-
Exploring the nanoscale origin of performance enhancement in Li$_{1.1}$Ni$_{0.35}$Mn$_{0.55}$O$_2$ batteries due to chemical do**
Authors:
Thomas Thersleff,
Jordi Jacas Biendicho,
Kunkanadu Prakasha,
Elias Martinez Moreno,
Leif Olav Jøsang,
Jekabs Grins,
Aleksander Jaworski,
Gunnar Svensson
Abstract:
Despite significant potential as energy storage materials for electric vehicles due to their combination of high energy density per unit cost and reduced environmental and ethical concerns, Co-free lithium ion batteries based off layered Mn oxides presently lack the longevity and stability of their Co-containing counterparts. Here, we demonstrate a reduction in this performance gap via chemical do…
▽ More
Despite significant potential as energy storage materials for electric vehicles due to their combination of high energy density per unit cost and reduced environmental and ethical concerns, Co-free lithium ion batteries based off layered Mn oxides presently lack the longevity and stability of their Co-containing counterparts. Here, we demonstrate a reduction in this performance gap via chemical do**, with Li$_{1.1}$Ni$_{0.35}$Mn$_{0.55}$O$_2$ achieving an initial discharge capacity of 159 mAhg$^{-1}$ at C/3 rate and a corresponding capacity retention of 94.3% after 150 cycles. We subsequently explore the nanoscale origins of this improvement through a combination of advanced diffraction, spectroscopy, and electron microscopy techniques, finding that optimized do** profiles lead to an improved structural and chemical compatibility between the two constituent sub-phases that characterize the layered Mn oxide system, resulting in the formation of unobstructed lithium ion pathways between them. We also directly observe a structural stabilization effect of the host compound near the surface using aberration corrected scanning transmission electron microscopy and integrated differential phase contrast imaging.
△ Less
Submitted 15 November, 2022;
originally announced November 2022.
-
Direct observation of Dirac states in Bi2Te3 nanoplatelets by 125Te NMR
Authors:
Wassilios Papawassiliou,
Aleksander Jaworski,
Andrew J. Pell,
Jae Hyuck Jang,
Yeonho Kim,
Sang-Chul Lee,
Hae ** Kim,
Yasser Alwahedi,
Saeed Alhassan,
Ahmed Subrati,
Michael Fardis,
Marina Karagianni,
Nikolaos Panopoulos,
Janez Dolinsek,
Georgios Papavassiliou
Abstract:
Detection of the metallic Dirac electronic states on the surface of Topological Insulators (TIs) is a tribune for a small number of experimental techniques the most prominent of which is Angle Resolved Photoemission Spectroscopy. However, there is no experimental method showing at atomic scale resolution how the Dirac electrons extend inside TI systems. This is a critical issue in the study of imp…
▽ More
Detection of the metallic Dirac electronic states on the surface of Topological Insulators (TIs) is a tribune for a small number of experimental techniques the most prominent of which is Angle Resolved Photoemission Spectroscopy. However, there is no experimental method showing at atomic scale resolution how the Dirac electrons extend inside TI systems. This is a critical issue in the study of important surface quantum properties, especially topological quasiparticle excitations. Herein, by applying advanced DFT-assisted solid-state 125Te Nuclear Magnetic Resonance on Bi2Te3 nanoplatelets, we succeeded in uncovering the hitherto invisible NMR signals with magnetic shielding influenced by the Dirac electrons, and subsequently showed how Dirac electrons spread and interact with the bulk interior of the nanoplatelets.
△ Less
Submitted 25 September, 2019;
originally announced September 2019.
-
Design and measurement methods for a lithium vapor box similarity experiment
Authors:
Jacob A Schwartz,
Eric D Emdee,
M A Jaworski,
R J Goldston
Abstract:
The lithium vapor box divertor is a concept for handling the extreme divertor heat fluxes in magnetic fusion devices. In a baffled slot divertor, plasma interacts with a dense cloud of Li vapor which radiates and cools the plasma, leading to recombination and detachment. Before testing on a tokamak the concept should be validated: we plan to study detachment and heat redistribution by a Li vapor c…
▽ More
The lithium vapor box divertor is a concept for handling the extreme divertor heat fluxes in magnetic fusion devices. In a baffled slot divertor, plasma interacts with a dense cloud of Li vapor which radiates and cools the plasma, leading to recombination and detachment. Before testing on a tokamak the concept should be validated: we plan to study detachment and heat redistribution by a Li vapor cloud in laboratory experiments. Mass changes and temperatures are measured to validate a Direct Simulation Monte Carlo model of neutral Li. The initial experiment involves a 5 cm diameter steel box containing 10 g of Li held at 650 degrees C as vapor flows out a wide nozzle into a similarly-sized box at a lower temperature. Diagnosis is made challenging by the required material compatibility with lithium vapor. Vapor pressure a steep function of temperature, so to validate mass flow models to within 10%, absolute temperature to within 4.5 K is required. The apparatus is designed to be used with an analytical balance to determine mass transport. Details of the apparatus and methods of temperature and mass flow measurements are presented.
△ Less
Submitted 8 August, 2018;
originally announced August 2018.
-
In-Datacenter Performance Analysis of a Tensor Processing Unit
Authors:
Norman P. Jouppi,
Cliff Young,
Nishant Patil,
David Patterson,
Gaurav Agrawal,
Raminder Bajwa,
Sarah Bates,
Suresh Bhatia,
Nan Boden,
Al Borchers,
Rick Boyle,
Pierre-luc Cantin,
Clifford Chao,
Chris Clark,
Jeremy Coriell,
Mike Daley,
Matt Dau,
Jeffrey Dean,
Ben Gelb,
Tara Vazir Ghaemmaghami,
Rajendra Gottipati,
William Gulland,
Robert Hagmann,
C. Richard Ho,
Doug Hogberg
, et al. (50 additional authors not shown)
Abstract:
Many architects believe that major improvements in cost-energy-performance must now come from domain-specific hardware. This paper evaluates a custom ASIC---called a Tensor Processing Unit (TPU)---deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOp…
▽ More
Many architects believe that major improvements in cost-energy-performance must now come from domain-specific hardware. This paper evaluates a custom ASIC---called a Tensor Processing Unit (TPU)---deployed in datacenters since 2015 that accelerates the inference phase of neural networks (NN). The heart of the TPU is a 65,536 8-bit MAC matrix multiply unit that offers a peak throughput of 92 TeraOps/second (TOPS) and a large (28 MiB) software-managed on-chip memory. The TPU's deterministic execution model is a better match to the 99th-percentile response-time requirement of our NN applications than are the time-varying optimizations of CPUs and GPUs (caches, out-of-order execution, multithreading, multiprocessing, prefetching, ...) that help average throughput more than guaranteed latency. The lack of such features helps explain why, despite having myriad MACs and a big memory, the TPU is relatively small and low power. We compare the TPU to a server-class Intel Haswell CPU and an Nvidia K80 GPU, which are contemporaries deployed in the same datacenters. Our workload, written in the high-level TensorFlow framework, uses production NN applications (MLPs, CNNs, and LSTMs) that represent 95% of our datacenters' NN inference demand. Despite low utilization for some applications, the TPU is on average about 15X - 30X faster than its contemporary GPU or CPU, with TOPS/Watt about 30X - 80X higher. Moreover, using the GPU's GDDR5 memory in the TPU would triple achieved TOPS and raise TOPS/Watt to nearly 70X the GPU and 200X the CPU.
△ Less
Submitted 16 April, 2017;
originally announced April 2017.
