Hardware and Information Security Primitives Based on 2D Materials and Devices

Abstract

Hardware security is a major concern for the entire semiconductor ecosystem that accounts for billions of dollars in annual losses. Similarly, information security is a critical need for the rapidly proliferating edge devices that continuously collect and communicate a massive volume of data. While silicon‐based complementary metal‐oxide‐semiconductor technology offers security solutions, these are largely inadequate, inefficient, and often inconclusive, as well as resource intensive in time, energy, and cost, leading to tremendous room for innovation in this field. Furthermore, silicon‐based security primitives have shown vulnerability to machine learning (ML) attacks. In recent years, 2D materials such as graphene and transition metal dichalcogenides have been intensely explored to mitigate these security challenges. In this review, 2D‐materials‐based hardware security solutions such as camouflaging, true random number generation, watermarking, anticounterfeiting, physically unclonable functions, and logic locking of integrated circuits (ICs) are summarized with accompanying discussion on their reliability and resilience to ML attacks. In addition, the role of native defects in 2D materials in developing high entropy hardware security primitives is also examined. Finally, the existing challenges for 2D materials, which must be overcome for large‐scale deployment of 2D ICs to meet the security needs of the semiconductor industry, are discussed.

A Peripheral-Free True Random Number Generator Based on Integrated Circuits Enabled by Atomically Thin Two-Dimensional Materials

“…Therefore, augmenting hardware security primitives has the potential to accelerate their adoption process and make them more attractive for IoT implementation. 31 In this context, some of the recently demonstrated hardware security applications 20,32−34 with MoS 2 as the channel material are encouraging. Before the operational assessment of our MoS 2 TSI-based TRNG, we first evaluated the quality of the grown MoS 2 film.…”

Dielectrics for Two-Dimensional Transition-Metal Dichalcogenide Applications

“…For example, bioinspired devices have been demonstrated using in a split-gated MoS 2 device with hydrogen silsesquioxane gate dielectric that can mimic the auditory cortex of a barn owl or be used for energy efficient and biorealistic spiking neural networks. , ALD grown 50 nm Al 2 O 3 on Pt/TiN/p +2 Si stack with 2D TMDs were used to demonstrate simulated annealing for Ising type spin systems, biomimetic collision detection based on the escape response in locusts, monolithically integrated, multipixel, bioinspired neural networks, bioinspired machine vision, and inspect-inspired collision detectors . Similarly, inherent stochastic charge trapping and detrapping in gate stacks that lead to cycle-to-cycle variation in device characteristics have been exploited for applications in stochastic computing , and hardware security. , Similar architectures based on ALD grown 50 nm Al 2 O 3 /Pt/TiN stack for MoS 2 memtransistor exploited the programmable stochasticity to demonstrate logic gate-based arithmetic operations, energy efficient hardware implementation of a Bayesian network, and probabilistic synapses and reconfigurable neurons for Bayesian neural networks . Hardware security applications can also utilize this architecture, with demonstrations of logic locking, physical unclonable functions, anticounterfeit measures, optical watermarks, camouflaging, and true random number generators …”

“…118 Similarly, inherent stochastic charge trapping and detrapping in gate stacks that lead to cycle-to-cycle variation in device characteristics have been exploited for applications in stochastic computing 119,120 and hardware security. 121,122 Similar architectures based on ALD grown 50 nm Al 2 O 3 /Pt/TiN stack for MoS 2 memtransistor exploited the programmable stochasticity to demonstrate logic gate-based arithmetic operations, 123 energy efficient hardware implementation of a Bayesian network, 124 and probabilistic synapses and reconfigurable neurons for Bayesian neural networks. 125 Hardware security applications can also utilize this architecture, with demonstrations of logic locking, 126 physical unclonable functions, 127 anticounterfeit measures, 127 optical watermarks, 127 camouflaging, 127 and true random number generators.…”

A Graphene-Based Straintronic Physically Unclonable Function

“…Hardware security is a pressing need for Internet of Things (IoT) edge devices, − and physically unclonable functions (PUFs) are relatively low-cost, low-power, and low-overhead hardware security solutions. PUFs exploit manufacturing process variation in devices and their interactions with external stimuli to generate unique and unpredictable digital signatures, which can be used for hardware authentication. ,− While silicon PUFs are the most prevalent in the semiconductor industry, their low entropy resulting from poor device-to-device variation and high power consumption owing to their need for complex error correcting peripheral circuits have led to the investigation of alternative PUF formats, , including nanowire PUFs, memristive PUFs, , graphene PUFs, CNT PUFs, polymer PUFs, and even optical and biological PUFs. , …”