Ferroelectric materials have switchable electrical polarization that is appealing for high-density nonvolatile memories. However, inevitable fatigue hinders practical applications of these materials. Fatigue-free ferroelectric switching could dramatically improve the endurance of such devices. We report a fatigue-free ferroelectric system based on the sliding ferroelectricity of bilayer 3R molybdenum disulfide (3R-MoS2). The memory performance of this ferroelectric device does not show the wake-up effect at low cycles or a substantial fatigue effect after 106 switching cycles under different pulse widths. The total stress time of the device under an electric field is up to 105 s, which is long relative to other devices. Our theoretical calculations reveal that the fatigue-free feature of sliding ferroelectricity is due to the immobile charge defects in sliding ferroelectricity.

Exploiting the excellent electronic properties of two-dimensional (2D) materials to fabricate advanced electronic circuits is a major goal for the semiconductor industry1,2. However, most studies in this field have been limited to the fabrication and characterization of isolated large (more than 1 µm2) devices on unfunctional SiO2–Si substrates. Some studies have integrated monolayer graphene on silicon microchips as a large-area (more than 500 µm2) interconnection3 and as a channel of large transistors (roughly 16.5 µm2) (refs. 4,5), but in all cases the integration density was low, no computation was demonstrated and manipulating monolayer 2D materials was challenging because native pinholes and cracks during transfer increase variability and reduce yield. Here, we present the fabrication of high-integration-density 2D–CMOS hybrid microchips for memristive applications—CMOS stands for complementary metal–oxide–semiconductor. We transfer a sheet of multilayer hexagonal boron nitride onto the back-end-of-line interconnections of silicon microchips containing CMOS transistors of the 180 nm node, and finalize the circuits by patterning the top electrodes and interconnections. The CMOS transistors provide outstanding control over the currents across the hexagonal boron nitride memristors, which allows us to achieve endurances of roughly 5 million cycles in memristors as small as 0.053 µm2. We demonstrate in-memory computation by constructing logic gates, and measure spike-timing dependent plasticity signals that are suitable for the implementation of spiking neural networks. The high performance and the relatively-high technology readiness level achieved represent a notable advance towards the integration of 2D materials in microelectronic products and memristive applications.

Ultra-scaled transistors are of interest in the development of next-generation electronic devices. Although atomically thin molybdenum disulfide (MoS2) transistors have been reported, the fabrication of devices with gate lengths below 1 nm has been challenging. Here we demonstrate side-wall MoS2 transistors with an atomically thin channel and a physical gate length of sub-1 nm using the edge of a graphene layer as the gate electrode. The approach uses large-area graphene and MoS2 films grown by chemical vapour deposition for the fabrication of side-wall transistors on a 2-inch wafer. These devices have On/Off ratios up to 1.02 × 105 and subthreshold swing values down to 117 mV dec–1. Simulation results indicate that the MoS2 side-wall effective channel length approaches 0.34 nm in the On state and 4.54 nm in the Off state. This work can promote Moore’s law of the scaling down of transistors for next-generation electronics.

Two-dimensional (2D) materials could potentially be used to develop advanced monolithic integrated circuits. However, despite impressive demonstrations of single devices and simple circuits—in some cases with performance superior to those of silicon-based circuits—reports on the fabrication of integrated circuits using 2D materials are limited and the creation of large-scale circuits remains in its infancy. Here we examine the development of integrated circuits based on 2D layered materials. We assess the most advanced circuits fabricated so far and explore the key challenges that need to be addressed to deliver highly scaled circuits. We also propose a roadmap for the future development of integrated circuits based on 2D layered materials.

Tin-based perovskites, renowned for their eco-friendliness, intrinsic high hole mobility, and low effective mass, hold great potential for p-type thin-film transistors (TFTs). However, their propensity for rapid crystallization and oxidation severely limits stability and carrier mobility. Here, we strategically enhance perovskite TFT performance by incorporating 2-thiopheneethylamine thiocyanate (TEASCN) into 3D tin-based perovskites. The induction of the pseudo-halide SCN− into a bilayer quasi-2D perovskite intermediate phase, combined with the strong interaction between sulfur-bearing thiophene rings (TEA+) and Sn-I octahedra, effectively reorients perovskite crystallization while inhibiting Sn2+ oxidation and reducing trap density. Consequently, TEASCN-based TFTs achieve an average hole mobility of more than 60 square centimeters per volt per second and an on/off current ratio surpassing 108, standing out among state-of-the-art p-type perovskite TFTs. Furthermore, unencapsulated devices preserve 84% of their initial mobility after 30 days in an N2 atmosphere, underscoring their remarkable stability. This work opens a straightforward path toward high-mobility and highly stable tin-based perovskite transistors.

The non-volatile spontaneous ferroelectric polarization field serves as a cornerstone for applying ferroelectric materials in electronic devices, yet it is frequently mitigated by charge trapping at defect sites. Achieving an effective transition between ferroelectric polarization and charge trapping is challenging due to the inherent opposition of the two mechanisms and the uncontrollable charge trapping types in ferroelectric materials. Here, we realized a polarity-dependent ferroelectric transition in two-dimensional ferroelectric heterojunction transistor by integrating a hybrid organic-inorganic ferroelectric layer embedded with electron trapping sites. Through theoretical calculations and experimental validation, we demonstrate a ferroelectric manifestation and elimination mechanism based on the polarity of the semiconductor layer. The electron-majority n-type semiconductor exhibits charge trapping behavior, while the electron-minority p-type transistor exhibits the ferroelectric control mechanism. Leveraging the mechanism transition, our bipolar heterojunction transistor enables synergistic heterogeneous control of non-volatile memory and volatile synaptic weight modulation within a single bipolar ferroelectric transistor. Based on the experimentally extracted parameters from the transistors, the device-informed simulation achieves a recognition accuracy of 92.9% and a 20.7-fold improvement in training efficiency of the transfer learning network.

Quasi-2D tin-based perovskites are promising p-type semiconductors due to their thermodynamic stability and suppressed ion migration tendencies. However, the competitive growth of low- and high-dimensional phases leads to pronounced structural disorder, increased defect density, and poor crystallographic orientation, thereby restricting charge transport. Here, phenethylammonium thiocyanate (PEASCN) is incorporated into the precursor to promote the preferential formation of PEA2FAn-1SnnI3n-1SCN2 (n = 2) templates. Substituting formamidinium iodide (FAI) with formamidinium formate (FAHCOO) and ammonium iodide (NH4I) suppresses the uncontrollable growth of 3D FASnI3 at room temperature, enabling precise crystallization control. These low-dimensional templates guide the growth of high-dimensional phases upon annealing, yielding vertically oriented films with reduced defects. The fabricated field-effect transistors exhibit mobility up to 43 cm2 V−1 s−1 and an on/off ratio exceeding 108, alongside nearly negligible hysteresis and enhanced stability. These results demonstrate a viable approach for regulating crystallization kinetics and realizing high-performance, stable tin-based perovskite devices.

Ferroelectric semiconductors have the advantages of switchable polarization ferroelectric field regulation and semiconductor transport characteristics, which are highly promising in ferroelectric transistors and nonvolatile memory. However, it is difficult to prepare a Sn-based perovskite film with both robust ferroelectric and semiconductor properties. Here, by doping with 2-methylbenzimidazole, Sn-based perovskite [93.3 mol% (FA0.86Cs0.14)SnI3 and 6.7 mol% PEA2SnI4] semiconductor films are transformed into ferroelectric semiconductor films, owing to molecular reconfiguration. The reconfigured ferroelectric semiconductors exhibit a high remanent polarization (Pr) of 23.2 μC/cm2. The emergence of ferroelectricity can be ascribed to the hydrogen bond enhancement after imidazole molecular doping, and then the spatial symmetry breaks causing the positive and negative charge centers to become non-coincident. Remarkably, the transistors based on perovskite ferroelectric semiconductors have a low subthreshold swing of 67 mv/dec, which further substantiates the superiority of introducing ferroelectricity. This work has developed a method to realize Sn-based ferroelectric semiconductor films for electronic device applications.

Accelerating the discovery of novel crystal materials by machine learning is crucial for advancing various technologies from clean energy to information processing. The machine-learning models for prediction of materials properties require embedding atomic information, while traditional methods have limited effectiveness in enhancing prediction accuracy. Here, we proposed an atomic embedding strategy called universal atomic embeddings (UAEs) for their broad applicability as atomic fingerprints, and generated the UAE tensors based on the proposed CrystalTransformer model. By performing experiments on widely-used materials database, our CrystalTransformer-based UAEs (ct-UAEs) are shown to accurately capture complex atomic features, leading to a 14% improvement in prediction accuracy on CGCNN and 18% on ALIGNN when using formation energies as the target, based on the Materials Project database. We also demonstrated the good transferability of ct-UAEs across various databases. Based on the clustering analysis for multi-task ct-UAEs, the elements in the periodic table can be categorized with reasonable connections between atomic features and targeted crystal properties. After applying ct-UAEs to predict formation energy in hybrid perovskites database, we realized an improvement in accuracy, with a 34% boost in MEGNET and 16% in CGCNN, showcasing their potential as atomic fingerprints to address the data scarcity challenges.

The polarization of HfO2-based ferroelectrics originates from the metastable orthorhombic phase formed during the tetragonal to monoclinic phase transition and is typically controlled by tuning the phase content. However, another way to control polarization via modulating ferroelectric domain orientations remains underexplored. This work uncovers a hidden tetragonal-orthorhombic phase transition pathway to engineer domain orientations and further polarization in polycrystalline Hf0.5Zr0.5O2 using single-crystalline TiN substrates. Specifically, (001)O and/or (010)O domains, which fully contribute to remanent polarization under an electric field, are controllable in Hf0.5Zr0.5O2 on TiN (001) and (111), enhancing remanent polarization compared to that on TiN (110). The key is the hidden transition from the tetragonal phase’s longest c-axis to the orthorhombic phase’s shorter bO/cO-axis, alongside the reported one to the longest aO-axis, assisted by periodic dislocations at the TiN/Hf0.5Zr0.5O2 interface. These findings shed light on governing the polarization of Hf0.5Zr0.5O2 films by controlling the interface dislocations and further domain orientations.