Nowadays, carbon dioxide (CO₂) capture and sequential catalytic conversion has been rationally proposed and deployed to address the issue of excessive CO₂ emission, as it is a nontoxic, easily available, and sustainable C1 building block in organic synthesis and a greenhouse gas that causes global warming and environmental crises [1-3]. In this context, a diversity of bifunctional materials with ordered porosity and efficient catalytic sites has been developed, such as metal-organic frameworks (MOFs), zeolites, and porous carbons [4-6]. Among these, MOFs and zeolites have shown excellent CO₂ capture performance due to the presence of metal cations, as well as donor ligands and polar functional sites, within the structure [7, 8]. However, their practical applications suffer from insufficient long-term stability and/or deactivation by moisture [9-12]. Notably, porous polymers, with the advantages of high stability toward heat and humidity, as well as low density and metal-free features, have been shown to be feasible and powerful as solid adsorbents for CO₂ capture and heterogeneous catalysts for CO₂ conversion [13, 14]. Therefore, the development of reliable technologies for CO₂ capture and conversion by such organic materials is certainly attractive and promising.

The 1, 3, 5-triazine moiety possesses a nitrogen-rich feature, as well as planar, rigid, and high-symmetry structural characteristics, rendering it desirable for the construction of extended CO₂-specific networks with permanent porosity [15, 16]. Meanwhile, as one commercial available industrial chemical, cyanuric chloride (CC) is the chlorinated derivative of triazine. Owing to the electron-deficient nature of carbons residing in the triazine ring, CC has been widely applied in the construction of porous triazine polymers by electrophilic substitution of its chlorine atoms with different nucleophiles [17-19]. For instance, porous covalent organic polymers from reaction of CC with piperazine have been devised, with CO₂ adsorption capacities up to 5616 mg/g at 65 ℃ and 20.0 MPa, which have been proved to be stable in boiling water for at least 1 week [20]. Porous sulfur-bridged covalent organic polymers derived from CC and 1, 3, 5-benzenetrithiol could also provide up to 3294 mg/g of CO₂ at 45 ℃ and 20.0 MPa, while being highly stable against heating up to 400 ℃ [21]. In addition, ferrocene-functionalized microporous aromatic polymers were synthesized by a one-step Friedel-Crafts reaction of ferrocene and CC and have been shown to have a good CO₂ adsorption capability of 16.9 wt% at 0℃ and 0.1 MPa [22].

However, although chemical fixation of CO₂ into value-added chemicals has been a long-sought goal in both academia and industry, the inherent thermodynamic stability and kinetic inertness of CO₂ pose a challenge for its chemical conversion under industrially viable conditions [23, 24]. In order to overcome this obstacle, employment of high-energy starting materials seems to be the wisest choice for utilization of inactive CO₂ as a reactant. Therefore, the atom-economical reaction involving the cyclo-addition of CO₂ with epoxides to render five-membered cyclic carbonates has been intensively studied [25-30]. Among the variety of catalysts developed, it is worth mentioning that porous organic polymers have been demonstrated as recyclable organocatalysts for this transformation [31-35]. However, the reaction protocols involving such polymers as heterogeneous catalysts often need high temperature [31], solvents [32, 33], and/or transition-metal components [34, 35] to enhance their reactivity. Hence, polymer-type metal-free catalysts that can operate under mild and solvent-free conditions are still in demand.

Herein, a novel hydrazine-bridged covalent triazine polymer (HB-CTP) was designed and synthesized through a simple nucleophilic substitution reaction of 2, 4, 6-trihydrazinyl-1, 3, 5-triazine with CC promoted by sodium carbonate at 110 ℃, as shown in Scheme 1. This new material was devised to acquire two outstanding advantages for CO₂ capture and conversion: (1) numerous triazine units to facilitate the reversible adsorption of CO₂ through their basic nitrogen sites that decorate the inside of the material's pores and (2) massive hydrazine moieties activated by electron-deficient triazines to coordinate with the substrates by formation of hydrogen bonds [33, 36, 37]. Indeed, HB-CTP was discovered to be not only a viable and stable CO₂ adsorbent but also an active and recyclable catalyst for cyclo-addition of CO₂ with various epoxides to afford cyclic carbonates at high efficiency. Hence, this work provides a facile method for the first synthesis of a hydrazine-bridged covalent triazine polymer, which shows promising applications in CO₂ capture and catalytic conversion.

CO₂ (99.999%) was used as received without further purification. The 1, 4-dioxane was distilled from sodium/benzophenone under N₂. CC was purchased from J & K Scientific (USA). Hydrazine hydrate was acquired from Alfa Aesar of ThermoFisher Scientific (USA). Tetra-n-butylammonium bromide (TBAB) was obtained commercially from Tianjin Guangfu Fine Chemical Research Institute (China). 2, 4, 6-trihydrazinyl-1, 3, 5-triazine was synthesized according to the established method [38]. Glycidyl propargyl ether and 4-(2, 3-epoxypropyl)-morpholine were prepared by coupling a reaction of corresponding terminal propargylic alcohol or morpholine with epichlorohydrin according to the literature methods with modifications [39, 40]. Other epoxides and the remaining reagents were obtained commercially and used without further purification.

Nuclear-magnetic-resonance (NMR) spectra were recorded on a Bruker AdvanceⅡ 400M-type spectrometer. Fourier-transform infrared (FTIR) spectroscopy was measured using a Nicolet NEXUS FTIR spectrophotometer. Thermal gravimetric analysis (TGA) was carried out by a Mettler Toledo TGA2 STAR* SYSTEM under a N₂ atmosphere with a ramp rate of 10 ℃/min. Elemental analysis was determined with an Elementar Vario EL Ⅲ elemental analyzer. Solid-state ¹³C CP/MAS NMR spectra were recorded on a Varian Infinity-Plus 400 spectrometer. X-ray powder diffraction (XRD) measurements were performed on a Rigaku D/MAX 2400 X-ray diffractometer. Adsorption-desorption measurements for N₂ and CO₂ were conducted on a Quantachrome Autosorb iQ² apparatus. The Brunauer-Emmett-Teller (BET) method was used to calculate the specific surface area. Scanning electron microscopy (SEM) investigations were performed on a Hitachi UHR FE-SEM SU8200 instrument. Transmission electron microscopy (TEM) images were obtained on a FEI TF30 apparatus.

In a Schlenk apparatus under dry nitrogen atmosphere, sodium carbonate (18 mmol, 1.91 g), CC (3 mmol, 0.55 g), and 2, 4, 6-trihydrazinyl-1, 3, 5-triazine (3 mmol, 0.51 g) was added to 1, 4-dioxane (20 mL) at room temperature. The above mixture was stirred at room temperature for 6 h and then at 110 ℃ for 48 h. Afterwards, the light yellow precipitate was collected by filtration and washed consecutively with dimethylformamide, water, and methanol. Finally, the quantitatively yielded product, designated HB-CTP, was dried at 120 ℃ under vacuum for 24 h. Elemental analysis results were as follows: calculated value (%) N 63.78, C 33.42, H 2.80; found value (%) N 60.75, C 27.89, H 2.65.

A 15-mL oven-dried autoclave containing a magnetic stir bar was charged with a corresponding epoxide (10 mmol), HB-CTP (50 mg), and TBAB (0.5 mmol, 0.16 g). The autoclave was then purged with CO₂ three times. The sealed autoclave was pressurized to 2 MPa CO₂ and continuously stirred at 80 ℃ for 12 h. Afterwards, the autoclave was cooled to room temperature, and the remaining CO₂ was vented slowly. Yields of the target products were determined by ¹H NMR with 1, 5-dichloro-2, 4-dinitrobenzene as an internal standard.

4-Ethyl-1, 3-dioxolan-2-one (2a). ¹H NMR (400 MHz, CDCl₃): δ = 4.68–4.63 (m, 1H), 4.52 (t, J = 6.4 Hz, ¹H), 4.07 (t, J = 6.0 Hz, 1H), 1.84–1.71 (m, 2H), 1.02 (t, J = 5.6 Hz, 3H). ¹³C NMR (100 MHz, CDCl₃):δ = 155.2, 78.1, 69.1, 27.0, and 8.5.

4-(chloromethyl)-1, 3-dioxolan-2-one (2b). ¹H NMR (400 MHz, CDCl₃): δ = 4.97–4.94 (m, 1H), 4.59 (t, J = 8.7 Hz, 1H), 4.42 (dd, J = 8.8 and 6.3 Hz, 1H), 3.80–3.71 (m, 2H). ¹³C NMR (100 MHz, CDCl₃): δ = 154.2, 74.3, 67.1, and 43.7.

4-(bromomethyl)-1, 3-dioxolan-2-one (2c). ¹H NMR (400 MHz, CDCl₃): δ = 4.98–4.92 (m, 1H), 4.59 (t, J = 8.8 Hz, 1H), 4.35 (dd, J = 8.9 and 5.9 Hz, 1H), 3.58–3.57 (m, 2H). ¹³C NMR (100 MHz, CDCl₃): δ = 154.2, 74.1, 68.2, and 31.5.

4-(hydroxymethyl)-1, 3-dioxolan-2-one (2d). ¹H NMR (400 MHz, DMSO-d₆): δ = 5.25 (t, J = 4.4 Hz, 1H), 4.81–4.77 (m, 1H), 4.52–4.45 (m, 1H), 4.29–4.27 (m, 1H), 3.68–3.64 (m, 1H), 3.53–3.48 (m, 1H). ¹³C NMR (100 MHz, DMSO-d₆): δ = 155.2, 70.0, 65.9, and 60.6.

4-methyl-1, 3-dioxolan-2-one (2e). ¹H NMR (400 MHz, CDCl₃): δ = 4.88–4.79 (m, 1H), 4.53 (t, J = 8.1 Hz, 1H), 4.00 (t, J = 8.3 Hz, 1H), 1.45 (d, J = 6.3 Hz, 3H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.1, 73.6, 70.7, and 19.4.

4-butyl-1, 3-dioxolan-2-one (2f). ¹H NMR (400 MHz, CDCl₃): δ = 4.73–4.66 (m, 1H), 4.52 (t, J = 8.1 Hz, 1H), 4.06 (t, J = 7.8 Hz, 1H), 1.81–1.67 (m, 2H), 1.40–1.37 (m, 4H), 0.91 (t, J = 6.9 Hz, 3H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.2, 69.5, 33.7, 26.5, 22.3, and 13.9.

4-hexyl-1, 3-dioxolan-2-one (2g). ¹H NMR (400 MHz, CDCl₃): δ = 4.83–4.66 (m, 1H), 4.51 (t, J = 8.0 Hz, 1H), 4.06 (t, J = 8.0 Hz, 1H), 1.81–1.63 (m, 2H), 1.46–1.29 (m, 8H), 0.88 (t, J = 6.4 Hz, 3H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.1, 77.1, 69.4, 33.9, 31.5, 28.8, 24.3, 22.5, and 14.0.

4-(methoxymethyl)-1, 3-dioxolan-2-one (2h). ¹H NMR (400 MHz, CDCl₃): δ = 4.81–4.77 (m, 1H), 4.48 (t, J = 8.4 Hz, 1H), 4.43–4.37 (m, 1H), 3.65–3.52 (m, 2H), 3.40 (s, 3H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.1, 75.2, 71.5, 66.3, and 59.7.

4-(isopropoxymethyl)-1, 3-dioxolan-2-one (2i). ¹H NMR (400 MHz, CDCl₃): δ = 4.80–4.77 (m, 1H), 4.48 (t, J = 8.4 Hz, 1H), 4.40–4.37 (m, 1H), 3.67–3.59 (m, 3H), 1.16 (d, J = 6.0 Hz, 6H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.2, 75.4, 72.8, 67.1, 66.4, 59.7, 21.9, and 21.8.

4-(butoxymethyl)-1, 3-dioxolan-2-one (2j). ¹H NMR (400 MHz, CDCl₃): δ = 4.82–4.76 (m, 1H), 4.47 (t, J = 8.4 Hz, 1H), 4.38–4.35 (m, 1H), 3.67–3.56 (m, 2H), 3.49 (d, J = 6.4 Hz, 2H), 1.57–1.50 (m, 2H), 1.38–1.29 (m, 2H), 0.89 (d, J = 7.6 Hz, 3H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.1, 75.2, 71.9, 69.7, 66.4, 31.6, 19.2, and 13.9.

4-((dodecyloxy)methyl)-1, 3-dioxolan-2-one (2k). ¹H NMR (400 MHz, CDCl₃): δ = 4.82–4.76 (m, 1H), 4.48 (t, J = 8.0 Hz, 1H), 4.40–4.36 (m, 1H), 3.67–3.58 (m, 2H), 3.49 (d, J = 6.8 Hz, 2H), 1.59–1.52 (m, 2H), 1.25 (m, 18H), 0.87 (d, J = 6.8 Hz, 3H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.1, 75.2, 72.4, 69.8, 66.5, 32.0, 29.8, 29.8, 29.7, 29.7, 29.6, 29.5, 29.5, 26.1, 22.8, and 14.2.

4-(phenoxymethyl)-1, 3-dioxolan-2-one (2l). ¹H NMR (400 MHz, CDCl₃): δ = 7.33–7.29 (m, 2H), 7.02 (t, J = 7.4 Hz, 1H), 6.92 (d, J = 8.0 Hz, 2H), 5.06–5.00 (m, 1H), 4.62 (t, J = 8.4 Hz, 1H), 4.54 (dd, J = 8.5 and 5.9 Hz, 1H), 4.24 (dd, J = 10.5 and 4.4 Hz, 1H), 4.16 (dd, J = 10.5 and 3.6 Hz, 1H). ¹³C NMR (100 MHz, CDCl₃): δ = 157.9, 129.9, 122.2, 114.8, 74.2, 67.1, and 66.4.

4-((benzyloxy)methyl)-1, 3-dioxolan-2-one (2m). ¹H NMR (400 MHz, CDCl₃): δ = 7.38–7.26 (m, 5H), 4.60–4.54 (m, 1H), 4.57 (q, J = 12.0 Hz, 2H), 4.47 (t, J = 8.4 Hz, 1H), 4.37 (dd, J = 8.3 and 6.0 Hz, 1H), 3.73–3.59 (m, 2H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.0, 137.2, 128.4, 127.9, 127.6, 75.1, 73.5, 68.9, and 66.2.

4-((allyloxy)methyl)-1, 3-dioxolan-2-one (2n). ¹H NMR (400 MHz, CDCl₃): δ = 5.90–5.83 (m, 1H), 5.30–5.21 (m, 2H), 4.82–4.81 (m, 1H), 4.52–4.48 (m, 1H), 4.06–4.04 (m, 1H), 3.70–3.67 (m, 2H), 3.63–3.59 (m, 2H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.0, 133.8, 118.1, 75.1, 72.7, 69.0, and 66.4.

4-((prop-2-yn-1-yloxy)methyl)-1, 3-dioxolan-2-one (2o). ¹H NMR (400 MHz, CDCl₃): δ = 4.86–4.83 (m, 1H), 4.50 (t, J = 8.4 Hz, 1H), 4.40–4.36 (m, 1H), 4.27–4.16 (m, 2H), 3.79–3.69 (m, 2H), 2.48 (t, J = 2.4 Hz, 1H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.0, 78.6, 75.7, 74.8, 68.5, 66.3, and 58.9.

4-phenyl-1, 3-dioxolan-2-one (2p). ¹H NMR (400 MHz, CDCl₃): δ = 7.46–7.42 (m, 3H), 7.38–7.35 (m, 3H), 5.68 (t, J = 8.0 Hz, 1H), 4.80 (t, J = 8.4 Hz, 1H), 4.37–4.35 (m, 1H). ¹³C NMR (100 MHz, CDCl₃): δ = 154.9, 135.9, 129.9, 129.4, 126.0, 78.1, and 71.3.

4-(morpholinomethyl)-1, 3-dioxolan-2-one (2q). ¹H NMR (400 MHz, CDCl₃): δ = 4.82–4.75 (m, 1H), 4.48 (t, J = 8.3 Hz, 1H), 4.18 (t, J = 7.7 Hz, 1H), 3.61 (t, J = 4.3 Hz, 4H), 2.62 (d, J = 5.3 Hz, 2H), 2.48 (m, 4H). ¹³C NMR (100 MHz, CDCl₃): δ = 154.7, 74.8, 67.6, 66.4, 69.8, and 54.0.

4, 4-dimethyl-1, 3-dioxolan-2-one (2r). ¹H NMR (400 MHz, CDCl₃): δ = 4.13 (s, 2H), 1.50 (s, 6H). ¹³C NMR (100 MHz, CDCl₃): δ = 154.7, 81.8, 75.5, and 26.1.

Hexahydrobenzo[d][1, 3]dioxol-2-one (2s). ¹H NMR (400 MHz, CDCl₃): δ = 4.69–4.64 (m, 2H), 1.87–1.85 (m, 4H), 1.63–1.54 (m, 2H), 1.43–1.35 (m, 2H). ¹³C NMR (100 MHz, CDCl₃): δ = 155.4, 75.8, 26.8, and 19.2.

The functional connectivity and structure of HB-CTP were investigated with FTIR spectroscopy by comparison of the spectra of CC, 2, 4, 6-trihydrazinyl-1, 3, 5-triazine, and the resultant HB-CTP (Fig. 1). The disappearance of the characteristic stretching vibration of C–Cl of CC at 850 cm^-1, together with the immense enhancement of N–H absorption around 3367 cm^-1, would reveal the total substitution of all three Cl atoms of CC, as well as the massive formation of hydrazine linkages. In addition, the strong bands in the 1200–1600 cm^-1 region correspond to the stretching modes of triazine units, while their breathing mode could be assigned near 805 cm^-1.

In the solid state ¹³C CP/MAS NMR spectroscopy (Fig. 2(a)), the only theoretical signal for aromatic triazine carbons of HB-CTP was observed at 168.1 ppm. Such a clean chart without any impurity peaks indicates the integrity of the corresponding triazine framework. Meanwhile, the XRD pattern was found to be featureless (Fig. 2(b)), illustrating an amorphous structure. The morphology of HB-CTP was further investigated by SEM and TEM as depicted in Fig. 2(c) and (d), which show that the HB-CTP material mainly exists in aggregate form. Finally, as shown in Fig. 2(e), the thermal stability of such hydrazine-rich polymer was disclosed, and the initial degradation could be observed at above 250 ℃.

The porosity of the HB-CTP material was measured by nitrogen-adsorption analysis carried out at –196 ℃. As shown in Fig. 3, the adsorption-desorption process displays a reversible type-Ⅱ isotherm, representing a nearly non-porous or macroporous adsorbent that tends to proceed through unrestricted monolayer-multilayer adsorption [41]. Indeed, the BET surface area was found to be 51.2 m²/g, and the total pore volume was 0.28 cm³/g at P/P₀ = 0.99. Such a low level of porosity should be attributed to the strong interaction between triazine framework layers through massive hydrogen bonds donated by the hydrazine linkages. Furthermore, the average pore size calculated from the Barrett-Joyner-Halenda method also lies largely in the macroporous region.

Despite poor surface area and porosity, the HB-CTP material has shown good CO₂ capture capacity (1.86 mmol/g, 8.2 wt%) at 0 ℃ and 0.1 MPa conditions (Fig. 4(a)). This abnormal achievement could be reasonable if one considers the fact that polymers based on triazine units and hydrazine linkages would hold an extremely high N content (60.75%, determined by elemental analysis) to intrinsically favor CO₂ adsorption through dipole-quadrupole interactions between the polarizable CO₂ molecule and the vast number of basic N sites [42-45]. Moreover, the HB-CTP still exhibited satisfactory capture performance after at least five consecutive adsorption-desorption cycles, rendering good recyclability (Fig. 4(b)).

The successful capture of CO₂ with 8.2 wt% capacity by HB-CTP gave us an opportunity to test its catalytic performance for conversion of CO₂ into cyclic carbonates by cyclo-addition reaction. Primary catalytic studies were performed with 2-ethyloxirane 1a and CO₂ (2 MPa) at 60 ℃ for 4 h (Table 1). Without the catalyst HB-CTP, only 15% yield of the corresponding cyclic carbonate 2a was observed by the co-catalyst TBAB (5 mol%) (entry 1). It was found that the introduction of HB-CTP would have an obviously positive effect on reaction outcome, indicating the feasibility of our catalyst design (entry 2). Raising reaction temperature from 60 to 80 ℃ resulted in further enhancement of catalytic activity (entry 3). Positively, a 97% yield of 2a could be obtained when the reaction time was extended to 12 h (entry 4). Furthermore, lowering either HB-CTP loading or CO₂ pressure would lead to a notable decrease in reaction efficiency (entries 5 and 6).

Subsequently, various functionalized terminal and internal epoxides were subjected to the optimized conditions in order to test the generality of the present reaction. As shown in Scheme 2, a number of cyclic carbonates were selectively synthesized with broad substrate scope owing to the remarkable catalytic activity of HB-CTP. Epoxides bearing alkyl, allyl, phenyl, benzyl, propargyl, ether, hydroxyl, halide, and morpholinyl groups reacted smoothly to give the corresponding products in good to excellent yields under metal- and solvent-free conditions (2a-2q). Notably, the less reactive substrate 1k with lauryl substituent as well as gem-substituted 2, 2-dimethyloxirane 1r both furnished the target cyclic carbonates (2k and 2r) in acceptable yields. Furthermore, the internal cyclohexene oxide 1s could also be successfully converted with high efficiency. It is worth mentioning that HB-CTP alone could act as sufficient catalyst for cyclo-addition of CO₂ with epichlorohydrin 1a, epibromohydrin 1b, or glycidol 1c affording comparable yields, although a higher reaction temperature should be provided (Scheme 3).

One advantage of the HB-CTP catalyst is its heterogeneous nature, which can operate under solvent-free conditions. Hence, recycling experiments were conducted to test its recyclability and stability. The results depicted in Fig. 5(a) demonstrate that, by simple centrifugation and vacuum drying, HB-CTP could be reused for at least five runs without significant loss of catalytic activity, rendering the present reaction potentially viable for practical applications. In addition, by the comparison of the FTIR spectra of HB-CTP before and after recycling (Fig. 5(b)), the structure of such catalyst was shown to be largely intact (the rare difference at 2960 cm^-1 was due to the residue of TBAB co-catalyst), which indicates its remarkable stability toward reaction conditions.

It has already been reported that hydroxyl groups attached to the catalyst could play a key role for the high efficiency in cyclo-addition of CO₂ with epoxides [46]. In addition, hydroxyl-functionalized porous polymers could also activate the epoxides by forming hydrogen bonds along the solid/liquid interface [33, 36, 37]. Therefore, on the basis of these results, a plausible reaction mechanism was proposed in which the catalytic cycle within this process tends to proceed via the synergistic operation of hydrogen-bond activation and bromide anion nucleophilic attack toward the epoxide substrate, as shown in Scheme 4. First, the hydrazine group within the structure of HB-CTP acts as donor to polarize the C–O bond of an epoxide through hydrogen bonding, illustrated as A. Then, a nucleophilic attack on the less sterically hindered carbon of epoxide made by the bromide anion would result in the ring-opening intermediate B. Subsequently, the alkoxide within B could be rapidly transferred into carbonate anion species C in the CO₂-rich environment. Finally, the intramolecular nucleophilic ring closure initiated by the carbonate anion of C renders cyclic carbonate and simultaneously releases the HB-CTP and bromide anion for the next catalytic cycle.

A hydrazine-bridged covalent triazine polymer was devised and applied for the first time in CO₂ capture and catalytic conversion. This novel polymer was prepared by a facile reaction based on CC and 2, 4, 6-trihydrazinyl-1, 3, 5-triazine, which have good CO₂ capture capacity as well as satisfactory recyclability. Moreover, provided with a hydrogen-bond donor property by the massive hydrazine linkages within its molecular structure, the heterogeneous HB-CTP exhibited high catalytic activity and reusability for cyclo-addition of CO₂ with epoxides under mild and solvent-free conditions with excellent tolerance to a wide range of functional groups. This HB-CTP may have various utilizations in base and/or hydrogen-bond-promoted reactions, and the exploration of its further applications is ongoing in our laboratory.