June 2022 ESRFnews
Users expose ageing of batteries Machine learning of X-ray data taken at the ESRF has revealed new patterns in the decay of lithium- ion battery electrodes. The results show that the decay is due to the properties of individual particles at first, but is then increasingly driven by multiple-particle interactions. Researchers at the SLAC
National Accelerator Laboratory in California, Purdue University in Indiana and Virginia Tech in the US, as well as the ESRF, performed X-ray nano-tomography at the ESRF s ID16A beamline to reconstruct 3D pictures of the cathodes of lithium-ion cells after they had gone through many charging cycles. The team then used computer learning to identify, in 2D slices of the images, more than 2,000 individual particles, along with their features such as size, shape and surface roughness, and global traits, such as how often they come into contact with one another. After 10 charging cycles, individual particle properties contributed most to the collective breakdown; but after 50 cycles, pair and group attributes were most
decisive (Science 376 6592). Battery particles are like people:
we all start out going our own way, says Keije Zhao, study author and mechanical engineer at Purdue. But eventually we encounter other people, and we end up in groups, going in the same direction. To understand peak efficiency, we need to study both the individual behaviour of particles, and how those particles behave in groups. For an in-depth look at machine learning and artificial intelligence at the ESRF, see p14.
Why diesel catalysts fail
Y IJ IN L IU /S LA
C N AT
IO N A L A C C E LE
O R L A B O R AT
O R YAn international team of scientists led
by the ESRF has discovered why the activity of catalysts for the removal of nitrogen oxides from diesel exhaust are easily restricted by the presence of just small amounts of sulphur dioxide. The results could help in the development of catalysts that work with a wider range of diesel fuels.
Nitrogen oxides present in diesel exhaust have adverse impacts on human respiration. They can be reduced by ammonia to harmless nitrogen and water in the presence of a catalyst. Among the promising catalysts for this reaction are copper-exchanged chabazite zeolites, which are stable and work well at low temperatures. But this catalyst is known to be suitable only for ultra-low sulphur diesel fuels, because even a small amount a few parts per million of sulphur dioxide present in common diesel fumes is known to drastically reduce the catalyst s activity.
Now Anastasia Molokova, a PhD student at the ESRF within the InnovaXN programme, together with colleagues from the ESRF, chemical manufacturer Umicore and the University of Turin, Italy, have determined the mechanism of poisoning by sulphur dioxide, by identifying the copper species in the catalyst that interact with it. They prepared copper-I and copper-II species with different ligands inside the pores of a copper-exchanged chabazite catalyst and exposed them to sulphur dioxide. By performing X-ray absorption and X-ray emission spectroscopy experiments, as well as measuring the sulphur-dioxide uptake, the researchers found that a complex of the copper-II species coordinated to both ammonia and oxygen reacted the most (JACS Au 2 787). The identified species are known to be the key intermediates in the nitrogen oxides reduction reaction, which explains why the catalyst deactivates upon exposure to sulphur dioxide. This finding is an important step towards developing catalysts resistant to sulphur poisoning.
A machine-learning algorithm s depiction of a battery cathode after 10 charging cycles, highlighting the most severely damaged particles.
S H U T TE
R S TO
C K /L A N A EL
C O VA
and transition-metal fuel cells Researchers led by the CNRS in France have found that novel fuel cells made of transition metals still perform after being aged artificially, despite changing in chemical composition. The results bode well for commercial use. Fuel cells are a promising means
of powering electric vehicles, but they often contain precious and expensive metals such as platinum. Those based on catalysts made of transition metals have become a promising alternative, including ones made of iron, nitrogen and carbon in alkaline electrolytes. However, scientists do not yet know which catalytic structures are most resistant to age-induced damage. The research group performed
experiments at the Université Grenoble Alpes as well as X-ray absorption spectroscopy at the ESRF beamline ID26 and wide-angle X-ray scattering at the ESRF beamline ID31 on iron-nitrogen-carbon catalysts in an alkaline medium. They aged the catalysts artificially, and found that iron that was atomically dispersed in the catalyst underwent irreversible changes, with 15% of the
F C A N D É
Study author and ESRF postdoc Viktoriia Saveleva at work at the ID26 beamline.
atoms dissolving in the electrolyte and 35% redepositing as iron-oxide nanoparticles. Meanwhile, iron- carbide nanoparticles in the catalyst transformed into iron-carbide nanoparticles with an iron-oxide shell. The changes did not destroy the catalytic behaviour, however (Appl. Catal. B. 311 121366). The technique of X-ray absorption
and emission spectroscopy on ID26 is a unique tool to investigate the active iron site of the catalyst, says Frédéric Maillard at the CNRS. It s provided us with clues about the steps to take to tweak the catalyst to make it better performing.