Experimental study on flame stability limits of lithium ion battery electrolyte solvents with organophosphorus compounds addition using a candle-like wick combustion system

Abstract

To evaluate the fire-retardant effectiveness of organophosphorus compounds (OPC) added to Li-ion battery electrolyte solvents, the limiting oxygen concentration (LOC) method is used in conjunction with a wick combustion system, called as wick-LOC method. With the wick-LOC method, two modes of stabilized flame are found, namely, wake flame and full flame. When OPC is added to the electrolyte, two distinct branches of extinction processes occur according to the different flame modes near extinction with no transition from the full flame to the wake flame in the case of higher OPC addition. The flame stability limits are measured as a function of OPC addition for both flame modes. The wake flame is shown to be consistently more stable at low levels of OPC addition. However, once the OPC addition exceeds a critical amount, the full flame shows higher stability with a lower LOC than the wake flame. These phenomena in the two regimes are also found in other cases of high OPC addition (different type of OPC and electrolyte solvent). In the most stable flame mode, the regime switches from the wake flame to the full flame with increasing OPC addition, and they are defined correspondingly as “blow-off regime” and “quenching regime”. To explain the presence of these two regimes, the thermal balance effect is considered in the discussion of flame extinction mechanisms. The difference in flame volume near the extinction limit shows that the quenching mechanism dominates flame extinction under higher OPC addition. The thermal balance effect on flame stabilization or extinction can be the additional impact on the fire retardation abilities of OPC itself.

Introduction

Lithium-ion batteries (LIB) have been the dominant type of rechargeable batteries used in electronic devices owing to their high energy density and portability. However, some serious fire and explosion accidents associated with LIBs have been reported in recent years, and fire safety issue has been emphasized in the LIB industry.

The organic electrolyte solvents used in commercial LIBs are mostly linear and cyclic alkyl carbonates and their combinations, and their volatility and flammability are the potential threat to the battery safety [1], [2], [3]. In case of battery failure with a high state of charge, the severe fire and explosion given by flammable electrolyte can be caused associated with the thermal runaway reaction [4], [5]. According to LIB designers and producers, adding fire-retardant additives into electrolytes is an effective means of improving the fire safety of electrolytes in LIBs. Organophosphorus compounds (OPCs) are among the promising candidates with high flame-retardant effectiveness and low environmental impact compared to halogenated fire retardants [6], [7], [8], [9]. While excessive OPC addition can considerably degrade the electrochemical performance of electrolyte and further impact on battery life and capacitance [2], [8], [10], [11], it is expected that a balance between fire safety and battery performance will be achieved.

To meet the above expectations, in [12], a candle-like wick combustion system was applied to quantify the flammability limit of electrolyte solvents. As it was optimized from the limiting oxygen index (LOI) method [13], [14], the wick burner was used to determine the limiting oxygen concentration of electrolyte solvents, called wick-LOC method. The fire-retardant effects of OPC additions were indicated by the change of LOC to sustain the wick flame. Using a similar principle to the LOI test for polymers and cup burner test for the gas-phase fire retardant effect [15], [16], the wick-LOC method was designed especially for mixed liquid fuels. A wick impregnated with the electrolyte solvent generates a candle-liked diffusion flame. The flame extinction limits (indicated by LOC) were obtained by decreasing the oxygen concentration under constant external flow velocity of N₂/O₂ atmosphere. The LOC results showed a significant fire-retardant effect with the addition of a small amount of OPC, while higher OPC addition (5–10 wt%) led to a marginal effect. This implies that a balance point can be found from the quantitative LOC results considering the positive (fire suppression) and negative (decreased battery performance) effects of OPC addition.

In a previous study [12], the LOC results were determined standardly under the same mode of stabilized flame before extinction, that is, the flame was stabilized in the wake region of the wick (wake flame). However, in addition to the wake flame, another mode in which a side-stabilized flame enveloped the entire wick (full flame) was observed during experiments. With increasing OPC addition (up to 10 wt%) into the electrolyte solvents, the flame could be extinguished directly from the full flame, and the corresponding LOC was lower than the LOC for extinction from the wake flame, which showed that full flame is more stable in the high OPC addition case.

The wake flame and the full flame modes were already discovered in the flame downward spread and extinction of solid material based on the LOI test method [17], [18], [19], and they have always been called wake-stabilized flame and side-stabilized flame for a candle-like solid slab/rod sample, respectively. By reducing the oxygen concentration of the external flow in the traditional LOI test, the flame extinction process has always been reported as follows. The base of the side-stabilized flame is pushed by external flow toward the downstream until the wake region with a shortening flame height; then, the flame proceeds to extinction. Extensive researches focusing on the change in extinction limits at different flow velocities, sample widths/thicknesses, and gravity [19], [20], [21], [22] conditions have been conducted, and in these researches, the wake flame was always more stable as a consequence of the side-stabilized flame (or full flame). This phenomenon is commonly ascribed to the higher residence time of the wake flame compared to the side-stabilized flame according to the Damköhler number, which is widely used to explain the blow-off mechanism of the side-stabilized flame [23], [24], [25], [26].

The flame in the wick-LOC method shows a configuration similar to that of downward flame propagation on a candle-like solid material during extinction. However, with increasing OPC addition, the flames are separated into two extinction processes given by the full and the wake flames near extinction. Moreover, the different extinction processes behave in a manner opposite to the common understanding mentioned in the previous paragraph, according to which the full flame becomes more stable with increasing OPC addition, and its LOC decreases to be lower than that of the wake flame (refer to Supplementary materials). To clarify the influence of OPC addition to electrolyte solvent on the wick flame extinction in the wick-LOC method, the flame stability limits of two modes of stabilized flame (full flame and wake flame) were determined depending on OPC addition in the present work. Then, the extinction mechanism of each flame mode considering electrolyte solvent added OPC was discussed.