Earth Day is a stark reminder that the climate crisis is one of the most urgent issues of our time. And while the challenges that lie ahead are great, the promise of new solutions driven by science and technology is even greater.
To successfully tackle climate change, we need a new, much more efficient way to discover new materials — using cutting-edge science and technology. The traditional trial-and-error, centuries-old scientific method has served society well, but it’s incredibly costly and time-consuming. With climate, we don’t have time to waste — we urgently need to accelerate discovery with the help of AI, hybrid cloud, robotics and eventually quantum computing. We need to supercharge the scientific method.
Accelerated discovery relies on AI-generated hypotheses, AI-enriched simulations and automated testing to speed up the discovery process. The approach will be very useful in the discovery of materials for carbon capture, nitrate-rich fertilizers, environmentally friendly batteries, and treatments to novel viruses. Accelerated discovery will become just as critical for increasing the sustainability and resiliency of global supply chains and operations, and preparation for impending global events: such as future weather events and potential global health crises caused by climate change.
For the past year, IBM scientists have been using this new supercharged scientific method to accelerate the discovery of sustainable materials, with promising results.
Cracking down on CO₂
Capturing and separating carbon dioxide (CO₂) emitted into the atmosphere is difficult, as is safely storing these captured gases long term. The IBM Research team has turned to AI to accelerate the design and discovery of better polymer membranes that can more efficiently capture carbon at its point of emission.1
Using generative AI modeling, our researchers identified several hundred molecular structures that could enable more efficient and cheaper alternatives to existing separation membranes for capturing CO₂.1 We are now evaluating these candidate molecules with the help of automated molecular dynamics simulation on high-performance computing (HPC) clusters.
IBM Research has also created an AI and cloud-based tool that simulates how carbon dioxide flows through specific types of rock,2 allowing scientists to evaluate CO₂ trapping and, eventually, conversion scenarios. Ultimately, this could enable the rapid analysis and optimization of rock-specific requirements for mineralizing and storing CO₂ efficiently, safely and long-term.
To fight climate change, IBM has already committed to reaching net zero greenhouse gas emissions by 2030 . But we can’t do this alone. Earlier this year , IBM joined Apple, Boeing, Cargill, Dow, PepsiCo, Verizon and others as the inaugural members of the MIT Climate and Sustainability Consortium (MCSC) to accelerate large-scale implementation of solutions to address climate change. IBM also joined the European Green Digital Coalition as a founding member.
Discovering more sustainable materials
To ensure the sustainability of our planet, we urgently need materials that are sustainable on many fronts, including environmentally and on a humanitarian level.
For example, as essential as batteries are to daily life, their production can take a terrible toll on the environment. IBM Research is using AI, quantum computing and other advanced technology to explore different battery alternatives that could lessen these concerns. This includes working with industry partners to discover and test materials that can eliminate the need for heavy metals in lithium-ion batteries, as well as using quantum computing to explore the next generation of sustainable batteries.
IBM scientists are also using generative AI models to design completely new materials with specific desired characteristics, including photoacid generators (PAGs) —a critical class of materials used in semiconductor manufacturing that can be used to build more sustainable computing devices. Our researchers used an end-to-end AI-powered workflow to synthesize three novel PAG candidates, accelerating a discovery process that usually takes up to 10 years and up to $100 million.
Preparing for future health crises
Global health crises are increasingly linked to the health of our planet. According to the World Health Organization, between 2030 and 2050 climate change is expected to lead to approximately 250,000 additional deaths per year, caused by malnutrition, malaria, diarrhea, and heat stress.
Just as AI can play an important role in the discovery of new materials, it has tremendous potential to accelerate the rate of discovery in the field of life sciences. This is the basis for a new 10-year collaboration with the Cleveland Clinic, which will apply advanced computational technology to generate and analyze data to help enhance research in areas including genomics, single cell transcriptomics, population health, clinical applications, and chemical and drug discovery.
And last month IBM researchers used an AI generative model known as a “deep generative autoencoder” to learn more about peptide molecules—short strings of amino acids that are also the building blocks of proteins. That knowledge helped researchers create two new non-toxic antimicrobial peptides (AMPs) with strong broad-spectrum potency,3 outperforming other leading AMP design methods by nearly 10 percent.
Our work in these and other areas will only intensify in the coming years, as IBM makes an urgent, science-based, coordinated effort to fuel a more sustainable future.
- Hsu, H., Giro, R., Steiner, M., Hama, T. & Takeda, S. Small Data Enabled Prediction and Verification of Potential Polymer Membranes for CO₂ Separation. APS March Meeting (2021).↩
- Steiner, M., Ferreira, R., Ronaldo, G. Simulating fluid flow with carbon dioxide in digital rock. APS March Meeting (2021).↩
- Das, P., Sercu, T., Wadhawan, K. et al. Accelerated antimicrobial discovery via deep generative models and molecular dynamics simulations. Nat Biomed Eng (2021).↩