Relevant technological advancements in recent years have established solar sailing as an attractive propulsion system for terrestrial and interplanetary space missions, particularly missions that require low-thrust propulsion for long periods of time. The initial stages in designing solar sail missions involve a thorough analysis of possible transfer trajectories. Traditional approaches primarily focus on deterministic scenarios where the spacecraft dynamics is precisely defined and fully characterized. Whereas, space trajectories are inevitably subject to various sources of uncertainty, including modeling errors in the dynamics, inaccurate measurements, and unpredictable environmental effects, such as fluctuations in solar irradiance. These uncertainties necessitate the development of robust control strategies that can effectively handle stochastic effects. In recent years, reinforcement learning (RL) offers a promising framework for addressing optimization problems characterized by significant uncertainty, which can be challenging to handle using traditional methods. Meanwhile, deep reinforcement learning (DRL) utilizes neural networks (NNs) to approximate complex policies, i.e. control laws. 

Einstein’s general theory of relativity reveals that spacetime can be warped by mass and energy. As a conjecture in this context, the evolution of space-time curvature affected by gravitation will propagate through the vast universe like ripples. There remains a wealth of mysteries about astrophysics that have not been explored yet, and these “ripples” called gravitational waves may revolutionize the investigation for this. Space-based gravitational-wave observations, which broadly adopt heliocentric, geocentric, or libration-point architectures, can extend the scope to celestial bodies that generate valuable millihertz frequency waves and escape the Earth-bound noise. The heliocentric gravitational-wave detection relying on laser interferometry faces dual challenges: 1) the attitude errors couple into the interferometric arm-length measurements, potentially corrupting the picometer-level displacement sensitivity, making the attitude accuracy exceptionally rigorous, while 2) the ultra-long interferometric arms (and thus million-kilometer-scale intersatellite communication distances) introduce significant communication delays. As a key to space-based gravitational-wave observations, the past few years have witnessed the development of attitude-orbit control technique, and numerous control methodologies have sprung up for various space missions. However, current communication delay models and their corresponding compensation methods prove inadequate for heliocentric gravitational-wave detection scenarios. Additionally, in line with adaptive SMC, there are few investigations on attitude-orbit coordinated control for heliocentric gravitational-wave detection spacecraft formation. 

As we venture deeper into space and establish bases and orbital stations on celestial bodies, it is imperative to develop sustainable space transportation methods that minimise propellant usage. Currently, the most common method for Earth-Mars transfers is through chemical propulsion systems. While the chemical propulsion system is well established, it has several drawbacks, including low efficiency, high propellant mass requirements, thermal limitations, and operational complexity. Momentum exchange tethers, which, in theory, can operate without the need for any propellant, offer an innovative alternative to these propulsion systems. Especially, in a symmetrical tether, the COM lies at the centre of the length. If both the equal mass payloads are released at the same time, then the orbit of the tether does not change. Additional delta-v can be generated by the libration (oscillation) or rotation of the tethers, which can be induced by a motor facility. Though there are various investigations into space tethers, there is a gap in the use of symmetrical tethers for interplanetary transfers. 

Sudden surface anomalies (SSAs)—ranging from earthquakes and floods to wildfires and industrial accidents—pose escalating threats to ecosystems and societies worldwide. With its advantages of wide coverage, high revisit frequency, and rich information content, remote sensing is regarded as an effective tool for rapid detection and early warning of such anomalies. However, significant technical bottlenecks remain in using remote sensing to monitor SSAs, such as delays due to revisit intervals and complex data processing chains and challenges for the integration and joint analysis of multi-source information, resulting in frequent occurrences of missed detections and false alarms. Real-time remote sensing has emerged as a transformative approach for monitoring and early warning of such abrupt events, driven by recent advances in satellite constellations, on-orbit artificial intelligence, and multi-source data fusion.

This perspective article proposes a paradigm shift from single-discipline pathogen control to an integrated, cross-disciplinary approach that combines earth science and environmental engineering. Led by Zheng-Yang Huo from Renmin University of China, an research team outlines how geochemical insights and advanced disinfection technologies can work together to achieve full-chain assessment and control of waterborne pathogens, addressing a persistent threat that causes over 10 million deaths annually worldwide.

Achieving regioselective C–H functionalization is a particularly difficult issue in organic chemistry when these C–H bonds have similar reactivity. Although significant progress has been made in regioselective C–H functionalization at one specific site, the tunable version of such processes is rarely documented. Here we describe a tunable meta- and para-selective C–H borylation of aryl and benzyl phosphonates, in which the regioselectivity can be switched by bifunctional ligands. A broad range of substrates are compatible and exhibit good to excellent regioselectivity under mild reaction conditions. Benzyl phosphonates can be conveniently transformed into styrene derivatives via Horner–Wadsworth–Emmons olefination. Gram-scale preparation and derivatization of borylated products demonstrate the important application of our method in synthetic chemistry. The orthogonality of two functional Bpin and phosphorylmethyl groups provides ample opportunity for increasing the diversity of styrene derivatives with precise site-control.

Organic clusters composed of heteroatoms and functional groups have been extensively employed to rationalize the intrinsic luminescence of nonconventional luminophores lacking aromatic structures. Nevertheless, in marked contrast to well-studied metal clusters, their structures and sizes remain poorly delineated, substantially impeding rigorous mechanistic understanding and rational performance optimization. Here, we successfully overcome this limitation by employing 3D-shaped cyclic hydrocarbon scaffolds to confine electron-rich moieties, in analogy to the ligand–metal coordination motif in metal clusters, thereby suppressing inter-cluster extended through-space conjugation (TSC) networks. This strategy enables, for the first time, the construction of organic clusters with a programmable and precisely defined number of electron-rich units (e.g., oxygen atoms). Our findings demonstrate that organic clusters with as few as four oxygen atoms can generate excitation-dependent emission with green afterglow, implying the presence of multiple emissive species with varied TSC even within these well-defined isolated clusters, likely associated with the intrinsic complexity and sensitivity of through-space electronic interactions. This strategy thus establishes a robust platform for excluding impurity-induced emission, precisely identifying organic cluster structures/sizes, and moreover providing a basis for correlating chemical structure, electronic configuration, and photophysical properties in nonconventional luminophores.

Mechanical force redistributes reactivity through anisotropic, topology-dependent load transmission. Predicting which covalent bond in a complex molecule ruptures first under tension remains a central problem in mechanochemistry, biomaterials and chemically recyclable polymers. Bond dissociation energy is an incomplete proxy because it describes zero-force thermodynamic stability, whereas mechanochemical site selectivity is governed more directly by the peak force sustained along a prescribed loading path. Here we introduce covalent-bond peak force network (CBPFNet), a graph-attention model that predicts covalent-bond peak force (CBPForce) from relaxed molecular structures through a virtual-stretching workflow and avoids the hundreds of constrained geometry optimisations typically required for each candidate bond. In the CHON benchmark studied here, CBPFNet reproduces bond-ranking trends from density functional theory (DFT) and can be used for bond-level screening. Post hoc attribution analyses indicate that rupture selectivity depends on both bond identity and local stiffness contrast, with relatively rigid segments concentrating tensile stress on softer adjacent bonds. In oligopeptides, the model recovers the DFT weakest bond and highlights a serine-containing motif as a possible topology-dependent weak point under the same protocol. CBPFNet therefore extends screening beyond equilibrium bond stability toward bond-resolved peak-force ranking in sustainable polymer design and biomaterials.