Differences in Reactivity of Epoxides in the Copolymerisation with Carbon Dioxide by Zinc‐Based Catalysts: Propylene Oxide versus Cyclohexene Oxide

The catalysis of the reaction of carbon dioxide with epoxides (cyclohexene oxide or propylene oxide) using the (salen)Cr(III)Cl complex as catalyst, where H(2)salen = N,N'-bis(3,5-di-tert-butylsalicylidene)-1,2-cyclohexenediimine (1), to provide copolymer and cyclic carbonate has been investigated by in situ infrared spectroscopy. As previously demonstrated for the cyclohexene oxide/CO(2) reaction in the presence of complex 1, coupling of propylene oxide and carbon dioxide was found to occur by way of a pathway first-order in catalyst concentration. Unlike the cyclohexene oxide/carbon dioxide reaction catalyzed by complex 1, which affords completely alternating copolymer and only small quantities of trans-cyclic cyclohexyl carbonate, under similar conditions propylene oxide/carbon dioxide produces mostly cyclic propylene carbonate. Comparative kinetic measurements were performed as a function of reaction temperature to assess the activation barrier for production of cyclic carbonates and polycarbonates for the two different classes of epoxides, i.e., alicyclic (cyclohexene oxide) and aliphatic (propylene oxide). As anticipated in both instances the unimolecular pathway for cyclic carbonate formation has a larger energy of activation than the bimolecular enchainment pathway. That is, the energies of activation determined for cyclic propylene carbonate and poly(propylene carbonate) formation were 100.5 and 67.6 kJ.mol(-1), respectively, compared to the corresponding values for cyclic cyclohexyl carbonate and poly(cyclohexylene carbonate) production of 133 and 46.9 kJ.mol(-1). The small energy difference in the two concurrent reactions for the propylene oxide/CO(2) process (33 kJ.mol(-1)) accounts for the large quantity of cyclic carbonate produced at elevated temperatures in this instance.

The copolymerization reactions of carbon dioxide and epoxides to give polycarbonates were examined by density functional theory (DFT), and chemically accurate thermochemical data (benchmarked to experimental values) were obtained via composite ab initio methods. All of the examples studied, i.e., formation of poly(ethylene carbonate), poly(propylene carbonate), poly(chloropropylene carbonate), poly(styrene carbonate), poly(cyclohexene carbonate), and poly(indene carbonate), exhibited enthalpies of polymerization of 21–23 kcal/mol, with the exception of poly(cyclopentene carbonate) (15.8 kcal/mol) which suffers both ring strain and intramolecular steric repulsion caused by the cyclopentane ring fused to the polymer chain. The metal-free carbonate backbiting reaction by a free anionic polycarbonate strand is inhibited by bulky groups at the methine carbon but is accelerated by resonance stabilization of the pentavalent transition state in the case involving poly(styrene carbonate). Nucleophilic attack at the methylene carbon of a substituted epoxide has a lower barrier than for the corresponding reaction involving ethylene oxide due to charges being distributed onto the pendant groups. The undesired backbiting reaction to afford cyclic organic carbonates observed under polymerization conditions for many systems due to the low activation barrier (ΔG‡ = 18–25 kcal/mol) was negligible for poly(cyclohexene carbonate) because, in this instance, it must overcome an additional endergonic conformational change (ΔG = 4.7 kcal/mol) before traversing the activation barrier (ΔG‡ = 21.1 kcal/mol) to cyclization. Backbiting from an alkoxide chain end is proposed to proceed via a tetrahedral alkoxide intermediate, where formation of this intermediate is barrierless. Further reaction of this intermediate to the cyclic carbonate has a free energy barrier 10 kcal/mol less than the carbonate chain end backbiting reaction.

The selective conversion of variously substituted epoxides into the corresponding cyclic carbonates under mild reaction conditions was achieved with mononuclear Fe(III) complexes bearing bis-thioether-diphenolate [OSSO]-type ligands, in combination with tetrabutylammonium bromide (TBAB). For example, propylene carbonate was obtained in 1 h at 35 °C (turnover frequency, TOF = 290 h–1), from propylene oxide and 1 bar of CO2 pressure, using 0.1 mol % of the Fe(III) complex and 0.5 mol % of TBAB. Product divergence is observed only for cyclohexene oxide toward the exclusive formation of the aliphatic polycarbonate (TOF = 165 h–1 at 80 °C and 1 bar of CO2 pressure, using 0.1 mol % of the Fe(III) complex and 0.1 mol % of tetrabutylammonium chloride). Kinetic investigations indicated reaction orders of two and one, with respect to the Fe(III) complex, for the production of propylene carbonate and the poly(cyclohexene carbonate), respectively. The enthalpy and entropy of activation were determined using the Eyring equation [for propylene carbonate: ΔH‡ = 8.4 ± 0.7 kcal/mol and ΔS‡ = −33 ± 3 cal/(mol·K); for poly(cyclohexene carbonate): ΔH‡ = 11.9 ± 0.3 kal/mol and ΔS‡ = −36 ± 2.2 cal/(mol·K)]. Supported by density functional theory based investigations, we propose a mechanistic scenario in which the rate-limiting step is the bimetallic ring opening of the epoxide, in the case of propylene carbonate, and the monometallic insertion of the epoxide in the growing polymer chain, in the case of poly(cyclohexene carbonate).

Block copolymerization of carbon dioxide with cyclohexene oxide and 4-vinyl-1-cyclohexene-1,2-epoxide in based poly(propylene carbonate) by yttrium–metal coordination catalyst

The terpolymerization of carbon dioxide (CO2), propylene oxide (PO), and cyclohexene oxide (CHO) were performed by both random polymerization and block polymerization to synthesize the random poly (propylene cyclohexene carbonate) (PPCHC), di-block polymers of poly (propylene carbonate–cyclohexyl carbonate) (PPC-PCHC), and tri-block polymers of poly (cyclohexyl carbonate–propylene carbonate–cyclohexyl carbonate) (PCHC-PPC-PCHC). The kinetics of the thermal degradation of the terpolymers was investigated by the multiple heating rate method (Kissinger-Akahira-Sunose (KAS) method), the single heating rate method (Coats-Redfern method), and the Isoconversional kinetic analysis method proposed by Vyazovkin with the data from thermogravimetric analysis under dynamic conditions. The values of ln k vs. T−1 for the thermal decomposition of four polymers demonstrate the thermal stability of PPC and PPC-PCHC are poorer than PPCHC and PCHC-PPC-PCHC. In addition, for PPCHC and PCHC-PPC-PCHC, there is an intersection between the two rate constant lines, which means that, for thermal stability of PPCHC, it is more stable than PCHC-PPC-PCHC at the temperature less than 309 °C and less stable when the decomposed temperature is more than 309 °C. Pyrolysis-gas chromatography/mass spectrometry (Py-GC/MS) and thermogravimetric analysis/infrared spectrometry (TG/FTIR) techniques were applied to investigate the thermal degradation behavior of the polymers. The results showed that unzipping was the main degradation mechanism of all polymers so the final pyrolysates were cyclic propylene carbonate and cyclic cyclohexene carbonate. For the block copolymers, the main chain scission reaction first occurs at PC-PC linkages initiating an unzipping reaction of PPC chain and then, at CHC–CHC linkages, initiating an unzipping reaction of the PCHC chain. That is why the T−5% of di-block and tri-block polymers were not much higher than that of PPC while two maximum decomposition temperatures were observed for both the block copolymer and the second one were much higher than that of PPC. For PPCHC, the random arranged bulky cyclohexane groups in the polymer chain can effectively suppress the backbiting process and retard the unzipping reaction. Thus, it exhibited much higher T−5% than that of PPC and block copolymers.

Current research efforts in the field of poly(propylene carbonate) prepared from carbon dioxide and propylene oxide are reviewed. Interest in the polymer has been revived in light of the current discussion on sustainability and biodegradability. The progress in understanding and increasing the activity of heterogeneous zinc catalysts is steadily increasing, but without a quantum leap. The microstructure, the properties of the melt, and the solid material are given. The material property profile can be expanded through the synthesis of terpolymers based on propylene oxide, carbon dioxide and other epoxides, lactones or anhydrides etc. with the same catalyst; or the preparation of poly(propylene carbonate) blends with biodegradable or biocompatible components like calcium carbonate, wood flour, fibers, or other biodegradable polymers.

A DFT-based description is given of the CO2/epoxide copolymerization with a catalyst system consisting of metal (chromium, iron, titanium, aluminum)-salen complexes (salen = N,N'-bis(3,5-di-tert-butylsalicyliden-1,6-diaminophenyl) in combination with either chloride, acetate, or dimethylamino pyridine (DMAP) as external nucleophile. Calculations indicate that initiation proceeds through nucleophilic attack at a metal-coordinated epoxide, and the most likely propagation reaction is a bimolecular process in which a metal-bound nucleophile attacks a metal-bound epoxide. Carbon dioxide insertion occurs at a single metal center and is most likely the rate-determining step at low pressure. The prevalent chain terminating/degradation-the so-called backbiting, a reaction leading to formation of cyclic carbonate from the polymer chain-would involve attack of a carbonate nucleophile rather than an alkoxide at the last unit of the growing chain. The backbiting of a free carbonato chain end is particularly efficient. Anion dissociation from six-coordinate aluminum is appreciably easier than from chromium-salen complexes, indicating the reason why in the former case cyclic carbonate is the sole product. Experimental data were gathered for a series of chromium-, aluminum-, iron-, and zinc-salen complexes, which were used in combination with external nucleophiles like DMAP and mainly (tetraalkyl ammonium) chloride/acetate. Aluminum complexes transform PO (propylene oxide) and CO2 to give exclusively propylene carbonate. This is explained by rapid carbonate anion dissociation from a six-coordinate complex and cyclic formation. CO2 insertion or nucleophilic attack of an external nucleophile at a coordinated epoxide (at higher CO2 pressure) are the rate-determining steps. Catalysis with [Cr(salen)(acetate/chloride)] complexes leads to the formation of both cyclic carbonate and polypropylene carbonate with various quantities of ether linkages. The dependence of the activity and selectivity on the CO2 pressure, added nucleophile, reaction temperature, and catalyst concentration is complex. A mechanistic description for the chromium-salen catalysis is proposed comprising a multistep and multicenter reaction cycle. PO and CO2 were also treated with mixtures of aluminum- and chromium-salen complexes to yield unexpected ratios of polypropylene carbonate and cyclic propylene carbonate.

Novel chromium (III) complexes with N4-donor ligands as catalysts for the coupling of CO2 and epoxides in supercritical CO2

The air-stable, chiral (salen)Cr(III)Cl complex (3), where H(2)salen = N,N'-bis(3,5-di-tert-butyl-salicylidene)-1,2-cyclohexene diamine, has been shown to be an effective catalyst for the coupling of cyclohexene oxide and carbon dioxide to afford poly(cyclohexenylene carbonate), along with a small quantity of its trans-cyclic carbonate. The thus produced polycarbonate contained >99% carbonate linkages and had a M(n) value of 8900 g/mol with a polydispersity index of 1.2 as determined by gel permeation chromatography. The turnover number (TON) and turnover frequency (TOF) values of 683 g of polym/g of Cr and 28.5 g of polym/g of Cr/h, respectively for reactions carried out at 80 degrees C and 58.5 bar pressure increased by over 3-fold upon addition of 5 equiv of the Lewis base cocatalyst, N-methyl imidazole. Although this chiral catalyst is well documented for the asymmetric ring-opening (ARO) of epoxides, in this instance the copolymer produced was completely atactic as illustrated by (13)C NMR spectroscopy. Whereas the mechanism for the (salen)Cr(III)-catalyzed ARO of epoxides displays a squared dependence on [catalyst], which presumably is true for the initiation step of the copolymerization reaction, the rate of carbonate chain growth leading to copolymer or cyclic carbonate formation is linearly dependent on [catalyst]. This was demonstrated herein by way of in situ measurements at 80 degrees C and 58.5 bar pressure. Hence, an alternative mechanism for copolymer production is operative, which is suggested to involve a concerted attack of epoxide at the axial site of the chromium(III) complex where the growing polymer chain for epoxide ring-opening resides. Preliminary investigations of this (salen)Cr(III)-catalyzed system for the coupling of propylene oxide and carbon dioxide reveal that although cyclic carbonate is the main product provided at elevated temperatures, at ambient temperature polycarbonate formation is dominant. A common reaction pathway for alicyclic (cyclohexene oxide) and aliphatic (propylene oxide) carbon dioxide coupling is thought to be in effect, where in the latter instance cyclic carbonate production has a greater temperature dependence compared to copolymer formation.

Tetramethylguanidine–silica nanoparticles as an efficient and reusable catalyst for the synthesis of cyclic propylene carbonate from carbon dioxide and propylene oxide

The catalytic activities of tetraphenylporphinatoaluminum chloride (TPPAlCl) and its propylene oxide adduct (TPPAl(PO)2Cl) were investigated in detail together with a quarternary salt Et4NBr for the copolymerization of carbon dioxide and propylene oxide. In addition, for the components and starting raw materials of the catalyst systems, catalytic activities were examined for the copolymerization. The TPPAlCl catalyst delivered oligomers containing ether linkages to a large extent, regardless of its PO adduction. And cyclic propylene carbonate, as byproduct, was formed in a very small portion. Using the TPPAlCl coupled with Et4NBr as a catalyst system, the formation of ether linkages was reduced significantly in the copolymerization; however, the obtained oligomer still contained ether linkages of 25.0 mol % in the backbone. On the other hand, the formation of cyclic carbonate was increased to 22.4 mol % relative to the oligomer product. The results indicate that the salt, which was coupled with the TPPAlCl catalyst, plays a key role in reducing the formation of ether linkage in the oligomer and, however, in enhancing the formation of cyclic carbonate. Similar results were obtained for the copolymerization catalyzed by the TPPAl(PO)2Cl/Et4NBr system. That is, the formation of ether linkages was not restricted further by the PO adduction of the TPPAlCl component in the catalyst system. Only oligomers with a relatively high molecular weight were produced. This indicates that the PO adduction of the TPPAlCl component contributes highly to the initiation and propagation step in the oligomerization, consequently leading to a relatively high molecular weight oligomer. In contrast, the Et4NBr, as well as the Et2AlCl, produced only cyclic carbonate in a very low yield. Furthermore, tetraphenylporphine exhibited no catalytic activity, regardless of using together with Et4NBr. On the other hand, the Et2AlCl coupled with Et4NBr provided a low molecular weight oligomer having ether linkages of 92.3 mol % in addition to the cyclic carbonate. © 1999 John Wiley & Sons, Inc. J Polym Sci A: Polym Chem 37: 3329–3336, 1999

Two dinuclear cobalt porphyrins comprising different structural tethering motifs at the porphyrin periphery were synthesised, along with a representative mononuclear cobalt porphyrin, and their catalytic activities tested towards carbon dioxide–propylene oxide copolymerisation in the presence of bis(triphenylphosphoranyl)ammonium chloride cocatalyst. The catalytic activities of the mononuclear and the bis‐para‐tethered dinuclear cobalt porphyrin with selective formation of poly(propylene carbonate) are largely comparable, showing no benefit of dinuclearity in contrast to the case of cobalt salen complexes and suggesting that polymer growth proceeds exclusively from one metal centre. The alternative bis‐ortho‐tethered porphyrin demonstrated considerably reduced activity, with dominant formation of cyclic propylene carbonate, as a result of hindered substrate approach at the metal centre. Time‐resolved UV/Vis spectroscopic studies suggested a general intolerance of the cobalt(III) porphyrin catalysts towards the copolymerisation conditions in the absence of carbon dioxide pressure, leading to catalytically inactive cobalt(II) species. In the presence of carbon dioxide, the bis‐ortho‐tethered catalyst showed the fastest deactivation, which is related to an unfavourable steric arrangement of the linker fragment, as was also confirmed by NMR spectroscopic measurements.

The synthesis of commercially relevant organic carbonates from CO2 can contribute to a sustainable utilization of this greenhouse gas. The catalytically controlled reaction with epoxides leads to the production of cyclic carbonates and aliphatic polycarbonates. In this work, we succeeded in finding innovative homogeneous catalysts for this task. The foundation was a very variable ligand system closely related to the salen compounds with an N2O2 framework, that was converted with the metals iron, zinc and aluminum to the corresponding metal complexes. The various combinations of ligand and metal yielded 66 different potential catalysts, of which some structures were investigated via x-ray structure analysis. High-pressure experiments were carried out to test the catalysts’ performance, allowing some insights in the relationship between structure and catalytic activity. The conversions were documented by means of the test substrates propylene oxide and cyclohexene oxide. Additionally a wider range of epoxides has been tested in some promising cases. One result is, that the control of the product spectrum via selection of the optimal epoxide-catalyst-combination is possible. The most active iron(III) catalyst was able to produce quantitative yields of propylene carbonate from propylene oxide without the addition of a cocatalyst. With the zinc catalysts the same result was possible, also under mild reaction conditions (40 ◦C, 2 bar CO2, cocatalyst Bu4NI). However, the aluminum catalysts were suitable for the quantitative conversion of cyclohexene oxide to fully alternating polycarbonates. Another important finding is, that the contributing nucleophiles in the reaction have a decisive influence on the formation of the product, regardless if they came from the catalyst itself or from a cocatalyst. While the more nucleophile and according to the HSAB-concept softer iodide anion ismore suitable for the synthesis of cyclic carbonates, the anions bromide and chloride are rather successful in the synthesis of polycarbonates. By the analysis of the many catalytic test results and additional kinetic measurements via in situ FT-IR spectroscopy, it was possible to postulate mechanisms for the different new catalysts of this work.

Two cobalt(II) oxaporphyrins (OTPPCoCl and OTPPCoSbF6) are synthesized in this work and are characterized, including by X‐ray crystallography. Both complexes are tested as catalysts in the copolymerization of propylene oxide (PO)/cyclohexene oxide (CHO) and CO2. Polycarbonate is obtained in CHO/CO2 copolymerization with OTPPCoCl as a catalyst, whereas in the case of PO, cyclic carbonate (CC) is majorly formed. An anion exchange from Cl− to SbF6 − of the cobalt(II) oxaporphyrin leads to a drastic change in the product selectivity: sole polyether is afforded for both epoxides. The polyether formation by OTPPCoSbF6 is postulated to proceed via a cationic mechanism. Further, an equivalent admixture of OTPPCoCl and OTPPCoSbF6 allows the formation of poly(propylene carbonate). In this copolymerization, the latter catalyst acts as a PO activator, while OTPPCoCl functions as the initiator. In addition, density functional theory (DFT) calculations reveal a lower ring‐opening energy of PO by OTPPCoSbF6, compared with the usual propagating species, OTPPCo‐alkoxides and OTPPCo‐carbonates.

Cycloaddition of carbon dioxide to propylene oxide catalysed by tetra- t-butylphthalocyaninatoaluminium(III) chloride