Accounts of materials research最新文献

The data-driven paradigm, represented by the famous machine learning paradigm, is revolutionizing the way materials are discovered. The inductive nature of the data-driven approach gives it great speed of prediction but also brings with it a heavy reliance on material data. However, unlike its success with text and images, which are supported by big data, materials data tend to be small data. Building a large database of materials is a good solution but not a permanent one. The cost of materials data is much higher than that of text or images, and the size of the materials database at this stage is far from sufficient. We will continue to face a shortage of materials data for a long time to come, making small data approaches necessary for machine learning based materials discovery.

In this Account, we focus on small data strategies developed over the past few years and the scenarios in which they are used. In the first part, we discuss two general strategies, active learning and transfer learning, which are ways of adding new data efficiently and using existing data, respectively. The key to active learning is the sampling strategy, which determines the speed of convergence and the predictive range of the machine learning model. For transfer learning, adversarial training is introduced to extend the scope of this strategy, allowing for knowledge transfer across materials and properties. We also discuss other small data approaches for special cases, such as material search with zero initial data and model training on multisource experimental data. In the second part, we focus on the construction of material descriptors and reduction of their dimensionality. We have developed a crystal-graph-based descriptor specifically for two-dimensional materials. It can encode both structural and atomic information and also has a flexible multilayer format for different target properties. Since the dimensionality of the material descriptor is limited by the amount of data, specially designed dimensionality reduction strategies are also discussed. In the third part, we discuss model interpretability. Several examples are given to illustrate how model-based and data-based interpretation strategies can be used to help us understand the machine learning model and its prediction results.

The Account concludes with our perspectives on the latest developments in generative AI (in particular, large language model and diffusion model) and explainable AI, which could be powerful tools in the future of machine learning assisted material discovery.

Organic battery electrode materials are key enablers of different postlithium cell chemistries. As a p-type compound with up to two reversible redox processes at relatively high potentials of 3.5 and 4.1 V vs. Li/Li⁺, phenothiazine is an excellently suited redox-active group. It can easily be functionalized and incorporated into polymeric structures, a prerequisite to obtain insolubility in liquid battery electrolytes. Phenothiazine tends to exhibit π-interactions (π*−π*-interactions) to stabilize its radical cationic form, which can increase the stability of the oxidized form but can also strongly influence its cycling performance as a battery electrode material. In recent years, we investigated a broad range of phenothiazine-based polymers as battery electrode materials, providing insight into the effect of π-interactions on battery performance, leading to design principles for highly functional phenothiazine-based polymers, and enabling the investigation of full cells. We observed that π-interactions are particularly expressed in “mono”-oxidized forms of poly(3-vinyl-N-methylphenothiazine) (PVMPT) and are enabled in the battery electrode due to the solubility of oxidized PVMPT in many carbonate-based liquid electrolytes. PVMPT dissolves during charge and is redeposited during discharge as a stable film on the positive electrode, however, still retaining half of its charge. This diminishes its available specific capacity to half of the theoretical value. We followed three different strategies to mitigate dissolution and inhibit the formation of π-interactions in order to access the full specific capacity for the one-electron process: Adjusting the electrolyte composition (type and ratio of cyclic vs. linear carbonate), encapsulating PVMPT in highly porous conductive carbons or cross-linking the polymer to X-PVMPT. All three strategies are excellently suited to pursue full-cell concepts using PVMPT or X-PVMPT as positive electrode material. The extent of π-interactions could also be modified by structural changes regarding the polymer backbone (polystyrene or polynorbornene) or exchanging the heteroatom sulfur in phenothiazine by oxygen in phenoxazine. By changing the molecular design and attaching electron-donating methoxy groups to the phenothiazine units, its second redox process can be reversibly enabled, even in carbonate-based electrolytes. Studies by us as well as others provided a selection of high-performing phenothiazine polymers. Their applicability was demonstrated as positive electrode in full cells of different configurations, including dual-ion battery cells using an inorganic or organic negative electrode, anion-rocking-chair cells as examples of all-organic batteries, or even an aluminum battery with a performance exceeding that of aluminum-graphite battery cells. In changing the design concept to conjugated phenothiazine polymers, a higher intrinsic semiconductivity can result, enabling the use of a lesse

Two-dimensional (2D) materials form a large and diverse family of materials with extremely rich compositions, ranging from graphene to complex transition metal derivatives. They exhibit unique physical, chemical, and electronic properties, making 2D materials highly promising in the fields of sustainable energy storage and electrocatalysis. Although significant progress has been made in the design and performance optimization of 2D materials, challenges persist, particularly in energy storage and electrocatalysis. A key issue is the restacking or aggregation of these materials in the powder form, which hinders ion transport and reduces their overall performance by limiting the effective surface area. In this Account, we delve into the latest advancements made by our team in the chemistry of 2D materials toward sustainable electrochemical energy storage and catalysis. We begin by highlighting some of the representative 2D materials developed by our team, such as fluorine-modified graphene and transition metal telluride nanosheets. These materials, with their atomic-scale thickness, offer significant advantages over traditional bulk materials by circumventing issues such as limited active surface area, extended ion transport pathways, and complex manufacturing processes, thereby providing innovative approaches for the development of high-performance materials. Next, the key synthesis strategies that have been pivotal in our research are summarized. Techniques such as electrochemical exfoliation, solid-state lithiation and exfoliation, and ion-adsorption chemical strategies have enabled precise control over the ionic and electronic conductivities, lateral dimensions, and internal atomic configurations of 2D materials. These methodologies not only facilitate the preparation of 2D materials with tailored properties, but also support the scalable production of high-quality materials. Furthermore, we outline the broad applications of 2D energy materials across various domains. In alkali-based batteries, these materials have been instrumental in enhancing battery performance, including extending the cycle life and improving the charge–discharge efficiency. They also contribute to increased energy and power densities in aqueous-based batteries and supercapacitor–battery hybrid devices. In the realm of metal-free anodes, they play a crucial role in inhibiting metal dendrite growth, thereby enhancing battery safety. Additionally, in energy catalysis, they demonstrate superior catalytic activity, promoting efficient energy conversion. In microscale electrochemical energy storage devices, they meet the demands for high power and energy density, propelling the advancement of miniaturized energy storage solutions. Lastly, we address the critical challenges confronting 2D energy materials and offer a perspective on future directions. While significant progress has been achieved in 2D material research, challenges persist in synthesis, performance optimizati