No abstract

Cellulose is crystallized by plants and other organisms into fibrous nanocrystals. The mechanical properties of these nanofibers and the formation of helical superstructures with energy dissipating and adaptive optical properties depend on the ordering of polysaccharide chains within these nanocrystals, which is typically measured in bulk average. Direct measurement of the local polysaccharide chain arrangement has been elusive. In this study, we use the emerging technique of scanning electron diffraction to probe the packing of polysaccharide chains across cellulose nanofibers and to reveal local ordering of the chains in twisting sections of the nanofibers. We then use atomic force microscopy to shed light on the size dependence of the inherent driving force for cellulose nanofiber twisting. The direct measurement of crystalline twisted regions in cellulose nanofibers has important implications for understanding single-cellulose-fibril properties that influence the interactions between cellulose nanocrystals in dense assemblies. This understanding may enable cellulose extraction and separation processes to be tailored and optimized.

No abstract

We propose a training method for deep neural network (DNN)-based source enhancement to increase objective sound quality assessment (OSQA) scores such as the perceptual evaluation of speech quality (PESQ). In many conventional studies, DNNs have been used as a mapping function to estimate time-frequency masks and trained to minimize an analytically tractable objective function such as the mean squared error (MSE). Since OSQA scores have been used widely for soundquality evaluation, constructing DNNs to increase OSQA scores would be better than using the minimum-MSE to create highquality output signals. However, since most OSQA scores are not analytically tractable, i.e., they are black boxes, the gradient of the objective function cannot be calculated by simply applying back-propagation. To calculate the gradient of the OSQA-based objective function, we formulated a DNN optimization scheme on the basis of black-box optimization, which is used for training a computer that plays a game. For a black-box-optimization scheme, we adopt the policy gradient method for calculating the gradient on the basis of a sampling algorithm. To simulate output signals using the sampling algorithm, DNNs are used to estimate the probability-density function of the output signals that maximize OSQA scores. The OSQA scores are calculated from the simulated output signals, and the DNNs are trained to increase the probability of generating the simulated output signals that achieve high OSQA scores. Through several experiments, we found that OSQA scores significantly increased by applying the proposed method, even though the MSE was not minimized.

No abstract

In materials science and crystallography, the true density is an important derived physical quantity of solids. Here we report the correlation of the true density of nanometer-wide fibrillar crystallites of cellulose with their purity, crystallinity, morphology, and surface functionality. In the single fibrils, all the cellulose molecules are uniaxiallly oriented. Thus, the true density indicates the molecular packing density in the single fibrils and is essential for the precise estimation of the volume fraction of cellulose in fibril-based composites or porous structures. We demonstrate that the true density of fibrillar crystallites of cellulose is approximately 1.60 g/cm3 irrespective of the biological origins of the cellulose (wood, cotton, or a tunicate) and the crystallinity. The true density is in fact independent of the dimension of the crystallites and the atomic conformation of the uniaxially oriented but noncrystalline molecules at the crystallite surface. In the single fibrils, all the cellulose molecules are densely packed from the crystalline core to the noncrystalline outermost regions. The value of 1.60 g/cm3 remains unchanged even when the fibrils are dispersed through the wet disintegration process of “nanocellulose” production. In contrast, tailoring the surface functionality of the fibrils by oxidation and/or adsorption results in a substantial change in the true density up to 1.8 g/cm3 or down to 1.3 g/cm3. The true density of nanocellulose is indeed governed by the surface functionality and has a strong gradient in the fibril cross-sectional direction.

No abstract

No abstract