Pyrazole Supported Zinc(ii) Benzoates as Catalysts for the Ring Opening Copolymerization of Cyclohexene Oxide and Carbon Dioxide

The bis(pyrazole)zinc(II) benzoate complexes bis(3,5-diphenylpyrazole)zinc(II) benzoate (1), bis(3,5-diphenylpyrazole)zinc(II) 3,5-dinitrobenzoate (2), bis(3,5-diphenylpyrazole)zinc(II) 4-hydroxybenzoate (3), and bis(3,5-di-tert-butylpyrazole)zinc(II) 2-chlorobenzoate (4) were synthesized from the reaction of 3,5-diphenylpyrazole (L1) or 3,5-di-tert-butylpyrazole (L2), zinc(II) acetate and the appropriate benzene carboxylic acid. The molecular structure of complex 2 confirmed that these zinc(II) benzoate complexes adopt a 4-coordinate tetrahedral geometry. All four complexes were screened as catalysts for the copolymerization of carbon dioxide (CO 2) and cyclohexene oxide (CHO) and were found to be active for the formation of poly(cyclohexene carbonate) (PCHC) at CO 2 pressures as low as 1.0 MPa under solvent-free conditions and without the use of a co-catalyst. At some reaction condition, most of the catalysts produced PCHC with high carbonate content of up to 98% and a good amount of cyclic cyclohexene carbonate (CCHC). The copolymers produced have low to moderate molecular weights (5200–12300 g/mol) and with polydispersity indices that vary from 1.19 to 2.50. Matrix Assisted Laser Desorption/Ionization-Time of Flight Mass Spectra (MALDI-TOF MS) of these copolymers showed they have benzoate and hydroxyl end groups.


Introduction
The fact that CO 2 is readily available, inexpensive, and non-toxic renders it advantageous for utilization as a C 1 synthetic feedstock, and thus its conversion to useful products has become highly desirable.The reaction of CO 2 that has been studied extensively in the past five decades is the metal catalyzed alternating copolymerization of CO 2 with epoxides to polycarbonates, some of which are biodegradable and are thus of great importance in the plastic industry.The alicyclic polycarbonate poly(cyclohexene carbonate) PCHC, resulting from the copolymerization of CO 2 and CHO, possesses physical attributes very similar to polystyrene.For example, the glass transition temperature of PCHC (T g = 115 ˝C) [1] is very close to that of polystyrene (T g = 100 ˝C), hence the polymer has good dimensional stability at temperatures above 100 ˝C.Furthermore, the T g of PCHC is slightly lower than that of the classical bisphenol-A polycarbonate (145 ˝C) [2].This allows for the application of PCHC in a number of industrial processes [3,4].In addition, PCHC displays excellent thermal performance with a one-step decomposition at ca. 310 ˝C [5], far higher than those of traditional CO 2 /epoxide copolymers like poly(propylene carbonate) (PPC); again allowing the use of PCHC as an environmentally friendly engineering thermoplastic [3].As a new biodegradable polycarbonate suitable for high-temperature applications, PCHC may provide a solution to the poor dimensional stability of PPC.
The copolymerization of CO 2 and epoxides is catalyzed by a number of metal complexes.Two reviews by Darensbourg [6] and Coates [1,7] and reports by Nozaki [8][9][10] and Lu [11,12] indicate that the best catalysts for this reaction are Schiff-base supported Cr, Co and Zn complexes.For homogeneous zinc-catalyzed copolymerization of CHO and CO 2 , three types of very active catalysts have been reported.These are: zinc carboxylates with fluorinated tails [13,14], bisphenoxides of zinc [15], and three-coordinate zinc complexes containing β-diiminate ligands [16].Zn-based catalyst systems have, thus far, remained one of the most active catalysts for the CO 2 and epoxide copolymerization reaction [17][18][19].In general, zinc(II) complexes containing a Zn-O bond and supported by N-donor bidentate ligand systems have been reported to be highly active for the copolymerization reaction of CO 2 and epoxides [1].However, a significant drawback associated with using CO 2 as a reagent in synthesis is the potential danger associated with operating at high temperatures and pressures.To overcome this drawback, it has been necessary to develop catalyst systems that can operate efficiently at low CO 2 pressures.We herein report the syntheses of very simple bis(pyrazole)zinc(II) benzoates that catalyze the copolymerization of CO 2 and CHO to both poly(cyclohexene carbonate) and cyclic cyclohexene carbonate), with some of the polymers having up to 98% carbonate content and at low CO 2 pressures; in some cases as low as 1.0 MPa.Although the catalytic activity is not as high as have been reported for some zinc [1,7] and cobalt [8] catalysts, the current zinc catalyst system is very simple to prepare and thus provide readily accessible non-toxic catalysts for CO 2 /epoxide copolymerization.

Synthesis and Characterization of Zinc(II) Complexes
The pyrazole ligands, 3,5-diphenylpyrazole (L1) or 3,5-di-tert-butylpyrazole (L2), were reacted with zinc(II) acetate in a 2:1 ratio of ligand to metal precursor to give the zinc(II) benzoate complexes shown in Scheme 1.The products were characterized by a combination of 1 H NMR spectroscopy, elemental analysis and, in the case of 2, by single crystal X-ray crystallography.The NMR data of these complexes have typical signals of the pyrazoles and the benzoates.
The crystal structure obtained for complex 2 (Figure 1) [20] shows that it is a monometallic zinc(II) tetrahedrally coordinated to two L1 ligands through the nitrogen atoms and to one oxygen atom of each of the two benzoate ligands.The crystallographic parameters for this complex are given in Table S1.Complex 2 was found to have Zn-O and Zn-N bond lengths of 1.94 and 2.00 Å, respectively.The angles around the zinc center varies from 100.70 ˝for O(1)-Zn(1)-O(3) to 116.78 ˝for O(3)-Zn(1)-N(3).These angles show the distorted tetrahedral geometry adopted by the complex (Figure 1).The structure of 2 can be compared to a similar complex of formula [Zn(L1) 2 (Bz) 2 ] (where Bz = benzoate group) synthesized by Baruah [21], which has an average Zn-O bond length of 1.93 Å and is very close to the average Zn-O bond length in 2 (1.94 Å).The angles around the zinc center in [Zn(L1) 2 (Bz) 2 ] are between 102.45 ˝and 114.59 ˝and for 2, these angles vary from 110.70 ˝to 116.78 ˝.Scheme 1. Synthesis of bis(pyrazole)zinc(II) benzoates pre-catalysts.

Ring-Opening Copolymerization of CO2 and CHO with Complexes 1-4
The four bis(pyrazole)zinc(II) benzoate complexes (1-4) were screened as catalysts for the copolymerization of CO2 and CHO.All four complexes (1-4) were found to be active for the

Ring-Opening Copolymerization of CO2 and CHO with Complexes 1-4
The four bis(pyrazole)zinc(II) benzoate complexes (1-4) were screened as catalysts for the copolymerization of CO2 and CHO.All four complexes (1-4) were found to be active for the

Ring-Opening Copolymerization of CO 2 and CHO with Complexes 1-4
The four bis(pyrazole)zinc(II) benzoate complexes (1-4) were screened as catalysts for the copolymerization of CO 2 and CHO.All four complexes (1-4) were found to be active for the copolymerization reaction at CO 2 pressure of 1.0 MPa.A variation of monomer:catalyst ([M]:[C]) ratio showed that complexes 1 and 2 were the most active at [M]:[C] ratio of 500:1 (Table 1).When the amount of catalyst was increased (i.e., 250:1), there was negligible increase in the catalyst activities of 1 and 2. Complex 4 showed no catalytic activity when [M]:[C] ratios of 1000:1 and 500:1 were used.This observation is in line with suggestion by Klaus [22] that low concentration catalyst lowers activity even in bimetallic catalyst systems.Catalytic activity was only observed when higher catalyst loadings of [M]:[C] ratio of 250:1 were used.The requirement for high catalyst loading of these complexes shows that for copolymerization to occur, the active metal centers (Zn-O bonds) should be present in excess compared to the CHO comonomer.The decreased activity of these complexes can also be attributed to the bulky tertiary-butyl groups in the 3,5-position of the pyrazole moiety in 4; and for 3, the presence of electron-donating hydroxyl group on the benzoate group which makes its metal center less electrophilic and therefore less active towards the ring-opening copolymerization of CHO with CO 2 .For these reasons, further investigations involving catalysis with these complexes were carried out using [M]:[C] ratio of 500:1 for 1 and 2 and 250:1 for 3 and 4. What is significant about using 1-4 as catalysts for copolymerization of CO 2 and CHO is that they did not require the use of solvents or co-catalyst to catalyze the reaction but produced PCHC with turnover frequencies (TOFs) that are comparable to structurally-related monomeric zinc(II) carboxylate complexes reported in the literature [1,23].Nevertheless, Coates' zinc(II) carboxylate complexes still remain one of the most active catalysts for CO 2 and epoxide copolymerization reaction, producing PCHC with carbonate content greater than 93% with narrow molecular weight distribution [24][25][26].In addition to producing PCHC, the copolymerization reactions catalyzed by 1-4 also led to the production of cyclic cyclohexene carbonate (CCHC) as a byproduct.This is generally the result of back-biting of the growing copolymer chain [27].
The PCHC produced were of moderate to high number average molecular weight (M n ), with M n in the range 5200-12300 g¨mol ´1.Molecular weight distribution in the copolymers, a measure of polydispersity index, was found to be moderate to wide in the range 1.19-2.50.The lower than expected M n values for some of the PCHC produced in this work shows that the copolymerization process is not living but rather immortal.The immortal nature of the copolymerization could be due to the presence of chain transfer agents (CTAs).One possible CTA could be the presence of trace amounts of water from insufficiently dry CO 2 , glassware and possibly other equipment.
Variation of CO 2 pressure while temperature (120 ˝C) and amount of CHO (9.88 mmol) were held constant resulted in significant increase in conversion of CHO to both PCHC and CCHC as CO 2 pressure increased.Also significant was the favored production of PCHC over CCHC (Table 2).
Although the increased formation of PCHC over CCHC with increasing CO 2 pressure is consistent with literature reports [28], it is only complex 4 that had large amounts of CCHC formed.Complexes 1-3 had selectivity for the formation of PCHC ranging from 71%-89% as pressure increased; but complex 2 was the most active for PCHC formation with TOF of up to 32 h ´1 and 89% selectivity for PCHC over CCHC at CO 2 pressure of 4.0 MPa.The higher catalytic activity of 2 could be attributed to the electron-withdrawing nature of the nitro groups in the 3,5-position of the phenyl ring that facilitates the initial binding of the epoxide; thus making the zinc(II) center in 2 highly electrophilic and more active in binding the epoxide before its ring-opening.This observation is in accordance with the high activities reported in literature for zinc(II) complexes bearing electron-withdrawing groups [16,23,29,30].In contrast, the complex bearing the OH group, 3, gave some of the lowest catalytic activities possibly because of a poorer electrophilic zinc center in this catalyst.
The kinetics of the copolymer catalyzed by 1 was studied using the in situ equipment ReactIR 45 that monitors stretching bands of IR active functional groups in the polymer.For PCHC and CCHC their carbonyl group stretching frequencies of 1750 cm ´1 and 1825 cm ´1, respectively, were followed as the reaction progressed (Figure S1). Figure S1 shows the asymmetric ν(C=O) stretch of PCHC as the stronger absorption peak at 1750 cm ´1 and that of CCHC as the weaker absorption peak at 1825 cm ´1 similar to observations made by Darensbourg and co-workers [31].It is also clear from the figure that the initiation step in the polymerization, which involves the ring-opening of the three-membered cyclic ether in CHO, is rather slow with induction time of ~6 h.Because carboxylates are known to be weak nucleophiles, catalysts containing carboxylates are usually pre-activated with Lewis acid, which may be a cocatalyst [22].Thus, in the absence of a cocatalyst, it is not surprising that reaction with complex 1 as catalysts had such long induction period.However, once the polymerization started the formation of PCHC was exponential with time, while the formation of CCHC increased linearly with time (Figure 2).The slower formation of CCHC than PCHC is attributed to the high activation energy barrier for CCHC formation, which is reported to be >80 kJ¨mol ´1 higher than that of PCHC production [29].The 1 H NMR data also indicate that the major linkage in the polymers were carbonate in most of the polymers (4.63 ppm); although in some of the polymers ether were substantial ether linkages (3.45 ppm).
Variation of CO2 pressure while temperature (120 °C) and amount of CHO (9.88 mmol) were held constant resulted in significant increase in conversion of CHO to both PCHC and CCHC as CO2 pressure increased.Also significant was the favored production of PCHC over CCHC (Table 2).Although the increased formation of PCHC over CCHC with increasing CO2 pressure is consistent with literature reports [28], it is only complex 4 that had large amounts of CCHC formed.Complexes 1-3 had selectivity for the formation of PCHC ranging from 71%-89% as pressure increased; but complex 2 was the most active for PCHC formation with TOF of up to 32 h −1 and 89% selectivity for PCHC over CCHC at CO2 pressure of 4.0 MPa.The higher catalytic activity of 2 could be attributed to the electron-withdrawing nature of the nitro groups in the 3,5-position of the phenyl ring that facilitates the initial binding of the epoxide; thus making the zinc(II) center in 2 highly electrophilic and more active in binding the epoxide before its ring-opening.This observation is in accordance with the high activities reported in literature for zinc(II) complexes bearing electron-withdrawing groups [16,23,29,30].In contrast, the complex bearing the OH group, 3, gave some of the lowest catalytic activities possibly because of a poorer electrophilic zinc center in this catalyst.
The kinetics of the copolymer catalyzed by 1 was studied using the in situ equipment ReactIR 45 that monitors stretching bands of IR active functional groups in the polymer.For PCHC and CCHC their carbonyl group stretching frequencies of 1750 cm −1 and 1825 cm −1 , respectively, were followed as the reaction progressed (Figure S1). Figure S1 shows the asymmetric ν(C=O) stretch of PCHC as the stronger absorption peak at 1750 cm −1 and that of CCHC as the weaker absorption peak at 1825 cm −1 similar to observations made by Darensbourg and co-workers [31].It is also clear from the figure that the initiation step in the polymerization, which involves the ring-opening of the three-membered cyclic ether in CHO, is rather slow with induction time of ~6 h.Because carboxylates are known to be weak nucleophiles, catalysts containing carboxylates are usually pre-activated with Lewis acid, which may be a cocatalyst [22].Thus, in the absence of a cocatalyst, it is not surprising that reaction with complex 1 as catalysts had such long induction period.However, once the polymerization started the formation of PCHC was exponential with time, while the formation of CCHC increased linearly with time (Figure 2).The slower formation of CCHC than PCHC is attributed to the high activation energy barrier for CCHC formation, which is reported to be >80 kJ•mol −1 higher than that of PCHC production [29].The 1 H NMR data also indicate that the major linkage in the polymers were carbonate in most of the polymers (4.63 ppm); although in some of the polymers ether were substantial ether linkages (3.45 ppm).In several of the reactions, the PCHC produced had relatively high carbonate content (some as high as 98%) and generally found to be dependent on CO 2 pressure; although there were a significant number of catalytic runs that produced polymers with significant ether content (Table 2, runs 1-3).A typical polymer showing high carbonate content is depicted by its 1 H NMR spectrum (Figure S2).Higher CO 2 pressures favored the formation of almost completely alternating copolymers with high carbonate content (Table 2, runs 10-16).In general, catalysts with less electrophilic metal centers gave PCHC with higher carbonate content.For example, copolymers with complex 4 possess higher carbonate content than those with complex 2 because the two chloro-substituted benzoates in 4 are less electron-withdrawing than the two nitro groups in 2. The relatively high carbonate content in PCHC obtained for catalysts with less electrophilic metal centers (1, 3 and 4) suggest that the insertion of CO 2 into the growing chain is greater or equal to the rate of CHO ring-opening; hence the copolymerization is almost completely alternating.However, in the more electrophilic catalyst, 2, the metal center is more active in binding and ring-opening CHO for copolymerization.For this reason, catalyst 2 opens the epoxide ring in CHO more rapidly and this results in rapid insertion of CHO fragments into the growing chain and slow incorporation of CO 2 ; hence the resulting copolymers have low carbonate content.The carbonate content of the copolymers was found to affect their glass transition temperature (T g ) (Table 2).Glass transition temperature in polycarbonates is affected by ether linkages in the polymer, so for our polymers with high carbonate content, it is not surprising that their T g (103-112 ˝C) were lower than the average T g in the literature (115 ˝C) [1].
The 13 C{ 1 H} NMR analyses (Figure S3) of all the PCHC produced indicate that the polymers are atactic and from MALDI-TOF mass spectrometry data, the polymer end-groups were either a benzoate ligand or a hydroxycyclohexyloxy group.Figure 3b is typical MALDI-TOF MS spectrum of our PCHCs produced with regular repeating mass units of 142.2. Figure 3a shows the benzoate group in our pre-catalyst 4 as the initiator of the copolymerization reaction.This particular polymer showed a significant ether linkages, depicted by peaks that conform to loss of CO 2 (e.g., peak 6256 from 6299 minus 44).So from the MALDI-TOF mass spectral data we were able to establish that some of the polymers produced have both carbonate and ether linkages; although some polymers have very high carbonate content.Figure 3b shows that water was as a competing initiator in the copolymerization reaction.The mass values of the stronger signals in Figure 3a,b match (211.0 (benzoate) + 142.2n (repeating units) + 115.1 (hydroxyclohexyloxy group)) and (17.0 (hydroxyl) + 142.2n (repeating units) + 115.1 (hydroxyclohexyloxy group)), respectively.The mass spectrometry data therefore suggests that the copolymerization is terminated by protolysis to give either copolymer I or II.Copolymer I shows that the possible initiating mechanism for the copolymerization is the ring-opening of CHO by nucleophilic attack.

Instrumentation
1 H and 13 C{ 1 H} NMR spectra were recorded in chloroform-d (CDCl 3 ) and dimethylsulphoxide-d 6 (DMSO) on a Bruker Ultrashield 400 instrument (Bruker, Karlsruhe, Germany) (400 MHz for 1 H NMR and 100 MHz for 13 C{ 1 H} NMR) at room temperature. 1H and 13 C{ 1 H} NMR chemical shifts were referenced to the residual signals of the protons or carbons of the NMR solvents and are quoted in δ: CDCl 3 at 7.26 and 77.16 ppm for 1 H and 13 C{ 1 H} NMR spectra respectively; DMSO at 2.50 and 39.52 ppm for 1 H and 13 C{ 1 H} NMR spectra respectively.Coupling constants are measured in Hertz (Hz).Infrared spectra were recorded on a Bruker FR-IR Tensor27 spectrometer fitted with an ATP-IR probe.Elemental analyses were performed on a Vario Elementar IIImicrocube CHNS analyzer at Rhodes University, South Africa.The crystal evaluation and data collection was performed on a Bruker Quazar SMART APEXII diffractometer using Mo radiation source.
For characterization of the copolymers produced, a JEOL JNM-ECP500 NMR spectrometer ( 1 H 500 MHz, 13 C{ 1 H} 126 MHz) was used.Infrared spectra were recorded on a Shimadzu FTIR-8400 spectrometer.Gel permeation chromatography (GPC) analyses of the copolymers were carried out with two columns (Shodex KF-804L) using tetrahydrofuran as eluent at 40 ˝C at a flow rate of 1 mL/min.The molecular weights were calibrated against polystyrene standards.Differential scanning calorimetry (DSC) measurements were performed on a Seiko DSC 7020 analyzer at a heating and cooling rate of 10 ˝C/min; the reported glass transition temperature (T g ) values were determined from the second heating scan.

Representative Procedure for CO 2 and CHO Copolymerization Reaction
In a typical reaction, to a clean autoclave (dried at 80 ˝C overnight in dynamic vacuum) containing the appropriate amount of the catalyst, was added 2.0 mL CHO under inert conditions.The reaction was initiated by pressurizing the autoclave with CO 2 and left to stir at the required temperature for the appropriate time.The autoclave was then vented, allowed to cool to room temperature and a small amount of phenanthrene (internal standard) added.A small aliquot was removed for 1 H NMR and IR analysis at the end of each reaction and the remaining mixture was quenched with five drops of a 5% HCl solution in methanol.
The copolymerization reaction that was monitored by IR spectroscopy was performed in a 150 mL Parr autoclave fitted with a ReactIR probe.The Parr autoclave, containing the appropriate amount of catalyst, was charged with toluene (7.0 mL) and CHO (1.0 mL) via the injection pot.A single 120 scan background spectrum was collected.The autoclave was pressurized to 3.0 MPa of CO 2 and heated at 120 ˝C as the IR probe began collecting scans.The probe was set up to collect one spectrum every minute for a period of 26 h 56 min.

Conclusions
We have prepared the first pyrazole zinc(II) benzoate complexes that are active and highly selective for the formation of PCHC from CO 2 and CHO copolymerization reaction.All the PCHC copolymers produced have glass transition temperatures that are indicative of almost completely alternating copolymers (high carbonate content).Our catalyst systems achieved TOFs that are higher than those reported by Darensbourg using monomeric zinc(II) carboxylate complexes containing electron-withdrawing substituents.The two most active catalysts are the ones with the electron-deficient 3,5-dinitrobenzoate and 2-chlorobenzoate groups (2 and 4), indicating that the higher electrophilicity of the zinc centers in these two catalysts is responsible for the higher activity.The results obtained also show that catalysts 1-3 are more selective for PCHC formation than catalyst 4, which only shows less than 70% selectivity for PCHC.We can thus conclude that catalysts 1-3 are best for the selective formation of PCHC while catalyst 4 would be best for CCHC formation; but, in general, we see these pyrazole supported zinc(II) benzoates as new catalysts systems in the copolymerization of epoxide and carbon dioxide that can be tuned to make either predominantly polycarbonates or cyclic carbonates and are currently working on various modifications of metal complexes using pyrazolyl ligands for this and other ring-opening polymerization reactions.

Table 1 .
Effect of catalyst loading on catalytic activity a .
a Reactions were carried out in 1.0 mL CHO using 1.0 MPa CO 2 pressure for 16 h at 120 ˝C; b % conversion, determined from 1 H NMR and IR spectroscopy; c moles of CHO consumed per mole of catalyst per hour.

Table 2 .
Effect of CO 2 pressure on CHO and CO 2 copolymerization a .
a Reactions were carried out in 1.0 mL CHO at 120 ˝C for 16 h using [M]:[C] ratio of 500:1 for 1-2 and 250:1 for 3-4; b % conversion, determined from 1 H NMR and IR spectroscopy; c determined from 1 H NMR spectra by comparing integration of proton peaks at 4.63 ppm corresponding to carbonate linkage and 3.45 ppm for ether linkage; d moles of CHO consumed per mole of catalyst per hour; e determined by GPC; f from DSC measurements.
Figure3a,b are assigned to the mass spectra of the copolymers using complexes 2 and 1 respectively.