ISSN 15600904, Polymer Science, Ser. B, 2015, Vol. 57, No. 4, pp. 298–303. © Pleiades Publishing, Ltd., 2015. Original Russian Text © V.A. Kuznetsov, M.I. Kodess, A.V. Pestov, 2015, published in Russian in Vysokomolekulyarnye Soedineniya, Ser. B, 2015, Vol. 57, No. 4, pp. 249–254.

CATALYSIS

Polymerization of εCaprolactone in the Presence of Tin(II) Chloride Complexes V. A. Kuznetsov*, M. I. Kodess, and A. V. Pestov Institute of Organic Synthesis, Ural Branch, Russian Academy of Sciences, ul. Sof’i Kovalevskoi 22, Akademicheskaya ul. 20, Yekaterinburg, 620990 Russia *email: [email protected] Received December 10, 2014; Revised Manuscript Received March 17, 2015

Abstract—The bulk polymerization of εcaprolactone at 110 and 155°C initiated by the tin complexes SnCl2 · 2H2O, SnCl4 · 5H2O, and SnCl2 · C4H8O2 (1,4dioxane) has been compared to that initiated by tin(II) 2ethyl hexanoate (octanoate). With 1H NMR spectroscopy, the monomer conversion and number average degree of polymerization of poly(εcaprolactone) samples have been determined and the effective rate constants of reaction have been calculated. All the studied initiators provide a conversion of 98.0–99.5%. The polymers with the highest molecular masses were obtained in the presence of SnCl4 · 5H2O at 155°C and in the presence of the tin(II) chloride complex with dioxane at 110°C. DOI: 10.1134/S1560090415040065

Poly(εcaprolactone) is a biodegradable polyester with a low melting temperature (56–65°C [1]) and a duration of hydrolytic degradation that ranges from several months to several years (depending on the deg radation conditions and polymer characteristics). The combination of these properties makes it possible apply the polymer as a material for prototyping (including that in 3D printers), for the preparation of biodegradable packaging and resolving surgical suture materials [2], and for articles for osteosynthesis [3, 4]. Poly(εcaprolactone) can be obtained via polycon densation of εoxycaproic acid or polymerization of ε caprolactone prepared through oxidation of cyclohex anone via the Bayer–Villiger reaction [5] or with the use of microbiological synthesis [6]. The polymeriza tion is the most suitable synthetic method because it allows easier control over the reaction and makes it

O

O

M

O

R

R

M O

O

O

R

possible to obtain a polymer with a high molecular mass and a narrow molecularmass distribution. Polymerization of εcaprolactone can performed via the anionic [7–12], cationic [13, 14], or coordination [15–22] mechanism in bulk, in a solution, or in a dis persion. The bulk polymerization is more attractive for industrial applications. In anionic polymerization of ε caprolactone, organometallic alkali metal compounds [7–9] and alkoxides of aluminum [10, 11] and tita nium(IV) [12] are used as initiators, whereas protonic acids act as initiators of cationic polymerization [13, 14]. The most widespread are the initiating systems composed of a Lewis acid and a nucleophilic com pound as a coinitiator (water, alcohol, or amine). In this case, the reaction mechanism is the subject of discus sion, but most authors suggest that the polymerization occurs via the coordination mechanism with the forma tion of a metal–oxygen covalent bond [1, 15–18].

M O

O

R

O

O M

O

O 298

R

O

M O O

R

O

O M O

POLYMERIZATION OF εCAPROLACTONE IN THE PRESENCE

As initiators of εcaprolactone polymerization via the coordination mechanism, organometallic com pounds and alkoxides of aluminum(III), tin(II), tita nium(IV), iron(III), magnesium(II), calcium(II), scandium(III), strontium(II), lanthanum(III), yttrium(III), and samarium(III) are used [1, 19–22]. These derivatives are characterized by high activity, but their industrial use is limited because of low hydro lytic stability. Thus, in polymerization of εcaprolac tone and other cyclic esters, initiating systems based on salts of transition and pmetals in the presence of alcohol are employed. In this case, metal alkoxides formed in situ in the reaction medium are also acting agents [23–25]. The most widespread systems are based on tin(II) 2ethyl hexanoate or octanoate [1, 15, 23–28]. It is believed that the optimum composition of the initiating system corresponds to a 2 : 1 alcohol– tin(II) octanoate ratio, which conforms to the com plete transesterification of the tin salt. The increase in the alcohol content in the reaction system results in polymerchain termination through transesterifica tion [15]. Among the complex compounds used for the poly merization of εcaprolactone, there are zirco nium(IV) acetylacetonate [29]; yttrium(III) amide complexes [30]; sterically hindered zinc(II), alumi num(III), yttrium(III), and scandium(III) complexes [1, 31–33]; and aluminum complexes with Schiff bases [1, 34]. In spite of the wide variety of studied ini tiators of εcaprolactone polymerization, in the litera ture, little attention has been paid to complexes of the simplest metal halogenides. Among these, aquacom plexes of tin(II), tin(IV), samarium(III), and ytter bium(III) are used predominantly. Moreover, com plexes of samarium(III) and ytterbium(III) halo genides with THF are employed [1, 35]. Generally, the majority of the described initiators of εcaprolactone polymerization are alkoxides or metal complexes with complicated ligands that are poorly suitable for practical use owing to hydrolytic instability and/or high cost. The widely used tin(II) octanoate is an imported reagent and, moreover, the preparation of a highmolecularmass polymer requires its additional purification via vacuum distillation and fractionation with about an 80% yield [34]. As available domestic Lewis acids, SnCl2 · 2H2O and SnCl4 · 5H2O crystalline hydrates may be used. However, the application of anhydrous reagents is favorable because of the possible hydrolytic opening of the monomer cycle and polymer degradation. In this context, anhydrous complexes of the simplest metal halogenides, which are practically unknown in the literature, are of undoubted practical and theoretical interest. In our opinion, the disadvantage of many studies is that comparison of the efficiency of the studied poly merization initiators with the efficiency of compounds described by other authors is impossible because of POLYMER SCIENCE

Series B

Vol. 57

No. 4

2015

299

differences in the reaction conditions and, moreover, under the conditions of one study, there are no com parative data for different initiators. Thus, a new initiator of εcaprolactone polymer ization—a tin(II) chloride complex with 1,4diox ane—was proposed and compared with the common initiators based on tin(II) and (IV), including tin(II) 2ethyl hexanoate. EXPERIMENTAL In this study, εcaprolactone (SigmaAldrich, ≥99%), crystalline hydrates of tin (II) chloride (ana lytical grade) and tin(IV) chloride (reagent grade), and dodecanol1 (specialpurity grade) were used without further purification. 1,4Dioxane (analytical grade) and benzene and ethyl acetate (specialpurity grade) were dried over zeolites, and tin(II) octanoate (Sigma Aldrich, 95%) was distilled under vacuum with separa tion of a fraction with Tb = 176–183°C/100 Pa [34]. Anhydrous tin(II) chloride was prepared as described in [36]. C,H,N elemental analysis was performed on a PerkinElmer automatic analyzer. 1H NMR spectra were recorded on a Bruker DRX400 spectrometer (400 MHz, CDCl3). The conversion of εcaprolac tone and the degree of polymerization were calculated from the ratios of integral intensities of signals due to methylene groups of the monomer and the polymer. For εcaprolactone, 1H NMR ( CDCl3): δ 1.77 (m, 4H, β(–CH 2–) + γ(–CH2–)), 1.86 (m, 2H, δ(–CH2–)), 2.64 (m, 2CH2, α(–CH2–)), 4.23 (m, 2CH2, ε(–CH2–)) ppm. For poly(εcaprolactone), 1 H NMR (CDCl3): δ 1.38 (m, 2H, γ(–CH2–)), 1.65 (m, 4H, β(–CH2–) + δ(–CH 2–)), 2.31 (t, 2H, α(–CH2–)), 3.65 (t, 2H, εend(⎯CH2–)), 4.02 (t, 2H, ε(–CH2–)) ppm [10]. Synthesis of the SnCl2 · C4H8O2 (1,4Dioxane) Complex Anhydrous tin(II) chloride (2.0 g, 22 mmol) and dried 1,4dioxane (23 mL) were placed into a flask. The mixture was boiled with refluxing for 1 h followed by filtration of the hot solution through a Schott filter. After cooling to room temperature, the precipitated crystals were filtered and dried under vacuum up to a constant mass. Anal. calcd. (%): C, 17.30; H, 2.90. Found (%): C, 17.43; H, 2.87. The product yield was 1.96 g (67%). The complex structure was described previously [37]. Polymerization of εCaprolactone Bulk polymerization of εcaprolactone was per formed at 110 and 155°C. εCaprolactone (2 g) was placed into 10mL tubes, and a solution of the initiator and the coinitiator in ethyl acetate or in benzene (in the presence of tin(II) octanoate) was added in

300

KUZNETSOV et al.

Effective rate constants of εcaprolactone polymerization at 110 and 155°C for initiators based on tin(II) and (IV) Кeff , mol–1 s–1 Initiator at 110°C

at 155°C

I1

8.96 × 10–7

5.44 × 10–5

I2

1.55 × 10–6

4.44 × 10–5

I3

7.45 × 10–7

3.90 × 10–5

I4

9.60 × 10–6

1.41 × 10–4

RESULTS AND DISCUSSION

amounts of 0.1 mL. The composition of the initiating system was chosen so that the initiator and coinitiator concentrations in the reaction mixture were 0.23 and 0.46 mol %, respectively. In the presence of water containing initiators, the coinitiator content was reduced in an equimolar ratio with added water. Once all the reagents were added, the mixture was evacuated for 30–40 min at room temperature and a pressure no higher than 50 Pa; then, the tubes were filled with nitrogen and immersed into an oil bath heated to 110°C and the polymerization was conducted at 110°C or the temperature was increased to 155°C. The polymerization was stopped by immersion of the tubes into liquid nitrogen. The reaction was performed with some initiators based on tin(II) or (IV) in the presence of dodecanol1 as a coinitiator. The resulting samples were estimated from the data on monomer Conversion, % 100 4 2

1

60

3

20

0

2

4

6

conversion, the numberaverage degree of polymer ization, and the effective polymerization rate constant calculated from the initial portion of the conversion– time plot.

8 τ, h

Fig. 1. Conversion of εcaprolactone vs. polymerization time at 110°C in the presence of initiators (1) I1, (2) I2, (3) I3, and (4) I4 and dodecanol1 as a coinitiator. (Here and in Figs. 2–4, cin = 0.23 mol % and ccoin = 0.46 mol %.)

First, the initiators were compared during εcapro lactone polymerization at 110°C. Among these, the complex of tin(II) chloride with 1,4dioxane (I1), tin(II) 2ethyl hexanoate (I2), tin(II) chloride dihy drate (I3) and tin(IV) chloride pentahydrate (I4) were estimated. The maximum reaction rate for εcaprolactone (table) within a polymerization time of 8 h was observed for tin(IV) chloride pentahydrate (I4); the minimum rate was observed for tin(II) chloride dihy drate (I3) (Fig. 1). During the use of I4, a higher amount of coinitiator was introduced into the reaction medium than those of other initiating systems at the cost of crystal water, a circumstance that supposedly resulted in a high reaction rate. The time dependences of the average degree of polymerization are patterns with the maximum near the point of ultimate conversion (Figs. 1, 2). This out come suggests the presence of chaintransfer reactions resulting in broadening of the molecularmass distri bution and a decrease in the average degree of poly merization. The cause is that the initiator of polymer ization, a Lewis acid, catalyzes the transesterification of ester groups of poly(εcaprolactone) as well. More over, the catalyst activity of the initiator increases with its activity in the initiation of polymerization. At the maximum point on the time dependence of the degree of polymerization, the equilibrium between the com petitive reactions is attained [38]. The maximum equilibrium degrees of polymer ization of poly(εcaprolactone) are achieved in the presence of the complex of tin(II) chloride with 1,4dioxane (I1) and tin(II) octanoate (I2): 118 and 110, respectively. In this case, the polymerization rate in the presence of tin(II) octanoate is higher than that of the complex of tin(II) chloride with 1,4dioxane, in spite of the induction period of ~30 min at the initial stage of synthesis, which is due to the low solubility of tin(II) octanoate in ε caprolactone and, consequently, to the heteroge neous character of initiation at the initial stage of the reaction. All the studied initiators make it pos sible to obtain conversions of 98–99%. The comparison of the data on the conversion of ε caprolactone and the degree of polymerization shows that, under the above conditions, the complex of tin(II) chloride with 1,4dioxane (I1) is the most efficient ini tiator. The most widely used initiator of polymerization of lactones and lactides—tin(II) 2ethyl hexanoate (I3)—gives a polymer with a lower molecular mass at a lower equilibrium conversion (Figs. 1, 2). POLYMER SCIENCE

Series B

Vol. 57

No. 4

2015

POLYMERIZATION OF εCAPROLACTONE IN THE PRESENCE Pn 125

301

Conversion, % 100 4 1 2

1

3 75

2

60

3 4 25 20 0

2

4

6

8 τ, h

0

20

40

60 τ, min

Fig. 2. Time dependence of the degree of polymerization, Pn, of poly(εcaprolactone) at 110°C in the presence of initiators (1) I1, (2) I2, (3) I3, and (4) I4 and dodecanol1.

Fig. 3. Conversion of εcaprolactone vs. polymerization time at 155°C in the presence of initiators (1) I1, (2) I2, (3) I3, and (4) I4 and dodecanol1.

The change in the composition of the metal coor dination sphere in the series of initiators I1, I2, I3, and I4 shows that the equilibrium conversions of εcapro lactone are comparable, whereas the average degrees of polymerization substantially depend on the initiator composition. The above results suggest that the pres ence of ether molecules in the metal coordination sphere provides not only protection against water, which is an active coinitiator capable of chain termi nation and macromolecule degradation, but also improves the activity of the metal atom as the Lewis acid. The first conclusion is based on the maximum average degree of polymerization for anhydrous com plexes and tin(II) octanoate (Fig. 2); the second con clusion is related to a higher rate of reaction of anhy drous complexes relative to that of tin(II) chloride crystalline hydrate (Fig. 1).

mined by the spatial accessibility of the central metal atom. The different regularities of polymerization in the presence of tin(IV) chloride pentahydrate result from introduction of large amounts of water (an active coinitiator of polymerization) to the reaction medium.

The polymerization at 155°C provides a suffi ciently low viscosity of the reaction medium for rapid chain propagation, whereas a significant degradation of the forming polymer is not observed (Figs. 3, 4). The maximum degree of polymerization was obtained in the presence of tin(IV) chloride pentahydrate (Fig. 4). At the same time, the polymerization rates in the presence of tin(II) octanoate, tin(II) chloride dihydrate, and the complex with dioxane are closely related (table, Fig. 3). Thus, the temperature elevation levels the influence of the structure of the complex on the polymerization rate. This may be due to changes in the mobile coordination sphere of the metal center and to attainment of the same coordination environ ment formed by 6hydroxyhexanoic acid. As a result, the effect of complexes is averaged. The increase in the molecular mass of the polymer supposedly is deter POLYMER SCIENCE

Series B

Vol. 57

No. 4

2015

The varying characters of the time dependences of the degree of polymerization with increasing temper ature (Figs. 2, 4) are indicative of different activities of the metal atom in the studied initiators acting as the Lewis acids in polymerization and transesterification reactions at various temperatures. At 155°C, the polymer with the highest molecular mass was obtained in the presence of aquacomplexes of tin(II) and (IV) chlorides (I3, I4). In addition, molecular mass grows once the equilibrium conver sion of εcaprolactone is attained. This result indicates the significant contribution of the transesterification reaction to the polymerization process. In this study, the polymerization of εcaprolactone initiated by a tin(II) chloride solvate complex with 1,4dioxane was first studied and this initiator was compared with the common initiators of polymeriza tion. In the temperature range 110–155°C, the use of the tin(II) chloride complex with 1,4dioxane results in a highermolecularmass polymer than that formed in the presence of traditional tin(II) octanoate. A similar regularity was found in our pre vious study of the polymerization of D,L and Llac tides in the temperature range 155–200°C [39]. In the studied temperature range, unlike the polymer

302

KUZNETSOV et al.

Pn 240 4 3 180 1 120

2

60

0

20

40

60 τ, min

Fig. 4. Time dependence of the degree of polymerization, Pn, of poly(εcaprolactone) at 155°C in the presence of initiators (1) I1, (2) I2, (3) I3, and (4) I4 and dodecanol1.

ization of εcaprolactone, the application of aqua complexes of tin(II) and (IV) chlorides, as well as anhydrous tin(II) chloride, results in polylactides with lower molecular masses. Thus, the tin(II) chloride complex with 1,4diox ane may be used as an efficient, available, and univer sal initiator of polymerization instead of tin(II) octanoate. ACKNOWLEDGMENTS This work was supported by the Government of Sverdlovsk oblast and the Russian Foundation for Basic Research (project no. 130396085 r_ural_a) as well as by the Ural Branch of Russian Academy of Sci ences (project no. 143IP34). REFERENCES 1. M. Labet and W. Thielemans, Chem. Soc. Rev. 38 (12), 3484 (2009). 2. M. Mochizuki, K. Nakayama, R. Qian, B.Z. Jiang, M. Hirami, T. Hayashi, T. Masuda, A. Nakajima, Pure Appl. Chem. 69 (12), 2567 (1997). ˆ

3. B. Mavis, T. T. Demirtas, M. Gümü¸s derelioglu, G. Gündüz, U. Colak, Acta Biomater. 5 (8), 3098 (2009). 4. P. Wutticharoenmongkol, N. Sanchavanakit, P. Pavas ant, and P. Supaphol, Macromol. Biosci. 6 (1), 70 (2006). 5. V. A. Kuznetsov, M. G. Pervova, and Yu. G. Yatluk, Russ. J. Appl. Chem. 86 (2), 176 (2013).

6. G. Chávez, R. HattiKaula, R. A. Sheldonc, and G. Mamo, J. Mol. Catal. B: Enzym. 89, 67 (2013). 7. A.F. Mingotaud, F. Dargelas, and F. L. Cansel, Mac romol. Symp. 153 (1), 77 (2000). 8. A. BhawLuximon, D. Jhurry, S. MotalaTimol, and Y. Lochee, Macromol. Symp. 231 (1), 60 (2006). 9. F. Sanda, M. Hitomi, and T. Endo, Macromolecules 34 (16), 5364 (2001). 10. A. Kiersnowski, P. Dabrowski, H. Budde, J. Kressler, J. Piglowski, Eur. Polym. J. 40 (11), 2591 (2004). 11. M. Endo, T. Aida, and S. Inoue, Macromolecules 20 (12), 2982 (1987). 12. D. Takeuchi, T. Nakamura, and T. Aida, Macromole cules 33 (3), 725 (2000). 13. S. GazeauBureau, D. Delcroix, B. MartínVaca, D. Bourissou, Ch. Navarro, S. Magnet, Macromole cules 41 (11), 3782 (2008). 14. F. Sanda, H. Sanada, Y. Shibasaki, and T. Endo, Mac romolecules 35 (3), 680 (2002). 15. R. F. Storey and J. W. Sherman, Macromolecules 35 (5), 1504 (2002). 16. D. Garlotta, J. Polym. Environ. 9 (2), 63 (2001). 17. A. P. Gupta and V. Kumar, Eur. Polym. J. 43 (10), 4053 (2007). 18. M. Labet and W. Thielemans, Chem. Soc. Rev. 38 (12), 3484 (2009). 19. Ph. Dubois, Ph. Degée, R. Jérôme, and Ph. Teyssié, Macromolecules 26 (11), 2730 (1993). 20. L. Zhang, Y. Niu, Y. Wang, P. Wang, L. Shen, J. Mol. Catal.: Chem. A 287 (1), 1 (2008). 21. J. Ling, W. Zhu, and Z. Shen, Macromolecules 37 (3), 758 (2004). 22. J. Cayuela, V. BounorLegaré, Ph. Cassagnau, A. Michel A, Macromolecules 39 (4), 1338 (2006). 23. A. Kowalski, A. Duda, and S. Penczek, Macromol. Rapid Commun. 19 (11), 567 (1998). 24. A. Kowalski, A. Duda, and S. Penczek, Macromole cules 33 (3), 689 (2000). 25. A. Kowalski, A. Duda, S. Penczek, and J. Libiszowski, Macromol. Symp. 153 (1), 41 (2000). 26. A. Kowalski, J. Libiszowski, T. Biela, M. Cypryk, A. Duda, S. Penczek, Macromolecules 38 (20), 8170 (2005). 27. C.M. Dong, K.Y. Qiu, Z.W. Gu, and X.D. Feng, Macromolecules 34 (14), 4691 (2001). 28. P. Dobrzynski, S. Li, J. Kasperczyk, M. Bero, F. Gasc, M. Vert, Biomacromolecules 6 (1), 483 (2005). 29. P. Dobrzynski, Polymer 48 (8), 2263 (2007). 30. L. J. E. Stanlake, J. D. Beard, and L. L. Schafer, Inorg. Chem. 47 (18), 8062 (2008). 31. A. Otero, J. FernándezBaeza, A. Antiñolo, A. Lara Sánchez, E. MartínezCaballero, J. Tejeda, L. F. SáanchezBarba, C. AlonsoMoreno, I. López Solera, Organometallics 27 (5), 976 (2008). POLYMER SCIENCE

Series B

Vol. 57

No. 4

2015

POLYMERIZATION OF εCAPROLACTONE IN THE PRESENCE 32. Y. Matsuo, K. Mashima, and K. Tani, Organometallics 20 (16), 3510 (2001). 33. D. J. Darensbourg and O. Karroonnirun, Macromole cules 43 (21), 8880 (2010). 34. US Patent No. 6448367 (1999). 35. D. BarbierBaudry, L. Brachais, A. Cretu, R. Gattin, A. Loupy, D. Stuerga, Environ. Chem. Lett. 1 (1), 19 (2003). 36. Yu. V. Karyakin and I. I. Angelov, Pure Chemical Com pounds (Khimiya, Moscow, 1974) [in Russian].

POLYMER SCIENCE

Series B

Vol. 57

No. 4

2015

303

37. S. N. Nitin, S. P. Dinkar, and M. B. Bhalchandra, Inorg. Chem. Commun. 11 (7), 733 (2008). 38. D. K. Gilding and A. M. Reed, Polymer 20 (12), 1459 (1979). 39. V. A. Kuznetsov, M. I. Kodess, and A. V. Pestov, Izv. Vyssh. Uchebn. Zaved., Khim. Khim. Tekhnol. 57 (2), 44 (2014).

Translated by L. Tkachenko