Computational Effort Research Articles

Evaluating an analytical prediction algorithm of positron emitter distributions in patient data for PET monitoring of carbon ion therapy: A simulation study

In vivo treatment monitoring in ion therapy is one of the key issues for improving the treatment quality assurance procedures. Range verification is one of the most relevant and yet complex task used for in vivo treatment monitoring. In carbon ion therapy, positron emission tomography is the most widely used method. This technique exploits the β+-activity of positron emitters created by nuclear interactions between the incoming beam and the irradiated tissue. Currently, high computational efforts and time-consuming Monte Carlo simulation platforms are typically used to predict positron emitter distributions. In order to avoid time-consuming simulations, an extended filtering approach was suggested to analytically predict positron emitter profiles from depth dose distributions in carbon ion therapy. The purpose of this work is to investigate such an analytical prediction model in patient anatomies of varying complexity, highlighting its potential and the need of further improvements, especially in highly heterogeneous anatomies where many air cavities are present in the beam path. The accuracy of range verification showed a mean relative error of ∼3% and a deviation between the simulation and the prediction below 2mm for the three patient cases analysed: a brain case and two head and neck cases. Additional investigations demonstrated the region of applicability of the method for cases of patient data. The analytical method enables range verification in carbon ion therapy by replacing computing-intensive Monte Carlo simulations and thus minimize the PET monitoring burden on the clinical workflow.

Short-term air pollution prediction using graph convolutional neural networks

Pollution is a major concern in the present day, causing multiple illnesses and deaths, specifically in developing countries in Asia and Africa. While it has drawn worldwide attention as governments try to issue laws to meet certain criteria for air pollution levels, pollution concentration forecasting has become a major challenge. Particularly, short term forecasting will help to gain information regarding concentrations of harmful pollutants for the upcoming hours and enable better decision-making with regards to controlling air pollution. In this paper, we investigate spatio-temporal graph-based models to determine the best methods for spatial and temporal analysis of data. The models have the additional capacity to perform multi-variate predictions of correlated data, i.e., predicting multiple pollutant concentrations simultaneously, thus requiring lower computational efforts. A real-world pollution dataset measured over Delhi, India, is used to comparing the proposed models with baselines, which shows the Spatio-Temporal Graph Convolution Neural Network (STGCN) models to be performing better than others. For a better understanding of model architectures with the most effective strategies for spatial and temporal data analysis, three models, namely STGCN-A, STGCN-B, STGCN-C have been developed. The models have been compared with 6 other baselines over multiple forecasting horizons of 1 h, 24 h, and 48 h timesteps using various metrics such as mean absolute error (MAE), root mean square error (RMSE), mean absolute percent error (MAPE). On the PM2.5 dataset of Delhi, STGCN-B achieves a performance of 10.53 MAE, 6.92 RMSE and 25.25 MAPE for a 1 h forecast, while STGCN-C achieves 20.18 MAE, 14.73 RMSE and 55.45 MAPE for a 24 h forecast. In general, both structures achieve similar results, with STGCN-C being better in many cases. They are further analysed through observation-prediction graphs and Taylor diagrams, which give an insight into our findings. The models are additionally validated on a benchmark real-world dataset from California, USA for better understanding of the spatio-temporal relations and model performances on a more stable dataset, where STGCN-C performs best for PM2.5 with 4.30 RMSE, 1.98 MAE, 25.96 MAPE for 1 h predictions for univariate data and 3.63 RMSE, 1.88 MAE and 25.91 MAPE in multivariate forecasting. The developed spatio-temporal graph-based models hold promising applications in urban air quality management, aiding policymakers in implementing targeted interventions to mitigate pollution-related health risks. Furthermore, these models can support public health agencies by providing timely and accurate forecasts of pollutant concentrations, enabling proactive measures to safeguard community well-being. Our study showcases the efficacy of spatio-temporal graph-based models in accurately forecasting air pollutant concentrations, with particular emphasis on short-term predictions. By leveraging multi-variate capabilities, our proposed models demonstrate superior performance compared to baseline approaches across various forecasting horizons.

Reliability estimation for individual predictions in machine learning systems: A model reliability-based approach

The conventional aggregated performance measure (i.e., mean squared error) with respect to the whole dataset would not provide desired safety and quality assurance for each individual prediction made by a machine learning model in risk-sensitive regression problems. In this paper, we propose an informative indicator ℛx to quantify model reliability for individual prediction (MRIP) for the purpose of safeguarding the usage of machine learning (ML) models in mission-critical applications. Specifically, we define the reliability of a ML model with respect to its prediction on each individual input x as the probability of the observed difference between the prediction of ML model and the actual observation falling within a small interval when the input x varies within a small range subject to a preset distance constraint, namely ℛx=Py∗−ŷ∗≤εx∗∈Bx, where y∗ denotes the observed target value for the input x∗,ŷ∗ denotes the model prediction for the input x∗, and x∗ is an input in the neighborhood of x subject to the constraint Bx=x∗x∗−x≤δ. The developed MRIP indicator ℛx provides a direct, objective, quantitative, and general-purpose measure of “reliability” or the probability of success of the ML model for each individual prediction by fully exploiting the local information associated with the input x and ML model. Next, to mitigate the intensive computational effort involved in MRIP estimation, we develop a two-stage ML-based framework to directly learn the relationship between x and its MRIP ℛx, thus enabling to provide the reliability estimate ℛx for any unseen input instantly. Thirdly, we propose an information gain-based approach to help determine a threshold value pertaing to ℛx in support of decision makings on when to accept or abstain from counting on the ML model prediction. Comprehensive computational experiments and quantitative comparisons with existing methods on a broad range of real-world datasets reveal that the developed ML-based framework for MRIP estimation shows a robust performance in improving the reliability estimate of individual prediction, and the MRIP indicator ℛx thus provides an essential layer of safety net when adopting ML models in risk-sensitive environments.

Modeling for sustainable groundwater management: Interdependence and potential complementarity of process-based, data-driven and system dynamics approaches

Groundwater systems are vast natural water reservoirs used to support human water demands and ecosystem services. Various modeling approaches have been developed to help manage these complex highly-dynamic systems. This paper discusses the strengths and limitations of three modeling approaches, namely: process-based, data-driven and system dynamics modeling. For demonstration purposes, the three modeling approaches are applied to the Konya Closed Basin, a large agricultural region with semi-dry climate located in central Turkey. Process-based modeling is grounded in the theory-based representation of the governing processes but is somewhat limited by the computational effort and the difficulty of defining the required input parameters that characterize the heterogeneous aquifer system. Process-based models are shown to be powerful tools for resource management purposes provided climatic and water demand scenarios are accurately defined. Data-driven models are efficient tools for the management of groundwater resources but are highly dependent on the availability of large training data sets encompassing the spectrum of possible system responses. The high efficiency of surrogate modeling approaches makes them ideal tools for incorporation into applications such as real-time decision support systems and digital twin platforms. System dynamics modeling examines the groundwater exploitation problem within a socio-economic context that involves multiple stakeholders and their decision making. It combines groundwater flow models with socio-economics and endogenous decision rules to conduct scenario analysis and support policy development. The analyses and model demonstrations presented in this paper underscore the interconnectedness and complementarity of these three modeling approaches and the need for more integrated use of these modeling approaches for enhanced multi-sectoral management of groundwater systems.

Improvement of Solar Farm Performance based on Photovoltaic Modules Selection

The emissions of greenhouse gases from conventional power plants are currently a significant cause for worry. In China, about 75% of total domestic energy is dependent on coal-fire power, which emits 50% of total SO2 and has a significant impact on the human respiratory system. Therefore, solar power plants are a viable option that can mitigate this problem. Furthermore, the efficiency of solar modules exhibits a progressive upward trend, while their price per watt experiences a corresponding decline, making it a promising source for future energy. This article examines the performance and effectiveness of several photovoltaic (PV) modules in designing solar plants on a certain land area measuring 10000 m2 (100 m * 100 m). The PV plant performance was evaluated by comparing occupation ratio (OR), PV power capacity, net energy production, performance ratio (PR) via PVsyst software, and lastly financial analysis. Consequently, the PV module (PV7), characterized by its high efficiency, low temperature coefficients, and affordable price, result in a significant OR (73.81%), increased installed PV power capacity (1568kW), enhanced net energy output (2269029 kWh/year), improved yearly PR (83.4%), and lastly, the shortest payback period (around two years). Instead of optimizing shadow length in existing research, this paper aims to improve the performance of large-scale solar farm based on PV module selection which results in less computation and structure installation efforts.

Conceptual design of furfural extraction, oxidative upgrading and product recovery: COSMO-RS-based process-level solvent screening

Liquid phase oxidation of furfural using hydrogen peroxide offers a promising route for bio-based C4 furanones and diacids; however, only dilute water-based process designs have been previously suggested that have limited techno-economic potential. In this study, a conceptual process design is presented, where aqueous furfural is extracted using an organic solvent, coupled with peroxide oxidation and product recovery in the presence of the solvent. To address the problem of solvent selection, the COSMO-RS-based solvent screening framework is applied, where quantum mechanics-based thermodynamics are utilized in pinch-based process models. About 2500 solvent candidates were identified as feasible. Focusing on a set of 400 solvent candidates revealed energy consumption values (Qreb,tot/ṁprod recov) between approximately 2 MWh/tonne and 33 MWh/tonne, signifying the potential of the solvent-based process in outperforming the reference aqueous process (49.4 MWh/tonne). The study provides potential solvent candidates and future directions to consider in more costly computational and experimental efforts.

Physics-Informed Neural Networks for Prediction of a Flow-Induced Vibration Cylinder

Abstract Flow-induced vibration (FIV) is a common phenomenon in ocean engineering for subsea structures with circular-shaped cross sections. A large amount of computational resources or experimental efforts are required to predict or measure the complicated motions and flow field surrounding a circular cylinder undergoing vortex-induced vibration (VIV). Physics-informed neural networks (PINNs) are powerful deep learning techniques for solving governing partial differential equations (PDEs) of dynamic systems as an alternative to complex numerical methods. In the present study, a framework is built employing PINNs for solving the Navier–Stokes equations to predict flows past an FIV cylinder using sparsely distributed spatiotemporal data inside the domain. The training process involves minimizing the supervised loss of flow data at these sparse points and the residuals of the governing PDEs. For the training of the PINN model, a moving frame around an FIV cylinder is used to collect the training flow data from two-dimensional direct numerical simulation results at a low Reynolds number. The structural displacements of the cylinder are also implemented in the residuals of the equations of the developed PINN. The performance of the PINN is evaluated by comparing the predicted contours of the surrounding flow velocities with the training data. The hydrodynamic forces prediction is achieved using the PINN-obtained flow field predictions combined with the force partitioning method.

RegMamba: An Improved Mamba for Medical Image Registration

Deformable medical image registration aims to minimize the differences between fixed and moving images to provide comprehensive physiological or structural information for further medical analysis. Traditional learning-based convolutional network approaches usually suffer from the problem of perceptual limitations, and in recent years, the Transformer architecture has gained popularity for its superior long-range relational modeling capabilities, but still faces severe computational challenges in handling high-resolution medical images. Recently, selective state-space models have shown great potential in the vision domain due to their fast inference and efficient modeling. Inspired by this, in this paper, we propose RegMamba, a novel medical image registration architecture that combines convolutional and state-space models (SSMs), designed to efficiently capture complex correspondence in registration while maintaining efficient computational effort. Firstly our model introduces Mamba to efficiently remotely model and process potential dependencies of the data to capture large deformations. At the same time, we use a scaled convolutional layer in Mamba to alleviate the problem of spatial information loss in 3D data flattening processing in Mamba. Then, a deformable convolutional residual module (DCRM) is proposed to adaptively adjust the sampling position and process deformations to capture more flexible spatial features while learning fine-grained features of different anatomical structures to construct local correspondences and improve model perception. We demonstrate the advanced registration performance of our method on the LPBA40 and IXI public datasets.

High-precision ultra-fast minimum cut approximation through aggregated hash of cut collection

Due to the wide application of s-t minimum cut (min-cut) in various scenarios, many acceleration algorithms have been proposed to solve it. However, the query times of the acceleration algorithms currently available are still high in large-scale graphs, rendering them useless in frequently solving scenarios. We re-examine the min-cut problem from a novel perspective of cut collection hash. By extracting aggregated hashes of mapped cut collections in one-dimensional space, a Monte Carlo-like method is used to quickly compare them and estimate the minimum cut between any two nodes with low computational effort and high accuracy. After the graph is preprocessed using a few hundred depth-first traversals, the time complexity of the min-cut solution can be logarithmic in terms of the average degree and capacity of the graph. Experiments on large-scale graphs show that compared to the fastest exact algorithm, the proposed algorithm can increase the speed of the min-cut solution by up to seven orders of magnitude, when only a few mathematical comparisons per pair are needed to obtain exact min-cut values of no less than 99.9% node pairs.

A probabilistic full waveform inversion of surface waves

AbstractOver the past decades, surface wave methods have been routinely employed to retrieve the physical characteristics of the first tens of meters of the subsurface, particularly the shear wave velocity profiles. Traditional methods rely on the application of the multichannel analysis of surface waves to invert the fundamental and higher modes of Rayleigh waves. However, the limitations affecting this approach, such as the 1D model assumption and the high degree of subjectivity when extracting the dispersion curve, motivate us to apply the elastic full‐waveform inversion, which, despite its higher computational cost, enables leveraging the complete information embedded in the recorded seismograms. Standard approaches solve the full‐waveform inversion using gradient‐based algorithms minimizing an error function, commonly measuring the misfit between observed and predicted waveforms. However, these deterministic approaches lack proper uncertainty quantification and are susceptible to get trapped in some local minima of the error function. An alternative lies in a probabilistic framework, but, in this case, we need to deal with the huge computational effort characterizing the Bayesian approach when applied to non‐linear problems associated with expensive forward modelling and large model spaces. In this work, we present a gradient‐based Markov chain Monte Carlo full‐waveform inversion where we accelerate the sampling of the posterior distribution by compressing data and model spaces through the discrete cosine transform. Additionally, a proposal is defined as a local, Gaussian approximation of the target density, constructed using the local Hessian and gradient information of the log posterior. We first validate our method through a synthetic test where the velocity model features lateral and vertical velocity variations. Then we invert a real dataset from the InterPACIFIC project. The obtained results prove the efficiency of our proposed algorithm, which demonstrates to be robust against cycle‐skipping issues and able to provide reasonable uncertainty evaluations with an affordable computational cost.

Data-driven computational mechanics: comparison of model-free and model-based methods in constitutive modeling

In computational homogenization approaches, data-driven methods entail advantages due to their ability to capture complex behavior without assuming a specific material model. Within this domain, constitutive model-based and model-free data-driven methods are distinguished. The former employ artificial neural networks as models to approximate a constitutive relation, whereas the latter directly incorporate stress–strain data in the analysis. Neural network-based constitutive descriptions are one of the most widely used data-driven approaches in computational mechanics. In contrast, distance-minimizing data-driven computational mechanics enables substituting the material modeling step entirely by iteratively obtaining a physically consistent solution close to the material behavior represented by the data. The maximum entropy data-driven solver is a generalization of this method, providing increased robustness concerning outliers in the underlying data set. Additionally, a tensor voting enhancement based on incorporating locally linear tangent spaces enables interpolating in regions of sparse sampling. In this contribution, a comparison of neural network-based constitutive models and data-driven computational mechanics is made. General differences between machine learning, distance minimizing, and entropy maximizing data-driven methods are explored. These include the pre-processing of data and the required computational effort for optimization as well as evaluation. Numerical examples with synthetically generated datasets obtained by numerical material tests are employed to demonstrate the capabilities of the investigated methods. An anisotropic nonlinear elastic constitutive law is chosen for the investigation. The resulting constitutive representations are then applied in structural simulations. Thereby, differences in the solution procedure as well as use-case accuracy of the methods are investigated.

An object-oriented implementation of a recursive “quantum network” solver and its application to district heating networks

With the increasing importance of energy efficiency and sustainability, the demand for high-performance district heating networks is also on the rise. As traditional engineering methods were often no longer sufficient, several packages for numerical simulation have evolved. Many of them offer high accuracy at the cost of considerable manual preparation and computational effort. However, there is still no user-friendly software that is equally suitable for simplified studies in the early project phase, for the rapid optimisation of multiple concepts and for subsequent detailed planning and dimensioning. For this reason and in cooperation with some companies from the energy sector, we had conceived a novel, highly flexible modular approach, the so-called “quantum networks”, where all parts of the district heating network are appropriately abstracted into quantum elements. Now we present our recent implementation of this model in an object-oriented C++ library. Starting from a generalised base class and making use of the concepts of inheritance and polymorphism, the idea of different levels of detail for the same element type is directly realised and can always be further refined. In addition, as all elements are derived from the same base class, they all share the same outer appearance and can thus be easily combined or interchanged. Based on these prerequisites, a dedicated thermo-hydraulic solver has then been developed. Thanks to its recursive design requiring neither outer iterations nor matrix inversions, it proves to be extremely fast and is therefore suitable for rapid design or optimisation studies in which a vast number of configurations has to be computed. To conclude this part of the work, the developed C++ library was benchmarked on two use cases covering a full year of operation that can now be computed in approximately one second of runtime.

SPTrack: Spectral Similarity Prompt Learning for Hyperspectral Object Tracking

Compared to hyperspectral trackers that adopt the “pre-training then fine-tuning” training paradigm, those using the “pre-training then prompt-tuning” training paradigm can inherit the expressive capabilities of the pre-trained model with fewer training parameters. Existing hyperspectral trackers utilizing prompt learning lack an adequate prompt template design, thus failing to bridge the domain gap between hyperspectral data and pre-trained models. Consequently, their tracking performance suffers. Additionally, these networks have a poor generalization ability and require re-training for the different spectral bands of hyperspectral data, leading to the inefficient use of computational resources. In order to address the aforementioned problems, we propose a spectral similarity prompt learning approach for hyperspectral object tracking (SPTrack). First, we introduce a spectral matching map based on spectral similarity, which converts 3D hyperspectral data with different spectral bands into single-channel hotmaps, thus enabling cross-spectral domain generalization. Then, we design a channel and position attention-based feature complementary prompter to learn blended prompts from spectral matching maps and three-channel images. Extensive experiments are conducted on the HOT2023 and IMEC25 data sets, and SPTrack is found to achieve state-of-the-art performance with minimal computational effort. Additionally, we verify the cross-spectral domain generalization ability of SPTrack on the HOT2023 data set, which includes data from three spectral bands.

The discontinuous strain method: accurately representing fatigue and failure

AbstractFatigue simulation requires accurate modeling of unloading and reloading. However, classical ductile damage models treat deformations after complete failure as irrecoverable—which leads to unphysical behavior during unloading. This unphysical behavior stems from the continued accumulation of plastic strains after failure, resulting in an incorrect stress state at crack closure. As a remedy, we introduce a discontinuous strain in the additive elasto-plastic strain decomposition, which absorbs the excess strain after failure. This allows representing pre- and post-cracking regimes in a fully continuous setting, wherein the transition from the elasto-plastic response to cracking can be triggered at any arbitrary stage in a completely smooth manner. Moreover, the presented methodology does not exhibit the spurious energy release observed in hybrid approaches. In addition, our approach guarantees mesh-independent results by relying on a characteristic length scale—based on the discretization’s resolution. We name this new methodology the discontinuous strain method. The proposed approach requires only minor modifications of conventional plastic-damage routines. To convey the method in a didactic manner, the algorithmic modifications are first discussed for one- and subsequently for two-/three-dimensional implementations. Using a simple ductile constitutive model, the discontinuous strain method is validated against established two-dimensional benchmarks. The method is, however, independent of the employed constitutive model. Elastic, plastic, and damage models may thus be chosen arbitrarily. Furthermore, computational efforts associated with the method are minimal, rendering it advantageous for accurately representing low-cycle fatigue but potentially also for other scenarios requiring a discontinuity representation within a plastic-damage framework. An open-source implementation is provided to make the proposed method accessible.

Advancing first-principles dielectric property prediction of complex microwave materials: an elemental-unit decomposition approach

Tungsten-bronze-type material Ba6-3xRE8+2xTi18O54, (RE = rare earth elements) is an important microwave dielectric that has shown great promises for future miniaturization of microwave devices because of its high dielectric constant, low loss, and tunabilities, and there is still much room for improvement. With their proven predictive power, first-principles calculations may greatly help accelerate materials optimization by reducing or eliminating the expensive and time-consuming experimental trial-and-error process. However, microwave dielectrics such as the tungsten-bronze-type materials are rather complex systems with unit cells containing hundreds or thousands of atoms, making ab initio calculations prohibitively expensive. In this work, we propose an elemental-unit decomposition (EUD) technique that can drastically reduce the computational effort of predicting the properties of complex microwave dielectrics and demonstrate its accuracy and efficiency. Our approach facilitates first-principles prediction and design of complex microwave dielectric materials that would otherwise be extremely difficult.

A deep learning-guided automated workflow in LipidOz for detailed characterization of fungal fatty acid unsaturation by ozonolysis.

Understanding fungal lipid biology and metabolism is critical for antifungal target discovery as lipids play central roles in cellular processes. Nuances in lipid structural differences can significantly impact their functions, making it necessary to characterize lipids in detail to understand their roles in these complex systems. In particular, lipid double bond (DB) locations are an important component of lipid structure that can only be determined using a few specialized analytical techniques. Ozone-induced dissociation mass spectrometry (OzID-MS) is one such technique that uses ozone to break lipid DBs, producing pairs of characteristic fragments that allow the determination of DB positions. In this work, we apply OzID-MS and LipidOz software to analyze the complex lipids of Saccharomyces cerevisiae yeast strains transformed with different fatty acid desaturases from Histoplasma capsulatum to determine the specific unsaturated lipids produced. The automated data analysis in LipidOz made the determination of DB positions from this large dataset more practical, but manual verification for all targets was still time-consuming. The DL model reduces manual involvement in data analysis, but since it was trained using mammalian lipid extracts, the prediction accuracy on yeast-derived data was reduced. We addressed both shortcomings by retraining the DL model to act as a pre-filter to prioritize targets for automated analysis, providing confident manually verified results but requiring less computational time and manual effort. Our workflow resulted in the determination of detailed DB positions and enzymatic specificity.

On an efficient global/local stochastic methodology for accurate stress analysis, failure prediction and damage tolerance of laminated composites

The quantification of uncertainties in the mechanical response of composite structures can be a computationally demanding task. This is due both to the number of uncertain parameters in a real study case and the complexity of the model to be analysed. In this paper, an efficient global/local approach to estimate the uncertainties of the quantities of interest in specific regions of interest with limited computational effort is proposed. This is achieved by refining only locally the model taking advantage of Refined Structural Theories. At the same time, since the variance of the uncertain parameters is usually relatively small, the stochastic analysis is dealt with a sensitivity study carried out both in the global and in the local model. In this way, it is possible to assess the influence of global and local uncertain parameters in the same submodeling analysis. The methodology presented is applied to several study cases of interest. The results focus on obtaining probabilistic distributions of the stress field that can be later used in failure criteria to evaluate the subsequent distribution of the failure index. Furthermore, a damage tolerance study case is investigated, showing good correlation with the reference Monte Carlo simulations.

Simplified Physics-Based Battery Model for Stationary Energy-Storage Applications

To perform safe battery operation and avoid unnecessary degradation, battery models are needed. How complex, or close to reality the model should be, depends on the purpose of the model and the computational power available. The possibility for accurate parameterization before model implementation is another crucial aspect for the validity. Physics-based models have the advantage to solve for internal battery states, which includes important information to avoid accelerated degradation and to evaluate state of health. But the parameterization effort can be extensive and the computational cost too high to solve for in real time. Identifying the most essential processes and if possible, simplify the model is therefore motivated.Parameter sensitivity is a powerful tool to investigate how much certain parameters and the related processes effect the output signal, i.e. battery voltage. Under normal operating conditions the cells are rarely completely discharged, and as a result the diffusion processes show low sensitivity during operation [1]. In this work we therefore explore the utilization of a physics-based model solving for the electrolyte dynamics, rather than the solid phase diffusion commonly applied in the so-called single particle model. By neglecting the diffusion processes, the second dimension in the pseudo-two-dimensional (P2D) model can be ignored and the computational complexity simplified. The parameterization strategy and validation of the model is further discussed.We evaluate our model by comparing simulations with experimental data from batteries subjected to stationary energy storage applications. The cells are commercial 18650-type NMC/Graphite cells with 2.6 Ah capacity. Different service cycles have different needs in terms of current and voltage, see examples in Figure 1. Capturing these behaviours with a physics-based model might pose different challenges and will further affect the degradation of the cells [2]. A wide range of models exist in literature to capture degradation mechanisms during battery operation [3]. To keep the computational and parameterization effort as low as possible, our methodology is not to include more physics, but instead update the parameters to the processes already included in the model [4]. We explore this strategy comparing the behaviour from different types of services, as well as batteries where the application changes, to see the model response and analyse the battery state of health, see Figure 2. Electrochemical techniques such as differential voltage analysis and electrochemical impedance spectroscopy is included to support our conclusions.Our results highlight the value of application considerations when designing battery models, rather than only focusing on physical extreme points. We present an alternative physics-based model with the ability to capture crucial degradation phenomena and validate it against operational data. We believe our findings can be of value for smarter battery operating strategies and a greater understanding of application dependant lithium-ion battery degradation.Figure 1. Some of the application duty cycles part of the experimental study. a) Current profile peak shaving b) Current profile frequency regulation c) Voltage response peak shaving d) Voltage response frequency regulation.Figure 2. Degradation trends from the experimental study. Purple cell initially performing PS following application change to FR_high cycling a) Capacity evolution of the cycled cells b) tortuosity parameter evolution obtained from model parameter fitting c) Measured impedance response during cell reference performance tests d) Calculated charge transfer resistance from measured impedance data.

Deep Learning Based 3D Resolution Recovery for Breakthrough Imaging of Electrochemical Energy Devices

Many energy materials can only be understood within the context of the (often sealed) devices they make up. Batteries and fuel cells contain intricate mixtures of materials with features and arrangements that span across many orders of magnitude in size and operate under non-ambient conditions. In these devices, the microstructures of the constituent materials and how they are arranged with respect to each other is directly linked to the device performance and degradation behaviors. Complicating matters, opening the devices for inspection and analysis can irreversibly alter the very microstructures in question and can pose safety hazards to researchers if not done correctly. Furthermore, critical material and device properties such as tortuosity or ion transport pathways can only be accurately quantified in 3D. As such, observing microstructures and their evolution in 3D and in their native state is essential for developing a robust understanding of how to engineer new battery and fuel cell materials and architectures with improved performance.X-ray microscopy (XRM) provides a unique method for observing these microstructures in 3D at high resolutions without needing to open them. However, microscopy techniques like XRM make tradeoffs between FOV and resolution, which can pose challenges for strongly multiscale systems like batteries and fuel cells. For example, if images are acquired at high enough resolution to observe critical features like electrode particle sizes in lithium-ion batteries or gas diffusion layer fiber weaves, catalyst distributions, and microporous layer cracks in polymer electrolyte fuel cells, the field of view of that image is restricted to a small local region of the sample. This means that broader defect distributions or larger scale inhomogeneities will be missed. Additionally, for quantitative analysis applications, this means that statistical relevance will be sacrificed because of the limited analysis volume at the appropriate resolution. And for computational modelling efforts, device level phenomena will not be captured well due to the restricted field of view.Advances in modern AI-based resolution recovery methods combined with the non-destructive multiscale imaging capabilities of X-ray microscopy have allowed researchers to overcome this dependency between resolution and field of view to produce 3D images with both high resolution and large field of view. We present here examples of applying these methods to applications in lithium-ion battery and polymer electrolyte fuel cell research [1].[1] Wang, Y.D., Meyer, Q., Tang, K. et al. Large-scale physically accurate modelling of real proton exchange membrane fuel cell with deep learning. Nat Commun 14, 745 (2023).

Investigation of Cu-Oxidation and Its Roles in Unassisted CO2 Reduction and Electrochemical Oxygen Evolution

Sunlight-driven photocatalytic CO2 reduction to CO under unassisted (unbiased) conditions was recently demonstrated using heterostructured catalysts that combine p-type GaN with plasmonic Au nanoparticles and Cu nanoparticle co-catalysts (p-GaN/Al2O3/Au/Cu) [1]. However, the exact oxidation state of Cu and the hole transfer directionality between Au and Cu remain under debate. Here, we combine computational and experimental efforts to investigate the different oxidation states of Cu under unassisted photocatalytic operating conditions. Experimentally we found that the Cu nanoparticles are primarily composed of Cu2O and CuO, CuCO3.Cu(OH)2 species, with no detectable metallic Cu present. The calculated bulk thermodynamics confirmed the high stability of CuCO3.Cu(OH)2 which is also validated by its largest contribution in the Pourbaix diagram, consistent with experimental observations. To gain insights into the origin of charge transfer, we simulate material interfaces namely, Au/CuO, Au/Cu2O, and Au/CuCO3.Cu(OH)2. Our theoretical analysis supports experimentally observed light-driven hole transfer from Au to Cu.More recently significant progress has been made in understanding copper oxide (CuO) behavior at oxidizing electrochemical conditions (> 1.5 V) [2]. In our study, we show CuO transforms into the OER active phase CuOOH, as evident in the bulk Pourbaix diagram presented here. Our calculations show a novel square planar, non-magnetic CuOOH phase. This discovery is supported by calculated Raman spectroscopic analysis, revealing distinct features corresponding to Cu3+ species at 567 and 585 cm-1, closely matching experimental peaks at 587 cm-1. These findings provide insights into the electrochemical properties of copper oxide phases and their relevance towards electrocatalytic OER. Acknowledgments:This material is based on work performed by the Liquid Sunlight Alliance, which is supported by the U.S. Department of Energy, Office of Science, Office of Basic Energy Sciences, Fuels from Sunlight Hub under Award Number DE-SC0021266. Reference:[1] H. A. Atwater et al. ACS Energy Lett, 6, 1849 (2021)[2] M J. Janik et al. ACS Appl. Mater. Interfaces, 15, 27878 (2023)