From mobile phones, trams to energy storage power stations, lithium batteries are everywhere in people’s lives, but due to the continuous loss of lithium ions during use, the longest life is only 6-8 years.
Peng Huisheng/Gao Yue’s team from the Department of Polymer Science at Fudan University broke the traditional design principles of lithium batteries. Through the combination of AI and organic electrochemistry, they successfully designed a lithium carrier molecule that allows waste batteries to be repaired without damage with a “shot”, increasing the life of lithium batteries by 1-2 orders of magnitude, and providing key technical support for the transformation of the battery industry.
“Precision treatment”: give old batteries a “heart shot” and increase the cycle life by 1-2 orders of magnitude
The active lithium ions in the battery are provided by the positive electrode material. The battery is scrapped after the loss of lithium ions reaches a certain level. This is the basic principle that lithium-ion batteries have been following since their advent in 1990. However, under this principle, lithium batteries can no longer meet the current and future electricity needs of humans.
For example, electric vehicle batteries can only guarantee a high-performance life of 6-8 years/1000-1500 charge and discharge cycles; low-temperature use will accelerate battery deterioration; energy storage power stations and extreme environment energy storage scenarios require battery life to be increased by an order of magnitude; the upcoming large-scale battery retirement and recycling may cause environmental pollution and waste of resources.
Faced with these realistic and urgent problems, the Peng Huisheng/Gao Yue team has been thinking about how to provide solutions through basic research innovation. Through principle analysis and a large number of experimental verifications, they found that the core reason for battery degradation is the reduction of active lithium ions, while other components remain intact.
“Why not develop transformative molecular drugs like treating diseases, and accurately and in situ non-destructively supplement lithium ions to batteries, thereby greatly extending their life and service time, rather than judging them as ‘dead’ and scrapping them for recycling?” Without the support of research precedents, the team boldly imagined breaking the theory that lithium ions depend on symbiosis with positive electrode materials in the basic design principles of batteries, designing a lithium carrier molecule, injecting it into the battery, and individually controlling the lithium ions in the battery.
This carrier molecule is like a drug. It can be injected into the old and decaying battery by “giving an injection”, accurately replenishing the lost lithium ions in the battery, realizing the non-destructive repair of the battery capacity, and providing a new way to deal with retired batteries.Using this technology, the battery still shows a healthy state close to that when it left the factory (96% capacity) after tens of thousands of charge and discharge cycles, and the cycle life has been increased from the current 500-2000 cycles to more than 12000-60000 cycles, which is the first case in the world. In addition, the binding rule that battery materials must contain lithium has also been broken, and it is possible to build batteries using green, heavy metal-free materials.
Out of nothing”: Innovative research paradigm, using AI to design complex organic molecules
The idea of ”giving an injection” to replenish lithium ions in batteries is clear, but the difficulty lies in how to “prescribe the right medicine”.To realize the concept of lithium carrier molecules, the molecules need to have strict and complex physical and chemical properties, including the electrochemical activity of the molecules, the range of decomposition voltage, solubility, air stability, chemical stability, acidity and alkalinity, the composition of decomposition products, reaction kinetics, molecular synthesizability and cost. Such molecular mechanisms have no precedent in the academic community and cannot be designed through traditional research paradigms, that is, relying on experience and intuition. To this end, the team adopted a new energy molecular design method assisted by artificial intelligence.After more than four years of exploration, the team successfully combined AI and organic electrochemistry to digitize the molecular structure and properties, and built a database by introducing a large number of related properties in organic chemistry, electrochemistry, and materials engineering technology. Using unsupervised machine learning, molecular recommendations and predictions were successfully obtained, and the lithium carrier molecule that has never been reported, lithium trifluoromethylsulfinate (CF3SO2Li), was successfully obtained, making the concept of AI for Science truly landed.
In traditional cognition, different molecules in the general organic molecule library “each have their own duties” in the fields of life health, chemistry and chemical engineering. “There is no organic molecule database in the battery field, so we use electrochemistry and chemical informatics knowledge to find and collect a large number of molecular fragments with potential functions, convert their chemical information into digital symbols, and recombine them to generate new molecules to form an energy molecule library with specific properties.” Gao Yue introduced.
Towards application: the cost accounts for less than 10%, with large-scale commercial potential
Exploring transformative basic research to solve practical problems and carrying out the full chain of “molecule-mechanism-material-device” work is the team’s firm goal.
Adhering to the purpose of solving practical problems, the research-related verification experiments are completed on real battery devices rather than models, so as to fully expose possible problems and solve them, thereby promoting the next step of industrial transformation, such as improving molecular reaction kinetics to avoid affecting the formation speed of the battery; exploring chemical preparation reaction paths, which can synthesize high-purity molecules at low cost and precision.
At present, lithium carrier molecules have passed initial experimental verification and are expected to account for less than 10% of the total battery cost. They have large-scale commercial potential and can be used for lithium replenishment, energy storage, and integrated light storage. The team is currently conducting large-scale preparation of lithium carrier molecules and cooperating with top international battery companies to strive to transform technology into products and commodities, and help the country’s leading development in the field of new energy.”If we can repair batteries through ‘injections’ in the future and make them recyclable, we can solve the problem of large-scale battery scrapping from the source and make the industrial ecology intelligent and environmentally friendly.” The team expects that this achievement will be applied as soon as possible to provide key technical support for promoting sustainable economic development and environmental protection.
