Search NASA⌕ Search

Engineering topics

Keffer, David

Publications and source records attributed to Keffer, David.

Local structure and distortions of mixed methane-carbon dioxide hydrates

A vast source of methane is found in gas hydrate deposits, which form naturally dispersed throughout ocean sediments and arctic permafrost. Methane may be obtained from hydrates by exchange with hydrocarbon byproduct carbon dioxide. It is imperative for the development of safe methane extraction and carbon dioxide sequestration to understand how methane and carbon dioxide co-occupy the same hydrate structure. Pair distribution functions (PDFs) provide atomic-scale structural insight into intermolecular interactions in methane and carbon dioxide hydrates. We present experimental neutron PDFs of methane, carbon dioxide and mixed methane-carbon dioxide hydrates at 10 K analyzed with complementing classical molecular dynamics simulations and Reverse Monte Carlo fitting. Mixed hydrate, which forms during the exchange process, is more locally disordered than methane or carbon dioxide hydrates. The behavior of mixed gas species cannot be interpolated from properties of pure compounds, and PDF measurements provide important understanding of how the guest composition impacts overall order in the hydrate structure.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Manufacturing of Monolithic Electrodes from Low-Cost Renewable Resources

Lignin, a low-cost, biomass derived precursor, was selected as an alternative for carbon based free standing anodes in Li-ion batteries. Industrially scalable melt-spinning and melt-blowing synthesis methods were developed at Oak Ridge National Laboratory that are compatible with industrially viable production. Engineering studies predict that LCFs can be manufactured at $3/lb using these technologies, which compares favorably to $12/lb for battery grade graphite. The physical properties of lignin carbon fibers, specifically the tunable electrochemical and thermal transport, are suitable for energy storage applications as both an active material and current collector. The elimination of inactive components in the slurry-coated electrodes was enabled by LCF processing parameters modifications to produce monolithic mats in which the fibers are electrically interconnected. These mats were several hundreds of micrometers thick, and the fibers functioned as both current collector and active material by virtue of their mixed ionic/electronic conductivities. The LCFs were coated onto copper current collectors with PVDF binder and conductive carbon additive through conventional slurry processing. Galvanostatic cycling of the LCFs against Li revealed reversible capacities greater than 300 mAh/g. The coulombic efficiencies were over 99.8%. The mats were galvanostatically cycled in half cells against Li. Specific capacities as high as 250 mAh/g were achieved approximately 17% lower than the capacities of the same fibers in slurries. However, there were no inactive materials reducing the practical specific capacity of the entire electrode construction. Lithiation and delithiation of the LCFs proceeded with coulombic efficiencies greater than 99.9%, and the capacity retention was greater than 99% over 100 cycles at a rate of 15 mA/g. Research sponsored by the Laboratory Directed Research and Development Program of Oak Ridge National Laboratory, managed by UT-Battelle, LLC, for the U. S. Department of Energy.

McNutt, Nichiolas William↗

A Multiscale Modeling Demonstration Based on the Pair Correlation Function

For systems with interatomic interactions that are well described by pair-wise potentials, the pair correlation function provides a vehicle for passing information from the molecular level to the macroscopic level of description. In this work, we present a complete demonstration of the use of the pair correlation function to simulate a fluid at the molecular and macroscopic levels. At the molecular level, we describe a monatomic fluid using the Ornstein-Zernike integral equation theory closed with the Percus-Yevick approximation. We show that all of the required thermodynamic properties can be evaluated knowing the pair correlation function. At the macroscopic level, we perform a multiscale simulation with macroscopic evolution equations for the mass, momentum, temperature, and pair correlation function, using molecular-level simulation to provide the boundary conditions. We perform a self-consistency check by comparing the pair correlation function that evolved from the multiscale simulation with the one evaluated at the molecular-level; excellent agreement is achieved.

97 MATHEMATICS AND COMPUTING↗