Relaxation Processes at Interfaces in Polymer Nanocomposites

Abstract

Polymer nanocomposites (PNC) where nanofillers such as nanoparticles and nanotubes are dispersed show better mechanical, electric and dielectric properties in comparison with the neat polymer matrix. The improvement of the properties is strongly dependent on the formation of interfacial regions between fillers and matrix polymer chains. A core-shell model [1], in which each nanofiller (core) is supposed to be covered with an interfacial polymer layer (shell) with slower dynamics than the bulk one, has been widely used to rationalize the property improvement. When the nanofiller fraction increases above a percolation threshold, the presence of nanofiller network can lead to the formation of interfacial regions with hierarchical dynamics. A quantitative understanding of the hierarchical nature in the relaxation dynamics of interfacial chains is critical for tailoring multiple properties of PNCs. In this study, the relaxation dynamics of a model rubber nanocomposite, nitrile-butadiene rubber (NBR) with silica (SiO2) nanoparticles, was characterized over broad temperature and frequency ranges by means of dielectric relaxation spectroscopy (DRS). As samples, NBR with a number-average molecular weight of 81.9k, a polydispersity index of 3.56 and a glass transition temperature (T g) of 242 K, and SiO2 nanoparticles with an average diameter of 18 nm, and their composites were kindly provided by Bridgestone Co., Ltd.. The weight content of acrylonitrile component in NBR was 34%, and the volume fraction of SiO2 nanoparticle in PNC (Φ) was varied from ~5 to ~30%. For DRS measurements, neat NBR and PNC films were prepared by a spin-coating method from tetrahydrofuran solutions onto quartz substrates coated with a 50 nm-thick Al layer. The films were dried in vacuum for 24 h at room temperature, and then a second 50 nm-thick Al layer was deposited on the surface of each film to form a parallel-plate capacitor. DRS measurements were conducted using a frequency response analyzer over the range from 10-1 to 106 Hz. Panels (a) and (b) of Figure 1 show representative examples of SEM images for PNC samples with Φ = 7 and 17 vol%, respectively. For the PNC sample with Φ = 7 vol%, SiO2 agglomerates with different sizes were distributed in the NBR matrix without the formation of long-range particle-particle structured network. In contrast, at Φ = 17 vol%, the SiO2 agglomerates tended to connect each other to form a long-range particle-particle network. This behavior has been experimentally confirmed in many nanocomposite systems. However, for a moment, it remains challenging to discuss a relationship between the particle-particle network and the relaxation dynamics of interfacial polymers. Panel (c) of Figure 1 shows dielectric loss (ε″) spectra for the neat NBR and PNCs as a function of frequency measured at 262 K. A strong peak representing the segmental relaxation process of NBR chains was observed around 5×102 Hz for each sample [2]. However, comparing with the neat NBR, the peak for the PNCs exhibited a decrease in amplitude and broadening on the low-frequency side. This trend became more remarkable for samples with higher Φ. This suggests the existence of a second relaxation process which represents the relaxation dynamics of the polymer shell layer surrounding SiO2 nanoparticle surface [1]. The ε′′ data for the neat NBR could be well fit using a Havriliak-Negami (HN) function with a single relaxation process [2]. However, a HN function with double relaxation processes and a conductivity term were necessary to well express the ε′′ data for the NBR-SiO2 films. The relaxation process of the shell layer was found to be slower than that of the bulk process by approximately two decades and almost independent of Φ. Panel (d) of Figure 1 shows temperature dependence of ε″. For all samples, a peak assignable to the segmental relaxation process was observed around 262 K. For the PNC films with Φ = 17 and 29 vol%, an additional peak was observed at a temperature higher than that for the bulk segmental relaxation peak by 50 ~ 80 K. The temperature elevation was more remarkable for the PNC sample with Φ = 29 vol%. Although the origin of this relaxation process can be hardly concluded, it seems reasonable to consider that it is arisen from slowed dynamics for chains restricted on the surface of SiO2 nanoparticles. [1] S. Cheng, S. Mirigian, J.-M. Y. Carrillo, V. Bocharova, B. G. Sumpter, K. S. Schweizer, and A. P. Sokolov, J. Chem. Phys. 143, 194704 (2015). [2] H. K. Nguyen, A. Konomi, S. Sugimoto, M. Inutsuka, D. Kawaguchi, and K. Tanaka, Macromol. Chem. Phys. 219, 1700329 (2018). Figure 1