A platform that integrates artificial intelligence (AI) with advanced computational simulations beyond DFT, designed to accelerate the design and optimization of materials for high-performance applications in electronics, photonics, and energy technologies is presented.
Abstract
The development of silicon-compatible materials for next-generation CMOS and post-CMOS technologies requires more precise and efficient methods for material discovery and optimization. Traditional computational approaches like density functional theory (DFT) often fail to capture the full complexity of materials at the atomic and electronic levels. To overcome these limitations, I will present a platform that integrates artificial intelligence (AI) with advanced computational simulations beyond DFT, designed to accelerate the design and optimization of materials for high-performance applications in electronics, photonics, and energy technologies.
This platform leverages both advanced computational techniques and machine learning to offer more accurate, scalable models for predicting the properties of novel materials. By going beyond traditional DFT, we can better capture complex phenomena such as excited-state properties and material behavior under different environmental conditions, which are critical factors for developing advanced materials like high-k dielectrics, wide-bandgap semiconductors (e.g., SiC, GaN), and ferroelectrics for neuromorphic computing.
The platform enables high-throughput simulations, allowing us to explore a broad material space and predict how materials will perform under various physical conditions. This capability is key for optimizing materials across a range of applications, from low-power electronics to advanced photonic and energy systems.
I will also focus on how this approach is applied to materials crucial for the advancement of future silicon-compatible technologies, such as those used in non-volatile memory, power electronics, and integrated photonics. Moreover, the platform facilitates the discovery of new materials with tailored properties, which are essential for applications like heterogeneous integration (HI), 2.5D/3D chip stacking, and optical interconnects, pushing the boundaries of high-bandwidth, high-performance systems.
By combining AI with beyond-DFT computational methods, this platform provides a powerful, scalable approach to accelerate the discovery, optimization, and integration of materials for next-generation semiconductor and photonic systems. It offers a solution to overcoming material-related challenges, helping to drive the evolution of next-generation electronics and photonics.
This presentation will explore the integration of AI-driven methods and advanced simulation techniques to accelerate the development of high-performance heterogeneous functional materials to enable the development of next-generation devices with high efficiency, scalability, and durability.
The growing computational demands of edge computing and AI applications are pushing semiconductor technologies to their physical and architectural limits, particularly in power efficiency and memory integration. Two-dimensional (2D) layered materials, with their unique (opto)electronic and mechanical properties, allow addressing the challenges of current Complementary Metal-Oxide-Semiconductor (CMOS) technology and enable emerging energy-efficient applications such as neuromorphic computing, integrated photonic devices, and non-volatile ferroelectric memory. Realizing these applications requires significant advances in the synthesis and CMOS-compatible integration of 2D materials, transitioning from lab-scale demonstrations to volume manufacturing.
In this talk, we present recent progress on synthesis and integration of n-type MoS
2
and p-type WSe
2
for CMOS applications in advanced transistor nodes using a Close Coupled Showerhead® (CCS) metal-organic chemical vapor deposition (MOCVD) system integrated into a 300 mm pilot line (Fig. 1A). Achieving uniformity and reproducibility at wafer-scale growth is not only critical for consistent 2D device performance at the circuit level, but also essential for the development of CMOS-compatible process flows. Using AIXTRON’s MOCVD technology, in situ monitoring techniques such as pyrometry and reflectometry are employed to optimize 2D growth and maintain statistical process control on a wafer scale (Fig. 1B). Furthermore, we establish wafer scale ex situ metrology protocols (Fig. 1C) to support the optimization of the growth process. We correlate wafer-scale 2D film properties with field-effect transistor (FET) performance, evaluating key metrics such as carrier mobility and drive current.
We further explore additional application spaces of 2D materials beyond CMOS, focusing on graphene for photonic integrated circuits and hexagonal boron nitride (h-BN) as a passive material that preserves clean van der Waals interfaces with other materials. We demonstrate controlled synthesis of monolayer graphene and multilayer hBN on 200 mm wafers and investigate its resistive-switching behavior in memristors, as well as its superior performance as a gate dielectric in FETs. Overall, the presented advancements in MOCVD technology demonstrate a decisive step toward the industrial integration of 2D materials in future (opto)electronic devices and CMOS fab-compatible manufacturing environments.
Fig. 1.
(A) Photograph of an AIXTRON 300 mm MOCVD system for 2D materials, featuring automated wafer loading and unloading, and capable of processing 200 mm and 300 mm silicon and sapphire wafers of varying thicknesses. (B) Monitoring 2D layer growth and statistical process control, including run-to-run repeatability assessment (C) Raman spectra of WS
2
within the 300 mm wafer demonstrating excellent uniformity .
Figure 1
Emre Yengel, C. Mauder, S. Pasko et al.· ECS Meeting Abstracts· 0 citations
This review traces the development of semiconductor equipment, from bipolar junction transistors (BJTs) to advanced architectures such as fin field‐effect transistors (FinFETs), nanowire transistors, and carbon nanotube FETs. We highlight major developments that have enabled improvements in performance, energy efficiency, and integration density. Special attention is paid to detailed bandgap materials (SiC, GaN), two‐dimensional materials, and van der Waals (VdW) heterostructures, which are ready to shape the next generation of power and high‐frequency electronics. Next‐generation trends beyond CMOS, such as neuromorphic engineering and quantum computing, are identified, with a focus on operating methodology and recent hardware realizations. The application ground is linked to industry‐specific needs in healthcare, energy, and AI hardware. Comparative data tables and combined figures are presented to further the performance benchmark. We consider existing challenges, fabrication intricacies, quantum decoherence, and the integration of materials and emphasize future paths, among them quantum–neuromorphic hybrid systems, green fabrication methods, and artificial intelligence‐based design approaches. Furthermore, 2026 predictions, in‐depth industrial applications, quantitative comparisons, and opportunities for further research efforts are critically reviewed.
M. Subramanian, Mahalakshmi Balan, Ananthi Kaliyamoorthy et al.· Physica Status Solidi (a)· 0 citations
This work pursues a multi-tier development strategy in which machine learning (ML) algorithms are combined with exact physical symmetries and constraints to significantly accelerate computations of electronic structure and atomistic dynamics.
Data movement has become a dominant bottleneck in modern artificial intelligence hardware, making in-memory computing a critical direction for energy-efficient and memory-centric architectures. Ferroelectric hafnium oxide provides a distinctive materials platform for this transition because field-driven polarization switching, non-volatility, CMOS compatibility, and nanoscale thickness scalability can be combined within a process-relevant oxide system. This review establishes a device-to-system perspective on HfO2-based and Hf0.5Zr0.5O2-based ferroelectric memories for in-memory computing. Instead of treating ferroelectric materials, memory devices, circuit primitives, and computing architectures as separate research topics, we examine how their mutual constraints define the achievable efficiency, precision, reliability, and scalability of hafnia-based computing systems. The discussion connects polarization engineering and defect control with charge-domain computation, threshold-state logic, associative search, analog weight representation, neuromorphic plasticity, and sensor-side processing. Particular emphasis is placed on the translation of ferroelectric functionality from individual devices to arrays, macros, and system-level accelerators. We identify variability, fatigue, charge trapping, multilevel-state uncertainty, peripheral overhead, and benchmarking inconsistency as the central barriers that prevent device-level advantages from directly becoming system-level gains. Finally, we outline a cross-layer roadmap in which ferroelectric stack engineering, variability-tolerant arrays, precision-scalable architectures, and SoC-level integration are co-optimized to enable reliable HZO-based memory-centric computing.
Chen He, Wei Li, Jianjun Li et al.· Micromachines· 0 citations
The continuous advancement of Very Large-Scale Integration (VLSI) technology has been primarily driven by the scaling of Complementary Metal-Oxide-Semiconductor (CMOS) devices. For several decades, Moore’s Law has enabled exponential growth in transistor density, leading to remarkable improvements in computational performance and energy efficiency. However, as CMOS technology approaches the sub-5 nm regime, several fundamental challenges such as short-channel effects, leakage currents, process variability, power density, and quantum mechanical limitations have become increasingly significant. These challenges threaten the sustainability of conventional transistor scaling and necessitate the development of innovative design methodologies and emerging device architectures. This paper presents a comprehensive review of advanced CMOS scaling challenges, surveys recent research developments, and proposes an adaptive multi-gate CMOS framework aimed at improving performance, reliability, and energy efficiency in nanoscale VLSI systems. The proposed methodology incorporates FinFET-based structures, high-k dielectric materials, and intelligent power optimization techniques. Simulation-based analysis demonstrates improvements in transistor performance, reduced leakage power, and enhanced scalability for future integrated circuits. The study concludes that advanced device engineering and architectural innovations are essential to extend CMOS technology beyond conventional scaling limits.
Puttapaka Lokesh Puttapaka Lokesh, Akula Sathvika Akula Sathvika, P. K. P Kavitha· International Journal of Sci...· 0 citations