Synthesis of Ultrahigh Molecular Weight PLAs Using a Phenoxy-Imine Al(III) Complex

Abstract

l- and d-lactide polymerization kinetics using phenoxy-imine ligands of the type Me₂Al[O-2-tert-Bu-6-(C₆F₅N=CH)C₆H₃] in the presence of n-butanol and benzyl alcohol by ring-opening polymerization into polylactide are investigated. Effects of initiator concentration, catalyst concentration, polymerization temperature, and time on the molecular weight of poly-l-lactide are also investigated. Purification and drying of l-lactide are found to significantly influence the polymerization kinetics and the final molecular weight achieved. Ultrahigh molecular weight poly(l-lactic acid) PLLA (M_w = 1.4 × 10⁶ g/mol with Đ = 1.8) and ultrahigh molecular weight poly(d-lactic acid) PDLA (M_w = 1.3 × 10⁶ g/mol with Đ = 2.0) are obtained when polymerization is performed with a molar ratio of monomer to catalyst (LA/Al) of 8000 for 72 h at 120 °C in the presence of benzyl alcohol with conversions of 96 and 91%, respectively. We report for the first time the synthesis of ultrahigh molecular weight poly-l- and d-lactide using the Me₂Al[O-2-tert-Bu-6-(C₆F₅N=CH)C₆H₃] catalyst. The identified catalyst is found to be suitable for the synthesis of a broad range of molecular weights.

1. Introduction

Polylactide (PLA) is known to be biodegradable (in a reduced time span), and compostable, making this material suitable for commodity applications to replace fossil-based polymers. In addition, the nontoxic and harmless effects on the human body and its degradation products like H₂O and CO₂ make this material also suitable for medical and pharmaceutical applications.¹⁻³ However, one of the main drawbacks of PLA is the brittle nature and relatively low mechanical properties for medical applications, often limiting its use.^1,4,5 It is generally accepted for polymeric materials that some mechanical properties of oriented structures can be improved by increasing the molecular weight of the chains produced. PLA can be produced by two synthetic routes. One being polycondensation⁶⁻⁸ of lactic acid and the other being ring-opening polymerization^9,10 (ROP) of lactide (LA). Due to the presence of a condensate, which takes part to trans-esterification side reactions for the former, the latter mechanism is often employed to achieve higher molecular weights.¹¹⁻¹⁴ Many organic compounds¹⁵⁻¹⁷ or inorganic complexes^18,19 can be used to achieve the ROP of lactides. However, purity of the monomer, catalyst concentration, polymerization temperature, and reaction time are important influencing factors on the molecular weight of PLA and its distribution. The key step in the preparation of high molecular weight poly(l-lactic acid) PLLA is the use of high purity lactide to prevent chain transfer and/or trans-esterification.²⁰⁻²² Therefore, it is of imperative importance to study the purification of the monomer and conditions leading to high/ultrahigh molecular weight PLAs. Although many papers on the synthesis of PLLA have been published, there is a serious lack of findings focusing on the synthesis of UHMWPLLA.^13,21,23,24 The present literature mainly focuses on the synthesis of relatively low molecular weights (in the order of 10³–10⁵ g/mol); while only limited works tackle the synthesis of ultrahigh molecular weight (UHMW) PLA. The common synthetic strategy to achieve high/ultrahigh molecular weights is by solid-state polymerization or by the use of multiols as initiators for the ROP catalysts resulting in highly branched PLAs. Indubitably, tin octanoate^25,26 (or tin (II) 2-ethylhexanoate) is the most commonly employed catalyst for the large scale production of PLA; however, several other catalysts can be industrially used to produce PLA.^15,19,27 Trans-esterification tendency of the tin octanoate and its severe cytotoxic to almost all kinds of living cells, question the use of tin for the synthesis of PLA.^5,11 Therefore, efficient, less toxic, and environmentally friendly alternatives to tin catalysts are sought. Various species, including alkoxy and alkyl complexes of aluminum, lanthanides, and transition metals used as initiators to promote the synthesis of PLA, have been reported in the literature.²⁸ In particular, aluminum-based complexes have been extensively studied and proved to be suitable for the synthesis of PLA with relatively narrow polydispersity, high conversion, and also high molecular weights (in the order of 10⁵ g/mol).^29,30 Nomura et al.³¹ reported a facile and efficient system of the ROP of ε-caprolactone at room temperature using the salicylaldiminato-Al complex. Pappalardo et al.³² found that salicylaldiminato-Al complexes are well performing and very versatile initiators to synthesize high molecular weight in the ROP of ε-CL and l-lactide. Normand et al.^33,34 revealed that dinuclear aluminum complexes are very prominent precursors for ROP of lactide. Fuoco et al.^35,36 reported that a series of aluminum-based organometallic compounds as initiators in ROP of cyclic esters shows the most convenient method for the preparation of these materials with controlled and predictable molecular weights, stereochemistry, and end groups. However, just a few reports concerned the synthesis of ultrahigh molecular weight with narrow polydispersity and high yield in the ROP of cyclic esters promoted by the related aluminum complex catalyst. Iwasa et al.³⁷⁻³⁹ investigated the ring-opening polymerization of various cyclic esters by Al complex catalysts containing a series of phenoxy-imine ligands. They found that the C₆F₅ analogue is the most effective in terms of catalytic activity in the ROP of cyclic esters and the fluorines into the aryl moiety drastically increases the stability of the catalytic active species. Weak H–F interactions were observed when C₆F₅ was used as the aryl substituent, which would form a favored geometry for subsequent cyclic esters coordination and insertion. Meanwhile, they also found various metal alkoxides supported by bulky ancillary ligands has the high Lewis acidity and low toxicity of the aluminum alkoxide-based systems bearing bulky phenolate ligands, it seem to be active and suitable in preparing well-defined polymers. Based on Iwasa’s^37,38 work on using a series of Al complexes bearing phenoxy-imine ligands in ROP, the Lewis acidity on the central metal center could be modulated by different initiators like 1-butanol (1B) or benzyl alcohol (BA).^40,41 In addition, the existence of initiators could play an important role in determining the constitution of the resulting complexes or affecting the reactivity of metal complexes in catalytic reactions.

In this paper, we report the ROP kinetics of l-lactide with the aluminum complex catalyst {Me₂Al[O-2-tert-Bu-6-(C₆F₅N = CH)C₆H₃]} initiated by 1B and BA to promote the synthesis of UHMWPLLA; additionally, we also investigated the polymerization of d-lactide using similar synthetic conditions. The relationship between different reaction conditions and the influences of the purity of l-lactide on the molecular weights of the polymer produced is also addressed. At the best of our knowledge, this report is the first example of the synthesis of UHMWPLLA and UHMWPDLA using an Al-based catalyst.

2. Experimental Section

2.1. Materials

All of the reactions and manipulations were conducted using Schlenk techniques under nitrogen/vacuum or in a glovebox (MBraun Unilab Plus). l-lactide and d-lactide ((S,S)-3,6-dimethyl-1,4-dioxane-2,5-dione and (R,R)-3,6-dimethyl-1,4-dioxane-2,5-dione, respectively) were kindly donated by Purasorb (Corbion Co., Ltd.). Toluene and tetrahydrofuran (THF) were purchased from LPS and purified by SPS Compact (Mbraun Solvent Purification System). n-Butanol and benzyl alcohol were purchased from Sigma-Aldrich and purified over molecular sieve under nitrogen. Chloroform (stabilized with amylene) used for polymer dissolution was purchased from Biosolve, while the chloroform used for chromatogram analyses (HPLC grade) was purchased from HiPerSolv and used as received. p-Toluenesulfonic acid, 2,3,4,5,6-pentafluoroaniline, and 3-tert-butylsalicylaldheyde were purchased from Acros Organics and used as received. CDCl₃ was purchased from Cambridge Isotope Laboratories and used as received. Aluminum trimethyl 2.0 M in toluene was purchased from Sigma-Aldrich and used as received.

2.2. Purification of l- and d-Lactides

Using a custom-made double Schlenk vial connected with a fritted disk, l-lactide and d-lactide (Corbion Co., Ltd.) were purified by the recrystallization method in toluene.^23,42 The as-received lactide was mixed at room temperature in a minimum amount of toluene and heated to 100 °C in a double Schlenk vial equipped with an in-built fritted glass filter. Further, the lactide is recrystallized at room temperature and filtered using the in-built fritted glass disk. The resulting purified lactide was dried in an oil bath at 40 °C for 6 h with a cold trap under vacuum, and then stored in a glovebox. These steps were repeated three times and six times, named “3x_LA” and “6x_LA”, respectively, for the l-lactide. The d-lactide used was recrystallized six times (Table 1).

Table 1. Chiral GC, and KFT Characterization of the l-Lactide and Toluene Together with the Yields of the Purifications.

samples	l-lactidea (%)	d-lactidea (%)	meso-lactide (%)	water contentb (ppm)	yield (%)
as received l-lactide	89.32	3.49	7.19	443
as received d-lactide	1.41	94.98	3.61	335
3x_l-lactide	93.47	2.12	4.41	228	84.7
6x_l-lactide	94.56	1.58	3.86	64	63.8
6x_d-lactide	1.17	95.65	3.18	184	80.0
toluene				7
1B				101
BA				17

^aChiral GC data in chloroform.

^bKFT data in THF.

2.3. Synthesis of {Me₂Al[O-2-tert-Bu-6-(C₆F₅N=CH)C₆H₃]}

The catalyst was synthesized partially following the procedure described in the literature.^37,39 To a stirred mixture of 3-tert-butyl-2-hydroxybenzaldehyde (3.90 mL, 21 mmol) and 2,3,4,5,6-pentafluoroaniline (25 mmol) in 100 mL of anhydrous toluene, 12 wt % solution of p-toluenesulfonic acid in acetic acid (62 μL, 47 μmol) was added at room temperature. The resultant yellow mixture was stirred at reflux temperature for 5 h in a Dean–Stark equipment, followed by stirring at room temperature for an additional 18 h. The solvent was removed under reduced pressure, giving a crude yellow oil, which was left at −20 °C for 20 h. Purification of the resulting ligand was performed by flash column chromatography using a Buchi Reverelis PREP purification system with an FP Ecoflex 80 g silica column. The eluent was a mix of n-hexane/ethyl acetate (100:1), giving 2-tert-Bu-6-(C₆F₅N=CH)C₆H₃OH, a yellow solid. Into a stirred solution containing AlMe₃ in toluene (0.825 mL of 2.0 M solution, 1 equiv) at −20 °C, 2-tert-Bu-6-(C₆F₅N=CH)C₆H₃OH (0.5668 g, 1 equiv) in toluene (1 mL) was added dropwise over 10 min period and was left to react for an additional 5 min at −20 °C using a cold bath made and controlled using a mixture of dry ice and acetone. The solution was allowed to warm to room temperature within 20–30 min and was stirred for 3 h. The mixture was then concentrated under vacuum, and the resultant yellow viscous oil was stored in a glovebox for months till completely crystallized into yellow crystals. ¹H NMR characterization of the resultant catalyst was provided in the Supporting Information.

2.4. Polymerization of l-Lactide

Typical polymerization procedures (Tables 2–5) were as follows. l-Lactide was polymerized as described in previous studies.^13,21 Briefly, PLLA was obtained by ring-opening polymerization of l-lactide ([LA]) using Me₂Al[O-2-tert-Bu-6-(C₆F₅N =CH)C₆H₃] (1) as catalyst, and 1B or BA as initiators. Schlenk vials for polymerization were dried in an oven at 130 °C overnight and quickly transferred in the glovebox. Then, the Schlenk vials were loaded with the desired amount of l-lactide. Separately, in a Schlenk vial, the Al complex was reacted with 1B or BA under nitrogen flow at room temperature. The solution was stirred for 10 min, followed by the addition of the desired amount of lactide in toluene. The reaction mixture was then placed into an oil bath preheated at the desired reaction temperature for the desired polymerization time. The resulting polymer was dissolved in chloroform and precipitated in cold methanol and dried until constant weight. The amount of CHCl₃ and CH₃OH used to dissolve and precipitate the PLAs are approximately 5 and 50 mL per gram of polymer obtained.

Table 2. Effect of Polymerization Time on the ROP of l-Lactide.

entry	time (h)	yieldc (%)	TONd	TOFe (h–1)	Mn (theor)f (kg/mol)	Mng (kg/mol)	Mwg (kg/mol)	Đg	Tm (°C)	χh (%)
1a	0.5	0
2a	24	4	80	3	12	58	66	1.1	174	57
3a	48	19	380	8	55	100	117	1.2	175	51
4a	72	69	1380	19	19	115	138	1.2	178	50
5a	96	22	440	5	63	96	110	1.1	175	53
6a	120	30	600	5	86	117	137	1.2	176	49
7b	0.5	0
8b	24	39	390	16	56	33	36	1.1	172	73
9b	48	69	690	14	99	47	55	1.2	172	68
10b	72	78	780	11	112	129	174	1.3	177	44
11b	96	87	870	9	125	68	103	1.5	174	57
12b	120	67	670	6	97	126	155	1.2	177	53

^aConditions: Al (1.73 μmol), 1B (1.73 μmol), LA (3.47 mmol), toluene, [LA]₀ =0.69 mmol/mL, LA/Al = 2000, 80 °C, 3x_L-LA.

^bConditions: Al (3.47 μmol), BA (3.47 μmol), LA (3.47 mmol), toluene, [LA]₀ = 0.69 mmol/mL, LA/Al = 1000, 80 °C, 3x_L-LA.

^cCalculated as follows: ((isolated weight of PLA)/(starting monomer amount)) × 100%.

^dTON = (molar amount of LA reacted)/(molar amount of Al).

^eTOF = TON/reaction time.

^fTheoretical molecular weight calculated using M_n (theor) = yield × [LA]₀/[Al] × M_LA.

^gGPC data in CHCl₃ vs polystryrene standards.

^hMelting enthalpy of 100% crystalline poly-l-lactide is 93.6 J/mol.⁴⁴

Table 5. Effect of Monomers in the Ring-Opening Polymerization of Lactides.

entry	LA/Al	monomer	conv.a (%)	yieldb (%)	TONc	TOFd (h–1)	Mn (theor)e (kg/mol)	Mnf (kg/mol)	Mwf (kg/mol)	Đf	Tm (°C)	χg (%)
1	8000	l-lactide	96	95	7600	106	1107	759	1397	1.8	174	52
2	8000	d-lactide	88	91	7280	101	1015	634	1261	2.0	162	33

^aConditions: Al (3.47 μmol), BA (3.47 μmol), LA (27.76 mmol) 6x_LA, in dry toluene, [LA]₀ = 5.55 mmol/mL, 72 h, 120 °C Monomer conversion determined by ¹H NMR spectroscopy (CDCl₃, 298 K).

^bCalculated as follows: ((isolated weight of PLA)/(starting monomer amount)) × 100%.

^cTON = (Molar amount of LA reacted)/(molar amount of Al).

^dTOF = TON/reaction time.

^eTheoretical molecular weight calculated using M_n (theor) = conversion × [LA]₀/[Al] × M_LA.

^fGPC data in CHCl₃vs polystryrene standards.

^gMelting enthalpy of 100% crystalline polylactide is 93.6 J/mol.⁴⁴

2.5. Chiral Gas Chromatography

Chiral gas chromatography (GC) was performed using a Hewlett-Packard 5890 Series II GC (Palo Alto, CA), equipped with a flame-ionization detector, having a CP-Chirasil-Dex CB capillary column (length = 25 m; internal diameter = 0.25 mm; film thickness = 0.25 mm). The heating program was 5 min at 40 °C; a ramp of 5 °C/min up to 200 °C; and a final 23 min at 200 °C. The operating conditions were as follows: detector, FID, at 200 °C; carrier gas, N₂, 2.0 mL/min flow rate; split ratio, 20:1; lactide sample concentration, 1 mg/mL in chloroform; injection volume, 0.2 μL.⁴³ Chromatograms are provided in the Supporting Information. The quantification was based on the area integration assuming the same response factor for l-, d-, and meso-lactides.

2.6. Gas Chromatography-Mass Spectrometry (GC-MS)

GC-MS was carried out on a GCMS-QP2010 Ultra from Shimadzu equipped with a DB-5 coated fused silica capillary column (30 m × 0.32 mm, 0.25 μm film thickness). One microliter of the derivative sample was injected into GC-MS using split mode (10:1). Ultrapure helium (constant flow, 1.5 mL/min) served as carrier gas with the purge flow of 1 mL/min. The injector temperature was 300 °C. Chromatograms and mass spectra are provided in the Supporting Information.

2.7. Gel Permeation Chromatography

The number-average molecular weight (M_n), weight-average molecular weight (M_w), and polydispersity index (Đ) of the polylactides were determined by gel permeation chromatography (GPC) on the Shimadzu LC-2030 GPC system. Chloroform was used as eluent at a flow rate of 1.0 mL/min. The temperature of the columns and detector was maintained at 40 °C. Molar masses were calculated against a reference curve performed using a calibration ranging from <10² to 2 × 10⁷ (M_w) using polystyrene standard samples. GPC traces are provided in the Supporting Information.

2.8. Karl Fischer Titration

The water content of monomers was determined by Karl Fischer Titration (KFT) C30S from Mettler Toledo Corporation. Samples, 50 mg/mL in tetrahydrofuran (THF); injection volume, 0.5 mL.

2.9. Differential Scanning Calorimetry

Differential scanning calorimetry (DSC) was performed on a TA Instruments DSC Q2000 to identify the glass-transition temperature, melting temperature/crystallization, and heat of melting/crystallization of the synthesized polymers. Heating and cooling runs were performed at rates of 10 °C/min. The melting temperature of the materials was determined from the second heating run. Crystallinity (χ) values were calculated according to the formula χ_c = (ΔH_m – ΔH_c/ΔH_m°) × 100%, where ΔH_m⁰ = 93.6 J/g represents the theoretical heat of melting of 100% crystalline PLA.^44,45

2.10. NMR

¹H NMR and ¹³C NMR spectra were recorded with a Bruker Ultrashield 300 spectrometer (300 MHz magnetic field). NMR samples were prepared by dissolving ca. 10 mg of sample in 0.5 mL of deuterated chloroform (CDCl₃). All spectra were referenced against tetramethylsilane or a residual solvent peak in the deuterated solvent.

3. Results and Discussion

3.1. Lactide Purification and Characterization

To study the effect of the monomer purity on the resulting polymer produced using the synthesized catalyst, a series of purification and analyses are performed. To estimate the chiral purity of the monomers from the chiral GC chromatogram, the area normalization method can be directly used to calculate the l-lactide and d-lactide contents, which are expressed as follows^13,43

where L% and D% are the l-and d-lactide contents; A is the peak area of the component; and impurity represents meso-lactide as the standard gas chromatography of the recrystallized lactides show only one peak. To recall, standard gas chromatography cannot discriminate between l-, d-, and meso-lactides.^13,21

A summary of the chiral GC and KFT characterization data obtained for the as-received l-lactide and the recrystallized l-lactides are summarized in Table 1. The GC chromatograms of the lactides are shown in Figures S7–S11. The retention time of the solvent used for the analyses is 3.87 min. The retention times of meso-lactide, l-lactide, and d-lactide are 29.39, 30.94, and 31.77 min, respectively. According to the results, as the recrystallization times of the as-received l-lactide increase, d-lactide and other components decrease, thus increasing the l-lactide concentration from 89.32 to 94.56%. The water content of the as-received l-lactide is 443 ppm and decreases to 64 ppm with the increasing number of recrystallizations. Thus, the recrystallization in toluene is an efficient method to reduce the water content and the impurities in the monomer. Chiral GC shows an increase in the chiral purities of the two lactides with the increasing number of recrystallization up to values of 94.56 and 95.65% for l- and d-lactides, respectively.

Standard gas chromatography on the as-received monomer shows the presence of an impurity (Figure S12); such impurities disappear from the chromatogram after three recrystallization in toluene (Figure S13).

3.2. Polymerization Kinetics

To understand the effect of the catalyst concentration on the molecular characteristics and the polymerization kinetic, a series of reactions varying the amount of catalyst used as a changing variable are performed. The results of the l-lactide polymerization, synthesized following ROP, when the catalyst is activated with 1B and BA at different catalyst concentrations are provided in the Supporting Information. As shown in Figure S26a,b, the highest molecular weight of PLLA as well as the highest polymer yield can be obtained for an LA/Al molar ratio of 2000 in the presence of 1B and at an LA/Al molar ratio of 1000 in the presence of BA. Therefore, the following kinetic study is performed using these molar ratios. Table 2 shows the polymerization conditions and molecular characteristics of the polymers synthesized by (1) in the presence of 1B and BA at different polymerization times and fixed LA/Al molar ratios.

Figure 1 shows the evolution of the molecular weight and PLLA yield as a function of reaction times in the presence of the two initiators used, 1B and BA, at LA/Al molar ratios of 2000 or 1000, respectively. For both initiators, the molecular weight increases with the polymerization time up to a value of 138 and 174 kg/mol when 1B and BA are used, respectively. For both initiators, the highest value of M_w, and polymer yield is achieved after 72 h of the polymerization time. Thus the chosen polymerization time, for both initiators, is 72 h in the following experiments. The polydispersity is found to be in a relatively narrow range (1.1 ≤ Đ ≤ 1.5) at all of the chosen polymerization times, reflecting the synthesis of unimodal PLA using a single-site catalyst.

Figure 1.

M_w (left axis) and yield (right axis) as a function of polymerization time using 1B (a) and BA (b).

Similar to the polymers of Table S1 in the Supporting Information, the polymers of Table 2 also show high melting temperatures (up to 178 °C) and crystallinities (up to 68%), which are indications of the high stereo-regularity of the PLLA synthesized. For most of the polymers tested, cold crystallization is observed between 100 and 120 °C and a classic glass-transition temperature around 53–60 °C (Figures S17 and S18).

3.3. Influence of Polymerization Temperature

Table 3 reports the polymerization conditions and molecular characteristics of PLLA synthesized using the two initiators, 1B and BA, at different polymerization temperatures. In this set of experiments, the LA/Al molar ratios and the polymerization time are fixed based on previous findings.

Table 3. Effect of Polymerization Temperature in the ROP of l-Lactide.

entry	T (°C)	yieldc (%)	TONd	TOFe (h–1)	Mn (theor)f (kg/mol)	Mng (kg/mol)	Mwg (kg/mol)	Đg	Tm (°C)	χh (%)
1a	60	0
2a	70	22	440	6	63	35	38	1.1	172	69
3a	80	69	1380	19	199	115	138	1.2	178	50
4a	90	77	1540	21	222	119	182	1.5	176	45
5a	100	77	1540	21	222	165	205	1.2	176	42
6a	120	74	1484	21	213	100	122	1.2	175	54
7b	60	16	160	2	23	33	36	1.1	170	64
8b	70	58	580	8	84	78	88	1.1	174	58
9b	80	78	780	11	112	128	175	1.4	177	44
10b	90	90	900	13	130	151	236	1.6	175	47
11b	100	86	860	12	124	109	177	1.6	174	46
12b	120	86	860	12	124	151	256	1.7	166 173	33

^aConditions: Al (1.73 μmol), 1B (1.73 μmol), LA (3.47 mmol), in dry toluene, [LA]₀ = 0.69 mmol/mL, LA/Al = 2000, 72 h, 3x_l-LA.

^bConditions: Al (3.47 μmol), BA (3.47 μmol), LA (3.47 mmol), in dry toluene, [LA]₀ = 0.69 mmol/mL, LA/Al = 1000, 72 h, 3x_l-LA.

^cCalculated as follows: ((isolated weight of PLA)/(starting monomer amount)) × 100%.

^dTON = (Molar amount of LA reacted)/(molar amount of Al).

^eTOF = TON/reaction time.

^fTheoretical molecular weight calculated using M_n (theor) = yield × [LA]₀/[Al] × M_LA.

^gGPC data in CHCl₃vs polystryrene standards.

^hMelting enthalpy of 100% crystalline poly-l-lactide is 93.6 J/mol.⁴⁴

Figure 2 graphically summarizes the polymerization data of Table 3. The yield increases with the polymerization temperature, up to approximately 90 °C, for both initiators. At the higher temperatures, at least between 90 and 120 °C, the yield seems to stay constant to values of approximately 77 and 87% for 1B and BA, respectively. The most reasonable explanation to the increase in the yield can be attributed to the increased polymerization rate achieved at higher temperatures. The molecular weight also increases as a function of polymerization temperature up to 100 °C for 1B and 120 °C for BA. By increasing the polymerization temperature, the molecular weight of the polymers synthesized increases (beyond the reported molecular weights in Table 2) up to 205 kg/mol using 1B and 256 kg/mol using BA. Polydispersity of the synthesized polymer also increases with the polymerization temperature up to a value of 1.5 for 1B and 1.7 for BA. An increase in the polydispersity, as a function of polymerization temperature, can be attributed to trans-esterification or hydrolysis that may result into different chains with different molecular weights. However, it cannot be excluded that depolymerization process might become relevant at higher polymerization temperatures. This temperature dependence is in good agreement with previous reported studies on ROP.^33,46

Figure 2.

M_w (left axis) and yield (right axis) as a function of polymerization temperature using 1B (a) and BA (b).

Similar to the polymers of Table 2, the polymers of Table 3 also show high melting temperatures (up to 178 °C) and crystallinities (up to 69%), which are indications of the high stereo-regularity of the PLLA synthesized. For most of the polymers tested, cold crystallization is observed between 90 and 110 °C and a classic glass-transition temperature around 53–60 °C (Figures S19 and S20).

To study the influence of water content on the kinetics of polymerization and maintain a relatively high catalytic productivity, further polymerizations are performed at the reaction temperature of 100 and 120 °C for 1B and BA, respectively.

3.4. Influence of l-Lactide Purification at Different LA/Al Molar Ratios

Table 4 summarizes the polymerization conditions and molecular characteristics of the polymers synthesized using (1) in combination with 1B and BA as initiators, at different LA/Al and monomer purities. The chosen polymerization conditions, such as catalyst concentration, polymerization time, and polymerization temperature, are the ones identified for securing the highest molecular weight and catalyst activity for the two initiators in the earlier sections.

Table 4. Influence of Monomer Purity at Different Molar Ratios of LA/Al in the ROP of l-Lactide on the Molecular Characteristics.

entry	LA/Al	initiator	monomer	conv.d (%)	yieldd (%)	TONe	TOFf (h–1)	Mn (theor)g (kg/mol)	Mnh (kg/mol)	Mwh (kg/mol)	Đh	Tm (°C)	χi (%)
1a	2000	1B	3x_LA	57	77	1540	21	164	165	205	1.2	176	42
2a	4000	1B	3x_LA	90	95	3800	53	519	204	357	1.7	176	37
3a	6000	1B	3x_LA	84	87	5220	73	726	235	346	1.5	176	39
4a	8000	1B	3x_LA	51	59	4720	66	588	243	302	1.2	177	39
5a	2000	1B	6x_LA	95	92	1840	26	274	189	351	1.9	174	41
6a	4000	1B	6x_LA	97	91	3640	51	559	327	672	2.1	173	38
7a	6000	1B	6x_LA	84	93	5580	78	726	465	924	2.0	176	39
8a	8000	1B	6x_LA	34	33	2640	37	392	153	194	1.3	175	53
9b	1000	BA	raw_LA	49	34	340	5	71	32	39	1.2	168	56
10b	1000	BA	3x_LA	93	86	860	12	134	151	256	1.7	166 173	33
11b	2000	BA	3x_LA	81	90	1800	25	233	233	416	1.8	175	40
12b	3000	BA	3x_LA	85	91	2730	38	368	344	576	1.7	176	37
13b	4000	BA	3x_LA	87	90	3600	50	502	369	694	1.9	175	38
14b	6000	BA	3x_LA	74	58	3480	48	640	292	317	1.1	173	46
15b	8000	BA	3x_LA	58	51	4080	57	669	226	290	1.1	173	42
16b	10 000	BA	3x_LA	25	25	2500	35	360	151	202	1.3	172	49
17b	1000	BA	6x_LA	90	85	850	12	130	167	291	1.7	163 169	37
18b	2000	BA	6x_LA	93	91	1820	25	268	323	621	1.9	169	35
19b	3000	BA	6x_LA	93	96	2880	40	402	423	783	1.9	170	34
20b	4000	BA	6x_LA	94	97	3880	54	542	451	902	2.0	173	37
21b	6000	BA	6x_LA	91	98	5880	82	787	608	1102	1.8	172	41
22b	8000	BA	6x_LA	96	95	7600	106	1107	759	1397	1.8	174	52
23b	10 000	BA	6x_LA	65	72	7200	100	937	372	547	1.5	172	46

^aConditions: 1B 1.0 equiv to Al (in dry toluene), Al (1.73 μmol), 1B (1.73 μmol), 72 h, 100 °C.

^bConditions: BA 1.0 equiv to Al (in dry toluene), Al (3.47 μmol), BA (3.47 μmol), 72 h, 120 °C.

^cMonomer conversion determined by ¹H NMR spectroscopy (CDCl₃, 298 K).

^dCalculated as follows: ((isolated weight of PLA)/(starting monomer amount)) × 100%.

^eTON = (molar amount of LA reacted)/(molar amount of Al).

^fTOF = TON/reaction time.

^gTheoretical molecular weight calculated using M_n (theor) = conversion × [LA]₀/[Al] × M_LA.

^hGPC data in CHCl₃vs polystryrene standards.

ⁱMelting enthalpy of 100% crystalline poly-l-lactide is 93.6 J/mol.⁴⁴

Figure 3 graphically summarizes the evolutions of the molecular weights and yields of the PLLA synthesized as a function of the LA/Al molar ratio in the presence of 1B and BA.

Figure 3.

M_w (left axis) and yield (right axis) as a function of LA/Al molar ratio in the presence of 1B using 3x_LA (a) and 6x_LA (b). M_w (left axis) and yield (right axis) as a function of LA/Al molar ratio in the presence of BA using 3x_LA (c) and 6x_LA (d).

The first observation is that by increasing the temperature beyond 80 °C (compared to the polymers of Table S1 in the Supporting Information), higher LA/Al molar ratios up to 10 000 can be used for both initiators. A tentative explanation of this observation could rise from the competition to coordinate the aluminum between the lactide and water; in particular, by providing higher energy (higher temperatures), the lactide can displace the water coordinated. In this explanation, it is assumed that the water molecules demand lower energy to coordinate the aluminum center compared to the bulkier lactide.

Clear differences between the two initiators on the catalytic activity are observed when the LA/Al molar ratios are increased beyond 2000. Under the isothermal polymerization conditions of 100 and 120 °C for 1B and BA, respectively, the catalyst favors the ROP of l-lactide, resulting in high-molecular-weight PLLA (M_w up to 357 kg/mol) for 1B and very high molecular weight PLLA (M_w up to 694 kg/mol) for BA as initiators. For both initiators, the polymer yield drops with a substantial increase in the LA/Al molar ratio. The drop in yield may be attributed to the catalyst deactivation promoted by the increase in water concentration present in the l-lactide (LA). Differences in the outcome of the polymers synthesized using the two initiators can arise from the different temperatures used.

Comparing entries 9 and 10 shows the effect of monomer purification; when the monomer is not purified, the M_w value decreases almost 7 times. Also, the obtained yield decreases by almost a factor of two when using the raw monomer. The reason is attributed to the higher increased amount of water in the raw monomer.

To gain further insight into the capabilities of the catalytic systems to produce higher molecular weights and elucidate the influence of the water content, polymerizations using l-lactide recrystallized six times (having lower water content) are performed. According to Figure 3, the molecular weights of polymers synthesized using the six times recrystallized l-lactide are always higher compared to the polymers synthesized using the three times recrystallized l-lactide. Using the catalytic conditions of entries 21 and 22, UHMWPLLA (M_w = 1100–1400 kg/mol) can be obtained at high yields ≥95%. At the best of our knowledge, there are no other examples of catalytic systems capable to achieve the same values of molecular weights with the same efficiency in yield. It seems clear that by reducing the water content a significant increase in polymerization is achieved. The higher yields, obtained using the six times recrystallized l-lactide, further confirm the hypothesis that the water can compete with the l-lactide to coordinate the aluminum metal center. The lower values of molecular weights obtained using the three times recrystallized l-lactide (higher water content) could be explained by considering the following three factors arising due to higher water content:^33,47,33,47(1) promotion in higher degree of hydrolysis of the ester bonds; (2) reduction in the polymerization rate due to the competition in coordination the aluminum metal center; (3) promotion of the chain termination.

Figure 4 summarizes the influence of the water content of the molecular weight of the polymers produced as a function of the LA/Al molar ratio using the two initiators investigated. Using the six times recrystallized lactide, the water content increase very gradually as a function of LA/Al molar ratio resulting in higher molecular weights produced; on a contrary, using the three times recrystallized lactide, the water content quickly increases as a function of the LA/Al molar ratio, leading to lower molecular weights produced.

Figure 4.

Summary of the molecular characteristics (M_w) and total water content as a function of LA/Al molar ratio using 3x_LA and 6x_LA for 1B (a) and BA (b).

It becomes apparent that the use of BA as initiator provides higher yields and molecular weight compared to 1B in almost all of the polymerization reported in this work. This observation is in agreement with a previous report from Iwasa.³⁸ A tentative explanation for the improved performance using BA can be found in the lower water content (17 ppm) compared to 1B (101 ppm).

Similar to the previous polymers synthesized, the polymers of Table 3 show high melting temperatures (up to 178 °C) and crystallinities (up to 52%), which are indications of the high stereo-regularity of the PLLA synthesized. For most of the polymers tested, cold crystallization is observed between 90 and 110 °C and a classic glass-transition temperature around 53–60 °C (Figures S21–S24). As an example, Figure 5 summarizes the DSC thermograms of the polymers synthesized with 6x_LA initiated with BA.

Figure 5.

DSC thermograms of the polymers of Table 4, entries 17 (a) to 23 (g).

Using the same complex, Isawa et al.³⁸ and Pappalardo et al.³² reported the ROP of rac- and l-lactide achieving less than 28 000 and 50 000 g/mol, respectively (corrected M_w measured in THF vs polystyrene standards). Our study shows the capability of the {Me₂Al[O-2-tert-Bu-6-(C₆F₅N =CH)C₆H₃]} complex to promote the synthesis of PLLA achieving molecular weight up to 1.4 × 10⁶ g/mol (M_w measured in CHCl₃vs polystyrene standards).

3.5. Polymerization of d-Lactide

Polymerization of d-lactide using similar conditions leading to the synthesis of UHMWPLLA is attempted. Table 5 reports the polymerization conditions and the polymer characterization results of the ROP of d-lactide initiated with the aluminum-based complex (1) in the presence of BA; for comparison, the same conditions using l-lactide is also reported. The use of BA is preferred compared to the 1B due to its capability to promote higher catalyst activity and higher molecular weights.

Under the optimal polymerization conditions, UHMWPLLA and UHMWPDLA with relatively narrow polydispersity index and high yield can be obtained using the Al-based catalyst. It means that the Al-based catalyst is an effective and suitable catalyst for the ring-opening polymerization of lactide to achieve ultrahigh molecular weights as high as 1397 kg/mol under the synthesis conditions tested.

4. Conclusions

In this paper, we have studied the ROP of l-lactide and d-lactide promoted by an aluminum-based complex {Me₂Al[O-2-tert-Bu-6-(C₆F₅N=CH)C₆H₃]} initiated with 1B or BA under different reaction conditions. The monomer purification and its implications on the resulting molecular weight have also been investigated. The chosen catalytic system is found to be a successful catalyst to promote the synthesis of PLLA having a broad range of molecular weight with relatively narrow polydispersity index (1.1 ≤ Đ ≤ 2.1). The synthesis of ultrahigh molecular weight polylactides has been reported for the first time using the {Me₂Al[O-2-tert-Bu-6-(C₆F₅N=CH)C₆H₃]} catalyst initiated with BA. Different molecular weights can be synthesized by changing the reaction conditions (T, LA/Al, and polymerization time), as well as monomer purity. The water content in the monomer is found to play a crucial role in achieving ultrahigh molecular weights as well as higher yields. It is suggested that the water plays multiple roles in catalyst deactivation as well as being responsible for chain termination and/or hydrolysis of the synthesized polymer.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c01952.

• Details on the catalyst characterization, GC monomer characterization, GPC traces of the polylactides, and DSC thermograms (PDF)

The authors declare no competing financial interest.