Explore the vast range of titles available.

Understanding the function of moisture on perovskite is challenging since the random environmental moisture strongly disturbs the perovskite structure. Here, we develop various N 2 -protected characterization techniques to comprehensively study the effect of moisture on the efficient cesium, methylammonium, and formamidinium triple-cation perovskite (Cs 0.05 FA 0.75 MA 0.20 )Pb(I 0.96 Br 0.04 ) 3 . In contrast to the secondary measurements, the established air-exposure-free techniques allow us directly monitor the influence of moisture during perovskite crystallization. We find a controllable moisture treatment for the intermediate perovskite can promote the mass transportation of organic salts, and help them enter the buried bottom of the films. This process accelerates the quasi-solid-solid reaction between organic salts and PbI 2 , enables a spatially homogeneous intermediate phase, and translates to high-quality perovskites with much-suppressed defects. Consequently, we obtain a champion device efficiency of approaching 24% with negligible hysteresis. The devices exhibit an average T 80 -lifetime of 852 h (maximum 1210 h) working at the maximum power point. Perovskite structure is disturbed by environmental moisture, limiting the device performance. Here, Wei et al. monitor the effect of moisture during the growth by N 2 -protected characterization techniques, and obtain an operationally stable perovskite solar cell with efficiency approaching 24%.

Wide-bandgap perovskite solar cells (WBG-PSCs) are critical for developing perovskite/silicon tandem solar cells. The defect-rich surface of WBG-PSCs will lead to severe interfacial carrier loss and phase segregation, deteriorating the device’s performance. Herein, we develop a surface reconstruction method by removing the defect-rich crystal surface by nano-polishing and then passivating the newly exposed high-crystallinity surface. This method can refresh the perovskite/electron-transporter interface and release the residual lattice strain, improving the charge collection and inhibiting the ion migration of WBG perovskites. As a result, we can achieve certified efficiencies of 23.67% and 21.70% for opaque and semi-transparent PSCs via a 1.67-eV perovskite absorber. Moreover, we achieve four-terminal perovskite/silicon tandem solar cells with a certified efficiency of 33.10% on an aperture area of one square centimeter. The defect-rich surface of wide-bandgap perovskite solar cells leads to severe interfacial carrier loss and phase segregation. Here, the authors reconstruct the surface through nano-polishing followed by passivation, achieving certified efficiency of 33.1% for perovskite/silicon tandem solar cells.

Constructing 2D/3D perovskite heterojunctions is effective for the surface passivation of perovskite solar cells (PSCs). However, previous reports that studying perovskite post-treatment only physically deposits 2D perovskite on the 3D perovskite, and the bulk 3D perovskite remains defective. Herein, we propose Cl 2 -dissolved chloroform as a multifunctional solvent for concurrently constructing 2D/3D perovskite heterojunction and inducing the secondary growth of the bulk grains. The mechanism of how Cl 2 affects the performance of PSCs is clarified. Specifically, the dissolved Cl 2 reacts with the 3D perovskite, leading to Cl/I ionic exchange and Ostwald ripening of the bulk grains. The generated Cl − further diffuses to passivate the bulk crystal and buried interface of PSCs. Hexylammonium bromide dissolved in the solvent reacts with the residual PbI 2 to form 2D/3D heterojunctions on the surface. As a result, we achieved high-performance PSCs with a champion efficiency of 24.21% and substantially improved thermal, ambient, and operational stability. Constructing 2D/3D perovskite heterojunctions is effective for the surface passivation of perovskite solar cells. Here, the authors apply Cl 2 -dissolved chloroform as a multifunctional solvent and achieve a champion device efficiency of 24.21% with improved thermal, ambient and operational stability.

Humanoid grasping is a critical ability for anthropomorphic hand, and plays a significant role in the development of humanoid robots. In this article, we present a deep learning-based control framework for humanoid grasping, incorporating the dynamic contact process among the anthropomorphic hand, the object, and the environment. This method efficiently eliminates the constraints imposed by inaccessible grasping points on both the contact surface of the object and the table surface. To mimic human-like grasping movements, an underactuated anthropomorphic hand is utilized, which is designed based on human hand data. The utilization of hand gestures, rather than controlling each motor separately, has significantly decreased the control dimensionality. Additionally, a deep learning framework is used to select gestures and grasp actions. Our methodology, proven both in simulation and on real robot, exceeds the performance of static analysis-based methods, as measured by the standard grasp metric Q 1 . It expands the range of objects the system can handle, effectively grasping thin items such as cards on tables, a task beyond the capabilities of previous methodologies.

Performing diverse motor skills with a universal controller has been a longstanding challenge for legged robots. While motion imitation-based reinforcement learning (RL) has shown remarkable performance in reproducing designed motor skills, the trained controller is only suitable for one specific type of motion. Motion synthesis has been well developed to generate a variety of different motions for character animation, but those motions only contain kinematic information and cannot be used for control. In this study, we introduce a control pipeline combining motion synthesis and motion imitation-based RL for generic motor skills. We design an animation state machine to synthesize motion from various sources and feed the generated kinematic reference trajectory to the RL controller as part of the input. With the proposed method, we show that a single policy is able to learn various motor skills simultaneously. Further, we notice the ability of the policy to uncover the correlations lurking behind the reference motions to improve control performance. We analyze this ability based on the predictability of the reference trajectory and use the quantified measurements to optimize the design of the controller. To demonstrate the effectiveness of our method, we deploy the trained policy on hardware and, with a single control policy, the quadruped robot can perform various learned skills, including automatic gait transitions, high kick, and forward jump.

Dexterous hands play vital roles in tasks performed by humanoid robots. For the first time, we quantify the correlation between design variables and the performance of 65 dexterous hands using Cramér’s V. Comprehensive cross-correlation analysis quantitatively reveals how the performance, such as speed, weight, fingertip force, and compactness are related to the design variables including degrees of freedom (DOF), structural form, driving form, and transmission mode. This study shows how various design parameters are coupled inherently, leading to compromise in performance metrics. These findings provide a theoretical basis for the design of dexterous hands in various application scenarios and offer new insights for performance optimization.

Robot control in complex and unpredictable scenarios presents challenges such as adaptability, robustness, and human-robot interaction. These scenarios often require robots to perform tasks that involve unknown objects in unstructured environments with high levels of uncertainty. Traditional control methods, such as automatic control, may not be suitable due to their limited adaptability and reliance on prior knowledge. Human-in-the-loop method faces issues such as insufficient feedback, increased failure rates due to noise and delays, and lack of operator immersion, preventing the achievement of human-level performance. This study proposed a shared control framework to achieve a trade-off between efficiency and adaptability by combing the advantages of both teleoperation and automatic control method. The proposed approach combines the advantages of both human and automatic control methods to achieve a balance between performance and adaptability. We developed a linear model to compare three control methods and analyzed the impact of position noise and communication delays on performance. The real-world implementation of the shared control system demonstrates its effectiveness in object grasping and manipulation tasks. The results suggest that shared control can significantly improve grasping efficiency while maintaining adaptability in task execution for practical robotics applications.

HighlightsThree bulky cation chlorides (PMACl, PEACl and NMACl) are used to modify the perovskite surface and form pure-anion 2D (PMA)2PbCl4, mixed-anion 2D (PEA)2Pb(IxCl4-x), and non-2D NMAI passivation layers, respectively.Intermolecular interactions between the bulky cations and the strength of cation-halide hydrogen bonds are critical to forming the three distinct passivation layers.Semi-transparent wide-bandgap perovskite solar cells (WBG-PSCs) with ITO as the back electrode show hysteresis-free PCE of 18.60% and VOC deficit of 0.49 V.Wide-bandgap (WBG) perovskite solar cells suffer from severe non-radiative recombination and exhibit relatively large open-circuit voltage (VOC) deficits, limiting their photovoltaic performance. Here, we address these issues by in-situ forming a well-defined 2D perovskite (PMA)2PbCl4 (phenmethylammonium is referred to as PMA) passivation layer on top of the WBG active layer. The 2D layer with highly pure dimensionality and halide components is realized by intentionally tailoring the side-chain substituent at the aryl ring of the post-treatment reagent. First-principle calculation and single-crystal X-ray diffraction results reveal that weak intermolecular interactions between bulky PMA cations and relatively low cation-halide hydrogen bonding strength are crucial in forming the well-defined 2D phase. The (PMA)2PbCl4 forms improved type-I energy level alignment with the WBG perovskite, reducing the electron recombination at the perovskite/hole-transport-layer interface. Applying this strategy in fabricating semi-transparent WBG perovskite solar cells (indium tin oxide as the back electrode), the VOC deficits can be reduced to 0.49 V, comparable with the reported state-of-the-art WBG perovskite solar cells using metal electrodes. Consequently, we obtain hysteresis-free 18.60%-efficient WBG perovskite solar cells with a high VOC of 1.23 V.

Mineral carbonation, which precipitates dissolved carbon dioxide (CO 2 ) as carbonate minerals in basaltic groundwater environments, is a potential technique for negative emissions. The Leizhou Peninsula in southwest Guangdong province has extensive basalt, indicating a promising potential for CO 2 storage through rapid mineralization. However, understanding of the basic geological setting, potential, and mechanisms of CO 2 mineralization in the basalts of the Leizhou Peninsula is still limited. The mineralization processes associated with CO 2 storage at two candidate sites in the area are investigated in this paper: Yongshi Farm and Tianyang Basin (of the dried maar lake). Petrography, rock geochemistry, basalt petrophysical properties, and groundwater hydrochemistry analyses are included in the study. Numerical simulation is used to examine the reaction process and its effects. The results show that basalts in the study areas mainly comprise plagioclase, pyroxene, and Fe–Ti oxides, revealing a total volume fraction exceeding 85%. Additionally, small amounts of quartz and fayalite are available, with volume fractions of 5.1% and 1.0%, respectively. The basalts are rich in divalent metal cations, which can form carbonate minerals, with an average of approximately 6.2 moles of metal cations per 1 kg of rock. The groundwater samples have a pH of 7.5–8.2 and are dominated by the Mg–Ca–HCO 3 type. The basalts demonstrate a porosity range of 10.9% to 28.8%, with over 70% of interconnected pores. A 20-year geochemical simulation revealed that CO 2 injection dissolves primary minerals, including anorthite, albite, and diopside, while CO 2 mineralization dissolves precipitation secondary minerals, such as calcite, siderite, and dolomite. Furthermore, a substantial rise in pH from 7.6 to 10.6 is observed in the vicinity of the injected well, accompanied by a slight reduction in porosity from 20% to 19.8%. Additionally, 36.8% of the injected CO 2 underwent complete mineralization within five years, revealing an increasing percentage of 66.1% if the experimental period is extended to 20 years. The presence of abundant divalent metal cations in basalts and water-bearing permeable rocks in the Leizhou Peninsula supports the potential for mineral carbonation in basalts, as indicated by the geochemical simulation results. Additional research is necessary to identify the factors that influence the CO 2 mineralization, storage, and sensitivity analysis of basalt in the Leizhou Peninsula.

Robotic racket sports provide exceptional benchmarks for evaluating dynamic motion control capabilities in robots. Due to the highly non-linear dynamics of the shuttlecock, the stringent demands on robots' dynamic responses, and the convergence difficulties caused by sparse rewards in reinforcement learning, badminton strikes remain a formidable challenge for robot systems. To address these issues, this study proposes DTG-IRRL, a novel learning framework for badminton strikes that integrates imitation-relaxation reinforcement learning with dynamic trajectory generation. The framework demonstrates significantly improved training efficiency and performance, achieving faster convergence and twice the landing accuracy. Analysis of the reward function within a specific parameter space hyperplane intuitively reveals the convergence difficulties arising from the inherent sparsity of rewards in racket sports and demonstrates the framework's effectiveness in mitigating local and slow convergence. Implemented on hardware with zero-shot transfer, the framework achieves a 90% hitting rate and a 70% landing accuracy, enabling sustained humanrobot rallies. Cross-platform validation using the UR5 robot demonstrates the framework's generalizability while highlighting the requirement for high dynamic performance of robotic arms in racket sports.