Hyper-Raman spectroscopy has been performed in PbMg1/3Nb2/3O3 (PMN) from 800K to 50K, with particular attention to the low frequency region, down to about 5 cm-1 [1]. The high temperature spectra exhibit two resolved peaks at low frequency confirming the doublet structure of the soft mode of PMN, as previously anticipated [2,3]. An even more pronounced doublet structure has also been observed in PbMg1/3Ta2/3O3 (PMT) emphasizing therefore a common property of cubic relaxors. In PMN the lowest frequency component becomes overdamped at the onset of the Burns temperature Td but very interestingly, it hardens again on cooling down to 50 K. The relaxational frequency 02/ mimics the behavior of the inverse dielectric constant over the full temperature range explored: it extrapolates to the Curie temperature T0  390 K in the high temperature region, and close to the freezing temperature Tf  220K in the low temperature region. In between, the dynamics is controlled by a strongly overdamped soft excitation defining thereby a relaxational regime of the polarization fluctuations, characteristic of the relaxor’s state.

Raman scattering is performed in three sets of aluminosilicate glasses with light cations and concentrations varying from peralcaline to peraluminate domain. The depolarized spectra highlight two cation modes below ∼400 cm(-1). Comparison with infrared data reveals very stringent selection rules providing as much additional information for a vibrational analysis. The latter suggests in-phase (network-coupled) and out-of-phase (network-decoupled) displacements of the cations relative to their adjacent negatively charged structures. The low frequency vibration involves all cations whatever their role in the glass, network modifiers or charge compensators. Very interestingly, the second mode originates mostly from cations at modifier's places, providing thereby a new support for structural and chemical analysis of silicate glasses using Raman scattering.

The mechanical properties of nanocomposite materials are controlled to a large extent by the filler-filler interactions. Nevertheless, another important contribution - less well understood - is due to the polymer chain-filler interactions. Experimentally, the radius of gyration (Rg) of polymer chains in nanocomposites can be measured by SANS using the zero average contrast conditions (mixing hydrogenated (H) and deuterated (D) chains) in order to match the filler signal and measure the chain form factor. However, many studies display an unexplained polluted SANS signal in the low-q range. In this talk, we will discuss the measurement of polymer chain signal in nanocomposites prepared from the drying of a colloidal dispersion of silica and polymer particles. In such samples, the mixture of H and D chains results from the dissolution of H and D latex beads. This dissolution was followed by SANS in the low q range, as function of thermal annealing (figure 1) and filler content using an original model [1]. Experimental results demonstrate that the dissolution dynamics of polymer chains is significantly slowed down by the presence of silica nanoparticles [1]. Besides, the effect of the filler size on the chain signal in the nanocomposites was studied by a combination of SAXS and SANS measurements and reveal that the filler contribute to the SANS signal in nanocomposites filled with small silica nanoparticles (compared to the latex beads) contrarily to nanocomposites filled with bigger silica particles. We rationalize this observation considering an attractive interaction between polymers and silica particles leading to an inhibited interdiffusion of latex particles in the vicinity of silica particles, and a statistical local assymmetry due to the small number of latex beads defining the environment of small silica particles. Finally, no evolution of the polymer Rg was found in our samples.References[1] Genix et al., Macromolecules (2012), 45, 1663.[2] Banc et al., Macromolecules (2015), in press.

Small Angle Scattering techniques are regularly used to probe the structure of proteins and proteins assemblies in a solvent. X-rays enable fast and localized measurements thanks to the high flux and the small beam size, but there is a risk of protein degradation that is prevented with neutron scattering. In order to increase the contrast between proteins and the solvent, deuterated solvents are usually used for SANS measurements. Normally, if the contrast is due to scattering density differences between proteins and the solvent, SAXS and SANS profiles are identical. In this talk, we will present the study of gluten protein gels that display radically different SAXS and SANS profiles when the solvent is deuterated. The shape of the high q signal is identical for both measurements, whereas totally different features are measured at low q (figure). The SAXS profile can be successfully described with a model combining a Lorentzian and a q-2 power law [1]. This model was previously used to describe polymeric gels under good solvent conditions and relates the random coil Gaussian chain conformation of proteins and large-scale concentration heterogeneities. For the range of q where a q-2 power law regime is measured for SAXS, SANS profile displays instead a scattered intensity that scales as q-4 and a shoulder at even smaller q from which a characteristic size can be extracted. Measurements with various levels of solvent deuteration were performed to understand these surprising scattering profiles. Quantitative analysis of the contrast with deuteration indicates a progressive H-D exchange of protein chains. Moreover, an increase of the low q q-4 contribution is observed with solvent deuteration. Results were rationalized considering a heterogeneous H-D exchange of protein protons that would be due to local high concentration of proteins where hydrogen would be involved in hydrogen bonds, preventing thus H-D exchange. In particular, the q-4 behavior would originate from the sharp interfaces between protein rich regions which are more hydrogenated and protein diluted regions which are more deuterated. The cross-over between the q-4 and the quasi-plateau at smaller q correspond to a distance of 60 nm, similar to the size of the protein assemblies measured in the dilute regime, reflecting the characteristic size over which H-D exchange can take place.

We have performed classical molecular dynamics simulations in order to study the changes under compression in the local and medium range structural properties of three sodium borosilicate glasses with varying sodium content. These glasses have been isostatically compressed up to 20 GPa and then decompressed in order to analyze the different mechanisms that affect densification, alongside with the permanent modifications of the structure after a full compression/decompression cycle. The results show that the atomic packing is the prominent characteristic that governs the amount of densification in the glass, as well as the setup of the permanent densification. During compression, the bulk modulus increases linearly up to approximately 15 GPa and more rapidly for higher pressures, a behavior which is reflected on the rate of increase of the average coordination for B and Na. Radial distribution functions at different pressures during the cycle help to quantify the amount of distortions in the elementary structural units, with a pronounced shortening of the Na–Na and Na–O bond lengths during compression. A subsequent decomposition of the glassy matrix into elementary Voronoi volumes verifies the high compressibility of Na-rich regions.

Many canonical processes in molecular biology rely on the dynamic assembly of higher-order nucleoprotein complexes. In bacteria, the assembly mechanism of ParABS, the nucleoprotein super-complex that actively segregates the bacterial chromosome and many plasmids, remains elusive. We combined super-resolution microscopy, quantitative genome-wide surveys, biochemistry, and mathematical modeling to investigate the assembly of ParB at the centromere-like sequences parS. We found that nearly all ParB molecules are actively confined around parS by a network of synergistic protein-protein and protein-DNA interactions. Interrogation of the empirically determined, high-resolution ParB genomic distribution with modeling suggests that instead of binding only to specific sequences and subsequently spreading, ParB binds stochastically around parS over long distances. We propose a new model for the formation of the ParABS partition complex based on nucleation and caging: ParB forms a dynamic lattice with the DNA around parS.This assembly model and approach to characterizing large-scale, dynamic interactions between macro-molecules may be generalizable to many unrelated machineries that self-assemble in superstructures.