A Highly Active and Selective Zirconium-Based Catalyst System for the Industrial Production of Poly(lactic acid)

Abstract

The biodegradable, aliphatic polyester poly(lactic acid), PLA, is a leading bio-based alternative to petrochemical-derived plastic materials across a range of applications. Widely reported in the available literature as a benchmark for PLA production via the bulk ring-opening polymerization of lactides is the use of divalent tin catalysts, and particularly tin(II) bis(2-ethylhexanoate). We present an alternative zirconium-based system that combines an inexpensive Group IV metal with the robustness, high activity, control, and designed compatibility with existing facilities and processes, that are required for industrial use. We have carried out a comprehensive kinetic study and applied a combined experimental and theoretical approach to understanding the mechanism by which the polymerization of lactide proceeds in the presence of this system. In the laboratory-scale (20 g) polymerization of recrystallized racemic d,l-lactide (rac-lactide), we have measured catalyst turnover frequencies up to at least 56,000 h^-1, and confirmed the reported protocols' resistance toward undesirable epimerization, transesterification, and chain scission processes, deleterious to the properties of the polymer product. Further optimization and scale-up under industrial conditions have confirmed the relevance of the catalytic protocol to the commercial production of melt-polymerized PLA. We were able to undertake the efficient preparation of high-molecular-weight PLA on the 500-2000 g scale, via the selective and well-controlled polymerization of commercial polymer-grade l-lactide under challenging, industrially relevant conditions, and at metal concentrations as low as 8-12 ppm Zr by weight ([Zr] = 1.3 × 10^-3 to 1.9 × 10^-3 mol %). Under those conditions, a catalyst turnover number of at least 60,000 was attained, and the activity of the catalyst was comparable to that of tin(II) bis(2-ethylhexanoate).

Scheme 1. General Mechanistic Pathways for the Activated Monomer- (Lewis Acid-Catalyzed) and Coordination–Insertion-Type ROP of Lactide in the Presence of a Metal-Based Catalyst, L_nM or L_nMOR, Respectively⁻

Figure 1

Zwitterionic zirconium amine tris(phenolate) complex Zr(HL^Me)₂, 1.

Scheme 2. Preparation of Catalyst Formulation f1 to Generate the Oligoester Solvent Phase and, Inset, Catalytic Application to the Immortal ROP of rac-LA

Figure 2

Semilogarithmic initial rate plots for determination of the effects of monomer stereochemistry and catalyst delivery method on the rate of the ROP of LA at [Zr] = 2.5 × 10^–3 mol % ([LA]:[Zr]:[ROH] = 40,000:1:100; 16 ppm Zr by weight). Labels refer to entry numbers in Table 1. R² values for IR-1–IR-5 = 0.98, 0.96, 0.99, 0.99, and 0.89, respectively.

Figure 3

Methine region of the ¹H NMR spectrum of the crude product from the ROP of l-LA where [Zr] = 2.60 × 10^–2 mol % ([LA]:[Zr]:[BnOH] = 3850:1:5; 160 ppm Zr by weight) and [BnOH] = 0.13 mol %, containing signals corresponding to isotactic PlLA and residual monomer species (top), and the methine signal from the homonuclear decoupled ¹H{¹H} NMR spectrum of the PlLA product of the same reaction, where all linkages are represented by the [iii] tetrad (bottom). 5 g of l-LA, solvent-free, 180 °C, 2 h, solid 1 and BnOH were used.

Figure 4

Semilogarithmic initial rate plots for comparison of f1, solid 1, and Sn(Oct)₂ at several metal loadings. Labels refer to entry numbers in Table 3. R² values for IR-6–IR-13 = 0.99, 0.99, 0.98, 0.98, 0.97, 0.97, 0.99, and 0.98, respectively.

Figure 5

VTNA plots for the ROP of rac-LA in the presence of various loadings of f1. Labels refer to entry numbers in Table 4.

Figure 6

A plot of k_obsversus [f1], where [f1] = [Zr], for determination of the propagation rate constant, k_p, for the ROP of rac-LA initiated by f1. Values of k_obs were obtained from semilogarithmic initial rate plots, constructed for determination of order in catalyst, 1. Error bars correspond to the root mean squared error.

Figure 7

Eyring plot for the determination of ΔG^⧧ for the solvent-free ROP of rac-LA at 174 °C in the presence of 1.3 × 10^–2 mol % Zr ([LA]:[Zr]:[ROH] = 7700:1:100; 15.6 ppm Zr by weight), dosed as f1. Experimental details in Table 5. Error bars correspond to the root mean squared error.

Figure 8

Semilogarithmic initial rate plots for the ROP of rac-LA in the presence of f1, f2, and f3, at equimolar concentrations of Zr and alcohol, respectively. Labels refer to entry numbers in Table 6. R² values for IR-6 and IR-21–IR-24 = 0.99, 0.99, >0.99, 0.99, and 0.98, respectively.

Figure 9

Free enthalpy profile of the favored mechanism for the initiation step of the ROP of l-LA in the presence of 1 and BnOH, calculated using the PBE0-D3 protocol. Propagation is anticipated to follow an analogous pathway.