Many modern technologies, from optical communications and artificial intelligence (AI) hardware to advanced sensors and medical imaging, depend on photonic and semiconductor devices that precisely control the interaction between light and electrons. Designing these devices, however, remains a major challenge because existing simulation tools often require researchers to choose between modelling an entire device or capturing the detailed behaviour of electrons. Few can do both within the same model.
Researchers from the Singapore University of Technology and Design (SUTD) and National University of Singapore (NUS) have developed a new computational approach that extends the widely used open-source Particle-in-Cell (PIC) method with condensed-matter physics. The result is a single platform that can simulate a much broader range of light–matter interactions in metals, semiconductors and emerging quantum materials.
Published in Computer Physics Communications , the research, “ Particle-in-cell simulations of quantum plasmas ”, demonstrates how an established plasma physics tool can be adapted to study condensed-matter systems, opening new possibilities for designing photonic and quantum technologies.
Seeing the big picture without losing the details
Simulation plays a critical role in developing new photonic and semiconductor technologies. Engineers rely on virtual models to predict how devices will behave before investing time and resources in fabrication.
However, current simulation tools typically specialise in only one aspect of the problem. Some can accurately simulate an entire device but provide limited insight into how electrons behave within it. Others capture electron behaviour in great detail but are only practical for structures that are much smaller than real-world devices.
Imagine using Google Maps. Zoom out and you can see the whole city, but not the individual vehicles. Zoom in and you can follow the vehicles but lose sight of how traffic is flowing across the city. Existing simulation tools face a similar trade-off. The new approach combines both perspectives, allowing researchers to see the bigger picture without sacrificing the finer details.
“For many years, researchers have had to choose between understanding how an entire device behaves and seeing what individual electrons are doing,” said Associate Professor Wu Lin from SUTD's Science, Mathematics and Technology Cluster. "We wanted to remove that trade-off. By extending an established simulation method rather than building a new one from scratch, we've created a framework that gives researchers a much richer picture of how light and electrons interact inside advanced materials.”
Building on an established tool
Rather than developing an entirely new simulator, the researchers enhanced an established computational method already trusted by plasma physicists around the world.
PIC simulations are widely used to model charged particles and electromagnetic fields. However, they were originally developed for classical plasmas and could not readily emulate the quantum behaviour of electrons in condensed materials such as metals, semiconductors and graphene.
To overcome these limitations, the team introduced four new physics modules that extend PIC simulations to capture important quantum effects while retaining the method's core strengths. Together, these additions allow researchers to study a much wider range of nanoscale light–matter interactions within a single computational platform.
The researchers demonstrated the framework across diverse condensed-matter systems, including plasmonic metals, ultraviolet silicon nanostructures and graphene-based materials, highlighting its versatility for studying complex electromagnetic phenomena.
Accelerating innovation across multiple industries
This work could benefit researchers and engineers working on next-generation semiconductor and photonic technologies. More realistic simulations could reduce the time and cost needed to develop photonic integrated circuits, optical sensors, lasers, LEDs and solar cells by allowing more design decisions to be tested virtually before manufacturing.
In the longer term, the researchers envision the framework contributing to digital twins of photonic and semiconductor devices—virtual models that closely mirror their real-world counterparts and allow engineers to optimise performance before fabrication. Such capabilities could help accelerate innovation in AI computing, telecommunications, renewable energy and healthcare.
By making the modified PIC codes open source, the team hopes researchers worldwide will continue expanding the platform. This collaborative approach could help avoid duplication of effort and accelerate progress across the field.
"This is only the beginning," said Associate Professor Wu. "Because the framework is open source, we hope researchers around the world will continue expanding it with new material models and applications. Ultimately, we want to help create digital twins that allow photonic and semiconductor devices to be designed, tested and refined virtually before they are built."
Their next goal is to incorporate additional physics into the platform — including complex many-body interactions and detailed electronic band structures not yet captured — while extending it to realistic device-scale simulations. The team is also developing AI accelerators based on physics-informed neural networks to reduce computation time without compromising physical accuracy, bringing the vision of practical digital twins for photonic and semiconductor devices a step closer to reality.
Computer Physics Communications
Particle-in-cell simulations of quantum plasmas