Lithium-Sulfur Battery Looks Promising For Electrified Vehicles
Jeff Cobb July 1, 2013
Could it be that the laboratory that once ushered in the atomic age has developed the battery chemistry that will enable affordable electric cars with 300-400 mile range?
This is one implication for Oak Ridge National Laboratory‘s solid state nanotechnology based lithium-sulfur chemistry developed between 2007-2013. As we reported last month, ORNL announced it as a patent-pending scientific success that’s theoretically safer and cheaper than lithium-ion.
At this point it’s up to a competent engineering company to license the “beyond lithium-ion” chemistry from Oak Ridge which started life in 1942 as a home to the Manhattan Project, is now the largest science and energy lab in the U.S. Department of Energy (DoE) system, and appears destined to become a national park as well.
ORNL’s mission is far more benign today, but its recent invention has had several interested parties “knocking on the door” including “at least two” in the automotive business, according to ORNL’s David L. Sims, technology commercialization manager.
A $30,000 electric car later this decade or early next that could outdistance today’s $90,000-plus 85-kwh Tesla Model S would be a significant milestone surpassing more modest electric cars which today may go only 80-100 miles on a charge.
Tesla’s galleries are already capturing peoples’ imaginations. The start-up promises much, and roves ahead for future technologies while pushing what can be done now.
Perhaps the biggest EV news on a more-predictable horizon is a $35,000 Tesla with 200-mile range said to be pending for 2016. This however is to use a bulky Panasonic Li-ion-based pack with one-quarter the energy density of what ORNL says is ready to go commercial.
As things stand, four-five times today’s EV range from a more elegant Li-S battery could happen within seven years, according to Altairnano engineer and senior director of IP & Technology, Jay Akhave.
Of course there are no guarantees, and there are hurdles to overcome, which Akhave is just as quick to observe.
The ‘Science Problem’ Is Now ‘Solved’
After an hour-long interview last week however, we summarized the discussion saying, “This may be one of the hottest contenders to change the paradigm. If everything goes well, it could lead to a Nissan Leaf that goes 300 miles on a charge.”
“Right. I concur with that,” said Akhave who was recommended as someone knowledgeable by the head of ORNL’s project. “Lithium-sulfur is a class I think that has the potential and some good battery designers and good practical designers can make that kind of difference in the range.”
Akhave is qualified to postulate this as he’s essentially a talent scout for intellectual property and technologies for which Altairnano may want to become involved.
“I study their entire portfolios and look at the pluses and the minuses and chart out a road map for us,” said Akhave. “Everybody is trying to get into that high energy storage business, you know, where you have capacity.”
Akhave said only “no comment” when asked whether Altainano was one of the companies intending to negotiate a license with ORNL.
Four-times greater energy density also stands to improve range-extended vehicles. It could mean a Volt with more all-electric range than today’s Leaf while still having combustion-powered backup.
In any case, ORNL is looking to give preference to a U.S. business to maximize return on taxpayer investment, according to Jennifer Caldwell, group leader, Technology Licensing.
“We want to deploy the technology as soon as possible and we do not want the technology to be shelved,” she said of intellectual property under control of UT-Battelle which operates ORNL for the DoE.
Sims said the first deal among multiple licensees working in various sectors – from automotive to consumer electronics to grid storage and more – could come later this year.
“So on this particular license, with the interest that we have so far I would hope – and again, stressing the word hope – I would hope that we could have a deal completed in two to three months,” Sims said, adding money has not yet been discussed.
Sims is confident because Dr. Chengdu Liang, the head ORNL project, is satisfied with the solid-state Li-S chemistry.
“We don’t see any problem where this battery can’t be used in all kinds of applications,” said Liang last week. “It really depends on the design of the battery, not the science. Now we’ve solved the science problem.”
The First Of More To Come?
There are other companies working on lithium-sulfur which has been challenging researchers for decades, said Akhave. These include DoE sponsored Polyplus and Sion – which has already seen a Li-S battery used in a record-setting solar aircraft –– and the UK’s Oxis.
It’s also nearly certain BMW and Toyota and perhaps others are quietly chipping away at lithium-sulfur’s technical hurdles as is the DoE-sponsored Joint Center for Energy Storage Research (JCESR) project at the Argonne National Lab.
JCESR is a collaborative between top research hubs around the country. It is actually in competition with ORNL, so you see, the Energy Department has bet taxpayer dollars on multiple contenders.
A total of $120 million was allocated for JCESR to develop a battery with five times the energy density of lithium-ion in five years and capable of 500 charge cycles.
“Out of the $120 million, most of the money has gone for lithium-sulfur and lithium-air,” said Akhave of another promising contender. “Little has gone for lithium-ion. Why? Because I think they consider lithium-ion a commercialized technology, the basic research is done, mostly. Research is focused on Beyond Lithium-Ion.”
Lithium-air has potentially 10 times the capacity of lithium-ion, but lithium-sulfur is perceived as having a shorter time to market.
“The pressure to do it, and the money being put into research has already been deployed in other companies,” said Akhave. “They have their ideas on how this is going to happen. Dr. Liang has come about with his solution – an inventive solution on top of that.”
And so far in this technological horse race, ORNL’s solid-state lithium-sulfur is ahead by well more than a nose.
Sidebar: Tech 101
(This is in simplified terms, but non-techies can skip it, if desired, for a quicker read.)
Lithium-ion ordinarily delivers 100-200 milliamp-hours (mAh) per gram, up to a theoretical limit of 300-320 mAh/g. A Chevy Volt might have 140 mAh/g and a Tesla Model S may be pushing a bit more.
Lithium-sulfur at ORNL has demonstrated 1,200 mAh/g for up to 300 cycles.
“I believe that the five year goal, 500 cycle goal that the DoE has is a good one and quite do-able,” said Akhave.
Dr. Liang has thus far reported 300 charge cycles with his solid-state chemistry, but says he is satisfied with this.
“I see 300 cycles as enough to prove the concept,” said Liang. More critical, he added, is his chemistry does not self discharge when left off of a charger on a shelf.
“Self discharge is almost completely eliminated,” Liang said.
Akhave said the DoE’s goal of 500 charge cycles means “you are in business,” agrees 300 is significant, but left some ambiguity open on this topic.
Essentially a charge cycle is defined as completely draining the battery to perhaps 20-percent charge. At this point the battery management system (BMS) shuts down the party, and the battery must be recharged.
Dr. Liang holds up ORNL’s lithium-sulfur coin cell.
If, for example, an EV has 80-miles range and you only travel 45, then plug it back in, that does not count toward the “charge cycle” count. Likewise, if you have 350-miles range, drive 100 miles and plug back in, that ought not to count either. Partial discharge may have some effect, but does not normally count toward the total.
These questions become critical when determining lifetime for an electric car. If it could be used just 300, 500 or even 1,000 times, we’d still have disposable cars only lasting a few years.
The prospect of “only” 300 charge cycles being good enough to commercialize is arguably validated given that with four or more times the energy density, the car could be realistically used for a normal lifespan. Most people will plug the car back in before depleting the battery. Plus, it appears there is room to exceed the 300 charge cycles thus far established. It’s still early, and as Liang said, the battery does not self-discharge, which he sees as most important along with benchmarks established to date.
Liang is understandably hesitant to speak beyond the basic science however. He also qualified that “four-times” energy density is “gravimetric” – that is, by weight, and not “volumetric” – by volume.
In other words, a solid-state Li-S pack in the floor of, for example, a Nissan Leaf, would look a lot different than Li-ion. ORNL’s energy density is not by volume, and Liang expressed uncertainty whether this would equate to four-times the range.
Akhave cleared this up, however.
Of a hypothetical electric car, we asked: “Is it really going to have four times the energy density and therefore four times the ell-electric range?”
“It will go up, but the factor depends on a lot of other design elements,” said Akhave. “Gravimetric energy density is just per kilogram … volumetric is per cubic foot or per liter,” he said, noting “gravimetric” is also called specific energy.
“Volumetric energy density is projected to equal and go beyond lithium-ion,” he said, so if the same size (volume) of battery were used, it would have the same power energy, but Li-S would weigh less. Akhave said it thus appears likely that a Li-S battery could be designed that delivers four-times the energy density by weight.
What’s more, it’s believed engineers will be able to pack Li-S more tightly into a given volume of space than a liquid electrolyte Li-ion pack assembly. The solid-state battery may not need liquid cooling from a thermal management system (TMS) and as much air space as would Li-ion, so this too will play into the final result.
A question then becomes: “Are you efficiently using every little volume in there and packing it with energy? That’s the issue so whether you compare gravimetric or volumetric they are correlated in that sense,” he said.
Chemistries have varying material densities so such a space (volume) choice depends upon the chemistry. Once battery volume is fixed (based on many other car design considerations), then the amount of storage is fixed for that chemistry. One generally does not place another chemistry in the same volume and look at tradeoffs. If you change the chemistry, you must go back to the earlier step and evaluate how much battery volume you want to now have for that new battery chemistry. At this point that for the same weight it’s safe to say Li-S is four-times better than Li-ion chemistry.
Another detail to be worked out is the need for quick-enough recharging.
Liang has demonstrated a 2C charge rate – with C rate being measured in units of 1/hr.
“If a battery charges its entire capacity in one hour, they call it a 1C charging. Same with discharging,” said Akhave. “If a battery charges in 30 minutes, then it is a 2C rate and if it does it in six minutes, you have a 10C rate.”
Liang said shorter recharging times for a full-scale EV pack could be achieved if one heats the Li-S battery to 100 deg C. This is counterintuitive to present Li-ion designs, but conductivity goes up with heat for solid state Li-S.
This heating could be accomplished by a built-in TMS to heat the battery. The battery also heats itself during its operation, so this too could be contemplated by engineers.
A Li-ion battery with liquid electrolyte cannot be allowed to get too hot. Present Li-ion electric car batteries are usually engineered with a liquid cooling TMS to prevent excessive vapor pressures, the electrolyte from evaporating, and in worst-case, fire.
In short, Liang’s solid-state Li-S chemistry works much differently.
“With this all-solid everything, the rules are going to change,” said Liang, adding there is no assembly line in the world that could yet build Li-S packs.
Presently, ORNL’s Li-S chemistry appears to be the most viable, but challenges remain and we asked its inventor if it will be part of an electric car one day?
“I think in theory it will be,” said Liang.
And Akhave does too, but said ORNL will want to take care with who it allows access to its new chemistry.
To successfully build an electric car battery pack will take “somebody who has been there done that and has sweated it out in terms of trying to make a practical battery work,” said Akhave noting a company experienced with Li-ion development is most qualified. “Someone who has just discovered the new lithium-sulfur solution at a chemical level will be hard pressed to convert that to a functioning battery in the industrial sector.”
Assuming the right people get the chemistry, the first thing they will need to do is research and build a coin-cell most likely, or possibly a 1-inch by 1-inch cell.
In testing, they will need to consistently demonstrate number of charge cycles – preferably over 500 – and desired operational temperature. An automobile needs a certain window of cycles and temperature.
The “skateboard” chassis design like this one for a Tesla Model S offers several advantages. The actual volume required for a Li-S battery pack is in question. This could be the best way to pack maximum energy into a Li-S design while giving automakers freedom to innovate suitable designs.
“That does take time. Honestly I’d be happy to see a lithium-sulfur cell performing 500 cycles with this kind of duty in three to five years,” said Akhave. “This requires solving several complex interlinked technical problems simultaneously. Once that performance is set at the research or pilot level, systems engineers don’t take a lot of time, but need to reconfirm performance at every stage of scale-up.”
Assuming their fundamental building block panned out, engineers would then build a “real world” battery. At this stage they’d test to re-confirm the performance observed at the pilot level.
Next, they’d scale up to module level, test and further confirm. These in turn would be taken by systems engineers and assembled serially or in what ever way they chose into usable battery packs for electric cars.
“At every level – cell, module, systems, you have to reestablish that and recalibrate and understand the performance,” Akhave said and this depends on “whatever design and improvements and niceties you are building in … and then you have to take the pack and put it into actual duty.”
Here’s where engineers may subject their Li-S pack to freezing Alaska or the baking Mojave. They may leave it unplugged, and otherwise test/abuse it until satisfied they have something acceptable for consumers.
We regularly see stories where the promise of “game changing” technology is so many years away, but unlike ORNL’s discovery, none have had the chemistry, anode, cathode and electrolyte all worked out and proven.
As always, nothing is certain until it happens, and this is actually only the first of other Li-S and Li-Air chemistries that are believed likely to come along.
The road ahead looks promising, but challenges remain.
Quadruple the energy density would dramatically improve all sorts of things that use batteries today including electrified bikes, trucks, aircraft, watercraft, laptops, smart phones, grid storage, not to mention applications for the military.
For electric cars, it could put them over a perception hump, leave far fewer would-be consumers sitting on the fence, and this possibility now appears closer than ever