The grandaddy of annual battery meetings turned 38 this year and retained its virtual status for a second year— but will the Battery Seminar and Exhibit be back to Florida for 2022? The event, although online, had a wealth of presentations showcasing the state of the energy storage technology industry for consumer, automotive, military and industrial applications. BEST technical editor, Dr Mike McDonagh reports on some of the key developments.
This year’s International Battery Seminar and Exhibit was a virtual event with speakers on-line making their presentations. Whilst the protocols still seem a little strange, it is something to which most of us are beginning to acclimatise. This year was again dominated by lithium-ion, as the world relies increasingly on this technology for its electrification. It does, however, have its problems and limitations, and the great hope of EV rollout in the world still rests on improving this battery chemistry. For this reason, a significant number of papers were concerned with the use and manufacture of solid-state batteries (SSBs).
Solid electrolytes
With this development it is hoped that the main source of problems, the volatile and flammable liquid electrolyte, can be removed, and that a stable solid electrolyte (SE) can be substituted. There were also a couple of other papers on alternative chemistries, notably Na-ion and zinc. The zinc, however, was a no-show, with the presenter unable to attend the conference. For these reasons my summary is centred around the use of SEs in SSBs.
Solid electrolytes in lithium-ion batteries have been investigated over the past decade with variable success rates. The main reason is that there are significant advantages with a solid electrolyte over a volatile liquid component:
- Safer, non-volatile components
- Simpler cell design, a possibility of bi-polar
- Can use a lithium anode with greater energy density
Kelsey Hatzell from Vanderbilt University showed a slide that mapped the progress of SS lithium-ion batteries and projected to 2025 and beyond, (Fig 1).
This shows that the reduction in anode size has driven the improvements in energy density for lithium-ion chemistry. It also shows the order of magnitude change possible for the SSB version of lithium-ion. We are still in the position, however, where fundamental research is required to overcome some of the difficulties. These can be listed as:
- Poor contact surface area between electrodes and the SE
- High interface energy, creating a barrier to the transfer of lithium ions
- High cell internal resistance, due to immobility of ions in the electrolyte
- Irregular deposition of ions on both electrodes, creating dendrites
- Uneven distribution of component phases within the electrolyte structure
- Breaking up of structure, particularly with thicker designs
- Manufacturing difficulties
- High cost
The presenter goes on to note that despite the interest in solid-state lithium-ion there is still, surprisingly, a problem with lithium dendrite growth creating intra-cell short circuits. Fig 2 shows experimentally-observed failure modes in solid electrolytes and asks the question: how does it occur?
The mechanism for this is still not yet clear. Despite the lack of a concentration gradient and a solid electrolyte wall to pass through, dendrites still form. The failure mechanisms in lithium-ion SSBs are discussed, and the weight of evidence seems to point to a chemo-mechanical process as a cause of the lithium dendrite growth. Fig 3 shows how an irregular interface on an electro-deposited surface can, under pressure, produce areas under compression and alternative areas that are subject to tensile forces. This creates a stress gradient that, under certain conditions, provides a deposition or growth path for lithium dendrite growth. More conventionally, the effects of current concentration at sharp points where irregular surfaces of SE and anode interfaces meet were also a proposed mechanism.
Some solutions to this were proposed, including the use of a polymer— ceramic composite electrolyte. Fig 4 shows the result of the combination with varying molecular weights of the polymer component. The lower MW (and softer) polymer gave the best results, most probably due to the ability to accommodate surface irregularities. The paper then proposed using this expedient and limiting the lithium-ion availability to avoid concentration hot-spots as a potential route forward for the manufacture of safe, energy-dense lithium-ion SSBs.
As well as the materials, manufacturing methods are another starting point to address the surface interface problems. For inorganic SSEs, there are two main materials: a garnet oxide and a sulphide, that dominate the current material choices for lithium-ion solid electrolytes. Fig 5, from the paper by JL Rupp of MIT entitled: Solid State Battery Material and Processing Choices, shows the pros and cons for either of these SE material choices. Because of the high electrochemical stability range and intrinsic stability against lithium, the garnet oxide was used in processing trials to manufacture a low interface resistance SE for lithium-ion SSBs. The other point to note is the clear energy density benefit of using a lithium metal anode, facilitated by the use of a solid electrolyte.
The various methods for producing a lithiated SSB that were trialled are listed in Fig 6.
Of these, a co-sintering approach was adopted as providing the best solution for obtaining a low interface specific area resistance (SAR). In order to save cost, a low temperature sintering process was thought best. This limited the cathode choices and a new approach of using a cathode precursor material into a pre-manufactured porous SE was adopted (Fig 7). This resulted in producing a combined SE and cathode with an SAR of 65Ωcm-2, sintered at the relatively low temperature of 700°C. The area resistance is still much higher than a standard lithium-ion cell but is the lowest value yet obtained for a lithium-ion SSB. This paper also looked at the opportunities for manufacturing thin film SSEs.
Fig 8 shows the thickness ranges and manufacturing routes for the various SSE materials currently under consideration. There is a consensus that using a thin-film approach for the SE will significantly reduce many of the problems of interface resistance, built in structural defects and lack of heterogeneity.
The paper by Richard Clark of Morgan Advanced materials titled: Comparing Thick and Thin Film Solid State electrolytes, lists the most widely accepted SE materials and their difficulties (Fig 9). As can be seen, these categories of materials have their unique problems: for the sulphides it is the chemistry and cost; the polymers have low ionic conductivity and the ceramic oxides have serious interface problems. Morgan have taken the view that the least worst is the LLZO (lithium lanthanum, zirconium oxide) ceramic option, and recognised that the difficulties of manufacturing a thick layer is at the route of the problems.
Their solution is a hybrid thin film lithium battery (H-TFLB). A schematic of the structure is given in Fig 10. It consists of conventional current collectors: aluminium at the carbon-based anode and copper at the lithium metal cathode. The methods used to ensure good bonding between the lithium metal and the electrolyte plus the copper current collector is where the fun starts. Clearly shown in the diagram are the different layers, consisting of: TiN (to crate a flat surface, Mg3N2 which protects the current collector then a wetting medium, Li3Al, Li2Zn to ensure a good bond with the lithium metal anode. On the cathode side, there is a protective layer for the SSE (in this case LLZO) of 20–50mm of Li3N). The cathode is an NCM composite filled with liquid electrolyte.
This is clearly a complex construction, requiring elaborate procedures to obtain a good, commercially viable production process. This will not be a cheap solution, which was reflected in their slide comparing the H-TFLB costs with those of a Panasonic NCR 18650B cell Fig 11. At 12 times the cost, this route seems, at first glance at least, commercially dubious. The saving grace appears to be an enhanced cycle life of around 4,500 cycles. This changes the economics. The $/Wh-cycle costs show that the H-TFLB works out only slightly more expensive. The main advantage being that of convenience— in effect you save 10 battery changes and have a lower weight and volume burden. And it may be that this is the deciding factor.
There are clear energy density benefits to switching to a lithium metal construction. Fig 12 shows the density savings and the enhanced cycle life benefits of this technology. At the moment the applications are for very small cells but Morgan are aspiring to manufacture larger, high-power cells suitable for aircraft or drones.
Another approach to improving energy density was proposed by Wilheim Pfleging of the Karlsruhe Institute of Technology in the paper: Laser Structuring of Composite Electrodes for Next Generation Lithium-Ion Batteries. In this presentation, Pfleging proposes the 3D concept for thick films. As most technologists are aware, thick films increase the energy requirement for ion transfer. The thicker the film the higher the resistance and energy needed to pass through. This paper proposes a solution, which is to use a laser cutting method to put channels into thick film that allow ions to cross a shorter path to get to the electrolyte (Fig 13). This diagram also shows the effect of this on the cell energy density. There is a clear increase in both volumetric and specific energy density the thicker the cathode film thickness.
There are also benefits for fast charging (Fig 14).
Sodium-ion battery
Going outside of the lithium-ion multiverse there was one paper by professor Yong-Sheng Hu proposing a new and improved sodium-ion battery. The benefits of cost and material availability for sodium when compared with lithium are self-evident; the design and manufacture of a viable Na-ion battery are less obvious. With an object lesson in physics, professor Hu uses ionic potential (charge number/ionic radius) to predict the most suitable cathode materials to contain the sodium ion. He also proposes the use of the Cu2+/Cu3+ reversible redox couple as a high-voltage enhancement in the cathode matrix as an energy booster. He also described the use of modified anthracite (coal) as the starting material to manufacture the graphite of the anode. In this way an energy density of 400mAh/g for the anode was shown to be achievable. Laboratory tests have shown that the battery has a specific energy density of 135Wh/kg and a cycle of approaching 4,000 cycles to 82%. The company HiNa was founded in 2017 to manufacture and to demonstrate the performance of pouch and cylindrical Na-ion cells, Fig 15 shows some of the products and applications currently on test. The performance and safety of the product appear to be demonstrated, but the final costs were not mentioned.
The other point is, in the final cathode material, there are significant quantities of lithium (around 1/3 of the mole fraction of lithium Fig 16), which does seem counter-productive, considering one of the stated aims at the start of the paper.
As a summary, it looks like the SSB lithium-ion is still at the development stage. So too, are fast charging and the substitution of the standard transition metals for more environmentally-friendly cathode materials. Energy densities are creeping up and manufacturing methods are progressing. However, we still do not have any technological breakthroughs of sufficient significance to greatly influence the uptake of electric vehicles.
