Saptarshi Das

In conventional field-effect transistors (FETs), the gate dielectric primarily provides electrostatic control of the channel and is engineered to remain chemically and electronically inert to ensure stable operation. Emerging studies, however, suggest that functional dielectrics integrated with two-dimensional (2D) semiconductors can introduce additional physical mechanisms that influence device behavior. Here, we investigate AgInP2S6 (AIPS), a high-κ van der Waals thiophosphate, as a functional top-gate dielectric in MoS2 field-effect transistors. The AIPS/MoS2 heterointerface exhibits coupled opto-ionic-electronic interactions, where defect-mediated sub-bandgap photoactivity and photon-assisted Ag+ ion migration independently occur within the dielectric yet collectively modulate the channel conductance. Optical excitation redistributes charge within the dielectric, gradually reconfiguring the ionic landscape and altering the interfacial electrostatic potential. The resulting conductance exhibits strong temperature dependence. At low temperatures (∼0 °C), reduced phonon activity suppresses Ag+ mobility, leading to persistent photoconductivity. At elevated temperatures (∼80 °C), thermally activated ion hopping accelerates relaxation of the stored state. This temperature-dependent retention behavior enables a thermally tunable fading memory response, where the persistence of the optically induced conductance state can be controlled through ionic dynamics. These results highlight how opto-ionic coupling at 2D/dielectric interfaces can introduce stimulus-responsive functionality in transistor architectures, providing a platform for devices that combine sensing and memory behavior within the gate stack.

Optical sensing in semiconductors is traditionally governed by photoconductive, photovoltaic, and photogating effects, in which optical inputs are encoded as continuous variations in average current. Here, we demonstrate a fundamentally distinct paradigm in which light is encoded in stochastic transport dynamics rather than steady-state conductance. Using a monolayer MoS2 field-effect transistor with an aggressively scaled gate stack (∼1.3 nm equivalent oxide thickness), we access a regime where interfacial defect states dominate transport and form gate-tunable potential wells. At low temperature, carrier emission from these states is suppressed, stabilizing trapped configurations in the dark, whereas optical excitation selectively activates emission, giving rise to large-amplitude (“giant”) random telegraph noise. This enables independent control of capture and emission processes through electrostatic gating and illumination, respectively, creating a direct mapping between optical input and stochastic observables such as switching rates, dwell times, and state occupancies. Interestingly, optical detection is achieved through the statistics of rare emission events, enabling sensitivity to ultralow illumination levels (∼50 fW). Spectral dependence further reveals photon energy–dependent emission kinetics, confirming the role of photon-assisted detrapping. These results show a light-induced nonequilibrium transport regime in which defect-mediated fluctuations, conventionally treated as noise, act as a functional signal and provide a stochastic sensing framework for ultralow-light detection in nanoscale electronic systems.

Reservoir computing is attractive for edge intelligence because it not only reduces training overhead but also turns transient material physics into computation. In in‐materio systems, memory, nonlinearity, separability, and efficiency are not added later at the circuit level; they emerge from defect kinetics, ionic motion, interfacial barriers, percolation pathways, and related dynamics. This perspective therefore treats material descriptors as design variables rather than post hoc explanations of device behavior. Accordingly, six design rules connect measurable material properties to useful reservoir function: timescale matching, operating‐regime nonlinearity, controlled stochasticity, observable state dimensionality, reversible and energy‐relevant operation, and robustness across an operating range. To make these rules actionable, the framework introduces a staged experimental workflow, a minimum quantitative characterization set, attribution controls separating material dynamics from input encoding and readout design, and a cross‐layer reporting checklist. The focus remains material‐ and device‐based in‐materio physical reservoirs for edge and in‐sensor processing, while the descriptor‐to‐dynamics logic is transferable to other physical reservoirs. The framework is then extended to reservoir‐by‐sensing platforms, where sensing and temporal processing coexist within one substrate. Finally, reporting of operating conditions, state observables, stability, and energy boundaries supports fair comparison and helps move proof‐of‐concept demonstrations toward deployable materials systems.

The combined effects of substrate miscut angle and growth temperature on the epitaxial growth and properties of MoS2 monolayers synthesized by metalorganic chemical vapor deposition (MOCVD) on c-plane sapphire substrates were investigated. Increasing the miscut angle toward the M-axis produces faceted step–terrace surfaces with higher step density and reduced terrace width on the sapphire surface, while promoting pronounced step bunching during growth. These miscut-induced surface reconstructions strongly modulate MoS2 growth, leading to systematic changes in bilayer domain density, size, morphology, and epitaxial alignment without significantly altering the growth rate. Comprehensive structural, optical, and electronic characterizations reveal that growth at 1000 °C on the lowest miscut (0.2°) sapphire yields the highest quality monolayer MoS2, as evidenced by the low full width at half-maximum of the in-plane X-ray diffraction φ-scan peaks, minimized defect-bound exciton emission in photoluminescence spectra at cryogenic temperatures, and superior field-effect transistor performance. In contrast, 1° miscut substrates promote increased bilayer domain density, likely due to increased step bunching of the sapphire, making them suitable for controlled bilayer MoS2 growth. However, further increasing the miscut to 4° or lowering the growth temperature to 950 °C results in significantly degraded structural, optical, and electrical properties. These results highlight the importance of substrate miscut and its role in governing the epitaxial growth of MoS2, offering valuable guidance for the scalable synthesis of high-quality two-dimensional MoS2 films.

Simulated annealing (SA), a metaheuristic algorithm inspired by physical annealing processes, attempts to solve combinatorial optimization problems by performing a stochastic search that iteratively explores the solution space. Probabilistic solution evaluation allows the algorithm to escape local minima and progressively converge toward a global optimum, even within complex solution landscapes. In this work, we present the hardware acceleration of SA for spatial optimization within constrained environments, specifically targeting drone placement that maximizes area coverage while avoiding no-fly zones. Stochasticity is introduced through a true random number generator (TRNG) based on two-dimensional (2D) materials. The system energy is evaluated using a 2D logic circuit and the acceptance of candidate solutions is governed by a programmable 2D circuit with tunable threshold behavior. For a representative drone placement task, our hardware-based annealing accelerator circuit modules consume 43 nJ of energy per iteration and achieve an 1800-fold search acceleration relative to brute-force exploration of the solution space. This work also underscores the potential of 2D-material-enabled stochastic and programmable hardware for real-time constrained optimization.