Synthesis of high-molecular-weight aliphatic polycarbonates from diphenyl carbonate and aliphatic diols by solid base

Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

Contents lists available at ScienceDirect

Journal of Molecular Catalysis A: Chemical journal homepage: www.elsevier.com/locate/molcata

Synthesis of high-molecular-weight aliphatic polycarbonates from diphenyl carbonate and aliphatic diols by solid base Wang Ziqing a,b,c , Yang Xiangui a,c , Li Jianguo a,c , Liu Shaoying a,c , Wang Gongying a,c,∗ a b c

Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences, Chengdu 610041, Sichuan, China University of Chinese Academy of Sciences, Beijing 100049, China Changzhou Institute of Chemistry, Changzhou 213164, Jiangsu, China

a r t i c l e

i n f o

Article history: Received 8 May 2016 Received in revised form 4 August 2016 Accepted 14 August 2016 Available online 16 August 2016 Keywords: Transesterification Diphenyl carbonate Aliphatic polycarbonate Magnesium oxide Basic sites

a b s t r a c t Various solid bases were synthesized and used as catalyst for direct transesterification of diphenyl carbonate (DPC) with aliphatic diols to synthesize high-molecular-weight aliphatic polycarbonates (APCs). Physical-chemical properties of these catalysts were characterized by means of different techniques to show the relationship between catalyst structure and catalytic performance. MgO prepared via simple coprecipitation method was found to be the most active catalyst among all the solid bases and magnesium compounds tested in the present study. The high-molecular-weight poly(1,4-butylene carbonate) (PBC) could be easily obtained over this MgO with Mw up to 182 200 g/mol under its optimization condition. Additionally, the amplification experiment in a 5.0 L stainless steel reactor also verified the reliability of this transesterification process with MgO as catalyst, giving maximum Mw value for PBC as high as 208 600 g/mol. It was found that medium and strong basic sites were responsible for this transesterification process. Simultaneously, strong basic sites also could favor the decomposition and depolymerization of the resultant PBC, leading to the decrease for Mw and yield at higher temperature. In addition, based on the experimental results and relevant literature, a plausible reaction mechanism involving the activation of diols via the abstraction of proton with basic site over MgO was proposed for this process. © 2016 Elsevier B.V. All rights reserved.

1. Introduction Aliphatic polycarbonates (APCs) have been used for a variety of applications in the field of biological medicine owing to their outstanding biocompatibility and biological reactivity [1]. They can also be used as raw material for waterborne polyurethane industry because of their good hydrolysis and oxidation resistance properties [2]. More importantly, APCs with number-average molecular weight (Mw ) greater than 70000 g/mol have also been thought to be promising alternatives to petrochemical monomers as biodegradable plastic [3,4]. Traditionally, these APCs can be synthesized via copolymerization of CO2 with epoxides and ring-opening polymerization of cyclic carbonates. Unfortunately, these routes are still suffering from various drawbacks, such as poor structure for resultant polymer, low effective for current catalysts as well as high cost

∗ Corresponding author at: Chengdu Institute of Organic Chemistry, Chinese Academy of Sciences, Chengdu 610041, Sichuan, China. E-mail addresses: [email protected], [email protected] (G. Wang). http://dx.doi.org/10.1016/j.molcata.2016.08.017 1381-1169/© 2016 Elsevier B.V. All rights reserved.

for raw feedstocks, which all restrict the production and application of these APCs on a large scale [5,6]. The melt transesterification of dimethyl carbonate (DMC) with aliphatic diols via a two-step condensation polymerization route has been regarded as a clean and sustainable synthetic route for the production of APCs with diverse structures, since this route is a high efficiency solvent-free process [7–10]. Recently, a simpler one-pot method was also developed in our previous work by replacing DMC with high boiling-point diphenyl carbonate (DPC), in which the by-product, phenol, can be easily separated and reused [11]. No matter which process we selected, highly effective catalyst is a key factor for the preparation of APCs with satisfactory Mw value. Therefore, a wide range of compounds, including metal salts [8,11–13], BMIM-2-CO2 [10], TiO2 /SiO2 (PVP) (TSP-44) [4,7], organic base [14] and even enzyme[15], have shown great promise for this reaction. Up to now, the most efficient catalysts proposed for this process have been confined to homogeneous basic catalysts [8,12]. However, the thorny issue associated with such strong bases is the presence of undesirable side reactions, and the residual of these catalysts also could decrease the thermal stability of polymer. Moreover, the structure-activity correlation of these basic catalysts for this process has not been well elucidated. In this con-

78

Z. Wang et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

text, it is highly desired to develop a new catalyst that can allow an excellent activity and well understand its action mechanism for this transesterification reaction. Most recently, magnesium compounds have been widely used as catalysts for the synthesis of sustainable polymers from renewable resources [16]. Particularly, MgO as a typical solid base catalyst has been extensively used in many transesterification processes. Using a triblock copolymer as soft template, Lee et al. have prepared a series of MgO through a surfactant-assisted route, which showed excellent catalytic performance in the transesterification of DMC with glycerol for the synthesis of glycerol carbonate (GLC) [17]. Wherein, the catalytically active sites were attributed to the higher basic site concentration on their surface. A K-doped MgO catalyst was also reported by Lago et al. for biodiesel synthesis through transesterification of soybean oil with methanol [18]. Also, magnesium-based mixed oxides (e.g. Mg-Al mixed oxides) have been found to be able to catalyze this reaction, in which the concentration of basic sites, specific surface area and crystallite size of MgO were thought to responsible for their excellent activity [19]. Di Serio and co-workers found that MgO was also the best single catalyst for the transesterification of dimethyl terephthalate with ethylene glycol for the production of poly(ethylene terephthalate) (PET)[20]. And the obtained PET polymer has chemical and physical properties very close to that of commercial sample. Additionally, MgO-catalyzed transesterification reaction is often used as a valuable tool for polymer modification due to its nontoxic and low cost [21]. Moreover, MgO itself is also outstanding inorganic filler for many polymers. The addition of MgO not only can improve their mechanical properties but also can provide them with new application performance [22,23]. In the present work, we also found that MgO exhibited remarkably higher activity than other solid bases and magnesium compounds for direct transesterification of DPC with aliphatic diols to synthesize APCs. Combining a variety of characterization results, the relationship between catalyst structure and catalytic performance was investigated by comparing MgO with other solid bases. Additionally, a plausible reaction mechanism involved the promotion effect of basic sites was also proposed.

2. Experimental 2.1. Materials and preparation of catalyst Commercial DPC was purified by recrystallization in absolute ethyl alcohol before use. All aliphatic diols were dehydrated by distillation over calcium hydride under dry nitrogen gas. Other reagents were all purchased from Chengdu Kelong Chemical Reagent Co., Ltd. (in China) and used as received without further treatment. In this study, two kinds of MgO were synthesized: direct thermal decomposition of Mg(NO3 )2 ·6H2 O (MgO-T) and coprecipitation of Mg(NO3 )2 ·6H2 O with Na2 CO3 (MgO-P). In direct calcination method, Mg(NO3 )2 ·6H2 O was calcined at 500 ◦ C in air atmosphere for 4.0 h. In coprecipitation method, Na2 CO3 (0.1 mol) in 50 g distilled water was dropwise added into the solution of Mg(NO3 )2 ·6H2 O until the pH reach 10.0 and stirred for 4.0 h at room temperature. The obtained white precipitate was filtered and washed with hot water until the pH value to 7.0, dried at 100 ◦ C overnight, and subsequently calcined at given temperature for 4.0 h. ZnO and ZrO2 were also obtained by thermal decomposition of Zn(OH)2 and Zr(OH)4 at 500 ◦ C for 4.0 h according to the same procedure as MgO-P, while CaO was prepared by calcination of calcium carbonate at 900 ◦ C in N2 for 4.0 h. The commercial source MgO (AR) labeled as MgO-C was also supplied by Chengdu Kelong Chemical Reagent Co., Ltd. To obtain MgO surface with dif-

ferent basic site amount, a series of Mg(Al)O samples with different content of Al were prepared by incipient impregnation route as described in literature [24]. Typically, a certain amount of aluminum isopropoxide was dissolved in 30 mL of benzene at 50 ◦ C. Then the solution was quickly added into Mg(OH)2 , dried at 100 ◦ C overnight and thermally decomposed in N2 at 500 ◦ C to get the impregnated catalyst. 2.2. Transesterification reaction The reaction was conducted in a 150 mL round-bottomed flask, equipped with a mechanical stirrer, reflux condenser and thermometer. In a typical process, DPC (100 mmol, 21.41 g), BD (100 mmol, 9.01 g) and a catalyst were successively charged into the reactor under N2 atmosphere. The reaction mixture was heated at 120 ◦ C and stirred continually for 30 min until it became homogeneous. Before starting the polymerization process, the reaction pressure was slowly reduced to 200 Pa. Then the temperature was further increased to a given value and maintained for a certain time to carry out the polymerization reaction. After reaction, the reaction mixture was cooled to room temperature and the residue was purified by dissolving in CH2 Cl2 and precipitating with ethanol. The PBC polymer was isolated by centrifugation and dried under vacuum at 65 ◦ C for 4.0 h. 2.3. Characterization of obtained polymers and catalyst Mw and PDI of polymers were determined by gel permeation chromatography (GPC). The GPC measurement was carried out at 30 ◦ C on a Waters 515 HPLC system equipped with a 2690D separation module and a 2410 refractive index det. Tetrahydrofuran (THF) was used as eluent at a flow rate of 0.5 mL/min. Polystyrene with a narrow molecular weight distribution was used as standard for calibration. XRD analysis was carried out on a PANalytical X’pert Pro diffractometer system, using CoKa radiation (0.1789 nm). N2 physisorption was measured on the gas adsorption instrument Nova Win2, then the surface area was obtained by BET equation. The basicity of the catalysts was evaluated by temperatureprogrammed desorption of carbon dioxide (CO2 -TPD) using an AutoChem 2910 instrument (Micromeritics) and the detector was MSC-200 quadruple mass analyzer from Balzer Company, Ltd. The sample was treated at 500 ◦ C for 2.0 h under a flow of He atmosphere (flow rate of 30 mL/min). After being cooled down to room temperature, pure CO2 was adsorbed at room temperature for 1.5 h. Then the physisorbed CO2 was removed by a flushing with He at room temperature for 1.5 h. TPD was carried out in the stream of He (20 mL/min) at a heating rate of 10 ◦ C up to 500 ◦ C (all the samples were 100 mg). 3. Results and discussion 3.1. Screening of catalyst The typical performances of different solid base catalysts in the transesterification of DPC with BD at given conditions are summarized in Table 1. One can see that ZrO2 , MgO-T and CaO were all active to this reaction at the same condition, giving PBC with Mw of 21 300, 36 600 and 47 800 g/mol, respectively. However, the polymerization rate of ZnO was so low that the Mw of its PBC polymer could not be detected under the given condition. Additionally, the catalytic performance and preparation method appeared to be significantly related for MgO, and Mw of PBC over MgO-P (115 200 g/mol) is much higher than those of MgO-C and MgO-T. Moreover, one also can find that the catalytic activity of MgO-P in this process is also evidently comparable to those of other magnesium

Z. Wang et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

79

Table 1 The catalytic performance of different catalysts for the transesterification and polymerization of DPC with BDa . Entry

Catalyst

Mw (×10−3 g/mol)

PDI(Mw /Mn )

Yield (%)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

ZnO MgO-T MgO-P MgO-C ZrO2 CaO Mg(NO3 )2 Mg(OH)2 MgCO3 MgSO3 Mg(OAc)2 c MgCl2 MgO-Pd MgO-Pe MgO-Pf

n.d.b 36.6 115.2 98.6 21.3 47.8 28.6 82.5 106.0 n.d. 102.0 54.1 51.6 125.1 109.8

n.d. 1.68 1.76 1.75 1.65 1.74 1.64 1.76 1.75 n.d. 1.78 1.73 1.59 1.78 1.78

n.d. 91.4 89.4 90.6 92.3 72.6 90.2 78.6 88.2 n.d. 89.5 92.0 92.5 83.2 78.6

a Reaction were carried out at 180 ◦ C and 200 Pa for 120 min and the molar ratio of catalyst to DPC was set at 0.1 mol%. b No detection. c Take from literature [13]. d Molar ratio of MgO to DPC 0.05 mol%. e Molar ratio of MgO to DPC 0.2 mol%. f Molar ratio of MgO to DPC 0.4 mol%.

compounds including Mg(OH)2 , Mg(NO3 )2 , MgCl2 and Mg(OAc)2 . Of course, not all magnesium compounds were active to this reaction; for instance, no any polymer and fraction could be detected in MgSO4 catalysis system. According to previous literature [13,25], the activity of metal salt for this reaction was closely related to their basicity of anion and the chelating ability of the metal species. Therefore, the poor catalytic performance of MgSO4 was mainly resulted from the low basicity of SO4 2− . Under the same condition, MgCO3 also possessed a relatively high performance with Mw of 106 000 g/mol, which was still inferior to that of MgO-P. However, the cost of MgO is also the cheapest among all the magnesium compounds and it can be easily available at industrial scale. Based on the results of entry 11–13 in Table 1, 0.1mol% of MgO is sufficient for this process.

(f) (e) (d) (c) (b) (a) 30

40

50

60

70

80

2 theta(degree) Fig. 1. XRD patterns of (a) ZnO, (b) ZrO2 , (c) CaO, (d) MgO-P, (e) MgO-C and (f) MgO-T.

(f) (e) (d) (c) (b) (a) 100

200

300

400

500

600

700

o

Temperature ( C) 3.2. Characterization of PBC As shown in Fig. S1, both FT-IR and 1 H NMR spectra of resultant copolymers are very close each other. All peaks of these polymers can be assigned to previously published spectrum for PBC and it is consistent with that expected for PBC structure [4,7–11]. However, obvious features for end-group can be detected in the 1 H NMR spectrum of PBC over ZrO2 with their chemical shift at 3.64 and 7.32–7.38 ppm for −CH2 OH and −OC6 H5 , respectively, suggesting that this polymer has rather low molecular weight. Additionally, Fig. S1(C) shows that the color of the PBC sample together with its polymer solution obtained over CaO is light yellow, while other samples are white solids. This result indicated that the use of CaO, a typical strong base, would induce some undesirable side reactions and lead to the colorization of the resultant polymers. 3.3. Characterization of catalysts XRD patterns of these solid bases obtained at given condition are shown in Fig. 1, one can see that ZrO2 and ZnO can be easily synthesized via the thermal dissociation of corresponding hydroxides at a high temperature. XRD pattern in Fig. 1(c) is cubic CaO without observation of CaCO3 . Fig. 2(d–f) reveal that these three samples have a clear MgO crystalline phase, though crystallite size of the MgO-C is smaller than those of MgO-P and MgO-T later. To elucidate these differences in catalytic performance with respect to the specific surface area (SBET ) of different catalysts, the textural properties

Fig. 2. CO2 -TPD profiles of various solid base catalysts (a) ZnO, (b) ZrO2 , (c) MgO-T, (d) MgO-P, (e) MgO-C and (f) CaO.

of these oxides were also measured by N2 adsorption-desorption. It can be observed in Table 2 that SBET of these six samples follows the sequence MgO-P > MgO-C > ZrO2 > MgO-T > ZnO > CaO, indicating that higher SBET seems to be in favor of the increase in catalytic performance. It is well known that transesterification reaction is a classical base catalysis process, in which the activity of a solid base is closely related to its basic site strength and amount [26–28]. To understand the catalytic activity of different catalysts, the basic site strength and amounts of these solid catalysts were measured by CO2 -TPD experiment, and the results are shown in Fig. 2, the amounts of chemisorbed CO2 over these samples are also listed in Table 2. According to literature [29], the amount of weak, medium and strong basic sites can be expressed in ␮mol desorbed CO2 /g over the range of 20–160, 160–400, and >400 ◦ C, respectively. As depicted in Fig. 2, only a trace amount of CO2 desorption can be observed for ZnO at a low temperature, which is ascribed to CO2 desorption on weak basic sites over ZnO, while CaO exhibits a sharp peak at 613 ◦ C for strong basic sites, and the amount of CO2 adsorbed on CaO is approximate 274 ␮mol/g. It is well accepted that the weak basic sites are probably associated with Brönsted basicity and generated by the surface hydroxyl (OH) groups, while the strong sites are probably associated with the surface isolated O2− anions [30].In

80

Z. Wang et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

Table 2 BET surface area and surface basicity of various solid bases. Catalysts

SBET (m2 /g)

ZnO ZrO2 MgO-T MgO-C MgO-P CaO

12.6 42.3 25.0 94.6 148.5 9.4

Desorption peaks (area%) weak

medium

strong

92.0 78.5 42.4 46.3 28.6 12.2

8.0 21.5 57.6 53.7 66.5 0

0 0 0 0 4.9 87.8

Total evolved CO2 (␮mol/g)

Basic site density (␮mol CO2 /m2 )

11 48 62 262 389 274

0.9 1.1 2.5 2.8 2.6 29.2

Fig. 3. (A) The CO2 -TPD of Mg(Al)O samples with Al/Mg molar ratio of (a)0.01 (b) 0.05 (c) 0.1 and (d) 0.2 and (B) the relationship between Mw and the total amount of medium basic sites over Mg(Al)O. Reaction conditions: temperature 180 ◦ C, pressure 200 Pa, time 2.0 h and catalyst concentration of 0.01 wt%.

SBET

100

Mw 150

2

80 60

100

40

50

Mw (Kg/mol)

200

SBET(m /g)

terms of MgO and ZrO2 , besides of low temperature peaks below 160 ◦ C, a very broad CO2 desorption peak also can be seen in the temperature range of 160–400 ◦ C, which could be assigned to the existence of the medium basic sites, associated to oxygen in Mg2+ O2− and Zr2+ -O2− pairs, respectively [29,31]. The amount of CO2 adsorbed on MgO-P was approximate 389 ␮mol/g, which is more than 7 and 8 times of those adsorbed over MgO-T (66.5 ␮mol/g) and ZrO2 (48 ␮mol/g), respectively. It is widely reported that transesterification reaction can also be promoted by acid sites [32]. Therefore, to examine the possible involvement of the acid property for this reaction, the NH3 -TPD measurement (See Supporting information for the detailed experiment) was also conducted to probe the activity of various solid bases above. One can see in Fig. S2, no obvious relevance can be found between the acid density and catalytic activity for these catalysts. Based on the results in Tables 1 and 2, one can see that MgO-P with the maximum amount of medium basic sites exhibits the highest polymerization rate, while the weak base, ZnO, was nearly inactive to this reaction at the same condition. Additionally, both Mw and yield of the obtained PBC over CaO were all inferior to MgO-P despite its strongest basicity compared to other catalysts. This result illustrates that the presence of strong basic sites may be disadvantageous to the improvement of Mw and corresponding yield. Therefore, it is concluded that basic site strength and amount both played an important role in this process, and the rich medium basic sites of MgO-P may be responsible for its excellent catalytic performance. In order to further insight into the basic property of MgO catalyst in relation to its activity in this transesterification reaction, a series of MgO with different amount of medium basic sites was prepared by introduction of different content of Al. It was found in Fig. 3(A) that the amount of medium basic sites over MgO decreased with the increase of Al content, but the addition of Al has a little effect on their basic strength. When these catalysts were employed in this reaction at a lower temperature, many side reactions could be inhibited effectively. Therefore, the results of transesterification

20 0 300

400

500

600

700

800

o

Calcination temperature ( C) Fig. 4. The effect of calcination temperature on SBET and catalytic activity of MgO. Reaction conditions: temperature 180 ◦ C, pressure 200 Pa, time 2.0 h and catalyst concentration of 0.01 wt%.

of DPC and BD with lower concentration Mg(Al)O as catalyst are shown in Fig. 3(B). Under the same condition, as shown in Fig. S3 that the highest Mw value is 150 600 g/mol, and these values in Fig. 3(B) did not reach this highest level. It can be seen that Mw could be linearly correlated with the total amount of medium base sites. Thus, this result can further confirm our hypothesis that basic sites with moderate strength can favor the transesterification of −OH and −OC(O)OC6 H5 end-group at a reduced pressure. In order to elucidate the dependence of the catalytic performance on the SBET of MgO, a series of MgO samples with different SBET were obtained by varying their calcination temperature. Fig. 4 illustrates the effect of calcination temperature on the SBET and catalytic performance of MgO. Obviously, SBET decreased with the increase of calcination temperature in the all range, which would

Z. Wang et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

81

Fig. 5. Effect of reaction temperature on (a) Mw and (b) yield of PBC polymer obtained over different solid base catalysts. Reaction were carried out at 200 Pa for 2.0 h and the molar ratio of catalyst to DPC was set at 0.1 mol%.

cause a negative influence to our reaction. The effect of calcination temperature on the base strength and basic site amount over MgO was also investigated, as shown in Fig. S4 that CO2 uptake desorbed from medium basic sites decreased gradually with the increase of calcination temperature. However, considering the variation of these MgO samples in terms of their textual properties, the basicity intensity for each catalyst is further calculated regarding SBET . These values for resultant intensity of MgO are fairly close, which is also in good agreement with previous literature [33]. Based on the discussion above, the variation of SBET for these MgO mainly originates from their basicity amount, and the higher SBET will offer abundant exposed basic sites for reaction proceed. As discussion in many studies with solid base as catalyst that the major control factors for transesterification reaction should be its base strength and basic site amount, SBET and crystallite size [19,26,34]. Therefore, calculated by Scherer equation based on the (1 0 1) plane in Fig. S5 (A), the relation between crystallite size of MgO and the Mw for its PBC was also discussed and the result is displayed in Fig. S5(B). Obviously, crystallite size of MgO samples significantly increased with elevating calcination temperature in all the range, leading to a decrease in Mw for PBC. 3.4. Effect of reaction conditions In order to obtain the optimum reaction condition and further understand the structure-reactivity correlations of solid base catalysts for this reaction, the transesterification of DPC with BD over MgO-P obtained through thermal decomposition Mg(OH)2 at 400 ◦ C was performed under various reaction parameters. The ZrO2

200 180

CaO

effect of polymerization temperature was firstly examined in the range of 160–220 ◦ C. The result of Fig. 5(a) indicated that Mw values of PBC over ZrO2 and MgO-P sharply increased with raising the polymerization temperature from 160 to 210 ◦ C, which can be ascribed to the acceleration of the diffusion-limited polycondensation kinetics due to the decrease of polymer viscosity at a higher temperature [11,13]. As for CaO, the optimum temperature for the highest Mw value of 102 600 g/mol can be observed at 190 ◦ C, and then the Mw was gradually decrease as the continuously elevating the temperature. Clearly, the Mw values for CaO and MgO-P at a lower temperature are much higher than that of ZrO2 , which illustrates the presence of strong basic sites over CaO is also good for the improvement of polymerization rate at a low temperature. However, as depicted in Fig. 5(b) that the yield of PBC over CaO is much lower than those of other catalysts at the identical conditions, implying that strong basic sites on the surface of CaO are also more easily to expedite the decomposition of the obtained PBC polymer. As proven by Liu et al. [30], the stronger the base strength of the basic sites, the more difficult is the desorption of the ester products. That is to say, isolated O2− anions are prone to attack the carbonyl carbon atoms in APCs and impede the growing of polymer chain to lower Mw value together with corresponding yield. Therefore, 200 ◦ C can be selected as suitable polymerization temperature for MgO-P catalyst considering its Mw and yield comprehensively. Fig. 6 shows the dependence of Mw and yield of PBC versus reaction time over different catalysts. As shown in Fig. 6(a), one can see that the Mw of PBC over different solid base catalysts increased in the order of ZrO2 < CaO < MgO under a short reaction time of 0.5 or 1.0 h, which is good accordance with the change trend for the total

MgO

100

(a)

(b)

80

160

Yield of PBC

-3

Mw (×10 g/mol)

140 120 100 80 60

60 40 20

40 20

0

0 0.5

1.0

1.5

2.0

2.5

Reaction time (h)

3.0

3.5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Reaction time (h)

Fig. 6. Effect of reaction time on (a) Mw and (b) yield of PBC polymer obtained over different solid base catalysts. Reaction were carried out at 200 ◦ C and 200 Pa, the molar ratio of catalyst to DPC was set at 0.1 mol%.

82

Z. Wang et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

Table 3 Melt transesterification of DPC with a verity of diols in the presence of MgO catalysta . Entry

diols

product

Mw (Kg/mol)

Yield (%)

PDI

1 2 3 4 5c 6d 7e 8f 9g

BD HD BPA ISB ISB BD BD BD BD

PBC PHC PC PSC PSC PBC PBC PBC PBC

182.6 159.2 8.2 12.8b 32.5b 165.7 208.6 171.8 158.4

83.7 92.3 95.4 95.8 92.2 78.5 66.8 85.2 82.8

1.80 1.79 2.12 1.64 1.78 1.78 1.82 1.80 1.78

a Reaction conditions: reaction temperature 200 ◦ C, reaction time 2.5 h, the molar ratio of MgO to DPC 0.1 mol%. b Determined by GPC in chloroform(1.0 mL/min) at 30 ◦ C. c The reaction was carried out at 240 ◦ C for 1.0 h. d The result for the large scale experiment in a 5.0 L stainless steel reactor with 2000 g DPC and 843 g BD as feedstocks using 0.36 g MgO as catalyst. The polymerization reaction was carried out at 200 ◦ C and 80 Pa for 2.0 h. e The DPC and BD were used in the same mass as entry 6, and polymerization reaction was carried out at 215 ◦ C and 80 Pa for 1.5 h. f The feedstocks used in this experiment including DPC (technical grade), BD (≥99% purity) and MgO (thermal decomposition of commercial Mg(OH)2 at 400 ◦ C for 4.0) were all industrial products without any purification and treatment. g The feedstocks used in this experiment is same as entry 8 with MgO-C as catalyst.

amount of medium and strong basic sites. This also supports our conclusion that both medium and strong basic sites can promote this transesterification process. One also can find that the Mw of PBC over MgO could rapidly increase up to its maximum value of 182 200 g/mol at 2.5 h. However, when the time was beyond 2.5 h, the value of Mw showed no significant improvement. In terms of CaO, the Mw value also increased as the reaction time prolonging, and Mw reached the maximum of 76 500 g/mol at 1.0 h, and then the Mw of PBC would rapidity decline to 12 800 g/mol with further increase of time to 3.0 h. This phenomenon can be well explained by the fact that this polymerization process is very easily proceed at the beginning, but the viscosity of reaction system would become higher and higher with the increase of molecular weight, which will cause a negative influence to the further polymerization. Therefore, excessive reaction time could enhance its reverse reaction, depolymerization process, leading to the decrease of Mw . Meanwhile, Fig. 6(b) illustrated that the yield of PBC over CaO also sharply decreased with the reaction proceed. As depicted in Fig. S6, PDI seems to only depend on the Mw of PBC, and the higher the Mw value, the wider the PDI. Considering the above results, a temperature of 200 ◦ C and 2.5 h were selected as the optimum reaction condition for realizing the highest Mw with satisfactory yield for MgO-P. 3.5. Catalytic activity towards other diols To evaluate the potential and general application range of MgO catalyst, the catalytic transesterification of DPC with a verity of diols, including 1,6-hexanediol (HD), bisphenol A (BPA) and isosorbide (ISB) for the synthesis of corresponding poly(hexamethylene carbonate) (PHC), aromatic polycarbonate (PC) and poly(isosorbide carbonate) (PSC) were also investigated. As shown in Table 3, another liner alkane aliphatic diols, HD, can also easily transesterification with DPC to high-molecular-weight PHC with a high yield. However, the Mw values of PC and PSC are much lower than those of liner alkane aliphatic polycarbonates at the identical condition. Nevertheless, elevating reaction temperature can also facilitate the improvement of Mw for PSC (see entry 5, Table 3). As reported in previous paper [7,11,12], the reactivity of diols was thought to be related to the nucleophilicity of hydroxyl group, and strong nucleophilicity had a beneficial impact on this process. According to Fig. S7, the electron densities of hydroxyl oxygen in BPA and ISB were

obviously lower than that of liner alkane aliphatic diols reported in literature [11]. That is to say, the hydroxyl groups in liner alkane aliphatic diols are much easier to react with DPC in the presence of a catalyst. Therefore, the poor catalytic performances of BPA and ISB are mainly resulted from the low nucleophilicity of their hydroxyl group, which is in good agreement with the conclusion proposed by Eo et al. [12]. One also can see that the maximum Mw value of PBC over MgO is 182 200 g/mol, which is also the highest value among all the PBC samples obtained in various catalysis systems reported so far for one-pot transesterification of DPC with BD. Moreover, when compared with the conventional TSP-44, metal salts and inorganic/organic base catalysts in two-step condensation polycondensation of DMC with BD, our MgO catalyst also obviously exhibits comparable or even higher Mw for the obtained PBC [8–15]. Moreover, as discussed in literature [8,35,36], the main drawback for such strong homogeneous acid/base catalysis systems lies in that the residual of catalyst would seriously affect the polymer properties, especially for the thermal stability. As for organic base catalyst, high cost would preclude its application on a large scale. As far as Ti-based catalysts concerned, they suffer from complicated preparation procedure and undesirable colorization reaction. Additionally, a large scale condensation polymerization using 2000 g DPC and 842 g BD as feedstocks in the presence of MgO also successfully obtained 850 g white PBC polymer in a 5.0 L stainless steel reactor (see Fig. S8). As shown from entry 6 in Table 3, the Mw of this PBC could arrive at 165 700 g/mol, which is also sufficient for the Mw requirement of biodegradable plastics [4]. Of course, much higher Mw value can also be acquired by optimizing reaction conditions (see entry 7, Table 3). Additionally, high-molecular-weight PBC polymer also can be produced by using industrial grade DPC and BD as raw materials directly, illustrating that MgO also possesses excellent impurity poison tolerance (see entry 8 and entry 9, Table 3). In sharp contrast, the present MgO material is fabricated via coprecipitation method, which is facile, simple and cheap. More importantly, no further treatment is demanded because MgO is nontoxic and also excellent inorganic filler for polymer. As widely reported that the thermal stability, mechanical performance, and even flame-retardant property of many polymers could be greatly improved by addition of MgO or its mixed oxides [22,37]. Therefore, further work aiming at investigating the effect of MgO content on the application properties of APCs/MgO composites prepared by in-situ polymerization method is underway in our group. Obviously, the comparison and discussion above strongly suggest that the single MgO catalyst, featuring low-cost preparation, environment-friendly and excellent catalytic performance, can serve as a promising heterogeneous catalyst for the practical production of APCs with satisfactory yield. 3.6. Possible mechanism The mechanism on the transesterification of dialkyl carbonates with hydroxyl group has been widely discussed in previous works, in which the catalyst must adsorb the substrate to yield an unstable temporary intermediate and then induce the transesterification reaction [38–40]. As proven in these studies, dialkyl carbonates are ambient electrophile and hardly activated by solid bases. Using solid base for the transesterification of DMC with glycerol, Simanjuntak et al. have proposed that the basic site would activate the alcohol molecular by abstraction of the proton from hydroxyl group, producing a nucleophilic temporary intermediate, which subsequently attacked the carbonyl carbon atom of DMC to complete the transesterification process [38]. Recently, the similar mechanistic pathway has also been observed in many solid base or acid catalysis systems [30,39–42]. In this case, we also

Z. Wang et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

HO

CH2

O δ-

OH m

OPh

PhO

H O CH2 OH m O Mg δ- δ+

83

C O

PhO

Mg δ+

O C

OPh

H O CH2 OH m O Mg δ- δ+

PhOH O C O CH2 mO n *

HO CH2 m O HO CH2 mOH O C PhO OPh

O C

OPh

O Mg δ- δ+

PhOH

Scheme 1. A plausible reaction mechanism for MgO-catalyzed transesterification of DPC with diols.

inferred that the reaction of DPC with diols also followed the similar mechanism, which started with the activation of diols through abstracting its proton (H) with basic site. Therefore, on the basis of the above-mentioned discussion and literature, a plausible mechanistic pathway on the transesterification of DPC with BD over MgO is also proposed as shown in Scheme 1. MgO adsorbed diols molecules on the surface through the interaction with −CH2 OH group. Then the adsorbed diols molecules could be activated by basic sites. So BD would release its proton from the hydroxyl group to produce a HOCH2 -CH2 -CH2 -CH2 O− , which attacked the carbonyl carbon of DPC to give the product of HOCH2 -CH2 -CH2 CH2 OC(O)OPh, and then the released proton reacted with the PhO− to give phenol. Therefore, the molecular chain would become longer and longer with another catalytic cycle proceed. 4. Conclusions The synthesis of high-molecular-weight APCs through the transesterification of DPC with aliphatic diols can be carried out over solid base catalysts. MgO prepared via simple coprecipitation route showed the best performance among the solid bases and magnesium compounds investigated. The active sites of solid base for this reaction have been specified as the medium and strong basicity rather than acid, which increase along with the decrease of the calcination temperature. Simultaneously, the strong basic sites were also inclined to promote its reverse reaction, decomposition and depolymerization process, leading to the decrease of Mw and yield. In addition, MgO catalyst can also catalyze the transesterification of DPC with non-linear alkane diols with higher yield. Compared with the previously reported catalytic systems, the heterogeneous MgO catalyst has remarkable merits in catalytic activity, cost and polymer quality. Therefore, it can be inferred that MgO materials are of practical potential for the synthesis and modification of APCs. Acknowledgements This work was financially supported by the National Key Technology Pillar Program (2013BAC11B05), Key Research and Innovation Program of Jiangsu Province (Grant no. BE2015055) and the Science &Technology Pillar Program in Sichuan Province (2016GZ0228).

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.molcata.2016.08. 017.

References [1] J. Feng, R.X. Zhuo, X.Z. Zhang, Prog. Polym. Sci. 37 (2012) 211–236. [2] J. Wang, H.M. Zhang, Y.Y. Miao, L.J. Qiao, X.H. Wang, F.S. Wang, Green Chem. 18 (2016) 524–530. [3] Y. Liu, W.M. Ren, W.P. Zhang, R.R. Zhao, X.B. Lu, Nat. Commun. 6 (2015) 8594. [4] W.X. Zhu, X. Huang, C.C. Li, Y.N. Xiao, D. Zhang, G.H. Guan, Polym. Int. 60 (2011) 1060–1067. [5] L. Mespouille, O. Coulembier, M. Kawalec, A.P. Dove, P. Dubois, Prog. Polym. Sci. 39 (2014) 1144–1164. [6] Y.S. Qin, X.H. Wang, F.S. Wang, Prog. Chem. 23 (2011) 613–622. [7] W.X. Zhu, W. Zhou, C.C. Li, Y.N. Xiao, D. Zhang, G.H. Guan, D.J. Wang, J. MacromolSci, Pure. Appl. Chem. 48 (2011) 583–594. [8] J.H. Park, J.Y. Jeon, J.J. Lee, Y. Jang, J.K. Varghese, B.Y. Lee, Macromolecules 46 (2013) 3301–3308. [9] H. Mutlu, J. Ruiz, S.C. Solleder, M.A.R. Meier, Green Chem. 14 (2012) 1728–1735. [10] P.U. Naik, K. Refes, F. Sadaka, C.H. Brachais, G. Boni, J.P. Couvercelle, M. Picquet, L. Plasseraud, Polym. Chem. 3 (2012) 1475–1480. [11] Z.Q. Wang, X.G. Yang, S.Y. Liu, J. Hu, H. Zhang, G.Y. Wang, RSC Adv. 5 (2015) 87311–87319. [12] Y.S. Eo, H.W. Rhee, S. Shin, J. Ind. Eng. Chem 37 (2016) 42–46. [13] Z.Q. Wang, X.G. Yang, S.Y. Liu, H. Zhang, G.Y. Wang, Chem. Res. Chin. Univ. 32 (2016) 512–518. [14] J.J. Sun, D. Kuckling, Polym. Chem. 7 (2016) 1642–1649. [15] Z.Z. Jiang, C. Liu, R.A. Gross, 41 (2008) 4671-4680. [16] K.F. Peng, Y. Chen, C.T. Chen, Dalton. Trans. 44 (2015) 9610–9619. [17] F.S.H. Simanjuntak, S.R. Lim, B.S. Ahn, H.S. Kim, H. Lee, Appl. Catal. A-Gen. 484 (2014) 33–38. [18] A.P.C. Teixeira, E.M. Santos, A.F.P. Vieira, R.M. Lago, Chem. Eng. J. 232 (2013) 104–110. [19] M. Hajek, P. Kutalek, L. Smolakova, I. Troppova, L. Capek, D. Kubicka, J. Kocik, D.N. Thanh, Chem. Eng. J. 263 (2015) 160–167. [20] M. Di Serio, R. Tesser, A. Ferrara, E. Santacesaria, J. Mol. Catal. A-Chem. 212 (2004) 251–257. [21] G.P. Lin, L. Lin, X.L. Wang, L. Chen, Y.Z. Wang, Ind. Eng. Chem. Res. 54 (2015) 1282–1291. [22] Y. Zhang, R.Q. Peng, G.D. Zhou, Z.P. Fang, X.N. Li, Chin. J. Polym. Sci. 33 (2015) 1683–1690. [23] G.F. Li, W.H. Luo, M. Xiao, S.J. Wang, Y.Z. Meng, Chin. J. Polym. Sci. 34 (2016) 13–22. [24] D. Wu, X. Fu, F. Xiao, J. Li, J.N. Zhao, W. Wei, Y.H. Sun, Catal. Commun. 9 (2008) 680–684. [25] Q. Li, W.X. Zhu, C.C. Li, G.H. Guan, D. Zhang, Y.N. Xiao, L.C. Zheng, J. Polym. Sci. Part A: Polym. Chem. 51 (2013) 1387–1397.

84

Z. Wang et al. / Journal of Molecular Catalysis A: Chemical 424 (2016) 77–84

[26] A. Takagaki, K. Iwatani, S. Nishimura, K. Ebitani, Green Chem. 12 (2010) 578–581. [27] Y.X. Feng, N. Yin, Q.F. Li, J.W. Wang, M.Q. Kang, X.K. Wang, Catal. Lett. 121 (2008) 97–102. [28] P.X. Wang, S.M. Liu, X.Y. Ma, Y.D. He, A.S. Alshammari, Y.Q. Deng, RSC Adv. 5 (2015) 25849–25856. [29] J.I. Di Cosimo, V.K. Diez, M. Xu, E. Iglesia, C.R. Apesteguia, J. Catal. 178 (1998) 499–510. [30] Q.H. Liu, B.C. Wang, C.X. Wang, Z.J. Tian, W. Qu, H.J. Ma, R.S. Xu, Green Chem. 16 (2014) 2604–2613. [31] G.M. Zhao, J.H. Shi, G. Liu, Y. Liu, Z.L. Wang, W.X. Zhang, M.J. Jia, J. Mol. Catal. A-Chem. 327 (2010) 32–37. [32] P.X. Wang, S.M. Liu, F. Zhou, B.Q. Yang, A.S. Alshammari, Y.Q. Deng, RSC Adv. 5 (2015) 84621–84626. [33] W.Y. Zhang, H. Wang, W. Wei, Y.H. Sun, J. Mol. Catal. A-Chem. 231 (2005) 83–88. [34] J.M. Montero, P. Gai, K. Wilson, A.F. Lee, Green Chem. 11 (2009) 265–268.

[35] J.J. Lee, J.Y. Jeon, J.H. Park, Y. Jang, E.Y. Hwang, B.Y. Lee, RSC Adv. 3 (2013) 25823–25829. [36] V.N. Ignatov, V. Tartari, C. Carraro, R. Pippa, G. Nadali, C. Berti, M. Fiorini, Macromol. Chem. Phys. 202 (2001) 1946–1949. [37] Y.L. Li, Y.Y. Rao, W. Wang, F.T. Tan, J.G. Chen, X.L. Qiao, Acta Polym. Sin. 7 (2014) 970–975. [38] F.S.H. Simanjuntak, T.K. Kim, S.D. Lee, B.S. Ahn, H.S. Kim, H. Lee, Appl. Catal. A-Gen. 401 (2011) 220–225. [39] X.G. Zhang, X.B. Ke, Z.F. Zheng, H.W. Liu, H.Y. Zhu, Appl. Catal. B-Environ. 150–151 (2014) 330–337. [40] Z.M. Liu, J.W. Wang, M.Q. Kang, N. Yin, X.K. Wang, Y.S. Tan, Y.L. Zhu, J. Ind. Eng. Chem. 21 (2015) 394–399. [41] S.T. Gadge, A. Mishra, A.L. Gajengi, N.V. Shahi, B.M. Bhanage, RSC Adv. 4 (2014) 50271–50276. [42] D. Cornu, H. Guesmi, G. Laugel, J.M. Krafft, H. Lauron-Pernot, Phys. Chem. Chem. Phys. 17 (2015) 14168–14176.