Engineering topics

Dufek, Eric J.

Publications and source records attributed to Dufek, Eric J..

At least 19 records

DOE OSTI2024

PINN surrogate of Li-ion battery models for parameter inference, Part I: Implementation and multi-fidelity hierarchies for the single-particle model

To plan and optimize energy storage demands that account for Li-ion battery aging dynamics, techniques need to be developed to diagnose battery internal states accurately and rapidly. Here, this study seeks to reduce the computational resources needed to determine a battery's internal states by replacing physics-based Li-ion battery models - such as the single-particle model (SPM) and the pseudo-2D (P2D) model - with a physics-informed neural network (PINN) surrogate. The surrogate model makes high-throughput techniques, such as Bayesian calibration, tractable to determine battery internal parameters from voltage responses. This manuscript is the first of a two-part series that introduces PINN surrogates of Li-ion battery models for parameter inference (i.e., state-of-health diagnostics). In this first part, a method is presented for constructing a PINN surrogate of the SPM. A multi-fidelity hierarchical training, where several neural nets are trained with multiple physics-loss fidelities is shown to significantly improve the surrogate accuracy when only training on the governing equation residuals. The implementation is made available in a companion repository (https://github.com/NREL/PINNSTRIPES). The techniques used to develop a PINN surrogate of the SPM are extended in Part II for the PINN surrogate for the P2D battery model, and explore the Bayesian calibration capabilities of both surrogates.

25 ENERGY STORAGE↗

DOE OSTI2024

PINN surrogate of Li-ion battery models for parameter inference, Part II: Regularization and application of the pseudo-2D model

Bayesian parameter inference is useful to improve Li-ion battery diagnostics and can help formulate battery aging models. However, it is computationally intensive and cannot be easily repeated for multiple cycles, multiple operating conditions, or multiple replicate cells. To reduce the computational cost of Bayesian calibration, numerical solvers for physics-based models can be replaced with faster surrogates. A physics-informed neural network (PINN) is developed as a surrogate for the pseudo-2D (P2D) battery model calibration. For the P2D surrogate, additional training regularization was needed as compared to the PINN single-particle model (SPM) developed in Part I. Both the PINN SPM and P2D surrogate models are exercised for parameter inference and compared to data obtained from a direct numerical solution of the governing equations. A parameter inference study highlights the ability to use these PINNs to calibrate scaling parameters for the cathode Li diffusion and the anode exchange current density. By realizing computational speed-ups of ~2250x for the P2D model, as compared to using standard integrating methods, the PINN surrogates enable rapid state-of-health diagnostics. Finally, in the low-data availability scenario, the testing error was estimated to ~2 mV for the SPM surrogate and ~10 mV for the P2D surrogate which could be mitigated with additional data.

25 ENERGY STORAGE↗

DOE OSTI2024

Rapid Inverse Parameter Inference Using Physics-Informed Neural Network

As Li-ion batteries become more essential in today's economy, tools need to be developed to accurately and rapidly diagnose a battery's internal state-of-health. Using a Li-ion battery's (high-rate) voltage response, it is proposed to determine a battery's internal state through Bayesian calibration. However, Bayesian calibration is notoriously slow and requires thousands of model runs. To accelerate parameter inference using Bayesian calibration, a surrogate model is developed to replace the underlying physics-based Li-ion model. Developing a surrogate model for rapid Bayesian calibration analysis is discussed for both the single particle model (SPM) and the pseudo two-dimensional (P2D) model. Surrogate models are constructed using physics-informed neural networks (PINNs) that encode the influence of internal properties on observed voltage responses. In practice, a neural network can be trained by: 1) using simulation results of the physics-based model (i.e., a data-loss approach); 2) using the residuals of the governing equations themselves (i.e., a physics-loss approach); or 3) using a combination of simulation results and governing equation residuals. In the present work, PINNs are developed using a variety of training losses and neural network architectures. In this analysis, it is shown that a PINN surrogate model can be reliably trained with only physics-informed loss. However, using a coupled data-informed and physics-loss approach produced the most accurate PINNs.

Bayesian calibration↗

DOE OSTI2024

Carbon-Binder Optimization for Lithium-Ion Battery Extreme Fast Charge

Battery performance is strongly correlated with electrode microstructure and weight loading of the electrode components. Among them are the carbon-black and binder additives that enhance effective conductivity and provide mechanical integrity. However, these both reduce effective ionic transport in the electrolyte phase and reduce energy density. Therefore, an optimal additive loading is required to maximize performance, especially for fast charging where ionic transport is essential. Such optimization analysis is however challenging due to the nanoscale imaging limitations that prevent characterizing this additive phase and thus quantifying its impact on performance. Herein, an additive-phase generation algorithm has been developed to remedy this limitation and identify percolation threshold used to define a minimal additive loading. Improved ionic transport coefficients from reducing additive loading has been then quantified through homogenization calculation, macroscale model fitting, and experimental symmetric cell measurement, with good agreement between the methods. Rate capability test demonstrates capacity improvement at fast charge at the beginning of life, from 37% to 55%, respectively for high and low additive loading during 6C CC charging, in agreement with macroscale model, and attributed to a combination of lower cathode impedance, reduced electrode tortuosity and cathode thickness.

carbon-binder additives↗

DOE OSTI2023

Calendar life of lithium metal batteries: Accelerated aging and failure analysis

Lithium metal batteries (LMBs) are prime candidates for next-generation energy storage devices characterized by their remarkable energy storage capabilities. Despite the importance of understating the calendar aging impacts in LMBs, cycle life and calendar life have received inconsistent attention across various research and development stages. For acceptance into an application, especially electric vehicles, batteries are required to have sufficient calendar life to ensure performance over the life of vehicles which often experience extended periods of inactivity. In this study, an in-depth exploration into the calendar aging of LMB (Li/ Li[Ni 0.8 Mn 0.1 Co 0.1 ]O 2 in pouch cell format) is conducted. Specifically, the impact of calendar aging factors, state-of-charge (SOC), temperature, pressure, and operating conditions (open-circuit voltage, OCV and constant voltage, CV) on the surface morphology and thickness change rate in lithium metal anode is investigated. Notably, the results show that storing LMB pouch cells at an OCV, 80 % SOC, 25 °C, and 10 psi external pressure leads to a <1% reduction in capacity over 18 months. The results also suggest routes which can be utilized in the future for accelerated aging without altering the primary degradation modes. The results also show an accelerated calendar aging at high temperatures (45 °C), 100 % SOC, and CV suggesting they may serve as key accelerated aging methods for future studies. During calendar aging, pressure mitigates dendrite growth and extends calendar life. In conclusion, the insights from this study reinforce the viability of LMBs as a compelling solution for addressing increased energy storage needs in real-world applications.

25 ENERGY STORAGE↗

DOE OSTI2023

Localized high-concentration electrolytes get more localized through micelle-like structures

Liquid electrolytes in batteries are typically treated as macroscopically homogeneous ionic transport media despite having a complex chemical composition and atomistic solvation structures, leaving a knowledge gap of the microstructural characteristics. Here, we reveal a unique micelle-like structure in a localized high-concentration electrolyte, in which the solvent acts as a surfactant between an insoluble salt in a diluent. The miscibility of the solvent with the diluent and simultaneous solubility of the salt results in a micelle-like structure with a smeared interface and an increased salt concentration at the centre of the salt–solvent clusters that extends the salt solubility. These intermingling miscibility effects have temperature dependencies, wherein a typical localized high-concentration electrolyte peaks in localized cluster salt concentration near room temperature and is used to form a stable solid–electrolyte interphase on a Li metal anode. Importantly, these findings serve as a guide to predicting a stable ternary phase diagram and connecting the electrolyte microstructure with electrolyte formulation and formation protocols of solid–electrolyte interphases for enhanced battery cyclability.

25 ENERGY STORAGE↗

DOE OSTI2023

Battery state-of-health diagnostics during fast cycling using physics-informed deep-learning

Rapid, in-situ Li-ion battery state-of-health (SOH) quantification is challenging. Li-ion battery aging can vary significantly with chemistry, operating conditions, cycling demands, electrode design, and operation history. As a cell ages, optimal and safe operating conditions need to be adapted to account for battery degradation by tracking critical aging modes such as loss-of-lithium-inventory (LLI), loss-of-active-material (LAM) in either electrode, and/or impedance rise. This manuscript describes a framework for identifying battery aging modes in-operando using fast-rate voltage charge/discharge responses. The framework uses a physically based Li-ion battery model to produce synthetic high-rate responses at aged states. The aging model is calibrated against experimental data from cells with different electrode loadings and cycled under a variety of fast-charging conditions (1 h, 15 min, 10 min, and 7 min charging). The synthetically generated high-rate responses at aged states are then used to train a deep-learning model to identify real cell state-of-health from fast charge/discharge battery voltage responses. The synthetically trained deep-learning model performance is validated by comparing to standard incremental capacity analysis and half-cell measurements. Finally, the framework demonstrates the benefits of using high-rate physics-based models to generate synthetic data for training deep-learning models.

25 ENERGY STORAGE↗

DOE OSTI2023

Methods and systems for diagnosis of failure mechanisms and for prediction of lifetime of metal batteries

Methods for diagnosing failure mechanisms and for predicting lifetime of metal batteries include monitoring rest voltage and Coulombic Efficiency over relatively few cycles to provide profiles that indicate, by the trends thereof, a particular failure mechanism (e.g., electrolyte depletion, loss of metal inventory, increased cell impedance). The methods also include cycling over relatively few cycles an anode-free cell, having the same cathode and electrolyte as the metal battery, but with a current collector instead of the anode. Discharge capacity is monitored and profiled, and a discharge capacity curve is fitted to the discharge capacity profile to discern a capacity retention per cycle. The lifetime of the metal battery is determined using the capacity retention per cycle discerned from the anode-free cell. Related systems include a metal-based battery and an anode-free cell or a battery cell reconfigurable between a metal-based and an anode-free cell.

Li, Bin↗

DOE OSTI2023

Rational designs to enable 10-min fast charging and long cycle life in lithium-ion batteries

A daunting challenge in the design of lithium ion batteries (LiBs) is enabling 10-min extreme fast charging (XFC) while achieving appreciable charge acceptance and cycle life. This desirable outcome requires both a comprehensive understanding of LiB operation and aging behavior at different length scales and careful optimization. Lithium plating has been a critical bottleneck because, at XFC rates, it consumes cyclable lithium causing distinct aging and safety concerns even in moderate-loading LiBs. Here, we propose combining multiple solutions, including materials-to-electrode design-to-charging protocols, that are intended to overcome limitations in lithium-ion transport in the electrolyte phase, thus enabling 10-min XFC in LiBs. Some implemented strategies include cathode chemistry, optimized carbon binder domain in the cathode, dual layer anode design, improved separator and advanced electrolyte. Further, innovative charging protocols in moderately loading (~3 mAh/cm 2 anode/2.7 mAh/cm 2 cathode) single-layer pouch cells are proposed, together with demonstrated 10-min XFC with higher charge acceptance between 87.3 and 92.1% (or 2–2.1 mAh/cm 2 ) for 600 cycles without lithium plating. This methodical study with well-defined cells shows promise in combining multiple solution strategies to enable 10-min XFC, charting a pathway to achieve XFC in higher-loading energy-optimized LiBs.

25 ENERGY STORAGE↗

DOE OSTI2023

Technology Strategy Assessment: Findings from Storage Innovations 2030 Lithium-ion Batteries

This report on accelerating the future of lithium-ion batteries is released as part of the Storage Innovations (SI) 2030 strategic initiative. The objective of SI 2030 is to develop specific and quantifiable research, development, and deployment (RD&D) pathways toward achieving the targets identified in the Long-Duration Storage Energy Earthshot, which seeks to achieve 90% cost reductions for technologies that can provide 10 hours or longer of energy storage within the coming decade. Through SI 2030, the U.S. Department of Energy (DOE) is aiming to understand, analyze, and enable the innovations required to unlock the potential for long-duration applications in the following technologies.

25 ENERGY STORAGE↗

DOE OSTI2023

Quantitative Analysis of Origin of Lithium Inventory Loss and Interface Evolution over Extended Fast Charge Aging in Li Ion Batteries

During the extreme fast charging (XFC) of lithium-ion batteries, lithium inventory loss (LLI) and reaction mechanisms at the anode/electrolyte interface are crucial factors in performance and safety. Determining the causes of LLI and quantifying them remain an essential challenge. We present mechanistic research on the evolution and interactions of aging mechanisms at the anode/electrolyte interface. We used NMC 532 /graphite pouch cells charged at rates of 1, 6, and 9 C up to 1000 cycles for our investigation. The cell components were characterized after cycling using electrochemical measurements, inductively coupled plasma optical emission spectroscopy, 7 Li solid-state nuclear magnetic resonance spectroscopy, and high-performance liquid chromatography/mass spectrometry. The results indicate that cells charged at 1 C exhibit no Li plating, and the increase of SEI thickness is the dominant source of the Li loss. In contrast, Li loss in cells charged at 9 C is related to the formation of the metallic plating layers (42%) the SEI layer (38.1%) and irreversible intercalation into the bulk graphite (19%). XPS analysis suggests that the charging rate has little influence on the evolution of SEI composition. The interactions between competing aging mechanisms were evaluated by a correlation analysis. In conclusion, the quantitative method established in this work provides a comprehensive analytical framework for understanding the synergistic coupling of anodic degradation mechanisms, forecasting SEI failure scenarios, and assessing the XFC lithium-ion battery capacity fade.

25 ENERGY STORAGE↗

DOE OSTI2023

Battery passports for promoting electric vehicle resale and repurposing

Starting in 2026, most batteries sold in the European Union will require a battery passport. This initiative is part of a broader legislative framework, led by the European Commission, to improve sustainable battery materials sourcing practices and enable a circular economy. All batteries having a capacity of greater than 2 kilowatt-hours will be covered under the legislation, effectively encompassing all electric vehicle (EV) batteries and stationary grid storage batteries. At its core, a battery passport defines a minimum, standard set of information that must be reported for every battery made and sold (see Figure 1). Legislative actions have so far focused on labeling standards enabling raw material traceability and recycling, which has received ongoing support from the European Council and Parliament. Furthermore, work remains to finalize the electrochemical performance and durability minimum reporting requirements by 2024, which has been challenging since performance and durability are subject to many different interpretations.

25 ENERGY STORAGE↗

DOE OSTI2023

Battery data integrity and usability: Navigating datasets and equipment limitations for efficient and accurate research into battery aging

A tremendous commitment of resources is needed to acquire, understand and apply battery data in terms of performance and aging behavior. There are many state of performance (SOP) and state of health (SOH) metrics that are useful to guide alignment of batteries to end-use, yet how these metrics are measured or extracted can make the difference between usable, valuable datasets versus data that lacks the necessary integrity to meet baseline confidence levels for SOP/SOH quantification. This work will speak to 1) types of data that support SOP and SOH evaluations on mechanistic terms, 2) measurement conditions needed to assure high data integrity, 3) equipment limitations that can compromise data high fidelity, and 4) the impact of cell polarization on data quality. A common goal in battery research and field use is to work from a data platform that supports economical paths of data capture while minimizing down-time for battery diagnostics. An ideal situation would be to utilize data obtained during normal daily use (“pulses or cycles of convenience”) without stopping the daily duty cycles to perform dedicated SOP/SOH diagnostic routines. However, difficulties arise in trying to make use of daily duty cycle data (denoted as cycle-by-cycle, CBC) that underscores the need for standardization of conditions: temperature and duty cycles can vary over the course of a day and throughout a week, month and year; polarization can develop within an immediate cycle and throughout successive cycles as a hysteresis. If CBC data is envisioned as a data source to determine performance and aging trends, it should be recognized that polarization is a frequent consequence of CBC and thus makes it difficult to separate reversible and irreversible components to metrics such as capacity loss and resistance increase over aging. Since CBC conditions can have a major impact on data usability, we will devote part of this paper to CBC data conditioning and management. Differential analyses will also be discussed as a means to detect changing trends in data quality. Our target cell chemistries will be lithium-ion types NMC/graphite and LMO/LTO.

25 ENERGY STORAGE↗

DOE OSTI2023

A mathematical approach to survey electrochemical impedance spectroscopy for aging in lithium-ion batteries

Electrochemical impedance spectroscopy (EIS) is a valuable technique to detect the health status and aging phenomena in lithium-ion batteries (LiB). Equivalent circuit modeling (ECM) is conventionally used when interpreting EIS data and gaining physical insights into the aging mechanisms. However, performing ECM is resource intensive and expert-level of knowledge is usually required to select suitable models and fitting parameters. This article presents a quick and user-friendly data analysis algorithm as an alternative to ECM by mathematically fitting geometric features in Nyquist plots and obtaining the growth trends of the features. The evolving trends in the Nyquist plots, such as chord lengths of the arcs and interception points, are consistent with the growth of resistance components obtained using ECM with R 2 values from 0.67 to 0.99, and therefore can be used as indicators of battery aging. Our results show that the quick-fitting approach is suitable for analyzing a series of EIS data acquired during battery cycling and identifying the underlying aging mechanisms.

Chen, Bor-Rong↗

DOE OSTI2022

The Evolution of LiNi 0.5 Mn 0.3 Co 0.2 O 2 Particle Damage from Fast Charging in Optimized, Full Li-Ion Cells

Fast charging batteries are critical to the widespread adoption of electric vehicles to compete with refueling times of combustion-based vehicles. In the near term, adapting current commercial battery technologies to perform better under fast charging conditions through engineering optimizations will greatly expedite the process while exploratory fast-charging electrode materials are being pursued. To do so, the degradation modes in optimized Li-ion batteries need to be completely explored to understand fast charging limits while maintaining a high energy density and a long cycle life. While lithium plating on graphite still remains a challenge, cathode degradation also plays a key role in battery performance. In this work, we used nano- and micro-X-ray computed tomography to characterize the mechanical degradation of LiNi 0.5 Mn 0.3 Co 0.2 O 2 (NMC532) in optimized Li-ion batteries cycled at three rates, 1C, 6C, and 9C, and at different stages of cycle life, 225 and 600 cycles. We report despite using a conservative upper voltage cutoff limit aimed to minimize extensive cathode degradation, higher charging rates and increased cycling caused the polycrystalline NMC532 particles to fracture and pulverize, which likely drives cathode capacity fade and contributes to the decrease in overall cell performance.

25 ENERGY STORAGE↗

DOE OSTI2022

Electrification and decarbonization of spent Li-ion batteries purification by using an electrochemical membrane reactor

The expanding electric vehicle market brings with it exponential growth in the use of lithium (Li)-ion batteries (LIB) for which a wave of spent LIB is expected to come within the next 5 to 10 years. Due to the economic and strategic value imbedded within the metals contained in LIB, different recycling technologies, including hydrometallurgy, pyrometallurgy and direct recycling, are under development. Being different from previous hydrometallurgical methods, which may have high chemical consumption and negative environmental impact, an electrochemical membrane reactor is designed and validated for the first time, to electrify and decarbonize the impurity removal process. This reactor electroplates copper (Cu) and electrochemically precipitates aluminum (Al) and iron (Fe) from simulated spent LIB leachates, by consuming only air, water, and electricity, and the impurities are reduced to <1 ppm. The purified leachate maintains 99.5 % of the nickel (Ni), 95.4 % of the cobalt (Co) and 99.14 % of manganese (Mn) from the original leachate solution, and then can be directly applied for cathode precursor synthesis. Additionally, the purification process doesn’t introduce extra impurity, and the reactor restoration process generates valuable by-product hydro sulfate (H2SO4). This electrochemical process can reduce the cost, because of the much less chemical consumption and the valuable by-product generation, and mitigates the waste emissions, because of no extra impurity introduced and no greenhouse gas (GHG) produced. In conclusion, the chemical precipitation method uses significant amount of NaOH, which induced GHG emission during the manufacturing process.

25 ENERGY STORAGE↗

DOE OSTI2022

Battery aging mode identification across NMC compositions and designs using machine learning

A comprehensive understanding of lithium-ion battery (LiB) lifespan is the key to designing durable batteries and optimizing use protocols. Although battery lifetime prediction methods are flourishing, diagnosis of the root causes of aging and degradation have not yet been well developed nor studied for a broad mixture of designs and use cases. Here, we create a machine-learning (ML)-based framework that distinguishes aging modes using multiple electrochemical signatures recorded cycle-by-cycle. The predominant aging behaviors include a combination of loss of active materials in cathode (LAMPE) and a loss of Li inventory (LLI) in Li plating or solid electrolyte interphase (SEI) formation, manifested from 44 batteries representing two cathode chemistries, two electrode loadings, and five charging rates. Here, the aging mode classification accuracy is 86% using features within the first 50 cycles and increases to 88% beyond 225 cycles. The same features can quantify the percentage of end-of-life LAMPE with only 4.3% of error.

25 ENERGY STORAGE↗

DOE OSTI2022

Simultaneous neutron and X-ray tomography for visualization of graphite electrode degradation in fast-charged lithium-ion batteries

Advanced battery characterization using neutron and X-ray-based imaging modalities is crucial to reveal fundamental degradation modes of lithium-ion batteries (LIBs). Taking advantage of the sensitivity of neutrons to some low-Z (Li) and X-rays to high-Z materials (Cu), here we demonstrate the viability of simultaneous neutron and X-ray tomography (NeXT) as a non-destructive imaging platform for ex-situ 3D visualization of graphite electrode degradation following extreme fast charging (XFC). In addition, we underscore the benefits of the simultaneous of NeXT by combining the neutron and X-ray tomography data from the same sample location for material identification and segmentation of one pristine and two XFC-cycled graphite electrodes (9C charge for 450 cycles). Our ex-situ results and methodology development pave the way for the design of NeXT-friendly LIB geometries that will allow operando and/or in-situ 3D visualization of electrode degradation during XFC.

25 ENERGY STORAGE↗