“All of this is connected together,” Larsen said. “There are microbumps, TSVs, microTSVs, and there are interfaces such as HBM and HBI. There are interconnections between die that have to be taken care of, as well.”

Low-k materials are essential for reducing parasitic capacitance and enhancing signal integrity in next-generation mixed-die packages. This research focuses on designing and analyzing the reliability of these materials under thermal and mechanical stress, crucial for improving performance and longevity in highly integrated systems. By optimizing the dielectric properties and addressing failure mechanisms, this work pushes the boundaries of microelectronic packaging, paving the way for faster, more efficient semiconductor devices.

Design Challenges Increasing For Mixed-Die Packages (semiengineering.com)

2. Innovative Process engineering for N+5 generation

"Innovative process engineering for the N+5 generation anticipates breakthroughs in semiconductor manufacturing nearly two decades ahead, where precision and scalability will define the future of nanoelectronics. This research tackles the extreme challenges of atomic-scale patterning, material integrity, and quantum effects, setting the stage for technologies that will shape industries 20 years from now. By developing processes that push beyond today's limits, we're laying the foundation for the semiconductor innovations of tomorrow."

https://semiwiki.com/semiconductor-manufacturers/tsmc/300552-vlsi-technology-forum-short-course-logic-devices/

(1) Innovation for Channel Mobility, Gate EOT Scaling

"In GAA technology, enhancing channel mobility involves optimizing the effective mass and scattering mechanisms within the channel. This is achieved through the incorporation of high-mobility materials like Ge or III-V semiconductors, which reduce the effective mass of carriers, and by applying tensile or compressive strain to further enhance carrier velocity. For gate EOT scaling, the physical challenge lies in reducing the thickness of the gate oxide without increasing leakage currents. This requires the use of high-k dielectrics like HfO₂ to maintain strong gate capacitance while minimizing quantum tunneling effects. The precise engineering of the gate dielectric interface, including reducing interfacial trap densities and maintaining a pristine high-k/semiconductor interface, is crucial for preserving electrostatic integrity and ensuring sub-threshold slope control at atomic scales."

https://semiwiki.com/semiconductor-services/techinsights/324910-iedm-2023-2d-materials-intel-and-tsmc/

(2) Contact process Innovations

"Reducing contact resistance in GAA transistors requires precise control of the metal-semiconductor interface, particularly at the source/drain (S/D) junctions and middle-of-line (MOL) structures. This involves advanced techniques such as selective epitaxy for minimizing Schottky barrier heights, and optimizing dopant activation to enhance carrier injection efficiency. The research focuses on mitigating Fermi level pinning and reducing interfacial states to lower contact resistivity, while maintaining high channel mobility. Careful MOL process integration, including metal line scaling and barrier layer engineering, is critical to achieving low-resistance contacts without compromising device electrostatics, ensuring the scalability of GAA for sub-3nm node technologies "

https://www.semianalysis.com/p/iedm2022p1

(3) BEOL Interconnect Innovations

"Innovations in BEOL interconnects for sub-3nm nodes focus on overcoming the critical limitations of resistivity and signal delay, which intensify at advanced scales. This research explores next-generation materials such as 2D materials, graphene, and carbon nanotubes, offering ultra-low resistivity and enhanced electromigration resistance. Advanced integration techniques like atomic layer deposition (ALD) for ultra-thin diffusion barriers and bottom-up metal filling are essential to minimize parasitics. Additionally, disruptive innovations such as 3D interconnect architectures, hybrid bonding, and quantum interconnects are being investigated to meet the demands of future high-density, low-power devices, where conventional copper interconnects may no longer suffice"

https://semiengineering.com/breaking-the-2nm-barrier/

3. Phase transition characteristics of inorganic/organic and hybrid materials for N+6 generation

"A phase transition is the transformation of a material from one state of matter to another, driven by changes in temperature, pressure, or other environmental conditions. For N+6 generation technologies, understanding the phase transition characteristics of inorganic, organic, and hybrid materials is critical. This research explores how these materials behave under extreme conditions, impacting their electronic, optical, and mechanical properties. These findings will guide the development of ultra-advanced devices and systems, projected for use nearly 25 years into the future, where material performance will be key to unlocking new technological capabilities."

(1) Organic Semiconductor phase transition

"In organic semiconductors, phase transitions offer a powerful means to tune electronic and optical properties, enabling adaptable and high-performance materials for next-generation devices. The ability to control molecular arrangement and phase behavior at the nanoscale allows for precise engineering of conductivity, charge mobility, and response to external stimuli, making organic semiconductors highly versatile for advanced applications."

https://onlinelibrary.wiley.com/doi/full/10.1002/smll.201906109

(2) Inorganic Semiconductor phase transition

"In contrast to organic semiconductors, inorganic semiconductor phase transitions are more rigid and predictable, typically offering stability under extreme conditions like high temperatures and pressures. While inorganic transitions focus on tuning electronic properties through crystalline lattice changes, organic semiconductors excel in flexibility, with phase transitions that allow for dynamic molecular rearrangement, enabling greater tunability in electronic, optical, and mechanical properties. This versatility makes organic materials ideal for flexible electronics and adaptive systems, where responsiveness to environmental changes is key."

https://pubs.acs.org/doi/full/10.1021/acs.nanolett.6b00719

(3) Hybrid Semiconductor phase transition

"Hybrid semiconductors combine the benefits of both organic and inorganic materials, with phase transitions that offer a unique interplay between flexibility and stability. These materials exhibit dynamic molecular behaviors typical of organic semiconductors, while maintaining the robust electronic properties of inorganics. Hybrid phase transitions can be finely tuned to control charge transport, optical responses, and mechanical resilience, making them ideal for advanced applications where both adaptability and durability are required, such as in next-generation optoelectronics and energy-efficient devices."

https://onlinelibrary.wiley.com/doi/full/10.1002/adfm.202207326

4. Materials design & Electric, Optical, and Structural Analysis for Advanced Light detection

(1) Helical Polymers for Circularly Polarized Light Imaging

"Chirality is a property of asymmetry important in several branches of science and widespread in nature and common in the most abundant forms of matter. Using their unique properties, chiral electronics have great potential for use in chemical synthesis, pharmaceutics, catalysis, fundamental physics, and biomedicine."

Helical polymers provide a cutting-edge platform for circularly polarized light imaging, with their unique chiral structures enabling selective interaction with circularly polarized light. This research investigates the optical activity and molecular alignment of these polymers, driving innovations in polarization-sensitive sensors. The ability to detect and manipulate circularly polarized light opens new avenues in areas such as bioimaging, optical communication, and quantum computing, making this a crucial development in advanced photonic materials.

Temperature-Tunable Circularly Polarized Light Detector Based on a Nodal-Line Semimetal ZrSiSe | ACS Photonics

(2) Twisted Crystalline Organic Semiconductor IR light detection

"Twisted crystalline organic semiconductors offer a revolutionary approach to infrared light detection by leveraging their unique molecular geometry to enhance charge transport and light absorption. This research explores the interplay of molecular twisting and crystalline order, unlocking new potentials for tunable and highly sensitive IR sensors. It addresses key challenges in organic electronics, offering solutions for low-cost, flexible, and efficient detection systems crucial for future optoelectronic applications"

Twisted Crystalline Organic Semiconductor Photodetectors - Jeong - 2023 - Advanced Functional Materials - Wiley Online Library

(3) Ultra fast Infrared photoresponse

"Ultra-fast infrared photoresponse delves into the precise mechanisms of carrier dynamics and photon-electron interactions at femtosecond timescales, critical for breakthroughs in high-speed optoelectronics and quantum technologies. By exploring the nonlinear optical properties of materials and ultrafast charge transport under extreme conditions, this research addresses the fundamental challenges of signal processing, energy efficiency, and the limits of detection in modern physics, making it a cornerstone of next-generation photonic systems"

Ultrafast Infrared Photoresponse from Heavily Hydrogen-Doped VO2 Single Crystalline Nanoparticles | Nano Letters (acs.org

5. Physics-Informed AI & Digital Twins for Materials and Devices.

(1) AI-Driven Inverse Design of Architected and Functional Materials

"Artificial intelligence enables materials design to evolve from empirical optimization toward predictive and inverse discovery."

Physics-informed artificial intelligence provides a powerful framework for connecting material architecture, processing conditions, and multifunctional properties across multiple length scales. Our research integrates experimental characterization, finite-element and multiphysics simulations, machine learning, and deep learning to uncover hidden structure–property relationships and inversely design materials with targeted mechanical, electronic, optical, and energy functionalities.

Rapid inverse design of metamaterials based on prescribed mechanical behavior through machine learning | Nature Communications

(2) Neuromorphic Electronics & Intelligent Memory Devices

"Brain-inspired electronic devices provide a physical platform for learning, memory, and information processing beyond conventional von Neumann architectures."

Our research explores emerging memristive and ferroelectric devices for neuromorphic and in-memory computing. By integrating device physics, nonlinear dynamics, reservoir computing, and machine learning, we investigate adaptive synaptic behavior, temporal information processing, and hardware-efficient artificial intelligence. These approaches provide pathways toward energy-efficient computing systems that directly exploit the intrinsic dynamics of functional materials.

Physical reservoir computing with emerging electronics | Nature Electronics

Ultralow energy adaptive neuromorphic computing using reconfigurable zinc phosphorus trisulfide memristors | Nature Communications

(3) Predictive Digital Twins for Phase-Transition and Optoelectronic Materials

"Digital twins connect experiments, physical models, and artificial intelligence to predict complex material states across multiple scales."

Our research develops physics-informed digital twins by combining structural, electrical, optical, and spectroscopic measurements with atomistic and continuum simulations. Deep learning and neural-operator frameworks are used to connect microstructure, phase evolution, transport behavior, and device response, enabling rapid prediction of material properties and unexplored operating regimes beyond conventional numerical simulation.

Digitalizing metallic materials from image segmentation to multiscale solutions via physics informed operator learning | npj Computational Materials

6. Autonomous Semiconductor Manufacturing & AI-Driven Process Intelligence

(1) Intelligent Plasma Processing & Endpoint Prediction

"Real-time process intelligence enables plasma processing to evolve from recipe-based operation toward predictive and adaptive control."

Our research integrates plasma diagnostics, equipment signals, process simulations, and deep learning to predict semiconductor etching processes in real time. Multimodal information from optical emission spectroscopy, electrical signals, mass spectrometry, and process sensors is used for endpoint prediction, plasma-state identification, fault detection, and feature-profile optimization.

Human–machine collaboration for improving semiconductor process development | Nature

(2) Virtual Metrology & Autonomous Process Control

"Virtual metrology transforms multidimensional equipment data into real-time predictions of wafer-level manufacturing outcomes."

Our research develops machine- and deep-learning frameworks for predicting critical process outcomes directly from equipment and sensor data. By integrating process history, equipment-state variables, physical models, and measured wafer data, virtual metrology can estimate film thickness, critical dimensions, uniformity, and process drift while reducing dependence on conventional post-process measurements. These predictive models provide the foundation for run-to-run optimization and closed-loop autonomous manufacturing.

Virtual metrology in semiconductor manufacturing: Current status and future prospects

(3) AI-Enabled Heterogeneous Integration & Advanced Packaging

"Data-driven co-optimization provides a pathway toward reliable, high-density three-dimensional integration beyond conventional device scaling."

Our research combines multiphysics simulations, semiconductor integration technologies, and data-driven modeling for next-generation heterogeneous systems. Key areas include Cu–Cu hybrid bonding, through-silicon vias, chiplet integration, thermal–mechanical reliability, and high-density three-dimensional interconnects. Machine-learning-assisted design and reliability prediction are used to explore complex interactions among process conditions, interfaces, architecture, thermal management, and electrical performance.

Advanced packaging of chiplets for future computing needs | Nature Electronics

Wafer-to-wafer hybrid bonding at 400-nm interconnect pitch | Nature Reviews Electrical Engineering

Page updated

Report abuse