Search

Thermoelectric oxide thin films are promising in chip cooling. The issues on the orientation of thin films are essential as they are related to the structures, morphologies, and thermoelectric properties. In this regard, the orientation modulation is conducted on La‐doped SrTiO3 thin films on (LaAlO3)0.3(Sr2TaAlO6)0.7 (LSAT) single crystal substrates. Layer‐by‐layer growth mode is found in (001)‐ and (110)‐ oriented thin films, resulting in few grain boundaries (GBs). In (111)‐oriented films, island growth mode leads to columnar grain boundaries that build up potential barriers for electrons to be strongly scattered and filtered, suppressing electron mobility and increasing effective mass. In addition, the GBs serve as oxygen vacancy diffusion paths when annealing, causing increased carrier concentration and lattice contraction. The weighted mobility of 71.9 cm2 V−1 s−1 and electrical conductivity of ≈600 S cm−1 are realized in the (001)‐oriented film at room temperature. Ultimately, outstanding power factor values of ≈569 µW m−1 K−2 (room temperature) and ≈791 µW m−1 K−2 (573 K) are successfully achieved, outperforming those in polycrystalline ceramics and (111)‐oriented films. This study systematically investigates the influence of grain boundaries and orientations on SrTiO3‐based thermoelectric films, which lays a solid foundation for improving thermoelectric performance in other oxide thin films. [ABSTRACT FROM AUTHOR]

The determinations of detailed stress states are of great importance for various environmental and engineering investigations, which makes numerical stress modelling a key issue in many fields. We developed a new stress modelling method governed by elastic wave equations using finite-difference scheme. By introducing an artificial damping factor to the particle velocity in wave modelling, the proposed method is able to solve both the dynamic stress evolution and the static stress state of equilibrium. We validate the proposed method both in body force and surface force benchmarks in different scales. With the proposed method, we are able to substantially improve the modelling accuracy of models in unbounded domain by using the perfectly matched layer as the artificial boundary conditions. A 3-D concrete-faced rockfill dam model is further presented as a numerical example of practical investigation. The consistent results with the finite-element method further illustrate the proposed method's applicability. As a minor modification to wave modelling scheme, the proposed stress modelling method is not only accurate for geological models through different scales, but also physically reasonable and easy to implement for geophysicists. [ABSTRACT FROM AUTHOR]

As a transparent thermoelectric oxide, gallium‐doped zinc oxide (GZO) has the potential to power wearable or portable electronics and may be used in the integrated circuits industry for chip cooling. Constructing ZnO–GZO interfaces has been proposed as an effective strategy for improving thermoelectric performance of GZO thin films. However, without the aid of band structure calculation for multilayered films, it is hard to directly elucidate the underlying mechanisms of carrier transport. Weighted mobility is an indicator that reveals the inherent electronic transport properties like carrier scattering, electronic band structure, and so on. Thus, to further investigate the effects of ZnO–GZO interfaces on electrical properties of GZO thin films, the structures containing different numbers of ZnO–GZO interfaces were designed and the correlations among numbers of ZnO–GZO interfaces, weighted mobility, and electrical properties were explored. It was found that with more ZnO–GZO interfaces, the weighted mobility increased, and the power factor values also improved as well. Consequently, an enhanced power factor value reached 439 μW m−1 K−2 at 623 K. This work demonstrated the beneficial effects of multiple interfaces on the improvements of electrical transport performance through analyzing weighted mobility, which laid a foundation for further optimization of thermoelectric performance. [ABSTRACT FROM AUTHOR]

Thermoelectric materials can realize direct conversion between heat and electricity, showing excellent potential for waste heat recovery. Cu2Se is a typical superionic conductor thermoelectric material having extraordinary ZT values, but its superionic feature causes poor service stability and low mobility. Here, we reported a fast preparation method of self-propagating high-temperature synthesis to realize in situ compositing of BiCuSeO and Cu2Se to optimize the service stability. Additionally, using the interface design by introducing graphene in these composites, the carrier mobility could be obviously enhanced, and the strong phonon scatterings could lead to lower lattice thermal conductivity. Ultimately, the Cu2Se-BiCuSeO-graphene composites presented excellent thermoelectric properties with a ZTmax value of ~2.82 at 1000 K and a ZTave value of ~1.73 from 473 K to 1000 K. This work provides a facile and effective strategy to largely improve the performance of Cu2Se-based thermoelectric materials, which could be further adopted in other thermoelectric systems. Here, the authors devise a synthesis strategy to optimize the stability and thermoelectric performance of Cu2Se-based materials. They obtain a maximum ZT value of ~2.82 at 1000 K on Cu2Se-BiCuSeO-graphene composites. [ABSTRACT FROM AUTHOR]

Abnormal event detection aims to automatically identify unusual events that do not comply with expectation. Recently, many methods have been proposed to obtain the temporal locations of abnormal events under various determined thresholds. However, the specific categories of abnormal events are mostly neglect, which are important to help in monitoring agents to make decisions. In this study, a Temporal Attention Network (TANet) is proposed to capture both the specific categories and temporal locations of abnormal events in a weakly supervised manner. The TANet learns the anomaly score and specific category for each video segment with only video‐level abnormal event labels. An event recognition module is exploited to predict the event scores for each video segment while a temporal attention module is proposed to learn a temporal attention value. Finally, to learn anomaly scores and specific categories, three constraints are considered: event category constraint, event separation constraint and temporal smoothness constraint. Experiments on the University of Central Florida Crime dataset demonstrate the effectiveness of the proposed method. [ABSTRACT FROM AUTHOR]