Indium selenides for next-generation low-power computing devices

Technical Summary: Indium Selenides for Next-Generation Low-Power Computing Devices

Problem Statement
The miniaturization of silicon-based microelectronics is approaching fundamental physical limits regarding energy efficiency, speed, and density. As silicon transistors face increasing energy consumption and limited scalability at advanced nodes, there is an urgent need for complementary materials that can overcome short-channel effects and enable ultra-low-power computing. While two-dimensional (2D) van der Waals materials, such as transition metal dichalcogenides (TMDs), have emerged as candidates, they often lack the combination of high carrier mobility and unique functional properties required to surpass silicon in both logic and memory applications.

Methodology and Scope
This perspective critically evaluates the material properties, fabrication challenges, and device applications of van der Waals indium selenides, specifically Indium Selenide (InSe) and Indium Selenide (In₂Se₃). The authors synthesize existing literature to analyze:

• Material Physics: The diverse polymorphic phases (e.g., β, γ, ε-InSe and α, β, β'-In₂Se₃), their electronic band structures, and thickness-dependent quantum confinement effects.
• Synthesis Techniques: A comparative analysis of top-down methods (mechanical and chemical exfoliation) versus bottom-up growth techniques (Metal-Organic Chemical Vapor Deposition [MOCVD], Molecular Beam Epitaxy [MBE], and Chemical Vapor Transport).
• Device Performance: The evaluation of InSe and In₂Se₃ in various transistor architectures, including ballistic MOSFETs, Tunnel Field-Effect Transistors (TFETs), and Ferroelectric Semiconductor Field-Effect Transistors (FeSFETs).
• Stability and Defects: An examination of oxidation mechanisms, native point defects, and strategies for encapsulation and doping to improve device stability and performance.

Key Contributions and Results

• Superior Transport Properties: InSe exhibits exceptional electron mobility exceeding 1,000 cm²V⁻¹s⁻¹ at room temperature, attributed to an unusually low effective electron mass (0.12 m₀) and a high thermal velocity (>2×10⁷ cm s⁻¹). These properties facilitate ballistic transport in nanoscale devices, allowing for high transconductance (6 mS) and current densities (1.2 mA μm⁻¹) in sub-10 nm channel devices, outperforming many 2D TMDs and approaching theoretical limits.
• Phase-Dependent Ferroelectricity: Unlike conventional semiconductors, specific phases of indium selenides (notably α-In₂Se₃ and sliding ferroelectric InSe) exhibit robust ferroelectricity persisting down to the monolayer limit. This enables the integration of non-volatile memory functionality directly within the semiconducting channel.
  • In₂Se₃: The α-phase demonstrates out-of-plane and in-plane ferroelectricity with a high Curie temperature (~700 K), suitable for high-temperature operation.
  • InSe: Multilayer InSe exhibits "sliding ferroelectricity," where polarization reversal occurs via interlayer sliding rather than atomic displacement, offering a defect-tolerant switching mechanism.
• Device Demonstrations:
  • Logic: Ballistic InSe transistors have demonstrated record-breaking performance metrics. InSe-based TFETs utilizing type-III heterojunctions with Silicon show sub-60 mV dec⁻¹ subthreshold swings, indicating potential for ultra-low-power switching.
  • Memory: Ferroelectric semiconductor junctions (FSJs) and FeSFETs based on In₂Se₃ have achieved non-volatile data storage with on/off ratios up to 10⁵ and switching speeds around 40 ns. These devices allow for logic-in-memory architectures, potentially overcoming the von Neumann bottleneck.
• Synthesis and Stability Challenges: The paper identifies that scalable synthesis remains a primary hurdle. While MOCVD offers a pathway to wafer-scale, phase-pure films, current synthetic films often suffer from lower mobility (~1–10 cm²V⁻¹s⁻¹) compared to exfoliated flakes due to defects, grain boundaries, and phase impurities. Furthermore, indium selenides are highly susceptible to oxidation in ambient conditions, forming In₂O₃ or amorphous oxides that degrade performance. Effective encapsulation (e.g., hBN, HfO₂) and oxygen-free processing are critical for stability.

Significance and Outlook
The paper posits that indium selenides represent a unique material platform capable of unifying high-performance logic and non-volatile memory within a single system, a capability absent in conventional silicon technology. By leveraging their high mobility for ballistic transport and their intrinsic ferroelectricity for memory, these materials could enable next-generation computing architectures that are both energy-efficient and dense.

The authors conclude that while significant progress has been made in material characterization and device prototyping, the transition to commercial viability depends on overcoming three critical challenges:

1. Scalable Synthesis: Developing robust, large-area growth methods (such as optimized MOCVD) that produce defect-free, phase-pure films with high mobility.
2. Phase Control: Precise engineering of polymorphs and stacking orders to tailor electronic and ferroelectric properties.
3. Oxidation Prevention: Implementing reliable encapsulation and processing strategies to maintain material integrity during device fabrication and operation.

If these challenges are addressed, indium selenides could serve as a cornerstone for post-silicon electronics, enabling monolithic 3D integration of logic and memory layers and facilitating novel computing paradigms such as neuromorphic and in-sensor computing.