Novel catalyst for polymerisation

The present invention relates to metal/organic complexes of Formula (I), (II) (III), (IV), (V) and (VI) that are useful as catalysts for the polymerisation of carbonyl- containing or cyclic monomers. Typical polymerisation reactions are, for example, those of lactides.

The compounds of the present invention are metal/organic complexes and are complexes are alkoxides or aryloxides formed from chiral, bidentate ligands. They are particularly useful for stereoselective polymerisation of these monomers. The complexes are alkoxides or aryloxides formed from chiral bidentate ligands and single metal cations and are of the general structures below where R may be selected from the group consisting of hydrogen, hydrocarbyl or substituted hydrocarbyl and M may be any Lewis-acidic metal, for example the s-block, f-block metals or scandium, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, yttrium, zirconium, tin or aluminium. Preferentially the metal may be an f-block metal. More preferably the metal may be from the lanthanide series, for example europium or erbium.

Description of the prior art

It is known that metal alkoxides are active ring-opening polymerisation catalysts. A number of metal alkoxides have been used in polymerisation reactions. Examples include tin, aluminium and zinc.

A widely used catalyst for the preparation of poly lactide is tin(II)octanoate [tin(II)bis(2-ethylhexanoate), Sn(OCt) ₂ ] (Chem. Rev. 104: 6147-6176 (2004)). However, the use of a tin-based catalyst may not be appropriate where the polymer is to be used in a biomedical application as tin is toxic and there may be traces of the tin catalyst in the polymer product. Also, tin(II)octanoate requires activation with an alcohol and activity of the catalyst is generally low. The structure of tin(II)octanoate is given below:

Aluminium alkoxides are less active than tin(II)octanoate (Am. Chem. Soc. 121 : 4072-4073 (1999)) and there are concerns about the use of aluminium as catalyst for polymerisation of biomedical polymers as it has been linked to Alzheimer's disease. The structure of an aluminium alkoxide is given below:

Zinc alkoxides are considered to be non-toxic, however their activity is low.

The use of yttrium and rare earth metals for the catalysis of lactone polymerisation is the subject of US patent applications 5,028,667 and 5,235,031 and PCT application number WO9619519. None of these documents report the use of chiral ligands to achieve stereoselective polymerisation and therefore the present invention is novel.

Commercial polylactides are synthesised from lactide monomers prepared from a single lactic acid enantiomer in order to obtain stereoregular polymers with a high degree of crystallinity. Polylactides derived from racemic lactide are amorphous with a lower glass transition temperature.

It has been reported that L-polylactide and D-polylactide form a stereocomplex with a melting temperature 5O ⁰ C greater than the homochiral polymers. Preparation of such a stereocomplex currently requires parallel ring-opening polymerisation of D- lactide and L-lactide and subsequent combination of the chiral polylactide chains. US patent applications 4,800,219, 4,766,182 and 4,719,246 describe polylactide compositions with enhanced physical properties. These compositions are obtained by mixing single enantiomers of D- and L-lactide in order to obtain stereocomplex polylactide.

Despite the improved physical properties of the stereocomplex, practical applications of the stereocomplex are restricted by the requirement for separate pools of enantiopure lactide monomers to generate enantiopure polymers i.e. there is a need to devise a method for preparing stereocomplex polylactide from racemic lactide monomer (J. Am. Chem. Soc. 122: 1552-1553 (2000)). An aluminium alkoxide catalyst has been generated that permits stereoselective polymerisation, however the activity of the polymer is low and the molecular weight of the resulting polymers is not sufficient for industrial applications such as packaging (Macromolecular Chemistry and Physics 197(9) : 2627-2637 (1996)).

It is therefore an object of the present invention to provide novel metal/organic complexes suitable for use as polymerisation catalysts. Another object of the present invention is to provide improved catalysts which are able to operate under more environmentally friendly conditions e.g. at lower temperatures or in more environmentally friendly solvents. It is a further object of the present invention to provide improved catalysts that are capable of rapidly polymerising a monomer. It is a further object of the present invention to provide improved catalysts with reduced toxicity. It is yet another object of the present invention to provide improved catalysts which are capable of producing higher molecular weight polymers. It is yet another object of the present invention to provide improved catalysts which are capable of producing low polymer dispersity polymers.

Summary of the invention

The present invention fulfils all or some of the above objects of the invention.

The present invention discloses new metal/organic complexes that are useful as catalysts for the polymerisation of carbonyl-containing or cyclic monomers, for example lactide. The complexes are particularly useful for stereoselective polymerisation of these monomers.

According to the first aspect of the present invention, there is provided a compound of Formula (I), (II), (III), (IV), (V) or (VI) :

(I) (H)

(IV) (V) (Vl) wherein R is independently selected at each occurrence from the group comprising : hydrogen, hydrocarbyl and substituted hydrocarbyl,

M is a Lewis-acidic metal and

X, if present, is any suitable counter ion.

In one embodiment, the complexes are alkoxides or aryloxides formed from chiral bidentate ligands and single metal cations. In an alternative embodiment, the complexes are alkoxides or aryloxides formed from chiral tridentate ligands and double metal cations. In another alternative embodiment, the complexes are alkoxides or aryloxides formed from a mixture of chiral bidentate and chiral tridentate ligands and single metal cations.

The drawings are not intended to limit the invention to any specific stereoisomer. All potential stereoisomers arising from planar, axial or centrosymmetric stereoelements are claimed herein.

In another aspect the present invention also discloses the use of these catalysts for stereoselective polymerisations of carbonyl-containing or cyclic monomers, for example lactide, glycolide, ε-caprolactone or ε-caprolactam.

The use of such stereoselective catalysts confers more precise control over the properties of a polymer and to allow more efficient polymer production. The resulting polymers have a number of applications in the biomedical industry e.g. surgery (tissue or bone repairing, sutures and controlled release drug delivery), food packaging (as a polyethylene alternative), agriculture and the engineering industry.

Inevitably trace amounts of catalyst are present in the resulting polymer and for this reason the catalysts of the present invention are particularly useful in producing polymers used in food and medical applications due to their low toxicity.

An example polymer which can be produced by a catalyst of the present invention is poly lactic acid (PLA). PLA is both biodegradable and bioassimilable. An additional environmental benefit with PLA is that the monomer, D,L-lactide is readily available by the fermentation of corn starch (a carbon neutral process). The molecular weight range of PLA is controllable between 1000 and 500000 g/mol and is dependent upon the catalyst used and conditions employed. The mechanical properties of PLA range from viscous oils and soft elastic plastics to stiff, high strength materials comparable to polyethylene.

In another aspect of the present invention, these catalysts may also be used for asymmetric Lewis-acid catalysed reactions, for example chiral Diels Alder reactions, asymmetric aldol (or aldol derivative) reactions.

Detailed description of the invention

In one aspect, the present invention relates to metal/organic complexes of Formula (I), (II) (III), (IV) (V) and (VI) that are useful as catalysts for the polymerisation of carbonyl-containing or cyclic monomers.

In any of the above embodiments, the substituted hydrocarbyl group may be substituted with one or more heteroatoms. Preferred heteroatoms include N, S, O, and Si.

M may be selected from s-block, p-block, d-block and f-block metals. M may be any Lewis-acidic metal, for example lithium, beryllium, sodium, magnesium, potassium, calcium, rubidium, strontium, caesium, barium, francium, radium, scandium, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, yttrium, zirconium, tin, aluminium, lanthanum, cerium, praseodymium, neodymium, samarium, europium, gadolinium, terbium, dysprosium, holmium, erbium, thulium, ytterbium, lutetium, thorium, protactinium, uranium, neptunium, plutonium, americium, curium, berkelium, californium, einsteinium, fermium, mendelevium, nobelium and lawrencium.

In an embodiment, the metal is selected from magnesium, calcium, scandium, titanium, vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, yttrium, zirconium, europium, erbium, tin or aluminium. Preferentially the metal may be an f-block metal. More preferably the metal may be from the lanthanide series, for example europium or erbium. Preferentially the metal is selected from the group comprising: magnesium, calcium, titanium, zinc, yttrium, europium, erbium, ytterbium, tin or aluminium.

In an embodiment, each R group is optionally substituted where chemically possible with 1 to 3 substituents selected from the group consisting of halo, hydroxy, oxo, cyano, mercapto, nitro, (Cl-C4)alkyl, and (Cl-C4)haloalkyl.

In an embodiment, each R is independently selected from the group comprising : a) (Cl-C6)alkyl, (C2-C6)alkenyl, (C2-C6)alkynyl, (Cl-C6)alkoxy, (Cl-C6)alkyl-S-, (Cl-Cδ)alkylamino, and di[(Cl-C6)alkyl]amino; wherein each of said groups may optionally be substituted where chemically possible with 1 to 3 substituents

independently selected from the group consisting of halo, hydroxy, cyano, mercapto, nitro, (Cl-C4)alkyl, and (Cl-C4)haloalkyl; or b) 5- to 10-membered heteroaryl containing 1 or 2 ring heteroatoms independently selected from the group consisting of N, S or O; wherein said heteroaryl ring may optionally be substituted with 1 to 3 substituents per ring independently selected from the group comprising: halo, hydroxy, cyano, mercapto, nitro, (Cl-C4)alkyl, (Cl-C4)haloalkyl, (C2-C4)alkenyl, (C2-C4)alkynyl, (Cl-C4)alkoxy, or c) phenyl, naphthyl, anthracenyl, phenanthranyl, and indenyl, wherein each of the foregoing groups is optionally be substituted with 1 to 3 substituents per ring independently selected from the group comprising: halo, hydroxy, cyano, mercapto, nitro, (Cl-C4)alkyl, (Cl-C4)haloalkyl, (C2-C4)alkenyl, (C2-C4)alkynyl, (Cl- C4)alkoxy.

In a preferred embodiment, each R is independently selected from the group comprising: a) (Cl-C6)alkyl, (C2-C6)alkenyl, and (C2-C6)alkynyl; or b) 5- or 6-membered-heteroaryl containing 1 or 2 ring heteroatoms independently selected from the group consisting of N, S or O; or c) phenyl and naphthyl.

When an individual R group is alkyl, it is preferably propyl or butyl. Most preferably it is t-butyl. When an individual R group is an aryl group, it is preferably a phenyl group which may be optionally substituted with 1 to 3 independently chosen substituents selected from halogen, CN, OH, NO ₂ , Ci _-4 alkyl and Ci _-4 alkoxy.

In a second aspect, the invention is related to the use of the catalysts of the present invention for stereoselective polymerisations of carbonyl-containing or cyclic monomers, for example lactide, glycolide, ε-caprolactone or ε-caprolactam.

Not meaning to be bound by theory, it is thought that the mechanism for the ring opening polymerisation (ROP) of D,L-lactide follows the route illustrated in scheme 3a:

Scheme 3a: ROP of D,L-lactide.

It is already known in the prior art that if one enantiomer of lactide is polymerised, e.g. D-lactide, then the resulting polylactide is the D enantiomer, D-polylactide. Likewise if L-lactide is polymerised the resulting PLA is L-polylactide. It is also known that if L-polylactide and D-polylactide are mixed and annealed, the L and D enantiomers form a more stable stereocomplex which has a melting point 50 ⁰ C higher than either L-lactide or D-lactide. The increase in melting point is believed to be due to the complementary interaction between each enantiomer. This is illustrated in scheme 3b:

plex C

Scheme 3b: Stereocontrol of lactide ring opening polymerisation.

If a racemic mixture of D, L-lactide is polymerised with a racemic mixture of a catalyst of the present invention, a mixture of D- and L-polylactide is produced. Annealing this mixture allows the formation of a stereocomplex. Figure 32 illustrates that after thermal annealing (180 ⁰ C, 5 min) the polymer exhibits a sharper T _q peak and a higher melting point suggesting the formation of the stereocomplex.

The increased stability and higher melting point of the stereocomplex increases the number of potential uses for the polymer. For example the polymer stereocomplex will have many useful applications in engineering.

General Procedures

In one embodiment, the novel catalysts are prepared from chiral bidentate ligands as described herein.

One method of preparing the chiral bidentate ligand is illustrated in reaction scheme Ia :

1 . n-BuLi

¹ BuCI

Scheme Ia: synthesis of a chiral bidentate ligand.

In another embodiment, the novel catalysts are prepared from chiral tridentate ligands.

One method of preparing the chiral tridentate ligands is illustrated in the reaction scheme Ib:

1 . n-BuLi

¹ BuCI

H ₂ L ²

Scheme Ib: synthesis of a chiral tridentate ligand.

Bimetallic, tridentate ligand complexes (of formula (V)) can be produced by reaction scheme 2a :

thf

2MN" ₃ + 3H ₂ L ² ► M ₂ H ₂ L ² ₄

-3HN"

Scheme 2a: synthesis of a chiral, bimetallic tridentate ligand complex.

In another embodiment, the novel catalysts are prepared from both chiral bidentate and chiral tridentate ligands.

Mixed bidentate/tridentate ligand complexes (of formula (VI)) can be produced by the reaction scheme 2b :

thf

2MN" ₃ + 2 HL ¹ + H ₂ L ² ► MHLS(HL ² )

-3HN"

Scheme 2b: synthesis of a chiral, bidentate/tridentate ligand complex. The invention is illustrated by way of example only by the following Figures:

Figure 1 : X-ray crystal structure of the bidentate ligand precursor, HL ¹ .

Figure 2 : X-ray crystal structure of the tridentate ligand precursor, H ₂ L ² .

Figure 3 : X-ray crystal structure of the bidentate ligand complex ML ^X ₃ , formula (II). Figure 4: X-ray crystal structure of the tridentate ligand complex M ₂ H ₂ L ² ₄ , formula (V). Figure 5 : X-ray crystal structure of the mixed bidentate/tridentate ligand complex ML^(HL ² ), formula (VI) Figure 6 : X-ray crystal structure of the amide ligand complex ML ^X ₂ N".

Figure 7A/B: X-ray crystal structures of (A) ligand 1 and (B) catalyst 1. Figure 8: M _n over time for reactions 1 - 7. Figure 9: M _n over conversion for reactions 1 - 7. Figure 10: Conversion over time for reactions 1 - 7. Figure 11 : GPC data for polymer samples from reaction 8. Figure 12: GPC data for polymer samples from reaction 9. Figures 13A/B/C: (A) ¹ H NMR, (B) homonuclear decoupled ¹ H NHR and (C) ¹³ C NMR spectra of polymer produced using D^-lactide and 1 % catalyst 2.

Figures 14A/B: (A) standard ¹ H NMR and (B) ¹³ C NMR spectra for polymer made using L-lactide and 1 % catalyst 2. Figure 15: electrospray mass spectrum of a relatively short chain polymer i.e. n = 3 to 9. Figure 16: (A) DSC data for PLA prior to annealing; T _g : 55 ⁰ C, Mp : 180- 190 ⁰ C (B) DSC data for PLA after annealing at 220 ⁰ C for 2 min; sharper T _g peak and higher Mp (21O ⁰ C).

Figure 17: Polymerisation results in THF for polymerisation of D, L-lactide using ErLS (1%) at 0 ⁰ C. Figure 18: Polymerisation results in DCM for polymerisation of D, L-lactide using ErLS (1%) at 0 ⁰ C.

Figure 19: ¹ H-NMR data for the polymerisation reaction. Figure 20: Gel permeation chromatography for the polymer detailed in figure 19

Figure 21 : Data comparison using 6 ^tBu with and without coinitiator benzyl alcohol.

Figure 22: GPC characterisation for complexes 4-6 and Sn(oct) ₂ . Figure 23: ¹ H NMR spectra (300 MHz in CDCI ₃ ) of PLA methine resonances with selective decoupling of PLA methyl resonances: (a) L-PLA prepared by ROP of L-lactide by 4 ^tBu ,(b) rac-PLA prepared by ROP of rac-lactide by 4 ^tBu and (c) rac-PLA prepared by ROP of rac-lactide with Sn(OCt) ₂ (tin (II)bis(2-ethylheanoate)). ¹⁸

Figure 24: ¹ H NMR spectra (300 MHz in CDCI ₃ ) of PLA methine resonances with selective decoupling of PLA methyl resonances: (a) L-PLA prepared by ROP of L-lactide by 4 ^tBu ,(b) rac-PLA prepared by ROP of rac-lactide by 4 ^tBu .

Figure 25: M _n and PDI versus conversion and Ln (l/(l-conv.)) versus the time of polymerisation for the polymerisation of D, L-lactide by 4 ^tBu .

Figure 26: Conversion versus the time of polymerisation for the polymerisation of D, L-lactide by 8 ^ph . Figure 27: Conversion versus the time of polymerisation for the polymerisation of D, L-lactide by ll ^tBu . Figure 28: ¹ H and ¹³ C NMR spectra (300 MHz in CDCI ₃ ) of PLGA. (a) PLGA prepared by ROP using 4 ^tBu after 6 h, (b) PLGA prepared by ROP using 4 ^tBu after 24 h, (c) PLGA prepared by ROP using 4 ^tBu after 24 h.

Figure 29: M _n and PDI versus conversion and conversion versus the time of polymerisation for the copolymerisation of D, L-lactide and glycolide by 4 ^tBu .

Figure 30: GPC chromatogram of the copolymerisation of glycolide and lactide using 4 ^tBu following the time of the polymerisation. Figure 31a-e: NMR Spectral characterization of polymers.

Figure 32: differential scanning calorimetry of D, L-PLA produced using a catalyst of the present invention (A) prior to annealing at 180 ⁰ C and (B) after annealing at 180 ⁰ C.

Figure 1 illustrates an x-ray crystal structure of a ligand used in the preparation of a catalyst of the present invention. The P - O bond length is 1.507 A, the P - C bond length is 1.816 A and the O - O bond length is 2.777 A. Additionally ³¹ P - NMR shows a P resonance at δ 65.8 ppm.

Figure 2 illustrates an x-ray crystal structure of another ligand used in the preparation of a catalyst of the present invention. The P - O bond length is 1.504 A, the P - C bond length is 1.816 A and the 0 - 0 bond length is 2.787 A. Additionally ³¹ P - NMR shows a P resonance at δ 63.9 ppm.

Figure 3 illustrates an x-ray crystal structure of a catalyst of the present invention. R is ¹¹ Bu and M can be any of Eu, Er, Y or Yb. The M - O= P bond length when M = Er is 2.32 A, when M = Eu 2.42 A and when M = Y is 2.37 A.

Figure 4 illustrates an x-ray crystal structure of another catalyst of the present invention. The Eu - Eu distance is 3.762 A.

Figure 5 illustrates an x-ray crystal structure of another catalyst of the present invention. This catalyst has both bidentate and tridentate ligands.

Figure 6 illustrates an x-ray crystal structure of another catalyst of the present invention. The Er - N bond length is 2.28 A [compared with 2.21 A in Er[N(SiMe ₃ ) ₂ ] ₃ , the Er - O=P bond length is 2.29 A and the Er - 0-C bond length is 2.09 A.

Figure 7A illustrates an x-ray crystal structure of ligand 1. Figure 7B illustrates an x-ray crystal structure of catalyst 1.

Figure 8 illustrates the M _n over time for reactions 1 - 7. This shows that after 8 minutes the molecular weight of the polymer has reached its maximum value of 130000 g/mol for reactions 1 - 7.

Figure 9 illustrates the M _n over conversion for reactions 1 - 7. This shows that the 100% conversion corresponds to a molecular weight of 130000 g/mol .

Figure 10 illustrates the conversion over time for reactions 1 - 7. This shows that 100% conversion is reached after 8 minutes reaction time.

Figures 11 illustrates gel permeation chromatography data from reactions 8 of example 4.

Figures 12 illustrates gel permeation chromatography data from reactions 9 of example 4.

Figure 13 illustrates (A) standard ¹ H NMR, (B) homonuclear decoupled ¹ H NHR and (C) ¹³ C NMR spectra of polymer produced using D,L-lactide and 1 % catalyst 2.

Figure 14 illustrates (A) standard ¹ H NMR and (B) ¹³ C NMR spectra for polymer made using L-lactide and 1 % catalyst 2.

Figure 15 illustrates electrospray mass spectrum of a relatively short chain polymer i.e. n = 3 to 9.

Figure 16 illustrates (A) DSC data for PLA prior to annealing; T _g : 55 ⁰ C, Mp : 180- 190 ⁰ C (B) DSC data for Pl-A after annealing at 220 ⁰ C for 2 min; sharper T _g peak and higher Mp (21O ⁰ C).

Figure 17 illustrates the polymerisation results in THF for the polymerisation of D, L - Lactide using ErL-S (1%) at 0 ⁰ C. This demonstrates the rapid conversion of D, L- lactide to Pl-A using ErL-S in THF (60% of the D,L-lactide is converted to Pl-A in under 10 minutes). The maximum conversion that can be achieved is approximately 65%. Figure 17 also illustrates the maximum molecular weight of Pl-A that can be achieved using THF as the solvent is 160000 g/mol . The molecular weight (length) of the polymer can be tailored by altering the reaction time.

Figure 18 illustrates the polymerisation results in DCM for the polymerisation of D, L - Lactide using ErL^ (1%) at 0 ⁰ C. This demonstrates that higher conversion levels (up to 100% conversion) can be achieved using ErLS in DCM (than for THF). However, the maximum molecular weight is lower when DCM is the solvent as opposed to THF. The reaction time required to achieve nearly full conversion is approximately 8 minutes.

Table 7 provides a comparison of the use of ErLS (the catalyst presented in figure 18) and prior art catalysts to catalyse the conversion of D,L-lactide to PLA. Much more rapid conversion is achieved irrespective of the solvent used (figures 17 and 18 illustrate the use of both coordinating and non-coordinating solvents) when ErLS is employed rather than a catalyst of the prior art. Additionally, the molecular weight of the polymer produced using this catalyst is much higher than for polymers produced using prior art catalysts. Higher molecular weight polymers hydrolyse slower than shorter polymers which is beneficial for important instances e.g. longer- lasting polymers for engineering applications. Other benefits of using ErLS include low polymer dispersion values and low toxicity.

Table 7

Table 8 provides examples of polymerization under different reaction conditions. The reactions for catalysts of the present invention (table 8, DCM) were carried out at - 18°C which is much lower than the temperature traditional methods employing

Sn(OCt) ₂ are carried out at. This illustrates the economic and environmental benefits of using a catalyst of the present invention e.g. greater energy efficiency. Additionally, because the reaction employing a catalyst of the present invention may be carried out in a range of solvents, (see figures 17 and 18) this allows a greater degree of choice with regard to other environmental and economic considerations.

Table 8: ROP of D,L-lactide

Figure 19 provides ¹ H-NMR data for the polymerisation reaction using 2% ErL-S in THF at 20 ⁰ C. The three portions of spectra at ca. 5ppm are for the C-H resonances and are well separated from the methyl (CH ₃ ) resonances at ca l.δppm. The left spectrum (marked "30s") corresponds to the monomer which possesses two close- lying resonances as seen in the spectrum. As the polymerisation progresses, the monomer is ring-opened (e.g. the mechanism given in scheme 3a). Only one IH resonance is obtained from the protons present in the polymer chain (attached to the same carbon atom as the methyl groups), indicating that the protons are equivalent due to the formation of an isotactic chain.

Figure 20 illustrates gel permeation chromatography for the polymer detailed in figure 19 (2% Erlλ, THF, 20 ⁰ C), using a CHCI ₃ / polystyrene standard. As can be seen, the peak at retention time ~13min is unsymmetrical and is near the "high limit" (for reliable detection). The following values can be derived from the graph Mw = 150000, Mn = 75000 and PD (Mw/Mn) = 2.0

Figure 21 provides data for the comparison of the reaction using 6 ^tBu with and without coinitiator benzyl alcohol.

Figure 22 illustrates GPC characterisation for complexes 4-6 and Sn(oct) ₂ .

Figure 23 illustrates ¹ H NMR spectra (300 MHz in CDCI ₃ ) of PLA methine resonances with selective decoupling of PLA methyl resonances: (A) shows L-PLA prepared by ROP of L-lactide by 4 ^tBu , (B) shows rac-PLA prepared by ROP of rac-lactide by 4 ^tBu and (C) shows rac-PLA prepared by ROP of rac-lactide with Sn(OCt) ₂ (tin (II)bis(2- ethylheanoate).

Figure 24 illustrates ¹ H NMR spectra (300 MHz in CDCI ₃ ) of PLA methine resonances with selective decoupling of PLA methyl resonances: (A) shows L-PLA prepared by ROP of L-lactide by 4 ^tBu and (B) rac-PLA prepared by ROP of rac-lactide by 4 ^tBu .

Figure 25 illustrates M _n and PDI versus conversion and Ln (l/(l-conv.)) versus the time of polymerisation for the polymerisation of D, L-lactide by 4 ^tBu .

Figure 26 illustrates the conversion versus the time of polymerisation for the polymerisation of D, L-lactide by 8 ^ph .

Figure 27 illustrates the conversion versus the time of polymerisation for the polymerisation of D,L-lactide by ll ^tBu .

Figure 28 illustrates ¹ H and ¹³ C NMR spectra (300 MHz in CDCI ₃ ) of PLGA. (A) shows PLGA prepared by ROP using 4 ^tBu after 6 h, (B) shows PLGA prepared by ROP using 4 ^tBu after 24 h and (C) shows PLGA prepared by ROP using 4 ^tBu after 24 h.

Figure 29 illustrates M _n and PDI versus conversion and conversion versus the time of polymerisation for the copolymerisation of D,L-lactide and glycolide by 4 ^tBu .

Figure 30 illustrates a GPC chromatogram of the copolymerisation of glycolide and lactide using 4 ^tBu following the time of the polymerisation.

Figure 31 illustrates NMR spectral characterization of polymers: a) methine region of the homonuclear decoupled ¹ H-NMR for entry 1. Integration of the iii peak corresponds to 26.2 %. ¹ H-NMR 5(CDCI ₃ ) : 5.146, 5.161, 5.171, 5.178, 5.181, 5.185, 5.202 [ppm]. b) methine region of the homonuclear decoupled ¹ H-NMR for entry 2. Integration of the iii peak corresponds to 88.8 %. ¹ H-NMR 5(CDCI ₃ ) : 5.103, 5.181, 5.200 [ppm]. c) methine region of the homonuclear decoupled ¹ H-NMR for entry 3. Integration of the iii peak corresponds to 78.7 %. ¹ H-NMR 5(CDCI ₃ ) : 5.144, 5.160, 5.178, 5.198, 5.211, [ppm]. d) methine region of the homonuclear decoupled ¹ H-NMR for entry 4. Integration of the iii peak corresponds >99 %. ¹ H-NMR 5(CDCI ₃ ): 5.151 ppm. e) methine region of the homonuclear decoupled ¹ H-NMR for entry 5. Integration of the iii peak corresponds to 36.1 %.

¹ H-NMR S(CDCI ₃ ): 5.097, 5.126, 5.139, 5.165, 5.179, [ppm].

Figure 32 illustrates differential scanning calorimetry of D, L-PLA produced using a catalyst of the present invention (A) prior to annealing at 180 ⁰ C and (B) after annealing at 180 ⁰ C.

Specific embodiments of the present invention are illustrated in the following examples. The examples should no be interpreted as limiting to the scope of the present invention.

Examples

Example 1 - this example illustrates the synthesis of proligands.

Synthesis of HL ^R

The synthesis of the proligand requires three steps. First a double Grignard reaction between magnesium tertiobutyl chloride and PBr ₃ yields ¹¹ Bu ₂ PBr (Scheme 4). The compound was obtained as a yellow oil and purified by distillation under reduced pressure (10 ^"2 mbar); pure ¹¹ Bu ₂ PBr was isolated as a colourless oil, characterised by ¹ H and ³¹ P NMR spectroscopy.

Bu' Mg PBr ₃ \

'BuQ * ^► 'BuMgCl ► I' Br

S ₂ O, reflux * H ₂ Q O ⁰ C /

Bu'

Scheme 4: Preparation of di-tert-butyl precursor

¹¹ Bu ₂ PBr was treated with LiAIH ₄ , yielding ¹¹ Bu ₂ PH, which was subsequently treated with nBuϋ to make LiP ¹¹ Bu ₂ which was treated with 3,3-dimethyl- epoxybutane, and the resulting compound oxidised with H ₂ O ₂ to give the targeted proligand HL ^R in a modified procedure based on that of Genov D., Kresinski R., Tebby J., J. Org. Chem, 1998, 63, 2574.

The general synthesis for HL ^R : R = ¹¹ Bu 1, R = Ph 2 is shown Scheme 5. An analogue R-HL ^R Ia was synthesised by a R-epoxide following the same procedure.

Scheme 5: Synthesis of HL ^R , R = ¹ Bu (1) and Ph (2). Synthesis of LiL ^R

A THF solution of HL ^ph 2 was treated with nBuLi, yielding LiL ^ph 3 (Scheme 6).

Scheme 6: Synthesis of lithium salt of the diphenyl proligand. Example 2 - this example illustrates the synthesis of catalysts.

A range of metal complexes of L ^R were synthesised using a variety of different metal starting materials, as shown in scheme 7. All reactions were conducted in toluene at 80 ⁰ C overnight.

5

Scheme 7: Synthesis of ML ₂ ^R and ML ₃ ^R complexes.

All the complexes were characterised by ¹ H and ³¹ P and some also by mass spectroscopy analysis and X-Ray crystallography.

Synthesis of catalysts from MCI ₂ /HL ^R

In this route, a metal dichloride salt was treated with two equivalents of the ligand in toluene at 70 ⁰ C overnight (scheme 8). It was envisaged that the elimination of HCI would provide a good driving force for the reaction.

MC1 ₂ + 2HL ^R R

ML2 ^K + 2 HC1

[M(HL^) ₂ (Cl) ₂ ]

Scheme 8: Syntheses of the complexes from MCI ₂ ZHLf-.

This reaction had limited success; the treatment of MCI ₂ (M = Mg, Zn, Sn) with two equivalents of HL ^R affords [M(HL ^R ) ₂ (CI) ₂ ] M = Mg (4), Zn (5), and Sn (6) respectively, in excellent yield.

Two magnesium complexes were synthesised from MgCI ₂ with two equivalents of HL ^R affords [Mg(HL ^R ) ₂ (CI) ₂ ] HL ^R = 1 (4 ^tBu ), Ia (4a) the R,R-4 ^tBu analogue and 2 (4 ^ph ).

Complex 4 ^tBu was isolated in a yield of 70.1 %, the ³¹ P NMR spectrum contains two resonances (70.0 and 70.6 pm) and ¹ H NMR spectrum contains a broad singlet at 5.22 ppm (0-H). The mass spectrum results shows m/z (11.5 %) = 582.6 [4 ^tBu - HCI] and m/z (7.1 %) = 546.6 [4 ^tBu - 2 HCI]. After contact of 4 ^tBu with water a new complex (scheme 9) is formed with a molecule of water coordinated to the magnesium.

The C ₂ -symmetric chirality is confirmed by a single crystal X-ray diffraction study of 4 ^tBu .H ₂ O; Scheme 9 shows the molecular structure of the SS-diastereomer.

The ³¹ P NMR spectrum of the diastereomerically pure complex 4a contains only one resonance at 69.8 ppm and the ¹ H NMR spectrum contains a broad resonance (OH) at 5.77 ppm.

The complex 4 ^ph was isolated in a yield of 75.5 %. The resonance for the OH is significantly changed upon complexation from 5.22 ppm (4 ^tBu ) to 3.65 ppm (4 ^ph ). The mass spectrometric analysis shows m/z (8.49 %) = 663.1 [4 ^ph - HCI].

Two zinc complexes were synthesised from ZnCI ₂ with two equivalents of HL ^R affords [Zn(HL ^R ) ₂ (CI) ₂ ] HL ^R = 1 (5 ^tBu ), and 2 (5 ^ph ).

Complex 5 ^tBu was isolated in a yield of 81.9 %; the ³¹ P NMR spectrum contains one resonance at 72.6 pm, and the ¹ H NMR spectrum contains a broad singlet at 4.63 ppm (OH) in, opposition at 5.22 ppm in ¹ H NMR for 4 ^tBu . The mass spectrum shows m/z (10.5 %) = 623.0 [5 ^tBu - HCI] and m/z (7.1 %) = 587.0 [5 ^tBu - 2 HCI]. A single tablet grown which is not representative of the bulk shows scheme 10.

Scheme 10: Crystal structure of 5 ^tBu . Displacement ellipsoid drawing of 5 ^tBu 50 % probability ellipsoids. P t-butyl Me groups omitted. Average distances (A): Zn-OR:

2.059, Zn-OP: 1.953, Zn-Cl: 2.217.

From scheme 10 and the presence of HCI, it is apparent that the formation of [Zn(HL ^R ) ₂ (CI) ₂ ] is certainly favourite instead of Znl_ ^R ₂ for the zinc as the magnesium.

Complex 5 ^ph was isolated in a yield of 85.0 %; the ³¹ P NMR spectrum contains one resonance at 41.6 pm, and the ¹ H NMR spectrum contains a broad singlet at 4.95 ppm (OH). The mass spectrum shows m/z (7.3 %) = 667.4 [5 ^ph - 2 HCI].

Two tin complexes were synthesised from SnCI ₂ with two equivalents of HL ^R affords [Sn(HL^) ₂ (CI) ₂ ] HL ^R = 1 (6 ^tBu ), and 2 (6 ^ph ). In opposition of the magnesium and zinc catalysts which were air and moisture sensitive, the both tin complexes were air and moisture stable.

Complex 6 ^tBu was isolated in a yield of 80.7 %, the ³¹ P NMR spectrum contains one resonance at 76.1 pm, and the ¹ H NMR spectrum doesn't show any broad singlet for OH. In opposition with 4 ^tBu at 5.22 ppm in the ¹ H NMR spectrum. Further more the two compounds were really different, 4 ^tBu was a colourless solid while 6 ^tBu was colourless glue but the mass spectrum shows m/z (39.1 %) = 677.3 [6 ^tBu - HCI], m/z (29.8 %) = 640.3 [6 ^tBu - 2 HCI].

Complex 6 ^ph was isolated in a yield of 88.5 %; the ³¹ P NMR spectrum contains one resonance at 39.7 pm, and the ¹ H NMR spectrum contains a broad singlet at 4.61 ppm (OH). In opposition with 6 ^tBu which possessed any OH bond in ¹ H NMR. The mass spectrum shows m/z (30.3 %) = 721.0 [6 ^ph - 2 HCI].

Synthesis of catalysts from MCI ₂ /LiL ^R

To avoid the presence of chloride in the final complexes, salt elimination method was carried out. The ligand 2 was treated with n-Buϋ to afford the lithium salt 3, which was treated with ¹ Z. an equivalent of ZnCI ₂ in toluene, overnight at - 78 ⁰ C (scheme 11).

MC1 ₂ + 2 LLL ^R ► ML ₂ ^R + 2Iiα

Scheme 11: Syntheses of the complexes from M Cl ₂ / Li L ^R .

Complex 7 ^ph was isolated in a yield of 74.2 %; the ³¹ P NMR spectrum contains one resonance at 40.0 pm, and the ¹ H NMR doesn't contains a resonance OH, in opposition of 5 ^ph (4.95 ppm); the aromatic resonances were broader than in the 5 ^ph .

The mass spectrum shows m/z (100.0 %) = 610.0 [7 ^ph - 'Bu].

Synthesis of catalysts from MN" ₂ /HL ^R

To avoid the presence of chloride in the final complexes, amine elimination method was carried out. Two equivalents of ligands 1 and 2 were added to a solution of one equivalent of Ca[N(SiMe ₃ ) ₂ ] ₂ (thf) ₂ in thf, overnight at - 78 ⁰ C (scheme 12).

MN" ₂ + 2 HL ^R ► ML ₂ ^R + 2 HN"

Scheme 12: Syntheses of the complexes from MN" ₂ /HL ^R .

For the complex 8 ^tBu ; the ³¹ P NMR spectrum contains one resonance at 69.4 pm, and the ¹ H NMR spectrum doesn't contain a resonance OH, just the resonances expected. The reaction was carried out in NMR so the yield wasn't optimised but it was possible to remove the volatile compound to afford colourless solid 8 ^tBu .

Complex 8 ^ph was isolated in a yield of 37.8 %, low yield due to a problem in the purification; the ³¹ P NMR spectrum contains one resonance at 20.0 pm, and the ¹ H NMR doesn't contain a resonance OH, just the resonances expected.

Some NMR experiments were carried out with CaCI ₂ /HL ^R to compare but they didn't get any concrete results to study due to the insoluble character of CaCI ₂ .

Synthesis of catalysts from MR,/HL ^R

To avoid the presence of chloride in the final complexes, alkyl elimination method was carried out. Two equivalents of ligands 1 and 2 were added to a solution of one equivalent of ZnEt ₂ /toluene in toluene, overnight at 70 ⁰ C (scheme 13).

MR ₂ + 2 HL ^R ► ML ₂ ^R + 2 HR

Scheme 13: Syntheses of the complexes from MR ₂ /HL ^R .

The complexes 9 ^tBu and 9 ^ph were difficult to isolate and characterise, due to the low quantity of starting material (0.17 ml and 0.15 ml for ZnEt ₂ in the synthesis of 9 ^tBu and 9 ^ph , respectively). Meanwhile, the ³¹ P NMR spectrum contains a resonance at 68.8 ppm for 9 ^tBu and at 52.0 ppm for 9 ^ph . The ¹ H NMR spectrum of 9 ^ph doesn't show any resonance for OH.

In comparison, the zinc complexes synthesised via MCI ₂ /HL ^R (5) have shown in the ¹ H NMR spectrum a OH resonance for the both ligands.

The ³¹ P NMR spectrum contains a higher resonance for 9 ^ph (52.0 ppm) than for 5 ^ph (41.6 ppm) or 7 ^ph (40.0 ppm).

After all the studies in the zinc complexes, it was choosing to concentrate the research on the method which has synthesised 7 ^ph . It's allowed a product without HCI 5 ^ph and it's safer than use diethyl zinc 9 ^ph .

Synthesis of catalysts from MR _a /HL ^R

Following previous research in our group, we are targeted C ₅ - symmetric racemic complexes with main group element by the utilisation of trisalkyl aluminium (AIMe ₃ and DABAL-Me ₃ )

MR ₃ + 3 HL ^R ► ML ₃ ^R + 3 HR

Scheme 14: Syntheses of the complexes from MR ₃ /HL ^R .

Firstly, a solution of three equivalents of 1 or 2 was added to a solution of one equivalent of AIMe ₃ /hexanes in deuterated benzene, overnight at 70 ⁰ C to afford complexes 10 ^tBu and 10 ^ph respectively (scheme 14) which were difficult to isolate and characterise, due to the low quantity of starting material (0.14 ml and 0.1 ml for AIMe ₃ . Meanwhile, the ³¹ P NMR spectrum contains a resonance at 79.3 ppm for lO ^tBu and at 51.0 ppm for 10 ^ph .

Secondly, a solution of six equivalents of 1 or 2 was added to a solution of one equivalent of DABAL-Me ₃ Jn toluene, overnight at 70 ⁰ C to afford complexes ll ^tBu and ll ^ph respectively (scheme 15).

Scheme 15: Syntheses of the complexes from DABAL-Me ₃ /HL ^R .

The ³¹ P NMR spectrum contains a resonance at 78.7 ppm for ll ^tBu and at 51.0 ppm for ll ^ph which are results close to these obtain with lO ^tBu (79.3 ppm) and 10 ^ph (51.0 ppm). The ¹ H NMR spectra contain no extra proton resonance for the both complexes 11.

In the case of the tris-tert- butyl aluminium complexes (10 ^tBu and ll ^tBu ) the phosphorus resonances were the highest obtained during theses complexations.

Synthesis of catalysts from MN'WHL ^R

Treatment of Ln(N{SiMe ₃ } ₂ ) ₃ (Ln = Y) with three equivalents of 1 in thf at low temperature affords LnL ₃ ^R Ln = Y (12), in excellent yield, after recrystallization from pentane (scheme 16), complex 12 is colourless.

MN" ₃ + 3 HL ^R ► ML ₃ * + 3 HN"

Scheme 16: Syntheses of the complexes from MW ₂ J HL ^R .

Complex 12 ^tBu was isolated in a yield of 90.0 %, the ³¹ P NMR spectrum contains two resonances at 70.5 pm and 70.1 ppm, the composition was confirmed by microanalysis, and complex 12a (made by Ia R-HL ^tBu ) was isolated in a yield of 86.5 %, the ³¹ P NMR spectrum contains one resonance at 68.6 ppm.

Comparison of the ¹ H and ³¹ P-C ¹ H) NMR spectra of solutions of 12 and 12a show what appears to be predominantly the same compound, save for an additional, minor set of resonances in the spectra of 12, which correspond to a minor diastereomer, RRS-/SSR-Y ^tBu , present in about 20 % of the total yield. The C ₃ - symmetric chirality is confirmed by a single crystal X-ray diffraction study of 12.

Complex 12 ^ph was isolated in a yield of 75.1 % (yield non-optimised); the ³¹ P NMR spectrum contains three 42.8 ppm (major), 42.3 and 42.0 ppm (minor); an additional, minor set of resonances in the ¹ H NMR spectrum of 12 ^ph , which correspond to a minor diastereomer, RRS-/SSR-12Ph, present in about 30 % of the total yield

Example 3 - Syntheses of other catalyst complexes

Ligand Catalyst

Ln = Eu (;atalvst 1 ) Ln = Er (catalyst 2)

Scheme 17: syntheses of complexes catalyst 1 and catalyst 2. Preparation of (t-Bu) ₂ P(O)CH ₂ CH(t-Bu)OH , HL (Ligand)

A 1.6 M hexane solution of n-BuLi (15 ml, 25 mmol) was added dropwise to a solution of 3,3-dimethyl-epoxybutane (2.1 g, 25 mmol) and F-Bu ₂ PH (3.6 g, 25 mmol) in 20 ml of THF at -78 ⁰ C, using a 250 ml 3-neck flask equipped with reflux condenser and dropping funnel. The reaction mixture was stirred for 2 hours at room temperature and boiled for 20 min at reflux. After cooling to 0 ⁰ C the solution was slowly hydrolysed with 10 ml of 10 % aqueous NH ₄ CI and oxidized by dropwise addition of 30 ml of 30 % H ₂ O ₂ . The organic layer was separated and the aqueous solution extracted with THF (3 x 10 ml). The combined organic layer was dried over Na ₂ SO ₄ , filtered and evaporated to dryness. The obtained colourless oil was dissolved in 10 ml CHCI ₃ and chromatographed on silica gel (60, 230-400 mesh) using 90 % CHCI ₃ / 10 % MeOH as eluent. Two bands were collected. The first band was identified as starting material (epoxide). The second band was collected and

evaporated to dryness. The white precipitate obtained was recrystallised from pentane. Yield 3.2 g (50 %).

¹ H-NMR 5(C ₆ D ₆ ) : 1.1 (18 H, dd, ² J _PC = 4.5 Hz, P-C(CH ₃ ) ₃ ); 1- 15 (9 H, s, C-CH ₃ ); 1.7 - 1.9 (2H, m, CH ₂ ); 4.0 - 4.1 (IH, m, CH) [ppm] . ¹³ C-NMR 5(C ₆ D ₆ ) : 22.2 (1 C, d, J _PC = 56.8 Hz, CH ₂ ); 25.7 (3 C, CH ₃ ); 25.9 (3 C, CH ₃ ); 26.3 (3 C, CH ₃ ); 35.3 (1 C, d, J _PC = 56.8 Hz, P-CMe ₃ ); 35.5 (1 C, CMe ₃ ); 36.1 (1 C, d, J _PC = 58.1 Hz, P-CMe ₃ ); 75.7 (1 C, d, ² Jpc = 5.7 Hz, C-OH) [ppm]. ³¹ P-NMR 5(C ₆ D ₆ ) : 77.6 ppm. MP: 98 ⁰ C. Analysis Found : C 63.22 %, H 11.72; calc. C 64.1 %, H 11.9 %

Preparation of EuL ₃ (Catalyst 1)

A solution of 3 equivalents (400 mg, 1.5 mmol) of HL in 10 ml of THF was added over 10 min to a solution of one equivalent (308 mg, 0.5 mmol) of Eu[N(SiMe ₃ ) ₂ ] ₃ in 10 ml of THF at 0 ⁰ C and stirred overnight at RT (scheme 17). All volatile compounds were removed under reduced pressure and the residual yellow solid recrystallised from pentane to afford pale yellow catalyst 1. Yield 440 mg (94 %).

¹ H-NMR 5(C ₆ D ₆ ): - 7.6 (3 H, CH); - 6.1 (27 H, 'Bu); - 4.6 (3 H, CH ₂ ); - 1.4 (3 H, CH ₂ ); 0.4 (27 H, 'Bu); 9.1 (27 H, 'Bu). 31P-NMR 5(C ₆ D ₆ ) : 69.9 ppm. Analysis Found : C 53.78 %, H 9.48 %; calc. C 53.9 %, H 9.6 %

Preparation of ErL ₃ (Catalyst 2)

A solution of HL (533 mg, 0.82 mmol) in 10 ml of THF was added over 10 min to a solution of one equivalent (647 mg, 2.5 mmol) of Er[N(SiMe ₃ ) ₂ ] ₃ in 10 ml of THF at 0 ⁰ C and stirred overnight at RT (scheme 17). All volatile compounds were removed under reduced pressure and the residual solid recrystallised from pentane to afford pale pink catalyst 2. Yield 720 mg (93 %).

¹ H-NMR 5(C ₆ D ₆ ) : - 9.15 (6 x t-Bu H); 24.14 (3 x t-Bu H). No other resonances observed. Analysis Found : C 52.90 %, H 9.61 %; calc. C 53.0 %, H 9.5 %.

Structure of the ligand-precursor and the complexes

Figure 7A shows the displacement ellipsoid drawing of compound 1 50 % probability ellipsoids. All hydrogens except alcohol OH omitted for clarity. Selected distance (A) : ligand Pl-Ol 1.5065(15) and figure 7B shows the displacement ellipsoid drawing of catalyst 1 (isostructural with compound 3) 50 % probability ellipsoids. All hydrogens except P t-butyl Me groups and all hydrogens except chiral CH omitted for clarity. Selected distances (A) : catalyst 1 Eu2-O7-2.449(4), Eu2-O8-2.191(4), Eu2-P4- 3.5627(17).

Example 4 - Experimental data for the ligand, catalyst 1 and catalyst 2

Reactions 1 - 9 : solvent = DCM (dichloromethane); Reactions 10 - 11 : solvent = THF (tetrahydrofurane);

Reaction 12 : melt polymerization ^a sample purified to remove shorter chains and monomer for NMR spectroscopy

Figures 8 - 12 illustrate the M _n over time for reactions 1 - 7, M _n over conversion for reactions 1 - 7, conversion over time for reactions 1 - 7 and GPC data for polymer samples from reactions 8 and 9.

NMR Spectra of polymers

Figures 13A-C illustrate ¹ H NMR and ¹³ C NMR spectra of polymer made from D ₁ L- lactide and 1 % catalyst 2; run 10 in table 1 of polymerisation data, ESI, after 8 minutes. M _n = 270300, PDI 1.24. The polymers were purified by precipitation from a dichloromethane solution with methanol, three times. Poly (D,L-lactide) : fwhm for the methine CH resonance is 29 Hz. The integration of the iii peak in the homonuclear decoupled ¹ H NMR spectrum immediately below it corresponds to 70 % of the combined peak areas.

For comparison, figures 14A and B illustrate spectra from polymer made using L- lactide and 1 % catalyst 2; run 11 in table 1, a 10 min run, same concentrations as above with M _n = 216000, PDI 1.34].

Mass Spectral analysis of polymer

Figure 15 illustrates the electrospray mass spectrum of a relatively short chain polymer, (cone voltage = 60 V): [L(CHMeCOO) _n H] ⁺ series: 479.6, 551.6, 623.7, 695.8, 767.8, 839.8, 911.9. i.e. n = 3 to 9.

Intensity data

EuL3-PLA Dolvmer olaudeck159-1 21 (0.456) AM (Cen ,2, 80.00 , Ar,5000.0,734.47); Sm (SG, 3x5.00); Cm (16:21 )

No Mass lrrten %BPI %TIC No Mass lnten %BPI %TIC No Mass lnten %BPI %TIC

1 : 432 8 2 86θ3 8 73 1 16 27: 932 6 3 74θ2 1 14 0 15 53: 1414 4 4 08e2 1 24 0 17

2: 478 9 5 53θ2 1 69 0 22 28: 952 9 4 78θ2 1 46 0 19 54: 1472 4 4 46e2 1 36 0 18

3: 504 7 2 15θ3 6 54 0 87 29: 953 5 6 98θ2 2 13 0 28

4: 550 8 2 45θ3 7 47 1 00 30: 982 6 2 33θ4 71 09 9 47

5: 576 7 5 36θ2 1 63 0 22 31 : 983 6 4 08θ3 12 44 1 66

6: 598 7 5 62θ2 1 71 0 23 32: 984 6 3 56θ2 1 08 0 14

7: 622 8 8 66θ3 26 40 3 52 33: 1023 9 3 78θ2 1 15 0 15

8: 623 8 7 52θ2 2 29 0 31 34: 1025 5 4 16θ2 1 27 0 17

9: 669 9 3 83θ2 1 17 0 16 35: 1040 6 4 05θ2 1 23 0 16

10: 694 7 1 85θ4 56 40 7 52 36: 1054 6 1 46θ4 44 53 5 93

11 : 695 8 1 75θ3 5 32 0 71 37: 1055 6 3 08θ3 9 39 1 25

12: 737 6 4 85θ2 1 48 0 20 38: 1 112 6 6 72θ2 2 05 0 27

13: 740 5 6 07θ2 1 85 0 25 39: 1 126 5 8 12θ3 24 75 3 30

14: 766 7 2 79θ4 85 04 11 33 40: 1 127 5 1 98θ3 6 04 0 80

15: 767 7 3 14θ3 9 57 1 28 41 : 1 184 5 7 85θ2 2 39 0 32

16: 788 7 3 33θ2 1 01 0 14 42: 1 198 5 4 08θ3 12 42 1 66

17: 809 6 5 89θ2 1 79 0 24 43: 1 199 5 1 25θ3 3 81 0 51

18: 811 3 7 13θ2 2 17 0 29 44: 1256 5 9 65θ2 2 94 0 39

19: 838 7 3 28θ4 100 00 13 33 45: 1257 5 3 82θ2 1 16 0 16

20: 839 7 4 39θ3 13 39 1 78 46: 1270 4 1 80θ3 5 49 0 73

21 : 860 6 4 23θ2 1 29 0 17 47: 1271 4 6 74θ2 2 05 0 27

22: 881 6 1 09θ3 3 32 0 44 48: 1328 4 8 14θ2 2 48 0 33

23: 882 1 8 43θ2 2 57 0 34 49: 1329 4 3 53θ2 1 07 0 14

24: 910 6 3 12θ4 94 98 12 66 50: 1342 4 8 28θ2 2 52 0 34

25: 911 6 4 77θ3 14 53 1 94 51 : 1343 4 3 96θ2 1 21 0 16

26: 912 6 3 57θ2 1 09 0 14 52: 1400 4 6 95θ2 2 12 0 28

DSC data for PLA (reaction 9)

Figure 16 illustrates DSC data for PLA.

Example 5 - Polymerisation of D,L-I_actide

Scheme 18: Ring Opening Polymerisation of D,L-lactide Using catalysts synthesis from MCb/HL ¹ *

The complexes 4 - 6 have been tested as initiators for the polymerisation of D _x Z--IaCtJcIe; two series of polymerizations were conducted: A: D^-lactide + [M(HL^) ₂ (CI) ₂ ]

B: D^-lactide + [M(HL ^R ) ₂ (CI) ₂ ] and benzyl alcohol. Benzyl alcohol was selected as coinitiator because its incorporation as benzylester end group is easily detectable by both ¹ H and ¹³ C NMR spectroscopy.

The polymerisations without coinitiator were conducted in toluene at 100 ⁰ C. The results obtained for series A are summarized in Table 2. Low yields were obtained in all experiments.

Cat Cat: monomer: T Conv. ^a t initiator ratio / ⁰ C / % / h

6 ^tBu 1 : 50:0 100 11.6 24

6 -P ^κ h ⁿ 1 : 50:0 100 4.4 24

Table 2: Polymerisation of rac-lactide using 6 without alcohol.

To compare, polymerisations using 6 ^tBu with coinitiator were conducted in toluene at 100 ⁰ C; the results are shown in figure 21.

To confirm that is not the benzylalcohol polymerise the D^-lactide, the proligand was treated with benzylalcohol which was use in polymerization of rac- lactide, the ¹ H NMR spectrum show no polymerization.

Despite the fact of using a coinitiator to improve the velocity of the polymerisations, these weren't good enough. So, All reactions were carried out at 140 ⁰ C, with benzyl alcohol as coinitiator to afford a melt polymerisation which the D^-lactide is the solvent and the monomer. In the Table 3, Sn(oct) ₂ is the abbreviation for Sn(octanoate) ₂ , the most widely industry catalyst, and thus a good reference.

Cat Cat:monomer: T Conv. ^a M _n ^b M _w /M _n ^c initiator ratio / ⁰ C / % g/mol

4«su 1 : 50: 1 140 89 5300 1.61

₄ Ph 1 : 50: 1 140 95 4500 1.88

5 ^ph 1 : 50: 1 140 55 1100 2.33

5 ^tBu 1 : 50: 1 140 97 1300 1.42 gtBu 1 : 50: 1 140 69 4400 1.19

6 ^ph 1 : 50: 1 140 55 1100 2.33

Sn(OCt) ₂ 1 : 50: 1 140 24 700 1.13

Table 3: Polymerisation of rac-lactide using 4-6. a: conversion of LA monomer (([LA] ₀ -[LA]V[LA] ₀ ), calculated by ¹ H NMR; b: measured by GPC, values based on polystyrene standards and corrected by multiplication by 0.47 (Mark-Houwink law); c: polydispersity index (M _w /M _n ), PDI, measured by GPC.

The polymerisations using 5 show that at 2 % catalyst loading the polymerisation are slow, the molecular weights are low (below 2000 g.mol ^"1 ), and the PDIs fluctuate between 1.3-2. On the other hand, the kinetic data for M _n versus conversion show that the kinetics for the three complexes appears to be living.

The polymerisations using 4 show the best results so far; high molecular weight (15000-20000 g.mol ^"1 ) although the polydispersities are not narrow around 1.6-1.8. Also, the kinetic traces show a living nature with a linear Mn versus conversion and PDI decreases with an increasing conversion.

The polymerisation using 6 are difficult to analyse and inconsistent; generally the polymerisation rates were slow and the molecular weights low. The polymerisations using Sn(OCt) ₂ are very slow in comparison, furthermore they are not living.

The GPC chromatogram of Figure 22 shows that the polymerisations with 4 ^tBu and 4 ^ph have the highest molecular weight, and 6 ^tBu has the narrow PDI. On the other hand 6 ^ph and Sn(OCt) ₂ have low molecular weight and high PDI.

The aim of this project is to polymerise a mixture of two stereocomplex PLA, poly-D-lactide and poly-Z_-lactide. Two separate control experiments were performed to confirm the tacticity, so it was decided that 4 ^tBu will be use to extend the studies

In the first control experiment, the ¹ H NMR spectra of the stereocomplex product should look like that of poly-Z_-lactide, with a single CHMe resonance (if the chains are infinitely long). If the polymerisation is less selective or transterification becomes a competing reaction at higher conversions, the original stereochemical control will be lost and the proton-decoupled spectra will show the different CH environments. L-lactide was polymerised using 4 ^tBu (Figure 23a), D^-lactide was polymerised using 4 ^tBu (Figure 23b) and was compared to the rac-PLA polymerised with Sn(OCt) ₂ (Figure 23c).

The shape of the NMR spectra samples of rac-lactide polymerised by 4 ^tBu at 89 % monomer conversion, (23b) are comparable to (23c) with the Hi resonance corresponding to 35 % of the combined peak areas, indicating a poor stereoselectivity of the polymerisation. It contains major additional resonances corresponding to unselective insertions.

In the second experiment, the ¹³ C NMR spectra of the stereocomplex product should look like that of poly-Z_-lactide, with a single CHMe resonance (if the chains are infinitely long). If there have been transferication reactions, or unselective insertions, the control will be lost and the NMR spectra will contain resonances for the different CH environments.

Spectra samples of rac-lactide polymerised by 4 ^tBu at 89 % monomer conversion, (24b) are a shape different to (24a) confirming a poor stereoselectivity of the polymerisation. It contains major additional resonances corresponding to unselective insertions.

The GPC data and ¹ H NMR spectra show a linear variation between M _n and conversion and between Ln (l/(l-conv.)) and the time of polymerisation that indicates a controlled, living polymerization (figure 25A and B); also the PDI is below 1.4. Furthermore, chains extension is possible by reactivation of the end groups. The theoretical molecular weights have been calculated using the formula:

^{Mn theo =} ^ ^{X l44' 13 X COnV} -

Any polymerisations were tried using a catalyst synthesise from MCI ₂ /LiL ^R

Using catalysts synthesis from MIM"n/HL ^R

The complex 8 ^ph was examined for polymerisation activity with rac-lactide (M/I = 50). The polymerisations were carried out in bulk at 140 ⁰ C with coinitiator. From the polymerisation data, it is apparent than the calcium complex shows at full conversion (> 95 %) a narrow distribution (1.2-1.3) but a low molecular weights (around 1000-2000 g.mol ^"1 ). Some studies are carrying out with 8 ^tBu .

The conversion versus the time of polymerisation using 8 ^ph is shown in figure 26.

Previously in our group (Robert Blaudeck), the complex 12 ^tBu has been tested as an initiator for the polymerization of rac-lactide (M/I = 100); even at - 18 ⁰ C in DCM, the polymerisation is rapid, and appears to be living in nature. The polymer weights are high (22 100 g.mol ^"1 at 35 % of conversion and 68 600 g.mol ^"1 at 99 %), and the polydispersities (PDI) of the polymers are narrow (1.3-1.5). Approximately half of the monomer is consumed after three minutes, during which the solution becomes extremely viscous. He also proved that with increasing M/I he obtained a decrease in the PDI (around 1.2).

Using catalysts synthesis from MRn/HL ^R

The complexes synthesis from ZnEt ₂ and AIMe ₃ yielded with so much compound that it was impossible to use the complexes 9 and 10 in polymerisation, only the complex 11 was used.

The complex ll ^tBu was examined for polymerisation activity with rac-lactide (M/I = 50). The polymerisations were carried out in toluene at 100 ⁰ C with coinitiator. From the polymerisation data, it is apparent than the aluminium complex shows a conversion > 90 %, a large distribution (1.7-1.9), and a low molecular weights (around 1000-2000 g.mol ^"1 ). The conversion versus the time of polymerisation using ll ^tBu is shown in figure 27.

Example 6 - Copolymerisation of L-lactide/glycolide

Scheme 19: Copolymerisation of glycolide and L-lactide.

All the copolymerisation between L-lactide and glycolide were carried out only with complexes 4-6: [M(HL^) ₂ (CI) ₂ ]

Kinetic Study

To understand the kinetic of copolymerisation between L-lactide and glycolide, different factors were changed, the metal (Mg, Zn, Sn); the ligand (tert- butyl, phenyl or octanoate); the time of polymerisation (from 10 seconds to 96 h); the feed composition (from 100 % of L-lactide to 100 % of glycolide); and the temperature (140 ⁰ C, 160 ⁰ C or 180 ).

A kinetic study was carried out to find the best combination of factors. The initial conditions were: 5 ^ph , 140 ⁰ C, 96 h, [Lac]/[Gly] = 4, [Lac]/[Cat] = 50.

Firstly, after 1.5 h reaction time the reaction is 64 % complete and after 24 h it is 85%. Secondly, the feed composition gives the best results for a ratio 60/40 (L- lactide/glycolide). Thirdly, the conversion rate increases with increasing temperature. And the rate is also dependent on the ligand following the order ¹¹ Bu > Ph > octanoate. Finally, the metal affects the rate following the order Mg > Zn > Sn.

The best combination found was: 4 ^tBu , 140 ⁰ C, 24 h, [Lac]/[Gly] = 1.5 [Lac]/[Cat] = 50, this combination was used in microstructure studies.

Copolymer microstructure

At the beginning of the polymerisation, the ¹ H NMR spectra of the copolymer product should show just -GGGGG- pentads because the glycolide, is polymerised faster than the L-lactide; with increasing time, some -LLGGL- pentads should emerge. If the copolymerisation is less selective, no stereochemical control will be observed and the microstructure will show a different tacticity.

By applying the probability theory to the estimation of copolymer sequence distribution we expected completely random copolymer with a « blocking » tendency (χ < 1). This is confirmed by the results of ¹ H NMR spectra which have shown a block tendency after 6 h (-GGGGG-) and some atactic pentades after 24 h (-LLGGL- + - LGGLL-), also confirmed by the presence of atactic tetrads after 24 h (-GGLL-) in the ¹³ C NMR (Figure 28).

GPC characterisations

The GPC data (figure 29) show a linear variation between M _n and conversion but not through 0 and that indicates a controlled, living polymerisation; also the PDI is below 1.6 that is good for a copolymerisation. The ¹ H NMR spectra show as predicted by theory, a polymerisation faster for the glycolide than for the L-lactide. The theoretical molecular weights have been calculated using the formula:

Mitteo = ( ^X _L - _L ac X M^ _L ae +X _fly X M^) X COnv. X —

The kinetic results are shown as a stacked plot on GPC chromatograms to demonstrate the dependence of the molecular weights with the conversion (Figure 30).

The GPC chromatograms confirm the results from figure 29; with increasing time of the polymerisation there is an increase in the molecular weights. Meanwhile, the PDI increases with an increase in the time of the polymerisation.

Example 7 - Polymerisation of lactide

entry Cat: monomer: T / ⁰ C Time /min Conv. ^b solvent ratio ^a / %

1 1 : 100 : 10000 - 10 0.33-1 >99

2 1 : 100 : 10000 - 10 3 >99

3 1 : 100 : 10000 - 10 6 >99

4 1 : 100 : 10000 - 10 10 >99 a. solvent = dichloromethane; b. conversion of LA monomer (([LA] ₀ -[I-A]V[I-A] _O ). c. measured by GPC, values based on polystyrene standards, weight corrected by multiplication by 0.47 [Mark-Houwink equation] d. polydispersity index (Mw/Mn), PDI, measured by GPC.

The polymers were characterized by NMR spectroscopy. The results are shown in figures 31a-e. a) methine region of the homonuclear decoupled ¹ H-NMR for entry 1. Integration of the iii peak corresponds to 26.2 %.

¹ H-NMR a(CDCI ₃ ): 5.146, 5.161, 5.171, 5.178, 5.181, 5.185, 5.202 [ppm]. b) methine region of the homonuclear decoupled ¹ H-NMR for entry 2. Integration of the iii peak corresponds to 88.8 %.

¹ H-NMR a(CDCI ₃ ): 5.103, 5.181, 5.200 [ppm]. c) methine region of the homonuclear decoupled ¹ H-NMR for entry 3. Integration of the iii peak corresponds to 78.7 %.

¹ H-NMR a(CDCI ₃ ): 5.144, 5.160, 5.178, 5.198, 5.211, [ppm]. d) methine region of the homonuclear decoupled ¹ H-NMR for entry 4. Integration of the iii peak corresponds >99 %.

¹ H-NMR a(CDCI ₃ ): 5.151 ppm.

Table 5. polymerization of rac-lactide by ZnL ₂ Ph

entry Cat: monomer: Conv. solvent ratio ^a / %

1 : 50 : 0 140 72 >99 a. solvent = dichloromethane; b. conversion of LA monomer (([LA] ₀ -[L-A]V[I-A] ₀ ). c. measured by GPC, values based on polystyrene standards, weight corrected by multiplication by 0.47 [Mark-Houwink equation] d. polydispersity index (Mw/Mn), PDI, measured by GPC.

e) methine region of the homonuclear decoupled ¹ H-NMR for entry 5. Integration of the iii peak corresponds to 36.1 %.

¹ H-NMR a(CDCI ₃ ): 5.097, 5.126, 5.139, 5.165, 5.179, [ppm].

Example 8 - Polymerisation of ε-caprolactone

Polymerization procedures

Solution :

A teflon valve-sealed ampoule was charged with 500 mg of the monomer which was dissolved in the volume of thf required to give the ratio in the table entry, and the solution stirred at the temperature given in the table. To this was added via cannula a solution of appropriate mass of catalyst (one of 1 to 4) in 2mls of thf (see table 6).

Melt:

The catalyst (one of 1 to 4) was ground using a pestle and mortar to a fine powder, which was mixed with the powdered monomer in a flask in the quantities 500 mg ε-caprolactone and the appropriate mass of catalyst (see table 6).

The mixture was heated in an ampoule in a sand bath to 180 centigrade. The powder melted into a viscous solution which solidified as it cooled down to RT. Yield 99 % (apparent complete conversion).

Cat: Monomer: Solvent Temp Time Mw Mn PDI Mp / ⁰ C /mins

_____________ _____ ______ _____ ε- TIO ^~ Y4900 ^{~ "} Tnoδδ ^{"" ~} ϊ20 caprolactone 0

1 : 100: 5,000 / ε- 20 240 38100 230000 1.65 200 caprolactone 0

1 : 50:5,000 / ε- 20 10 14900 112000 1.32 - caprolactone 0

1 : 50:0* / ε- 20 10 28400 155000 1.83 180 caprolactone 0

*melt: powdered catalyst dissolved in monomer - solidified after 10 mins

2

1 = Er 4 ( 1)

Example 9 - Preparation of poly ε-caprolactam

A vigorously stirred solution of 0.5 g (4.4 mmol) ε-caprolactam in 50 ml thf was treated with an solution of 5 mg ErL ₃ in 1 ml thf at room temperature. After 30 min the reaction mixture was quenched with 5 drops of MeOH. Removing the solvent yielded white amorphous polymer. M _n = 101000 g/mol, PDI = 1.4