Search for: All records

Understanding interparticle interactions has been one of the most important topics of research in the field of micro/nanoscale materials. Many significant characteristics of such materials directly stem from the way their building blocks interact with each other. In this work, we investigate the efficacy of a specific category of Machine Learning (ML) methods known as interaction networks in predicting interparticle interactions within colloidal systems. We introduce and study Local Neighborhood Graph Neural Networks (LN-GNNs), defined according to the local environment of colloidal particles derived from particle trajectory data. The LN-GNN framework is trained for unique categories of particle neighborhood environments in order to predict interparticle interactions. We compare the performance of the LN-GNN to a baseline interaction network with a simpler architecture and to an Instance-Based ML algorithm, which is computationally more expensive. We find that the prediction performance of LN-GNN measured as an average normalized mean absolute error outperforms the baseline interaction network by a factor of 2–10 for different local neighborhood configurations. Furthermore, LN-GNN’s performance turns out to be very comparable to the instance-based ML framework while being an order of magnitude less expensive in terms of the required computation time. The results of this work can provide the foundations for establishing accurate models of colloidal particle interactions that are derived from real particle trajectory data.

 

Dynamic covalent Diels–Alder chemistry was combined with multiwalled carbon nanotube (CNT) reinforcement to develop strong, tough and conductive dynamic materials. Unlike other approaches to functionalizing CNTs, this approach uses Diels–Alder bonds between diene pendant groups on the polymer and the CNT surface πσ bonds acting as dienophiles. Experimental and simulation data align with the CNT reinforcement coming from dynamic covalent bonds between the matrix and the CNT surface. The addition of just 0.9 wt% CNTs can lead to an almost 3-fold increase in strength and 6–7 order of magnitude increases in electrical conductivity, and materials with 0.45 wt% CNTs show excellent strength, self-healing and conductivity. 

Dynamically crosslinked polymers and their composites have tremendous potential in the development of the next round of advanced materials for aerospace, sensing, and tribological applications. These materials have self-healing properties, or the ability to recover from scratches and cuts. Applied forces can have a significant impact on the mechanical properties of non-dynamic systems. However, the impacts of forces on the self-healing ability of dynamically bonded systems are still poorly understood. Here, we used a combined computational and experimental approach to study the impact of mechanical forces on the self-healing of a model dynamic covalent crosslinked polymer system. Surprisingly, the mechanical history of the materials has a distinct impact on the observed recovery of the mechanical properties after the material is damaged. Higher compressive forces and sustained forces lead to greater self-healing, indicating that mechanical forces can promote dynamic chemistry. The atomistic details provided in molecular dynamics simulations are used to understand the mechanism with both non-covalent and dynamic covalent linkage responses to the external loading. Finite element analysis is performed to bridge the gap between experiments and simulations and to further explore the underlying mechanisms. The self-healing behavior of the crosslinked polymers is explained using reaction rate theory, with the applied force proposed to lower the energy barrier to bond exchange. Overall, our study provides fundamental understanding of how and why the self-healing of cross-linked polymers is affected by a compressive force and the force application time. 

Recent progress on stretchable, tough dual-dynamic polymer single networks (SN) and interpenetrated networks (IPN) has broadened the potential applications of dynamic polymers. However, the impact of macromolecular structure on the material mechanics remains poorly understood. Here, rapidly exchanging hydrogen bonds and thermoresponsive Diels–Alder bonds were included into molecularly engineered interpenetrated network materials. RAFT polymerization was used to make well-defined polymers with control over macromolecular architecture. The IPN materials were assessed by gel permeation chromatography, differential scanning calorimetry, tensile testing and rheology. The mechanical properties of these IPN materials can be tuned by varying the crosslinker content and chain length. All materials are elastic and have dynamic behavior at both ambient temperature and elevated temperature (90 °C), owing to the presence of the dual dynamic noncovalent and covalent bonds. 100% self-healing recovery was achieved and a maximum stress level of up to 6 MPa was obtained. The data suggested the material's healing properties are inversely proportional to the content of the crosslinker or the degree of polymerization at both room and elevated temperature. The thermoresponsive crosslinker restricted deformation to some extent in an ambient environment but gave excellent malleability upon heating. The underlying mechanism was explored by the computational simulations. Furthermore, a single network material with the same crosslinker content and degree of polymerization as the IPN was made. The SN was substantially weaker than the comparable IPN material. 

Dynamically cross‐linked polymer networks have attracted significant interest in recent years due to their unique and improved properties including increased toughness, malleability, shape memory, and self‐healing. Here, a computational study on the mechanical behavior of dynamically cross‐linked polymer networks is presented. Coarse grained models for different polymer network configurations are established and their mechanical properties using molecular dynamics (MD) simulations are predicted. Consistent with the experimental measurements, the simulation results show that interpenetrating networks (IPN) withstand considerably higher stress compared to the single networks (SN). Additionally, the MD results demonstrate that the origin of this variation in mechanical behavior of IPN and SN configurations goes back to the difference in spatial degrees of freedom of the noncovalent cross‐linkers, represented by nonbonded interactions within the two system types. The results of this work provide new insight for future studies on the design of systems with dual dynamic cross‐linkers where one linkage exchanges rapidly and provides autonomic dynamic character, while the other is a stimulus responsive dynamic covalent linkage that provides stability with dynamic exchange on‐demand.