Could JCESR’s Li-Sulfur Battery Revolutionize EVs?

Just when automakers are hunkering down to mass produce electric cars with good-enough lithium-ion batteries, a federally funded “dream team” of researchers says in one year its lithium-sulfur prototype may have triple the energy density of the best available today.

Their U.S. Department of Energy sponsored hope is that what they are onto may usher in an era in the next decade of long-range $20,000 electric cars with no real downside compared to petroleum powered ones they will rapidly displace.

That’s the goal, anyway, with an eye toward projecting U.S. industry ahead in global leadership, alleviating energy insecurity, and environmental concerns. And unlike pipe dreams you may have heard from cash-strapped startups, this proposed battery is the brainchild of several years of intense research by a “Manhattan Project” for energy storage.

George Crabtree is an Argonne National Laboratory Distinguished Fellow. As JCESR director, he directs its strategy and goals and acts as liaison to executives of JCESR partner organizations.

JCESR Director George Crabtree is an Argonne National Laboratory Distinguished Fellow. He directs JCESR’s strategy, goals, and acts as liaison to executives of JCESR partner organizations.

Followers of electrified cars may recall when the U.S. Department of Energy in Nov. 2012 announced the Joint Center for Energy Storage Research (JCESR), overseen by Argonne National Laboratory. This collaboration of 10 universities, five national laboratories, and five industrial firms was mandated to accelerate “beyond lithium ion” batteries toward commercialization. With $120 million in funding, their goal was to cut by one fifth costs for a 2011 Nissan Leaf-spec battery, provide five-times the energy density, do it in five years, and JCESR’s Director George Crabtree says it will make good on the promise.

“What we will produce in December of 2017, what we’ve promised to produce and we expect to produce, is a proof of principle battery that I can hold in my hand – so it’s small,” said Crabtree of a prototype verified with JCESR’s techno-economic modeling, “but that performs in a way that validates everything you need to make a $100 kWh/kg lithium-sulfur battery at the pack level.”

JCESR’s transportation prototype will contain a lithium metal anode and a sulfur cathode, taking advantage of the system’s high theoretical capacity for energy storage and the low cost of sulfur.

In contrast to making do with laptop batteries that have both enabled and limited electric cars, JCESR was commissioned to purpose-make batteries with the urgency of “national security” behind its mission. Pictured: JCESR’s transportation-focused prototype will contain a lithium metal anode and a sulfur cathode, taking advantage of the system’s high theoretical capacity for energy storage and the low cost of sulfur. Work at Argonne National Lab which oversees JCESR has already developed technology making possible the Chevy Volt.

Does that mean it’s guaranteed? Of course not. Nothing is but death and taxes, and Crabtree said trying to predict the future can be tricky. A commercially viable battery may take five years, 10 years, or may never happen, but ruling things out at this stage is not on the agenda for JCESR which has achieved goals to date with more successes in sight.

Speaking of which, JCESR’s goal of $100 kWh/kg for a completed battery pack is one Tesla hopes to achieve a few years from now for its Li-ion packs. JCESR’s prototype thus stands to be cheaper than today’s best Li-ion batteries which may be a bit under $200 kWh/kg if not well over that.

Sulfur is among the cheapest and most abundant active battery materials. It’s much cheaper than cobalt, the most expensive part of a Li-ion battery. A byproduct of oil refining, its availability is dramatized by the image of piles of sulfur stored outdoors produced in treating heavy oil sands in Fort McMurray, Canada. Photo: GlobalForestWatch.ca

Sulfur is among the cheapest and most abundant active battery materials. It’s much cheaper than cobalt, the most expensive part of a Li-ion battery. A byproduct of oil refining, its availability is dramatized by piles of sulfur produced in treating heavy oil sands in Fort McMurray, Canada. Photo: GlobalForestWatch.ca

That’s impressive in itself, as it took 25 years for Li-ion to whittle costs by a factor of 10 and this is where Li-S may start with room to improve from there, but that’s just one benefit. JCESR’s work-in-progress lithium-sulfur prototype is also to be far more energy dense – thus lighter and more compact – and with much more upside development potential than lithium-ion.

In other words, the federal brain trust may deliver a prototype battery whose proverbial floor is above the ceiling of the best Li-ion batteries in cars like the Tesla Model S, Chevy Bolt, and, well, anyone.

What this could mean is lighter electric cars with greater range, lower costs, fewer packaging challenges in manufacturing, and more.

As for that “energy density,” Crabtree said an initial prototype pack will have at least 400 Wh/kg which is five times more than a 2011 Leaf’s 80 Wh/kg. This was the minimum goal set for JCESR’s five-year charter which could be renewed or expire at the end of next year, but he and researchers are hopeful for more.

What can we expect from Li-S? USABC's 2020 goals are an authoritative reference for pack versus cell gravimetric energy density. Cell-level goals are 50-percent larger than the pack level (“system”). This reflects expectations based on Li-ion. Using this criteria, JCESR’s pack level goal of 400 Wh/kg translates to 600 Wh/kg at the cell level. Theoretical gravimetric energy density of Li-ion is 625 Wh/kg, and commercial availability is 250 Wh/kg. The theoretical gravimetric energy density of Li-S is 2,567 Wh/kg. Assuming Li-S scales from theoretical to cell as Li-ion does, Crabtree says one might expect Li-S cells to be 1026 Wh/kg. Using USABC goals, this might translate to 684 Wh/kg at the pack level. "This suggests that our commitment of 400 Wh/kg at the pack level might be significantly exceeded by Li-S, an outcome we would welcome," he said.

What can we expect from Li-S? USABC’s 2020 goals are an authoritative reference for pack versus cell gravimetric energy density. Cell-level goals are 50-percent larger than the pack level (“system”). This reflects expectations based on Li-ion. Using this criteria, JCESR’s pack level goal of 400 Wh/kg translates to 600 Wh/kg at the cell level. Theoretical gravimetric energy density of Li-ion is 625 Wh/kg, and commercial availability is 250 Wh/kg. The theoretical gravimetric energy density of Li-S is 2,567 Wh/kg. Assuming Li-S scales from theoretical to cell as Li-ion does, Crabtree said one might expect Li-S cells to be 1,026 Wh/kg. Using USABC goals, this might translate to 684 Wh/kg at the pack level. “This suggests that our commitment of 400 Wh/kg at the pack level might be significantly exceeded by Li-S, an outcome we would welcome,” he said.

“We haven’t done it yet, but in December 2017, our goal is to demonstrate – again with a little battery I can hold in my hand – a proof of principle prototype that gets 400 Wh/kg at the pack level,” said Crabtree. “If Li-S scales from theoretical energy densities like Li-ion did, even more should be possible – up to 680 Wh/kg at the pack level or 1,000 Wh/kg at the cell level, that could be within range. But we want to learn to walk before we try to fly”

In fact, Crabtree is being conservative. Researchers have postulated 1,200 Wh/kg at the cell level is possible for a chemistry with more than double the theoretical peak Wh/kg potential, and Li-S has huge implications – if it comes to pass.

Tech Challenges

JCESR is actually at work on a slew of energy storage projects, including for grid storage and it’s left a trail of 52 invention disclosures, 27 patent applications, and written a small library’s worth of research documents.

For transportation purposes, it has narrowed down contenders to speed forward a “transformative” change in the kinds of cars and trucks people will eventually buy.

So, the Lithium-Sulfur transportation battery prototype project being headed up by Sandia National Lab with other JCESR researchers is what the team has decided has the best potential among more than three chemistries in consideration.

Energy density in a 2011 Leaf battery pack was ~ 80 kWh/kg, and in a Model S it is ~ 130 kWh/kg. Projected Li-S energy density ranges from ~ 200 Wh/kg to ~ 450 Wh/kg at the pack level. Thus, one might expect a factor of 2-3 more gravimetric energy density for Li-S at pack level than for Model S. The cost of Li-S at the target of $100/kWh is likely to be lower than Tesla’s cost for Li-ion batteries in the Model S, though their cost now and in the future is not definitively known.

Energy density in a 2011 Leaf battery pack was ~ 80 kWh/kg, and in a Model S it is ~ 130 kWh/kg. Projected Li-S energy density ranges from ~ 200 Wh/kg to ~ 450 Wh/kg at the pack level. Thus, one might expect a factor of 2-3 more gravimetric energy density for Li-S at pack level than for Model S. The cost of Li-S at the target of $100/kWh is likely to be lower than Tesla’s cost for Li-ion batteries in the Model S, though their cost now and in the future is not definitively known.

“On launching in December 2012, we were considering several technologies, including Li-S,” said Crabtree. “In January 2016, we selected our prototype targets for transportation, Li-S as our top priority and Mg intercalation as our second priority.”

Lithium-air is a third possibility, but of the Li-S battery, Crabtree says the researchers see light at the end of the tunnel, but must overcome hurdles yet barring the way. Though they have not gotten there yet, they have four potential pathways toward development, as Crabtree explained.

If Li-S does not pan out: JCESR's plan B for $100/KWh is a multivalent batteries. In this, JCESR's scientists replace singly charged lithium ions, which are used in the lithium-ion battery, with doubly or triple charged working ions. This could increase the battery energy storage capacity by a factor of two or three and is attractive for transportation applications.

If Li-S does not pan out: JCESR’s plan B for $100/KWh is a multivalent battery. In this, JCESR’s scientists replace singly charged lithium ions, which are used in the lithium-ion battery, with doubly or triple charged working ions. This could increase the battery energy storage capacity by a factor of two or three and is attractive for transportation applications.

“The four pathways are (a) binding the intermediate states of the Li-S reaction, the polysulfides of chemical formula Li2Sx where x can be, for example 2, 4 or 6 (among other values) at the cathode, (b) choosing an electrolyte that does not dissolve the polysulfides, so that they never migrate through the electrolyte to the Li metal anode, (c) protecting the Li metal anode with a blocking layer that does not conduct polysulfides, so that any polysulfides that may have dissolved in the electrolyte do not penetrate to the Li metal anode and react with it. The last pathway (c) has several options to block polysulfides from penetrating to the cathode: (1) a membrane of graphene oxide or Al2O3, or (2) a naturally formed layer from the reaction of Li metal with LiNO3 as an additive to the electrolyte. So far all of these are promising, and we hope and expect that more than one will prove successful (and we do not want to predict the future at this stage).”

In Simple Terms

If the above explanation went over your head, here’s the skinny: Crabtree says JCESR has strong reason to be sure one or more of its tech “pathways” to overcoming present shortcomings will pan out.

By Dec. 2017, JCESR wants its prototype to be capable of 100 charge cycles with “minimum fade,” and potentially much more room to grow.

Of course only 100 charge cycles is not enough for an electric car, and Crabtree knows this, but it is a five-times leap above present commercially available Li-S batteries which may die after 20 charge cycles or up to 50 at the high end.

JCESR's benchmark: The 2011 Leaf. Of course costs have plummeted, and superior Li-ion batteries have come along, but Li-S has quadruple the energy density  potential. Environmental concerns from Sulfur are negligible. Thermal management would be engineered as needed for a finished end product. Volume and weight – and Achilles heel for EVS – could be halved or better.

JCESR’s benchmark: 2011 Leaf. Of course costs have plummeted, and superior Li-ion batteries have come along, but Li-S has many times the energy density potential and the initial prototype should have triple that of a Model S. Environmental concerns from sulfur are negligible. Thermal management would be engineered as needed for a finished end product. Volume and weight – an Achilles’ heel for EVs – could be slashed.

“You need hundreds of cycles, not one hundred,” he said of a viable electric car battery. “You would like to have hundreds, ideally a thousand. And I wouldn’t want to predict when one would achieve that, but if you can make progress from 20 to 100 – a factor of five – it seems reasonable that by further pursuing those pathway you’ll get it well beyond 100.”

The key, he said, is getting over an initial hump of limiting “polysulfide migration” that is the death knell of present-tech Li-S batteries.

“Once you have a technique that limits the polysulfide migration from the cathode to the anode, you can always make it better,” he said.

Assuming this is done, further refinement and an eye toward commercialization would follow a sequence of steps JCESR can foresee.

“We would then seek a manufacturing partner to make a prototype 10 times larger, to explore the challenges of scaling up,” said Crabtree assuming the prototype with 100 cycles next year is successful. “The next step would be a product prototype again larger by a factor of 10, to demonstrate a manufacturing process. This would be followed by actual commercial products, perhaps in several different formats and for several different uses (around town and long distance driving, for example). This process could reasonably take 5-10 years.”

About Lithium-Ion

Crabtree likens development of lithium-sulfur to the “circuitous” path lithium-ion took.

Lithium-ion batteries have their roots back to theories postulated before the 1960s. Believe it or not, it was Exxon corporation that commercialized its Li–TiS 2 battery in 1977 less than a decade after lab breakthroughs were made by others. Those Exxon batteries were fatally flawed however due to “dendrite formation” which killed them prematurely.

Collaboration is key. Li-ion technology has taken decades to mature in part because multiple competitors were independently following their own paths, and communications was slower or non-existent. By pooling the best and brightest, the Energy Department hoped to make for close communications and cooperation enabling research on steroids. Apparently the formula is working.

Collaboration is key. Li-ion technology has taken decades to mature in part because multiple competitors were independently following their own paths, and communications was slower or non-existent. By pooling the best and brightest, the Energy Department hoped to make for close communications and cooperation enabling research on steroids. Apparently the formula is working.

The first successful Li-ion batteries – and precursors to what we have today – were commercialized by Sony in 1991. These got around the problem of premature dendritic death using a carbon host structure. This contained lithium embedded in graphite at the anode instead of metallic lithium and Sony happily sold its first Li-ion camcorder that was lightweight enough to walk around with, and do the job.

Sony’s Li-ion battery was twice as energy dense as the best nickel-metal hydride and nickel cadmium batteries, and had less of a “memory” issue from partial charging and discharging.

From this minimum level, researchers have been tinkering with Li-ion chemistries as costs have come down, and previously unforeseen uses for the better batteries came as opportunities were presented.

Jump a decade-and-a-half ahead to around 2005, and the first smartphones began appearing which soon led to ubiquitous “personal devices” we’re all so familiar with today.

Along the way, in 2008, Tesla strung together more than 7,000 Li-ion laptop batteries to prove a desirable car could be so powered; in 2011 Nisan and Chevrolet introduced the Leaf and Volt, and you know the rest of the story.

Present Realities

This week the first sub-$40,000 EV with over 200-miles range – the Chevy Bolt – was delivered to customers, and next year Tesla’s Model 3 and Nissan’s generation-two Leaf are expected to follow.

First Chevrolet Bolt EVs Delivered

Automakers including Daimler, BMW, and VW Group are announcing their own competitive plug-in hybrids and all-electric cars, with projections that 15-25 percent of their global sales will be plug-in cars. On top of that, China is saber rattling for its now-number-one plug-in industry, Japan, Inc., may be coming to terms with plug-ins, but all are bound at this point to lithium-ion.

SEE ALSO: What’s Really Motivating Automakers To Build Electric Cars?

As things stand, despite periodic excitement for news of hopeful next-generation batteries, the reality is costs have come down faster for lithium-ion than anyone projected they would five years ago, and energy density has inched upwards.

And, with govenment regulations putting a proverbial gun to the head of carmakers now suddenly far more on board with the EV agenda, they are going with what they have, and for all anyone knows, Li-ion may be good enough for government work.

But not necessarily, says the government – or rather, the government-sponsored JCESR, and there are those who say it may be right.

The Model 3 shows demand for a high-value car, but folks wanting cars priced from the teens to low 20s have nothing so engaging.

The Model 3 shows demand for a high-value car, but folks wanting cars priced from the teens to low 20s have nothing so engaging.

Despite hoopla for Chevrolet’s “under $30,000” (after federal credit) Bolt, green car analyst Alan Baum does not project more than 23,000 first year sales, and comparable numbers in following years.

Tesla has over 400,000 reservations for the Model 3 which is unprecedented, shows outstanding potential, but even Tesla has said it wants a lower priced car for mass appeal.

SEE ALSO: Musk Looks Beyond Model 3 To More-Affordable Fourth Generation Car

While EV fans see these new price-for-performance benchmarks as major leaps forward, automakers have made it plain they are concerned about mandated cars that do not fully convince their buyers to forgo petroleum technology, and sales are the bottom line.

Today’s U.S. plug-in market share hovers near 1 percent of 17.4 million passenger vehicles annually sold. To move the dial beyond 5 percent to as much as 50 percent and more – as the federal government wants – cheaper electric cars in greater varieties and styles are required.

What about charge times? Li-S batteries are typically charged at rates from C/5 to C (C/5 corresponds to five hours and C to one hour for a full charge), or approximately in the same range as Li-ion batteries -   "One might expect similar performance in a future high energy density, long lifetime Li-S battery," said Crabtree.

What about charge times? Li-S batteries are typically charged at rates from C/5 to C (C/5 corresponds to five hours and C to one hour for a full charge), or approximately in the same range as Li-ion batteries – “One might expect similar performance in a future high energy density, long lifetime Li-S battery,” said Crabtree.

“You can achieve a lot with a price reduction, but I think to be transformational you need a $20,000 car that goes 200 miles,” said Crabtree observing at this stage a bit over 200 miles costs $40,000-plus out the door while cognizant of the history and ultimate potential of present technology. “And that’s probably out of reach with Li-ion batteries.”

For now a federal $7,500 federal tax credit helps, but whether this is extended a couple years from now for Tesla, General Motors and Nissan which will be first to reach a 200,000 unit ceiling, it always has been a temporary prop.

Meanwhile automakers are now making major commitments – but really because they have to and they also know advanced research, some of it being done by themselves – is underway.

While there is a back-and-forth debate on this, most have said if better batteries were available, it would be a good thing.