Degradable polymers from ring-opening polymerization of α-angelica lactone, a five-membered unsaturated lactone

Abstract

A degradable polymer was prepared from α-angelica lactone, a five-membered unsaturated lactone, by ring-opening polymerization (ROP). The polymerizability of α-angelica lactone was explained by a DFT calculation. The degradability of the resultant polymer and the reaction kinetics of α-angelica lactone ROP were also considered. Owing to the presence of a C=C bond in α-angelica lactone, the ROP of five-membered cyclic lactone becomes feasible under moderate conditions and the resultant polyester exhibits good degradability under light or acidic/basic circumstances. Since α-angelica lactone can be easily obtained from the commercially available green bio-platform chemical levulinic acid, its ROP may provide a potential route to produce functionalized aliphatic polyesters from renewable resources.

---

1. Introduction

There has been great interest in the medical application of aliphatic polyesters in recent decades owing to their biodegradability and biocompatibility.^1,2Ring-opening polymerization (ROP) of cyclic lactones or functionally related compounds is a well known approach for obtaining such degradable polymers.^3,4 Among the usual lactones (Fig. 1), four-, six-, and seven-membered ones can be easily polymerized via ring-opening and the ROP of β-butyrolactone^5,6 (1), δ-valerolactone^7,8 (5) and ε-caprolactone^9,10 (6) has been widely researched. However, it is generally considered that the five-membered cyclic lactones like γ-butyrolactone (2) and γ-valerolactone (3) are thermodynamically stable and are impracticable for ROP except under rather strict conditions.^11–13 The oligomer of 2 with a molecular weight (M_n) up to 3.5 × 10³ was prepared at high temperature and pressure (e.g. 160 °C and 2 × 10³ MPa).¹⁴

Fig. 1 Structures of lactones with 4–7 members.

In practice, however, it is generally accepted that the five-membered lactones cannot form high molecular weight homopolymersviaROP, because they have very small ring strain energy (SE) and the Gibbs free energy of polymerization (ΔG_p) are positive.^11–13,15 It was assumed that the ester in the five-membered lactone ring was not likely to break down and form a polymer chain under normal conditions; and therefore, little attention was paid to the ROP of five-membered lactones in the past. Obviously, this view neglected the feasibility of ROP of five-membered unsaturated lactones.

In this work, we describe an encouraging result for the ROP of α-angelica lactone (4), a bio-derived monomer, which can be easily obtained from the commercially available green bio-platform chemical levulinic acid^16,17 (7), as shown in Scheme 1. As a five-membered unsaturated lactone, 4 can undergo ROP to produce functionalized poly(α-angelica lactone) (8). The polymer8 has carbon–carbon double bonds (C=C) in its iterative units, which makes it degradable and/or possible to form numerous polymers by chemical modification. As a result, ROP of 4 will provide a potential route to produce functionalized aliphatic polyesters from renewable resources.

Scheme 1 Synthesis of α-angelica lactone (4) and its ring-opening polymerization to polymer8. Reaction conditions: (a) H₃PO₄ as catalyst, 140 °C, 9 kPa, 3 h. (b) Sn(OCt)₂ as initiator, toluene as solvent, 130 °C.

On account of this, the ROP of 4 in the presence of stannous octoateinitiator was investigated in this work. The polymerizability of 4 was explained by a DFT calculation. The degradability of the resultant polymer8 and the reaction kinetics of α-angelica lactone ROP were also considered.

2. Experimental section

General procedure for the synthesis of poly(α-angelica lactone)

5 ml of 4 (AL, 5.44 g, 0.056 mol) was first put in a flask under a flow of nitrogen. 5 ml of dry toluene and 0.186 mmol (0.075 g) of stannous octoate (Sn(OCt)₂, Sn) were added with stirring under a nitrogen atmosphere. The mixture was intensively stirred and rapidly heated to 130 °C, and then kept at this temperature for 30 h to perform the ring opening polymerization (ROP). The polymerization reaction was quenched by exposing the reaction mixture to air outside the flask. After that, the solvent was evaporated. The remaining polymer8 (poly(α-angelica lactone), 4.08 g) was washed with cold hexane and then vacuum dried at 50 °C.

Characterization of the polymer

The molecular weight (M_n) and polydispersity index (PDI) of the resultant polymer were determined by gel permeation chromatography (GPC) using polystyrene standards. The GPC measurements were made at 30 °C on Waters 150-C gel permeation chromatograph equipped with a differential refractometer. The chromatograph was fitted with a set of three columns. Tetrahydrofuran was used as the mobile phase with a flow rate of 1 ml/min. The FT-IR spectra of the polymers were collected in the range of 400–4000 cm⁻¹ on a Nicolet-380 FT-IR spectrometer by using KBr pellet technique. ¹H-NMR and ¹³C-NMR spectra of the polymers were acquired at 25 °C with a Bruker DRX 300 NMR spectrometer.

3. Results and discussion

ROP of α-angelica lactone

For a typical polymerization, α-angelica lactone4 was first dissolved in toluene (5.0 mol L⁻¹), Sn(OCt)₂ (Sn) was added as initiator ([4]₀/[Sn]₀ = 300), and then the reaction was conducted at 130 °C under nitrogen atmosphere. After stirring for 30 h, the mixture is exposed to air and nearly 79.8% of 4 is converted to the polymer8 with a molecular weight (M_n) of 29357 g mol⁻¹ and a polydispersity index (PDI) of 1.25 by gel permeation chromatography (GPC) using polystyrene standards (Table 1). After reaction for 50 h, the conversion of 4 reaches 85.6% and remains essentially unchanged with further prolonging the reaction time and the polymerization may achieve an equilibrium state, as observed for other lactone ROP. By varying the loading of the initiator, it is possible to control the M_n and PDI of the resultant polymer.¹⁸ As listed in Table 1, M_n and PDI values of 8 decrease with the increase of monomer/initiator ratio ([4]₀/[Sn]₀). The polymerization rate at 130 °C is far faster than that at 110 °C and does not take place below 80 °C.

Table 1 Polymerization of α-angelica lactone with Sn(OCt)₂ as initiator

T/°C	[4]0/[Sn]0	Reaction time (h)	Conversion (%)	Molecular weight (g mol^−1)	PDI
130	300	30	79.8	29357	1.25
130	500	30	67.3	24390	1.22
130	1000	30	56.6	17175	1.15
130	300	50	85.6	21076	1.57
120	300	30	51.4	15903	1.16
110	300	30	18.2	8869	1.12
80	300	70	0	—	—

---

The molecular weights and PDI values of resultant polymer8 are dependent on the reaction time, as shown in Fig. 2. The molecular weight increases at first with the reaction time, reaches a maximum after reaction for 30 h and then decreases with further prolonging the polymerization; meanwhile, the PDI value increases consistently from 1.09 to 1.57 with the reaction time from 10 h to 50 h. The molecular weight reduction and PDI broadening with the extension of reaction time can be ascribed to the chain transfer in the polymer by transesterification;^9,10 the intramolecular chain transfer results in cyclic oligomers while the intermolecular transfer results in scrambling of chain segments.

Fig. 2 Dependence of molecular weight (M_n) and polydispersity index (PDI) of poly(α-angelica lactone) on reaction time. Polymerization conditions: 4 in toluene (5.0 mol L⁻¹), 130 °C, [4]₀/[Sn]₀ = 300.

Characterization of poly(α-angelica lactone)

As shown in Fig. 3, ¹H NMR spectrum of 8 is consistent with the iterative units resulting from the expected ROP of 4. The signal at 12.3 ppm can be assigned to the terminal carboxyl and hydroxylgroups. The peak at 7.2 ppm indicates that the carbon–carbon double bonds (C=C) survive the polymerization reaction and are present in the polymer8, which are prone to generating HO˙ and CH=CH˙ free radicals under light or heat and to trigger a series of reactions. The iterative units in the polymer8 from ROP of 4 are further confirmed by the ¹³C-NMR spectrum, as shown in Fig. 4.

Fig. 3 ¹H NMR spectrum (DMSO) of poly(α-angelica lactone) (29357 g mol⁻¹, PDI = 1.25).

Fig. 4 ¹³C NMR spectrum (DMSO) of poly(α-angelica lactone) (29357 g mol⁻¹, PDI = 1.25).

The structure of the polymer8 was further characterized by the FT-IR spectrum, as shown in Fig. 5. The functional groups –OH, CH₃–, –CH₂–, C=O, C=C, and C–O are identified in the FT-IR spectrum. The results further indicate that the iterative units resulting from the expected ROP of 4 and the carbon–carbon double bonds (C=C) are present in polymer8.

Fig. 5 FT-IR spectra of poly(α-angelica lactone) (29357 g mol⁻¹, PDI = 1.25).

As shown in Scheme 1, as a five-membered unsaturated lactone, 4 can undergo ROP to produce the functionalized polymer8. Similar to other lactones, the ROP of 4 may also follow a coordination-insertion mechanism with acyl-oxygen bond cleavage in which the chain grows via attaching the metal with an alkoxide bond.^19,20 The polymer8 has carbon–carbon double bonds (C=C) in its iterative units, which makes it degradable and/or possible to form numerous polymers by chemical modification. As a result, ROP of 4 will provide a potential route to produce functionalized aliphatic polyesters from renewable resources.

Degradability of poly(α-angelica lactone)

The degradability of the resultant polymer8 was then examined. Because the carbon–carbon double bonds (C=C) present in the polymer8 are prone to generate HO˙ and CH=CH˙ free radicals under light or heat and to trigger a series of reactions, the polymer8 becomes degradable. As shown in Fig. 6, the polymer8 is stable under dark circumstances; however, the weight loss reaches 34.8% when it is exposed to daylight for 49 days. Moreover, the polymer is easily degraded under acidic or basic circumstances. As shown in Fig. 7, when it is placed in solutions with the pH values of 2.5 and 10.0, the weight losses in 49 days reach 63.3% and 52.0%, respectively, much higher than that of 13.5% in a neutral solution.

Fig. 6 Comparison of the weight losses of poly(α-angelica lactone) by degradation under daylight illumination and dark conditions.

Fig. 7 Weight losses of poly(α-angelica lactone) by degradation in solution of different pH values.

Kinetics for ROP of α-angelica lactone

The reaction kinetics for the polymerization of 4 catalyzed by Sn(OCt)₂ in toluene was further determined. As shown in Fig. 8, the logarithm of the [AL]₀/[AL]_t ratio is linearly correlated with the reaction time, suggesting that the polymerization follows first order kinetics with respect to the concentration of monomer 4; the apparent rate constant (k) is 0.063 h⁻¹ with a halftime (t_1/2) of 11 h at 130 °C. Apparent activation energy (E_a = 110.1 kJ mol⁻¹) was also determined through a correction of rate constant with temperature (110–130 °C) by the Arrhenius expression, as shown in Fig. 9.

Fig. 8 Semilogarithmic plot of the relative α-angelica lactone concentration against the reaction time under different temperatures (Reaction conditions: [AL]₀ = 5.0 mol L⁻¹ in toluene, [AL]₀/[Sn]₀ = 500): k = 3.07 × 10⁻⁶ s⁻¹ (R = 0.9988) at 110 °C; k = 8.78 × 10⁻⁶ s⁻¹ (R = 0.9975) at 120 °C; k = 1.70 × 10⁻⁵ s⁻¹ (R = 0.9964) at 130 °C.

Fig. 9 Plot of ln(k) versus 1/T. E_a = 110.1 kJ mol⁻¹ (linear fit, R = −0.99912). Reaction conditions: [AL]₀ = 5.0 mol L⁻¹ in toluene, [AL]₀/[Sn]₀ = 500. (△, ○ and ▽ represent 3 sets of independent experiments performed under the same conditions).

Molecular modeling of lactones

Why does α-angelica lactone4 polymerize under moderate conditions but γ-butyrolactone2 and γ-valerolactone3 do not? Thermodynamically, the polymerizability of cyclic esters is determined by the Gibbs' free energy change of polymerization (ΔG_p).^{11–13,21,22} Because the entropy of polymerization (ΔS_p) is almost independent of the ring size, ΔG_p at given temperature depends mainly on the enthalpy of polymerization (ΔH_p) which is related to the ring strain.^{11–13,15,21–23} The release of ring strain through the polymerization process provides the principal driving force for the polymerization of cyclic monomers.^24,25 Molecular modeling with the program suite of Gaussian 03 was used to evaluate the ring strain and the enthalpy change for lactonepolymerization. Geometry optimizations and frequency calculations of lactones were performed by using the B3LYP density functional theory (DFT) in conjunction with the 6-31G(d) basis set.^26,27 The thermodynamic parameters were evaluated by the CBS-Q method. Strain energy (SE, CBS-Q) for the carbonyl derivatives is based upon the insertion/extrusion of –CH₂–/–O– energy equivalents,²⁸ as shown in Fig. 10.

Fig. 10 Strain energies (CBS-Q, kJ mol⁻¹) calculated for different lactones.

Table 2 lists the calculated results of a series of cyclic esters.^15,29 Various lactones exhibit little difference in the bond lengths of O–[C=O] and O–C, whereas they are significantly different in the bond angles of ∠C–O–[C=O] and ∠O–[C=O]–C, which show the distortion around the breaking bond for ROP. This suggests that the major contribution to the ring strain is bond angle distortion. The bond angles for ∠C–O–[C=O] and ∠O–[C=O]–C in 4 are 105.2° and 106.9°, which are 4.3° and 3.1° deviation from the forcefield equilibrium values, respectively. In contrast, the corresponding bond angles in 3 are only 0.3° and 0.9° deviation from the forcefield equilibrium values, respectively.

Table 2 Calculated structure and thermodynamic parameters of several lactones

Lactone	Bond angle (°)		Bond length (Å)		SE (kJ mol^−1)	ΔHp ^a (kJ mol^−1)	ΔHp ^b (kJ mol^−1)
	C–O–(C=O)	O–(C=O)–C	O–C	O–(C=O)
a From reference of 29. b This work. c Forcefield equilibrium value.^15
2	109.1	110.9	1.442	1.397	32.6	−15.4	−13.2
3	109.2	110.9	1.442	1.367	30.8	−7.1	−7.4
4	105.2	106.9	1.440	1.392	39.5		−26.3
5	121.5	108.7	1.444	1.360	41.4	−26.8	−27.6
6	122.5	117.8	1.439	1.361	48.8	−35.9	−35.9
7	109.5	110.0	1.425	1.370

---

The minimized energy of both the lactone monomer (M) and the 3- and 2-mers of the polymer (P₃ and P₂) were obtained, as listed in Table 3. The internal energy of a single polymer unit was estimated as the difference in energy between the 3- and 2-mers, and so the change in internal energy upon polymerization could be calculated from: ΔH_p = (E(P₃) − E(P₂)) − E(M).

Table 3 Calculated energy (CBS-Q) of the monomer (M) and 2- (P₂) and 3-mers (P₃) of the polymer from several lactones and the polymerization enthalpy (ΔH_p)^a

Lactone entry	Energy (a.u.)			ΔHp (a.u.)
	M	P2	P3
a 1 a.u. = 2625.5 kJ mol^−1; ΔHp = (E(P3) − E(P2)) − E(M).
2	−306.47181	−612.94985	−919.42669	−0.00503
3	−305.85986	−611.72398	−917.586657	−0.00282
4	−344.07452	−688.16901	−1032.25354	−0.01001
5	−345.26643	−690.73132	−1036.00827	−0.01052
6	−366.85785	−733.79357	−1100.66510	−0.01368

---

The bond angles for ∠C–O–[C=O] and ∠O–[C=O]–C in 4 are far distorted due to the presence of C=C in the cyclic ring, suggesting a higher strain than that in 3. As shown in Table 2, the strain energy (SE) in 4 is comparable to that in a six-membered lactone (5), much higher than that in 3. Accordingly, the enthalpy change (–ΔH_p) for ROP of 4 is close to that of 5, much higher than that of 3; actually, 4 also exhibits similar ROP ability to 5. All these indicate that the presence of a C=C bond in the unsaturated lactone enhance the strain energy and –ΔH_p for ROP, which makes the ROP of five-membered α-angelica lactone feasible under moderated conditions.

4. Conclusions

A degradable polymer was prepared from α-angelica lactone, a five-membered unsaturated lactone, by ring-opening polymerization. Owing to the presence of C=C bond in α-angelica lactone that significantly enhances the strain energy of pentagon ring, the ROP of five-membered cyclic lactone becomes feasible under moderate conditions and the resultant polyester exhibits good degradability under light or acidic/basic circumstances.

α-Angelica lactone as a bio-derived monomer can be easily obtained from the commercially available green bio-platform chemical levulinic acid. The ring-opening polymerization of it produces functionalized poly(α-angelica lactone) with carbon-carbon double bonds (C=C) in its iterative units, which makes it degradable and/or possible to form numerous polymers by chemical modification. This may provide a potential route to produce functionalized aliphatic polyesters from renewable resources.

Acknowledgements

The authors are grateful for the financial support of the National Basic Research Program of China (2006CB202504) and the Knowledge Innovation Program of the Chinese Academy of Sciences (KJCX2.YW.H16, YZ200933).

References

1. A. C. Albertsson and I. K. Varma, Biomacromolecules, 2003, 4, 1466–1486.
2. L. S. Nair and C. T. Laurencin, Prog. Polym. Sci., 2007, 32, 762–798.
3. J. Wu, T.-L. Yu, C.-T. Chen and C.-C. Lin, Coord. Chem. Rev., 2006, 250, 602–626.
4. O. Dechy-Cabaret, B. Martin-Vaca and D. Bourissou, Chem. Rev., 2004, 104, 6147–6176.
5. C. Guillaume, J.-F. Carpentier and S. M. Guillaume, Polymer, 2009, 50, 5909–5917.
6. P. Kurcok, P. Dubois and R. Jérôme, Polym. Int., 1996, 41, 479–485.
7. X. Lou, C. Detrembleur and R. Jérôme, Macromolecules, 2002, 35, 1190–1195.
8. R. R. Gowda and D. Chakraborty, J. Mol. Catal. A: Chem., 2009, 301, 84–92.
9. J. H. Khan, F. Schue and G. A. George, Polym. Int., 2009, 58, 296–301.
10. M. Labet and W. Thielemans, Chem. Soc. Rev., 2009, 38, 3484–3504.
11. R. Seymour and C. Carraher, Structure Property Relationships in Polymers, Plenum Press, New York, 1984.
12. H. R. Allcock, F. W. Lampe and J. E. Mark, Contemporary Polymer Chemistry, Pearson Education, Upper Saddle River, New Jersey, 3rd. edn, 2003.
13. K. N. Houk, A. Jabbari, H. K. Hall Jr and C. Alemán, J. Org. Chem., 2008, 73, 2674–2678.
14. F. Korte and H. Glet, J. Polym. Sci., Part C: Polym. Lett., 1966, 4, 685–689.
15. W. Saiyasombat, R. Molloy, T. M. Nicholson, A. F. Johnson, I. M. Ward and S. Poshyachinda, Polymer, 1998, 39, 5581–5585.
16. R. H. Leonard and Pensacola, U.S. Pat., 2 809 203, 1957.
17. T. Chen, T. Deng and X. Hou, Fine Chemicals, 2009, 26, 885–888.
18. M. Tang, A. J. P. White, M. M. Stevens and C. K. Williams, Chem. Commun., 2009, 941–943.
19. Y. Nakayama, K. Sasaki, N. Watanabe, Z. Cai and T. Shiono, Polymer, 2009, 50, 4788–4793.
20. N. Ropson, Ph. Dubois, R. Jérôme and Ph. Teyssié, Macromolecules, 1994, 27, 5950–5956.
21. J. Casanovas and C. Alemán, J. Comput.-Aided Mol. Des., 1994, 8, 441–448.
22. C. Alemán, Biopolymers, 1994, 34, 841–847.
23. A. Duda, J. Libiszowski, J. Mosnacek and S. Penczek, Macromol. Symp., 2005, 226, 109–119.
24. H. R. Allcock, Polym. Rev., 1970, 4, 149–189.
25. H. Sawada, Thermodynamics of Polymerization. Marcel Dekker, New York, 1976.
26. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
27. P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213–222.
28. R. D. Bach and O. Dmitrenko, J. Am. Chem. Soc., 2006, 128, 4598–4611.
29. J. Brandrup, E. H. Immergut, E. A. Grulke, A. Abe and D. R. Bloch, Polymer Handbook, John Wiley & Sons, New York, 4th edn, 1999.

---