Li-ion batteries and fire safety: Part 2
By Apoorv Shaligram
In the previous part, we looked at the potential causes of a lithium-ion battery (LIB) fire. In this part, we will go through the various measures that can be taken to prevent LIB fires.
As I mentioned previously, the various potential starters for LIB fires only provide the spark for the fire to begin. However, LIB fires sustain themselves because they are the source of both the fuel (electrolyte) and the oxidant (oxygen released by the cathodes). For the oxidant to be released, one of the two conditions needs to be satisfied.
- Excess de-lithiation of the layered metal-oxide type cathodes (LCO, NMC, NCA), which happens when LIBs are overcharged significantly.
- Exposure of these cathodes to high temperatures (>300°C) at high state-of-charge. Under these conditions, the transition metals (Co, Ni) go from +3 oxidation state to +2 oxidation state (which is energetically favorable at such temperatures) and releases oxygen.
A bit technical, but the point is that unless batteries are overcharged beyond reason, i.e., a complete failure of BMS + incorrect charger usage, for LIBs to have uncontrolled fire incidents, they need to go through a step that causes such high temperatures, irrespective of the starter involved. The only way the cells can see such temperatures (except of course, being kept in an oven or thrown into an existing fire) is if the energy stored in the cell is suddenly released. How would this happen? if there is a total internal short-circuit in the cells. Remember, that even a hard external short-circuit would cause the current to flow through a long-ish path, thereby releasing heat a bit slowly. There would be fuses that would break the path, wires that would melt down etc. Even an internal short-circuit via dendritic growth cannot remain for too long, as the tiny lithium spikes would melt eventually before the entire cell reaches such high temperatures. In almost all LIB fires, the sudden release of heat occurs because of a breakdown of the separator membrane in the cell. These membranes are typically made of polyethylene or polypropylene and shrink at temperatures between 90°C and 120°C. Thus, if the starter is capable of raising the temperature locally above this range, the separator will start to shrink, and thus cause the two electrodes to come directly in contact with each other, causing a hard internal short-circuit that suddenly releases heat and shrinks more and more separator around it causing a chain reaction.
Prevention:
Since we know that LIB fires are notoriously difficult to stop once started, our next best bet is to design LIBs in such a way that fires are prevented. So, what are the ways to prevent these LIB fires? They rely on cutting off the steps in the mechanism of fire propagation. Remember the fire triangle? Eliminate any one side, and fires are prevented. In practice, it is all about reducing the probabilities…
- Cut off the oxidant: Changing cathode chemistry from LCO/NMC/NCA to one which is not prone to releasing oxygen, e.g., LiFePO4 or other olivine cathodes
- Cut off the possibility of dendritic growth or over-charging: Switching to an anode chemistry that is electrochemically a bit further away from Li metal allows some room for overcharge protection. e.g., If graphite were to be completely replaced by Silicon, each cell would have room for overcharging by 0.2V before the damage starts. If it were to be replaced by LTO or other such anodes, the room for overcharging increases to 1.5V per cell in series. This also prevents/reduces the possibility of dendritic growth.
- Cut off the possibility of damage from external short-circuits: Two ways to do this. One is the use of fuses or wire-bonding in battery pack construction. The other is use of built-in PCBs or current interruption devices/fuses or PTCR elements in cell construction. These elements can break the circuit internally in case of short-circuits. The elements built-into the cells are more potent as they can prevent damage from individual cells that face short-circuits e.g., from loose wire strands or condensed water droplets getting between the cell cap (+ve terminal) and the crimped wall (-ve terminal).
- Prevent temperature rise: Addition of thermal mass in form of materials with high latent heat, e.g., phase change materials (PCMs) to prevent temperature from crossing certain levels. However, in case temperature does cross those levels and a fire does occur, these organic PCMs also become fuel for that fire.
- Cut off the source of fuel: One rather unique approach that has found a lot of popularity has been the use of solid-state electrolytes in place of organic liquid-state electrolytes. The logic being that if there is no fuel, there cannot be a fire. However, after over a decade since work started on SSEs, we are yet to see commercially proven solid-state batteries.
- Cut off the major source of heat: One approach that has found takers quite quickly has been the use of thermally stable (or at least more stable than previously) separator membranes that resist shrinkage till higher temperatures. This is achieved by a thin coating of ceramic nanoparticles on the membrane surfaces. When the polymer membrane starts to shrink, the ceramic particles hold their positions and thus, there is no major arial shrinkage. If the separator does not shrink or degrade, the complete internal short-circuit is prevented and thus the fire will not occur. Another approach to the same end has been use of highly thermally stable polymers such as cellulose (300°C) and para-aramids (500°C) to make the skeleton of separators. These offer even better resistance to complete internal short-circuits.
- Break internal electrical pathway: A different approach to quickly stop internal short-circuits arising out of penetration/crushing is the use of metallized polymer films as the current collector substrate. If there is penetration, these films have localized shrinkage due to heat from excess current flow, thereby breaking the current flow pathway.
Based on these general principles, methods can be developed to prevent LIB fires or at the very least, reduce their probability. Based on these same principles, we aim at eliminating fire safety issues associated with LIBs entirely with our battery cell technology that is under development. Looking forward to making a real difference to how batteries and electric vehicles are perceived! Until next time, adios…
P.S.: SSBs are projected to solve other problems such as improving life (as LIBs life is a result of liquid electrolyte degradation) and reducing weight (as metallic lithium anodes can now be used instead of graphite). However, increasing life was a good goal to have a decade ago, as opposed to now when life of liquid electrolyte LIBs has been increased manifold since. Additionally, while the electrolyte degradation mechanism for capacity fade might be removed, we do not yet know if there are any new degradation mechanisms that may be specific to SSEs. As for weight reduction, replacing liquid electrolytes with much denser ceramic SSEs leads to weight addition, while the weight of the anode is reduced. The much hyped protection against dendritic penetration also hasn’t seen conclusive proof yet. Thus everything considered, SSBs may not bring as much of a benefit as has been the hope for the last decade. Perhaps more on this later, in a separate blog…
