About

Agenda

Seminar Information

Seminar 1 - Upcycling Virgin and Waste Thermoplastic Polyethylene, Polypropylene, and Related Copolymers into Covalent Adaptable Networks that Recover Cross-link Density upon Recycling, John Torkelson

Nine hundred billion pounds of synthetic polymers/plastics are produced annually worldwide. Unfortunately, less than 10% of spent polymers and plastics are effectively recycled. Polyethylene (PE) and polypropylene (PP) account for more than half of global polymer production; therefore, approaches to address this recycling crisis should consider PE, PP, and their copolymers. We will describe how thermoplastic PE, PP, and associated copolymers, constituted of linear or branched chains, can be upcycled into covalent adaptable networks (CANs) using a one-step, radical-based reactive processing method that is a simple “drop-in” modification of the commercial process used to make non-recyclable, permanently cross-linked PE (XLPE) networks or thermosets from thermoplastic PE. The commercial process to make XLPE thermosets, which have enhanced properties relative to thermoplastic PE and are produced at levels equivalent to thermoplastic polystyrene or poly(vinyl chloride), two of the six most widely produced polymers, involves melt processing PE with a low level of radical initiator, resulting in the transfer of a hydrogen atom from PE to the initiated radical. This transfer leaves a radical on the PE backbone which can react with another PE backbone radical, forming a permanent cross-link. We have developed several dynamic covalent cross-linkers with two carbon-carbon double bonds that can be “dropped into” the reactive process at several weight percent relative to PE or ethylene-based copolymer, leaving other conditions unchanged and resulting in CANs with cross-links that are robust at use conditions but dynamic at melt reprocessing conditions, leading to recyclability. Our dynamic covalent cross-linkers yield PE CANs that quantitatively recover the original cross-link density and associated properties after multiple recycling steps and lead to massive reductions in elevated-temperature creep relative to thermoplastic PE. We will also show that fully recyclable PE CANs can be produced via direct free-radical copolymerization of ethylene with low levels of a dynamic covalent cross-linker. Our approach has also enabled us to overcome the challenge of forming networks via reactive melt-state processing of PP and propylene-based copolymers. No commercial cross-linked PP thermoset is produced by simple radical-based reactive processing due to accompanying chain scission. We overcame this problem by developing approaches that stabilize the radicals via resonance, yielding recyclable PP CANs that recover cross-link density upon recycling. Finally, we will discuss studies supporting the utility of propylene-ethylene copolymer CANs and CANs made from PE/PP blends as blend compatibilizers, including for sorted recyclate contaminated with other polyolefins.

Seminar 2 - Sustainable, Recyclable Alternatives to Polyurethane Networks: Biobased and Biowaste-based Non-isocyanate Polyurethane (NIPU) and Non-isocyanate Polythiourethane (NIPTU) Covalent Adaptable Networks (CANs) as Bulk Materials and Foams, John Torkelson

Polyurethane (PU) ranks among the top six polymers in annual global production, with the vast majority being made as cross-linked networks, often as foams. There are several sustainability concerns with PUs. Among these, PU is made by reacting toxic diisocyanates with polyols, with diisocyanate derived from even more toxic phosgene. Furthermore, PU networks are not easily recyclable or reprocessable. The Torkelson group has aimed to overcome PU's sustainability challenges by developing biobased and biowaste-based, circularly recyclable and reprocessable non-isocyanate polyurethane (NIPU) materials of two types: polyhydroxyurethane (PHU) and non-isocyanate polythiourethane (NIPTU). PHUs can be synthesized by reacting cyclic carbonates with amines while NIPTUs are made by reacting cyclic dithiocarbonates with amines. We employed a rheological strategy to develop the rapid synthesis of self-blowing PHU network foams, reducing the reaction time from hours to minutes. We then demonstrated that biowaste-based, self-blowing PHU network foams could be made from renewable starting materials, e.g., cashew nutshell liquid and rice husks, and reprocessed multiple times into bulk materials with full recovery of cross-link density. We also addressed the intrinsic limitations of PHU, including its high hydrophilicity, by developing NIPTU cross-linked networks. The NIPTU networks have dynamic covalent cross-links of two types: thionourethane and disulfide, the latter achieved by auto-oxidation of NIPTU pendant thiol groups. The NIPTU networks exhibit outstanding repocessability, with full recovery of cross-link density. We compared structurally analogous biowaste-based PHU and NIPTU networks, demonstrating that NIPTU networks exhibit more rapid synthesis, higher cross-link density, enhanced tensile properties, and better water resistance. We also developed the first NIPTU foams. We leveraged the rapid chain growth to form NIPTU linear backbones and the slower thiol auto-oxidation to form inter-chain disulfide cross-links. We then balanced the gelling reaction with the foaming process, resulting in relatively homogeneous NIPTU foams. By exploiting the disulfide dynamic chemistry, we developed NIPTU foams that can be reprocessed into bulk NIPTU by both compression molding and melt extrusion. We also demonstrated the first proof-of-principle NIPU foam-to-foam recycling via melt extrusion of a NIPTU foam, with bicarbonate salts added as a blowing agent. Finally, by exploiting our discovery and understanding of the NIPTU dynamic chemistry, trans(thio)carbamoylation, we successfully depolymerized a biowaste-derived NIPTU network via methanolysis, yielding high-purity small molecules at greater than 90% recovery.

Seminar 3 - Dissociative vs. Associative Dynamics in Covalent Adaptable Networks (CANs): Fundamental Differences in What Controls the Relaxation Times and Activation Energies as Well as the Reprocessability and Self-Healing Natures of CANs, John Torkelson

Covalent adaptable networks (CANs) have the properties and performance of thermosets at use conditions but, unlike thermosets, allow for melt-state reprocessing at elevated temperature (T) with complete recovery of cross-link density and related properties after reprocessing. The reprocessability of CANs is due to the dynamic nature of covalent bonds within the cross-links, which may be dissociative in nature, associative in nature, or, in select cases, a combination of dissociative and associative. At use conditions, the dynamic nature of such bonds is dormant, but at elevated-T, it is sufficient in many cases to allow reprocessing and recovery of cross-link density upon cooling to use conditions. For some years, it has been recognized that the apparent activation energies (representing the T-dependence) of CAN stress relaxation and elevated-T viscous creep response, which are assumed to be dynamic underpinning the reprocessability and self-healing natures of CANs, vary significantly across different polymeric CANs. Several years ago, Filip du Prez and co-workers concluded that “no straightforward prediction of relaxation times or activation energies (of CANs) is possible at this stage (Polym. Chem.  2020, 11, 5377).” The Torkelson group has undertaken dynamic mechanical analysis and rheological studies on an array of CANs with dissociative or strictly associative dynamic bonds to address this issue. In general, if the CAN cross-linker dynamics are dissociative, the cross-linker bond dissociation energy determines the activation energy (and the T-dependence) of CAN stress relaxation and viscous creep. If the CAN cross-linker dynamics are associative and the cross-linker is directly attached to the CAN backbone, the activation energy of the polymer backbone alpha-relaxation (cooperative segmental mobility) defines the activation energy of stress relaxation and viscous creep. There are other important fundamental differences between associative and dissociative CANs that affect their utility. If a CAN has exclusively or predominantly dissociative dynamics, its dynamic nature will allow reprocessing at relatively low cross-link densities but not at high ones. If a CAN has exclusively or predominantly associative dynamics, its dynamic nature will allow reprocessing at relatively high cross-link densities but not at low ones. The seemingly surprising outcome for associative CANs is due to the associative dynamics becoming increasingly more rapid with increasing cross-link density (the associative reaction is second-order), thereby accommodating both elevated-T reprocessing and self-healing character.

Seminar 4 - Fragility is a Key Controlling Factor in the T_g-Confinement Effects of Polymers: How Do We Know This and How May We Exploit This Knowledge to Eliminate the Effect?, John Torkelson

The glass transition temperature (T_g)-confinement effect was first described in the research literature more than 30 years ago. In 1994, Keddie et al. (Europhys. Lett. 1994, 27, 59) reported that, relative to bulk response, polystyrene (PS) films supported on Si substrates, with no attractive polymer-substrate interfacial interaction, exhibited a nearly 30 °C reduction in 10-nm-thick films. Since then, such trends in T_g reduction in nanoconfined PS films have been approximately replicated by many other groups, including the Torkelson group which has used a combination of fluorescence and ellipsometric techniques to characterize and investigate such confinement effects, including via novel multi-layer film studies. We will review some of those ground-breaking studies in this seminar. Because thin and ultrathin films have great commercial relevance, e.g., in photolithography, drug delivery, sensors, and nanoimprinting applications, there have been intense research efforts over the past 25 years to understand, control, and even eliminate such effects. In PS films, it is now well known that the T_g-confinement effect is primarily associated with the polymer-air interface, i.e., the free surface. We have conducted extensive research demonstrating that the physics underlying the control and elimination of the effect is closely related to the reduction of bulk fragility or the suppression of the fragility-confinement effect. High-fragility glass-forming polymers require more neighboring repeat units to participate in a structural relaxation and thus exhibit a greater perturbation to T_g caused by a free surface, leading to greater T_g-confinement effects. Strategies used by the Torkelson group to suppress the T_g-confinement effect in PS include incorporating surfactants or small-molecule plasticizers or modifying PS architecture from linear chains into more unusual forms, e.g., star-shaped, cyclic, and bottlebrush polymers as well as dense polymer brushes. Very recently, we demonstrated that both the T_g-confinement effect and the fragility-confinement effect in PS and related films can be eliminated by incorporating several mol% of branched comonomers, such as 2-ethylhexyl acrylate, via simple free-radical copolymerization with styrene. Such simple, inexpensive approaches for eliminating the T_g-confinement effect may be particularly amenable to wide-scale commercial application.