中央研究院 原子與分子科學研究所

We here report on the direct observation of ferroelectric properties of water ice in its 2D phase. Upon nanoelectromechanical confinement between two graphene layers, water forms a 2D ice phase at room temperature that exhibits a strong and permanent dipole which depends on the previously applied field, representing clear evidence for ferroelectric ordering. Characterization of this permanent polarization with respect to varying water partial pressure and temperature reveals the importance of forming a monolayer of 2D ice for ferroelectric ordering which agrees with ab-initio and molecular dynamics simulations conducted. The observed robust ferroelectric properties of 2D ice enable novel nanoelectromechanical devices that exhibit memristive properties. A unique bipolar mechanical switching behavior is observed where previous charging history controls the transition voltage between low-resistance and high-resistance state. This advance enables the realization of rugged, non-volatile, mechanical memory exhibiting switching ratios of 106, 4 bit storage capabilities and no degradation after 10,000 switching cycles.

We here demonstrate the multifunctional properties of atomically thin heterojunctions that are enabled by strong interfacial interactions and their integration into ultra-high performance, self-powered sensors. Epitaxial alignment between tin diselenide and graphene through direct growth produces thermoelectric and mechanoelectric properties beyond the ability of either component. An unprecedented ZT of 2.43 originated from the synergistic combination of graphene’s high carrier conductivity and SnSe2 mediated thermal conductivity lowering. Moreover, strong interaction at the SnSe₂/graphene interface produces stress localization that results in a novel 2D-crack-assisted strain sensing mechanism whose sensitivity (GF=450) is superior to all other 2D materials. Finally, the graphene-assisted growth process, permits the formation of high-quality heterojunctions directly on polymeric substrates for flexible and transparent self-powered sensors that achieve fast and reliable strain sensing from a small temperature gradient. Our work enhances the fundamental understanding of multifunctionality at the atomic scale and provide a route towards structural health monitoring through ubiquitous and smart devices.

Atomic‐precision patterning at large scale is a central requirement for nanotechnology and future electronics that is hindered by the limitations of lithographical techniques. Historically, imperfections of the fabrication tools have been compensated by multi‐patterning using sequential lithography processes. The realization of nanometer‐scale features from much larger patterns through offset stacking of atomically thin masks is demonstrated. A unique mutual stabilization effect between two graphene layers produces atomically abrupt transitions that selectively expose single‐layer covered regions. Bilayer regions, on the other hand, protect the underlying substrate from removal for several hours permitting transfer of atomic thickness variations into lateral features in various semiconductors. Nanoscopic offsets between two 2D materials layers could be introduced through bottom‐up and top‐down approaches, opening up new routes for high‐resolution patterning. A self‐aligned templating approach yields nanometer‐wide bilayer graphene nanoribbons with macroscopic length that produces high‐aspect‐ratio silicon nanowalls. Moreover, offset‐transfer of lithographically patterned graphene layers enables multipatterning of large arrays of semiconductor features whose resolution is not limited by the employed lithography and could reach <10 nm feature size. The results open up a new route to combining design flexibility with unprecedented resolution at large scale.

Carrier transport processes in assemblies of nanostructures rely on morphology-dependent and hierarchical conduction mechanisms, whose complexity cannot be captured by current modelling approaches. Here we apply the concept of complex networks to modelling carrier conduction in such systems. The approach permits assignment of arbitrary connectivity and connection strength between assembly constituents and is thus ideal for nanostructured films, composites and other geometries. Modelling of simplified rod-like nanostructures is consistent with analytical solutions, whereas results for more realistic nanostructure assemblies agree with experimental data and reveal conduction behaviour not captured by previous models. Fitting of ensemble measurements also allows the conduction properties of individual constituents to be extracted, which are subsequently used to guide the realization of transparent electrodes with improved performance. A global optimization process was employed to identify geometries and properties with high potential for transparent conductors. Our intuitive discretization approach, combined with a simple solver tool, allows researchers with little computational experience to carry out realistic simulations.

The carrier conduction in 2D materials is more sensitive to surface-bound disorder than bulk materials which is thought to limit their achievable performance in electronic devices. To date, charged impurity scattering is considered the main source of interaction between ionic adsorbates and carriers in 2D materials. We here observe a previously unknown source of carrier scattering in graphene upon interaction with ionic impurities. Different from charged impurity scattering, these “neutral scatterers” do not depend on carrier concentration and result in a sixfold mobility variation at similar doping. Comparison of different ionic residue from various metal etchants reveals a universal mechanism that controls the carrier mean free path. Raman spectroscopy suggests that inhomogeneous charge distribution is the source of neutral scatterers and we extract an optical fingerprint for their presence. The charge heterogeneity thus acts as an additional degree of freedom in graphene's carrier transport and its consideration can explain the transition from ambipolar to unipolar charge transport in graphene. Our results not only provide new insight into the carrier transport of 2D materials in the presence of disorder and provide guidelines for enhancing the performance of graphene devices but also enable novel device concepts in graphene.

Lateral heterojunctions in two-dimensional (2D) materials have demonstrated potential for high-performance sensors because of the unique electrostatic conditions at the interface. The increased complexity of producing such structures, however, has prevented their widespread use. We here demonstrate the simple and scalable fabrication of heterojunctions by a one-step synthesis process that yields photodetectors with superior device performance. Carrier transport in these heterojunctions was found to proceed by minimizing the path through the amorphous carbon barriers, which results in a self-selective Schottky emission process with high uniformity and low emission barriers. We demonstrate the potential of thus produced  heterojunctions by realizing a photodetector that combines an ultrahigh detectivity of 1013 Jones with microsecond response time, which represents the highest performance of 2D material heterojunction devices. These attractive features are retained even for millimeter-scale devices, and the demonstrated ability to produce transparent, patterned, and flexible sensors extends lateral heterojunction sensors toward wearable and large-scale electronics.