Li-S battery with novel solid-state electrolyte shows capacity approaching theoretical value and high Coulombic efficiency
30 January 2015
|
Voltage profiles of charge-discharge cycles of the solid-state Li-S battery. Current density of 0.05 C). The specific capacity is given per g of sulfur. Yamada et al. Click to enlarge. |
A team from Samsung R&D and the University of Rome “La Sapienza” have fabricated a novel all solid-state Li-S battery that exhibits a capacity (∼ 1600 mAhg−1) approaching the theoretical value and an initial charge-discharge Coulombic efficiency approaching 99% (the average in ten cycles was 98%). An open access paper on their work is published in the Journal of The Electrochemical Society.
In addition to these and its other favorable properties (ie.e, smooth stripping-deposition of lithium), the activation energy of the charge transfer process was 44.5 kJmol−1—much smaller than that of a corresponding liquid electrolyte Li-S cell. These results, the team concluded, “are convincing in demonstrating that the solid electrolyte is very effective in physically preventing polysulfide migration.”
Overcoming the polysulfide shuttle is a significant advantage since it is a major drawback for a typical liquid electrolyte based Li-S battery. Further work is in progress in our laboratories to elucidate the behavior of our battery, and also to improve its construction. Nevertheless, we believe that the data here reported, even if still at a preliminary stage, are of importance for the progress of the lithium-sulfur battery technology.
—Yamada et al.
While Li-sulfur batteries are of keen interest as a high energy density successor to current Li-ion technology, there are significant challenges to overcome, including the red-ox shuttle of polysulfides originating originates from the dissolution of the sulfur cathode material into the organic electrolyte and poor lithium cycle performance due to the consumption of lithium metal during the charge-discharge process.
Solid-state electrolytes, based on both inorganic and organic compounds, are valid alternatives to develop lithium batteries with high safety and long cycle life, as in fact practically demonstrated. … It is important to point out that these “solid-state batteries” have the favorable characteristic of avoiding lithium dendrite deposition and hence, of preventing short circuits, in cells using lithium metal as the anode active material. In addition, solid electrolytes are expected to be safer than common non-aqueous electrolyte media, because of their low or negligible vapor pressure.
… In a previous work, we have demonstrated a solid-state Li-S battery based on 0.8Li2S-0.2P2S5 electrolyte. However, even cycling under shallow depth of discharge (DOD), dendrite short circuits were indeed observed. In this work we have extended the investigation, by developing and testing a solid-state Li-S battery using a stoichiometric composition of 0.75Li2S-0.25P2S5, Li3PS4 as the electrolyte.
—Yamada et al.
The team prepared the sulfide solid electrolyte using a high energy ball milling method. A cathode composite was made from mixing sulfur powder and carbon nanofibers in the ratio of 3:1 (w/w). A lithium foil served as the anode.
In electrochemical testing, the Li/Li3PS4/S solid-state cell achieved the high specific capacity at both 25 ˚C and 80 ˚C. The team concluded that the high value of the coulombic efficiency was clear evidence that polysulfide shuttle was prevented by the solid electrolyte layer.
Another important remark is that the discharge plateau typically reported for Li-S batteries was not seen at 25 ˚C where a large discharge-charge polarization was also observed. The electrochemical reaction was significantly accelerated at 80 ˚C, where the cycling nearly evolved along the expected plateau. Although the reaction in the solid electrolyte Li-S cell is still unclear, its kinetics are expected to be much slower in the liquid electrolyte cell.
… Realistically, the risk of dendrite formation and of cathode and electrolyte degradation cannot be excluded and they might shorten the cycle life of the battery. Long cycle life and high coulombic efficiency have been reported for thin film lithium batteries where the electrolyte layers are generally prepared by a vapor deposition process and are quite dense. This suggests that a dense solid electrolyte is a key for making a high performance solid state battery.
—Yamada et al.
Resources
Takanobu Yamada, Seitaro Ito, Ryo Omoda, Taku Watanabe, Yuichi Aihara, Marco Agostini, Ulderico Ulissi, Jusef Hassoun, and Bruno Scrosati (2015) “All Solid-State Lithium–Sulfur Battery Using a Glass-Type P2S5–Li2S Electrolyte: Benefits on Anode Kinetics” J. Electrochem. Soc. 162(4): A646-A651; doi: 10.1149/2.0441504jes
I have always thought that low current solid state batteries with thick electrodes (minimizing non-energy containing components in the batteries- fewer current collectors etc.) would be a good way to lower costs of materials and potentially make these good grid or home storage technologies. C-rates are insignificant for home solar storage. For instance, suppose you have a 200 kWhr battery for home solar and wind storage. Once fully charged you would have (assuming 10kW per day usage) you have a battery that can supply 20 days of electricity. With 24 hours in a day, that is a C/480 rate. In cars we need C-rates of 1 or C/2. The solid state electrolytes may not be capable of these vehicle requirements, but they should have no problem with home solar storage. Also, thick electrodes are more capable at low C-rates too. It is largely diffusion limitations that make simply thickening the electrodes not a viable method of increasing energy density and lowering cost for automobiles, since they require high currents, but again home storage can be low current, which minimizes the diffusion limitations issue. Big changes to our energy systems are scary for those that are already making the money, but we should act in the best interest of the people, however we don't, because money can buy influence.
Posted by: Brotherkenny4 | 30 January 2015 at 06:54 AM
@Brotherkenney4: "...suppose you have a 200 kWhr battery for home solar ..."
A quick check of Wholesale Solar gives me a current technology battery cost alone of $50,000 - $80,000 for that much storage (not counting controller(s), etc). Unless newer technology is an order of magnitude cheaper, I don't see much of a market for such a large amount of storage.
I have grid-connected solar panels, and might some day consider 10-20 kWhr of storage to keep things running during the occasional blackout. 20 days of storage is not necessary or economically sensible for an individual household. Now for a neighborhood, perhaps...
Posted by: Nick Lyons | 30 January 2015 at 07:39 AM
"..smooth stripping-deposition of lithium.."
This allows a lithium anode, which increases energy density further without dendrite formation. Solid state eliminates the sulfide migration.
Posted by: SJC | 30 January 2015 at 09:38 AM
Why not combine this with the Kevlar-based nanocomposite dendrite-suppressing Li-ion battery separator (http://www.greencarcongress.com/2015/01/20150127-kevlar.html).
Posted by: Account Deleted | 30 January 2015 at 12:44 PM
@BrotherKenny and Nick
I think Brotherkenny has a good point. Yes, the cost of batteries is too high for that now, but if the capacity is 16x the current capacity for Li-FEPO4, then the price will be much lower per kWhr. I could see these batteries costing $5000 - $10000. Although I have no data for that other than other chemistry's that are cheaper per kWhr when hey have higher specific energy. I think we can take this one step further. While vehicles currently need higher C rates, that's because the packs are so small. If you make an 800 mile vehicle, you want something like 100kWh of pack. Again, when you get to that level, you don't need the high C rates anymore. So if the price comes down enough, this battery works for vehicles also.
Posted by: Guetenburg | 31 January 2015 at 08:44 AM
I've noticed that the comments on this article are pretty informed technically. I was wondering which companies / chemistry's you think have the best chance of making a splash in the next 2 years? So bringing the battery to market before 2017.
Posted by: Guetenburg | 31 January 2015 at 08:46 AM
@Guetenburg, your idea about bigger batteries not requiring high C rates is generally correct, but the numbers are off a bit.
For a very efficient vehicle like the BMW i3 which uses about 272 watt per mile, you would need a battery of 217 kWh to go 800 miles. The good news is, passenger cars almost never drive 800 miles without a break, unless the driver wears astronaut diapers. Even if you drive 8 hours straight at an average speed of 65 mph that's only 520 miles.
If you drive four hours at 65 mph that's 260 mi - almost exactly what a Tesla Model S is capable of today. A meal break at the mid-point of an 8 hour trip is probably a good trade-off for a fuel cost savings of ~ $35 on that trip. (Maybe there is a free lunch!)
(above model compares gas at $2.50/gal to electricity at $.012 per kWh. When gas goes back up to $4gal the fuel cost savings will be ~$66.)
You could see a 260 mile battery for $5,000 - $10,000 by 2017. That's about $75-$150 per kWh.
Posted by: electric-car-insider.com | 31 January 2015 at 09:43 AM
@eci:
Very funny.
It appears Tesla worshippers never drive in cold weather, never drive faster than 65 mph and certainly not in hilly areas, and universally suffer from incontinence.
As for $75kwh in 2017, you have a real talent for innumeracy, or simply fantasy - wonderful stuff!
Do you have a personal as well as a corporate identity, or do you just like to share the blame?
Posted by: Davemart | 31 January 2015 at 10:10 AM
Davemart, your fixation on my identity is a little creepy. Get over it old man.
As energy density goes up (e.g. doubles) the materials cost will decrease by half, assuming reasonably priced or even cheaper materials - like sulfur for example. That's true not only for the anode and cathode materials, but also for things like casings and monitoring/balancing electronics. Even the effective cost of production equipment will drop as the total kWh throughput doubles on the assembly line.
This is assuming that exotic materials or processes are not introduced, but much of the promising research we've read about over the past few years does not use expensive materials or processes.
Whether these costs fall on the low side of that estimate or the high side by 2017 is almost irrelevant. The cost curve is going the right direction. Within the next few years an all-electric drive train will leave very little room for ICE or H2 to be competitive on a TCO basis.
Posted by: electric-car-insider.com | 31 January 2015 at 11:56 AM
"UBS, in a report based around a discussion with Navigant Research, says the $230/kWh mark will be reached by the broader market within two to three years, and will likely fall to $100/kWh.
Navigant estimates the cost of materials going into a battery at the Tesla Gigafactory on a processed chemical basis (not the raw ore) is $69/kWh.
The cost of the battery is only ~10-20% higher than the bill of materials – suggesting a potential long-term competitive price for Lithium Ion batteries could approach ~$100 per kWh.
Navigant says that the broader market place will reach the levels Tesla is paying in the next two to three years."
http://reneweconomy.com.au/2014/battery-storage-costs-plunge-below100kwh-19365
Posted by: electric-car-insider.com | 31 January 2015 at 09:16 PM
I agree with ECI, $100 per kWh is where we're headed in the next 3-4 years, driven mainly by Mr Musk's big workshop. After that, costs will fall more slowly, but progress in battery size and mass (at the pack level) will continue with good pace.
Posted by: clett | 01 February 2015 at 12:54 AM
"Within the next few years an all-electric drive train will leave very little room for ICE or H2 to be competitive on a TCO basis."
When will battery get to cost $0.5 / kWh the equivalent of gasoline tank, or $15/kWh, the equivalent of a hydrogen fuel tank?
Posted by: Roger Pham | 01 February 2015 at 03:03 PM
fixing italic.
Posted by: Roger Pham | 01 February 2015 at 03:07 PM
With this type of battery, affordable extended range (500+ Km) BEVs become a possibility by 2020 or so?
Posted by: HarveyD | 01 February 2015 at 03:34 PM
Roger, I think you already realize that comparing a gasoline fuel tank, essentially a metal cannister, with a battery is simplistic to the point of irrelevance.
On a Total Cost of Ownership (TCO) basis, that metal can is going to cost a whole lot more to fill over
Posted by: electric-car-insider.com | 01 February 2015 at 07:13 PM
That gas tank will cost $14,000 to fill for 100,000 miles at 25 mpg and $3.50 gallon average gas price over 10 years.
The battery will cost $3,600 in fuel over 100k miles at a national average of $0.12 per kWh and 300W per mile.
It's unknown what the hydrogen tank will cost to fill over the next 10 years, but it's extraordinarily unlikely to cost less than gasoline. Current price to automakers who buy in bulk is $13.59 per kg, ~ equivalent to $6.80 gasoline in real world use.
Posted by: electric-car-insider.com | 01 February 2015 at 07:30 PM
It will be interesting to see if anyone builds a ~ 300 mile range "affordable" BEV in that timeframe Harvey but I think it unlikely. I don't think many automakers will be interested in stuffing that large of a battery in a low cost car. 200 mile BEVs will probably be selling very well by then, with battery supply being the chief constraint on sales.
Unless other automakers commit to charge rates competitive with Superchargers, long distance travel is just not going to be that practical in a non Tesla BEV.
PHEVS will be available in virtually every car configuration made according to already announced road maps. I believe that will be the predominant long distance solution for the next 15 years at least. 50-80 mile AER PHEV is a pretty comprehensive solution with a relatively small incremental cost for virtually unlimited range, perhaps $3k. Until you're getting ~ 150-200 miles at a cost under that, there's not a lot of incentive to go all-electric for the occasional long distance trip.
Posted by: electric-car-insider.com | 01 February 2015 at 08:14 PM
@eci,
Those who cares about fuel cost will buy a Prius, capable of 50 mpg, so fuel cost will be $7,000.
Those who charge BEV at home at 85% charging efficiency at electricity rate of $0.12/kWh will pay $0.14 per kWh received by the battery pack. At 100,000 miles, the electricity cost will rack up $4,200, for a cost differential of $2,800 per 100,000 miles. This is excluding much higher fast-charging fees on when travel out of town! If you travel out of town 10% of the mileage, at $0.36 per kWh fast charge rate, figure this in, also.
However, a BEV with a 60-kWh battery pack at $100/kWh will incur $6,000 cost disadvantage, vs. the $50 cost of the fuel tank of the Prius...This money won't be recoverable until the car will run past 200,000 miles...By then, the car will be worn out, in the interior, fabric, structures, suspension...etc...that it will usually be retired.
So, a Prius at $24,000 list price vs. a comparable BEV with a 60-kWh battery pack for 200-mi range, costing $30,000 due to the $6,000 cost premium of the battery pack over the cost of a 7-gallon fuel tank...
Will this be the end of ICE as you predicted?
I don't think so! Most people will rather pay $24k for the equivalent of the Prius instead of paying $30,000 for a comparable BEV that they won't be able to recoup this cost difference until the end of the life of the car...
If you invest $6,000 for the 15 years life of the car, with 7% annual return on investment, 3% inflation and 15% tax rate, how much return do you get out of that?
Answer: Investment return totals $14,278 after 15 years
Source: http://www.bankrate.com/calculators/retirement/roi-calculator.aspx
Posted by: Roger Pham | 01 February 2015 at 08:17 PM
Continued from above:
To be fair, those BEV owners can also invest the yearly savings on fuel cost, over those 15 years.
So, HEV fuel cost $14,000 over 15 years, or $933 per year.
BEV electricity cost assuming 10% fast charging at higher rates of $0.36 and at 0.8 charging efficiency ($0.36 / .8 = $0.45/kWh). (54,000 kWh x $0.14) + (6,000 kWh x $0.45) = $10,260 /15yrs = $684 .
Saving in fuel cost of BEV over HEV, $933 - $684= $249 per year. If this is invested yearly, for 15 years, at the above calculation, "Investment return totals $6,117 after 15 years."
So, $14,278 return from investing $6,000 principal for 15 years, vs. $6,117 return for yearly cumulative investment of $249 from fuel cost savings, the difference is: $8161 in favor of owners of HEV who had wisely investing the cost differential between a BEV and HEV. Both BEV and HEV buyers forked out $30,000 out of pocket initally.
Posted by: Roger Pham | 01 February 2015 at 09:18 PM
Continued from above:
Both vehicles BEV and HEV above are assumed to be driven 200,000 miles over 15 years' time.
What I have neglected to mention is battery degradation, such that after 10 years, the BEV's battery pack may have aged so much that it may not be useable, forcing another $6,000 cost for a battery replacement pack. So, the last 5 years of the BEV's life, the second battery pack may have to be pro-rated to cost $3,000 extra expense for BEV owner. Adding this cost disadvantage to the $8,161 cost disadvantage above for BEV owners would mean $11,161 cost disadvantage for BEV owners.
"The rumor of my death [ICE] have been greatly exaggerated!"
Posted by: Roger Pham | 01 February 2015 at 09:27 PM
What about the economics of PHEV like the 4-seat Volt, at $34,000 with questionable profitability, vs the 5-seat Prius at $24,000 at assured profitability, the Prius being best-selling models in California and Japan for several years? Well, you can the run the same calculation like above to see for yourself.
Even if the Volt's price differential of $10,000 vs the Prius will be reduced by half, which will require a major miracle, the cost differential of $5,000 will still favor heavily the Prius, especially the 20% of the time that the Volt will be driven on gasoline, with poorer efficiency than the Prius, 37 mpg vs 50 mpg!
Posted by: Roger Pham | 01 February 2015 at 09:39 PM
What about the economics of H2-fueled vehicles vs BEV?
We have to make several fact-based assumptions:
1) The cost of H2 will be ~$3.5 per kg, from RE. The factual basis can be based from googling ITM-power/H2 cost projection.
2) Continual improvement in the fuel efficiency of the Prius, including new engine technology with cooled exhaust valve area, and Toyota's new SiC inverter technology, etc., can boost efficiency from 50 mpg to 60 mpg.
3) When run on H2 fuel, the ICE can be expected to gain efficiency by 20%, from 38% of current Prius engine, to 45.5% efficiency. So, 60 mpg can be boosted to 72 mpg.
http://www.transportation.anl.gov/engines/efficiency_goals_h2_combust_engine.html
Now then, let's start with the hypothetical improved Prius capable of 72 mpg on H2 fuel. The H2 fuel tanks with total capacity of 4 kg, cost $1,500 to add to the cost of $24,000 of current Prius = $25,500.
The BEV will still cost $30,000 as in previous case.
Over 15 years and 200,000 miles, the H2-Prius will cost $9722 from 2777 kg of H2 fuel, or $648 per year in fuel cost.
The BEV will cost $684 per year in electricity cost, including 10% of mileage fast charged when driven out of town, as above.
Right of the bat, there is no saving in energy cost of BEV vs H2-HEV, but $35 increase in energy cost.
So, plug in $4,500 initial investment with $35 yearly additional investment from savings in fuel cost, will give "Investment return totals $11,568 after 15 years"
http://www.bankrate.com/calculators/retirement/roi-calculator.aspx
Posted by: Roger Pham | 01 February 2015 at 10:07 PM
@Roger
Real world case: Last summer I got a car for my daughter as a graduation present. The Prius had a $5K premium over the Leaf after factoring in tax incentives. I don't see anywhere where you factor in tax incentives. I opted for the Leaf.
Nissan pays the fees for charging network memberships as well as charges for power from those networks for the first couple of years.
Charges vary per network but $0.25 / KWHr of FC is more realistic....at least here in Northern California. You may want to adjust your calculations.
I believe that later this year we will see that the default choice for those who are fuel cost conscious will be the 2016 Volt which after tax incentives will essentially be the same cost as the 2016 Prius but will have much lower fuel costs. If the Prius wants to remain viable it will have to become a plug in and offer battery range approaching that of the Volt. Less than 30 mAER won't cut it.
Posted by: Gasbag | 01 February 2015 at 10:48 PM
Roger,
By many accounts the Tesla Model S (especially the SD) is a much more fun car to drive than a Prius. The BEV max Torque at 0 RPM is a killer feature.
Different strokes for different folks and I agree that ICE's will be around for the foreseeable future.
What is cool is that we're on the cusp of not NEEDING them so we can reduce CO2 emissions without killing the global economy.
Posted by: msevior | 02 February 2015 at 02:53 AM
Roger, always nice to see you weighing in, especially with a thoughtful argument. But as Gasbag points out, you've omitted the tax and rebate incentives given to those who choose to drive a zero emission vehicle.
In California, it's $10,000. In Georgia, it's $12,500, making that state one of the highest per-capita adopters of ZEVs in the country. If you lease you can get those incentives up front, so that you can drive $0 down and less than $200 per month for a very well appointed brand new car.
My fundamental premise is that the cost curve for BEVs and PHEVs is going down rapidly. There will be an inflection point. That inflection point is being acellerated by the low cost and favorable cost curve of solar, which savvy homeowners can use to create very low cost transportation fuel with very stable prices.
Gasoline and Hydrogen will not be able to compete with those advantages.
As much as you'd like to speculate about low cost hydrogen, nothing about its manufacture, distribution or dispensing is cheap or easy. The likelihood you'd be able to get a permit for a rooftop system that compresses to 12,000 PSI (required for dispensing) within the next ten years is virtually nil.
Posted by: electric-car-insider.com | 02 February 2015 at 07:46 AM