New analytical methods help researchers peek inside energy devices

The devices around us that generate and store the energy that powers our electronics and propels our vehicles rely on intricate chemistry that often happens in tiny spaces. But those sometimes-nanoscale reactions are incredibly difficult to study, making scientists wish they could magically shrink to watch the chemistry in real time.

For example, Jasna Jankovic, a specialist in imaging methods at the University of Connecticut, would want to spy on hidden nanoscale pockets of water in materials used in automobile fuel cells. And Ana Flávia Nogueira, a solar-cell chemistry expert at Brazil’s University of Campinas, would try to observe how moisture and oxygen break down some solar-cell materials.

But these researchers don’t need magic to aid them in their scientific quest. They are part of a large group of scientists who design cutting-edge methods based on microscopy, spectroscopy, and other types of analysis to reveal hard-to-observe phenomena in the electrodes, catalyst layers, and other critical materials in batteries, fuel cells, and photovoltaic devices. Their aim is to use the new information to boost device performance, extend service life, and reduce costs.

Many of the latest studies employ new analytical procedures to interrogate some constituent, often a single component of a thin layer within a multilayer device, in its natural or working state. For example, scientists have probed chemistry in battery electrodes while they charge and discharge. Known as in situ or operando methods, these techniques often depend on cleverly designed energy devices and analytical instruments, as well as on sophisticated imaging software.

“These in situ and operando methods are absolutely transformative,” says veteran battery scientist Esther S. Takeuchi, who leads research groups at Stony Brook University and Brookhaven National Laboratory. Takeuchi has witnessed significant improvements in analytical methods over her long career, which includes more than 20 years as the head of R&D at a medical battery company before moving to academia in 2007. During her industry years, she invented the highly commercialized lithium-silver-vanadium oxide battery that powers implantable cardiac defibrillators. The science underpinning that lifesaving invention depended on laboratory methods that, by today’s standards, seem old fashioned and limited in analytical capability.

“It wasn’t that long ago that the only recourse we had was to test a battery under various conditions, cut the thing apart, take the pieces out, analyze them, and try to decipher what happened to the battery’s components,” Takeuchi says. Researchers would then track chemical and physical changes by comparing the used component with a pristine one. But that approach is not ideal.

“Before-and-after analyses are so unsatisfying,” Takeuchi says, because they leave researchers wondering whether the changes they detected were caused or influenced by exposing a component to air or removing it from the device.

Scientists often questioned whether they were monitoring the most relevant processes. “In some cases, the mechanisms that cause change are transient” and occur only while the device is running, she says. “If you take a battery apart and look at the pieces, that signature may be long gone.”

Thankfully, analysis of these energy devices is now much more sophisticated.

Looking inside batteries

Chongmin Wang and coworkers at Pacific Northwest National Laboratory (PNNL) came up with one of those sophisticated techniques to study a problem that afflicts lithium batteries: lithium dendrites.

Overcharging a battery can cause metallic lithium to accumulate on the battery anode surface. Under some conditions, the accumulated lithium forms spiky structures called dendrites that can grow long enough and forcefully enough to pierce the insulating separator film between electrodes, “like grass poking through a concrete walkway,” Wang says. If that happens, the wirelike metallic structure can reach from the anode to the cathode, short-circuit the battery, and start a fire or cause an explosion.

To reduce that risk, lithium-ion battery manufacturers use anodes made of graphite instead of reactive lithium metal, which exacerbates dendrite growth. If they could master the dendrite problem, however, battery makers could use anodes made of metallic lithium and boost charge-storage capacity by almost 10-fold.

To better understand the dendrite problem, Wang’s team wanted to watch the spiky structures form and grow in a running, custom-made electrochemical cell housed inside a transmission electron microscope (TEM). Wang, Yang He, Wu Xu, and PNNL coworkers outfitted their TEM with a specially designed atomic force microscope (AFM) cantilever. The ultrathin AFM arm functioned as one of the cell’s electrodes. It also applied force to—and measured the pushback from—a growing dendrite, simulating the forces exerted by and on the separator film.

The researchers controlled the voltage on the battery-in-a-TEM setup to stimulate dendrite growth, allowing them to capture the growth process in a high-resolution video. They found that as lithium accumulates on the anode, the metal forms a tiny spherical particle that grows fairly uniformly and sluggishly to nanometer dimensions. Then, as the particle exceeds a critical size, it suddenly spawns a long, skinny protrusion to minimize surface energy. Depending on the cantilever force, the dendrite buckled, kinked, or stopped growing, providing clues about barriers that could prevent the dendrite from piercing the separator and short-circuiting the battery (Nat. Nanotechnol. 2019, DOI: 10.1038/s41565-019-0558-z).

Another clue about how dendrites form came from the gas composition inside the TEM. The spiky lithium grew in a carbon dioxide atmosphere but not in nitrogen. That made the team wonder whether lithium carbonates on the dendrite surface promoted dendrite growth. If they did, that would imply that ethylene carbonate, a widely used lithium battery electrolyte solvent, is a bad actor. So the group conducted additional experiments, varying the concentration of ethylene carbonate in the electrolyte solution. As suspected, more carbonate spelled more dendrites. Omitting that solvent prevented dendrite growth.

Wang says the take-home message is that the battery community should experiment with electrolyte chemistry to find a composition that gives good battery performance and eliminates dendrites.

Takeuchi’s group has also developed new operando methods for ferreting out hidden battery secrets. In one recent study, her team used its new synchrotron-based energy-dispersive X-ray diffraction (EDXRD) technique to generate tomographic-like images of thick iron oxide electrodes in Li/Fe₃O₄ batteries.

Typical battery electrodes consist of a thin (less than 50 µm) film of an electroactive material mixed with a binder and other inactive materials applied to a metal-foil current collector. Takeuchi’s group aims to boost the amount of charge that can be stored in a battery and simultaneously reduce the overall cost by increasing the active fraction in electrodes by using thick films (about 500 µm) of an inexpensive material—in this case iron oxide.

But thick electrode films can be problematic. The components may not disperse uniformly, and ions and electrons may be unable to travel through the entire thickness, leaving much of the active material unused. Standard analytical techniques offer limited help in determining how extensive the problem is and whether modifications to the battery preparation are helpful.

So Takeuchi, Stony Brook’s Amy C. Marschilok, Kenneth J. Takeuchi, and others used their EDXRD method to look through metal-encased batteries and record time-lapse images that show how the electrochemical reaction progresses over time and moves throughout the electrode’s volume during battery discharge.

The team found that thick electrodes can work well if they’re designed properly. The team prepared some electrodes with needlelike, hollow carbon nanotubes, which the researchers thought would boost the porosity of the electrode material and improve electron and ion transport. The imaging method showed that those electrodes reacted uniformly throughout their entire thickness. In contrast, very little of the electrode material underwent electrochemical reaction when other forms of carbon replaced the carbon nanotubes (J. Phys. Chem. C 2019, DOI: 10.1021/acs.jpcc.9b04977).

Imaging techniques are also helping researchers improve sodium-ion batteries (SIBs). Replacing lithium with abundant and inexpensive sodium to make SIBs may sound like an obvious route to low-cost energy storage devices. Indeed, these batteries are being studied for grid-scale electricity use. But so far, they don’t work well. SIBs’ charge-storage capacity falls quickly after just a few charging cycles.

To help understand why, Jiajun Wang of Harbin Institute of Technology and coworkers used a synchrotron-based X-ray nanotomography technique and other methods to peer inside test cells as copper oxide, a model material, reacted electrochemically with sodium. The researchers found that CuO undergoes multistep reactions, forming Cu₂O and copper metal. By probing individual particles, the team found that these species adopt a core-shell structure with an unreactive CuO core surrounded mainly by copper metal. The outer structure acts as a diffusion barrier, impeding sodium-ion insertion and preventing the battery from holding charge.

The Harbin team experimented with an engineered form of CuO that has a pom-pom-like structure with microchannels that remain open to sodium-ion diffusion, which boosted the battery’s performance, at least initially (ACS Energy Lett. 2019, DOI: 10.1021/acsenergylett.9b01347).

Probing perovskite photovoltaics

Photovoltaic devices made with sunlight-absorbing metallo-organic perovskite materials have rocked the solar-cell world in recent years. Methylammonium lead trihalides and other perovskites are much less expensive to make and process than crystalline silicon, the standard photovoltaic material, yet they offer comparable performance. Both types of cells can achieve a sunlight-to-electricity conversion efficiency of roughly 23%.

But perovskite cells don’t stand the test of time; within days, the photoactive layer decomposes, especially in the presence of moist air, ruining device performance. In commonly studied materials such as formamidinium methylammonium lead trihalide (FAMA), the breakdown causes organic cations to diffuse through the film, leaving behind inactive regions enriched in PbI₂ and other compounds. The lack of a solution to this shortcoming has impeded commercialization of these solar cells and has led to numerous proposed strategies for stabilizing them.

Nonetheless, detailed knowledge of the degradation mechanism continues to elude scientists. “These inorganic-organic hybrid materials exhibit rapid lattice movement, and the electrical properties are modified by ion motion,” says Chris Case, chief technology officer of Oxford PV, a company commercializing the technology. Case adds that analytical techniques that can study ion migration in perovskite devices will lead to a better understanding of their operation and ultimately allow for maximum stability and performance.

Not just any technique will do. “It should combine spectroscopy for chemical sensitivity with nanoscale resolution,” says Hélio C. N. Tolentino, a scientist at the Brazilian Synchrotron Light Laboratory. The technique needs both because the degradation process begins within a hodgepodge of nonuniform, nanosized perovskite grains in the thin photoactive layer. Surface-averaging methods cannot track grain-to-grain variations.

To tackle that tough problem, Tolentino, together with University of Campinas scientists Flávia and Rodrigo Szostak, and others, has just developed a technique called nano-FTIR that combines scanning nanoprobe microscopy and infrared vibrational spectroscopy for mapping the chemical diversity of individual perovskite grains. The method calls for shining synchrotron light on an AFM tip as it scans the film. The tip functions like an antenna, confining the area interrogated by the method to a nanometer spot, comparable to the AFM tip radius (Sci. Adv. 2019, DOI: 10.1126/sciadv.aaw6619).

The team studied FAMA and other perovskites and found that the method can use spectroscopy signatures associated with cation depletion and the accumulation of breakdown products to pinpoint nanoscale regions undergoing early stages of degradation. Now that the team can spot exactly where the trouble begins, the next step is to use the method to understand the damage mechanism and how to stop it.

Analyzing fuel cells

After going through years of testing, hydrogen-powered fuel-cell cars are just starting to be commercialized for private use. Mass adoption of these cars could reduce global dependence on petroleum and benefit the environment because water, produced via the electrochemical reaction of hydrogen and oxygen, is the only tailpipe emission from a fuel-cell car. But because fuel-cell components have limited durability, manufacturers are proceeding slowly in rolling out the vehicles.

One factor that controls fuel-cell performance is the ease with which reactants reach the catalytic electrodes and products leave them. That parameter is hard to control and even harder to probe, especially at the nanoscale. The task is difficult because the electrodes are covered with a heterogeneous layer of nanosized, porous clumps—agglomerates—of a charged polymer called an ionomer and precious-metal catalysts deposited on a carbon support.

To make headway into analyzing nanoscale mass transport through that complex material, researchers at five US national labs—Los Alamos, Argonne, Lawrence Berkeley, and Oak Ridge National Laboratories and the National Renewable Energy Laboratory—are pooling their talents.

In a recently published study, Deborah J. Myers, a research group leader at Argonne, and coworkers reported on a new X-ray computed tomography method with nanometer resolution, called nano-CT, that the team developed and applied to studying the cathode in a Toyota Mirai fuel cell.

The imaging technique allowed the researchers to analyze the role played by carbon morphology and agglomerate size in controlling reactant-product transport and electrode performance. Using the method, the researchers simulated the complex interplay between fuel-cell humidity levels and the size, shape, and structure of agglomerates, and they identified conditions to optimize fuel-cell performance (J. Electrochem. Soc. 2019, DOI: 10.1149/2.0082001JES).

By any measure, today’s energy devices are sophisticated and complex. And they often depend on ultrathin layers of nanostructured materials. Improvements that increase the efficiency and reduce the costs of these devices are likely to come with additional complexity. And that’s going to require ever-more-sophisticated analytical tools. Researchers won’t be able to magically shrink themselves to the nanoscale to watch these devices’ chemistries in action. But with ongoing advances in instrumentation, they’re getting closer all the time.