleophile in the binary catalytic system makes
the ring-opening process, which is often considered the rate-
determining step, less energetically demanding. Our
consideration was that, due to their coordination geometry, metal-
complexes 2a-c were perfectly suited to act as bifunctional
systems, possessing an outer sphere nucleophilic halogen anion
and a free coordination site on the zinc. To verify this hypothesis
we tested all metal complexes both in the presence and in the
3
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
(99%) and excellent purity by simple filtration followed by
precipitation with n-hexane. No reaction was observed when
using ligand 1 as the catalyst, nor with ZnCl2 alone (Table S1).
Encouraged by these preliminary results, we decided to optimize
the reaction conditions with complexes 2b and 2c in order to
maintain high conversions and selectivities under milder reaction
conditions. First we decreased the catalyst loading, maintaining
the CO2 pressure of 0.8 MPa, the temperature at 125 °C and the
reaction time set at three hours (for full experimental details see
Supporting Information, page ). Results are summarised in Figure
3.
Table 2. Dependence of the temperature in [Zn(II)(X)(Pc-L)]X, 2b, catalysed
cycloaddition of CO2 to styrene epoxide[a]
Entry
T (°C)
Selectivity[b]
(%)
Conversion[b]
(%)
TOF[c]
(h-1)
1
100
75
61
10
97
82
30
10
58
39
15
5
2
3[d]
4[d]
100
75
29
19.5
[a] Reaction conditions: neat, 250 L of 3a (2.19 mmol), 2b = 0.5 mol%; P(CO2)
= 0.8 MPa; reaction time = 4 h. [b] Conversions and selectivities determined by
1H NMR using 1,3,5-trimethylbenzene as internal standard. [c] Turnover
frequency (mol3a(converted)·molcat-1·reaction time-1). [d] The catalyst was
previously dissolved in 250 µL of dimethyl sulfoxide.
100
90
80
70
Since the best results in terms of selectivity were always obtained
with the cheaper bromide complex, we decided to complete the
optimization of the reaction conditions only using complex 2b. We
next looked at the reaction outcome at different times under fixed
catalyst loading (1 mol%), temperature (125 °C) and CO2
pressure (0.8 MPa). Then we compared these results with those
obtained under otherwise identical conditions but using 2 mol% of
TBAB instead of 2b as catalyst (Figure 4). We were pleased to
find that the Zn complex 2b alone outperforms the quaternary
ammonium salt by being more active and more selective.
Conversion 2b
60
Selectivity 2b
50
Conversion 2c
40
Selectivity 2c
30
20
10
0
0,1
0,2
0,3
0,4
0,5
0,6
0,7
0,8
0,9
1
Catalyst loading (mol%)
100
90
Figure 3. Conversion vs selectivity at different catalyst loading with complexes
2b and 2c. Reaction conditions: neat, 250 L of 3a (2.19 mmol), at 125 °C;
P(CO2) = 0.8 MPa; reaction time = 3h.
80
70
60
Conversion 2b
The selectivity does not seem to be affected by the catalyst
loading while the conversion starts to suffer below 0.5 mol%.
Complex 2c maintains a good activity even at very low catalyst
loading and with a remarkable TOF of 277 mol3a·molcat-1·reaction
time-1 (83% conversion with 0.1 mol% catalyst loading in 3h).
Despite the catalyst loading, higher selectivities were always
observed with complex 2b. With this complex, in all cases, the
desired cyclic carbonate 4a was obtained as the only reaction
product. Noteworthy, an almost quantitative conversion (99%)
with complete selectivity (>99%) towards 4a was obtained with
0.5 mol% loading of the [Zn(II)(Br)(Pc-L)]Br, 2b, complex.
We next monitored the effect of the temperature and, especially
with complex 2b, we observed a dramatic drop in the conversion
at lower temperatures (Table 2). We supposed that this drop in
conversion was most probably due to the lower solubility of the
metal complexes in the neat epoxide 3a at lower temperatures.
To overcome this problem, we added 250 l of DMSO to the
reaction mixture and, as expected, even at 75 °C a 39%
conversion with good selectivity for compound 4a was obtained
(entry 4, Table 2).
50
Selectivity 2b
40
Conversion TBAB
30
Selectivity TBAB
20
10
0
0
1
2
2,5
3
4
Reaction time (h)
Figure 4. Conversion vs selectivity at different reaction times with complexes
2b and TBAB. Reaction conditions: neat, 250 L of 3a (2.19 mmol),
[Zn(II)(Br)(Pc-L)]Br, 2b (1 mol%) or TBAB (2 mol%), at 125 °C.
Finally, we investigated the effect of CO2 pressure on conversion
and selectivity, and we disclosed that the catalytic activity was
retained also at very low pressures (Figure 5). In fact, even
charging the autoclave under CO2 at atmospheric pressure (0.101
MPa) at 125 °C, we observed a moderate conversion (31%) in
just 3 h, with a good selectivity (77%). It should be noted that, at
such low conversion the errors we introduce in measuring the
yield in 4a is quite high. In fact, in the 1H NMR we just observe the
starting epoxide and the target cyclic carbonate.
4
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
100
90
80
70
60
50
40
30
20
10
0
Table 3. Substrate scope of the [Zn(II)(Br)(Pc-L)]Br, 2b, catalysed cycloaddition
of CO2 to epoxides[a]
Conversion
Selectivity
Selectivity[b] Conversion[b] TOF[c]
Entry
Substrate
(%)
(%)
(h-1)
1
2
>99
>99
>99
>99
>66
0,1
0,3
0,5
0,7
0,9
CO2 pressure (MPa)
>66
Figure 5. Conversion vs selectivity at different CO2 pressures. Reaction
conditions: neat, 250 L of 3a (2.19 mmol), [Zn(II)(Br)(Pc-L)]Br, 2b (1 mol%), at
125 °C; reaction time = 3h.
3
4[d]
5
74
>99
19
96
15
94
-
>99
>99
42
>66
>66
28
65
17
66
-
Scope of 2b-catalysed cyclic carbonates synthesis. With
these results in our hand, we decided to study the scope of the
reaction by using complex 2b under the optimised reaction
conditions (0.5 mol% of the zinc complex at 125 °C, P(CO2) = 0.8
MPa; reaction time = 3h). Whenever this was possible, reactions
were carried out in neat epoxide (250 L); on the other hand, in
the case of solid epoxides (stilbene oxide, 3l) or in cases where
we found a very low solubility of 2b in the reaction media, 125 L
of propylene carbonate (PC) were added (method B, see
Supporting Information for details). We choose PC due to its high
polarity and non-toxic nature, and because we previously notice
that the complex solubility was improved during the course of the
6[d]
7
97
26
8[d]
9
99
n.d.
n.d.
0
10[d]
11
12[d]
13
14[d]
-
-
-
-
40
99
15
10
17
53
reaction, when the cyclic carbonate is formed. Table
3
25
summarizes the results obtained. Conversion and selectivities
1
>99
80
reported have been calculated by H NMR of the crude reaction
mixture using 1,3,5-trimethylbenzene as internal standard, but
products can be easily isolated after the reaction and when
necessary purified by column chromatography (see Supporting
Information). An excellent correlation between isolated and 1H
NMR calculated yield has been found in all cases.
15
-
7
5
16[d]
17
-
0
-
Reactive epoxide such as (+/-)-epichlorohydrin 3b, due to
electronegative effect of its substituent that facilitated the
nucleophilic attack to open the epoxide ring,[27] gave quantitative
yield of cyclic carbonate (entry 2, Table 3). On the other hand, the
low solubility of complex 2b in alkyl epoxides even at high
temperatures caused a drop in the reaction yield, whose effect is
more pronounced on increasing chain length, as exemplified by
the difference between propylene oxide and 1,2-hexene oxide
(compare entry 3 with entry 7 in Table 3). However, as cited
above, complex 2b is very soluble in more polar cyclic carbonates
even at room temperature. Thus, if the metal complex was
dissolved previously in PC, high conversions with excellent
selectivities were restored even with alkyl epoxides (entries 4, 6
and 8, Table 3).
>99
>99
98
65
18
71
47
[a] Reaction conditions: neat, 250 L of 3, 2b = 0.5% at 125 °C; P(CO2) = 0.8
MPa; reaction time = 3h. [b] Conversions and selectivities determined by 1H
NMR using 1,3,5-trimethylbenzene as internal standard. [c] Turnover frequency
(mol3a·molcat-1·reaction time-1). [d] Complex 2b was dissolved in 125 L of
propylene carbonate.
Worth to note, when the reaction product is the same as the cyclic
carbonate employed as the solvent, as is the case of entry 4,
Table 3, the reaction product can be isolated in almost
quantitative yield and very high purity by addition of chloroform to
the reaction mixture because complex 2b is insoluble in
chlorinated solvents and precipitates quantitatively. A simple
5
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
filtration, followed by evaporation of chloroform under reduced
pressure yielded pure 4c in almost quantitative yield.
styrene epoxide and charging again the reaction vessel under
CO2 pressure. Results of four catalytic cycles are summarised in
Table 4.
On the other hand, the reactions of 1,1- and 1,2-disubstituted
epoxides failed to give the desired cyclic carbonates (entries 9-
13, Table 3). Actually, in the case of 1,1-dimethyloxirane (1,2-
epoxy-2-methyl propane), 3f, we were not able to determine the
conversion of the starting product because we noticed a high loss
in weight of the sample even when pre-dissolving the catalyst in
propylene carbonate. This result seemed to be due to the high
volatility of 1,1-dimethyloxirane that evaporated, at least in part,
under the reaction conditions (entries 9 and 10, Table 3). A very
low conversion (15%), and only prior dissolution of the catalyst in
propylene carbonate, was observed in the case of cyclohexene
oxide (entry 12, Table 3). However, cyclic carbonate 4g was
obtained with a modest selectivity (40%) beside some unidentified
by-products. The strong preference for the cycloaddition reaction
of CO2 towards terminal vs. internal double bonds was further
demonstrated in the case of 4-vinylcyclohexene dioxide 3h,
where only the terminal exocyclic epoxide highlighted in red
reacted to give in good yields 4-(7-oxabicyclo[4.1.0]heptan-3-yl)-
1,3-dioxolan-2-one, 4h (entries 13 and 14, Table 3).
Table 4. Recyclability of the [Zn(II)(Br)(Pc-L)]Br, 2b, catalysed cycloaddition of
CO2 to styrene epoxide[a]
Run
Selectivity[b]
(%)
Conversion[b]
(%)
TOF[c] (h-1)
TON[d]
1
>99
>99
n.d.
>99
>99
95
>66
63
200
390
n.d.
756
2
3[e]
4
n.d.
88
n.d.
59
No reaction was observed with sterically hindered internal
epoxides such as cis-limonene 1,2-oxide, 3i, or trans stilbene, 3l
(entries 15 and 16 Table 3).
Unsaturated cyclic carbonates are currently gaining increasing
attention both from academic and industrial communities as
reactive monomers for copolymerization reactions, since they
allow for a good control in terms of spontaneous crosslinking
reactions.[28] When we reacted allyl glycidyl ether, 3m, we were
pleased to find an almost quantitative formation of (2-oxo-1,3-
dioxolan-4-yl)methyl vinyl ether, 4m (entry 17, Table 3).
To demonstrate further the tolerability of electron-donating
substituents, eugenol epoxide 3n was converted in a satisfying
71% yield into the corresponding cyclic carbonate (entry 18, Table
3). To the best of our knowledge, cyclic carbonate 4n was never
characterized before, but it can be considered as an interesting
building block for eugenol-based non-isocyanate polyurethane
from renewable resources.[29]
[a] Reaction conditions: neat, 250 L of 3a, 2b = 0.5 mol% at 125 °C; P(CO2) =
0.8 MPa; reaction time = 3h. At each run 250 L of 3a were added. [b]
Conversions and selectivities determined by 1H NMR using 1,3,5-
-
trimethylbenzene as internal standard. [c] Turnover frequency (mol3a·molcat
1·reaction time-1). [d] Moles of styrene epoxide converted per mole of catalyst.
[e] Due to the experimental setup (see Supplementary Information), we were
not able to determine the conversion of the 3rd run without altering the yield of
the 4th.
Only at the fourth run, we observed a decrease in catalytic activity,
but again this result might be due to a less efficient stirring due to
the increased solvent volume or even to a higher dilution of the
catalyst. In any case, a remarkable TON of 756 mol of styrene
oxide per mole of catalyst was achieved.
Conclusion
In summary, we have shown that well-defined pentacoordinated
Zn(II) complexes of basic Pc-L macrocyclic ligands are efficient
catalyst, without the need of any Lewis base as co-catalyst, for
the cycloaddition of CO2 to epoxide under solvent-free conditions
and under mild reaction conditions. The bifunctional nature of the
catalyst is assured by the peculiar geometry imposed by the
tetracoordinate ligand to the cationic Zn(II) complex. The best
catalytic performances have been obtained by using complex 2b,
which can be conveniently prepared in high yields starting from
cheap and available reagents. The metal complexes are easier to
handle and far less hygroscopic than quaternary ammonium salts.
Moreover, at the end of the reaction, they can be simply filtered
off, thus allowing for the isolation of cyclic carbonates in high
yields and excellent purities by simple recrystallization.
Remarkably, quantitative conversions of terminal epoxides with
full selectivity towards the cyclic carbonate have been obtained
without the need of any co-catalyst at 0.8 MPa in 3h at 125 °C,
with a 0.5 mol% of catalyst loading.
Catalyst recovery and recycle. The stability and recyclability of
a catalyst are of fundamental importance for practical application.
In order to gain insights to the catalyst stability and its potential
reuse, two different tests were conducted under the optimised
reaction conditions (2b = 0.5 mol% at 125 °C; P(CO2) = 0.8 MPa;
reaction time = 3 h). As we already pointed out, at the end of the
reaction, dilution with chlorinated solvent allows for the
precipitation of the zinc complex that can be collected by filtration
as a white solid. In order to set up a more reproducible procedure,
we performed the reaction on 2.5 mL of styrene oxide, 3a. We
noticed a scale up effect, probably due to a less efficient stirring
in the reaction vessel, and the conversion was only of 86%, but
again the only product formed was styrene carbonate 4a (>99%
selectivity). At the end of the catalytic reaction, 4 mL of chloroform
were added and the solid metal complex 2b (76%) was recovered
by filtration followed by a double washing with 2 mL of chloroform
(see Supporting Information for details). The 1H NMR spectrum of
the recovered solid is perfectly consistent with the starting
complex [Zn(II)(Br)(Pc-L)]Br, 2b, while the elemental analysis
showed a slightly higher content of carbon than expected.
Bearing in mind that our metal complex is more soluble in cyclic
carbonate than in the starting epoxide, we next decided to restore
a catalytic cycle without isolating the catalyst but just adding fresh
Finally, the reusability of the metal complex has been assessed
by restoring the catalytic cycle for four consecutive runs. Based
on these results, we think that among the several homogeneous
catalytic systems reported in the last years for the synthesis of
cyclic carbonates by cycloaddition of CO2 to epoxides, the
6
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
from an aqueous solution at room temperature while prismatic colourless
crystals of the compound 2b were obtained from slow evaporation of
mother liquor after purification process. Single crystal X-ray diffraction
experiments were performed on a Bruker Smart APEX II diffractometer
with graphite monochromated Mo-Kα radiation (λ = 0.71073 Å) by the ω-
scan method, within the limits 3.4° < 2θ < 52.7° < θ < 22.7 for 2a and 3.9°
< 2θ < 52.7° for 2b. The frames were integrated and corrected for Lorentz-
polarization effects with the Bruker SAINT software package1073 Å) by
the ω-scan method, within the limits 3.4° < 2θ < 52.7° < θ < 22.7 for 2a
and 3.9° < 2θ < 52.7° for 2b. The frames were integrated and corrected for
Lorentz-polarization effects with the Bruker SAINT software package.[30]
The intensity data were then corrected for absorption by using SADABS.[31]
No decay correction was applied. The structure of both the compounds in
this study was solved by direct methods (SIR-97)[32] and refined by iterative
cycles of full-matrix least-squares on Fo2 and ΔF synthesis with SHELXL-
97[33] within the WinGX interface.[34]
pyridine containing macrocyclic Zn(II) halide complexes represent
a considerable case of study and surely deserve the development
in further applications.
Experimental Section
All the reactions that involved the use of reagents sensitive to oxygen or
to moisture were carried out under an inert atmosphere employing
standard Schlenk techniques. All chemicals and solvents were
commercially available and used as received except where specified.
NMR spectra were recorded at room temperature with 300 or 400 MHz
spectrometers. Chemical shifts (δ) are expressed in ppm and they are
reported relative to TMS. The coupling constants (J) are expressed in hertz
(Hz). The 1H NMR signals of the compounds described in the following
were attributed by 2D NMR techniques. Assignments of the resonances in
13C NMR were made by using the APT pulse sequence, HSQC and HMBC
techniques. Catalytic tests were analysed by 1H NMR spectroscopy: the
+
-
̅
Crystal data for 2a: (C11H18ClN4Zn) (Cl) ·3/2H2O, triclinic space group 푃1
(n° 2), Z = 2, a = 7.8000(4), b = 8.9923(7), c = 12.1580(9) Å, α = 78.255(1),
β = 88.493(1), γ = 75.167(1)°, V = 806.79(1) Å. Dc = 1.517 g cm−3, F(000)
= 380, T = 298(2) K, μ(Mo-Kα) = 1.855 mm−1. Total number of reflections
1
reaction crude was diluted in chloroform and analysed by H NMR, after
the addition of an internal standard (ISTD). Low resolution MS spectra
were recorded with instruments equipped with ESI/ion trap sources. The
values are expressed as mass−charge ratio and the relative intensities of
the most significant peaks are shown in brackets. Elemental analyses were
recorded in the analytical laboratories of Università degli Studi di Milano.
Syntheses of the ligand and zinc complexes are described in the
experimental section. X-ray data collection for the crystal structure
determination was carried out by using Bruker Smart APEX II CCD
diffractometer with the Mo K radiation ( = 0.71073) at 296 K.
recorded, to θmax = 26.369°, was 6179 of which 3277 were unique (Rint
=
0.0098); 3144 were ‘observed’ with I > 2σ(I). Final R-values: wR2 = 0.0579
and R1 = 0.0229 for all data; R1 = 0.0216 for the ‘observed’ data. All atoms
are in general position. A water molecule (oxygen atom O2) is located very
̅
close to a cell vertex (site symmetry 1, Wyckoff letter a) and refined with
an occupancy of 0.5. A second water molecule is present that shows a
disorder over two positions (site symmetry 1, Wyckoff letter i) of its oxygen
atom. The two sites, O1A and O1B, share the same hydrogens and were
refined with a sof of 0.5. The hydrogen atoms of the ligand were placed in
geometrically calculated positions and then refined using a riding model
based on the positions of the parent atoms with Uiso = 1.2 Ueq. The
hydrogen atoms (H12 and H13) attached to O1A and O1B were located
from a difference Fourier map and refined isotropically, whereas it was not
possible to determine the position of the hydrogens bound to O2. Restrains
were applied to the distances between the oxygen and the hydrogen atoms
(H12 and H13) of the O1A and O1B water molecules. A restrain was
applied to the H12…H13 distance as well in order to obtain a reasonable
value for the H-O-H angle.
General procedure for the synthesis of Zn(II) complexes. A zinc (II)
salt (0.65 mmol) was slowly added to a solution of ligand 1 (0.1340 g, 0.75
mmol) in DCM (20,0 mL). The mixture was left to react at room temperature
for 1.5 hours. The reaction mixture was then filtered over a Büchner and
washed with DCM (5,0 mL x 2), with cold MeOH (few drops), and again
with DCM (5,0 mL). The product was recovered as a white powder and it
was dried under vacuum (see Supporting Information for details).
General catalytic procedures. Method A: In a 2.5 mL glass liner
equipped with a screw cap and glass wool, the catalyst (0.1, 0.5 or 1 mol%,
see Table captions) and the epoxide (250 µL) were added. The vessel was
transferred into a 250 mL stainless-steel autoclave; three vacuum-nitrogen
cycles were performed and CO2 was charged at room temperature (0.1,
0.2, 0.4 or 0.8 MPa). The autoclave was placed in a preheated oil bath (at
75 or 100 or 125 °C) and it was let to react under stirring (for 2 to 4 hours),
then it was cooled at room temperature in an ice bath and slowly vented.
The crude was treated with chloroform (700 µL) and a sample (150 µL)
+
-
̅
Crystal data for 2b: (C11H18BrN4Zn) (Br) ·H2O triclinic space group 푃1 (n°
2), Z = 2, a = 8.3278(8), b = 9.2464(9), c = 10.631(1) Å, α= 81.269(1), β =
89.239(1), γ = 89.991(1)°, V = 809.0(1) Å3.. Dc = 1.845 g cm−3, F(000) =
444, T = 298(2) K, μ(Mo-Kα) = 6.454 mm−1. Total number of reflections
recorded, to θmax = 26.368°, was 6583 of which 3317 were unique (Rint
=
0.0124); 2786 were ‘observed’ with I > 2σ(I). Final R-values: wR2 = 0.0744
and R1 = 0.0355 for all data; R1 = 0.0275 for the ‘observed’ data. All atoms
are in general position. The hydrogen atoms of the ligand were placed in
geometrically calculated positions and then refined using a riding model
based on the positions of the parent atoms with Uiso = 1.2 Ueq. The
hydrogen atoms of the water molecule (H1W and H2W) were located from
a difference Fourier map and refined isotropically. Their distance to the
oxygen atom O was restrained to be equal within a standard deviation of
0.02. Furthermore, the distance H1W…H2W was restrained to a target
value in order to retain a reasonable value for the H-O-H angle.
1
was analysed by H NMR spectroscopy by using 1,3,5-trimethylbenzene
(0.11 eq.) as the internal standard. Method B: In a 2.5 mL glass liner
equipped with a screw cap and glass wool, the catalyst (0.5 mol%) and the
epoxide (250 µL) were added and they were suspended in propylene
carbonate (125 µL) as the solvent. The vessel was transferred into a 250
mL stainless-steel autoclave; three vacuum-nitrogen cycles were
performed and CO2 was charged at room temperature (0.8 MPa). The
autoclave was placed in a preheated oil bath (125 °C) and let react under
stirring (for 2 to 4 hours), then it was cooled at room temperature in an ice
bath and it was slowly vented. The crude was treated with chloroform (700
µL) and a sample (150 µL) was analysed by 1H NMR spectroscopy by
using 1,3,5-trimethylbenzene (0.11 eq.) as the internal standard.
Full crystallographic data have been deposited with the Cambridge
Crystallographic Data Centre (CCDC No. 2064861 for complex 2a and
2064860 for complex 2b). A copy of the data can be obtained free of
charge on application to CCDC, 12 Union Road, Cambridge CB2 IEZ, UK
(Fax: +44 1223 336 033; e-mail: deposit@ccdc.cam.ac.uk).
Procedure for the isolation of styrene carbonate, 4a. At the end of the
reaction, after venting the autoclave, 700 L of DCM were added, the
reaction mixture was filtered then n-hexane (30,0 mL) was added and the
vessel was stored at 4 °C overnight. A white precipitate fell off that was
collected and dried under reduced pressure (99% yield).
Acknowledgements ((optional))
We thank the MUR-Italy (Ph.D. fellowship to N. P., A. di B. and M.
C.) and the University of Milan (PSR 2020 – financed project
Crystal structure determination. Prismatic colourless crystals of the
compound 2a suitable for X-ray analysis were grown by slow evaporation
7
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
[22] a) K. M. Lincoln, M. E. Offutt, T. D. Hayden, R. E. Saunders, K. N. Green,
Inorg. Chem. 2014, 53, 1406-1416; b) K. M. Lincoln, P. Gonzalez, T. E.
Richardson, D. A. Julovich, R. Saunders, J. W. Simpkins, K. N. Green,
Chem. Commun. 2013, 49, 2712-2714; c) N. W. a. B. Alcock, D. H. and
Liu, C. Y., CSD Communication 2006, CCDC 628572.
‘‘Catalytic strategies for the synthesis of high added-value
molecules from bio-based starting materials’’) for financial
support.
Keywords: carbon dioxide fixation • zinc complexes •
cycloaddition • cyclic carbonates • pyridine containing
macrocyclic ligands
[23] A. W. Addison, T. N. Rao, J. Reedijk, J. van Rijn, G. C. Verschoor, J.
Chem. Soc., Dalton Trans. 1984, 1349-1356.
[24] F. Castro-Gómez, G. Salassa, A. W. Kleij, C. Bo, Chem. Eur. J. 2013,
19, 6289-6298.
[25] S. Wang, J. Peng, H.-J. Yang, B. Ban, L. Wang, B. Lei, C.-Y. Guo, J. Hu,
J. Zhu, B. Han, J. Nanosci. Nanotechno. 2019, 19, 3263-3268.
[26] For considerations on TOF calculations in the CO2 reaction with epoxide
in the presence of catalysts and co-catalysts, see: V. Campisciano, C.
Calabrese, F. Giacalone, C. Aprile, P. Lo Meo, M. Gruttadauria, J. CO2
Util. 2020, 38, 132-140.
[1]
[2]
[3]
J. A. Martens, A. Bogaerts, N. De Kimpe, P. A. Jacobs, G. B. Marin, K.
Rabaey, M. Saeys, S. Verhelst, ChemSusChem 2017, 10, 1039-1055.
A. Modak, P. Bhanja, S. Dutta, B. Chowdhury, A. Bhaumik, Green Chem.
2020.
a) C. Claver, M. B. Yeamin, M. Reguero, A. M. Masdeu-Bultó, Green
Chem. 2020, 22, 7665-7706; b) L. Guo, R. Zhang, Y. Xiong, D. Chang,
H. Zhao, W. Zhang, W. Zheng, J. Chen, X. Wu, Molecules 2020, 25,
3627; c) F. Della Monica, A. Buonerba, C. Capacchione, Adv. Synth.
Catal. 2019, 361, 265-282.
[27] M. K. Leu, I. Vicente, J. A. Fernandes, I. de Pedro, J. Dupont, V. Sans,
P. Licence, A. Gual, I. Cano, Appl. Catal., B 2019, 245, 240-250.
[28] V. Besse, F. Camara, C. Voirin, R. Auvergne, S. Caillol, B. Boutevin,
Polym. Chem. 2013, 4, 4545-4561.
[4]
[5]
[6]
M. North, R. Pasquale, C. Young, Green Chem. 2010, 12, 1514-1539.
M. Aresta, A. Dibenedetto, E. Quaranta, J. Catal. 2016, 343, 2-45.
G.-Y. Hwang, R. Roshan, H.-S. Ryu, H.-M. Jeong, S. Ravi, M.-I. Kim, D.-
W. Park, J. CO2 Util. 2016, 15, 123-130.
[29] C. Cheng, Y. Li, X. Zhang, J. Li, Iran. Polym. J. 2017, 26, 821-831.
[30] Bruker AXS Inc. SAINT. Madison, WI, USA 2007.
[31] Bruker AXS Inc. SADABS Area-Detector Absorption Correction
Program. Madison, WI, USA 2001.
[7]
[8]
H. Yasuda, L.-N. He, T. Sakakura, J. Catal. 2002, 209, 547-550.
L. Liu, S.-M. Wang, Z.-B. Han, M. Ding, D.-Q. Yuan, H.-L. Jiang, Inorg.
Chem. 2016, 55, 3558-3565.
[32] A. Altomare, G. Cascarano, C. Giacovazzo, A. Guagliardi, M. C. Burla,
G. Polidori, M. Camalli, J. Appl. Crystallogr. 1994, 27, 435-436.
[33] G. Sheldrick, Acta Crystallogr. A 2015, 71, 3-8.
[9]
J. Sun, W. Cheng, W. Fan, Y. Wang, Z. Meng, S. Zhang, Catal. Today
2009, 148, 361-367.
[34] L. J. Farrugia, J. Appl. Crystallogr. 1999, 32, 837-838.
[10] W. Zhang, Q. Wang, H. Wu, P. Wu, M. He, Green Chem. 2014, 16, 4767-
4774.
[11] J.-Q. Wang, X.-D. Yue, F. Cai, L.-N. He, Catal. Commun. 2007, 8, 167-
172.
[12] K. R. Roshan, G. Mathai, J. Kim, J. Tharun, G.-A. Park, D.-W. Park,
Green Chem. 2012, 14, 2933-2940.
[13] J. Sun, S.-i. Fujita, M. Arai, J. Organomet. Chem. 2005, 690, 3490-3497.
[14] J. Byun, K. A. I. Zhang, ChemCatChem 2018, 10, 4610-4616.
[15] J.-Q. Wang, K. Dong, W.-G. Cheng, J. Sun, S.-J. Zhang, Catal. Sci. &
Technol. 2012, 2, 1480-1484.
[16] A. Kilic, B. Sobay, E. Aytar, R. Söylemez, Sustainable Energy & Fuels
2020, 4, 5682-5696.
[17] A. Monfared, R. Mohammadi, A. Hosseinian, S. Sarhandi, P. D.
Kheirollahi Nezhad, RSC Adv. 2019, 9, 3884-3899.
[18] G. Tseberlidis, L. Demonti, V. Pirovano, M. Scavini, S. Cappelli, S.
Rizzato, R. Vicente, A. Caselli, ChemCatChem 2019, 11, 4907-4915.
[19] a) N. Panza, A. di Biase, S. Rizzato, E. Gallo, G. Tseberlidis, A. Caselli,
Eur. J. Org. Chem. 2020, 2020, 6635-6644; b) G. Tseberlidis, A. Caselli,
R. Vicente, J. Organomet. Chem. 2017, 835, 1-5; c) G. Tseberlidis, M.
Dell'Acqua, D. Valcarenghi, E. Gallo, E. Rossi, G. Abbiati, A. Caselli,
RSC Adv. 2016, 6, 97404-97419; d) T. Pedrazzini, P. Pirovano, M.
Dell'Acqua, F. Ragaini, P. Illiano, P. Macchi, G. Abbiati, A. Caselli, Eur.
J. Inorg. Chem. 2015, 2015, 5089-5098; e) B. Castano, E. Gallo, D. J.
Cole-Hamilton, V. Dal Santo, R. Psaro, A. Caselli, Green Chem. 2014,
16, 3202-3209; f) M. Dell’Acqua, B. Castano, C. Cecchini, T. Pedrazzini,
V. Pirovano, E. Rossi, A. Caselli, G. Abbiati, J. Org. Chem. 2014, 79,
3494-3505; g) M. Trose, M. Dell’Acqua, T. Pedrazzini, V. Pirovano, E.
Gallo, E. Rossi, A. Caselli, G. Abbiati, J. Org. Chem. 2014, 79, 7311-
7320; h) B. Castano, S. Guidone, E. Gallo, F. Ragaini, N. Casati, P.
Macchi, M. Sisti, A. Caselli, Dalton Trans. 2013, 42, 2451-2462; i) B.
Castano, P. Zardi, Y. C. Honemann, A. Galarneau, E. Gallo, R. Psaro,
A. Caselli, V. Dal Santo, RSC Adv. 2013, 3, 22199-22205; j) B. Castano,
T. Pedrazzini, M. Sisti, E. Gallo, F. Ragaini, N. Casati, A. Caselli, Appl.
Organomet. Chem. 2011, 25, 824-829; k) A. Caselli, F. Cesana, E. Gallo,
N. Casati, P. Macchi, M. Sisti, G. Celentano, S. Cenini, Dalton Trans.
2008, 4202-4205.
[20] J. Serrano-Plana, A. Aguinaco, R. Belda, E. García-España, M. G.
Basallote, A. Company, M. Costas, Angew. Chem. Int. Ed. 2016, 55,
6310-6314.
[21] V. Felix, J. Costa, R. Delgado, M. G. B. Drew, M. T. Duarte, C. Resende,
J. Chem. Soc., Dalton Trans. 2001, 1462-1471.
8
This article is protected by copyright. All rights reserved.

10.1002/ejoc.202100409
European Journal of Organic Chemistry
FULL PAPER
Entry for the Table of Contents
Pentacoordinated [ZnII(X)(Pc-L)]X (with X = Cl, Br, I) complexes, which can be easily prepared in high yields starting from cheap and
available reagents, are efficient bifunctional catalyst for the cycloaddition of CO2 to epoxide under solvent-free and mild reaction
condition (0.8 MPa in 3h at 125 °C, with a 0.5 mol% of cat.). The reusability of the metal complex has been assessed by restoring the
catalytic cycle for four consecutive runs.
Institute and/or researcher Twitter usernames: Università degli Studi di Milano @LaStatale
9
This article is protected by copyright. All rights reserved.