Planck Technologies: Shaping the Future of Energy Storage by Computational Material Screening

With wind turbines and solar panels dotting landscapes, the transition to cleaner energy production is well underway. However, a critical challenge remains: the efficient storage of energy.

Researchers and engineers are working towards better energy storage solutions, mainly by improving battery technology and hydrogen storage—and at the heart of this is finding the right materials, as they fundamentally determine what energy storage systems are capable of. 

Planck Technologies was founded by Maryam Ghadrdan and Hamid Mehdizadeh in the spring of 2023 to leverage computational chemistry methods and machine learning to find better materials for green energy storage. With an unwavering commitment to sustainability, Planck Technologies is poised to redefine the material stack and boost renewable energy storage.

Learn more about the future of computational material screening from our interview with the co-founder and CEO, Maryam Ghadrdan

Why Did You Start Planck Technologies?

I am a chemical engineer by training, and I worked in research and industry before joining the innovation department of Equinor. There I had the chance to work on different projects aiming to develop better materials for the green energy transition. 

One day, a trader in energy marketing pitched us the idea of developing a better material specifically for gas storage. She had no background in materials but thought there could be a way to use metal–organic frameworks to make energy storage more efficient. This sounded interesting. I took a leap internally and became the project leader for that idea, managing various stakeholders, researchers, and engineers while also focusing on the financial and application side. 

This project showed me what a powerful tool computational material screening could be. While synthesizing a material in the lab can take ages, using computational methods, we could screen millions and millions of material candidates and optimize them for specific properties matching the needs of a given application. While the project in the innovation department was still early stage and mostly academic, I was excited about taking it one step further and bringing such materials to the market, and so I founded Planck Technologies. 

How Does Computational Material Development Work?

The quest for net-zero and renewable energies has opened up plenty of opportunities. But for now, we focus on just one problem: improving the performance of electrochemical and chemical energy storage, specifically battery cycle life and energy density.  

There are lots of battery materials already commercially available, so our focus is on specific parts of the battery where we can make the most impact. For everything else, we collaborate with established industry players. One example is finding a new cathode material. In this instance, we are working with a material manufacturing company to improve lithium iron phosphate (LiFePo4, in short LFP) batteries with a new formula, LFPX. The challenge is to find X, the additional ingredient that can make batteries better—in terms of higher cycle life—than the original LFP batteries. 

We’re using a mix of different methods, as each comes with unique advantages and disadvantages. I’m a big fan of first principles methods like density functional theory (DFT); however, they require a lot of simulation time to model a single material. That’s too slow for material screening. Instead, we use DFT to generate our own training data, combine it with more data from public material databases, and train neural networks to infer the properties of materials and thus screen quickly through loads of them. 

The main challenge is getting good, reliable training data. Publicly available databases are a good starting point but may include incomplete or incorrect data. So, we need to validate and clean the data and use simulations to generate our own training data. 

We use machine learning models to screen for material candidates with desired properties. Due to the vastness of the chemical search space, we start with the criteria of what we’re looking for and then tailor our screening process to meet those criteria. For example, we may search for materials that meet a certain level of conductivity or porosity.  

Then, we rough screen millions of materials and narrow it down to thousands and eventually a few candidates that we analyze comprehensively. Material screening is not only about the material’s performance at the nanoscale but also about its performance in real-world applications, commercial potential, and sustainability footprint.

Finally, we take into account the ambient conditions in our simulations, like temperature and pressure, since they can drive phase transitions that can alter the material properties abruptly and depend on the use case of the material. For example, in hydrogen storage, you can accommodate higher pressures, but you need high mechanical robustness for the materials to withstand those pressures. So we include these parameters in our problem definition.

How Did You Evaluate Your Startup Idea?

Building a material screening tool might not be the most sexy thing in the world, but it’s the best way to pursue our mission to introduce next-generation energy storage materials to the market. 

The overall market potential for energy storage is huge, so initially limiting our scope helps us address more concrete problems, be more interesting to investors, and get industry advisor’s knowledge, as each area has its own challenges and set of stakeholders. 

We use standard financial modeling techniques like estimating the net present value of a material to decide which projects we should pursue first. Whether it’s a startup finding its first project or an established company adding another one, both need to predict their future value and translate them to a net present value today. 

We are not a SaaS company, as we are not selling software but materials licenses. Our approach is to start projects with customers. So we approached them early on, getting the cathode project and another hydrogen project going. From there, we can go deeper and eventually license materials for mass production to really drive change in energy storage. 

What Advice Would You Give Fellow Deep Tech Founders?

Build a good network and reach out to people, as founding a startup can be lonely, and in the end, no one knows everything. I’ve been working in the industry for over ten years, and I am a chemical engineer at heart, but I had to learn a lot about the commercial side. The leadership program like Stanford Lead helped me a lot, and it’s a good idea to be part of a startup incubator or accelerator program like AdMaLab. Don’t be afraid to ask people for advice.

Comments are closed.