Structure and dynamics of polymer networks using non-Euclidean geometry

SNIC 2018/3-76


SNAC Medium

Principal Investigator:

Christer Elvingson


Uppsala universitet

Start Date:


End Date:


Primary Classification:

10402: Physical Chemistry

Secondary Classification:

10105: Computational Mathematics

Tertiary Classification:

10407: Theoretical Chemistry



Polymer gels in the form of covalently cross-linked networks are important in, e.g., tissue engineering and for drug delivery, as well as in different types of actuators. These systems are also used for fundamental studies of transport processes in, e.g. cells. We have developed a novel algorithm to create a closed network on S3 without any underlying bias from a starting structure in R3, randomly distributing the crosslinking nodes directly on the hypersphere. Looking at the diffusion of tracer particles in these gels showed the importance of the dynamics of the network on the diffusion process, and that particles sizes well above the maximum in the pore size distribution had a finite diffusion coefficient (ref. 1). Another important feature of directly distributing the crosslinking nodes before forming a closed network, is that one can construct gel nanoparticles. We have recently used this to also look at the dynamics during collapse transitions for core-shell nanoparticles (ref. 2). With these tools now available, we will continue to study the influence of inhomogeneities on different length scales, and how this will affect the diffusion of tracer particles. This is most relevant comparing with experimental data since it is know that crosslinked gel materials usually have a rather high degree of heterogeneity. In parallel with this, we will also develop tools to analyse the structure of closed networks and gels living on e.g. a hypersphere. In a first step, we have investigated the mechanical properties of inhomogeneous crosslinked networks (refs. 3-4), and will now combine these efforts when looking at diffusion in crowded systems. The need for cpu-time would be in the range 5-10000 cpu hours per month during autumn, but considerably less during the spring, since a new PhD student will first acquire the theoretical background taking full time courses during the spring. The disk space needed to store the resulting trajectories during the analysis stage would be at least 500Gb during the present part of the project. 1. N. Kamerlin et. al., J. Phys., Cond. Matter 28 (2016) 475101 2. N. Kamerlin, Macromolecules 49 (2016) 5740 3. N. Kamerlin Macromolecules 50 (2017) 7628 4. N. Kamerlin Macromolecules 50 (2017) 9353