Conventional artificial intelligence (AI) systems are facing bottlenecks due to the fundamental mismatches between AI models, which rely on parallel, in-memory, and dynamic computation, and traditional transistors, which have been designed and optimized for sequential logic operations. This calls for the development of novel computing units beyond transistors. Inspired by the high efficiency and adaptability of biological neural networks, computing systems mimicking the capabilities of biological structures are gaining more attention. Ion-based memristive devices (IMDs), owing to the intrinsic functional similarities to their biological counterparts, hold significant promise for implementing emerging neuromorphic learning and computing algorithms. In this article, we review the fundamental mechanisms of IMDs based on ion drift and diffusion to elucidate the origins of their diverse dynamics. We then examine how these mechanisms operate within different materials to enable IMDs with various types of switching behaviors, leading to a wide range of applications, from emulating biological components to realizing specialized computing requirements. Furthermore, we explore the potential for IMDs to be modified and tuned to achieve customized dynamics, which positions them as one of the most promising hardware candidates for executing bioinspired algorithms with unique specifications. Finally, we identify the challenges currently facing IMDs that hinder their widespread usage and highlight emerging research directions that could significantly benefit from incorporating IMDs.
Understanding and improving the contact resistance of two-dimensional materials for the fabrication of next-generation devices is of vital importance to be able to fully utilize the new physics available in these materials. In this work, eight different contact metals (Ag, Au, Cr, Cu, In, Mo, Ni, and Ti) have been investigated using the same sample of monolayer MoS2. Through the fabrication and testing of multiple, identically sized field-effect transistor devices per contact metal, we compensate for large variability in electrical properties of as-grown chemical vapor deposition MoS2 and deduce the relative performance of each metal. The general trend of lower work function metals having lower contact resistance holds with In, Ag, and Ti performing the best of the metals tested. Our results are compatible with recent research suggesting that the contact resistance in undoped, monolayer MoS2 is dominated by a lateral junction resistance, and we provide context for how this manifests in device-to-device variation. Multiple orders of magnitude differences in contact resistance are observed between metals and can be explained by this lateral barrier operating in the thermionic-field emission regime.
Autonomous electronic microsystems smaller than the diameter of a human hair (<100 μm) are promising for sensing in confined spaces such as microfluidic channels or the human body. However, they are difficult to implement due to fabrication challenges and limited power budget. Here we present a 60 × 60 μm electronic microsystem platform, or SynCell, that overcomes these issues by leveraging the integration capabilities of two-dimensional material circuits and the low power consumption of passive germanium timers, memory-like chemical sensors, and magnetic pads. In a proof-of-concept experiment, we magnetically positioned SynCells in a microfluidic channel to detect putrescine. After we extracted them from the channel, we successfully read out the timer and sensor signal, the latter of which can be amplified by an onboard transistor circuit. The concepts developed here will be applicable to microsystems targeting a variety of applications from microfluidic sensing to biomedical research.
Self-assembly of vertically aligned III-V semiconductor nanowires (NWs) on two-dimensional (2D) van der Waals (vdW) nanomaterials allows for integration of novel mixed-dimensional nanosystems with unique properties for optoelectronic and nanoelectronic device applications. Here, selective-area vdW epitaxy (SA-vdWE) of InAs NWs on isolated 2D molybdenum disulfide (MoS2) domains is reported for the first time. The MOCVD growth parameter space (i.e., V/III ratio, growth temperature, and total molar flow rates of metalorganic and hydride precursors) is explored to achieve pattern-free positioning of single NWs on isolated multi-layer MoS2 micro-plates with one-to-one NW-to-MoS2 domain placement. The introduction of a pre-growth poly-l-lysine surface treatment is highlighted as a necessary step for mitigation of InAs nucleation along the edges of triangular MoS2 domains and for NW growth along the interior region of 2D micro-plates. Analysis of NW crystal structures formed under the optimal SA-vdWE condition revealed a disordered combination of wurtzite and zinc-blend phases. A transformation of the NW sidewall faceting structure is observed, resulting from simultaneous radial overgrowth during axial NW synthesis. A common lattice arrangement between axially-grown InAs NW core segments and MoS2 domains is described as the epitaxial basis for vertical NW growth. A model is proposed for a common InAs/MoS2 sub-lattice structure, consisting of three multiples of the cubic InAs unit cell along the [21̄1̄] direction, commensurately aligned with a 14-fold multiple of the Mo-Mo (or S-S) spacing along the [101̄0] direction of MoS2 hexagonal lattice. The SA-vdWE growth mode described here enables controlled hybrid integration of mixed-dimensional III-V-on-2D heterostructures as novel nanosystems for applications in optoelectronics, nanoelectronics, and quantum enabling technologies.
Materials with multiple structural phases can be highly desirable because the transition can drive properties between phases. Low enthalpy transitions can particularly enable reconfigurable behaviors due to their low energy cost. In this work, a reversible thermally induced phase transformation is found near room-temperature between centrosymmetric and a previously unobserved non-centrosymmetric hafnium disulfide polytype. The symmetry-breaking mechanism requires no diffusion, intercalation, or charge transfer. A combination of experiments and simulations is used to determine that anharmonic phonons, weaker interlayer interactions, and variable metal atom coordination play important mediating roles in creating a pathway accessible to a lower free energy configuration. The transition occurs at 300 K and is marked by stacked AAA planes shifting into an ABC stacking and the shift is reversible in repeated cool-down experiments. The results demonstrate a reversible a phase transition that offers a unique mechanism for reconfigurability in a two dimensional material.
The ALD process of deposition of ultrathin high-κ HfO2 on chlorinate graphene.
Neuromorphic hardware implementation of Boltzmann Machine using a network of stochastic neurons can allow non-deterministic polynomial-time (NP) hard combinatorial optimization problems to be efficiently solved. Efficient implementation of such Boltzmann Machine with simulated annealing desires the statistical parameters of the stochastic neurons to be dynamically tunable, however, there has been limited research on stochastic semiconductor devices with controllable statistical distributions. Here, we demonstrate a reconfigurable tin oxide (SnO x )/molybdenum disulfide (MoS 2 ) heterogeneous memristive device that can realize tunable stochastic dynamics in its output sampling characteristics. The device can sample exponential-class sigmoidal distributions analogous to the Fermi-Dirac distribution of physical systems with quantitatively defined tunable “temperature” effect. A BM composed of these tunable stochastic neuron devices, which can enable simulated annealing with designed “cooling” strategies, is conducted to solve the MAX-SAT, a representative in NP-hard combinatorial optimization problems. Quantitative insights into the effect of different “cooling” strategies on improving the BM optimization process efficiency are also provided.
Dynamic reconfigurability of material properties is essential to enabling innovative neuromorphicand quantum-computing paradigms. The unique structure of van der Waals layers can facilitate a robust mechanism for this desired reconfigurability. Here, we present a highly versatile and effective approach, based on electrochemical intercalation of organometallics, to control the electron and phonon behavior in hafnium disulfide. Computational and experimental exploration of the physical properties in the intercalated material indicates a significant and measured change. Furthermore, the weak chemical interactions between the organometallics and hafnium disulfide enable an electric-field mediated intercalant drift and charge-discharge process. The control of organometallic concentration in this way provides a dynamic 400-fold control of cross-plane electrical conductivity (1.8 mu S/cm - 741 mu S/cm) and a corresponding 4-fold control of cross-plane thermal conductivity in hafnium disulfide (0.35 Wm(-1) K-1-1.45 Wm(-1) K-1) Our findings unveil a broad approach to dynamically design layered-material properties for high-performance electronic and phononic applications.
Atomic vacancies related to structural disorder and doping variation influence carrier transport in monolayer transition-metal dichalcogenide devices. Here, we investigate the effect of hydrogen plasma exposure (HPE) on monolayer MoS2 field-effect transistors (FETs). We observe that a 1% increase in sulfur vacancy after HPE results in incremental 0.06 eV of the Schottky barrier. Short-range scattering from the sulfur vacancies reduces the carrier mobility of monolayer MoS2 by 2 orders of magnitude. Despite the defects and grain boundaries formed during the chemical vapor deposition and transferring process, the surface desulfurization induced by the proton exposure and thermally accelerated oxidation can be blocked by monolayer graphene cladding with a van der Waals contact distance of 2.5 Å. The material-level study indicates a promising route for a low-cost and robust fabrication of smart sensor circuits on a monolithic MoS2 wafer, where the bare MoS2 FETs can serve as proton sensors, with their electronic readout processed by a logic circuit of graphene-protected pristine FETs with a high on/off ratio.
Through a systematic study of HfS2 based on density functional theory calculations using the quasiharmonic approximation, we show that out-of-plane ZA phonons alone are responsible for 80% of the thermal transport—both in-plane and out-of-plane. The calculations determine the temperature-dependent structural, phonon, and thermal properties in HfS2. The cause is the uncommonly strong interlayer van der Waals interaction relative to its in-plane interaction. The highly anisotropic interactions also give rise to a combination of effects including anisotropic phonon group velocities, a large ZA-optical phonon gap, and a vanishing Grüneisen parameter in ZA modes that leads to large phonon lifetimes. In turn, these enable other unusual properties. HfS2 can have either a negative (T<40 K) or positive (T>40 K) temperature-dependent thermal expansion coefficient and an exceptionally large LO-TO splitting.
The state-of-the-art heterostructure-based devices often involve stacks of epilayers of few nanometer thick crystals. However, the ultimate limit would be a hitherto single-atomic-layer structure. Using material-by-design approach, the flexibility of two-dimensional (2D) materials could be explored to develop a multilayered artificial van der Waals (vdW) heterostructure where the synergistic coupling between the individual materials and the underlying interface creates new and novel functionalities. Herein, based on first-principles calculations, we report that vertically stacked few-layer vdW heterostructure of monolayer molybdenum disulfide (MoS2) and hexagonal boron phosphide (BP) is a strong electrically narrow direct bandgap material. More intriguingly, few-layer vdW heterostructure of monolayer MoS2 and BP exhibits superior mechanical properties. In striking contrast to heterostructures of MoS2/graphene, MoS2/WS2, and few-layer graphene, where the 2D moduli are lower than the sum of the 2D modulus of each nanosheet, the mechanical strength of few-layer hybrid vdW heterostructure of monolayer MoS2 and BP rather increased with increasing thickness. We attribute this difference to the unique interlayer and interface coupling, which affected the vibrational properties of the heterostructures including the emergence of shear and breathing phonon modes as well as the transformation of flexural phonon modes. Such superior mechanical properties could be explored for nanoelectromechanical device applications, e.g., as a nanoresonator. We demonstrate that the electronic structure could as well be tuned with both increasing numbers of monolayer stacks and defect-engineering promising for low-power applications.
Integration schemes are implemented with a plane-wave basis in the context of real-time time-dependent density functional theory. Crank-Nicolson methods and three classes of explicit integration schemes are explored and assessed in terms of their accuracy and stability properties. Within the framework of plane-wave density functional theory, a graphene monolayer system is used to investigate the error, stability, and serial computational cost of these methods. The results indicate that Adams-Bashforth and Adams-Bashforth-Moulton methods of orders 4 and 5 outperform commonly used methods, including Crank-Nicolson and Runge-Kutta methods, in simulations where a relatively low error is desired. Parallel runtime scaling of the most competitive serial methods is presented, further demonstrating that the Adams-Bashforth and Adams-Bashforth-Moulton methods are efficient methods for propagating the time-dependent Kohn-Sham equations. Our integration schemes are implemented as an extension to the Quantum ESPRESSO code.
We have developed a theory for the photoluminescence and absorption coefficient in nanohybrids made of an ensemble of metallic nanoparticles and the core–shell quantum emitter. The core–shell quantum emitter is made of a quantum emitter core and a dielectric shell. When a probe laser light falls on metallic nanoparticles, electric dipoles are induced in the ensemble. Hence, these dipoles interact with each other via the dipole–dipole interaction. The surface plasmon polaritons are also present in metallic nanoparticles. Excitons in the quantum emitter interact with these surface plasmon polaritons and the dipole–dipole interaction electric fields. Using the quantum mechanical density matrix method, we have developed a theory for the photoluminescence quenching and enhancement, the nonradiative decay rate, and the absorption coefficient for the quantum emitter in the ensemble of metallic nanoparticles. We showed that the nonradiative energy loss is mainly due to the exciton coupling with the dipole–dipole interaction and it is responsible for the power loss in the quantum emitter. This in turn produces anomalous photoluminescence enhancement and quenching. We have compared our theory with experimental data of core–shell CdSe/ZnS quantum dots embedded in an ensemble of Au nanoparticles. A good agreement between theory and experiment is found. We showed that there are an energy shift and an enhancement in the absorption peak due to the dipole–dipole interaction. Finally, we showed that there is the anomalous quenching and enhancement in the photoluminescence spectrum of the CdSe/ZnS quantum dot embedded in the ensemble of Au nanoparticles. This phenomenon also occurs mainly due to the dipole–dipole interaction in the ensemble of Au nanoparticles. These are interesting results and can be used to fabricate nanosensors for applications in nanomedicine and nanotechnology.
The mechanical and electronic properties of two-dimensional materials make them promising for use in flexible electronics1-3. Their atomic thickness and large-scale synthesis capability could enable the development of 'smart skin'1,3-5, which could transform ordinary objects into an intelligent distributed sensor network6. However, although many important components of such a distributed electronic system have already been demonstrated (for example, transistors, sensors and memory devices based on two-dimensional materials1,2,4,7), an efficient, flexible and always-on energy-harvesting solution, which is indispensable for self-powered systems, is still missing. Electromagnetic radiation from Wi-Fi systems operating at 2.4 and 5.9 gigahertz8 is becoming increasingly ubiquitous and would be ideal to harvest for powering future distributed electronics. However, the high frequencies used for Wi-Fi communications have remained elusive to radiofrequency harvesters (that is, rectennas) made of flexible semiconductors owing to their limited transport properties9-12. Here we demonstrate an atomically thin and flexible rectenna based on a MoS2 semiconducting-metallic-phase heterojunction with a cutoff frequency of 10 gigahertz, which represents an improvement in speed of roughly one order of magnitude compared with current state-of-the-art flexible rectifiers9-12. This flexible MoS2-based rectifier operates up to the X-band8 (8 to 12 gigahertz) and covers most of the unlicensed industrial, scientific and medical radio band, including the Wi-Fi channels. By integrating the ultrafast MoS2 rectifier with a flexible Wi-Fi-band antenna, we fabricate a fully flexible and integrated rectenna that achieves wireless energy harvesting of electromagnetic radiation in the Wi-Fi band with zero external bias (battery-free). Moreover, our MoS2 rectifier acts as a flexible mixer, realizing frequency conversion beyond 10 gigahertz. This work provides a universal energy-harvesting building block that can be integrated with various flexible electronic systems.
Spin coating and drop casting are viable methods for rapid and low-cost additive manufacturing of components for flexible devices and sensors. We investigated the cumulative effects of layering a conductive polymer composite 2 wt% MWCNT filler in PEDOT:PSS on a Mylar substrate for application to electromagnetic interference (EMI) shielding. The optical transmittace of spin coated composite layers is 90%, 45%, and 20% with a thickness of 0.05 µm, 0.15 µm, and 0.45 µm respectively. Drop cast composite layers have 0% transmittance due to their much greater starting thickness of 4.4 µm. The addition of isopropyl alcohol (IPA) to the solution mixture and substrate heating to 40 °C improves the conductivity, and drying time of the cured composite layers to 10 min. This study shows that the cumulative effects of composite layering are additive, but the electrical properties do not scale the same way. A significant increase in the EMI SE is mainly attributed to the enhanced electrical conductivity of the composite. The insertion of a 50 µm gap in between two 15 µm composite layers accentuates the EMI shielding effectiveness (SE) significantly to a peak of 21 dB within a narrow frequency range in the Ku-band tested.
As two-dimensional (2D) electronic devices continue to advance, the need for integrating high-quality, high-κ nanoscale dielectrics becomes more essential. Plasmaenhanced atomic layer deposition (PEALD) is a promising approach for depositing ultrathin dielectrics directly onto the surface of 2D materials. However, the mechanism for PEALD film growth on the van der Waals materials, along with the impact of the plasma process on structural and interfacial properties of 2D materials, has not been fully explored. In this work, we demonstrate the effects of the plasma process on monolayer, bilayer, and trilayer MoS2. Back-gated MoS2 transistors of varying thickness were tested before and after ALD/PEALD HfO2, and it was verified that plasma damage does occur, predominantly in the surface layer of the MoS2, leading to significantly greater impact in monolayers. By increasing the thickness of the MoS2, the adverse effects of the plasma process are reduced appreciably. This observation is further supported by Raman and transmission electron microscopy analysis. In addition to providing information about defect generation and morphology, this study provides key insights into the charge transfer between HfO2 and MoS2. Overall, this detailed analysis of the impact of the PEALD plasma process on MoS2 contributes to the reliable integration of ultrathin, high-κ dielectrics in 2D devices.