Bilu Liu

Perovskite–2D material heterostructures have evolved from empirical interface engineering motifs to model systems for controlling interfacial photophysics in solution-processed optoelectronics. Graphene, MXenes, transition-metal dichalcogenides (TMDs), black phosphorus, and related layered materials can improve device operation through charge extraction, contact stabilization, recombination suppression, morphology control, or radiative-state management. However, these outcomes cannot be understood from the material identity perspective alone. This review organizes perovskite–2D material interfaces according to the interfacial variable they control and the photoexcited-state pathway they enable. Electrode-facing graphene and MXene contacts primarily regulate tunneling, work-function alignment, carrier extraction, and stability. Semiconducting perovskite/TMD heterostructures introduce finite-bandgap interfaces where selective transfer, interfacial coupling, and excitonic or valley-sensitive responses become important. Structure- and field-mediating interfaces act through growth control, strain release, grain-boundary carrier pathways, or local electrostatic fields rather than through uniform planar charge transfer. We then connect these archetypes to three interfacial photophysical regimes: extraction-dominated, loss-dominated, and coupling-sensitive. Particular emphasis is laid on spectroscopic readout, including time-resolved photoluminescence (TRPL), transient absorption, ultraviolet photoelectron spectroscopy (UPS)/Kelvin probe force microscopy (KPFM), photocurrent dynamics, and second harmonic generation (SHG) or time-resolved SHG, because similar observables can arise from different mechanisms. Finally, we discuss how these pathways determine solar-cell, photodetector, and emission- or gain-device performance as well as outline future directions toward cleaner, more programmable, and more diagnosable perovskite–2D material interfaces.

Twisted bilayer MoS₂ forms a quantum-engineered sensing platform, where the moiré superlattice creates a configurable surface potential for molecular control. Here we present a periodically modulated landscape in bilayer MoS₂ by twist-angle tuning, and use it for molecular trapping and selective sensing. The native 54 meV moiré potential confines aromatic molecules (e.g., rhodamine 6G, R6G; methylene blue, MB) at AA-stacking sites. Introducing extended vacancies deepens confinement to 238 meV, enhancing adsorption. Twist angle enables band alignment with molecules, driving charge-transfer resonances. This integrated design, moiré potentials, defect-enhanced trapping and band-aligned charge transfer, synergistically amplifies Raman signals, achieving single-molecule sensitivity down to 10⁻²⁰ M. Importantly, the platform exhibits molecule-specific selectivity: R6G is optimally detected at 10° and 50° bilayer MoS₂, while MB responds at 15° and 45°, demonstrating twist-angle encoding of molecular recognition. Our twisted MoS₂ with extended vacancies establishes a moiré system as a versatile platform for selective molecular sensing.

Seawater electrolysis is promising for green hydrogen production, while intensified chloride corrosion at high‐current density (HCD) poses significant challenges to the activity and stability of the anodes. Conventional anode catalysts leverage anion electrostatic repulsion to repel chloride but often fail under industrially relevant ampere‐level HCD conditions, restricting their applications. Here we report a corrosion‐resistant cobalt‐vanadium‐oxide catalyst featuring in situ formed, current‐adaptive polyoxovanadates (POVs) for efficient and stable seawater electrolysis at HCD. As the current densities increase, the vanadate monomers dynamically self‐assemble into POVs with significantly increased steric hindrance to Cl ⁻ , thereby enhancing chloride repulsion. The catalyst works stably for over 5000 h at 1000 mA cm ⁻² with a low decay rate of 0.83 µV h ^{− 1} in seawater electrolyte. Furthermore, the anion exchange membrane seawater electrolyzer only needs 1.64 V at 1000 mA cm ⁻² and maintains stable operation for over 1500 h. Technoeconomic analysis results show that H ₂ production cost is $1.17/GGE H ₂ , which surpasses the U.S. Department of Energy's $2.0/GGE H ₂ target. This work reports a novel dynamic chloride repulsion behavior at HCD conditions and paves the way for robust corrosion‐resistant anode catalysts for seawater electrolysis.

2D non‐layered ferrimagnetic materials have opened up opportunities to explore new physics and develop spintronic devices. Among them, 2D Fe ₃ O ₄ with a single magnetic domain has attracted great attention due to its high remanence ratio and data retention ability. Thickness reduction is effective to achieve single magnetic domain structure, which is challenging for non‐layered materials. Here, we established the relationship between the thickness of 2D non‐layered Fe ₃ O ₄ and growth parameters by a Bayesian nonlinear regression model based on machine learning. We studied the effect of concentration, solution volume, temperature, and substrate wettability on the thickness of 2D Fe ₃ O ₄ and found that the combined effect of concentration and substrate wettability has the greatest impact. Importantly, the converged and accurately predicted model suggests the optimal condition for the required thickness. Guided by this prediction, we grew single‐crystalline 2D Fe ₃ O ₄ with a thickness down to 4 nm, exhibiting a single magnetic domain structure. This work obtains single magnetic domain 2D ferrimagnetic materials and provides a feasible model to guide growth with the required thickness, promoting the preparation and application of 2D ferrimagnetic materials.

Electrosynthesis of cyclohexanone oxime, the vital precursor for producing nylon-6, through C−N coupling reaction from cyclohexanone and nitrate serves as an environmentally friendly route to substitute for conventional technology. However, an electrocatalyst with a single active site is not sufficient to realize the co-adsorption and co-activation of various reactants, which is not conducive to the C−N coupling processes, resulting in the unsatisfactory selectivity and Faraday efficiency of cyclohexanone oxime. Here, we present a geometrically ordered PdCu (O-PdCu) catalyst with high-density dual active sites to efficiently synthesize cyclohexanone oxime via co-reduction of cyclohexanone and nitrate. Operando measurements and control experiments reveal that O-PdCu can generate more NH2OH intermediates to react with cyclohexanone, and the geometric structure with highly ordered and high-density Pd-Cu dual active sites is more favorable to the C−N coupling process, thereby enhancing the performance of cyclohexanone oxime electrosynthesis. Consequently, the O-PdCu achieves selectivity in organic products and Faradaic efficiency of 100% and 70.75% at −0.4 V vs RHE, respectively. Furthermore, the O-PdCu catalyst facilitates the production of 2.20 g of cyclohexanone oxime with a yield rate of 4.82 mmol cm−2 h−1 at a current of 1.12 A through a large-scale synthesis process. Technoeconomic analysis further indicates that the electrosynthesis of cyclohexanone oxime in our system represents a feasible and economically profitable strategy.

Transition metal dichalcogenide (TMD) membranes offer a promising solution to freshwater scarcity by desalinating seawater through ion blocking. However, research has largely focused on MoS₂ due to its efficient exfoliation via n-butyl lithium intercalation, unlike other TMDs. Here, we report scalable production of WS₂ nanosheets by increasing the Li⁺ intercalation driving force (increasing current density up to 10 mA g^-1 and decreasing cutoff voltage down to 0.7 V) of the electrochemical intercalation method. WS₂ membranes were then fabricated by organohalide functionalization followed by vacuum filtration. The resulting membranes exhibit a high rejection rate of up to 90% for divalent and trivalent cations. The selectivity for monovalent/divalent and monovalent/trivalent cations reaches up to 8.7 and 9.9, respectively, in both single and binary cation solutions. Theoretical calculations reveal that the improved rejection rates of divalent and trivalent cations compared to those of monovalent cations originate from the stronger binding preferences of monovalent cations with functionalized WS₂ membranes.

Pulsed electrolysis is widely regarded as a promising strategy for enhancing the performance of the electrocatalytic nitrate reduction reaction (NO₃^-RR). However, research on the effect of symmetric and asymmetric pulsed electrolysis on the NO₃^-RR remains elusive. Here, the CuMn alloy serves as a model catalyst to systematically investigate the impact of symmetric and asymmetric pulsed electrolysis on the performance of NO₃^-RR. A series of operando measurements and control experiments indicate that a suitable asymmetric pulsed electrolysis can optimize the local microenvironment of the NO₃^-RR, including optimizing the coverage of NO₃^-, enhancing the generation of key N-containing intermediates, and inhibiting the hydrogen evolution reaction, thereby improving the Faradaic efficiency and yield rate of NH₃. Consequently, the Faradaic efficiency and yield rate of NH₃ for the optimal asymmetric pulsed electrolysis between a less negative potential time of 2 s and a more negative potential time of 4 s are 94.03% and 9.13 mg h^-1 cm^-2, respectively, which are 1.42-fold and 2.10-fold higher than those of static electrolysis. Moreover, the Faradaic efficiency of NH₃ for the optimal asymmetric pulsed electrolysis was 66.32% at a lower concentration of NO₃^- (2000 ppm), which is 1.73-fold higher than that of static electrolysis, indicating that the performance of asymmetric pulsed electrolysis is further enhanced in the lower concentration of NO₃^-, further confirming that asymmetric pulsed electrolysis is a feasible strategy to improve the performance of the NO₃^-RR.