Residual stresses often arise after glass crystallization due to the residual glass and crystal phases' thermal and elastic properties mismatch. Using the Positional Finite Element Method, we developed a computational tool for nonlinear 2D geometric analysis to simulate residual stresses from crystal growth in a glass matrix. We implemented a damage model in which interface elements are introduced into the mesh. These elements' elastic modules were penalized after reaching a tensile stress limit to simulate crack propagation. The crystal mesh was generated separately and included in the glass mesh via kinematic compatibility without introducing degrees of freedom nor requiring mesh conformity. Results were consistent with analytical calculations using Selsing's model. The fracture mode in crystal-glass composite microstructures with different signs of thermal expansion mismatch is closely in accordance with the microstructure pictures in the literature. The adapted damage model showed excellent potential in evaluating and predicting crack patterns due to residual stresses in glass-ceramics.

Chemical phase equilibria and phase diagrams are well established and have served as indispensable guides for designing materials through chemical tuning. However, a general thermodynamic theory for strain phase equilibria remains elusive despite extensive studies which have revealed dramatic impacts of mechanical strain on the stability of phases and domain states in solid-state materials. Here, we establish the thermodynamic theory of strain equilibria and a general framework for efficiently constructing multidomain and multiphase diagrams of arbitrarily strained solids under incoherent conditions. As examples, we obtain temperature-strain phase diagrams of ferroelectric PbTiO3, strongly correlated VO2, and unconventional ferroelectric Hf0.5Zr0.5O2. We reveal the analogs of the Gibbs phase rules, multiple- and multi-critical points, common-tangent construction, and level rule in strain equilibria to those in the familiar chemical phase equilibria. Our strain equilibria theory offers a powerful framework for predicting the thermodynamic stability of structural phases and domains in strained solids and for guiding the strain engineering of their functional properties.

Surface relief in forms of extrusions and intrusions as the substantial feature of early fatigue damage is one of the most important phenomena studied in metal fatigue. The most common surface relief models in bulk metals are agreed to be correlated with the formation of typical dislocation patterns as persistent slip bands (PSBs), while little is known about the fundamental mechanisms at submicron and even nanometer scales where dislocation pattern formation is fully inhibited. Here, as exampled with thin Au films, the underlying fatigue damage mechanism at small scales is investigated through the quantitative characterization of fatigue damages. Continuous generation and migration of vacancies is found to be crucial for the shape of extrusion/intrusions and kinetics of their growth at submicron and even nanometer scales. Due to the degraded dislocation interaction and intensified vacancy diffusion, the delayed vacancy accumulation in the small-scale metal interior suppresses the extrusion and interface void formation in thinner films, which finally leads to the superior ability to support tremendous surface relief and strong fatigue resistance. The finding of the vacancy-dominated fatigue mechanism at small scales extends our understanding of the metal fatigue mechanisms down to the submicron and even nanometer scales and suggests a novel interface engineering strategy by vacancy behavior modulation for fatigue-tolerance material design.

Micro- and nanolattices can achieve simultaneously both lightweight and ultra-high strength due to a combination of resilient architecture and enhanced mechanical properties of small-scale unit constituents. However, understanding the fundamental deformation mechanism for these novel materials remains intriguing due to the intricate interplay of various geometric characteristics and intrinsic microscopic defects. Here, we investigate the fundamental mechanism of plastic deformation in terms of dynamic evolution of defects and corresponding mechanical responses in aluminum nanolattices using a mesoscale defect dynamics model which couples three-dimensional dislocation dynamics and finite element method. Our concurrently coupled model could capture detailed dislocation motion under complex loading conditions and predict the plastic flow stress of the nanolattices. We demonstrate that the size of individual constituent beam plays a critical role in the properties of nanolattices through small-scale strengthening, showing the strength of the nanolattices increases with decreasing beam size with the scaling law of σ∝db−0.89. As a result, the scaling law of plastic yielding with material density drastically increases from the continuum prediction. In addition, our modeling could allow us to study the geometric effect with various nanolattices. These findings establish a foundation for the fundamental understanding of the governing mechanism of plastic deformation, allowing access to a new regime in designing optimal structures using nanolattices.

The design of alloys has typically involved adaptive experimental synthesis and characterization guided by machine learning models fitted to available data. A bottleneck for sequential design, be it for self-driven or manual synthesis, by Bayesian Global Optimization (BGO) for example, is that the search space becomes intractable as the number of alloy elements and its compositions exceed a threshold. Here we investigate how reinforcement learning (RL) performs in the compositional design of alloys within a closed loop with manual synthesis and characterization. We demonstrate this strategy by designing a phase change multicomponent alloy (Ti27.2Ni47Hf13.8Zr12) with the highest transformation enthalpy (ΔH) -37.1 J/g (-39.0 J/g with further calibration) within the TiNi-based family of alloys from a space of over 2 × 108 candidates, although the initial training is only on a compact dataset of 112 alloys. We show how the training efficiency is increased by employing acquisition functions containing uncertainties, such as expected improvement (EI), as the reward itself. Existing alloy data is often limited, however, if the agent is pretrained on experimental results prior to the training process, it can access regions of higher reward values more frequently. In addition, the experimental feedback enables the agent to gradually explore new regions with higher rewards, compositionally different from the initial dataset. Our approach directly applies to processing conditions where the actions would be performed in a given order. We also compare RL performance to BGO and the genetic algorithm on several test functions to gain insight on their relative strengths in materials design.

The region of permanent densification beneath a Berkovich indentation imprint in silica glass is investigated using a novel chemical dissolution technique. The use of the similitude regime in sharp indentation testing allows one to record reliable data with a good spatial resolution that makes it possible to deal with low loads (typically below 10 mN) and, more importantly, crack-free imprints. The densified zone dissolves more quickly than the non densified regions. The analysis of the results, along the vertical axis, indicates that the densification zone is rather homogeneous with a steep transition to the non densified zone. The size of the densification zone, with respect to the initial free surface, is estimated to be around two times the maximum penetration depth of the instrumented indentation test. These findings are compared with those obtained by numerical simulations using different constitutive equations from the literature. A very good concordance between Raman spectroscopy and chemical probe results is found for imprints made with no or few cracking events during indentation testing.

The work clarifies several key questions in shape memory research that have eluded previous studies. The findings show that dislocation slip emanates at austenite-martensite interfaces during unloading and aligns with the internal twin boundary interface of martensite. It was observed that the type II internal twins of the martensite become parallel dislocations in the austenite. During reloading, these dislocations act as nucleation sites for the martensitic twins, reducing the nucleation barrier and the transformation stress. The precipitates facilitate martensite nucleation but also act as an obstacle to martensite front motion, restrict detwinning, and pin the interfacial dislocations during unloading, thereby contributing to residual strains and martensite stabilization. Martensite nucleation is not suppressed by the size of the thin film, which is of the order of 85 to 105 nanometers thick, and repeated transformation occurred cycle after cycle. Single crystals deformed in the <101>LD exhibited the best recoverability of up to 5.5 % and tensile stresses of up to 1.4 GPa. It was demonstrated for the first time that, when favorably oriented, Ni4Ti3 precipitates undergo a reversible phase transformation to R-phase and can accommodate up to 4 % reversible strains.

Advanced nanofabrication exploiting e-beam lithography has been used to prepare nanocomposites consisting of periodic arrays of soft magnetic FeCo-based nano-inclusions of variable dimensions, embedded in a hard magnetic FePt matrix. Nanocomposites with non-magnetic (Pt) inclusions and single phase hard magnetic FePt microstructures were prepared as reference samples. The formation of Kirkendall voids in the soft-in-hard nanocomposites, because of annealing-induced diffusion, has been identified through high resolution imaging and chemical analysis. Replication of the μm-scaled nanocomposite and hard magnetic units over mm-scaled surfaces allowed global magnetic characterisation of magnetisation reversal using standard magnetometry. Magnetic force microscopy imaging was used for spatially resolved magnetic characterisation in different remanent magnetic states. Both of these experimental techniques were used to study the influence of the size, volume content and nature of the nano-inclusions on magnetisation reversal. The well-defined geometry and nano-scaled size of the inclusions render these nanocomposites as model systems for micromagnetic simulations and very good agreement was achieved between measured and simulated magnetisation reversal curves. This accordance motivates future in-silico optimisation of such soft-in-hard nanocomposites, which may serve as model systems to guide the design of new high performance bulk magnets which are less dependent on critical elements.

Defects in ferroelectric materials can cause microscopic strain and electron doping, which opens up new possibilities for unique physical properties in nanoelectronics. However, comprehending the relationship between these deep-buried defect configurations and the resulting physical properties is a long-term challenge. Here, utilizing the integrated differential phase imaging technique equipped in scanning transmission electron microscopy combined with energy dispersive X-ray spectroscopy, we show the full element-resolved atomic configuration of characteristic extrinsic defects in Nb-doped epitaxial BiFeO3 thin films. Atomic-resolution characterization reveals that the Nb doping substitutes Fe at the center of an oxygen octahedron and creates Fe vacancies near the defect core, leading to the formation of one perovskite cell-wide BiNbO3 chain that is expanded by 19 % in perovskite cell volume and rotated by 45° relative to the matrix lattice. Further quantitative analysis of the defect demonstrates that the ferroelectric polarization distribution around the dopant-induced defects displays a “head-to-head” configuration, indicating the accumulation of negative charges. Supported by electron energy loss spectroscopy and density functional theory calculations, the linear defect leads to one-dimensional metallic channels embedded within the ferroelectric BiFeO3 matrix. These atomic-scale findings establish the structure-functionality relationship of extrinsic defect in Nb-doped BiFeO3 thin films, providing new insight into their fundamental understanding and potential use in low-dimensional nanoelectronics devices.

The discovery of highly efficient near-infrared (NIR) emitting phosphors is essential for multiple spectroscopic applications. Pentavalent-manganese (Mn5+) is a rarely-reported NIR-II activator due to the its instability in inorganic solids. In this study, Mn5+-activated apatite phosphors A5(B1-xMnxO4)3X (A = Sr, Ba; B = P, V; X = OH, F, Cl, Br) were reported. Benefiting from the discrete P/VO4 tetrahedron and stabilization effect of alkaline-earth cation, Mn5+ existed as MnO43− anion can be effectively incorporated into the A5(P/VO4)3X. A selection of 16 different apatites is conducted to test the structural and compositional effect on Mn5+ stabilization and luminescence. The structure-property relationship was discussed with a detailed analysis on Mn5+ Tanabe-Sugano diagram and Dq/B calculation. Particularly, the optimized Ba5(VO4)3Cl: 0.01Mn5+ (λex = 650 nm, λem = 1181 nm) reveals an exceptional external quantum efficiency of 61.1 % with vivid turquoise body color. The multiparametric thermometry by luminescence intensity ratio (LIR) and 1E lifetime were demonstrated with excellent linear regressions (R2 > 99.9 %) and high relative sensitivities of > 1.0% K−1 in the range of physiological temperatures. A NIR-II phosphor-converted light-emitting diode (pc-LED) is fabricated with a NIR output power of 30.18 mW at 60 mA driven current and photoelectric conversion efficiency of 6.63 %.