Two-dimensional(2D)transition-metal dichalcogenides(TMDCs)have quickly become key in the development of next-generation semiconductor technologies.

The epitaxial growth of wafer-scale single-crystal two-dimensional transition metal dichalcogenides (TMDs) is essential for advancing beyond-silicon electronics. While the c-plane of sapphire has been the standard substrate for epitaxy, exploiting the richness of crystallographic planes may help understand universal mechanisms governing van der Waals epitaxy. The vast spectrum of crystallographic planes, which offer diverse surface symmetries and atomic configurations, remains underexplored due to the complex potential energy landscape at each epitaxial interface. Here, we develop a theoretical framework to explore the vast sapphire crystallographic planes for deterministic MoS2 epitaxy. By evaluating interfacial strain and surface energy as key descriptors across 33 crystallographic planes, we identify five optimal candidates, namely, A(112̅0) C(0001), P(112̅3), R(11̅02), and S(11̅01) We establish that unidirectional alignment is governed by reducing the surface symmetry of the substrate, achieved either through the plane's intrinsic low symmetry (P, R, S) or via engineered step-edges on high-symmetry surfaces (C, A). This universal principle enables the successful growth of wafer-scale single-crystal MoS2 on all five predicted planes.

Interlayer stacking is an important degree of freedom to tune the properties of two-dimensional materials and offers enormous opportunities for designing functional devices. As a classic example, rhombohedral-stacked (3R) two-dimensional materials exhibit ferroelectricity and optical nonlinearity that are non-existent in naturally abundant hexagonal-stacked (2H) counterparts. However, the ability to grow stacking-controlled large-area films remains challenging due to the thermodynamic competition of different polytypes. Here we report the chemical vapour deposition growth of two-inch wafer-scale 3R-MoS2 films with high phase purity by homoepitaxy on top of a crystalline monolayer MoS2. A defect-promoted nucleation mechanism was proposed, in which Mo-substituted sulfur vacancy is identified as one of the possible defects promoting 3R stacking. We fabricate ferroelectric semiconductor field-effect transistors with 3R-MoS2 channels and demonstrate non-volatile memory characteristics. The control of interlayer stacking is an essential step towards the large-scale production of two-dimensional materials for multifunctional integration.

Transition metal dichalcogenides are a potential alternative to silicon and could be used to create transistors with a contacted gate pitch below 40 nm as required by the ångström-node transistor technology. However, it remains challenging to maintain an ohmic contact when the contact length is reduced to less than 20 nm. Here we show that crystalline semi-metallic antimony contacts can be epitaxially grown on molybdenum disulfide (MoS2) by molecular beam epitaxy, creating ohmic contacts with a resistance of 98 Ω µm at a contact length of 18 nm. We use the contacts to build scaled field-effect transistors with a contacted gate pitch of 40 nm with drive currents of 0.85 mA µm−1, 0.95 mA µm−1 and 1.08 mA µm−1 for monolayer, bilayer and trilayer MoS2 channels, respectively. Statistical analysis of transistor arrays confirms that the crystalline antimony contacts are reproducible and stable. Semi-metallic single crystals of antimony can be deposited using molecular beam epitaxy on molybdenum disulfide to create ohmic contacts with resistance of under 100 Ω µm at a contact length of 18 nm.

The fine-grained dynamic sparsity in biological synapses is an important element in the energy efficiency of the human brain. Emulating such sparsity in an artificial system requires off-chip memory indexing, which has a considerable energy and latency overhead. Here, we report an in-memory sparsity architecture in which index memory is moved next to individual synapses, creating a sparse neural network without external memory indexing. We use a compact building block consisting of two non-volatile ferroelectric field-effect transistors acting as a digital sparsity and an analogue weight. The network is formulated as the Hadamard product of the sparsity and weight matrices, and the hardware, which is comprised of 900 ferroelectric field-effect transistors, is based on wafer-scale chemical-vapour-deposited molybdenum disulfide integrated through back-end-of-line processes. With the system, we demonstrate key synaptic processes—including pruning, weight update and regrowth—in an unstructured and fine-grained manner. We also develop a vectorial approximate update algorithm and optimize training scheduling. Through this software–hardware co-optimization, we achieve 98.4

We report on the first monolayer MoS2 FETs scaled to 40 nm contacted gate pitch (CGP), corresponding to 1 nm node as defined by IRDS. The FETs deliver drive current of 0.79 $\text{mA}/\mu \mathrm{m}$ under 0.6 V drain bias with on/off ratio $> 10^{7}$ and SS = 62 mV/dec. We achieve this by developing a new Sb deposition technology which reduces $R_{\mathrm{c}}$ to below 100 $\Omega\cdot\mu \mathrm{m}$ at deep $L_{\mathrm{c}}$ scaling regime $(L_{\mathrm{c}}=20\ \text{nm})$. TCAD simulations predict that quad-stacked Mos2 nanosheet FET employing the new Sb contact reduces the gate delay by 46.6% compared to Si CMOS at 1 nm node and demonstrates superior scaling potential to the end of roadmap.

The prediction of load-dependent losses of gear pairs is a topic of constant concern, which depends largely on the modelling of friction coefficient in tooth meshing. Although the current load-sharing based models can handle the friction coefficient in mixed elastohydrodynamic lubrication regimes, dynamics behaviour of each tooth pair or the time variation of direction of normal contact load is not taken into consideration. This study proposes a new method for modelling efficiency of spur gear pairs in transient operating conditions by coupling the multi-tooth meshing model with the friction coefficient model. The friction coefficient is modelled through load-sharing function, and the gear dynamics model is established with regard to alternate meshing of single-double tooth pair and time variation of direction of normal load and mesh stiffness. The model is validated by comparing the calculated results of friction coefficients and load-dependent losses with experimental data in published literatures, showing good agreement. At the end, the efficiency model is applied to two types of gear pairs to investigate the influence of tooth shapes and operating parameters on the load-dependent losses of gear pairs.

Single-crystalline transition metal dichalcogenides (TMD) films are of potential application in future electronics and optoelectronics. In this work, a halide vapor phase epitaxy (HVPE) strategy was proposed and demonstrated for the epitaxy of molybdenum diselenide (MoSe2) single crystals, in which metal halide vapors were in-situ produced by the chlorination of molybdenum as sources for the TMD growth. Combined with the epitaxial sapphire substrate, unidirectional domain alignment was successfully achieved and monolayer single-crystal MoSe2 films have been demonstrated on a 2-inch wafer for the first time. A series of characterizations ranging from centimeter to nanometer scales have been implemented to demonstrate the high quality and uniformity of the MoSe2. This work provides a universal strategy for the growth of TMD single-crystal films.

Nitroanilines are important building blocks in pharmaceuticals, materials and dyes. Nitration methods for anilines under mild conditions are highly desired. Herein, we report a photochemical method for the nitration of anilines bearing various protecting groups by 5-methyl-1,4-dinitroimidazole as a new type of nitro source. This method is light-controlled and proceeds under mild reaction conditions with high efficiency. Fmoc-, Ts- and alkyl-protected anilines are all well nitrated with good functional group tolerance.

Two-dimensional transition metal dichalcogenides could potentially be used to create transistors that are scaled beyond the capabilities of silicon devices. However, despite progress on the single-transistor level, the development of high-frequency integrated circuits remains a challenge and the operating frequency of integrated circuits based on transition metal dichalcogenides has so far been limited to the megahertz regime; this is well below the silicon complementary metal–oxide–semiconductor technology, as well as emerging technologies such as carbon nanotubes. Here we report two-dimensional semiconductor integrated circuits—five-stage ring oscillators—that operate in the gigahertz regime (up to 2.65 GHz) and are developed using a design-technology co-optimization process. The circuits are based on monolayer molybdenum disulfide field-effect transistors that have an air-gap structure, which leads to doping-free ohmic contacts and low parasitic capacitance. Technology computer-aided design simulations also suggest that our air-gap structure can potentially be scaled to the 1 nm technology node and could reach the targets set out in the IEEE International Roadmap for Devices and Systems for 2031.

The development of next-generation electronics requires scaling of channel material thickness down to the two-dimensional limit while maintaining ultralow contact resistance 1 , 2 . Transition-metal dichalcogenides can sustain transistor scaling to the end of roadmap, but despite a myriad of efforts, the device performance remains contact-limited 3 – 12 . In particular, the contact resistance has not surpassed that of covalently bonded metal–semiconductor junctions owing to the intrinsic van der Waals gap, and the best contact technologies are facing stability issues 3 , 7 . Here we push the electrical contact of monolayer molybdenum disulfide close to the quantum limit by hybridization of energy bands with semi-metallic antimony ( 011̅2 ) through strong van der Waals interactions. The contacts exhibit a low contact resistance of 42 ohm micrometres and excellent stability at 125 degrees Celsius. Owing to improved contacts, short-channel molybdenum disulfide transistors show current saturation under one-volt drain bias with an on-state current of 1.23 milliamperes per micrometre, an on/off ratio over 10 8 and an intrinsic delay of 74 femtoseconds. These performances outperformed equivalent silicon complementary metal–oxide–semiconductor technologies and satisfied the 2028 roadmap target. We further fabricate large-area device arrays and demonstrate low variability in contact resistance , threshold voltage, subthreshold swing, on/off ratio, on-state current and transconductance 13 . The excellent electrical performance, stability and variability make antimony ( 011̅2 ) a promising contact technology for transition-metal-dichalcogenide-based electronics beyond silicon.

The tooth form is the most influential factor on the bearing capacity and operation performance of a gear drive. Tooth profiles of spur gears can be designed based on control of relative curvature of conjugate profiles, but how to conceive appropriate control strategies and appraise gear pairs accurately needs to be investigated in depth. This paper presents the control strategies through analysis of the effect mechanisms of relative curvature on the bearing capacity and lubrication performance. Evaluation approaches of the gear pairs are then proposed by establishing the analytical models (AMs). In case studies, gear pairs with optimal comprehensive performance are designed through five control strategies. The AMs for contact and bending capacity are validated with the finite element models. The effects of the control strategies are discussed through comparative analysis. Results reveal that the control strategies lead to remarkable enhancement of bearing capacity and lubrication performance.

The growing computational demand in artificial intelligence calls for hardware solutions that are capable of in situ machine learning, where both training and inference are performed by edge computation. This not only requires extremely energy-efficient architecture (such as in-memory computing) but also memory hardware with tunable properties to simultaneously meet the demand for training and inference. Here we report a duplex device structure based on a ferroelectric field-effect transistor and an atomically thin MoS 2 channel, and realize a universal in-memory computing architecture for in situ learning. By exploiting the tunability of the ferroelectric energy landscape, the duplex building block demonstrates an overall excellent performance in endurance (>10 13 ), retention (>10 years), speed (4.8 ns) and energy consumption (22.7 fJ bit –1 μm –2 ). We implemented a hardware neural network using arrays of two-transistors-one-duplex ferroelectric field-effect transistor cells and achieved 99.86% accuracy in a nonlinear localization task with in situ trained weights. Simulations show that the proposed device architecture could achieve the same level of performance as a graphics processing unit under notably improved energy efficiency. Our device core can be combined with silicon circuitry through three-dimensional heterogeneous integration to give a hardware solution towards general edge intelligence.

Two-dimensional (2D) semiconductors are promising channel materials for next-generation field-effect transistors (FETs). However, it remains challenging to integrate ultrathin and uniform high-κ dielectrics on 2D semiconductors to fabricate FETs with large gate capacitance. We report a versatile two-step approach to integrating high-quality dielectric film with sub-1 nm equivalent oxide thickness (EOT) on 2D semiconductors. Inorganic molecular crystal Sb2O3 is homogeneously deposited on 2D semiconductors as a buffer layer, which forms a high-quality oxide-to-semiconductor interface and offers a highly hydrophilic surface, enabling the integration of high-κ dielectrics via atomic layer deposition. Using this approach, we can fabricate monolayer molybdenum disulfide-based FETs with the thinnest EOT (0.67 nm). The transistors exhibit an on/off ratio of over 106 using an ultra-low operating voltage of 0.4 V, achieving unprecedently high gating efficiency. Our results may pave the way for the application of 2D materials in low-power ultrascaling electronics. A van der Waals buffer layer of Sb2O3 enables the integration of high-κ dielectric layer with sub-1 nm equivalent oxide thickness on two-dimensional semiconductors, resulting in high performance of two-dimensional field-effect transistors.

Abstract The growing computational demand in artificial intelligence (AI) calls for hardware solutions that are capable of in-situ machine learning, where both training and inference are performed by edge computation. This not only requires extremely energy-efficient architecture (such as in-memory computing, IMC) but also memory hardware with tunable properties to simultaneously meet the demand for training and inference. Here, we report a duplex device structure based on ferroelectric field-effect transistor (FeFET) and atomically thin MoS2 channel and realize a universal IMC architecture for in-situ learning. By exploiting the tunability of ferroelectric energy landscape, the duplex building block demonstrates overall excellent performance in endurance (>1013), retention (>10 years), speed (4.8 ns) and energy consumption (22.7 fJ/(bit·μm2)). We implemented a hardware neural network using arrays of two-transistor-one-duplex-FeFET (2T1D) cells and achieved 99.86% accuracy in non-linear localization task with in-situ trained weights. Simulations show that the proposed device architecture could achieve the same level of performance as graphics processing unit under notably improved energy efficiency. Our device core can be combined with silicon circuitry through three-dimensional heterogeneous integration to give a hardware solution toward general edge intelligence (EI).

Icing and anti-icing tests were carried out on a rotating conical spinner in an icing wind tunnel under almost identical incoming flow conditions including temperature and LWC and MVD of supercooled water droplets but with different mass-flow rates of hot air. During the icing test, the ice accretion process was recorded, and the results indicate that milky white rime ice accretes on the entire spinner, with its shape conforming to that of the spinner’s surface. In the tip area, the ice is thick, gradually decreasing in thickness along the generatrix; in the rear part of the spinner, ice thickness increases. The icing rate is almost constant with time. In the hot-air anti-icing test the surface temperature of the rotating spinner was measured using an infrared thermal camera calibrated with thermocouples, and the distribution of water film on the spinner surface was also measured. The results show that after spaying begins the surface temperature first decreases rapidly and then drops slowly with time. From about t =60s, the surface temperature of the spinner tends to stability. Along the spinner generatrix, the surface temperature increases first and then decreases, reaching its maximum near the tip. The water film appears on the rear of the spinner, while the front of the spinner remains dry. From about t = 40s the coverage and flow of water film remain almost constant with time. In particular, when the mass-flow rate of the hot air is low, the spinner surface is completely covered by the water film, while there is a complete evaporation area around the tip as the mass-flow rate is high.

Transition metal-catalyzed C-H activation is a step-economical strategy for peptide functionalization. Herein, we report the method of late-stage peptide ligation and macrocyclization through rhodium-catalyzed alkylation of tryptophan residues at the C7 position. This method utilizes a N-Pt Bu2 directing group and tolerates various peptide and alkene substrates. Utilizing internal olefins, this study represents the first example of site-selective peptide C-H alkylation through deconjugative isomerization. Furthermore, our method provides access to peptide macrocycles with unique Trp(C7)-alkyl crosslinks and potent cytotoxicity towards cancer cells.

Two-dimensional transition-metal dichalcogenides (TMDs) are of interest for beyond-silicon electronics 1 , 2 . It has been suggested that bilayer TMDs, which combine good electrostatic control, smaller bandgap and higher mobility than monolayers, could potentially provide improvements in the energy-delay product of transistors 3 – 5 . However, despite advances in the growth of monolayer TMDs 6 – 14 , the controlled epitaxial growth of multilayers remains a challenge 15 . Here we report the uniform nucleation (>99%) of bilayer molybdenum disulfide (MoS 2 ) on c -plane sapphire. In particular, we engineer the atomic terrace height on c -plane sapphire to enable an edge-nucleation mechanism and the coalescence of MoS 2 domains into continuous, centimetre-scale films. Fabricated field-effect transistor (FET) devices based on bilayer MoS 2 channels show substantial improvements in mobility (up to 122.6 cm 2 V −1 s −1 ) and variation compared with FETs based on monolayer films. Furthermore, short-channel FETs exhibit an on-state current of 1.27 mA μm −1 , which exceeds the 2028 roadmap target for high-performance FETs 16 .