Tuesday, March 26, 2019

AI Accurately Predicts the Useful Life Of Batteries, Will Help Develop Better Batteries

In an advance that could accelerate battery development and improve manufacturing, scientists have found how to accurately predict the useful lifespan of lithium-ion batteries, used in devices from mobile phones to electric cars.

If manufacturers of cell-phone batteries could tell which cells will last at least two years, then they could sell only those to phone makers and send the rest to makers of less demanding devices. New research shows how manufacturers could do this. The technique could be used not only to sort manufactured cells but to help new battery designs reach the market more quickly.

MIT professor Richard Braatz, left, and William Chueh, assistant professor in materials science and engineering at Stanford, led researchers at their institutions who developed a better battery testing technique.

 Image credit: Amos Enshen Lu

Combining comprehensive experimental data and artificial intelligence revealed the key for accurately predicting the useful life of lithium-ion batteries before their capacities start to wane, scientists at Stanford University, the Massachusetts Institute of Technology and the Toyota Research Institute discovered. After the researchers trained their machine learning model with a few hundred million data points of batteries charging and discharging, the algorithm predicted how many more cycles each battery would last, based on voltage declines and a few other factors among the early cycles.

The predictions were within 9 percent of the number of cycles the cells actually lasted. Separately, the algorithm categorized batteries as either long or short life expectancy based on just the first five charge/discharge cycles. Here, the predictions were correct 95 percent of the time.

Published March 25 in Nature Energy, this machine learning method could accelerate research and development of new battery designs and reduce the time and cost of production, among other applications. The researchers have made the dataset – the largest of its kind – publicly available.

Stanford researchers developed a machine learning technique to identify how long batteries will last.
Credit: David Dewhurst

“The standard way to test new battery designs is to charge and discharge the cells until they fail. Since batteries have a long lifetime, this process can take many months and even years,” said co-lead author Peter Attia, Stanford doctoral candidate in materials science and engineering. “It’s an expensive bottleneck in battery research.”

The work was carried out at the Center for Data-Driven Design of Batteries, an academic-industrial collaboration that integrates theory, experiments and data science. The Stanford researchers, led by William Chueh, assistant professor in materials science and engineering, conducted the battery experiments. MIT’s team, led by Richard Braatz, professor in chemical engineering, performed the machine learning work. Kristen Severson, co-lead author of the research, completed her doctorate in chemical engineering at MIT last spring.
Optimizing fast charging

One focus in the project was to find a better way to charge batteries in 10 minutes, a feature that could accelerate the mass adoption of electric vehicles. To generate the training dataset, the team charged and discharged the batteries until each one reached the end of its useful life, which they defined as capacity loss of 20 percent. En route to optimizing fast charging, the researchers wanted to find out whether it was necessary to run their batteries into the ground. Can the answer to a battery question be found in the information from just the early cycles?


Stanford graduate students Nicholas Perkins, left, Peter Attia and Norman Jin are among the researchers who found the key for accurately predicting the useful life of lithium-ion batteries.

Image credit: Dean Deng

“Advances in computational power and data generation have recently enabled machine learning to accelerate progress for a variety of tasks. These include prediction of material properties,” Braatz said. “Our results here show how we can predict the behavior of complex systems far into the future.”

Generally, the capacity of a lithium-ion battery is stable for a while. Then it takes a sharp turn downward. The plummet point varies widely, as most 21st-century consumers know. In this project, the batteries lasted anywhere from 150 to 2,300 cycles. That variation was partly the result of testing different methods of fast charging but also due to manufacturing variability among batteries.

“For all of the time and money that gets spent on battery development, progress is still measured in decades,” said study co-author Patrick Herring, a scientist at the Toyota Research Institute. “In this work, we are reducing one of the most time-consuming steps – battery testing – by an order of magnitude.”
Possible uses

The new method has many potential applications, Attia said. For example, it can shorten the time for validating new types of batteries, which is especially important given rapid advances in materials. With the sorting technique, electric vehicle batteries determined to have short lifespans – too short for cars – could be used instead to power street lights or back up data centers. Recyclers could find cells from used EV battery packs with enough capacity left for a second life.

Yet another possibility is optimizing battery manufacturing. “The last step in manufacturing batteries is called ‘formation,’ which can take days to weeks,” Attia said. “Using our approach could shorten that significantly and lower the production cost.”

The researchers are now using their model to optimize ways of charging batteries in just 10 minutes, which they say will cut the process by more than a factor of 10.

Chueh is also a center fellow at Stanford’s Precourt Institute for Energy, which funded his exploratory work for this project. Other co-authors are Stanford students Norman Jin, Nicholas Perkins and Michael Chen; MIT Professor Martin Bazant, postdoc Benben Jiang and student Dimitrios Fraggedakis; Muratahan Aykol at Toyota Research Institute; Stephen Harris at Lawrence Berkeley National Laboratory and a visiting scholar at Stanford; and University of Michigan student Zi Yang, a Stanford intern.

This work was supported by the Toyota Research Institute, the Thomas V. Jones Stanford Graduate Fellowship, the National Science Foundation, SAIC through Stanford Energy 3.0, and the U.S. Department of Energy.


Contacts and sources:
Mark Golden
Stanford University




Researchers Create Hydrogen Fuel from Seawater



Splitting water into hydrogen and oxygen presents an alternative to fossil fuels, but purified water is a precious resource. A Stanford-led team has now developed a way to harness seawater – Earth’s most abundant source – for chemical energy.

Stanford researchers have devised a way to generate hydrogen fuel using solar power, electrodes and saltwater from San Francisco Bay.

Hongjie Dai and his research lab at Stanford University have developed a prototype that can generate hydrogen fuel from seawater

 Courtesy of H. Dai, Yun Kuang, Michael Kenney

The findings, published March 18 in Proceedings of the National Academy of Sciences, demonstrate a new way of separating hydrogen and oxygen gas from seawater via electricity. Existing water-splitting methods rely on highly purified water, which is a precious resource and costly to produce.

Theoretically, to power cities and cars, “you need so much hydrogen it is not conceivable to use purified water,” said Hongjie Dai, J.G. Jackson and C.J. Wood professor in chemistry in Stanford’s School of Humanities and Sciences and co-senior author on the paper. “We barely have enough water for our current needs in California.”

Hydrogen is an appealing option for fuel because it doesn’t emit carbon dioxide, Dai said. Burning hydrogen produces only water and should ease worsening climate change problems.

Dai said his lab showed proof-of-concept with a demo, but the researchers will leave it up to manufacturers to scale and mass produce the design.
Tackling corrosion

As a concept, splitting water into hydrogen and oxygen with electricity – called electrolysis – is a simple and old idea: a power source connects to two electrodes placed in water. When power turns on, hydrogen gas bubbles out of the negative end – called the cathode – and breathable oxygen emerges at the positive end – the anode.

But negatively charged chloride in seawater salt can corrode the positive end, limiting the system’s lifespan. Dai and his team wanted to find a way to stop those seawater components from breaking down the submerged anodes.

The researchers discovered that if they coated the anode with layers that were rich in negative charges, the layers repelled chloride and slowed down the decay of the underlying metal.

They layered nickel-iron hydroxide on top of nickel sulfide, which covers a nickel foam core. The nickel foam acts as a conductor – transporting electricity from the power source – and the nickel-iron hydroxide sparks the electrolysis, separating water into oxygen and hydrogen. During electrolysis, the nickel sulfide evolves into a negatively charged layer that protects the anode. Just as the negative ends of two magnets push against one another, the negatively charged layer repels chloride and prevents it from reaching the core metal.

Without the negatively charged coating, the anode only works for around 12 hours in seawater, according to Michael Kenney, a graduate student in the Dai lab and co-lead author on the paper. “The whole electrode falls apart into a crumble,” Kenney said. “But with this layer, it is able to go more than a thousand hours.”

Previous studies attempting to split seawater for hydrogen fuel had run low amounts of electric current, because corrosion occurs at higher currents. But Dai, Kenney and their colleagues were able to conduct up to 10 times more electricity through their multi-layer device, which helps it generate hydrogen from seawater at a faster rate.

“I think we set a record on the current to split seawater,” Dai said.

The team members conducted most of their tests in controlled laboratory conditions, where they could regulate the amount of electricity entering the system. But they also designed a solar-powered demonstration machine that produced hydrogen and oxygen gas from seawater collected from San Francisco Bay.

And without the risk of corrosion from salts, the device matched current technologies that use purified water. “The impressive thing about this study was that we were able to operate at electrical currents that are the same as what is used in industry today,” Kenney said.
Surprisingly simple

Looking back, Dai and Kenney can see the simplicity of their design. “If we had a crystal ball three years ago, it would have been done in a month,” Dai said. But now that the basic recipe is figured out for electrolysis with seawater, the new method will open doors for increasing the availability of hydrogen fuel powered by solar or wind energy.

In the future, the technology could be used for purposes beyond generating energy. Since the process also produces breathable oxygen, divers or submarines could bring devices into the ocean and generate oxygen down below without having to surface for air.

In terms of transferring the technology, “one could just use these elements in existing electrolyzer systems and that could be pretty quick,” Dai said. “It’s not like starting from zero – it’s more like starting from 80 or 90 percent.”

To read all stories about Stanford science, subscribe to the biweekly Stanford Science Digest.

Other co-lead authors include visiting scientist Yun Kuang from Beijing University of Chemical Technology and Yongtao Meng of Shandong University of Science and Technology. Additional authors include Wei-Hsuan Hung, Yijin Liu, Jianan Erick Huang, Rohit Prasanna and Michael McGehee.

This work was funded by the U.S. Department of Energy, National Science Foundation, National Science Foundation of China and the National Key Research and Development Project of China.


Contacts and sources:
Erin I. Garcia de Jesu / Amy Adams
Stanford University


Citation: Solar-driven, highly sustained splitting of seawater into hydrogen and oxygen fuels.
Yun Kuang, Michael J. Kenney, Yongtao Meng, Wei-Hsuan Hung, Yijin Liu, Jianan Erick Huang, Rohit Prasanna, Pengsong Li, Yaping Li, Lei Wang, Meng-Chang Lin, Michael D. McGehee, Xiaoming Sun, Hongjie Dai. Proceedings of the National Academy of Sciences, 2019; 201900556 DOI: 10.1073/pnas.1900556116