Reconfigurable Compute-In-Memory on Field-Programmable Ferroelectric Diodes.

The deluge of sensors and data generating devices has driven a paradigm shift in modern computing from arithmetic-logic centric to data-centric processing. Data-centric processing require innovations at the device level to enable novel compute-in-memory (CIM) operations. A key challenge in the construction of CIM architectures is the conflicting trade-off between the performance and their flexibility for various essential data operations. Here, we present a transistor-free CIM architecture that permits storage, search, and neural network operations on sub-50 nm thick Aluminum Scandium Nitride ferroelectric diodes (FeDs). Our circuit designs and devices can be directly integrated on top of Silicon microprocessors in a scalable process. By leveraging the field-programmability, nonvolatility, and nonlinearity of FeDs, search operations are demonstrated with a cell footprint <0.12 µm2 when projected onto 45 nm node technology. We further demonstrate neural network operations with 4-bit operation using FeDs. Our results highlight FeDs as candidates for efficient and multifunctional CIM platforms.

TLFND: A Multimodal Fusion Model Based on Three-Level Feature Matching Distance for Fake News Detection.

In the rapidly evolving information era, the dissemination of information has become swifter and more extensive. Fake news, in particular, spreads more rapidly and is produced at a lower cost compared to genuine news. While researchers have developed various methods for the automated detection of fake news, challenges such as the presence of multimodal information in news articles or insufficient multimodal data have hindered their detection efficacy. To address these challenges, we introduce a novel multimodal fusion model (TLFND) based on a three-level feature matching distance approach for fake news detection. TLFND comprises four core components: a two-level text feature extraction module, an image extraction and fusion module, a three-level feature matching score module, and a multimodal integrated recognition module. This model seamlessly combines two levels of text information (headline and body) and image data (multi-image fusion) within news articles. Notably, we introduce the Chebyshev distance metric for the first time to calculate matching scores among these three modalities. Additionally, we design an adaptive evolutionary algorithm for computing the loss functions of the four model components. Our comprehensive experiments on three real-world publicly available datasets validate the effectiveness of our proposed model, with remarkable improvements demonstrated across all four evaluation metrics for the PolitiFact, GossipCop, and Twitter datasets, resulting in an F1 score increase of 6.6%, 2.9%, and 2.3%, respectively.

Post-CMOS Compatible Aluminum Scandium Nitride/2D Channel Ferroelectric Field-Effect-Transistor Memory.

Recent advances in oxide ferroelectric (FE) materials have rejuvenated the field of low-power, nonvolatile memories and made FE memories a commercial reality. Despite these advances, progress on commercial FE-RAM based on lead zirconium titanate has stalled due to process challenges. The recent discovery of ferroelectricity in scandium-doped aluminum nitride (AlScN) presents new opportunities for direct memory integration with logic transistors due to the low temperature of AlScN deposition (approximately 350 °C), making it compatible with back end of the line integration on silicon logic. Here, we present a FE-FET device composed of an FE-AlScN dielectric layer integrated with a two-dimensional MoS2 channel. Our devices show an ON/OFF ratio of ∼106, concurrent with a normalized memory window of 0.3 V/nm. The devices also demonstrate stable memory states up to 104 cycles and state retention up to 105 s. Our results suggest that the FE-AlScN/2D combination is ideal for embedded memory and memory-based computing architectures.

Discovery of Six Ramoplanin Family Gene Clusters and the Lipoglycodepsipeptide Chersinamycin*.

Ramoplanins and enduracidins are peptidoglycan lipid intermediate II-binding lipodepsipeptides with broad-spectrum activity against methicillin- and vancomycin-resistant Gram-positive pathogens. Targeted genome mining using probes from conserved sequences within the ramoplanin/enduracidin biosynthetic gene clusters (BGCs) was used to identify six microorganisms with BGCs predicted to produce unique lipodepsipeptide congeners of ramoplanin and enduracidin. Fermentation of Micromonospora chersina yielded a novel lipoglycodepsipeptide, called chersinamycin, which exhibited good antibiotic activity against Gram-positive bacteria (1-2 µg/mL) similar to the ramoplanins and enduracidins. The covalent structure of chersinamycin was determined by NMR spectroscopy and tandem mass spectrometry in conjunction with chemical degradation studies. These six new BGCs and isolation of a new antimicrobial peptide provide much-needed tools to investigate the fundamental aspects of lipodepsipeptide biosynthesis and to facilitate efforts to produce novel antibiotics capable of combating antibiotic-resistant infections.

Multistate, Ultrathin, Back-End-of-Line-Compatible AlScN Ferroelectric Diodes.

The growth in data generation necessitates efficient data processing technologies to address the von Neumann bottleneck in conventional computer architecture. Memory-driven computing, which integrates nonvolatile memory (NVM) devices in a 3D stack, is gaining attention, with CMOS back-end-of-line (BEOL)-compatible ferroelectric (FE) diodes being ideal due to their two-terminal design and inherently selector-free nature, facilitating high-density crossbar arrays. Here, we demonstrate BEOL-compatible, high-performance FE diodes scaled to 5, 10, and 20 nm FE Al0.72Sc0.28N/Al0.64Sc0.36N films. Through interlayer (IL) engineering, we show substantial improvements in the on/off ratios (>166 times) and rectification ratios (>176 times) in these scaled devices. These characteristics also enable 5-bit multistate operation with a stable retention. We also experimentally and theoretically demonstrate the counterintuitive result that the inclusion of an IL can lead to a decrease in the ferroelectric switching voltage of the device. An in-depth analysis into the device transport mechanisms is performed, and our compact model aligns seamlessly with the experimental results. Our results suggest the possibility of using scaled AlxSc1-xN FE diodes for high-performance, low-power, embedded NVM.

Tuning Polarity in WSe2/AlScN FeFETs via Contact Engineering.

Recent advancements in ferroelectric field-effect transistors (FeFETs) using two-dimensional (2D) semiconductor channels and ferroelectric Al0.68Sc0.32N (AlScN) allow high-performance nonvolatile devices with exceptional ON-state currents, large ON/OFF current ratios, and large memory windows (MW). However, previous studies have solely focused on n-type FeFETs, leaving a crucial gap in the development of p-type and ambipolar FeFETs, which are essential for expanding their applicability to a wide range of circuit-level applications. Here, we present a comprehensive demonstration of n-type, p-type, and ambipolar FeFETs on an array scale using AlScN and multilayer/monolayer WSe2. The dominant injected carrier type is modulated through contact engineering at the metal-semiconductor junction, resulting in the realization of all three types of FeFETs. The effect of contact engineering on the carrier injection is further investigated through technology-computer-aided design simulations. Moreover, our 2D WSe2/AlScN FeFETs achieve high electron and hole current densities of ∼20 and ∼10 µA/µm, respectively, with a high ON/OFF ratio surpassing ∼107 and a large MW of >6 V (0.14 V/nm).

Nonvolatile Control of Valley Polarized Emission in 2D WSe2-AlScN Heterostructures.

Achieving robust and electrically controlled valley polarization in monolayer transition metal dichalcogenides (ML-TMDs) is a frontier challenge for realistic valleytronic applications. Theoretical investigations show that the integration of 2D materials with ferroelectrics is a promising strategy; however, an experimental demonstration has remained elusive. Here, we fabricate ferroelectric field-effect transistors using a ML-WSe2 channel and an Al0.68Sc0.32N (AlScN) ferroelectric dielectric and experimentally demonstrate efficient tuning as well as non-volatile control of valley polarization. We measure a large array of transistors and obtain a maximum valley polarization of ∼27% at 80 K with stable retention up to 5400 s. The enhancement in the valley polarization is ascribed to the efficient exciton-to-trion (X-T) conversion and its coupling with an out-of-plane electric field, viz., the quantum-confined Stark effect. This changes the valley depolarization pathway from strong exchange interactions to slow spin-flip intervalley scattering. Our research demonstrates a promising approach for achieving non-volatile control over valley polarization for practical valleytronic device applications.

Scalable CMOS back-end-of-line-compatible AlScN/two-dimensional channel ferroelectric field-effect transistors.

Three-dimensional monolithic integration of memory devices with logic transistors is a frontier challenge in computer hardware. This integration is essential for augmenting computational power concurrent with enhanced energy efficiency in big data applications such as artificial intelligence. Despite decades of efforts, there remains an urgent need for reliable, compact, fast, energy-efficient and scalable memory devices. Ferroelectric field-effect transistors (FE-FETs) are a promising candidate, but requisite scalability and performance in a back-end-of-line process have proven challenging. Here we present back-end-of-line-compatible FE-FETs using two-dimensional MoS2 channels and AlScN ferroelectric materials, all grown via wafer-scalable processes. A large array of FE-FETs with memory windows larger than 7.8 V, ON/OFF ratios greater than 107 and ON-current density greater than 250 µA um-1, all at ~80 nm channel length are demonstrated. The FE-FETs show stable retention up to 10 years by extension, and endurance greater than 104 cycles in addition to 4-bit pulse-programmable memory features, thereby opening a path towards the three-dimensional heterointegration of a two-dimensional semiconductor memory with silicon complementary metal-oxide-semiconductor logic.

Vertical and Lateral Etch Survey of Ferroelectric AlN/Al1-xScxN in Aqueous KOH Solutions.

Due to their favorable electromechanical properties, such as high sound velocity, low dielectric permittivity and high electromechanical coupling, Aluminum Nitride (AlN) and Aluminum Scandium Nitride (Al1-xScxN) thin films have achieved widespread application in radio frequency (RF) acoustic devices. The resistance to etching at high scandium alloying, however, has inhibited the realization of devices able to exploit the highest electromechanical coupling coefficients. In this work, we investigated the vertical and lateral etch rates of sputtered AlN and Al1-xScxN with Sc concentration x ranging from 0 to 0.42 in aqueous potassium hydroxide (KOH). Etch rates and the sidewall angles were reported at different temperatures and KOH concentrations. We found that the trends of the etch rate were unanimous: while the vertical etch rate decreases with increasing Sc alloying, the lateral etch rate exhibits a V-shaped transition with a minimum etch rate at x = 0.125. By performing an etch on an 800 nm thick Al0.875Sc0.125N film with 10 wt% KOH at 65 °C for 20 min, a vertical sidewall was formed by exploiting the ratio of the 1011¯ planes and 11¯00 planes etch rates. This method does not require preliminary processing and is potentially beneficial for the fabrication of lamb wave resonators (LWRs) or other microelectromechanical systems (MEMS) structures, laser mirrors and Ultraviolet Light-Emitting Diodes (UV-LEDs). It was demonstrated that the sidewall angle tracks the trajectory that follows the 1¯212¯ of the hexagonal crystal structure when different c/a ratios were considered for elevated Sc alloying levels, which may be used as a convenient tool for structure/composition analysis.