How Supercomputers and Quantum Tech Are Revolutionizing Materials Science
At Lawrence Livermore, cutting-edge supercomputing drives rapid, high-fidelity predictions of material properties, accelerating the design of batteries, capacitors, hydrogen storage systems, and ultra-wide bandgap electronics. By simulating billions of atoms, researchers explore materials under extreme conditions, model energetic materials safely, and connect atomic-scale interactions to device-level performance. Multiscale computational frameworks guide experiments, uncovering why materials behave as they do and informing next-generation designs. Looking ahead, quantum computing promises to revolutionize materials modeling and even optimize qubit technologies, creating a powerful feedback loop between computation and materials innovation. And at the heart of it all is a vibrant, interdisciplinary community pushing the boundaries of what’s possible.
Transcript
00:00:00 The extreme high performance computing facilities at Livermore allow us to predict very rapidly material properties at the very high fidelities. And these simulations and results then can be used to guide material selections and also optimizations processed. So in this way simulations and modeling
00:00:27 can reduce the cost and also reduce the time required for material development. Example include uh predicting advanced materials for lithium ion batteries, super capacitor, CO2 capture and conversions, hydrogen's production and storage and also ultra wide band gap materials for energy saving electronic device. One area of a particular interest for
00:00:54 our research at MSD is the simulation of materials at extreme conditions. By leveraging the supercomputers at the lab, we have been able to run some of the largest molecular dynamic simulations, sometimes in excess of 10 billion atoms. We've been able to simulate microscopic phenomena that relies on atomistic descriptions. The chemistry and performance of energetic
00:01:16 materials is another area of interest for MSD where we study the performance and safety of how to handle explosive via computational techniques. Why would one want to model high explosives on a computer? Uh this question actually has uh deep historical roots. Some of our earliest understanding of how explosives actually operate, what causes them to detonate, go back to early computer
00:01:45 simulations in the 1960s that were performed on Department of Energy computers. Those simulations which were looking at how material microructural defects interact with a shock wave uh were instrumental in understanding the actual physics of what makes an explosive detonate. The ideal use case of a Livermore weapon product is that it's never used. It is assembled. It
00:02:11 sits on a shelf. It acts as a deterrent and it's disassembled. And across that entire time history, which is many decades, we want no safety surprises and no uh aging surprises. So computation fills both the understanding how materials age but also projecting those forecasts forward in time so that we can make some sort of assessment closely coupled with experiment in most cases
00:02:38 about the longevity of our products. Here at the lab we we turn to really bridge the scales and trying to develop a multiscale modeling framework to really understand you know material performance. Uh for example we do a very detailed large scale automistic simulations and trying to understand atomic level interactions and how this leads to the desired structure property
00:03:06 relationship that we want to obtain for uh the hands of performance. And then we move to a larger scale of looking to the particle particle interactions and how you want to manipulate the microructures to give you the desired performance of the um of the material and eventually all this information will be collected and contributes to device level of design that how you want to um optimize
00:03:30 the the material's components and its functionality. The computational research is really important um for helping us on the experimental side uh for a couple of reasons. So one is that a lot of times when we make the batteries and we test them um we might get a certain performance um maybe it's not performing so well but we don't really know why. And there are a million
00:03:53 different reasons why it could be performing that way but we can't really test all of those experimentally. Um so it's really helpful to have the modeling uh to where they can look at things on a smaller scale on the atomic scale and really understand what's going on. Um if we know why our batteries are performing that way then we can try and develop new solutions to make them perform better.
00:04:19 So, we're really excited about uh the the use of quantum computing in the future for materials research uh because it will allow us to uh compute chemistry and materials properties at an unprecedented accuracy and with system sizes that really are not possible with classical computing. We're also engaging currently in using our material science
00:04:42 uh expertise to advance the state-of-the-art uh enabling technologies for quantum computing itself. Uh for example, we're investigating how to optimize fabrication and material selection for a range of quantum bit technologies including superconducting, semiconducting and ion trap uh cubits. So that you know we we expect our work
00:05:05 will advance the state of quantum computing so that we in the future can use it back to impact our material science in a in a really nice uh closed cycle.
>> What excit me the most about being in Lawrence Livermore National Lab is is people. We are surrounded by thousand of excellent energetics researcher with a very diverse background from physics,
00:05:32 chemistries, material science to computer science and that wide range of expertise makes Livermore become one of the idea placed for team work and you cannot get bored because there are always opportunities, new ideas, new directions that pop up almost every days.