Accelerating materials research by integrating automation with artificial intelligence is increasingly recognized as a grand scientific challenge to discover and develop materials for emerging and ...future technologies. While the solid state materials science community has demonstrated a broad range of high throughput methods and effectively leveraged computational techniques to accelerate individual research tasks, revolutionary acceleration of materials discovery has yet to be fully realized. This perspective review presents a framework and ontology to outline a materials experiment lifecycle and visualize materials discovery workflows, providing a context for mapping the realized levels of automation and the next generation of autonomous loops in terms of scientific and automation complexity. Expanding autonomous loops to encompass larger portions of complex workflows will require integration of a range of experimental techniques as well as automation of expert decisions, including subtle reasoning about data quality, responses to unexpected data, and model design. Recent demonstrations of workflows that integrate multiple techniques and include autonomous loops, combined with emerging advancements in artificial intelligence and high throughput experimentation, signal the imminence of a revolution in materials discovery.
Integrating automation with artificial intelligence will enable scientists to spend more time identifying important problems and communicating critical insights, accelerating discovery and development of materials for emerging and future technologies.
The discovery and optimization of new materials for energy storage are essential for a sustainable future. High-throughput experimentation (HTE) using a scanning droplet cell (SDC) is suitable for ...the rapid screening of prospective material candidates and effective variation of investigated parameters over a millimeter-scale area. Herein, we explore the transition and challenges for SDC electrochemistry from aqueous toward aprotic electrolytes and address pitfalls related to reproducibility in such high-throughput systems. Specifically, we explore whether reproducibilities comparable to those for millimeter half-cells are achievable on the millimeter half-cell level than for full cells. To study reproducibility in half-cells as a first screening step, this study explores the selection of appropriate cell components, such as reference electrodes (REs) and the use of masking techniques for working electrodes (WEs) to achieve consistent electrochemically active areas. Experimental results on a Li–Au model anode system show that SDC, coupled with a masking approach and subsequent optical microscopy, can mitigate issues related to electrolyte leakage and yield good reproducibility. The proposed methodologies and insights contribute to the advancement of high-throughput battery research, enabling the discovery and optimization of future battery materials with improved efficiency and efficacy.
As the materials science community seeks to capitalize on recent advancements in computer science, the sparsity of well-labelled experimental data and limited throughput by which it can be generated ...have inhibited deployment of machine learning algorithms to date. Several successful examples in computational chemistry have inspired further adoption of machine learning algorithms, and in the present work we present autoencoding algorithms for measured optical properties of metal oxides, which can serve as an exemplar for the breadth and depth of data required for modern algorithms to learn the underlying structure of experimental materials science data. Our set of 178 994 distinct materials samples spans 78 distinct composition spaces, includes 45 elements, and contains more than 80 000 unique quinary oxide and 67 000 unique quaternary oxide compositions, making it the largest and most diverse experimental materials set utilized in machine learning studies. The extensive dataset enabled training and validation of 3 distinct models for mapping between sample images and absorption spectra, including a conditional variational autoencoder that generates images of hypothetical materials with tailored absorption properties. The absorption patterns auto-generated from sample images capture the salient features of ground truth spectra, and band gap energies extracted from these auto-generated patterns are quite accurate with a mean absolute error of 180 meV, which is the approximate uncertainty from traditional extraction of the band gap energy from measurements of the full transmission and reflection spectra. Optical properties of materials are not only ubiquitous in materials applications but also emblematic of the confluence of underlying physical phenomena yielding the type of complex data relationships that merit and benefit from neural network-type modelling.
Sequential learning (SL) strategies,
i.e.
iteratively updating a machine learning model to guide experiments, have been proposed to significantly accelerate materials discovery and research. ...Applications on computational datasets and a handful of optimization experiments have demonstrated the promise of SL, motivating a quantitative evaluation of its ability to accelerate materials discovery, specifically in the case of physical experiments. The benchmarking effort in the present work quantifies the performance of SL algorithms with respect to a breadth of research goals: discovery of any "good" material, discovery of all "good" materials, and discovery of a model that accurately predicts the performance of new materials. To benchmark the effectiveness of different machine learning models against these goals, we use datasets in which the performance of all materials in the search space is known from high-throughput synthesis and electrochemistry experiments. Each dataset contains all pseudo-quaternary metal oxide combinations from a set of six elements (chemical space), the performance metric chosen is the electrocatalytic activity (overpotential) for the oxygen evolution reaction (OER). A diverse set of SL schemes is tested on four chemical spaces, each containing 2121 catalysts. The presented work suggests that research can be accelerated by up to a factor of 20 compared to random acquisition in specific scenarios. The results also show that certain choices of SL models are ill-suited for a given research goal resulting in substantial deceleration compared to random acquisition methods. The results provide quantitative guidance on how to tune an SL strategy for a given research goal and demonstrate the need for a new generation of materials-aware SL algorithms to further accelerate materials discovery.
Benchmarking metrics for materials discovery
via
sequential learning are presented, to assess the efficacy of existing algorithms and to be scientific in our assessment of accelerated science.
Multimetallic nanoclusters (MMNCs) offer unique and tailorable surface chemistries that hold great potential for numerous catalytic applications. The efficient exploration of this vast chemical space ...necessitates an accelerated discovery pipeline that supersedes traditional “trial-and-error” experimentation while guaranteeing uniform microstructures despite compositional complexity. Herein, we report the high-throughput synthesis of an extensive series of ultrafine and homogeneous alloy MMNCs, achieved by 1) a flexible compositional design by formulation in the precursor solution phase and 2) the ultrafast synthesis of alloy MMNCs using thermal shock heating (i.e., ∼1,650 K, ∼500 ms). This approach is remarkably facile and easily accessible compared to conventional vapor-phase deposition, and the particle size and structural uniformity enable comparative studies across compositionally different MMNCs. Rapid electrochemical screening is demonstrated by using a scanning droplet cell, enabling us to discover two promising electrocatalysts, which we subsequently validated using a rotating disk setup. This demonstrated high-throughput material discovery pipeline presents a paradigm for facile and accelerated exploration of MMNCs for a broad range of applications.
BATTERY 2030+ targets the development of a chemistry neutral platform for accelerating the development of new sustainable high‐performance batteries. Here, a description is given of how the ...AI‐assisted toolkits and methodologies developed in BATTERY 2030+ can be transferred and applied to representative examples of future battery chemistries, materials, and concepts. This perspective highlights some of the main scientific and technological challenges facing emerging low‐technology readiness level (TRL) battery chemistries and concepts, and specifically how the AI‐assisted toolkit developed within BIG‐MAP and other BATTERY 2030+ projects can be applied to resolve these. The methodological perspectives and challenges in areas like predictive long time‐ and length‐scale simulations of multi‐species systems, dynamic processes at battery interfaces, deep learned multi‐scaling and explainable AI, as well as AI‐assisted materials characterization, self‐driving labs, closed‐loop optimization, and AI for advanced sensing and self‐healing are introduced. A description is given of tools and modules can be transferred to be applied to a select set of emerging low‐TRL battery chemistries and concepts covering multivalent anodes, metal‐sulfur/oxygen systems, non‐crystalline, nano‐structured and disordered systems, organic battery materials, and bulk vs. interface‐limited batteries.
The large‐scale European research initiative BATTERY 2030+ strives to create a modular materials acceleration platform and an AI‐assisted toolkit, which will facilitate accelerated closed‐loop discovery of new battery concepts, materials, and interfaces. Here, an outline is presented of how the developed AI‐assisted tools and procedures can be applied to resolve some of the main challenges for future, low‐technology readiness level battery concepts and materials.
Amorphous Al2O3 film that naturally exists on any Al substrate is a critical bottleneck for the cyclic performance of metallic Al in rechargeable Al batteries. The so‐called electron/ion insulator Al ...oxide slows down the anode's activation and hinders Al plating/stripping. The Al2O3 film induces different surface properties (roughness and microstructure) on the metal. Al foils present two optically different sides (shiny and non‐shiny), but their surface properties and influence on plating and stripping have not been studied so far. Compared to the shiny side, the non‐shiny one has a higher (~28 %) surface roughness, and its greater concentration of active sites (for Al plating and stripping) yields higher current densities. Immersion pretreatments in Ionic‐Liquid/AlCl3‐based electrolyte with various durations modify the surface properties of each side, forming an electrode‐electrolyte interphase layer rich in Al, Cl, and N. The created interphase layer provides more tunneling paths for better Al diffusion upon plating and stripping. After 500 cycles, dendritic Al deposition, generated active sites, and the continuous removal of the Al metal and oxide cause accelerated local corrosion and electrode pulverization. We highlight the mechanical surface properties of cycled Al foil, considering the role of immersion pretreatment and the differences between the two sides.
Al2O3 film on Al substrates affects anode activation and hinders Al plating/stripping. Al foils have distinct sides with different surface properties. Pretreatment with EMImC/AlCl3 (1 : 1.5) ILE forms an electrode‐ electrolyte interphase layer rich in Al, Cl, and N, improving Al diffusion. This study revealed that after 500 cycles, dendritic Al deposition and electrode pulverization occur. This study highlights the mechanical surface properties of cycled Al foil and the impact of immersion pretreatment on both sides.
Electrocatalysis of the oxygen evolution reaction is central to several energy technologies including electrolyzers, solar fuel generators, and air-breathing batteries. Strong acid electrolytes are ...desirable for many implementations of these technologies, although the deployment of such device designs is often hampered by the lack of non-precious-metal oxygen evolution electrocatalysts, with Ir-based oxides comprising the only known catalysts that exhibit stable activity at low overpotential. During our exploration of the Mn–Sb–O system for precious-metal-free electrocatalysts, we discovered that Mn can be incorporated into the rutile oxide structure at much higher concentrations than previously known, and that these Mn-rich rutile alloys exhibit great catalytic activity with current densities exceeding 50 mA cm–2 at 0.58 V overpotential and catalysis onset at 0.3 V overpotential. While this activity does not surpass that of IrO2, Pourbaix analysis reveals that the Mn–Sb rutile oxide alloys have the same or better thermodynamic stability under operational conditions. By combining combinatorial composition, structure, and activity mapping with synchrotron X-ray absorption measurements and first-principles materials chemistry calculations, we provide a comprehensive understanding of these oxide alloys and identify the critical role of Sb in stabilizing the trivalent Mn octahedra that have been shown to be effective oxygen evolution reaction (OER) catalysts.
Manual cell assembly confounds with research digitalization and reproducibility. Both are however needed for data-driven optimization of cell chemistries and charging protocols. Therefore, we present ...herein an automatic battery assembly system (AutoBASS) that is capable of assembling batches of up to 64 CR2023 cells. AutoBASS allows us to acquire large datasets on in-house developed chemistries and is herein demonstrated with LNO and Si@Graphite electrodes with a focus on formation and manufacturing data. The large dataset enables us to gain insights into the formation process through d Q /d V analysis and assess cell to cell variability. Exact robotic electrode placement provides a baseline for laboratory-scale manufacturing and reproducibility towards the accelerated translation of findings from the laboratory to the pilot plant scale.
Electrolyte additives in liquid electrolyte batteries can trigger the formation of a protective solid electrolyte interphase (SEI) at the electrodes
e.g.
to suppress side reactions at the electrodes. ...Studies of varying amounts of additives have been done over the last few years, providing a comprehensive understanding of the impact of the electrolyte formulation on the lifetime of the cells. However, these studies mostly focused on the variation of the mass fraction of additive in the electrolyte while disregarding the ratio (
r
add
) of the additive's amount of substance (
n
add
) to the electrode area (
A
electrode
). Herein we utilize our accurate automatic battery assembly system (AUTOBASS) to vary electrode area and amount of substance of the additive. Our data provides evidence that reporting the mass ratios of electrolyte components is insufficient and the amount of substance of additive relative to the electrodes' area should be reported. Herein, the two most utilized additives, namely fluoroethylene carbonate (FEC) and vinylene carbonate (VC) were studied. Each additive was varied from 0.1 wt-%-3.0 wt-% for VC, and 5 wt-%-15 wt-% for FEC for two electrode loadings of 1 mA h cm
−2
and 3 mA h cm
−2
. To help the community to find better descriptors, such as the proposed
r
add
, we publish the dataset alongside this manuscript. The active electrode placement correction reduces the failure rate of our automatically assembled cells to 3%.
Electrolyte additives in liquid electrolytes can trigger the synthesis of a protective solid electrolyte interphase (SEI) at the electrodes. This robot can taylor such synthesis through combinatorial electrolyte formulation.