By Kyle Proffitt
March 5, 2019 | Researchers from Argonne National Laboratory have used high-intensity X-rays to quantify how lithium ions arrange themselves into the graphite anode during a fast charge, revealing discrete “phases” that can be monitored during charging and discharging of Li-ion batteries. The work was published last month in Energy & Environmental Science (DOI: 10.1039/C8EE02373E).
A pervasive obstacle to making electric vehicles viable competitors with conventional gas-powered vehicles is the time needed to recharge batteries. For the common Li-ion batteries, charging involves the relocation of lithium ions from a metal oxide cathode, through a porous separator material, and into a graphite anode. Maximal energy density is achieved when these lithium ions homogeneously access and fill the graphite anode, and a slow charge allows this to occur. However, fast charging prevents the lithium from uniformly filling the graphite electrode.
Daniel Abraham and his team at the U.S. Department of Energy’s Argonne National Laboratory used the high-intensity and spectrally-focused X-rays of the Advanced Photon Source to study the real-time behavior of lithium ions as they migrate from one electrode to another in a coin cell-type battery. These powerful X-rays penetrate the stainless steel case of the batteries, and the diffraction patterns produced from different wavelength photons reveal specific quantized phases of lithium intercalation between graphene planes as the graphite crystal swells.
Using this technique, several snapshots can be acquired during the charging and discharging of batteries, and the local lithium ion concentrations can be spatially and temporally quantified. “We looked at the cross-section of the electrodes. We can see the lithium ions actually moving from one electrode to the other,” Abraham tells Battery Power Online. “Now, we obviously cannot see the lithium ions by themselves, but what we do see are the changes in the X-ray patterns of the graphite and of the NMC [the Ni-Mn-Co oxide cathode].”
Diffraction patterns reveal specific phases, or concentrations of lithium in the graphite anode. “You start off with graphite, which is nothing but carbons, and then you start putting lithium in it… it starts off with LiC48, then it goes to 24, 12, and 6; LiC6 is eventually the final part.”
Why Speed Kills
Unfortunately, a limit of intercalation into graphite is reached when lithium reaches a ratio of one lithium for each six-carbon hexagon in the honeycomb lattice of graphite—the LiC6 phase; additional lithium is more likely to deposit as metal.
Abraham explains: “If we try to push the lithium ions a bit too quickly, what happens is that, instead of it going between the graphene planes, it tends to plate on the graphite particles.” Plating involves the formation of metallic needles or dendrites on these graphite particles, and the free metal can react with the electrolyte to form a solid-electrolyte interface (SEI). “When metal deposition occurs, it tends to reduce the electrolyte and immobilizes a portion of the lithium; you cannot use it to cycle back and forth anymore,” Abraham says. The energy density of the battery is degraded over time.
And of course, if a little battery life depreciation isn’t enough, dendrites can create short circuits that start fires.
The speed of charging wouldn’t be such a problem with thinner electrodes, but battery manufacturers want to maximize energy density by increasing electrode thickness. “It takes a certain finite amount of time for [the lithium ions] to traverse the electrode thickness. This time is dependent on various factors, including things like tortuosity—how fast can it go through the electrode? What is preventing it from going? How fast does it intercalate into the graphite particles?—and the porosity of the graphite particles.” Tortuosity is technically a measure of the “twistiness” of the pathway through which the lithium ions must travel.
Abraham’s results were not wholly unanticipated. At a moderately fast rate of charging—1C, which corresponds to a 1-hour full charge or discharge—significant lithium ion concentration gradients formed across the graphite electrode. “The portions closest get filled up first. The portions further away get filled up later,” Abraham says. The key importance of the study, according to Abraham, is that “This is the first time we have been able to quantify… how the lithium ions traverse across the section”.
Models and some other optical studies have predicted that the higher-order lithiated graphite species (LiC6) would predominate at the separator-graphite interface. These gradients are dangerous because they decrease cell capacity, cause uneven cell aging, and create conditions favorable for lithium deposition. But Abraham emphasizes that his team’s quantifications are useful to build and validate models.
“Typically when people talk about these models, they’re talking about lithium concentrations in the electrolyte, whereas we are actually showing lithium concentrations in the graphite particles themselves,” he says. The results highlight the magnitude of the gradient problem. As the original paper states, “Our study suggests that due to the concentration heterogeneity, Li plating can be difficult to avoid near the surface even when care is taken to keep the average lithiation sufficiently low during the rapid charge.”
Surprises and Next Steps
There were some unexpected findings from the study. Lithiated species persisted in the deepest graphite layers during discharging. Instead of the lithium ions returning completely to the cathode, some lithium remained within the graphite, even after an extended low-voltage hold. Abraham calls this, “a very surprising finding. This is something that we have not really talked about too much. Our expectation was that once we charge at a fast rate, eventually the gradients will equalize over the cross-section.” In an ideal world, 100% of the lithium ions would cycle back and forth. “But we find that it doesn’t happen as readily as we thought it would,” Abraham says. The paper also emphasizes the surprising asymmetry between the charging and discharging profiles.
Having captured a more complete picture of lithium ion behavior, Abraham has ideas for future battery improvements. “The bottom line is that we have to look at different electrode designs. Can you create electrodes of different levels of porosity or tortuosity? If the lithium ions are able to move across the cross section very rapidly, then the likelihood of plating at the surface is much smaller,” he says. “Electrode tortuosity can be manipulated by coming up with different electrode designs. Can we create channels in the electrode that allow the lithium ions to move from front to back at a very rapid rate? Can we create electrolytes that can conduct lithium ions fast through the electrode?… Ideally, we obviously don’t want any gradients.”
Abraham and colleagues are also extending their work in other directions, with interest in studying pouch cells and analyzing these gradients in aged batteries that have experienced hundreds of charge cycles. They are investigating lithium behavior in the NMC. “It also forms gradients, but we don’t have it published yet,” Abraham says. A separate report from the Abraham group used a complementary method to detect lithium plating, but Abraham says that the X-rays can also reveal plating (DOI: 10.1021/acsaem.8b01975).
“If you had asked me this question two months ago, I would have said it’s not visible, but now we can see lithium plating… We are now refining the techniques. And once we are confident that we can see lithium plating consistently, reproducibly, and systematically we will write a separate article.”