Making Batteries Denser And Safer

Battery technology is improving swiftly, driven by the rapidly rising demand for electric vehicles and the vast body of knowledge developed by the semiconductor industry.

The market for electric vehicles (EVs) is on a fast upward trajectory, with global sales predicted to grow more than 12 times to more than 31 million vehicles. In fact, EVs will account for almost a third of new vehicle sales by 2030, up from approximately 2.5 million vehicles in 2020, according to Deloitte. Alongside of this shift, consumer demand for better vehicle performance and sustainability means OEMs and automotive ecosystem must deliver better batteries and the electronics to manage them.

The history of EVs dates back more than 130 years. A series of breakthroughs with batteries and electric motors in the 1800s led to the first electric vehicle on the road in the U.S. about 1890. By 1900, electric cars made up about one-third of all vehicles on the road, and continued to grow steadily for more than a decade until the mass-produced internal combustion engine-based Ford Model-T outpaced its electric counterparts.

Another confluence of technology advancements, as well as environmental and cost concerns, have again brought EVs to the forefront of automotive development.

Fig. 1: Better and cheaper batteries might have made a difference. A Riker electric vehicle, circa 1900. Source: Smithsonian/National Museaum of American History

Next-generation batteries aim to target EV adoption barriers such as cost, carbon footprint, and range per charge. That requires advanced battery lifecycle management and data collection/analysis to achieve sustainability goals, along with reuse and recycling. And due to close adjacencies between semiconductors and vehicle batteries, semiconductor ecosystem players are well-positioned to play a role in next-generation battery developments.

“It’s the same physics that drive a battery and a semiconductor device,” said Søren Smidstrup, senior manager for R&D at Synopsys. “It’s the same fundamental equations. We sometimes refer to the simulations we do as ‘ab initio’ simulations, meaning from the beginning. From a simulation point of view, it’s all of the same quantum mechanical equations we solve. The application area is different, so there’s a lot of domain knowledge required, but it is the adjacency that we’re trying to level using the ability to simulate this, using only the fundamental physical laws that drives this. So, while batteries and EDA may seem distant in terms of physics, they are adjacent.”

The end goal is increased energy density in batteries, and that comes down to choosing the proper chemistries. Chemistry has provided continuous increases in storage capacity using the same size battery cell.

“That pursuit continues today in the industry, and there have been a lot of improvements in that area with companies across the automotive ecosystem getting involved,” said Puneet Sinha, director of new mobility at Siemens Digital Industries Software. “Even in the last 10 years, the energy density of batteries has almost tripled. That means in the same weight of battery, we are putting in three times more energy.”

This has been achieved through improving the chemistry within each cell, which has improved significantly, and it explains why EVs today that provide around 250 miles of drive range are at the same price point as much shorter ranges in 2010. For example, a Nissan Leaf 11 years ago had about a 100 mile range. A Nissan Leaf today has a 2X driving range for the same price, Sinha said.

Within the chemistries of both the anode and cathode, improvements are needed to make cells more energy dense. “In the past, some graphite-based or carbon-based material have been common for anodes, whereas some combination of nickel manganese cobalt was utilized for the cathode to increase the energy density. Currently, one thing that has been identified in the industry, especially on the supplier side, is the need to unshackle the chemistries from graphite only as one of the anodes.”

Here, various players within the automotive ecosystem are looking into using new materials, like a combination of silicon and carbon in the form of graphite, as an anode so that it can accept a lot more lithium in the same space. This is key to getting higher energy density for the cells. But it also adds new challenges, given that silicon expands as it takes on lithium.

“Think of a board, which is expanding and contracting continuously depending on whether it is taking lithium or giving back lithium,” Sinha said. “Due to this shrinking and swelling, silicon particles can start decomposing or breaking apart, and as that happens the energy density goes down and the life of the cell starts decreasing rapidly. Because of that, there are a lot of activities in the industry to make silicon more robust so that when it is shrinking and swelling, cracks are not developing and the integrity of material continues to give the high energy density for a long period of time.”

Being able to simulate this on an atomistic level is essential. “We start on the raw material side and look at what type of materials, what type of new ideas could work on anode or cathodes,” said Ronald Gull, director, consulting and engineering services for TCAD at Synopsys. “What type of new structures in the separators could help to mitigate some of the problems like the spiking of the dendrites that basically destroy the battery? Or what type of electrolytes could be used that are more efficient from the fundamental physical properties of the materials point of view?”

As with many complex designs, technology is drawing more heavily from across periodic table than ever before. The good news for the battery industry is that the chip industry has explored this in depth.

“We can look to the periodic table to see what’s happening, and this is our own turf, so we are looking at silicon carbide, germanium, gallium arsenide, phosphates, boron,” Gull said. “This is where we do most of our work in the semiconductor space. Even in the battery optimization for the classical batteries, there is some movement where we try to get away from carbon more to silicon or aluminum. These atomistic properties change, and they change behavior, so we need to see what can actually work and what cannot work.”

Another area with a lot of activity right now is to replace cobalt in standard batteries, mostly because it’s very expensive, but also because it is limited in quantity. “There are material candidates in this domain that can be looked at, and combinations thereof,” he said. “Then we look toward the ions, since most batteries are constructed around lithium ions. It’s a very small atom. It’s very mobile. It’s very quick. But it also includes the risk that it is very reactive, flammable, and can lead to explosions. Because of this, there is research into finding new ions, such as sodium, magnesium, or even cadmium.

Solid-state batteries also are being considered as an alternative, and there is research at the atomistic level to try to find out what’s happening at the interfaces of different chemicals, how it can be optimized, and how industry can work with them.

“Research into solid-state batteries has received a lot of funding in the startup world,” Gull noted. “The big hope is that we can go to a solid-state electrolyte, going away from the liquid state, which would allow for a very safe battery with high performance at a similar cost, but much higher reliability — especially against the explosions or thermal issues.”

Still, it is really hard to go to new battery architectures, so simulation is a good way to try new material configurations quickly.

“If you have four or five different ingredients with different compositions and different crystalline structures, the exploration space gets very big. You’ve got to try the combinations of materials in different constellations, at different temperatures, under different pressures, to see how they behave. By doing this computationally, you can narrow down the promising approaches and promising materials without going through experimentation, without building batteries and testing [in the real world], and come up with new creative ways to create the next material for your battery,” he explained.

Electrical-thermal considerations
When the battery design is then scaled up to large-scale cells, electrical-thermal behavior comes into play. Every electrochemical reaction generates heat, and depending on how the heat is transporting and propagating in a given area as the temperature rises up, that increases the rate of the electrochemical reaction, or vice versa.

“Heat feeds into the reaction, and if the temperature is not distributed for any reason — it can be the design or the packaging — it can create even more non-uniformity in the operation of the cells,” Sinha said. “So we have to look at the cell in terms of what is happening at the material level and the thermal behavior. Simulations are a great way to connect all of this together, and to complement the testing.”

However, it’s one thing to look at electro-thermal interactions in one cell, and quite another in a battery pack that goes into a vehicle where hundreds and thousands of these cells are packaged.

“There you have to manage, even thermally, how all of these need to be cooled,” Sinha noted. “Even before getting to the question about what issues need to be dealt with, how the pack needs to be cooled, there is a fight over volume. The vehicle guys are saying, ‘I can give you a 10-cm thick battery pack, which is going to fit into the bottom of the car.’ Whatever that thickness or volume is, the whole idea is to maximize the number of cells to maximize energy. You also have to account for structural reinforcement that you have to put in the battery pack for the safety worthiness. You may have to take out battery volume to put in the coolant channels that need to run through the battery pack. Everything you’re doing is volume that you cannot give to the cells. This is a very big packaging challenge, and it’s always there.”

To properly account for all of these aspects, tools needed include CAD, CFD, electrical, and structural simulation. All of that needs to happen together so that within the CAD environment, the tradeoffs can be made.

Charles Poon, global director of electrified systems engineering at Ford Motor Co., said during a recent presentation that Ford sees a lot of opportunities when electrifying vehicles, as well as challenges and pitfalls.

“When it comes to the technology, in terms of the electrode manufacturing process, how you are able to control every element of the manufacturing of the stacking machines?” he asked. “How are they welded together, whether it’s a cylindrical or a prismatic or a pouch cell? Each one of those steps is very critical in ensuring the quality output, and specifically the noise factors that could create a potential short within the cell itself. There have been a lot of learnings over the course of the 20-plus years that Ford has been working in the battery space, starting with its first hybrid electric vehicles, and now with full battery electric vehicles. It does take quite a bit of experience in some cases, and selecting the appropriate technology partner goes along with it as we learn together as part of that manufacturing and launch phase to pay attention to those details of the manufacturing.”

Battery management
On top of these thermal, packaging, and manufacturing issues, the batteries themselves need to be managed.

A battery management system is essentially the brain of the battery, and it’s absolutely critical because it has to measure behavior of every cell, from temperature to voltage, and whether the charge in every cell is balanced, among other considerations. To optimize batteries, the engineering team must determine the best algorithm, and then bring it within the embedded software framework.

“Once it is there, then of course you want to test it in a hardware-in-the-loop framework before it is ready to go,” Sinha said.

As part of this, there is a lot of movement towards overall battery management systems. Today, the common practice is a board sitting on top of the battery that manages it, but new techniques such as wireless battery management are coming into play.

“From the hardware perspective this is very different,” he explained. “Now, instead of an actual board that is hardwired, you are relying on the wireless aspects of it. But the actual software is still the same — the same algorithms and embedded software are used.”

Ford’s Poon agreed the development in software in detecting issues and the ability to mitigate those issues is ever growing. “Ford has made a significant investment in its in-house battery management software. One of the main objectives is to be able to monitor the health of that battery, all of the arrays, and each individual cell, and to be able to create ideal conditions to extend the life, extend all of the capabilities, manage all of the thermal issues associated with it, so that we can ensure safe operation through the life of the vehicle.”

Battery management concepts also play into possible reuse or recycling of the battery, particularly since some estimates say 70% of an EV battery’s life is left when the vehicle is retired. The question now is whether all lithium-ion batteries can and will be recycled.

Poon said Ford’s batteries are designed on the mechanical side, as well as the chemistry side, to be able to be recycled, and a new working laboratory has been set up to specifically address this point. In the future, the automaker has plans to make recyclability a significant aspect of the input raw material that will be utilized in its cell manufacturing process.

Gina Aquilano, technology director for Analog Devices’ Automotive Business Unit, said during a recent presentation that there’s no doubt about the massive investments in recycling efforts, noting the automotive ecosystem is trying to prepare for the tsunami of used batteries coming off the line within the next decade. “But given the advancements in battery technology, and battery management, there is going to be a lot of life left to give when it comes out of the EV. There’s a lot of dynamics at play in terms of the cost tradeoff analysis of whether to recycle directly, or re-use and then recycle. There are battery reuse processes that can deliver value to OEMs or battery owners and help recover some of the initial costs that goes into the battery production.”

Companies like Analog Devices have been paying close attention to this area since the electronics attached to the battery can aid in achieving the circularity goals, whichever path that ends up being.

Also, on the repair side, Aquilano noted that if there’s an issue with a battery in a pack, a decision must be made whether to replace the whole pack or just swap out a module. Wireless battery management can show details on what needs to be replaced, such as a section of a battery pack, a module within a battery pack, rather than the whole pack. “Monitoring batteries in a highly accurate way, collecting that information over the entire lifecycle, can help with these decisions and predicting a repair before it becomes downtime.”

Tim Grewe, general director for electrification strategy and cell engineering at General Motors, said this is happening today. “When you pull into a GM service center, we’ve demonstrated that wireless capability, which has hardware encryption so it’s cyber-secure. We can take that data, put it up into our cloud, and use it not only for the direct service, but since it’s got sort of a blockchain detail to it, we have this information up in the cloud. Based on the telematics, we know the mine that the lithium came from, we know the characteristics of that lithium. We know how it was used for its primary life, we know how it was used for its secondary life, and then we can feed that information to the recyclers. One of the key innovations happening in the recyclers now is they’re not breaking the battery down to the base metal sulfates, but new development allow them to go straight to cathode active material because of some of this information that’s coming out of these block chains.”

The future
As the automotive ecosystem invests more in electrification, there is also the realization that with all of the new developments and technologies, they not only need more people because of new programs and new vehicles, but they need new expertise, as well.

“This is about material-level expertise and electrochemistry expertise — and also embedded software expertise, which is there but it’s much more prevalent as we look into electric vehicles and batteries, no exception. Finally, battery technology especially is an area where, it’s not just about the problems of physics, bit different teams need to collaborate,” Siemens EDA’s Sinha noted.

From the battery team, to the body in white team, to the software, electronics, and battery management teams, he said, “If they don’t collaborate, the end result is really bad because everything is interdependent. There is a necessity to have an engagement, and all speak very different languages. And while we can’t make somebody an expert in all of these areas, through simulations, and the digital thread, we can connect the different groups so that they can collaborate as needed to get the right system.”

All of it comes back around to the connection with semiconductors, and associated materials.

“Everything we do now is based on that revolution around silicon,” Synopsys’ Smidstrup observed. “It’s one single element, and the understanding of one element’s properties drove the entire IT revolution. The same goes for batteries. But if one wants to move to the next generation, the number of combinations becomes vastly more. Back when they started making chips, it took 50 years to develop to get to where we are today. Today the competition is tougher, both from a climate point of view and from the technology race point of view, so whatever you can do to get ahead is needed.”

Managing complexity becomes critical. “When there are so many more elements involved, the complexity simply scales out of hand for you to do experimental work on all conditions,” Smidstrup said. “This is a place where simulations could accelerate the research into new materials, and especially since many of these materials like cobalt are very toxic, simulations can call away some possibly unsafe experiments. Simulations can also help define the boundaries at which we should explore more, and where should we not explore more. Sometimes simulations can reveal underlying physical limits that cannot be crossed to some criteria and that can also guide where to optimize in the future. You could indicate a theoretical maximum, for example, and when you’re close to that, that means you should search other places so it can accelerate but also guide the direction,” he said.