Catalytic Conversion of Carbon Dioxide Using Binuclear Double-Stranded Helicates: Cyclic Carbonate from Epoxides and Diol

Abstract

The construction of sophisticated molecular architectures from chemical subunits requires careful selection of the spacers, precise synthetic strategies, and substantial efforts. Here, we report a series of binuclear double-stranded helicates synthesized from different combinations of pyridyl hydrazone-based multidentate ligands (H₂1, H₂2, H₂3) by increasing the methylene spacer and transition metals (Co, Ni, and Zn). The ligands H₂1 (N′1,N′3-bis((E)-pyridin-2-ylmethylene)malonohydrazide), H₂2 (N′1,N′4-bis((E)-pyridin-2-ylmethylene)succinohydrazide), and H₂3 (N′1,N′5-bis((E)-pyridin-2-ylmethylene)glutarohydrazide) and their respective complexes with Co, Ni, and Zn were obtained. Single-crystal X-ray diffraction studies of these binuclear metallohelicates confirm the double-stranded helical structure of the complexes derived from H₂2. The set of helicates Co-1, Co-2, and Co-3; Ni-1, Ni-2, and Ni-3; and Zn-1, Zn-2, and Zn-3 were investigated for its catalytic activity in the cyclic carbonate formation reaction. Intriguingly, among the synthesized catalyst, Co-1 was found to be better in terms of conversions with the calculated TOF (turnover frequency) of 128/h. The catalytic performance was significantly improved by adding 0.2 mmol of tetrabutylammonium bromide by achieving 76% conversion in 30 min, with the observed TOF of 15,934 h^–1/molecule and 7967 h^–1/Co center. The results obtained herein show that the double-stranded helicates are effective catalysts for converting both terminal and non-terminal epoxides into their corresponding cyclic carbonates. The striking feature of this catalytic protocol lies in demonstrating the catalytic activity for the conversion of diol to cyclic carbonate, and the detailed kinetic experiments tempted us to propose a tentative reaction mechanism for this conversion.

Introduction

The utilization of carbon dioxide is of scientific and societal interest due to its direct implications on economic and environmental aspects.¹⁻⁴ Though CO₂ is deemed to be a major contributor to climate change, on the contrary, it also serves as an advantageous and inexpensive C1 source for a diverse array of chemical synthesis.^5,6 However, CO₂ being a fully oxidized molecule, its inherent thermodynamic stability necessitates the need for a catalyst to activate and explore its potential applications.⁷ From the viewpoint of atom economy and low energy requirements, utilization of CO₂ for the synthesis of cyclic carbonate has attracted great attention from industry and academics.

Among the pioneering studies that utilize CO₂, the formation of cyclic carbonates has received growing interest due to its wide spectrum of applications ranging from electrolytes for lithium-ion batteries to synthetic targets for versatile materials. The reaction is 100% atom economical and deserves a wide variety of industrial applications.⁸ Attempts have been made to identify a suitable catalytic system, including homogeneous and heterogeneous catalysts⁹⁻¹¹ such as ammonium salts,¹⁰ metal complexes,¹² functional polymers,¹³ supported nanoparticles,¹⁴ metal organic frameworks,¹⁵ and ion exchange resins. Considering the reaction mechanism involved in the reaction of CO₂ and epoxide, it requires a bifunctional catalyst for higher cyclic carbonate yields.

On the other hand, mimicking the unparalleled efficiency of multiple metal-centered enzyme catalysis has motivated chemists to design and develop artificial helical architectures for its potential applications.¹⁶ Helicates, the fascinating structural motif, are well-explored for its application in biological activity due to its structural resemblance with DNA.¹⁷⁻¹⁹ Ironically, these interesting tunable chemical strands are rarely deployed as catalysts for organic transformations.²⁰ Among the best known catalytic systems,²¹ aluminum–triphenolate²² with a TOF up to 36,000/h and magnesium–porphyrin²³ that showed TOF = 46,000/h are reported to be the highest with excellent conversion. These reports identified and dictated that the catalyst with suitable Lewis acidic sites (Al/Mg) are in a multinuclear form and responsible for the efficient catalytic activity. Helicates being known for its multimetallic center and our long-standing interest in helicate chemistry^24,25 together prompted us to consider that this strategy would be amenable to achieve cyclic carbonate with high yield and selectivity. Herein, we demonstrate the catalytic activity of binuclear helicates with different Lewis acidic centers for the synthesis of cyclic carbonate. Recently, biodegradable diols are identified as an alternative to epoxide for the synthesis of cyclic carbonate. However, the formation of the water molecule during the progress of the reaction promoting reverse reaction makes it highly challenging.²⁶ The less-reactive hydroxyl group may be activated by adapting a harsh reaction condition.

With this understanding, the present manuscript demonstrates the catalytic activity of a series of binuclear double-stranded helicates for two important reactions: (i) epoxide and CO₂ to cyclic carbonate (ii) in single-phase conversion of diol to cyclic carbonate, and subsequently, the by-product water is used for hydration reaction in the conversion of 2-cyanopyridine to 2-aminopyridine.

The ligands H₂1 (N′1,N′3-bis((E)-pyridin-2-ylmethylene)malonohydrazide) H₂2 (N′1,N′4-bis((E)-pyridin-2-ylmethylene)succinohydrazide), and H₂3 (N′1,N′5-bis((E)-pyridin-2-ylmethylene)glutarohydrazide) and their respective complexes Co-1, Co-2, and Co-3; Ni-1, Ni-2, and Ni-3; and Zn-1, Zn-2, and Zn-3 were obtained as shown in Figure 1. The multidentate ligands (H₂1-H₂3) and their cobalt helicate formation are already well-established, whereas the helicate formation nickel and zinc and their catalytic application are included in this manuscript.

Figure 1.

(a) Structure of multidentate ligands (H₂1 to H₂3). (b) Structure of binuclear double-stranded metallohelicates with Co, Ni, and Zn.

Results and Discussion

Pyridylhydrazone-based multidentate ligands H₂1–H₂3 (Figure 1a) and their respective binuclear double-stranded Co(III), Ni(II), and Zn(II) helicates were synthesized (Figure 1b) as per our earlier-reported procedure and are found matching with the analytical data reported therein.^24,25 The multidentate ligands H₂1, H₂2, and H₂3 are obtained by varying malonate, succinate, and glutarate spacers with one, two, and three −CH₂– groups, respectively, as shown above. In addition to the Co(III) recently reported by us,^24,25 we have extended the synthesis to Ni(II) and Zn(II) helicates adapting a similar synthetic protocol (Figures S1–S9). Meanwhile, in the cases of helicates Co-1, Co-2, and Co-3 complexes, the cobalt metal were found to exist in the +3 state, whereas in the nickel and zinc complexes, the respective metal are found to be in the +2 oxidation state.

In the absorption spectra (Figure 2), the three well-resolved metal-centered transitions for Ni-1, Ni-2, Ni-3 at 1364, 1423, and 1390 and 1130, 1140, and 1154 nm, respectively, correspond to ³A_2g → ³T_2g, and the third electronic transitions at 872, 872, and 868 nm correspond to ³A_2g → ³T_1g. These transitions match the distorted octahedral geometry of related Ni(II) complexes.²⁷ The absorption spectra for the corresponding Zn(II) helicates show a similar pattern; the ligand-centered transitions occurred at 290 nm. The transition at 365 nm represents σ → π*, corresponding to ketonic oxygen to metal, while the metal-centered region is flat.

Figure 2.

UV–vis spectra for (a) Ni(II) helicates and (b) Zn(II) helicates recorded in water

X-ray Structures for Ligand and Helicates

The ligand H₂1 was crystallized by slow evaporation of its methanolic solution at room temperature. The crystal structure was resolved in the monoclinic system of the P2₁/c space group with four molecules in the unit cell. The ligand possesses “S” conformation, showing the twist angle of the alkyl −CH₂ group ∠C9C8C7 = 108.10° (Figure S10). The single-crystal analyses for cobalt helicates are reported.²⁴ The X-ray structures for both Co-2 and Co-3 are existing in meso conformation and ionic with two anions in the lattice. In addition, all the labile −NH’s are in a deprotonated form, and the Co²⁺ adopted is oxidized to Co³⁺, similar to our earlier observation.

The suitable crystals for Ni-2 are obtained by diffusing the THF into the aqueous solution of the sample (Figure 3). The molecule is found to be discretely ionic [Ni₂(H₂2)₂]Cl₄ with four chloride anions in the lattice and all the −NH in the protonated form. In addition, eight molecules of water are observed in the lattice as a solvent of crystallization. The ligand strands are coordinated with metal through the N2O cleft, and each metal center exhibits octahedral coordination. In addition, the ligand possesses the S conformation, leading to the formation of helicates. Both the ligand strands are twisted in a helical pattern upon coordination with two metal ions with bond distances (Ni-1–O1 = 2.120; Ni-1–O3 = 2.132; Ni-1–N1 = 2.070; Ni-1–N2 = 1.972; Ni-1–N7 = 2.060; Ni-1–N8 = 1.985 Å) and (Ni-2–O1 = 2.116; Ni-2–O4 = 2.157; Ni-2–N5 = 1.994; Ni-2–N6 = 2.069; Ni-2–N11 = 1.978; Ni-2–N12 = 2.095 Å). Both ligand strands showed a helical twist with a torsion angle measured on C7C8C9C10 = 56.14° and 57.36° on the opposite strands revolves around the M–M axis to provide a helical conformation with a Ni–Ni distance of 5.931 Å (Table S1).

Figure 3.

X-ray structure of cationic [Ni₂(H₂2)₂] (chloride counter ions are omitted for clarity).

Similarly, the better quality crystals for [Zn₂2₂] with nitrate anions were harvested by slow evaporation of water. The helical molecules resolved in the C2/c space group are discretely ionic with two nitrate anions in the lattice (Figure 4). Similar to Ni-2, the molecule exhibited helical conformation, and the double-stranded helicate is cationic with nitrate as counter anions. Each metal center exhibits an octahedral geometry with bond distances (Zn-1–O1 = 2.169; Zn-1–O2 = 2.252; Zn-1–N1 = 2.141; Zn-1–N4 = 2.147; Zn-1–N2 = 2.087; Zn-1–N5 = 2.075 Å) (Figure 4). Both the ligand strands are twisted in a helical pattern and possess coordination with two metal ions through their terminal N2O cleft.

Figure 4.

X-ray structure of cationic [Zn₂(H₂2)₂] (nitrate counter ions are omitted for clarity).

[Co₂2₂] exists in the mesocate form, whereas [Ni₂(H₂2)₂]⁴⁺ and [Zn₂(H₂2)₂]⁴⁺ exhibits helical conformation.²⁴ Interestingly, in the former case, the di-anionic ligands are assembled in the mesocate form, while in the latter, the neutral ligands are assembled in helicate conformation.²⁸

Solution-State Spectral Analysis of Metallohelicates

The ¹H NMR recorded in DMSO for cobalt helicate showed a narrow spectral resonance in the diamagnetic range, which confirms the existence of Co(III).²⁴ The ¹H NMR spectra for Zn-1 showed unresolved peaks with slight broad features indicating their coordination (Figure S11). The signals at 8.52–7.65 ppm correspond to aromatic and azomethine protons, and the peak at 3.80–3.78 ppm is attributed to the methylene proton. In the case of Zn-2, the aromatic and aliphatic region shows well-resolved signals (Figures S12–S13). The ligand H₂2 in DMSO gives two unresolved triplets at 2.95–2.97 ppm and 2.56–2.53 ppm, which correspond to two methylene protons (Figure 5a). The same methylene group in complex Zn-2, showing four well-resolved signals 3.00–2.98 (t, 1H), 2.94 (s, 1H), 2.63 (s, 1H), 2.61–2.58 (t, 1H) (Figure 5b), suggests that the methylene protons become diastereotopic upon coordination to metal ions.²⁸ It is also observed that the molecule is behaving highly dynamic in the solution state; two singlets found at δ = 2.94 and 2.63 ppm correspond to the mesocate structure, and the triplets at δ = 3.00 and 2.61 ppm strongly supports the helicate structure (Figure 5b). Interestingly, upon deuterium exchange, the intensity of one set of signals is greatly reduced (Figure 5c). However, upon recording the ¹H NMR solely in D₂O solvent, all those signals attributed to −CH₂– protons get merged into a single peak at δ = 2.85 ppm (Figure 5d). This apparent change in the spectral pattern suggests that the conformational changes occurred around the −CH₂– group and thus led to the transformation of nonsymmetric to highly symmetric ligand strands. These spectral observations suggest the existence of a solely mesocate structure in water.²⁹ However, in the case of Zn-3, there are no significant changes in ligands, and helicate patterns in ¹H NMR other than minor resonance shift indicates that the ligand upon coordination does not exhibit reasonable structural modification (Figures S14–S15).

Figure 5.

¹H NMR spectra for Zn-2 with the aliphatic region showing a diastereotopic peak pattern. (a) Ligand H₂2 in DMSO. (b) [Zn(H₂2)₂]⁴⁺ in DMSO. (c) D₂O exchange for [Zn(H₂2)₂]⁴⁺. (d) NMR recorded in D₂O for [Zn(H₂2)₂]⁴⁺.

Cyclic Carbonate from Epoxide

Addition of CO₂ to epoxide permits the synthesis of cyclic carbonate or polycarbonate in the presence of an appropriate catalyst. Driven by the tunable nature of metallohelicates, we attempted here to study its catalytic activity for the synthesis of cyclic carbonate. In this regard, we screened the metallohelicates as catalyst for the reaction of styrene oxide and CO₂ under solvent-free condition at 100 °C and 1 MPa for 4 h. Beforehand, we examined the reaction in blank condition (without catalyst) using just ligands H₂1 to H₂3 (without metal) for the conversion of styrene oxide and observed no product and a trace amount of conversion, respectively (Table 1, entries 1–4). These results indicate the importance of catalyst and metal centers for this reaction. Subsequently, the screening of metallohelicates as catalysts and the respective yields of cyclic carbonates are included in Table 1 (Table 1, entries 8–13). Intriguingly, the cobalt-based catalysts demonstrate excellent catalytic activity and resulted in the desired cyclic carbonate with high yields of 90–98% (Table 1, entries 5–7). The change of the metal center to Ni and Zn resulted in slightly decreased yields ranging from 82 to 95% (Table 1, entries 8–10) and from 76 to 94% (Table 1, entries 11–13), respectively. This observation matches well with the literature precedents that the cobalt centers in the +3 oxidation state are found to be the most active catalyst for this reaction.³⁰ Among the catalysts screened, Co-1 has produced the maximum yield of styrene carbonate (98%) with TOF of 128/h (Table 1, entry 5), while the other helicate systems resulted in the desired product in the range of 76–94% yield with a TOF range of 100–128/h (Table 1, Entries 6–13). Thus, the metallohelicate Co-1 has been selected as a suitable catalyst for further optimization.

Table 1. Catalyst Screening for the Reaction of CO₂ to Epoxide^a.

entry	catalyst	yield (%)b	TONc	TOFd (h–1)
1	blank
2	H21	trace
3	H22	trace
4	H23	trace
5	Co-1	98	513	128
6	Co-2	94	492	123
7	Co-3	90	472	118
8	Ni-1	82	430	108
9	Ni-2	93	487	122
10	Ni-3	95	498	124
11	Zn-1	93	488	122
12	Zn-2	76	398	100
13	Zn-3	94	492	123

^aAll reactions were carried out using 100 mmol of styrene oxide, 0.2 mmol of catalyst, and 1 MPa CO₂ at 100 °C for 4 h under solvent-free condition.

^bIsolated yield.

^cTON = mmol of product formed/mmol of catalyst used,

^dTOF (TON/reaction time) calculated after the completion of the reaction time. All the catalysts are selectively giving cyclic carbonate as per ¹H NMR spectral data.

Optimization of Reaction Conditions

With the identified choice of catalyst, the reaction parameters such as CO₂ pressure, temperature, catalyst loading, and effect of tetrabutylammonium bromide (TBAB) in the reaction of styrene oxide and CO₂ were studied to achieve the best catalytic activity. Figure 6 represents the optimization of temperature and pressure with respect to the TOF number for the formation of styrene carbonate. The reaction with varying CO₂ pressures ranging from 0.2 to 1.2 MPa at 100 °C, were performed, and the results are summarized in Figure 6a. The maximum conversion was obtained at 1 MPa CO₂ with a TOF value of 128/h (per molecule), and 64/h (per Co centre) is found to be the optimized pressure. Then, reaction at different temperatures ranging from 30 to 120 °C revealed that the catalytic activity increases relative to the rise in temperature (Figure 6b).

Figure 6.

(a) Pressure versus conversion and TOF plot. (b) Temperature versus conversion and TOF plot for the catalytic synthesis of cyclic carbonates using 0.2 mmol catalysts and 100 mmol of the styrene oxide for 4 h. The conversion was determined by ¹H NMR, and TOF = TON/4.

Then, we studied the effect of catalyst loading keeping the other parameters constant and observed a maximum conversion with 0.2 mmol of catalyst (Table 2). Surprisingly, by decreasing the catalyst loading to 0.1 mmol, a slight increase in TOF to 136/h was observed (Table 2, entry 3); however, with further decrease, it showed a detrimental effect on both conversion and TOF (Table 2, entries 4–7). Here, 0.2 mmol of catalyst is fixed as the optimized catalyst load.

Table 2. Optimization of Catalyst Loading and Effect of TBAB^a.

entry	catalyst (mmol)	TBAB (mmol)	time (h)	conversion (%)b	TONc	TOFd (h–1)
1	0.2		4	98	513	128
2	0.15		4	75	524	131
3	0.1		4	52	545	136
4	0.08		4	14	183	46
5	0.06		4	3	52	13
6	0.04		4	2	52	13
7	0.02		4	2	105	26
8	0.02	0.2	3	99	5189	1730
9	0.02	0.2	2	99	5189	2595
10	0.02	0.2	1	98	5150	5150
11	0.02	0.2	0.5	88	4612	9225
12	0.01	0.2	1	80	8386	8386
13	0.01	0.2	0.5	76	7967	15,934
14	blank	0.2	1	5

^aReactions were carried out using 100 mmol of styrene oxide, 0.2–0.01 mmol of Co-1, 0.2 mmol of TBAB, and 10 bar CO₂ at 100 °C for 30 min to 4 h under solvent-free condition.

^bConversion determined by ¹H-NMR; the selective cyclic carbonate formation is also observed.

^cTON = mmol of product formed/mmol of catalyst used,

^dTOF = TON/reaction time. TOF is calculated per molecule of the catalyst.

Furthermore, to improve the TOF, attempts were made by using TBAB (0.2 mmol) as an additive, and to our delight, it has a pronounced positive effect and resulted in an excellent conversion and TOF. With varied reaction times, we could observe a significant change in TOF, which motivated us to have a further improvisation on TOF.²² Interestingly, even with a decreased catalyst loading of 0.01 and 0.2 mmol of TBAB at 0.5 h, 76% of styrene carbonate with 15,934/h of TOF (7967 h^–1/Co center) was achieved (Table 2, entry 13). The obtained TOF is comparable and/or higher than those of most active catalytic systems.³¹ These results suggest that the cooperative effect of the metallohelicate catalyst and TBAB reflects a reduction in the reaction time with better catalytic activity. It is imperative to mention here that the plain catalyst (only helical complexes under additive-free condition) is proven to give better conversion; the addition of additive increased the efficiency of the catalytic activity to a remarkable raise in the TOF value.

With the optimized reaction conditions, we examined the substrate scope with various alkyl- or aromatic-substituted terminal epoxides (a–j) and non-terminal epoxides (k,l), and the reactions were performed without additives (see Supporting Information for details). As shown in Chart 1, the Co-1 has demonstrated better catalytic activity with many epoxides of varying electronic and steric effects. Also, the TOF of the catalytic system for various substituted epoxides was calculated to be in the range of 90–220/h (per molecule) without the use of any additives (a-j). Among the different substrates studied, internal epoxides exhibited lower conversions (k,l). However, these more difficult substrates afforded fairly good yields (88–99%) in the presence of 0.2 mmol of TBAB (k,l). Besides, we also attempted for a substrate containing bis-epoxides (1,2,7,8-diepoxy octane) and observed a good yield of 94% (j) with a TOF of 98/h (Figure S16).

Chart 1. Screening of Substrates Using Catalyst Co-1.

All reactions were carried out using 100 mmol of the substrate, 0.2 mmol of catalyst, and 1 MPa CO₂ at 100 °C for 4 h under solvent-free condition. Isolated yield TON = mmol of product formed/mmol of catalyst used, TOF = TON/reaction time. The 0.2 mmol of TBAB was used as an additive for internal epoxide, and the yield given in the braces are in absence of additive. The catalysts are giving selective cyclic carbonates confirmed by ¹H NMR spectra. The TON and TOF for the conversion are calculated per molecule.

Cyclic Carbonate from Diol

The catalytic activity found above for Co-1 prompted us further to study its catalytic performance using biodegradable 1,2-diols as an alternative precursor to epoxide for the synthesis of cyclic carbonate with CO₂. By considering the significance and challenges involved in this reaction, we attempted with the same catalytic system in converting diol to cyclic carbonate. Ethylene glycol was chosen as a model substrate and acetonitrile as a solvent using 2.5 MPa CO₂. Initially, the blank reaction was carried out and monitored up to 8 h, which showed no conversion of cyclic carbonate (Table 3, entry 1). Among the cobalt-based metallohelicates, Co-1 was found to be better in terms of conversion (34%), whereas others resulted in trace conversion (Table 3, entries 2–5).

Table 3. Synthesis of 5-Membered Cyclic Carbonate from Ethylene Glycol and Carbon Dioxide^a.

entry	catalyst	dehydrating agents	carboxylation reactionb (%)	hydration reactionb (%)
1	blank	MeCN	no reaction
2	Co-1	MeCN	34
3	Co-2	MeCN	trace
4	Co-3	MeCN	trace
5	Co-1	2-CNPy	15	15
6	Co-2	2-CNPy	19	42
7	Co-3	2-CNPy	65	46
8	Ni-1	2-CNPy	16	20
9	Ni-2	2-CNPy	34	29
10	Ni-3	2-CNPy	19	24
11	Zn-1	2-CNPy	34	30
12	Zn-2	2-CNPy	21	28
13	Zn-3	2-CNPy	27	25

^aAll reactions were carried out using 0.25 mmol of catalyst, 32 mmol of ethylene glycol, 96 mmol of 2-pyridine carbonitrile, at 2.5 MPa CO₂, 100 °C for 8 h.

^bConversion was calculated using ¹H-NMR.

The challenge involved in the reaction is the formation of water as a by-product, which in turn maintains equilibrium and reduces the conversion of the desired product. This problem with the interference of water molecules can be by-passed by using a suitable dehydrating agent.³² With the literature knowledge, we replaced acetonitrile with 2-cyanopyridine as a dehydrating agent.³³ For this conversion, Co-3 worked well with the combination of 2-cyanopyridine and enhanced the conversion of diols to cyclic carbonate (65%), and the results are summarized in Table 3. The formed by-product water is also subsequently been used for the effective conversion of 2-cyanopyridine to 2-picolinamide. Thus, Co-3 helicate can simultaneously catalyze both the reaction in the same reactor vessels (Figure S17).

Reaction Kinetics

In order to determine the path of the reaction, kinetic experiments were performed with styrene oxide as a model substrate under optimized reaction conditions.

As an attempt to understand the rate of the reaction,^34a rate eq 1 was simplified to eq 2, considering the excess use of CO₂ and the concentration of catalyst remains constant during the reaction. The natural logarithm of the rate law (eq 2) results in eq 3, which is possible to afford the order “c” of the reaction with respect to the catalyst concentration by examination of a double logarithmic plot.³⁴

1

Since [epoxide] is a variable, [CO₂] and [cat] are constant, the rate can be expressed as equation 2.

2

By taking natural logarithmic on both sides of eq 2,

3

To investigate the mechanism of the reaction, the co-catalyst free reaction was studied to understand the order of the reaction using Co-1 as a catalyst, and the conversion was recorded in the interval of 30 min. The linear fit plot drawn against ln[Epo] versus time (Figure 7a) confirms that the reaction follows first-order kinetics.^34d,34e

Figure 7.

Reaction kinetics plots. (a) Linear fit plot ln[Epo] versus time to determine the order of the reaction. (b) Catalyst concentration plot ln[rate] versus ln[cat]. (c) Arrhenius plot ln[rate] versus T^–1 (K) to determine the activation energy. (d) Schematic representation of probable mechanism for the helicate catalyzed cyclic carbonate formation. All the reactions for kinetic experiments were carried out using 100 mmol of substrate, 1 MPa CO₂, and 100 °C for plot (a), and the conversion is recorded in an interval of 30 min. For plot (b), the catalyst loading varies from 0.2 mmol to 0.02 mmol. For plot (c), the temperature varied from RT to 120 °C.

In addition, to evaluate the catalytic function of the helicate Co-1, the kinetic experiments were carried out by varying the catalyst concentration from 0.02 to 0.2 mmol. All the reactions were carried out using 1 MPa CO₂, 100 mmol of styrene oxide at 100 °C for 4 h. The plot shown in Figure 7b drawn against ln[Cat] versus ln[rate] suggests that as the concentration of the catalyst increases, the rate of the reaction also increases (Figure 7b).

The activation energy in the conversion of epoxide to cyclic carbonate was established by using the Arrhenius equation.

where Ea is the activation energy (kJ.mol^–1), R is the universal gas constant (8.314 Jmol^–1K), and temperature T is the absolute temperature. The activation energy for the CO₂ insertion into styrene oxide catalyzed by Co-1 was determined from the range of 30–120 °C. The plot between the rates with respect to temperature fitted in a straight line and obeyed the Arrhenius equation (Figure 7c). So, the slope of the straight line is related to −(Ea/R). By using the slope, the activation energy for this reaction was calculated as 53.8 kJmol^–1.

Mechanism

The catalytic mechanism in the reaction is explained in four steps as shown in Figure 7d. The labile NH proton on the ligand leads to the enhanced the activity of the catalyst without any external additives. Based on the kinetic experiments, the mechanism for the catalytic conversion is described as follows. The epoxide is activated by the metal centers interacting via an oxygen atom and leads to the formation of metal–alkoxide in step 1. The epoxide oxygen is electron rich when compared to ketonic oxygen of the ligand, which favors the metal–alkoxide interaction to be more feasible. Simultaneously, the nitride (N^–) ion of the ligand activates the CO₂, implying that the carbon atom has a partial positive charge (δ⁺) and the oxygen atoms have a partial negative charge (δ^–). Thus, each metal center and the labile NH in the ligand are involved in the catalytic reaction. The metal–alkoxide intermediate is quite stable for a rapid CO₂ attack, leading to the CO₂ insertion to form the carbonate intermediate. This nucleophilic attack on the carbonate in step 4 facilitates the ring closure and produces the end product as cyclic carbonate as depicted in Figure 7d.³⁵

Conclusions

In summary, a series of Co-, Ni-, and Zn-based binuclear double-stranded helicates of pyridylhydrazone-based ligands were synthesized. The d–d transitions in the UV–vis spectra indicate an octahedral geometry at the metal center. The X-ray structure of the molecules confirms the formation of a helical type of molecule. Ni-2 and Zn-2 end-up with helicate formation in the solid state, and all the helicates are applied as a catalyst for the preparation of cyclic carbonates. The catalyst converts epoxide to cyclic carbonate selectively, giving better conversion with most of the aliphatic epoxides, under solvent-free conditions. The use of 0.2 mmol TBAB significantly improved the conversion of 76% (in 30 min reaction time) and TOF 15934 h^–1/molecule, and 7967 h^–1/Co center was achieved, which is comparable to most of the reported active metal catalysts. Based on the kinetic studies, a probable mechanism is proposed. The catalyst was also found to be active to convert biodegradable 1,2-diols to corresponding cyclic carbonate using 2-cyanopyridine as a dehydrating agent.

Experimental Section

Materials and General Methods

All the chemicals were purchased from Aldrich & Co. IR spectra were recorded using KBr pellets (1% w/w) on a Perkin-Elmer Spectrum GX FT-IR spectrophotometer. Electronic spectra were recorded on a Shimadzu UV 3101PC spectrophotometer. Mass analyses were performed using the positive and negative ion spray ionization technique on a Waters Q Tof-micro mass spectrometer for all these complexes upon dissolving in methanol–water solvents. CHNS analyses were done using a Perkin-Elmer 2400 CHNS/O analyzer. Single crystal structures were determined using a BRUKER SMART APEX (CCD) diffractometer. ¹H and ¹³C NMR spectra were recorded on a Bruker Avance II 500 MHz or Jeol 600 MHz FT-NMR spectrometer. Chemical shifts for proton resonances are reported in ppm (δ) relative to tetramethylsilane, and ¹³C spectra are calibrated with reference to DMSO-d₆. All the catalytic products were established based on ¹H NMR spectra. The 100 mL high-pressure catalytic reactor was from AMAR equipment.

Synthesis of Nickel and Zinc Metallohelicates

Ni-1. [Ni₂(H₂1)₂]Cl₄(H₂O)₅: To a methanolic solution of ligand H₂1 (652 mg, 2.10 mmol), nickel chloride hexahydrate (500 mg, 2.10 mmol) in methanol was added dropwise. The colorless solution of the ligand turns into pale green and then dark green. The reaction mixture is heated to 60 ^οC for 2 h and then allowed to stir at RT for 14 h. The solution was evaporated under reduced pressure to yield a greenish black solid. Yield: 47%. IR (KBr): ν = 3402, 1634, 1542, 1498, 1356, 1222, 1154, 775 cm^–1. UV–vis [water, λ_nm, (ε/M^–1)]: 1365(906), 1126(141), 870(162), 318(22723), 255(862). Elemental analysis: calcd for C₃₀H₃₈Cl₄Ni₂N₁₂O₉: C, 37.15, H, 3.95; N, 17.33%. Found: C, 37.21; H, 4.425; N, 17.05%. ESI(+)-MS: m/z calcd for [Ni₂1₂]H⁺ 733.08, expt 733.46.

Ni-2. [Ni₂(H₂2)₂]Cl₄(H₂O)₄: A similar procedure adapting H₂2 (1364 mg, 4.20 mmol) and nickel chloride hexahydrate (1000 mg, 4.20 mmol) and evaporated to obtained pale green powder. Yield: 52%. IR (KBr): ν = 3404, 3076, 2900, 1612, 1530, 1468, 1230, 1170, 983, 772 cm^–1. UV–vis [water, λ_nm, (ε/M^–1)]: 1422(1357), 1139(1693), 870(1693). Elemental analysis: calcd for C₃₂H₄₀Cl₄Ni₂N₁₂O₈: C, 39.22; H, 4.11; N, 17.15%. Found: C, 38.90; H, 4.045; N, 17.08%. ESI(+)-MS: m/z calcd for [Ni₂2₂]H⁺ 761.11, expt 761.51.

Ni-3. [Ni₂(H₂3)₂]Cl₄(CH₃OH)(H₂O)₄: A similar procedure adapting H₂3 (712 mg, 2.10 mmol) and nickel chloride hexahydrate (500 mg, 2.10 mmol to yield green solid. Yield: 55%. IR (KBr): ν = 3395, 3023, 2936, 1625, 1534, 1355, 1153, 935, 778 cm^–1. UV–vis [water, λ_nm, (ε/M^–1)]: 1386(1986), 1153(847), 865(515), 318(1003), 255(33165). Elemental analysis: calcd for C₃₄H₄₄Cl₄Ni₂N₁₂O₈: C, 40.42; H, 4.65; N, 16.16%. Found: C, 40.59; H, 5.00; N, 16.16%. ESI(+)-MS: m/z calcd for [Ni₂3₂]H⁺ 789.14, expt 789.13.

Zn-1. [Zn₂(H₂1)₂]Cl₄(H₂O)₇: A similar procedure adapting H₂1 (720 mg, 2.32 mmol) and zinc chloride dihydrate (400 mg, 2.32 mmol), and a yellow solid is obtained. ¹H NMR (D₂O, TMS, 600 MHz): δ = 8.52 (s, 1H), 8.46 (s, 1H), 8.18–8.15 (t, 2H), 8.09(s, 2H), 7.70–7.68 (d, 2H), 7.65(s, 2H), 3.80–3.78 (t, 2H) ppm. IR (KBr): ν = 3413, 3192, 2925, 2852, 2345, 1607, 1560, 1440, 1384, 1305, 1230, 1155, 1100, 1018, 776, 638 cm^–1. UV–Vis [water, λnm, (ε/M^–1)]: 367(8645), 293(40405), 265(32635). Elemental analysis: calcd for C₃₀H₄₂Cl₄Zn₂N₁₂O₁₁. C, 35.35; H, 4.15; N, 16.49%. Found: C, 35.00; H, 3.41; N, 16.08%. ESI(+)-MS: m/z calcd for [Zn₂1₂]K⁺ 783.03, expt 783.70.

Zn-2. [Zn₂(H₂2)₂]Cl₄(H₂O)₈: A similar procedure adapting H₂2 (753 mg, 2.32 mmol) and zinc chloride dihydrate (400 mg, 2.32 mmol) to yield a colorless solid. ¹H NMR (DMSO, TMS, 500 MHz): δ = 11.55–11.52 (d, 1H), 8.61–8.59 (t, 2H), 8.25–8.24 (d, 1H), 8.05 (s, 1H), 7.93–7.88 (m, 2H), 7.48–7.37 (m, 4H), 3.04–3.01 (t, 1H), 3.00 (s, 1H), 2.65 (s, 1H), 2.63–2.61 (t, 2H) ppm. IR (KBr): ν = 3429, 3130, 3035, 2877, 2780, 1656, 1540, 1470, 1443, 1363, 1306, 1282, 1239, 1164, 1096, 997, 943, 779 cm^–1. UV–vis [water, λ_nm, (ε/M^–1)]: 293(102115), 265(60940). Elemental analysis: calcd for C₃₂H₄₈Cl₄Zn₂N₁₂O₁₂: C, 36.08; H, 4.54; N, 15.78%. Found: C, 36.25; H, 4.83; N, 15.48%. ESI(+)-MS: m/z calcd for [Zn₂2₂]H⁺ 775.10, expt 775.70.

Zn-3. [Zn₂(H₂3)₂]Cl₄(H₂O)₆: A similar procedure adapting H₂3 (786 mg, 2.32 mmol) and zinc chloride dihydrate (400 mg, 2.32 mmol) and evaporated to yield a colorless solid. ¹H NMR (D₂O, TMS, 500 MHz): δ = 11.51–11.47 (d, 1H), 8.61–8.56 (d, 2H), 8.28–8.26 (d, 1H), 8.04–8.01 (d, 1H), 7.99–7.94 (m, 1H), 7.91–7.87 (m, 1H), 7.82–7.77 (m, 1H), 7.56–7.53 (t, 1H), 7.39–7.33 (m, 2H), 2.77–2.75 (t, 2H), 2.46–2.40 (m, 2H), 1.97–1.94 (t, 2H) ppm. IR (KBr): ν = 3510, 3131, 2939, 2859, 1651, 1620, 1554, 1468, 1441, 1358, 1275, 1232, 1154, 1098, 1045, 1015, 931, 781 cm^–1. UV–vis [water, λ_nm, (ε/M^–1)]: 290 (59540), 266 (40393). Elemental analysis: calcd for C₃₄H₄₈Cl₄Zn₂N₁₂O₁₀: C, 38.42; H, 4.58; N, 15.90%. Found: C, 38.25; H, 5.03; N, 16.42%. ESI(+)-MS: m/z calcd for [Zn₂3₂]H⁺ 803.13, expt 803.75.

Zn-2a. [Zn₂(H₂2)₂](NO₃)₄: A similar protocol adapting ligand H₂2 (1090 mg, 3.36 mmol) and zinc nitrate hexahydrate (1000 mg, 3.36 mmol) was used. The colorless precipitate obtained was filtered, washed with methanol, and dried to yield a colorless solid. Yield: 72%. IR (KBr): ν = 3420, 1636, 1540, 1380, 1233, 1172, 982, 779 cm^–1. UV–vis [water, λ_nm, (ε/M^–1)]: 292 (81366), 269 (47196), 206 (80093). Elemental analysis: calcd for C₃₂H₃₈Zn₂N₁₆O₁₉: C, 35.54; H, 3.54; N, 20.72%. Found: C, 35.41; H, 3.59; N, 20.72%. ESI(+)-MS: m/z calcd for [Zn₂2₂]H⁺ 775.0981, expt 775.6658.

General Protocol for Cycloaddition of CO₂ to Epoxide

To a 100 mL stainless steel high-pressure reactor vessel, 0.2 mmol of catalyst and 100 mmol of substrate were charged and connected. The reactor was purged with carbon dioxide 2–3 times, and 1 MPa pressure of carbon dioxide was fixed and the reaction mixture was heated at 100 °C for 4 h. The reactor was cooled, and the reaction mixture was passed through a silica column to separate the catalyst and isolate it.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.9b04241.

• Characterization data of all the ligands and the complexes such as ESI-MS, UV–vis, FT-IR, NMR, and crystal data and characterization of catalytic product brief description (PDF)

• Crystallographic data of H₂1 (CIF)

• Crystallographic data of Ni-2 (CIF)

• Crystallographic data of Zn-2 (CIF)

The authors declare no competing financial interest.