Poly­(lactide-co-glycolide) (PLGA) is an important component in many pharmaceutical applications. A detailed understanding of repeat unit ordering in PLGA is complicated by a prohibitively large number of equations that would be required for a conventional deterministic mathematical description of the reversible lactide/glycolide copolymerization. Established copolymerization models apply only to irreversible polymerization chemistry. Here, we proposed a numerical copolymerization model that includes forward reactivity ratios (r G, r L) and a third parameter to account for lactide reversibility (r –L/L). We combined a simplified deterministic model and full stochastic model to extract reactivity ratios from experimental time-evolution data taken at several initial feed compositions. We fit the kinetic data at four initial glycolide compositions (f G 0 = 0.10, 0.15, 0.20, and 0.25) to the simplified deterministic model to determine the full set of reactivity ratios: r G = k GG/k GL = 4.5 ± 0.3, r L = k LL/k LG = 0.32 ± 0.01, and the reversibility parameter for lactide: r –L/L = k L–L/k LL = 0.15 ± 0.02 mol/L. Glycolide was consumed early in the copolymerization and did not exhibit reversibility. The stochastic model was used to verify that the reactivity ratios provided an accurate description of the partially reversible copolymerization by comparison of the fully parameterized model against experimental data.

Small-molecule battery electrolytes are composed of mixtures of high-polarity and low-viscosity solvents at compositions that optimize ionic conductivity. In this work, we examined analogous polymer blends composed of one component with rapid segmental dynamics to provide low viscosity and another with a high dielectric constant (ε) to enhance ion dissociation. We investigated the inherent tradeoff between polymer polarity and segmental dynamics limiting ionic conductivity through the analysis of ionic conductivity of electrolytes containing lithium bis(trifluoromethanesulfonyl)imide (LiTFSI) with different polarity hosts: poly(allyl glycidyl ether) (PAGE) (ε ≈ 9), poly[(cyanoethyl glycidyl ether)-co-(n-butyl glycidyl ether)] [P(CEGE-co-nBGE)] (ε ≈ 24), and poly(cyanoethyl glycidyl ether) (PCEGE) (ε ≈ 36). Two high-polarity-contrast polymer blends, PAGE/P(CEGE-co-nBGE)/LiTFSI and PAGE/PCEGE/LiTFSI, were prepared. While PAGE/PCEGE/LiTFSI blends were immiscible at all compositions, PAGE/P(CEGE-co-nBGE)/LiTFSI blends were miscible at LiTFSI concentrations above r = 0.065. The immiscibility of PAGE/PCEGE/LiTFSI blends imposed a negative deviation in ionic conductivity from a calculated linear average of the two single-polymer electrolytes. This negative deviation was decreased in magnitude to less than 10% in miscible PAGE/P(CEGE-co-nBGE)/LiTFSI blends between 30 and 90 °C. To understand the changes in the effective interaction parameter in the presence of LiTFSI, we investigated the disordered-state small-angle X-ray scattering (SAXS) of a diblock polymer, PAGE-b-PCEGE, across a range of LiTFSI concentrations. By fitting SAXS profiles of this copolymer using the random phase approximation and an adjustable contrast model, we found that the effective interaction parameter decreased monotonically as the LiTFSI concentration increased. At low concentrations, LiTFSI was primarily solvated in the high-polarity PCEGE-rich phase.