'So many opportunities, so little time'

While scientists and the private sector continue to invent and deploy new clean energy technologies at an impressive clip, early career researchers in the field face no shortage of problems to solve.

Postdoctoral scholars and doyens in energy research explored advances and remaining challenges for powering artificial intelligence, next-generation batteries, fuel cells, sustainable aviation, low-carbon cement, and other topics this spring at the second annual symposium of the Stanford Energy Postdoctoral Fellowship. One broad piece of advice to the postdocs came from Yet-Ming Chiang, the Kyocera Professor of Materials Science & Engineering at MIT, who has invented materials to overcome disparate barriers to sustainable, affordable, and secure energy.

“Don’t stay in your lane,” Chiang said. “My original lane was batteries, but now I’m also working on low-carbon transportation, sustainable manufacturing, mining, etc.” At the same time, his talk title underscored the urgency in solving the problems: “So many opportunities, so little time.”

Several of the Stanford University energy fellows may have taken the time element to heart as they were soon to set out on the next chapter of their careers. “This year, five of our postdocs are heading to faculty positions, and two will soon begin fellowships abroad,” said Audrey Yau, who runs the Precourt Institute for Energy program. “It’s very rewarding to see the impact of our fellows, who will go on to positions in Singapore, Japan, the U.K., and across the United States.”

The Precourt Institute, which resides in the Stanford Doerr School of Sustainability, opened applications for the 2027 cohort on July 1.

Energy for AI, and AI for energy

Debates about the wildly unknown future electricity needs of artificial intelligence distract attention from the impacts AI computing is already having, CalTech computing and mathematical sciences professor Adam Wierman said.

Early-career researchers might consider setting aside macro analyses, like country-wide forecasts, and instead focus their work on the problematic impacts of data centers on local communities today, Wierman said. Water taps ran dry in one town with a new “zero-water” data center, and tech companies are building many data centers in drought-stricken areas in Arizona and Texas. Data centers are noisy and can increase the air temperature six degrees for several miles around. In addition, few regional electricity grids in the United States and elsewhere can handle the huge surge in data center power demand.

“If a location has a grid voltage instability, the data centers disconnect and fire up their own generation,” Wierman said. “The sudden absence of that demand worsens the instability of the local grid.”

However, new data centers are bringing wide attention to an existing problem: dilapidated U.S. electricity infrastructure. Energy researchers could figure out ways to turn data centers from liabilities into assets that help manage aging infrastructure until it gets modernized, said Wierman. Data centers have flexibility in their geographic location and in the timing of their electricity and water use. If data centers stored local electricity supplies in big battery systems rather than build on-site generation, that could make local grids more reliable and alleviate some local environmental impacts.

Researchers could also investigate non-evaporative water cooling for data centers and locating new compute facilities in areas where solar and wind power get wasted due to transmission bottlenecks. Other local challenges require research, too, and almost all potential solutions will require motivating technology companies to deploy them with price incentives and shorter permitting times, Wierman said.

“I hope data centers can be a nexus for the sustainability community,” he said, “a living testbed for new technologies and a partner in infrastructure modernization.”

Increasingly, researchers and others are using AI to improve energy production and use. Postdoctoral fellow Nico Christianson described at the symposium his research developing AI-based methods to help utilities and large energy consumers better manage risk in the selling and buying of electricity. This can help them avoid the very large losses that are seen occasionally in U.S. power markets. The methods can also be used to help manage risk and promote resilience in broader energy systems applications, said Christianson, a member of the fellowship’s 2025 cohort who is headed to a faculty position at Johns Hopkins University.

Zisheng Zhang presented his research using AI and physics-based models to design less expensive boron-based catalysts for electrochemical energy storage and conversion, critical processes for advancing the energy transition. Zhang, a 2024 cohort fellow, will begin a professorship in Stanford’s Department of Chemical Engineering in the fall.

Energy storage

Linda Nazar, an award-winning professor at Canada’s University of Waterloo, is a leading voice in using solid electrolytes – typically ceramic – to transport electric charges between the positive and negative electrodes. Solid electrolytes generally conduct ions more efficiently than liquid ones, and they are expected to be safer in terms of combusting and leaking.

Researchers have converged on chloride-based solid electrolytes as particularly attractive, Nazar said. These electrolytes are mechanically soft (unlike most ceramic conductors), oxidatively stable, inexpensive to make, and free of rare earth elements, with electrochemical stability better than today’s lithium batteries. However, this technology, like others she discussed, “is not there yet” for a few reasons, so the topic offers plenty of opportunities for early-career researchers.

Meanwhile, many researchers continue to work on improving next-generation lithium-ion batteries, said Khalil Amine, leader of the Advanced Battery Technology team at the U.S. Department of Energy’s Argonne National Laboratory. His group's work on a new type of negative electrode – two decades in development – is evolving toward higher nickel, lower cobalt formulations for greater energy density, with high-manganese variants under development for improved safety and cyclability. However, Amine said, finding the right silicon-based material for the positive electrode to replace graphite – a significant supply chain problem – in lithium-ion batteries remains a long-standing barrier to silicon-based batteries.

Learning exactly how various charged molecules react with other chemicals will likely pave the way for future advances across many energy technologies. 2024 cohort’s Yecun Wu discussed his research using quantum sensors to optically capture and visualize the behaviors of individual single ions in liquid. Wu, who will join the Vanderbilt University faculty in 2027 after completing his Stanford fellowship, also translates fundamental discoveries into applications. He is applying atomic-level insights to engineer the ionic pathway in next-generation lithium-metal batteries with no dendrite formation over 1,000 cycles.

Kate Reidy, a Miller Postdoctoral Fellow at UC-Berkeley who will become an assistant professor of materials science and engineering at Stanford in September, also works at the micro level. Reidy discussed her "reverse design" research approach. Starting with a desired target property, she designs the atomic architecture that delivers it and then creates a recipe to build it atom-by-atom. The nanomaterials she creates have broad applications in energy, including for energy-efficient computing and other electronic devices.

Backing up the grid

Lithium-ion and similar batteries are unlikely to be the answer to grid-scale, multi-day or seasonal energy storage, a requirement for relying heavily on renewable energy that must be met inexpensively. Nevertheless, grid-scale storage technologies are making progress, said MIT’s Chiang. Form Energy, of which Chiang is a co-founder and chief scientific officer, is building a large iron-air battery system as part of replacing a coal-fired power plant in Minnesota. The system can cost-effectively store enough energy to overcome power supply problems lasting several days, he said. That's a need lithium-ion batteries almost certainly cannot fill economically.

No technology exists yet for seasonal storage, which saves excess solar power in the summer for use through the winter, said Chiang. Again, expectations for lithium-ion batteries are very low. “If you’re going to do anything in energy storage,” Chiang advised, “make it something you think lithium ion will never be able to do."

Policy and economics

While the symposium made clear that the technical pathways to solutions exist, paths to solving political and business challenges to the energy transition remain unclear, particularly in the United States. MIT emeritus professor and CEO of the EFI Foundation Ernest Moniz and Stanford Nobel prize-winning physicist Steven Chu – both former U.S. secretaries of energy – discussed the big picture for the energy transition toward the end of the symposium's first day. Whether the industry moves quickly enough and whether the political environment allows the right incentive structures to take shape remains genuinely uncertain, Moniz and Chu said. Some of the biggest obstacles to the energy transition, as is frequently the case, are not in laboratories, the two agreed.

Public-private partnerships are the key to deploying clean energy technologies quickly and on a large scale, said Karen Skelton, senior policy advisor to Secretary of Energy Jennifer Granholm during the Biden administration. Capitalizing on economic opportunities and avoiding catastrophic climate change depends on advocacy for government action at federal, state, and local levels, as well as the private sector’s continued commitment in developing and utilizing new technologies in response to government signals, said Skelton, a visiting scholar at the Precourt Institute.

Chenfei Qu’s work has directly influenced policy recommendations on China’s carbon emissions trading system and renewable policies. The 2025 energy fellow shared her research on applying the lessons learned from China’s emissions trading system for policies in emerging economies.

As the fellowship’s faculty director Yi Cui pointed out, good communication skills are critical for researchers in helping develop good policies. From former Secretary of State George Shultz, Cui learned to practice his presentations several times before testifying on Capitol Hill. “If you know how to communicate, you can have tremendous impact,” he advised his postdocs. “You’ve got to know how to sell. As a professor I have to do this with students every day.”

The application completion deadline for the 2027 cohort is Oct. 1, 2026. Application information and portal can be found on this page.