Polymer Dynamics in PEG-Silica Nanocomposites: Effects of Polymer Molecular Weight, Temperature and Solvent Dilution

Abstract

The mechanical properties of particulate nanocomposites strongly depend upon the particle dispersion, as well as on the closely related properties in thin polymer films covering the particle surface. The length scale of such changes is relevant for the understanding of particle−particle interactions, which ultimately dominate the mechanical response. Using well-defined 44 nm diameter silica nanoparticles dispersed in poly(ethylene glycol), we focus on surface-induced changes in polymer dynamics. Using proton time-domain NMR, we distinguish three polymer phases of different mobility, i.e., a strongly adsorbed, solid-like fraction, a fraction with intermediate relaxation times and a highly mobile fraction. We explore how these fractions change as we vary polymer molecular weight from 300 to 20 000 and particle volume fraction up to 0.3. A multiple-quantum experiment enables a closer analysis of the mobile component which we show consists of two fractions, one resembling the bulk melt-like and another one showing network-like properties. We demonstrate that above a polymer molecular weight-dependent volume fraction, polymers form elastically active links between particles, resulting in the physical gelation observed in such systems. Our results provide a quantitative picture of network formation, which is described by the amount and length of network-like chains as well as heterogeneities in the polymer dynamics. We relate changes in polymer dynamics to particle microstructure obtained from small angle neutron scattering.

Reduced-mobility layers with high internal mobility in poly(ethylene oxide)–silica nanocomposites

“…Therefore, conclusions based upon neutron-scattering observations at comparably high temperatures, stating the absence of a "glassy" layer 9 or even claiming "contradiction" 14 to earlier NMR results, 13 originate partially from the simplistic nomenclature and, specifically, a lack of a more detailed discussion of the layer properties in Ref. 13.…”

“…They are specific for polymers displaying attractive interactions with either the bare or the surface-modified filler. Bare oxide fillers such as silica and alumina, displaying surface hydroxy groups, interact favorably with polymers capable of acting as hydrogen-bond acceptor, such as poly(dimethyl siloxane), PDMS, [8][9][10] poly(ethylene oxide), PEO, [11][12][13][14][15] poly(2-vinyl pyridine), P2VP, 16,17 or different acrylate polymers. 5,[18][19][20][21][22] The amount of dynamically modified polymer, and with this the nominal thickness of an assumed-to-be contiguous reduced-mobility layer, can be estimated on the basis of results from different techniques.…”

“…URL: www.physik.uni-halle.de/nmr system and the same method. [12][13][14] In this context DSC, 10,16,19 dielectric spectroscopy, 23,24 NMR spectroscopy, 13,18,20,21 and static 22 as well as dynamic 9,11,12,14,15 neutron scattering techniques have been used. Apart from purely structural (=density) information obtained by static scattering, 22 most other techniques are sensitive to dynamic properties.…”

“…For example, the notion of a "glassy layer" has emerged 1,4,13,18,20 but is subject to ambiguity as this implies literally that the reduced-mobility components are below their T g , i.e., have a cooperative segmental relaxation time (τ α ) of at least 100 s at the temperature of measurement, and may exhibit only rather localized faster motions ( β and higher relaxations). Therefore, conclusions based upon neutron-scattering observations at comparably high temperatures, stating the absence of a "glassy" layer 9 or even claiming "contradiction" 14 to earlier NMR results, 13 originate partially from the simplistic nomenclature and, specifically, a lack of a more detailed discussion of the layer properties in Ref. 13.…”

“…We compare samples previously studied by neutron spin-echo (NSE) and time-of-flight (TOF) spectroscopy 14,15 with a sample from a series previously studied by NMR 13 and highlight the qualitative differences. We can thus reconcile previous conclusions on a potential disagreement of neutron-scattering and NMR results-which turn out to be only apparent.…”