Exploring How AI Contributes to the Unearthing of Novel Materials
Your Weekly Shortcut to Deeptech Investing—Exclusive Trends & Startup Reports for VCs & Angels in Just 5 Minutes
By Vishal Katariya, Deeptech Contributor - Deep Science and Technology Investments at Ankur Capital.
Modern problems require modern materials, and modern materials come from computational discovery algorithms.
What is going on?
The rapid rise of artificial intelligence (AI) has promised to upend, at a fundamental level, the way we go about our tasks. GPT, to name a prominent example, has changed the way some of us go about our day-to-day activities. The impact of advanced computation and AI is felt across sectors, though, and materials science is one such sector where we see it having a paradigm-shifting impact.
What does it mean?
Materials science, just like other fields of science, has rapidly entered the third and fourth paradigms where it’s incorporated advanced computation and big data approaches respectively. This development, over the past three decades, has allayed fears of a stagnation in materials science. The biggest impact has been in the process of materials discovery; a process that used to be guided earlier by heuristics and rules of thumb is now accelerated and focused by computational tools.
A computation-based materials discovery process is not just an improvement over previous approaches, but marks an “inversion” of the problem. Instead of synthesizing and only then testing materials for certain desired properties, candidate materials can be computationally screened to have a high likelihood of those desired properties.
Why does it matter?
💸For markets:
AI-powered materials discovery is still in its nascent stages when it comes to commercial uptake, with most of its noteworthy discoveries thus far coming from academia. Recent research advances include AWS’ project on scaling up computational chemistry with the aim of supporting a circular economy and a new MIT project for material development that improves upon current deep learning methods.
However, both large industrial players as well as startups are quickly becoming active in this space. Given the difficulty of the task and the fact that expertise is required in both materials science and computational techniques, startups are generally spinning out of prominent research groups around the world. These startups have built AI-powered platforms to accelerate materials discovery, of course with different strengths and specialties. These platforms are used both by academic labs and industrial corporations. Citrine Informatics (founded 2013), Mat3ra (founded 2015), and Kebotix (founded 2017) work on this model and have successfully partnered with large industrial and chemical players across the world such as BP, Bayer and Mitsubishi Chemical. Just this year, Citrine Informatics, a platform for computational materials discovery, raised $16M in its Series C fundraise.
🧑🏿🤝🧑🏻For society:
There was a genuine fear in the mid- to late-20th century that materials science was entering a period of stagnation after its meteoric growth earlier. This concern has been put to rest by the new paradigms of computational and data-driven materials discovery, and the computational revolution in materials discovery comes at an opportune time. Our global climate change mitigation effort consists of a number of novel technologies to be deployed, each of which requires specialized materials to be developed, ranging from advanced battery electrodes and membrane-based electrolyzers for hydrogen production to coatings for solar panels.
We have already seen a number of successful academic projects using computational materials science to make important discoveries that are currently being deployed in the real world, some of which are polar metals, self-assembling polymers, and OLEDs. The maturation and commercialization of this science comes at a crucial time, just as we seek to turn the tide on climate change.
🔮What’s next?
Well, what’s not? In our modern age, we live surrounded by man-made materials, both structural and functional. As we develop further in a more sustainable manner, we will need newer, better, and greener materials that serve their purpose optimally. This was once seen as a daunting challenge because of the perceived stagnation in materials science, but these fears can be laid to rest as materials discovery has embraced a computational and data-driven approach.
Given the huge demand for new materials, there is enormous scope for both startups and established industry players to build competency in this space, and to discover new materials. The global climate change mitigation effort will require a bulk of these new materials, but many more will be needed in healthcare, biotechnology, and even agriculture. We are super excited about the developments in the space, and the increasing role that computational tools are playing in the deeptech innovation sphere.
✨ That’s all for today. Thanks for reading !
💖Like, follow and subscribe to our Community ! Stay tuned for our next Deeptech Analysis coming up with our Deeptech Insights Newsletter.
Much love Deeptechers!👋