Towards a universal size distribution in a polymer network. Implications for drug delivery and plasmonic nanoparticle transport phenomena in polysaccharide and synthetic hydrogels.

Abstract

Polymeric hydrogels are paramount to outstanding applications in biology, medicine, pharmacy. Their similarity to living tissues is leveraged in clinical branches (oncology, cardiology, immunology, neurology, wound healing) for delivering a large range of drugs (encompassing DNA, RNA, protein molecules) and realizing in-vivo models of stimuli-responsive or controlled drug release. Rubber elasticity theory and the swollen network hypothesis are key for properly designing the geometric and mechanical features of hydrogels and polymer networks. The assumption of a Gaussian distribution of end-to-end lengths in a polymer molecule, however, can break down in several cases. Here, strongly supported by Low field NMR and rheology experiments, we propound the generalized Weibull law of extreme value statistics (EVS) to have universal validity in hydrogel materials. Mesh size values that account for an intrinsic statistical dependence between monomeric positions (or stiffness) show much better agreement with measurements conducted on physically crosslinked samples (agar, alginate and scleroglucan), including sputum specimens (rich in mucins) from patients affected by chronic respiratory conditions (cystic fibrosis) and on chemically crosslinked samples (poly-vinylpyrrolidone, PVP; poly-(ethylene-glycol/propylene-glycol), PEG/PPG). Across all ten gels, the Gaussian distribution yields the smallest average mesh size, ranging roughly from 7 nm for the densest alginate 2 % (9 gl^-1) hydrogel to about 80 nm for one of the sputum. Working with the pierced Gaussian inflates the mesh size to ≈1.5 × the Gaussian value, with increases from a modest +4 % in alginate 1 % up to nearly +100 % in the open PVP network (48 → 98 nm). The generalized Weibull distribution usually falls between the two Gaussians, yet in agar 1 % and scleroglucan 2 % it overtakes the pierced Gaussian (e.g. 20.2 > 15.8 nm for agar 1 %), reflecting a strong heavy-tailed distribution. The predicted mesh order therefore is Gaussian < generalized Weibull ≈ pierced Gaussian, with the precise ranking ruled by the width and skewness of each network statistics. Overall, our findings - being straightforward to apply - will profoundly impact on the description, conception and control of polymer networks, which often demand advanced instrumental techniques for compensating the lack of adequate predictive models. Among other relevant implications, aside from drug delivery, we highlight the characterization of the photothermal (or thermoplasmonic) response of hydrogel matrices hosting metal nanoparticles (e.g. with applications in hyperthermia cancer treatment and enhanced chemical processes). On the theoretical side, we emphasize the study of transport and thermomechanical properties of polymeric networks.