Solventless Coupling of Epoxides and CO 2 in Compressed Medium Catalysed by Fluorinated Metalloporphyrins

Metal complexes of meso-arylporphyrins (Cr(III), Fe(III), and Zn(II)) were evaluated in the coupling reaction of cyclohexene oxide (CHO) with CO2 in compressed medium, where the Cr complexes were demonstrated to be the most active systems, leading predominantly to copolymerisation products. It is noteworthy that no addition of solvent was required. To improve the catalytic activity, and to simultaneously increase the solubility in compressed CO2, a new fluorinated catalyst, tetrakis(4-trifluoromethylphenyl)porphyrinatochromium(III) chloride (CrCl-pCF3TPP), was applied to this reaction. The alternating copolymerisation of CHO with CO2, using the Cr(III) fluorinated porphyrin catalyst, required the use of a co-catalyst, bis(triphenylphosphine)iminium chloride (PPNCl), with the best yields of copolymers being obtained at 80 ◦C, and CO2 pressures in the range of 50–110 bar, over a period of 24 h, with a low catalyst/substrate molar ratio (0.07%). The polycarbonate’s structure was analysed by 1H NMR, 13C NMR, and MALDI-TOF spectroscopy, which demonstrated high carbonate incorporations (98–99%). Gel permeation chromatography revealed number-average molecular weights (Mn) in the range of 4800–12,800 and narrow molecular weight distributions (Mw/Mn ≤ 1.63).


Introduction
Carbon dioxide in compressed form is an attractive green alternative to organic solvents in synthetic chemistry due to its intrinsic low toxicity and availability.Particularly, the coupling of carbon dioxide (CO 2 ) with epoxides [1][2][3][4][5] is a topic of great interest due to the economic and environmental benefits arising from the utilisation of renewable sources for the preparation of polycarbonates or cyclic carbonates [6][7][8][9], and the growing concern on the greenhouse effect [10][11][12][13][14][15].There is substantial literature on catalyst development for CO 2 insertions into epoxides, most of which consist of transition-metal complexes modified with N-donor ligands [16][17][18][19][20][21][22][23][24][25][26][27].However, the low polarity of CO 2 often generates insolubility problems when transition-metal complexes are employed as catalysts.To address this issue, the introduction of fluorinated groups in the ligands is, in general, a promising strategy to improve the solubility of organometallic catalysts in compressed CO 2 [28].
In this context, metalloporphyrins are particularly attractive among the homogeneous catalysts, since their aromatic heterocycle can be easily modulated by the introduction of several peripheral substituents that may increase their solubility in compressed CO 2 , and also because they are able to coordinate with a wide range of different transition metals, thus offering the possibility of fine-tuning the catalytic activity and selectivity [29][30][31][32][33][34][35][36].
Recently, we reported the application of meso-phenyl Mn(III) porphyrins as highly efficient catalysts for coupling the reactions of epoxides with CO 2 , showing that the introduction of electron withdrawing halogen atoms in the ortho positions of meso-tetraphenylporphyrin manganese(III) complexes strengthens the Lewis acidity of the metal centre, thereby increasing their catalytic activity [37].Earlier, in 2000, Holmes reported the use of tetrakis(pentafluorophenyl)porphyrinato chromium(III) chloride (6) in the copolymerisation reaction of cyclohexene oxide with CO 2 under supercritical conditions [20].However, these catalysts originated polycarbonates with a low number-average molecular weight (M n ≈ 3500).In addition, the preparation of such a catalyst involves the use of an expensive perfluorinated reagent and low synthetic yield is obtained, resulting from its laborious purification procedure.So, we hypothesised that a convenient alternative could be achieved by the introduction of trifluoromethyl groups in the meso-aryl porphyrin's backbone.Such electron withdrawing groups would not only promote the catalyst's solubility in compressed CO 2 , but could also increase the Lewis acidity of the metal centre and consequently improve its catalytic properties for CO 2 /epoxide couplings.
In this paper, we present the catalytic evaluation of metal complexes of meso-arylporphyrins (Cr(III), Fe(III), and Zn(II)) in the coupling reaction of epoxides with compressed CO 2 , in the absence of added solvent.For the Cr(III) catalysts, the effect of the fluorine atoms on the meso-aryl porphyrins are appraised regarding catalytic activity and selectivity for the alternating copolymerisation of cyclohexene oxide and CO 2 .We further describe the synthesis of the fluorinated Cr(III)-porphyrin catalyst, tetrakis(4-trifluoromethylphenyl)porphyrinatochromium(III) chloride (7), whose potentialities are explored in this reaction.

Synthesis of the Catalysts
The preparation of metalloporphyrin complexes 1-6 (Figure 1) was accomplished based on recently developed methods by Coimbra's group [38,39] through the condensation of pyrrole with the desired aldehydes, followed by metal insertion through the reaction of free-base porphyrins with the appropriate metal salts [40,41].
Catalysts 2017, 7, 210 2 of 14 In this context, metalloporphyrins are particularly attractive among the homogeneous catalysts, since their aromatic heterocycle can be easily modulated by the introduction of several peripheral substituents that may increase their solubility in compressed CO2, and also because they are able to coordinate with a wide range of different transition metals, thus offering the possibility of fine-tuning the catalytic activity and selectivity [29][30][31][32][33][34][35][36].
Recently, we reported the application of meso-phenyl Mn(III) porphyrins as highly efficient catalysts for coupling the reactions of epoxides with CO2, showing that the introduction of electron withdrawing halogen atoms in the ortho positions of meso-tetraphenylporphyrin manganese(III) complexes strengthens the Lewis acidity of the metal centre, thereby increasing their catalytic activity [37].Earlier, in 2000, Holmes reported the use of tetrakis(pentafluorophenyl)porphyrinato chromium(III) chloride (6) in the copolymerisation reaction of cyclohexene oxide with CO2 under supercritical conditions [20].However, these catalysts originated polycarbonates with a low number-average molecular weight (Mn ≈ 3500).In addition, the preparation of such a catalyst involves the use of an expensive perfluorinated reagent and low synthetic yield is obtained, resulting from its laborious purification procedure.So, we hypothesised that a convenient alternative could be achieved by the introduction of trifluoromethyl groups in the meso-aryl porphyrin's backbone.Such electron withdrawing groups would not only promote the catalyst's solubility in compressed CO2, but could also increase the Lewis acidity of the metal centre and consequently improve its catalytic properties for CO2/epoxide couplings.
In this paper, we present the catalytic evaluation of metal complexes of meso-arylporphyrins (Cr(III), Fe(III), and Zn(II)) in the coupling reaction of epoxides with compressed CO2, in the absence of added solvent.For the Cr(III) catalysts, the effect of the fluorine atoms on the meso-aryl porphyrins are appraised regarding catalytic activity and selectivity for the alternating copolymerisation of cyclohexene oxide and CO2.We further describe the synthesis of the fluorinated Cr(III)-porphyrin catalyst, tetrakis(4-trifluoromethylphenyl)porphyrinatochromium(III) chloride (7), whose potentialities are explored in this reaction.

Synthesis of the Catalysts
The preparation of metalloporphyrin complexes 1-6 (Figure 1) was accomplished based on recently developed methods by Coimbra's group [38,39] through the condensation of pyrrole with the desired aldehydes, followed by metal insertion through the reaction of free-base porphyrins with the appropriate metal salts [40,41].The metalloporphyrin 7, previously reported in [42][43][44][45] was synthesised, in high yields, by the reaction of equimolar amounts of pyrrole with 4-(trifluoromethyl)benzaldehyde in a The metalloporphyrin 7, previously reported in [42][43][44][45] was synthesised, in high yields, by the reaction of equimolar amounts of pyrrole with 4-(trifluoromethyl)benzaldehyde in a nitrobenzene/acetic acid mixture, using NaY zeolite as solid catalyst [39], followed by the complexation of p-CF 3 TPP with CrCl 2 and in situ air oxidation [40] in refluxing DMF (Scheme 1).

Effect of the Metal
The effects of the metal (M) and co-axial ligand (X) on the catalytic activity and selectivity in the coupling reaction of cyclohexene oxide (CHO) and CO2 using tetraphenylporphyrin (TPP) as a model N-donor ligand (Figure 1) were first appraised.The reactions were performed in the absence of solvent, under previously optimised conditions (50 bar CO2, 80 °C), using a 0.07 mol % catalyst [37], with and without bis(triphenylphosphine)iminium chloride (PPNCl) as a co-catalyst.The results are presented in Table 1.Under these conditions, all of the TPP-based metalloporphyrin catalysts required the use of a co-catalyst to promote the coupling reaction of CHO with CO2.Among them, Cr(III) catalytic systems 1 and 2 (either with Cl or OAc as co-axial ligands) presented the highest activity, with high selectivity for the formation of copolymerisation products (Table 1, entries 1-3).On the other hand,

Effect of the Metal
The effects of the metal (M) and co-axial ligand (X) on the catalytic activity and selectivity in the coupling reaction of cyclohexene oxide (CHO) and CO 2 using tetraphenylporphyrin (TPP) as a model N-donor ligand (Figure 1) were first appraised.The reactions were performed in the absence of solvent, under previously optimised conditions (50 bar CO 2 , 80 • C), using a 0.07 mol % catalyst [37], with and without bis(triphenylphosphine)iminium chloride (PPNCl) as a co-catalyst.The results are presented in Table 1.nitrobenzene/acetic acid mixture, using NaY zeolite as solid catalyst [39], followed by the complexation of p-CF3TPP with CrCl2 and in situ air oxidation [40] in refluxing DMF (Scheme 1).

Effect of the Metal
The effects of the metal (M) and co-axial ligand (X) on the catalytic activity and selectivity in the coupling reaction of cyclohexene oxide (CHO) and CO2 using tetraphenylporphyrin (TPP) as a model N-donor ligand (Figure 1) were first appraised.The reactions were performed in the absence of solvent, under previously optimised conditions (50 bar CO2, 80 °C), using a 0.07 mol % catalyst [37], with and without bis(triphenylphosphine)iminium chloride (PPNCl) as a co-catalyst.The results are presented in Table 1.Under these conditions, all of the TPP-based metalloporphyrin catalysts required the use of a co-catalyst to promote the coupling reaction of CHO with CO2.Among them, Cr(III) catalytic systems 1 and 2 (either with Cl or OAc as co-axial ligands) presented the highest activity, with high selectivity for the formation of copolymerisation products (Table 1, entries 1-3).On the other hand, Under these conditions, all of the TPP-based metalloporphyrin catalysts required the use of a co-catalyst to promote the coupling reaction of CHO with CO 2 .Among them, Cr(III) catalytic systems 1 and 2 (either with Cl or OAc as co-axial ligands) presented the highest activity, with high selectivity for the formation of copolymerisation products (Table 1, entries 1-3).On the other hand, the iron(III) and zinc(II) complexes were less active, but highly selective for the cycloaddition reaction, producing exclusively the cyclohexane carbonate (CHC).

Effect of the Fluorinated Groups in Cr(III)-Porphyrin Catalysts
The catalytic performance of fluorinated Cr(III)-porphyrin catalysts was then investigated in cyclohexene oxide/CO 2 copolymerisation.For the sake of comparison, the three catalysts 5, 6, and 7 were tested under the same reaction conditions (50 bar CO 2 , 80 • C), to evaluate the effects of the meso-arylporphyrin's different fluorinated groups (Table 2).the iron(III) and zinc(II) complexes were less active, but highly selective for the cycloaddition reaction, producing exclusively the cyclohexane carbonate (CHC).

Effect of the Fluorinated Groups in Cr(III)-Porphyrin Catalysts
The catalytic performance of fluorinated Cr(III)-porphyrin catalysts was then investigated in cyclohexene oxide/CO2 copolymerisation.For the sake of comparison, the three catalysts 5, 6, and 7 were tested under the same reaction conditions (50 bar CO2, 80 °C), to evaluate the effects of the meso-arylporphyrin's different fluorinated groups (Table 2).From the results presented in Table 2, we see that catalyst 7, in the presence of PPNCl as co-catalyst, has efficiently catalysed the ring-opening copolymerisation of CHO and CO2, which proceeded with 86% conversion along 24 h, with practically exclusive formation of copolymers (99%) (Table 2, entry 1).A slightly lower conversion of 79% was obtained with catalyst 5, based on 2,6-difluorophenyl porphyrin, and the selectivity for copolymerisation products was 92%, also slightly lower than with catalyst 7 (Table 2, entry 2).However, the application of perfluorinated catalyst 6 under the same conditions has led to a significantly lower conversion (only 50% in 24 h), with 91% of selectivity for copolymers (Table 2, entry 3).Therefore, under these conditions, the trifluoromethyl containing catalyst 7 appears to be the most effective, both in terms of activity and selectivity.

Effect of Temperature and CO2 Pressure
The CO2/CHO copolymerisation reactions were carried out along 24 h, under conditions where the reaction mixture was initially homogeneous, i.e., a single homogeneous phase was observed in the reactor prior to polymerisation (Figure 2a), although polymer precipitation began after just a few hours (Figure 2b).The effects of varying reaction temperature and CO2 pressure were studied for catalyst 7, as shown in Table 3. From the results presented in Table 2, we see that catalyst 7, in the presence of PPNCl as co-catalyst, has efficiently catalysed the ring-opening copolymerisation of CHO and CO 2 , which proceeded with 86% conversion along 24 h, with practically exclusive formation of copolymers (99%) (Table 2, entry 1).A slightly lower conversion of 79% was obtained with catalyst 5, based on 2,6-difluorophenyl porphyrin, and the selectivity for copolymerisation products was 92%, also slightly lower than with catalyst 7 (Table 2, entry 2).However, the application of perfluorinated catalyst 6 under the same conditions has led to a significantly lower conversion (only 50% in 24 h), with 91% of selectivity for copolymers (Table 2, entry 3).Therefore, under these conditions, the trifluoromethyl containing catalyst 7 appears to be the most effective, both in terms of activity and selectivity.

Effect of Temperature and CO 2 Pressure
The CO 2 /CHO copolymerisation reactions were carried out along 24 h, under conditions where the reaction mixture was initially homogeneous, i.e., a single homogeneous phase was observed in the reactor prior to polymerisation (Figure 2a), although polymer precipitation began after just a few hours (Figure 2b).The effects of varying reaction temperature and CO 2 pressure were studied for catalyst 7, as shown in Table 3.

Effect of Temperature and CO2 Pressure
The CO2/CHO copolymerisation reactions were carried out along 24 h, under conditions where the reaction mixture was initially homogeneous, i.e., a single homogeneous phase was observed in the reactor prior to polymerisation (Figure 2a), although polymer precipitation began after just a few hours (Figure 2b).The effects of varying reaction temperature and CO2 pressure were studied for catalyst 7, as shown in Table 3.At a lower temperature (50 °C), the reaction proceeded with a slightly lower conversion (77%) than that obtained at 80 °C, but the selectivity for copolymers and the carbonate content was similar (Table 3, entries 1 and 2).This observation differs from the results obtained with a CrCl-TPFPP/DMAP catalytic system under supercritical conditions [20], where the outcome of the reaction was strongly temperature-dependent, with only oligomeric polyether formation being observed at lower temperatures (70 °C), while higher reaction temperatures (95-110 °C) were found, in general, to lead to higher CO2 incorporations and higher polymer yields.With the reaction temperature set on 80 °C, the effect of CO2 pressure was then evaluated in the range of 10-150 bar (Table 3, entries 2-6).We observed that, at 10 bar CO2, a conversion of 70% is achieved, with almost full selectivity for polymers, containing 99% of carbonate content (Table 3, entry 3).This result seems also to contrast with the results obtained with a CrCl-TPFPP/DMAP catalytic system, for which a low-pressure CO2 atmosphere resulted in only oligomeric products (Mn ~ 600 g•mol −1 ) obtained in poor yields and with low CO2 incorporation [20].In the range of 50-110 bar CO2, conversions are considerably higher (84-86%) than at 10 bar, with no significant effects being observed in this pressure range, either in catalytic activity (TON's = 1008-1224) or selectivity for copolymers (96-99%) (Table 3, entries 2-5).However, a further increase of CO2 pressure to 150 bar led to a significant decrease in polymer yield, resulting from lower catalytic activity (Table 3, entry 6).This outcome can be directly observed from Figure 3, which illustrates the influence of CO2 pressure in the isolated yields of the copolymer PCHC, showing a maximum value at 110 bar CO2, and a visible drop at 150 bar CO2.At a lower temperature (50 • C), the reaction proceeded with a slightly lower conversion (77%) than that obtained at 80 • C, but the selectivity for copolymers and the carbonate content was similar (Table 3, entries 1 and 2).This observation differs from the results obtained with a CrCl-TPFPP/DMAP catalytic system under supercritical conditions [20], where the outcome of the reaction was strongly temperature-dependent, with only oligomeric polyether formation being observed at lower temperatures (70 • C), while higher reaction temperatures (95-110 • C) were found, in general, to lead to higher CO 2 incorporations and higher polymer yields.With the reaction temperature set on 80 • C, the effect of CO 2 pressure was then evaluated in the range of 10-150 bar (Table 3, entries 2-6).We observed that, at 10 bar CO 2 , a conversion of 70% is achieved, with almost full selectivity for polymers, containing 99% of carbonate content (Table 3, entry 3).This result seems also to contrast with the results obtained with a CrCl-TPFPP/DMAP catalytic system, for which a low-pressure CO 2 atmosphere resulted in only oligomeric products (M n ~600 g•mol −1 ) obtained in poor yields and with low CO 2 incorporation [20].In the range of 50-110 bar CO 2 , conversions are considerably higher (84-86%) than at 10 bar, with no significant effects being observed in this pressure range, either in catalytic activity (TON's = 1008-1224) or selectivity for copolymers (96-99%) (Table 3, entries 2-5).However, a further increase of CO 2 pressure to 150 bar led to a significant decrease in polymer yield, resulting from lower catalytic activity (Table 3, entry 6).This outcome can be directly observed from Figure 3, which illustrates the influence of CO 2 pressure in the isolated yields of the copolymer PCHC, showing a maximum value at 110 bar CO 2 , and a visible drop at 150 bar CO 2 .This effect is probably due to the phase behaviour, since the increase of CO2 pressure leads to variations in both the density of the reaction mixture and the molar ratio of CHO to CO2.Therefore, the decrease in yield and molecular weight when the reaction is carried out at 150 bar CO2 pressure might be attributed to a reduction in effective CHO concentration, which occurs when the phase boundary for the system is crossed, in agreement with previous studies [20,46].Moreover, a notable effect of CO2 pressure was observed in the molecular weight of the polymers, with the highest number-average molecular weight (in the range of 12,500 g•mol −1 ) being obtained when the reaction is performed at a CO2 pressure of 70 bar (Table 3, entry 4), while the lowest polydispersity value (Mw/Mn = 1.25) is obtained when the reaction is carried out at a 50 bar CO2 pressure (Table 3, entry 2).In this case, the molecular weight of the polymer obtained showed a monomodal distribution, while in the rest of the other experiments bimodal distributions were obtained (see discussion in Section 2.2.4).

Characterisation of PCHC Copolymers
The reaction's crude was dried in vacuum at 100 °C for 5 h, to eliminate unreacted CHO.After washing with n-hexane to remove possible traces of CHC, the residue was filtered and dried under vacuum at 100 °C for 12 h.Finally, the purified copolymer was characterised by 1 H and 13 C NMR spectroscopy and gel permeation chromatography (GPC).
The 1 H NMR spectrum of PCHC (Figure 4) shows a signal at δ = 4.64 ppm, typical of the methine proton (H1 in PCHC) of the repeating oxycarbonyloxy unit, while a very narrow singlet at δ = 3.57 ppm corresponding to oxy units was scarcely perceptible, which confirmed that the polymers were comprised mostly of carbonate content (98%) and contained a diminutive percentage of ether linkages.3, entry 2).
The 13 C NMR spectrum (Figure 5), performed in CDCl3, presents a dominant C=O signal at δ = 153.9ppm, which has previously been assigned to the carbonate carbon atom of m-centred structural units ((mmm), (mmr), (rmr)), while minor resonances at δ = 153.2and 153.4 ppm can be assigned to r-centered units ((rrr), (rrm), (mrm)) [47][48][49], suggesting that the ring-opening copolymerisation mechanism occurs without a defined stereocontrol.This effect is probably due to the phase behaviour, since the increase of CO 2 pressure leads to variations in both the density of the reaction mixture and the molar ratio of CHO to CO 2 .Therefore, the decrease in yield and molecular weight when the reaction is carried out at 150 bar CO 2 pressure might be attributed to a reduction in effective CHO concentration, which occurs when the phase boundary for the system is crossed, in agreement with previous studies [20,46].Moreover, a notable effect of CO 2 pressure was observed in the molecular weight of the polymers, with the highest number-average molecular weight (in the range of 12,500 g•mol −1 ) being obtained when the reaction is performed at a CO 2 pressure of 70 bar (Table 3, entry 4), while the lowest polydispersity value (M w /M n = 1.25) is obtained when the reaction is carried out at a 50 bar CO 2 pressure (Table 3, entry 2).In this case, the molecular weight of the polymer obtained showed a monomodal distribution, while in the rest of the other experiments bimodal distributions were obtained (see discussion in Section 2.2.4).

Characterisation of PCHC Copolymers
The reaction's crude was dried in vacuum at 100 • C for 5 h, to eliminate unreacted CHO.After washing with n-hexane to remove possible traces of CHC, the residue was filtered and dried under vacuum at 100 • C for 12 h.Finally, the purified copolymer was characterised by 1 H and 13 C NMR spectroscopy and gel permeation chromatography (GPC).
The 1 H NMR spectrum of PCHC (Figure 4) shows a signal at δ = 4.64 ppm, typical of the methine proton (H 1 in PCHC) of the repeating oxycarbonyloxy unit, while a very narrow singlet at δ = 3.57 ppm corresponding to oxy units was scarcely perceptible, which confirmed that the polymers were comprised mostly of carbonate content (98%) and contained a diminutive percentage of ether linkages.This effect is probably due to the phase behaviour, since the increase of CO2 pressure leads riations in both the density of the reaction mixture and the molar ratio of CHO to CO2.Therefor e decrease in yield and molecular weight when the reaction is carried out at 150 bar CO2 pressu ight be attributed to a reduction in effective CHO concentration, which occurs when the pha undary for the system is crossed, in agreement with previous studies [20,46].Moreover, a notab fect of CO2 pressure was observed in the molecular weight of the polymers, with the highe mber-average molecular weight (in the range of 12,500 g•mol −1 ) being obtained when the reactio performed at a CO2 pressure of 70 bar (Table 3, entry 4), while the lowest polydispersity val w/Mn = 1.25) is obtained when the reaction is carried out at a 50 bar CO2 pressure (Table 3, entry this case, the molecular weight of the polymer obtained showed a monomodal distribution, wh the rest of the other experiments bimodal distributions were obtained (see discussion in Sectio 2.4).

Characterisation of PCHC Copolymers
The reaction's crude was dried in vacuum at 100 °C for 5 h, to eliminate unreacted CHO.Aft ashing with n-hexane to remove possible traces of CHC, the residue was filtered and dried und cuum at 100 °C for 12 h.Finally, the purified copolymer was characterised by 1 H and 13 C NM ectroscopy and gel permeation chromatography (GPC).
The 1 H NMR spectrum of PCHC (Figure 4) shows a signal at δ = 4.64 ppm, typical of t ethine proton (H1 in PCHC) of the repeating oxycarbonyloxy unit, while a very narrow singlet a 3.57 ppm corresponding to oxy units was scarcely perceptible, which confirmed that the polyme ere comprised mostly of carbonate content (98%) and contained a diminutive percentage of eth kages.The 13 C NMR spectrum (Figure 5), performed in CDCl3, presents a dominant C=O signal at δ 3.9 ppm, which has previously been assigned to the carbonate carbon atom of m-centr ructural units ((mmm), (mmr), (rmr)), while minor resonances at δ = 153.2and 153.4 ppm can signed to r-centered units ((rrr), (rrm), (mrm)) [47][48][49], suggesting that the ring-openin polymerisation mechanism occurs without a defined stereocontrol.
CO 2 pressure (bar) The 13 C NMR spectrum (Figure 5), performed in CDCl 3 , presents a dominant C=O signal at δ = 153.9ppm, which has previously been assigned to the carbonate carbon atom of m-centred structural units ((mmm), (mmr), (rmr)), while minor resonances at δ = 153.2and 153.4 ppm can be assigned to r-centered units ((rrr), (rrm), (mrm)) [47][48][49], suggesting that the ring-opening copolymerisation mechanism occurs without a defined stereocontrol.Catalysts 2017, 7, 210 7 of GPC analysis, performed in toluene, and calibrated against a polystyrene standard, allowed t calculation of the number-average molecular weights (Mn) of the copolymers, which were in t range of 4800-12,500 g•mol −1 .Furthermore, the molecular weight distributions (Mw/Mn) we determined to be in the range 1.25-1.63,which represent significant progress over distributio previously reported for some scCO2-based systems [20,50].GPC chromatograms of the copolyme (Figure 6) showed elution curves with unimodal (in the case of entry 2, Table 3) or bimod distributions (entries 1, 3-6, Table 3), which are indicative of a chain transfer process or a differe chain termination step.3, 2).GPC analysis, performed in toluene, and calibrated against a polystyrene standard, allowed the calculation of the number-average molecular weights (M n ) of the copolymers, which were in the range of 4800-12,500 g•mol −1 .Furthermore, the molecular weight distributions (M w /M n ) were determined to be in the range 1.25-1.63,which represent significant progress over distributions previously reported for some scCO 2 -based systems [20,50].GPC chromatograms of the copolymers (Figure 6) showed elution curves with unimodal (in the case of entry 2, Table 3) or bimodal distributions (entries 1, 3-6, Table 3), which are indicative of a chain transfer process or a different chain termination step.GPC analysis, performed in toluene, and calibrated against a polystyrene standard, allowed the calculation of the number-average molecular weights (Mn) of the copolymers, which were in the range of 4800-12,500 g•mol −1 .Furthermore, the molecular weight distributions (Mw/Mn) were determined to be in the range 1.25-1.63,which represent significant progress over distributions previously reported for some scCO2-based systems [20,50].GPC chromatograms of the copolymers (Figure 6) showed elution curves with unimodal (in the case of entry 2, Table 3) or bimodal distributions (entries 1, 3-6, Table 3), which are indicative of a chain transfer process or a different chain termination step.3).

MALDI-TOF Analysis and Mechanistic Considerations
The matrix assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF) mass spectra of the polymers resulting was performed in order to analyse the chain end groups of the polycarbonates and draw conclusions about the probable reaction mechanism.
The MALDI-TOF mass spectra of the polymers obtained present repeating units of 142 m/z corresponding to a cyclohexylcarbonate -C6H10C(O)O-fragment (Figures S1-S6, Supplementary Information).In the case of the polymer obtained from entry 2 in Table 3, the main distribution presented peaks at m/z 5981.6,6124.1, and 6266.0 (Figure S2) that correspond to chains containing  3).

MALDI-TOF Analysis and Mechanistic Considerations
The matrix assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF) mass spectra of the polymers resulting was performed in order to analyse the chain end groups of the polycarbonates and draw conclusions about the probable reaction mechanism.
The MALDI-TOF mass spectra of the polymers obtained present repeating units of 142 m/z corresponding to a cyclohexylcarbonate -C 6 H 10 C(O)O-fragment (Figures S1-S6, Supplementary Information).In the case of the polymer obtained from entry 2 in Table 3, the main distribution presented peaks at m/z 5981.6,6124.1, and 6266.0 (Figure S2) that correspond to chains containing two -OH terminal groups (a + K in Scheme 2, expected for n = 41-43 HO(C 7 H 10 O 3 ) 13 C 6 H 10 OH m/z 5981. 1, 6123.2, 6265.3).This suggests that the initiation step involves the epoxide opening by a nucleophilic attack, with -OH arising from water traces present in the reactor, and water is also involved in the final step to form a -OH chain end (Scheme 2).
Minor peaks distributions fitted with the presence of Cl -as end group as well as -OH (b, Scheme 2, observed m/z 5963.79,6105.63 (Figure S2); expected for n = 41, 42; m/z 5960.6,6102.7).In this case the presence of a main chain corresponded with the observation of a single band in the GPC analysis and low M w /M n polydispersity.
In the experiments run at different pressure and temperature (entries 1,3-6, Table 3), which presented bimodal distribution in the GPC analysis, the MALDI-TOF mass spectra showed mixtures of three species (Figures S1, S3-S6, Supplementary Materials).The results indicated chains a and b were present (Scheme 2), and the third peak distribution with peaks observed at m/z 6000.91 and 6141.42 may correspond to polymers with two -Cl final groups (c + Na, Scheme 2, expected for n = 41, 42; m/z 6000.9, 6141.4).

Reagents
Cyclohexene oxide was purchased from Sigma-Aldrich, dried over alumina and stored under Scheme 2. Proposed mechanism for the polymerization of CHO with CO 2 using 7/PPNCl catalytic system.
Catalysts 2017, 7, 210 Consequently, the mechanism may involve the opening of the epoxide, presumably after coordinating to the metal center, by the Cl − from PPNCl or OH − from water.The alkoxo species thus formed inserts CO 2 and produces the cyclic carbonate by intramolecular attack, generating the nucleophile that, in this case, may be Cl − since OH − is not a good leaving group (i, in Scheme 2).
An alternate insertion of epoxide and CO 2 leads to the polycarbonate growing chain (ii, in Scheme 2).As evidenced by the chain end group analysis, termination may occur by different pathways, including by the combination of two growing chains or by the hydrolysis of water.The polymer obtained under optimal reaction conditions (80 • C and 50 bar CO 2 ) shows a monomodal distribution, which suggests that the chain transfer, responsible for different length polymers, is probably a minor process.Instead, using higher CO 2 pressures, precipitation of the polymer in the compressed CO 2 medium may generate the formation of different chains' length, which explains the bimodal molecular weight distributions.At lower CO 2 pressure (10 bar) or temperature (50 • C), the lower conversions obtained may induce a similar effect.

Reagents
Cyclohexene oxide was purchased from Sigma-Aldrich, dried over alumina and stored under inert atmosphere.Carbon dioxide (SCF Grade, 99.9993%, Abelló Linde, Barcelona, Spain) was previously passed by an Agilent oxygen/moisture trap.The rest of reagents and solvents were purchased form Sigma-Aldrich (Madrid, Spain) unless otherwise stated.

Equipment
The 1 H NMR and 13 C NMR spectra were recorded in a 400 MHz Mercury instrument (VARIAN, Palo Alto, CA, USA), in CDCl 3 using tetramethylsilane as internal standard.Gel permeation chromatography (GPC) measurements (URV, Tarragona, Spain) were made in toluene, versus polystyrene standards, on a Millipore-Waters 510 HPLC Pump device (Milford, MA, USA) using a three-serial column system (MZ-Gel 100 Å, MZ-Gel 1000 Å, MZ-Gel 10,000 Å linear columns, Millipore-Waters, Milford, MA, USA) with UV-Detector (ERC-7215) and IR-Detector (ERC-7515a, Millipore-Waters, Milford, MA, USA).The software used to get the data was NTeqGPC 5.1 (Millipore-Waters, Milford, MA, USA).Samples were prepared as follows: 5 mg of the copolymer was dissolved with 2 mL of toluene (HPLC grade).MALDI-TOF analyses were performed on a Voyager System 4412 instrument equipped with a 337 nm nitrogen laser (Applied Biosystems, Foster City, CA, USA).All spectra were acquired in the positive ion reflector mode.Dithranol (Sigma-Aldrich, Madrid, Spain) was used as matrix, which was dissolved in MeOH (Panreac, Barcelona, Spain) at a concentration of 10 mg•mL −1 .The polymer (5 mg) was dissolved in 1 mL of CHCl 3 (Euriso-Top, Saint-Aubin, France).One microliter (1 µL) of the sample, 1 µL of the matrix, and 1 µL of potassium trifluoroacetate (KTFA) solution in the case of polymers (1 mg of KTFA in 1 mL of THF) were deposited consecutively on the stainless steel sample holder and allowed to dry before introduction into the mass spectrometer Three independent measurements were made for each sample.For each spectrum, 100 laser shots were accumulated.Elemental analysis were done at the Serveis Tècnics de Recerca (Universitat de Girona, Spain).

Preparation of Metalloporphyrin Catalysts
The porphyrins 5,10,15,20-tetraphenylporphyrin (TPP), 5,10,15,20-tetra(2,6-difluorophenyl) porphyrin (TDFPP), 5,10,15,20-tetra(2,3,4,5,6-pentafluorophenyl)porphyrin (TPFPP), and 5,10,15,20-tetra(4-trifluoromethylphenyl)porphyrin (p-CF 3 TPP) were prepared by reacting pyrrole with the corresponding aldehydes in equimolar amounts, in a nitrobenzene/acetic acid mixture, according to previously reported methods [38,39,51,52].Then, the resulting free-base porphyrins were reacted with the desired metal salt to achieve the corresponding metalloporphyrin catalysts 1-7 [53].Synthesis of 5,10,15,20-tetrakis(4-trifluoromethylphenyl)porphyrinatochromium(III) Chloride (7) The free base porphyrin p-CF 3 TPP [51] (0.400 g, 0.45 mmol) was dissolved in 15 mL of dimethylformamide (DMF).The mixture was stirred under reflux at 170 • C.After approximately 10 min, CrCl 2 (0.095 g, 0.77 mmol) was added to the refluxing solution.After 30 min, an aliquot was taken from the mixture and analysed by UV-VIS spectroscopy, which indicated that free p-CF 3 TPP still remained in solution.So, a further amount of CrCl 2 (0.073 g, 0.59 mmol) was added to the solution, refluxing for another 20 min, until no free p-CF 3 TPP was found by UV-VIS.At the end, the reaction mixture was allowed to cool to room temperature and poured into 400 mL of ice-cold water.After the solid was filtered out and washed three times with water, it was dried under vacuum at 100 • C. The crude product was purified by column chromatography over alumina with CHCl 3 as the eluent.After solvent evaporation, the product was dried overnight under vacuum at 100 • C, yielding the chromium porphyrin complex CrCl-p-CF 3 TPP (7) as a dark green solid, in 95% yield (0.415 g, 0.43 mmol).UV-VIS (CH 2 Cl 2 ): λ max /nm (ε/dm 3  The reactions were carried out in a 100 mL Berghof stainless steel autoclave (Berghof, Eningen, Germany).The catalyst (0.07%) and co-catalyst (0.07%) (when indicated) were placed inside the autoclave and kept for 3 h, under vacuum, at 80 • C.Then, the epoxide substrate (4 mL), previously dried over alumina, was injected into the autoclave via cannula.The autoclave was then pressurized with CO 2 , and the reaction proceeded at the desired temperature.At the end of the reaction, the autoclave was cooled and slowly depressurized.The % of conversion was determined by 1 H NMR of the crude mixture, using mesitylene as standard.Selectivity was calculated by integral ratio between polycarbonate and cyclic carbonate.The work-up was performed only when polymeric products were obtained, and is described below.

Conclusions
Metal complexes of meso-arylporphyrins (Cr(III), Zn(II), and Fe(III)) have efficiently catalysed the coupling reaction between CO 2 and cyclohexene oxide, with Cr(III)porphyrin 7/PPNCl being the most active and selective for the formation of copolymers.This trifluoromethyl-based catalytic system is an attractive alternative to other fluorinated catalysts, considering its easy synthesis, high selectivity for copolymerisation, and high solubility in the substrate and compressed carbon dioxide medium, avoiding the use of any additional harmful solvent.Using this new catalytic system, the CO 2 pressure was found to affect the outcome of the reaction, with the best copolymer yield being obtained at 110 bar CO 2 , the higher number-average molecular weight (12,500 g•mol −1 ) being obtained at a CO 2 pressure of 70 bar, and the lowest polydispersity value (1.25) at 50 bar CO 2 pressure.The formed copolymers

Figure 3 .
Figure 3.Effect of CO2 pressure in isolated yields of the copolymer PCHC.

Figure 3 .
Figure 3.Effect of CO 2 pressure in isolated yields of the copolymer PCHC.

Figure 3 .
Figure 3.Effect of CO2 pressure in isolated yields of the copolymer PCHC.

Table 1 .
Metal effect on the selectivity of coupling reactions of CO2 with cyclohexene oxide.

Table 1 .
Metal effect on the selectivity of coupling reactions of CO 2 with cyclohexene oxide.a

Table 1 .
Metal effect on the selectivity of coupling reactions of CO2 with cyclohexene oxide.

Table 2 .
Effect of meso-arylporphyrin's fluorinated groups on catalytic activity and selectivity.a

Table 2 .
Effect of meso-arylporphyrin's fluorinated groups on catalytic activity and selectivity.a

Table 3 .
Effects of temperature (T) and pressure (P) in CO 2 /cyclohexene oxide copolymerisation using catalyst 7. a

Table 3 .
Effects of temperature (T) and pressure (P) in CO2/cyclohexene oxide copolymerisation using catalyst 7. a

3 f M w /M n f
a Reaction conditions: CHO (39.5 mmol, 4 mL); catalyst 7: 0.07 mol %; co-catalyst: PPNCl 0.07 mol %; t = 24 h.b % Conversion determined by 1 H NMR. c turnover number calculated as mol (converted substrate) /mol (catalyst) .dIsolatedyield for poly(cyclohexane)carbonate, based on weighted mass of obtained product, relatively to weighted mass of epoxide and CO 2 .e%CO 2 content in polymers determined by1H NMR integral ratio of carbonate linkages/(carbonate linkages+ether linkages).f Number average molecular weight and polydispersity determined by GPC, using polystyrene as standard.