In July 2023, the US Geological Survey released a report showing that at least 45 percent of the nation’s tap water is estimated to have one or more types of per- and polyfluorinated alkyl substances, or PFAS. These substances, known as “forever chemicals” because they break down very slowly, have been linked to adverse health effects, including a higher risk of low birth weight in children, certain types of cancer, greater obesity, and decreased fertility.
PFAS are not generally biodegradable, so figuring out how to break down their chemical bonds efficiently and eliminate them from the environment has become a major priority for scientists around the world. The problem is urgent not just because of the harmful health effects of PFAS but also due to their widespread use—in waterproof clothing, food packaging, firefighting foams, paints, adhesives, lubricants, and many other products.
The team at Good Chemistry, founded in 2021, believes that computational chemistry is the key to improving the quality of drinking water and removing forever chemicals from the environment. The company uses cloud-based high performance computing (HPC) technology to analyze PFAS molecules and determine how they can be broken apart. It also uses this technology to identify the molecules so that they can be removed from manufacturing processes.
“The accurate understanding of chemical reactions is the key to finding the best solution to break PFAS apart and remove them from the environment,” says Arman Zaribafiyan, founder of Good Chemistry.
A Computational Approach to Chemistry
After the introduction of supercomputers in the 1960s, their cost and size confined them to government agencies and research labs. But over the past 15 years, the cloud has made massive computing power more affordable and accessible to a wide range of organizations, including Good Chemistry. The Vancouver-based company—composed of computational and quantum chemists, software developers, machine-learning engineers, and quantum computing scientists—focuses on using computational chemistry, running digital simulations instead of traditional lab experiments, to change how new materials are discovered and designed.
As part of its efforts to deliver breakthrough levels of accuracy and speed, Good Chemistry has developed QEMIST Cloud, a computational chemistry platform that uses HPC clusters to expand what is possible with quantum chemistry simulations. Built on AWS infrastructure, QEMIST Cloud is a high-throughput, cloud-native software-as-a-service (SaaS) solution that computational chemistry developers can use to create chemical simulation applications and workflows for their own experiments and research. The platform brings supercomputing capabilities to research organizations, regardless of their size or resources.
“We can now [analyze] chemical reactions at a tremendous volume because of the unprecedented scale of the cloud and accuracy of our algorithms,” Zaribafiyan says.
With this approach, which leverages artificial intelligence and quantum computing, the company can anticipate how a molecule will behave before it is ever created in a lab.
Scaling Up to Combat Forever Chemicals
Since high-accuracy chemistry simulations require massive amounts of computing power, Good Chemistry partnered with AWS to get access to HPC technology without the burden of maintaining infrastructure. This helped Good Chemistry multiply the scaling capabilities of its QEMIST Cloud platform while reducing costs.
While customers can use AWS to run tightly coupled jobs, similar to those run on
traditional supercomputers, Good Chemistry takes a different approach. QEMIST Cloud can rapidly and automatically scale up to use more than a million distributed CPU cores—just by tapping into scalable, flexible AWS infrastructure rather than relying on fixed-size, on-premises computing clusters.
“Good Chemistry can take advantage of the massive capacity that goes beyond even the biggest traditional supercomputers,” says Dr. Neil Ashton, principal computational engineering specialist with the AWS Advanced Computing and Simulation team. “With the scale of AWS infrastructure, we could support a million cores in a single region. This means Good Chemistry can do more experiments and do them with higher accuracy, which wouldn't be possible without access to this level of capacity.”
AWS offers a wide range of pricing options for compute capacity, providing Good Chemistry and others the performance they need at an optimal price. “You don’t have to spend millions of dollars in infrastructure to get computing capability at this scale,” says Philip Ifrah, head of product at Good Chemistry. “Our solution on AWS orchestrates millions of computing resources on demand to perform experiments that push the boundaries of what’s possible.”
With access to scalable HPC, Good Chemistry has been able to accurately calculate the energy required to break the chemical bonds in three PFAS molecules: TFA, PFBA, and PFOA. The company’s work with PFOA, one of the largest and most notorious PFAS molecules, is especially noteworthy. Good Chemistry calculated the bond-breaking energy for PFOA in 37 hours, with only four hours at the peak of one million cores, a process that would have taken several years if the company had run the same simulations sequentially. According to the company, PFOA is the largest PFAS to be simulated with near-exact accuracy to date.
“By determining how to break down this molecule, Good Chemistry lays the foundation for creating engineering processes that can break down these chemicals outside the lab,” Ashton says. “This is just the first step. The speed with which Good Chemistry was able to do this opens the door to understanding and removing thousands of PFAS at an unprecedented rate, a process which could be scaled up and used around the world to improve the health of millions.”
Says Zaribafiyan, “We dynamically scaled QEMIST Cloud to one million cores on AWS, and by the next day, we were able to create a new solution that we couldn’t do before. All it took was the on-demand scalability of the cloud. It’s a game changer for HPC in material science and chemistry.”
New Scientific Discoveries to Improve the Planet
Good Chemistry’s work on QEMIST Cloud is a critical step forward for the remediation of PFAS and will likely play a major role in the discovery of new methods of breaking down the harmful molecules.
“Through this PFAS project, we demonstrated that we could run very high-accuracy calculations on AWS,” says Takeshi Yamazaki, director of research and development at Good Chemistry. “We are creating lots of high-quality data that will, in turn, help us offer differentiated machine learning models for material discovery.”
Good Chemistry is expanding QEMIST Cloud to support more industries, such as pharmaceuticals, advanced chemicals, energy, and automotive. Use cases in progress—including crystal structure prediction, virtual screening, and reaction pathway prediction—will significantly reduce the cost, time, and risk associated with new drug development. The company says that the platform will lead to the development of better batteries, more effective carbon capture, and better solar panels.
“Our infrastructure is entirely composable, so you can easily build the environment you need, tailored to your specific application,” Ashton says. “Using AWS technology, you can run millions of loosely coupled tasks, like Good Chemistry does, and the next day build up a cluster of thousands of GPUs working tightly together to train the latest generative AI model, like Stability.AI.”
Good Chemistry has been selected to join the AWS Clean Energy Accelerator, where it will work with leading energy organizations to solve pressing clean energy and decarbonization challenges.
“Right now, we’ve only scratched the surface,” Ifrah says. “We’re excited to extend our capabilities in computational chemistry, machine learning, and quantum computing to bring many new use cases to life.”
Learn more about how AWS cloud infrastructure can help you innovate to build and run secure and performant applications that create new customer experiences, improve efficiencies, and scale your business faster.
This story was produced by AWS and edited by WIRED Brand Lab.
