The structure of polymer nanocomposites has important consequences on final properties, like e.g.mechanical reinforcement. The structure of the hard filler phase is usually characterized by electronmicroscopy and small-angle X-ray scattering (SAXS), and we will review recent examples of fillerstructure in model and industrial nanocomposites. The chain conformation can only bemeasured by small-angle neutron scattering (SANS) with isotopic substitution. Over the past 15 years,considerable efforts have been dedicated to such measurements. Results appear to converge towardGaussian statistics of chains, unperturbed by the presence of filler NPs, and exceptions seem to be dueto system-dependent effects. In nearly all studies, however, unexplained pollution of the scatteringdata in the low-q range has been observed, in spite of careful contrast-matching.In this study, the conformation of polymer chains in silica-latex nanocomposites has been studiedunder zero-average contrast (ZAC) conditions using SANS. Samples have been prepared by dryingcolloidal suspensions of silica and polymer nanoparticles (NPs) for two different silica NPs (radius of5 and 15 nm) and two chain masses (17 and 100 kg/mol). By mixing appropriate amounts ofhydrogenated and deuterated polymer, chain scattering contrast is introduced, and in principle silicascattering is suppressed. However, the SANS scattering intensities displayed distinct deviations at lowangles. Comparison of the SANS and SAXS intensities proves that the measurement of chain structureby SANS is perturbed by unwanted silica contributions, i.e., the silica is ‘mysteriously’ visible. Aquantitative model rationalizing these contributions for the first time is proposed. It is based on astatistical H/D-mismatch in the local environment of the silica NPs. Such a mismatch leads to theexistence of a nanometer-thick polymer shell surrounding silica NPs, which carries H/D contrast andthereby produce an undesired scattering contribution. These shells are presumably maintained by thereduced mobility of polymer close to interfaces or attractive polymer-silica interactions. The effect ofunwanted filler scattering is NP size-dependent − it is pronounced only for smaller beads. This isprobably why ZAC turned out to be so difficult with polymer nanocomposites.

A large body of experimental work on the microstructure and dynamics of simplifiedindustrial nanocomposites made of disordered silica filler in a styrene-butadiene matrixby solid-phase mixing is regrouped and critically discussed in this feature article. Recentresults encompass systems with varying polymer mass, grafting functionality, and fillercontent. They have been obtained by simulation-based structural modelling of nanoparticleaggregate size and mass deduced from small-angle scattering and transmission electronmicroscopy. Our model has been validated by independent swelling experiments.Comparison of structurally-close nanocomposites of widely different chain mass led tothe identification of a unique structure-determining parameter, the grafting density, aswell as to a unified picture of aggregate formation mechanisms in complex nanocompositesduring mixing. In addition, low-field proton NMR allowed for the characterization ofdynamically slowed-down (‘glassy’) polymer layers, which were shown not to dominatethe rheological response, unlike the structural contribution. Finally, broadband dielectricspectroscopy was used in an innovative manner to identify filler percolation – also identifiedby rheology – via dynamics along filler surfaces.

Alumina-coated silica nanoparticles (NPs) grafted with phosphonic acids of different hydrophobicitywere used as filler in poly(ethylacrylate) nanocomposites. Phosphonic acids bearing short alkyl chains ora diethylene glycol group have been grafted at densities up to 3.2 P/nm2 on NPs (20 nm) dispersed inwater. Nanocomposites at particle fractions up to 10 vol% have been formulated by casting from thecolloidal mixtures of modified NPs and nanolatex in water. The dispersion of the NPs in the polymermatrix has been studied by TEM combined with small-angle scattering, evidencing aggregation of NPs.TEM shows micrometer-scale inhomogeneities depending on the surface/polymer matrix compatibility.For the local interparticle correlations, a quantitative analysis of the intensity based on the mapping ontothe effective structure factor of polydisperse hard spheres is developed. This mapping allows the modelfreedetermination of the internal volume fraction of aggregates, termed compacity k, to between 10%and 30%, compatible with the TEM analysis. k is found to increase for the higher particle volume fractions,to decrease with grafting density, and to be mostly independent of the nature and mass of thegraft. Preliminary evidence for an improved compatibility of grafted with respect to bare NPs is found, asopposed to their aqueous precursor suspensions where some pre-aggregation is induced by grafting.

Polymer nanocomposites are used widely, mainly for the industrial application of car tyres. The rheological behavior of such nanocomposites depends in a crucial way on thedispersion of the hard filler particles – typically silica nanoparticles embedded in a soft polymer matrix. It is thus important to assess the filler structure, which may be quitedifficult for aggregates of nanoparticles of high polydispersity, and with strong interactions at high loading. This has been achieved recently using a coupled TEM/SAXSstructural model describing the filler microstructure of simplified industrial nanocomposites with grafted or ungrafted silica of high structural disorder. Here, wepresent an original method capable of reducing inter-aggregate interactions by swelling of nanocomposites, diluting the filler to low-volume fractions. Note that this isimpossible to reach by solid mixing due to the large differences in viscoelasticity between the composite and the pure polymer. By combining matrix crosslinking,swelling in a good monomer solvent, and post-polymerization of these monomers, it isshown that it is possible to separate the filler into small aggregates. The latter have then been characterized by electron microscopy and small-angle X-ray scattering,confirming the conclusions of the above mentioned TEM-SAXS structural model applied directly to the highly loaded cases.

The structure of polymer nanocomposites has important consequences on final properties, like for instance mechanical reinforcement. While the structure of the hard filler phase is usually characterized by electron microscopy and small-angle X-ray scattering, the chain conformation can only be measured by small-angle neutron scattering (SANS).