Search for: All records

Despite groundbreaking advances in the additive manufacturing of polymers, metals, and ceramics, scaled and accurate production of structured carbons remains largely underdeveloped. This work reports a simple method to produce complex carbon materials with very low dimensional shrinkage from printed to carbonized state (less than 4%), using commercially available polypropylene precursors and a fused filament fabrication-based process. The control of macrostructural retention is enabled by the inclusion of fiber fillers regardless of the crosslinking degree of the polypropylene matrix, providing a significant advantage to directly control the density, porosity, and mechanical properties of 3D printed carbons. Using the same printed plastic precursors, different mechanical responses of derived carbons can be obtained, notably from stiff to highly compressible. This report harnesses the power of additive manufacturing for producing carbons with accurately controlled structure and properties, while enabling great opportunities for various applications.

 

The choice of learning rate (LR) functions and policies has evolved from a simple fixed LR to the decaying LR and the cyclic LR, aiming to improve the accuracy and reduce the training time of Deep Neural Networks (DNNs). This article presents a systematic approach to selecting and composing an LR policy for effective DNN training to meet desired target accuracy and reduce training time within the pre-defined training iterations. It makes three original contributions. First, we develop an LR tuning mechanism for auto-verification of a given LR policy with respect to the desired accuracy goal under the pre-defined training time constraint. Second, we develop an LR policy recommendation system (LRBench) to select and compose good LR policies from the same and/or different LR functions through dynamic tuning, and avoid bad choices, for a given learning task, DNN model, and dataset. Third, we extend LRBench by supporting different DNN optimizers and show the significant mutual impact of different LR policies and different optimizers. Evaluated using popular benchmark datasets and different DNN models (LeNet, CNN3, ResNet), we show that our approach can effectively deliver high DNN test accuracy, outperform the existing recommended default LR policies, and reduce the DNN training time by 1.6-6.7× to meet a targeted model accuracy.

 

The diversity of cell types is a challenge for quantifying aging and its reversal. Here we develop ‘aging clocks’ based on single-cell transcriptomics to characterize cell-type-specific aging and rejuvenation. We generated single-cell transcriptomes from the subventricular zone neurogenic region of 28 mice, tiling ages from young to old. We trained single-cell-based regression models to predict chronological age and biological age (neural stem cell proliferation capacity). These aging clocks are generalizable to independent cohorts of mice, other regions of the brains, and other species. To determine if these aging clocks could quantify transcriptomic rejuvenation, we generated single-cell transcriptomic datasets of neurogenic regions for two interventions—heterochronic parabiosis and exercise. Aging clocks revealed that heterochronic parabiosis and exercise reverse transcriptomic aging in neurogenic regions, but in different ways. This study represents the first development of high-resolution aging clocks from single-cell transcriptomic data and demonstrates their application to quantify transcriptomic rejuvenation.

 

Energy transport in proteins is critical to a variety of physical, chemical, and biological processes in living organisms. While strenuous efforts have been made to study vibrational energy transport in proteins, thermal transport processes across the most fundamental building blocks of proteins, i.e. helices, are not well understood. This work studies energy transport in a group of “isomer” helices. The π-helix is shown to have the highest thermal conductivity, 110% higher than that of the α-helix and 207% higher than that of the 3 10 -helix. The H-bond connectivity is found to govern thermal transport mechanisms including the phonon spectral energy density, dispersion, mode-specific transport, group velocity, and relaxation time. The energy transport is strongly correlated with the H-bond strength which is also modulated by the H-bond connectivity. These fundamental insights provide a novel perspective for understanding energy transfer in proteins and guiding a rational molecule-level design of novel materials with configurable H-bonds. 

Efficient heat dissipation in batteries is important for thermal management against thermal runaway and chemical instability at elevated temperatures. Nevertheless, thermal transport processes in battery materials have not been well understood especially considering their complicated microstructures. In this study, lattice thermal transport in lithium cobalt oxide (LiCoO 2 ), a popular cathode material for lithium ion batteries, is investigated via molecular dynamics-based approaches and thermal resistance models. A LiCoO 2 single-crystal is shown to have thermal conductivities in the order of 100 W m −1 K −1 with strong anisotropy, temperature dependence, and size effects. By comparison, polycrystalline LiCoO 2 is more isotropic with much lower thermal conductivities. This difference is caused by random grain orientations, the thermal resistance of grain boundaries, and size-dependent intra-grain thermal conductivities that are unique to polycrystals. The grain boundary thermal conductance is calculated to be in the range of 7.16–25.21 GW m −2 K −1 . The size effects of the intra-grain thermal conductivities are described by two empirical equations. Considering all of these effects, two thermal resistance models are developed to predict the thermal conductivity of polycrystalline LiCoO 2 . The two models predict a consistent thermal conductivity–grain size relationship that agrees well with molecular dynamics simulation results. The insights revealed by this study may facilitate future efforts on battery materials design for improved thermal management.