O‐to‐S Substitution Enables Dovetailing Conflicting Cyclizability, Polymerizability, and Recyclability: Dithiolactone vs. Dilactone

Abstract

Developing chemically recyclable polymers represents a greener alternative to landfill and incineration and offers a closed‐loop strategy toward a circular materials economy. However, the synthesis of chemically recyclable polymers is still plagued with certain fundamental limitations, including trade‐offs between the monomer's cyclizability and polymerizability, as well as between polymer's depolymerizability and properties. Here we describe the subtle O‐to‐S substitution, dithiolactone monomers derived from abundant feedstock α‐amino acids can demonstrate appealing chemical properties different from those of dilactone, including accelerated ring closure, augmented kinetics polymerizability, high depolymerizability and selectivity, and thus constitute a unique class of polythioester materials exhibiting controlled molecular weight (up to 100.5 kDa), atactic yet high crystallinity, structurally diversity, and chemical recyclability. These polythioesters well addresses the formidable challenges of developing chemically recyclable polymers by having an unusual set of desired properties, including easy‐to‐make monomer from ubiquitous feedstock, and high polymerizability, crystallinity and precise tunability of physicochemical performance, as well as high depolymerizability and selectivity. Computational studies explain why O‐to‐S modification of polymer backbone enables dovetailing desirable, but conflicting, performance into one polymer structure.

Toward Recyclable Polymers: Ring-Opening Polymerization Enthalpy from First-Principles

“…While this picture may look straightforward and simple, reaching a predictive level in computing Δ H ROP for realistic situations is nontrivial. The impression of simplicity may come from the only procedure that has been adopted to-date to model ROP in the scientific literature. − In this approach, cyclic monomers are opened, and the two dangling bonds are saturated by two suitable end groups derived from neutral molecular initiators, for example, methanol CH 3 –OH or methanethiol CH 3 –SH, so that the reactions are isodesmic, that is, the numbers and types of bonds are conserved. Then the ground-state atomic configurations of the monomer, the “polymer”, that is, the ring-opened capped monomer, and the initiator molecule, are guessed and optimized using first-principles computations.…”

“…These are very small energy values, considering that 1 kcal/mol, or ≃5 kJ/mol, is designated as “chemical accuracy”. Quantitatively predicting such a tiny energy value across diverse chemical spaces is extremely difficult, and it remains a longstanding challenge for computations. − In the context of Δ H ROP computations, ab initio methods like DFT , and beyond must be used. − As these methods are computationally expensive, the atomic models adopted thus far are typically small and simple. As mentioned above, cyclic monomers are opened, and the two dangling bonds are saturated by two suitable end groups of a neutral molecular initiator, for example, methanol CH 3 –OH or methanethiol CH 3 –SH.…”

“…This approach ignores the polymer energy spectrum, which is essentially continuous, representing numerous atomic configurations that are very close to the ground state in energy. 35 Such strategies have been used mostly for understanding the atomic-level geometric details of the involved species, 23,24,26 while computed values of ΔH ROP using this approach can, at best, be used to understand trends within a restricted chemical class. 22,25 The present scheme, portrayed in Figure 3, was designed to compute ΔH ROP in a more realistic manner.…”