Revealing the multilevel thermal safety of lithium batteries

Abstract

The ever-growing “endurance mileage” anxiety has been stimulating the continuous energy density raising of Li-ion batteries (LIBs) and the burgeoning of battery chemistries “beyond Li-ion”. However, if operated under abuse conditions, LIBs are easy to get thermal runaway. Encouragingly, great efforts have been devoting for achieving safer LIBs. However, the pace of thermal safety assessment has obviously lagged behind the energy density improvement of LIBs. Therefore, it is urgently needed to systematically, comprehensively and timely study the thermal safety issue of LIBs by combining different testing methods. In this review, we briefly summarize the varied testing methods for thermal safety evaluation of LIBs. Wherein, the highly-integrated accelerating rate calorimetry (ARC, an adiabatic calorimeter) technology is widely used to study the “worst case” thermal safety of LIBs at multilevel, ranging from ranging from battery materials to single cells and even battery packs. Combined with examples, the realizable specific functions of ARC in evaluating the thermal safety of LIBs are systematically summarized, and it reveals that the thermal safety of LIBs are rather complicated. Finally, critical perspectives in studying and improving thermal safety of LIBs and battery chemistries “beyond Li-ion” are provided.

Introduction

Nowadays, rechargeable Li-ion batteries (LIBs) have been widely used in varied fields, ranging from small consumer electronics to large-scale electric vehicles (xEVs) and renewable energy storage systems (ESSs) [[1], [2], [3]]. The ever-growing “endurance mileage” anxiety has been stimulating the continuous energy density raising of conventional LIBs [[4], [5], [6], [7]], and the burgeoning of battery chemistries “beyond Li-ion” (such as lithium metal based batteries, sodium based batteries, multivalent secondary batteries (such as magnesium, calcium, aluminum, and zinc based batteries), dual-ion batteries, capacitors, etc.) [[8], [9], [10], [11]]. However, if operated under mechanical, electrical, and thermal abuse conditions, LIBs are easy to get thermal runaway (smoke, fire, and even explosion), threating human life and property [12,13]. Every year, there are many serious accidents worldwide reported to be associated with smokes, fires and explosions of LIBs. Therefore, safety issue is a prerequisite for the practical application of LIBs. Encouragingly, at multilevel of battery material, single cell and pack, great efforts have been devoting to understand and improve the safety of LIBs [[12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23]]. However, the pace of thermal safety assessment has obviously lagged behind the energy density improvement pace of LIBs, and is always biased to one point of view. Adhering to the concept of “reaching every aspect of a matter”, what we need to do is to study the LIB safety systematically, comprehensively and timely by combining different testing methods.

Presently, researchers have developed varied testing methods to study the thermal safety of LIBs (Fig. 1a), such as accelerating rate calorimetry (ARC) [[24], [25], [26], [27]], vent sizing package 2 (VSP2) adiabatic calorimetry [[28], [29], [30], [31]], isothermal microcalorimetry (IMC) [[32], [33], [34], [35], [36], [37], [38], [39]], differential scanning calorimetry (DSC) [[40], [41], [42], [43], [44]], C80 micro-calorimeter [[45], [46], [47], [48], [49], [50]], fire propagation apparatus (FPA, also called Tewarson calorimeter) [[51], [52], [53], [54]], and in-situ high-energy X-ray diffraction technique (HEXRD) [[55], [56], [57]], etc. In general, IMC, DSC and C80 present high accuracy on detection of both exothermic and endothermic reactions, while ARC and VSP2 only detect exothermic reactions. Specifically, at a constant temperature, combining with an electrochemical cycler, IMC was used to study the heat generation (including reversible heat generation and irreversible heat generation) of LIBs during charge-discharge processes. At battery material level, DSC is usually employed to reveal the thermal stability and compatibility of electrolyte and/or electrodes. Compared to conventional DSC, a Calvet calorimeter of C80 is also used to study thermal safety at battery material level, but possesses higher accuracy, better tightness and bigger vessel volume. VSP2 is a commercially available adiabatic calorimeter, which is designed for determination of pressure related parameters at battery cell level (especially 18,650-type LIBs). FPA belongs to a fire testing equipment and it focuses more on the combustion behavior (such as fire/heat releasing rate, toxic gas releasing rate and battery mass loss). HEXRD often is adopted as an alternative to DSC for deciphering the internal chemical reactions between the electrolyte and electrode during thermal ramping. Obviously, DSC, C80 and HEXRD are only suitable for thermal safety study at battery material level, while IMC, VSP2 and FPA only serve for thermal safety study at battery cell level. Of course, it is difficult, costly and not necessary to let one laboratory have all aforementioned equipment. Except collaboration between laboratories, what we expect is the versatility of a single test method. Encouragingly, due to its versatile testing modes, ARC is considered as the most powerful technology to evaluate thermal safety of batteries at multilevel, ranging from battery materials to single cells and even battery packs, also ranging from normal battery charge/discharge conditions to complicated battery thermal runaway under abuse conditions.

As an adiabatic calorimeter, ARC (Fig. 1b) is a pivotal integrated technology to study the “worst case” thermal safety of LIBs at multilevel, ranging from battery materials to varisized single cells and even battery packs. ARC is initially developed by Dow Chemical, then firstly commercialized by Columbia Scientific Industries, and presently manufactured mainly by three companies of HEL (Fig. 1b), THT and Netzsch. Using top, side and bottom heaters, ARC simulates an accurate adiabatic condition by keeping the cavity temperature consistent with the sample temperature, preventing the self-generated heat loss of sample (inset in Fig. 1b). The realizable specific functions of ARC (Fig. 1c) in evaluating thermal safety of LIBs are preliminarily summarized as: (Ⅰ) thermal stability evaluation of electrode, electrolyte, and electrode/electrolyte, deciphering the key role of electrode engineering (material type, morphology, coating, doping, binder, etc.), electrolyte formulation (lithium salt, solvent, functional additive), etc.; (Ⅱ) thermal runaway features of any (any available type, any available size, and any available capacity) LIB under any state of charge (SOC) and any state of health (SOH); (Ⅲ) thermal runaway features of any LIB under abuse conditions, such as mechanical abuse (nail penetration, crush), electrical abuse (short circuit, overcharge, overdischarge) and thermal abuse (high temperature storage, rapid thermal shock, subzero temperature usage); (Ⅳ) specific heat capacity (Cp) and heat (including reversible and irreversible heat) generation determination of any LIB under adiabatic conditions. The focused parameters or testing items of (Ⅰ) (Ⅱ) (Ⅲ) include self-heating onset temperature, self-heating time before thermal runaway onset temperature, thermal runaway onset temperature, maximum temperature, self-heating rate (SHR, dT/dt), pressure (also pressure growth rate (PGR, dP/dt)), temperature distribution, toxic gas collection and online visual viewing. Etc. Cp and heat determination in (Ⅳ) are needed in the thermal simulating process of single cell and battery pack, serving the rational design of battery management system (BMS).

For thermal safety testing items of (Ⅰ) and (Ⅱ) in ARC, typical Heat-Wait-Search (HWS) mode is usually adopted (Fig. 1d) [15]. During the heating stage, the temperature of the whole cavity will increase by 5 or 10 ​°C, followed by the searching mode to identify whether exothermic reactions will happen when waiting. If the heat generation rate (dT/dt) of testing sample (battery materials or battery single cell) is smaller than the set detection limit, then next heating step starts (the stage below T₁ temperature in Fig. 1d). Once the exothermic reactions cause a self-heating rate (dT/dt) larger than the set detection limit, the heaters (top, side and bottom) of ARC will work to heat cavity according to the sample self-heating rate (dT/dt), eliminating any heat dissipation into the surrounding and providing a quasi-adiabatic condition until the sample finally gets thermal runaway (the stage between T₁ and T₂ temperature in Fig. 1d). In the representative curve of ARC using typical Heat-Wait-Search (HWS) mode, three critical temperatures are chosen to clearly understand the thermal runaway behavior of LIBs: T₁ denotes the self-heating onset temperature, reflecting the loss of overall thermal stability; T₂ is the triggering temperature point, where the sample get severe thermal runaway accompanying with rapid temperature increase (self-heating rate (dT/dt) will increase by several orders of magnitude); T₃ represents the maximum temperature, contributing to determination of the total heat generation during thermal runaway.

In this review, we will systematically and comprehensively review the thermal safety research progress of LIBs by using the highly-integrated ARC technology. This review aims to reveal the critical role of ARC in evaluating the thermal safety of conventional LIBs and battery chemistries “beyond Li-ion”. More importantly, this review will provide meaningful perspectives to study and improve the thermal safety of batteries.