Jean-Charles
Buffet
a,
Ashley N.
Martin
a,
Moshe
Kol
b and
Jun
Okuda
*a
aInstitute of Inorganic Chemistry, RWTH Aachen University, Landoltweg 1, D-52056, Aachen, Germany. E-mail: jun.okuda@ac.rwth-aachen.de; Fax: +49 241 80 92 644
bSchool of Chemistry, Raymond and Beverly Sackler Faculty of Exact Sciences, Tel Aviv University, Ramat Aviv, Tel Aviv 69978, Israel
First published on 15th August 2011
Group 4 metal complexes Zr-1 to Zr-4, Ti-4 and Ti-4a that contain an (OSSO)-type tetradentate bis(phenolate) ligand were found to initiate the ring-opening polymerization of meso-, rac-, and L-lactides in toluene solution. The polymerizations were controlled with initiator efficiency values around one and gave polymers with low polydispersity (Mw/Mn < 1.2). The polylactides were atactic when rac-lactide was used as the monomer and heterotactic biased when meso-lactide was used. Kinetic measurements showed that meso-lactide was polymerized generally faster than rac- and L-lactide.
![]() | ||
Scheme 1 Lactide monomers. |
Group 4 metal complexes are widely used as single-site catalysts for the polymerization of olefins.4 Currently, there is interest in the so-called post-metallocene complexes which may serve as new types of single-site polymerization catalysts.5 Polydentate ligands for metal complexation are crucial for structurally well-defined initiators for lactide polymerization as well: Davidson et al.6a and Eisen et al.6b recently reported titanium(IV) and zirconium(IV) complexes with two bidentate ligands, whilst Harada et al.7a and Mountford et al.7b introduced tridentate ligands. Recently, Ishii et al. reviewed group 4 metal complexes with (OSSO)-type bis(phenolato) ligands and their use as polymerization catalysts for olefins.8Kolet al. used an (OSSO)-type ligand that is analogous to the (ONNO)-Salan type ligand, containing a methylene group between the sulfur atoms and the phenol groups. The resulting group 4 metal complexes polymerized 1-hexene, affording stereoirregular poly(1-hexene).9 Furthermore, group 4 metal complexes containing an (OSSO)-type dithiobis(phenolate) ligand polymerized rac- and L-lactide efficiently, but with a low level of control.10 In contrast, initiators based on group 4 metal bis(phenolate) complexes catalyze the syndioselective polymerization of meso-lactide efficiently and in a controlled fashion.11
We report here the use of further group 4 metal complexes with an (OSSO)-type ligand framework as initiators for the polymerization of lactide monomers. The complexes reported are structurally classified according to their chelate ring size (5-5-5, 5-6-5 and 6-5-6) at the metal center.8,9
![]() | ||
Scheme 2 Group 4 metal complexes. |
Complexes Zr-1 to Zr-4 were synthesized by the reaction of [Zr(OtBu)4] with one equivalent of proligand H2-1 to H2-4, respectively, in n-pentane at 25 °C for 18 h. Complexes Ti-4 and Ti-4a were synthesized by the reaction of [Ti(OiPr)4] or [Ti(OiPr)3Cl] with one equivalent of H2-4 in toluene at 25 °C for 18 h. Complexes Zr-1 and Zr-3 were isolated as colorless solids after crystallization from a cold n-pentane solution (Fig. S1–S4†). Complexes Zr-1 and Zr-3 were obtained in 57% and 30% yields, respectively. The 1H NMR spectrum for complex Zr-1 shows a diagnostic doublet of doublets at 2.5 ppm for the bridging CH2 protons. In addition, a singlet at 1.85 ppm represents the methyl protons from the tert-butoxy group.
Stoichiometric reaction of proligand H2-4 with [Ti(OiPr)3Cl] at 25 °C in n-pentane afforded complex Ti-4a, as orange crystals in 78% yield after crystallization from a cold n-pentane solution (Fig. S11–S12†). A diagnostic multiplet signal is found at 5.10 ppm (1H) which corresponds to OCH(CH3)2. The bridging protons corresponding to SCH2CH2S and SCH2Ph are shifted downfield from 3.76 to 3.50 ppm and 2.57 to 1.92 ppm, respectively, upon complexation.
![]() | ||
Scheme 3 Stereoselective polymerization of meso-lactide to afford syndio- or hetero-tactic PLAs including their respective microstructures. |
Entry | Init. | Time/h | Conv.b (%) | M n,exp c/g mol−1 | f d | M w/Mnc | P s e |
---|---|---|---|---|---|---|---|
a Polymerization conditions: [LA]0/[init]0 = 100, [LA]0 = 0.520 M, T = 100 °C, toluene, 2 mL. b Conversion of monomer (([LA]0 − [LA]t)/[LA]0). c Measured by GPC with PS standards in THF, corrected by multiplication by 0.58.12 d Efficiency calculated using f = Mn,exp/Mn,theo with Mn,theo = [LA]0/[init]0 × 144.13 × conv. e P s is the probability of a new s dyad.13 f M n,theo = [LA]0/2[init]0 × 144.13 × conv. | |||||||
1 | Zr-1 | 0.5 | 29 | 6500 | 1.29 | 1.03 | 0.18 |
2 | Zr-1 | 1 | 55 | 9800 | 1.24 | 1.02 | 0.16 |
3 | Zr-1 | 4 | 95 | 19![]() |
1.23 | 1.07 | 0.20 |
4 | Zr-2 | 4 | 67 | 9200 | 0.95 | 1.06 | 0.24 |
5 | Zr-3 | 48 | 84 | 15![]() |
1.05 | 1.16 | 0.36 |
6 | Zr-4 | 24 | 47 | 8100 | 1.19 | 1.03 | 0.50 |
7 | Ti-4 | 72 | 55 | 4800 | 1.21f | 1.14 | 0.42 |
8 | Ti-4a | 72 | 48 | 4500 | 0.65 | 1.08 | 0.44 |
The polymerization of meso-lactide by complex Zr-1, based on 5-5-5 (OSSO)-type ligand 1, proceeded with efficiency values f of around 1.2. The efficiency value decreased with an increase in the steric bulk of the ligand's ortho- and para-substituents from tert-butyl, Zr-1 (f = 1.23), to cumyl (CMe2Ph), Zr-2 (f = 0.95). Complex Zr-3, based on flexible (OSSO)-type ligand 3 with a 5-6-5 chelate array, polymerized meso-lactide the fastest with efficiency value f = 1.05. However, complex Zr-4, based on (OSSO)-type ligand 4 with a 6-5-6 chelate array, polymerized meso-lactide the slowest with less control over the polymerization (f = 1.19). The poly(meso-lactides) synthesized by group 4 metal complexes with an (OSSO)-type ligand showed low polydispersity (1.02 < Mw/Mn < 1.16).
Poly(meso-lactides) produced using complexes Zr-1 and Zr-2 with the 5-5-5 chelate based on (OSSO)-type ligands 1 and 2 showed heterotactic biased PLAs (0.18 < Ps < 0.24). Heterotacticity decreased from a Ps of 0.20 to 0.50 when initiators 3 and 4 with more flexible ligands were used. When the metal is changed from Zr-4 to Ti-4, complexes with the 6-5-6 chelate based on the (OSSO)-type ligand, the heterotacticity did not change (Ps of 0.50 and 0.42, respectively). Previously, we reported that the group 4 metal bis(phenolate) complex with 5-5-5 and 5-6-5 (OSSO)-chelates led to syndiotactic biased PLAs.11 The somewhat unexpected formation of heterotactic PLA from meso-lactide is ascribed to site epimerization (Λ–Δ inversion) or to chain transfer after each ring-opening step, triggered by traces of acids.17c
In Fig. 1, 13C{1H} NMR spectra of the methine region are shown which indicate the decrease of heterotacticity with increasing flexibility of the ligand.
![]() | ||
Fig. 1
13C{1H} NMR spectra of the methine region of poly(meso-lactides) obtained at 100 °C with an initiator![]() ![]() ![]() ![]() |
![]() | ||
Scheme 4 Stereoselective polymerization of rac-lactide to afford atactic PLAs including its respective microstructure. |
Entry | Init. | Time/h | Conv.b (%) | M n,exp c/g mol−1 | f d | M w/Mnc | P s e |
---|---|---|---|---|---|---|---|
a Polymerization conditions: [LA]0/[init]0 = 100, [LA]0 = 0.520 M, T = 100 °C, C6D6, 0.5 mL. b Conversion of monomer (([LA]0 − [LA]t)/[LA]0). c Measured by GPC with PS standards in THF, corrected by multiplication by 0.58.12 d Efficiency calculated using f = Mn,exp/Mn,theo with Mn,theo = [LA]0/[init]0 × 144.13 × conv. e P s is the probability of a new s dyad.13 f M n,theo = [LA]0/2[init]0 × 144.13 × conv. | |||||||
1 | Zr-1 | 15 | 87 | 21![]() |
1.65 | 1.15 | 0.48 |
2 | Zr-4 | 141 | 90 | 8150 | 0.63 | 1.11 | 0.49 |
3 | Ti-4 | 255 | 69 | 4430 | 0.71f | 1.18 | 0.54 |
4 | Ti-4a | 255 | 86 | 5740 | 0.58 | 1.08 | 0.54 |
The polymerization of rac-lactide by complex Zr-1 led to a high efficiency value of PLA, indicating the possible occurrence of transesterification reactions (f = 1.65). The introduction of the flexible linker in complex Zr-4 decreased the efficiency of the polymerization of rac-lactide (f = 0.63). This mirrors the results found with meso-lactide. The polymerization of rac-lactide showed a slightly higher efficiency value when the metal was changed from Zr-4 to Ti-4 (f = 0.71). However, on going from the two isopropoxide groups to one (Ti-4a), the poly(rac-lactides) produced after 255 h but at a lower conversion level showed the lowest efficiency value (f = 0.58), indicating that the presence of the chloro ligand has a strong effect over the polymerization. Similar efficiency values for the ROP of rac-lactide were reported in the literature.7b,14 All the poly(rac-lactides) exhibited low polydispersities (1.08 < Mw/Mn < 1.18).
The ROP of rac-lactide using complexes Zr1, Zr4, Ti-4 and Ti-4a give an essentially atactic microstructure of poly(rac-lactides). The introduction of a methylene linker, change in the metal from zirconium to titanium, or in the number of initiating groups did not influence the microstructure (Fig. S17–S20†). The poly(rac-lactides) synthesized by using group 4 metal complexes have been shown in the literature to give isotactic,15 heterotactic,6a,14b,e or atactic polymers.6a,14c
![]() | ||
Fig. 2 Semilogarithmic plots of meso-lactide conversion vs. time, [LA]0/[init]0 = 100, [LA]0 = 0.520 M, T = 100 °C, benzene-d6 (0.5 mL): (a) using zirconium complexes: Zr-1 (kobs = 423 × 10−3 h−1), Zr-2 (kobs = 271 × 10−3 h−1), Zr-3 (kobs = 535 × 10−3 h−1), and Zr-4 (kobs = 38 × 10−3 h−1); and (b) using the flexible ligand 4: Ti-4 (kobs = 16 × 10−3 h−1) and Ti-4a (kobs = 10 × 10−3 h−1). |
The C3 bridged complex, Zr-3, based on the 5-6-5 (OSSO)-type ligand, showed the highest activity in the ROP of meso-lactide (kobs = 535 × 10−3 h−1), followed by the C2 bridged complex, Zr-1, based on the 5-5-5 (OSSO)-type ligand (kobs = 423 × 10−3 h−1). This is in agreement with our previous observation that the slowest among all of the titanium complexes tested was the one that possessed a rigid trans-1,2-cyclohexanediyl backbone and the fastest was the titanium complex with a C3 bridged backbone.11
Higher activity is observed with the complexes containing tert-butyl substituents, Zr-1 (kobs = 423 × 10−3 h−1), rather than cumyl (CMe2Ph), Zr-2 (kobs = 271 × 10−3 h−1). This is to be expected as coordination of the monomer to the Zr-1 complex with tert-butyl substituents is less sterically hindered, allowing easier monomer addition.
The introduction of a flexible linker (SCH2Ph) between the phenol ring and the sulfur bridge of the ligand has a significant effect on the polymerization rates of meso-lactide. Complex Zr-1, based on the 5-5-5 (OSSO)-type ligand, has 11 times higher activity (kobs = 423 × 10−3 h−1) than Zr-4, based on the 6-5-6 chelate array (kobs = 38 × 10−3 h−1).
This slow rate with the flexible ligand can also be observed by changing the metal center to titanium using complex Ti-4 (kobs = 16 × 10−3 h−1) which has twice as many initiating OiPr groups than Ti-4a which showed ca. 1.6 times faster rate for the polymerization of meso-lactide (kobs = 10 × 10−3 h−1). This is attributed to a combination of steric and strong electronic effects of the chloro ligand in comparison with those from a second growing chain (i.e. starting from the bis(isopropoxide) titanium initiator Ti-4).
The polymerization of meso-lactide using Zr-1 in toluene (kobs = 1360 × 10−3 h−1) showed a faster rate than in benzene-d6 (kobs = 423 × 10−3 h−1) (Fig. S25†). This rate for the polymerization of meso-lactide in toluene is comparable to the fastest rate reported in the literature by Feijen et al. with salan aluminium complexes.17c The polymerization rates for the polymerization of meso-lactide in benzene-d6 are the highest for group 4 metal initiators,11 but lower than those of other initiator metals reported in the literature.17
As with the ROP of meso-lactide, Zr-1 is faster than Zr-4 towards the ROP of rac-lactide (kobs = 142 × 10−3 h−1 and kobs = 21 × 10−3 h−1 respectively). Likewise, Zr-4 is faster than Ti-4. However, the rate of the polymerization of rac-lactide using Ti-4 showed an initiation period with a higher rate constant (kobs = 14 × 10−3 h−1) followed by a slower period after around 50% conversion (kobs = 5 × 10−3 h−1). Interestingly, the rate for the slower second phase was equal to that of propagation using Ti-4a (kobs = 5 × 10−3 h−1). This could be due to the growing polymer chains becoming too large for initiation by both isopropoxide groups. A rate similar to that of the mono(isopropoxide) complex Ti-4a may result. The titanium and zirconium complexes polymerized rac-lactide more slowly than other group 4 metal complexes reported in the literature.6a,7b,14b
Kinetic studies of the polymerization of L-lactide at 100 °C in benzene-d6 with an initiator:
monomer ratio of 1
:
100 using Zr-1 and Zr-4 are shown in Fig. S27 and Table S11–S14†. As was the case for meso- and rac-lactides, Zr-1 polymerizes L-lactide faster (kobs = 164 × 10−3 h−1) than does Zr-4 (kobs = 24 × 10−3 h−1). In addition, zirconium initiators polymerized L-lactide faster than the titanium homologs (kobs = 3 × 10−3 h−1 for Ti-4, kobs = 7 × 10−3 h−1 for Ti-4a) (Fig. S28–S29†).
The kinetic results of the polymerization of meso-, rac- and L-lactides using Zr-1 and Zr-4 are given in Fig. 3. meso-Lactide is polymerized faster than rac- and L-lactides, when using Zr-1 and Zr-4 (Table 3). This can be attributed to the fact that the RS configuration of meso-lactide is more strained in comparison to the RR and SS configuration found in rac-lactide.18 Furthermore, rac- and L-lactide are polymerized with similar propagation rates relative to each other (Table 3).
![]() | ||
Fig. 3 Semilogarithmic plots of lactide monomer conversion vs. time, [LA]0/[init]0 = 100, [LA]0 = 0.520 M, T = 100 °C, benzene-d6 (0.5 mL): (a) Zr-1meso-lactide (kobs = 423 × 10−3 h−1), L-lactide (kobs = 164 × 10−3 h−1), and rac-lactide (kobs = 142 × 10−3 h−1); and (b) Zr-4meso-lactide (kobs = 38 × 10−3 h−1), L-lactide (kobs = 24 × 10−3 h−1), and rac-lactide (kobs = 21 × 10−3 h−1). |
Entry | Initiator | Monomer | Solvent | k obs/×10−3 h−1 |
---|---|---|---|---|
a Polymerization conditions: [LA]0/[init]0 = 100, [LA]0 = 0.520 M, T = 100 °C. | ||||
1 | Zr-1 | meso | C6D6 | 423 |
2 | Zr-2 | meso | C6D6 | 271 |
3 | Zr-3 | meso | C6D6 | 535 |
4 | Zr-4 | meso | C6D6 | 38 |
5 | Ti-4 | meso | C6D6 | 16 |
6 | Ti-4a | meso | C6D6 | 10 |
7 | Zr-1 | meso | Toluene | 1360 |
8 | Zr-1 | rac | C6D6 | 142 |
9 | Zr-4 | rac | C6D6 | 21 |
10 | Ti-4 | rac | C6D6 | 5 |
11 | Ti-4a | rac | C6D6 | 5 |
12 | Zr-1 | L | C6D6 | 164 |
13 | Zr-4 | L | C6D6 | 24 |
14 | Ti-4 | L | C6D6 | 3 |
13 | Ti-4a | L | C6D6 | 7 |
Complexes with the 5-6-5 chelate derived from the C3-bridged (OSSO)-type ligand 3 led to higher rates of the polymerization of the lactide monomers than those based on C2 bridged ligands 1 and 2 with the 5-5-5 chelate array. The introduction of the methylene linker between the sulfur donor and the phenolate group in the complexes based on C2 bridged ligand 4 with the 6-5-6 chelate resulted in lower polymerization rates.
Chemical shifts for 1H and 13C{1H} NMR spectra were recorded on a Bruker Avance 400 MHz spectrometer or a Varian NMR 200 MHz spectrometer at room temperature in 5 mm NMR tubes. Chemical shifts were reported in parts per million and referenced against TMS using the residual proton signal of the solvent (1H: benzene-d6, δ = 7.16 ppm; chloroform-d1, δ = 7.26 ppm), (13C{1H}: benzene-d6, δ = 128.06 ppm; chloroform-d1, δ = 77.00 ppm).
Molecular weights and polydispersities were determined by size exclusion chromatography (SEC) in THF at 35 °C, at a flow rate of 1 mL min−1 utilizing an Agilent 1100 Series HPLC, a G1310A isocratic pump, an Agilent 1100 Series refractive index detector and 8 × 600 mm, 8 × 300 mm, 8 × 50 mm PSS SDV linear M columns. Calibration standards were commercially available narrowly distributed linear polystyrene samples that cover a broad range of molar masses (103 < M < 2 × 106 g mol−1). The molecular weights were multiplied by 0.58 according to a method in the literature.12
Footnote |
† Electronic supplementary information (ESI) available: NMR data of complexes Zr-1 to Zr4, Ti-4 and Ti-4a; NMR of the microstructures of polylactides and tables for the kinetic of polymerization data. See DOI: 10.1039/c1py00266j |
This journal is © The Royal Society of Chemistry 2011 |