Search

In this paper, we develop a data-driven machine learning (ML) approach to predict the adiabatic temperature change (ΔT) in BaTiO3-based ceramics as a function of chemical composition, temperature, and applied electric field. The data set was curated from a survey of published electrocaloric measurements. Each chemical composition was represented by elemental descriptors of A-site and B-site elements. Pair-wise statistical correlation analysis was used to remove linearly correlated descriptors. We trained two separate regression-based ML models for indirect and direct measurements and found that both are capable of capturing the general trend of the temperature vs ΔT curve for various applied electric fields. We then complemented the regression models with a classification learning model that predicts the expected phase as a function of chemical composition and temperature. The combined regression and classification learning ML models predict a global maxima in ΔT near rhombohedral to cubic or tetragonal to cubic phase transition regions. An interactive, open source web application is developed to enable interested users to query our trained models and accelerate the design of novel BaTiO3-based ceramics with targeted phase and ΔT properties for electrocaloric applications.

Descriptors play a central role in constructing composition-structure-property relationships to guide materials design. We propose a materials descriptor, δτ , for the composition dependence of the...

Traditional strategies for designing new materials with targeted property including methods such as trial and error, and experiences of domain experts, are time and cost consuming. In the present study, we propose a machine learning design system involving three features of machine learning modeling, compositional design and property prediction, which can accelerate the discovery of new materials. We demonstrate better efficiency of on a rapid compositional design of high-performance copper alloys with a targeted ultimate tensile strength of 600–950 MPa and an electrical conductivity of 50.0% international annealed copper standard. There exists a good consistency between the predicted and measured values for three alloys from literatures and two newly made alloys with designed compositions. Our results provide a new recipe to realize the property-oriented compositional design for high-performance complex alloys via machine learning.

Lattice mismatch induced epitaxial strain has been widely used to tune functional properties in complex oxide heterostructures. Apart from the epitaxial strain, a large lattice mismatch also produces other effects including modulations in microstructure and stoichiometry. However, it is challenging to distinguish the impact of these effects from the strain contribution to thin film properties. Here, we use La0.9Sr0.1MnO3 (LSMO), a lightly doped manganite close to the vertical phase boundary, as a model system to demonstrate that both epitaxial strain and cation stoichiometry induced by strain relaxation contribute to functionality tuning. The thinner LSMO films are metallic with a greatly enhanced TC which is 97 K higher than the bulk value. Such anomalies in TC and transport cannot be fully explained by the epitaxial strain alone. Detailed microstructure analysis indicates La deficiency in thinner films and twin domain formation in thicker films. Our results have revealed that both epitaxial strain and strain relaxation induced stoichiometry/microstructure modulations contribute to the modified functional properties in lightly doped manganite perovskite thin films.

Herein, we demonstrate how to predict and experimentally validate phase diagrams for multi‐component systems from a high‐dimensional virtual space of all possible phase diagrams involving several elements based on small existing experimental data. The experimental data for bulk phases for known systems represents a sampling from this space, and screening the space allows multi‐component phase diagrams with given design criteria to be built. This approach uses machine learning methods to predict phase diagrams and Bayesian experimental design to minimize experiments for refinement and validation, all within an active learning loop. The approach is proven by predicting and synthesizing the ferroelectric ceramic system (1‐ω)(Ba0.61Ca0.28Sr0.11TiO3)‐ω(BaTi0.888Zr0.0616Sn0.0028Hf0.0476O3) with a relatively high transition temperature and triple point, as well as the NiTi‐based pseudo‐binary phase diagram (1‐ω)(Ti0.309Ni0.485Hf0.20Zr0.006)‐ω(Ti0.309Ni0.485Hf0.07Zr0.068Nb0.068) designed for high transition temperature (ω ⩽ 1). Each phase diagram is validated and optimized through only three new experiments. The complexity of these compounds is beyond the reach of today's computational methods., A machine learning‐based approach is proposed to predict all possible phase diagrams of a given multi‐component system from a high‐dimensional virtual space. By quickly screening the space, a specific phase diagram with given design criteria can be constructed. Bayesian experimental design is then employed to refine the phase diagram with as few experiments as possible.

Hamiltonian parameters estimation is crucial in condensed matter physics, but is time‐ and cost‐consuming. High‐resolution images provide detailed information of underlying physics, but extracting Hamiltonian parameters from them is difficult due to the huge Hilbert space. Here, a protocol for Hamiltonian parameters estimation from images based on a machine learning (ML) architecture is provided. It consists in learning a mapping between spin configurations and Hamiltonian parameters from a small amount of simulated images, applying the trained ML model to a single unexplored experimental image to estimate its key parameters, and predicting the corresponding materials properties by a physical model. The efficiency of the approach is demonstrated by reproducing the same spin configuration as the experimental one and predicting the coercive field, the saturation field, and even the volume of the experiment specimen accurately. The proposed approach paves a way to achieve a stable and efficient parameters estimation., Magnetic parameters can be efficiently estimated from an experimental observation of spin configuration by combining machine learning and micro‐magnetic simulation. The method includes learning a mapping from spin configurations to magnetic parameters on a small amount of micro‐magnetic simulated images and applying the trained machine learning model to a single unexplored experimental image to estimate its key parameters.

Designing a material with multiple desired properties is a great challenge, especially in a complex material system. Here, we propose a material design strategy to simultaneously optimize multiple targeted properties of multi-component Co-base superalloys via machine learning. The microstructural stability, γ′ solvus temperature, γ′ volume fraction, density, processing window, freezing range, and oxidation resistance were simultaneously optimized. A series of novel Co-base superalloys were successfully selected and experimentally synthesized from >210,000 candidates. The best performer, Co-36Ni-12Al-2Ti-4Ta-1W-2Cr, possesses the highest γ′ solvus temperature of 1266.5 °C without the precipitation of any deleterious phases, a γ′ volume fraction of 74.5% after aging for 1000 h at 1000 °C, a density of 8.68 g cm−3 and good high-temperature oxidation resistance at 1000 °C due to the formation of a protective alumina layer. Our approach paves a new way to rapidly design multi-component materials with desired multi-performance functionality.

The problem that is considered is that of maximizing the energy storage density of Pb‐free BaTiO3‐based dielectrics at low electric fields. It is demonstrated that how varying the size of the combinatorial search space influences the efficiency of material discovery by comparing the performance of two machine learning based approaches where different levels of physical insights are involved. It is started with physics intuition to provide guiding principles to find better performers lying in the crossover region in the composition–temperature phase diagram between the ferroelectric phase and relaxor ferroelectric phase. Such an approach is limiting for multidopant solid solutions and motivates the use of two data‐driven machine learning and design strategies with a feedback loop to experiments. Strategy I considers learning and property prediction on all the compounds, and strategy II learns to preselect compounds in the crossover region on which prediction is carried out. By performing only two active learning loops via strategy II, the compound (Ba0.86Ca0.14)(Ti0.79Zr0.11Hf0.10)O3 is synthesized with the largest energy storage density ≈73 mJ cm−3 at a field of 20 kV cm−1, and an insight into the relative performance of the strategies using varying levels of knowledge is provided.

The damping and transformation behaviors of Ti50(Pd50−xCrx) shape memory alloys with x ranging from 4.0 to 5.0 are systematically investigated. The damping capacity (Q−1) at the martensitic transformation is found to be inversely proportional to the square root of frequency, i.e., Q−1∝ω−0.5. A relaxation peak or shoulder is observed slightly below the martensitic transformation damping peak for compositions within the compositional crossover region (4.5 ⩽ x ⩽ 4.8). Furthermore, the damping capacity at the martensitic transformation is smaller within the compositional crossover region (4.5 ⩽ x ⩽ 4.8), compared with that of compositions at both sides ( x = 4.0 and x = 5.0 ). These observations can be ascribed to the hysteretic motion of interfaces between different phases near the compositional crossover region.

The corrosion behavior of pure Molybdenum (Mo) and Mo doped with 0.3 wt.% La2O3 was investigated in aerated 3.5 wt.% NaCl at 25 °C. Compared with the pure Mo, the doped Mo exhibits significantly increased corrosion resistance, with a smaller current density during anodic polarization and a 2∼3 times larger charge transfer resistance. Such an enhancement originates from the refinement of grains and purification of the grain boundary due to the addition of La2O3, which facilitates the formation of a compact and protective oxide film. Our results provide a recipe to improve the corrosion resistance of Mo alloys.

A key aspect of the developing field of materials informatics is optimally guiding experiments or calculations towards parts of the relatively vast feature space where a material with desired property may be discovered. We discuss our approach to adaptive experimental design and the methods developed in decision theory and global optimization which can be used in materials science. We show that the use of uncertainties to trade-off exploration versus exploitation to guide new experiments or calculations generally leads to enhanced performance, highlighting the need to evaluate and incorporate errors in predictive materials design. We illustrate our ideas on a computed data set of M2AX phases generated using ab initio calculations to find the sample with the optimal elastic properties, and discuss how our approach leads to the discovery of new NiTi-based alloys with the smallest thermal dissipation.

In this paper, we propose a general experimental design framework for optimally guiding new experiments or simulations in search of new materials with desired properties. The method uses the knowledge of previously completed experiments or simulations to recommend the next experiment which can effectively reduce the pertinent model uncertainty affecting the materials properties. To illustrate the utility of the proposed framework, we focus on a computational problem that utilizes time-dependent Ginzburg-Landau (TDGL) theory for shape memory alloys to calculate the stress-strain profiles for a particular dopant at a given concentration. Our objective is to design materials with the lowest energy dissipation at a specific temperature. The aim of experimental design is to suggest the best dopant and its concentration for the next TDGL simulation. Our experimental design utilizes the mean objective cost of uncertainty (MOCU), which is an objective-based uncertainty quantification scheme that measures uncertainty based upon the increased operational cost it induces. We analyze the performance of the proposed method and compare it with other experimental design approaches, namely random selection and pure exploitation.

One of the main challenges in materials discovery is efficiently exploring the vast search space for targeted properties as approaches that rely on trial-and-error are impractical. We review how methods from the information sciences enable us to accelerate the search and discovery of new materials. In particular, active learning allows us to effectively navigate the search space iteratively to identify promising candidates for guiding experiments and computations. The approach relies on the use of uncertainties and making predictions from a surrogate model together with a utility function that prioritizes the decision making process on unexplored data. We discuss several utility functions and demonstrate their use in materials science applications, impacting both experimental and computational research. We summarize by indicating generalizations to multiple properties and multifidelity data, and identify challenges, future directions and opportunities in the emerging field of materials informatics.

Ferroelectric-field switching in $\mathrm{BaTi}{\mathrm{O}}_{3}$ generates ``Barkhausen noise'' when domain walls are displaced. We show by acoustic emission spectroscopy that electric-field switching of ${90}^{\ensuremath{\circ}}$ boundaries generates large strain fields, which emit acoustic phonons during ferroelectric hysteresis measurements. We use highly sensitive receivers (microphones) to measure the time sequences of noise in close analogy to noise patterns in ferroelastic and magnetic materials. Domain-wall interactions and jamming generate the ``crackling noise'' that follows scale invariant avalanche dynamics: the avalanche energy and amplitude probability distribution functions follow power laws with exponents $\ensuremath{\varepsilon}=1.65$ (energy) and \ensuremath{\tau}\ensuremath{'} = 2.25 (amplitudes). Aftershocks are very common and follow Omori law with probability $\ensuremath{\sim}{t}^{\ensuremath{-}p}$ where $t$ is the time elapsed after the main shock and $p$ is the Omori exponent $p\ensuremath{\sim}1$. The interevent times follow a double power-law distribution with exponents 0.9 for small times and 2.2 for the larger times. The scaling behavior is consistent with predictions of mean field theory.

The functional properties of shape memory alloys are strongly sensitive to the composition of alloys. In this study, the effect of Ti/Ni ratio and Hf/Zr ratio on the martensitic transformation behavior, shape memory effect and microscopic structures of two different quaternary TiNiHfZr alloys (TixHf15Zr5Ni80-x and Ti31.5HfyZr20-yNi48.5) have been investigated systematically. It is found that increasing Ti/Ni ratio of TixHf15Zr5Ni80-x alloys dramatically increases the martensitic transformation temperature and a maximum value (5%) of the recovered strain during shape recovery on heating is obtained for the alloy with x = 30. On the contrary, changing Hf/Zr ratio of Ti31.5HfyZr20-yNi48.5 alloys does not result in big change of martensitic transformation temperature or recovered strain. The evolution of macroscopic properties of TixHf15Zr5Ni80-x alloys with x can be understood by considering the balanced effect between the Ni content change of the matrix, and the volume fraction of Ti2Ni-like precipitates, which are often brittle and not desired for SMAs. Our results suggest that the TixHf15Zr5Ni80-x (x > 30.5 at. %) alloys are promising shape memory alloys for high temperature applications above 250 °C.

The oxidation mechanism of refractory Molybdenum exposed to oxygen-saturated lead-bismuth eutectic (LBE) at 600 °C was investigated. The results show a hierarchical structure of multiple corrosion layers. Specifically, a nanostructured tetragonal MoO2 is the first scale that forms on the Mo substrate, followed by development of an ultrafine-grained tetragonal lead molybdate (PbMoO4) layer. On the top of this ultrafine-grained layer, there is a thicker scale that consists of micron-sized columnar PbMoO4 and its upper part can be further oxidized into blocky monoclinic-structured Pb2MoO5. The formation of these multilayered scales is likely predominated by inward diffusion of oxygen and Pb.

The precision control of thermal expansion is of fundamental interest and desirable for applications that require materials to retain their shape at different temperatures. In the present study, we propose that the thermal expansion coefficient can be tuned by uniaxial external stress for a single material embedded with nanoclusters, if the formation, growth or alignment of the nanoclusters depends on the external stress. We demonstrate the idea in a prototype alloy (Ti50(Pd40Cr10)) located at the compositional crossover region between martensite and strain glass in the temperature-composition phase diagram of Ti50(Pd50−xCrx). Its thermal expansion coefficient varies linearly from positive, through zero to negative values with increasing uniaxial tensile stress within 200 K ∼ 300 K. The phase field simulations show that the volume fraction of nanoscale martensite variant favored by the external stress increases with stress, producing extra strain and compensating for the contraction of the austenite matrix on cooling. The degree of compensation leads to different thermal expansion coefficients. Such a tunable thermal expansion behavior occurs only in the crossover compositions between martensite and strain glass, providing a design recipe for searching new systems with similar behavior.

Shape memory alloys (SMAs) have attracted considerable attention for their high damping capacities. Here we investigate the damping behavior of Ti 50 (Pd 50− x D x ) SMAs (D=Fe, Co, Mn, V) by dynamic mechanical analysis. We find that these alloys show remarkably similar damping behavior. There exists a sharp damping peak associated with the B2–B19 martensitic transformation and a high damping plateau ( Q −1 ~0.02–0.05) over a wide ambient-temperature range (220–420 K) due to the hysteretic twin boundary motion. After doping hydrogen into the above alloys, a new relaxation-type damping peak appears in the martensite phase over 270–360 K. Such a peak is considered to originate from the interaction of hydrogen atoms with twin boundaries and the corresponding damping capacity ( Q −1 ~0.05–0.09) is enhanced by roughly twice that of the damping plateau for each alloy. Moreover, the relaxation peaks are at higher temperatures for the TiPd-based alloys (270–370 K) than for the TiNi-based alloys (190–260 K). We discuss the influence of hydrogen diffusion, mobility of twin boundaries and hydrogen–twin boundary interaction on the temperature range of the relaxation peak. Our results suggest that a martensite, with appropriate values for twinning shear and hydrogen doping level, provides a route towards developing high damping SMAs for applications in desired temperature ranges.

Ferroic glasses (strain glass, relaxor and cluster spin glass) refer to frozen disordered states in ferroic systems; they are conjugate states to the long-range ordered ferroic states—the ferroic crystals. Ferroic glasses exhibit unusual properties that are absent in ferroic crystals, such as slim hysteresis and gradual property changes over a wide temperature range. In addition to ferroic glasses and ferroic crystals, a third ferroic state, a glass-ferroic (i.e., a composite of ferroic glass and ferroic crystal), can be produced by the crystallization transition of ferroic glasses. It can have a superior property not possessed by its two components. These three classes of ferroic materials (ferroic crystal, ferroic glass and glass-ferroic) correspond to three transitions (ferroic phase transition, ferroic glass transition and crystallization transition of ferroic glass, respectively), as demonstrated in a generic temperature vs. defect-concentration phase diagram. Moreover, through constructing a phase field model, the microstructure evolution of three transitions and the phase diagram can be reproduced, which reveals the important role of point defects in the formation of ferroic glass and glass-ferroic. The phase diagram can be used to design various ferroic glasses and glass-ferroics that may exhibit unusual properties.

The alloy Ti50(Pd40Cr10) undergoes a strain glass transition around room temperature evidenced by frequency dispersion of dynamic mechanical properties and lack of average structure change from that of the high symmetry austenite phase. However, since the strain glass transition is not a thermodynamic phase transition but a dynamic freezing process governed by the kinetics, a quantitative characterization of the slowing down of dynamics during the strain glass transition is still lacking. In the present study, the probability distribution function (PDF) of the relaxation time of the strain glass alloy is investigated spanning the whole transition temperature range (253 K-313 K). The slowing down of dynamics of the strain glass is indicated by the rapid increase of the characteristic relaxation time ([Formula: see text]) upon cooling. The [Formula: see text], as a function of temperature, shows a transition from Vogel-Fulcher relationship to an Arrhenius relationship. Such a change suggests two fundamentally different states: unfrozen strain glass state and frozen strain glass state. Furthermore, the spread of the PDF is connected to the fraction of quasi-static nanodomains, which helps the understanding of the dynamic freezing process in the strain glass.

A key challenge in guiding experiments toward materials with desired properties is to effectively navigate the vast search space comprising the chemistry and structure of allowed compounds. Here, it is shown how the use of machine learning coupled to optimization methods can accelerate the discovery of new Pb-free BaTiO3 (BTO-) based piezoelectrics with large electrostrains. By experimentally comparing several design strategies, it is shown that the approach balancing the trade-off between exploration (using uncertainties) and exploitation (using only model predictions) gives the optimal criterion leading to the synthesis of the piezoelectric (Ba0.84 Ca0.16 )(Ti0.90 Zr0.07 Sn0.03 )O3 with the largest electrostrain of 0.23% in the BTO family. Using Landau theory and insights from density functional theory, it is uncovered that the observed large electrostrain is due to the presence of Sn, which allows for the ease of switching of tetragonal domains under an electric field.

Alloys exhibiting high damping capacity in ambient-temperature range (250–400 K) are rare. Here it is reported that a martensitic alloy Ti–45Pd–5Cr shows twin-boundary-related high damping capacity over this temperature range. In the hydrogen-doped martensitic state, a relaxation-type high damping peak (Q−1 ≈ 0.09) exists, extending over 305–370 K. In the hydrogen-free martensitic state, there is a fascinating high-damping plateau (Q−1 ≈ 0.05) over 250–450 K. This work provides new insight into how to develop high damping alloys for the desired temperature range.

The time dependence of physical properties in the paraelectric phase was probed recently in a Mn(3+) doped (Ba0.8Sr0.2)TiO3 ceramic, providing a simple situation (without spontaneous polarization or domain walls) to quantify the role of the oxygen vacancy during aging. In the present study, we propose a quantitative model for paraelectric aging to understand how the distribution of the oxygen vacancy evolves with time and consequently influences the dielectric response of the paraelectric phase. First, by comparing dielectric behavior of paraelectric aging in a Mn(3+) doped (Ba0.75Sr0.25)TiO3 ceramic and the dielectric tunable effect, an internal bias field E(in) related to the oxygen vacancy is shown to exist in the paraelectric phase. Second, by introducing such a time dependent E(in) in a Landau-type model, we reproduce the dielectric response of Mn(3+) doped (Ba0.8Sr0.2)TiO3 ceramic during paraelectric aging. It is suggested that the increase of dielectric permittivity can be ascribed to the decrease of E(in) with time. The investigation of paraelectric aging is helpful for understanding the role of the oxygen vacancy in influencing the physical properties of ferroelectric materials.

The time evolution of the physical properties of martensite during martensite ageing is traditionally explained by the symmetry-conforming short-range order (SC-SRO) principle, which requires the spatial configuration of crystal defects to follow the symmetry change of the host lattice. In the present study, we show that the volume change of the host lattice also contributes to the ageing effects in Cu-Al-Ni shape memory alloy besides the symmetry change. To substantiate this statement the gradual increase of the storage modulus with time at constant temperature was measured by dynamic mechanical analysis (DMA) and the experimental results were quantitatively described in the framework of the symmetry-conforming Landau theory of martensitic transformations in a crystal with defects. The comparison of experimental and theoretical results confirmed that the time dependence of the storage modulus is caused by two different physical mechanisms. Evaluations showing that the first mechanism is driven by the spontaneous symmetry change and the second mechanism is caused by the volume change after the martensitic transformation was carried out.

Different ferroelectric domains are degenerate states of the same ferroelectric phase; thus they are energetically equivalent and, in principle there exists no preference for a particular domain pattern. However, the existence of point defects is considered to stabilize certain preferential domain states. In order to study the temperature violation on such stabilized domains, we performed in?situ observation on an aged K + -doped (Ba, Sr)TiO3 single crystal and found that both the domain configuration and polarization state can be memorized after experiencing a thermally activated ferro?para?ferro transition cycle, as manifested by a reappearance of the same domain pattern and double P?E hysteresis loop. In contrast, after the sample was aged in the paraelectric state (>10?min), these memory effects disappeared. The above memory effects are considered to originate from the interaction between point defects and the crystal symmetry driven by a symmetry-conforming tendency of point defects. Such a mechanism suggests that the memory effects are relevant to the existence of acceptor dopant and associated mobile oxygen vacancies, and they are not restricted to a particular dopant. Thus similar memory effects are expected to exist in a wide range of ferroelectric materials with acceptor doping.