Thermal Safety of 1-Ethyl-3-Methylimidazolium Bis(trifluoromethylsulfonyl)imide for Construction-Related Safety Process

Abstract

The surge in demand for sustainable materials has instigated significant research into versatile substances applicable in fields ranging from everyday commodities to construction and energy. Among these, ionic liquids, notably 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ([EMIM][Tf2N]), have risen to prominence as green solvents. However, an urgent demand exists to comprehend their thermal safety characteristics, particularly for energy applications. Contrary to previous research, which predominantly employed linear fitting or empirical formulas, our study presents a novel non-linear fitting approach to investigate the thermal behavior of [EMIM][Tf2N]. It yields new insights into its activation energy value, marking a significant advance in attaining precise thermal safety data for sustainable construction applications. To ensure safety at elevated temperatures, [EMIM][Tf2N] was selected for comprehensive analysis. Our research evaluated the kinetic model using thermogravimetric analysis coupled with assessing fundamental reaction parameters and simulating thermodynamic equations by identifying hazardous temperatures. This study revealed that the reactivity hazard of [EMIM][Tf2N] escalated considerably when the temperature surpassed 280 °C, emphasizing the importance of process safety. Furthermore, when the temperature exceeded 287 °C, the time to reach the maximum reaction rate (TMR) diminished to less than a day—an aspect crucial to process safety. At temperatures beyond 300 °C, around 70% of the substance was consumed, further underlining the need for stringent safety measures in processing environments. We also considered the impact of different storage containers on thermal safety. The potential runaway temperatures for box-shaped and cylindrical storage containers were established at 270 °C and 280 °C, respectively, providing valuable data for designing safe storage environments. Our research significantly contributes to the prudent utilization and sustainable application of ionic liquids like [EMIM][Tf2N] by considering various safety scenarios and establishing safe temperature ranges.

1. Introduction

Ionic liquids (ILs) are a unique class of organic salts known for their remarkable physicochemical properties, such as low volatility, high thermal stability, and excellent solvation capabilities. Their potential as alternative solvents and materials in various domains, including energy storage, electrochemistry, organic synthesis, materials science, and environmental remediation, has garnered much attention [1].

As we strive towards sustainable practices, ILs have shown considerable potential in the domain of renewable energy [2], due to their ability to maintain the long-term stability of energy materials under environmentally favorable conditions. However, despite their potential, there is a critical need to explore their safety parameters and thermodynamic properties, particularly for [EMIM][Tf2N] which, although promising, is comparatively less explored [3].

This study aims to explore the advanced thermal properties and hazard assessment of the widely used ionic liquid, [EMIM][Tf2N], employing both calorimetry and calculation techniques. Previous works [4,5,6,7,8,9] have made notable contributions to the understanding of this subject. The intricate relationship between reaction kinetics and thermodynamics has long been an area of scientific research. Particularly, the thermal behavior of [EMIM][Tf2N] has garnered attention, with various studies offering disparate insights into its thermal decomposition and behavior.

The unique physicochemical properties of ILs enable the development of advanced materials with specific characteristics. For instance, ILs have facilitated the creation of self-assembled structures and composite materials [10], providing enhanced mechanical, thermal, and environmental properties suitable for various building applications. The IL [EMIM][Tf2N] has shown potential in developing such assembly materials. Its unique properties have facilitated the creation of self-assembled structures such as supramolecular gels, which can serve as self-healing or shape-memory materials in building applications. The use of [EMIM][Tf2N] in synthesizing composite materials has also generated enhanced properties in these materials, making them suitable for a wide range of building applications, including structural components, insulation, and protective coatings [11,12].

However, while [EMIM][Tf2N] has potential across many applications, its potential hazards have not been comprehensively assessed. As we continue to explore its uses in building materials, it is essential to analyze the thermal hazards associated with [EMIM][Tf2N] to ensure its safe and effective use [13,14].

Thermodynamic models can provide critical safety assessments in chemical processes. For instance, critical parameters such as the time to maximum reaction rate (TMR), the time for a substance to be consumed by the reaction (TCL), and the runaway temperature can help characterize the transition into hazardous states caused by reaction-based thermal instability [15,16].

By incorporating basic analysis techniques such as thermogravimetric analysis (TG) as the foundational source of thermodynamic analysis and calculation, this study aims to bridge the knowledge gap regarding the safety and performance of ILs. This paper focuses specifically on the thermal properties of [EMIM][Tf2N] and aims to evaluate its thermal hazards using calorimetry and calculations [17,18,19,20] to supplement its safety parameters. This research incorporates the assessment results of an analytical model to evaluate the reaction kinetics, highlighting the role of kinetic parameters in identifying and interpreting the corresponding thermal hazards. By doing so, this study aims to contribute to the safe and effective application of ILs in various industries, including construction.

2. Experimental Methods

2.1. Sample

The experimental sample, 98% [EMIM][Tf2N], was produced by Beijing Hwrkchemical company, Beijing, China, and its chemical formula is shown in Figure 1. The nitrogen gas used in the instrument was purchased from Xi’an Weiguang Gas Company, Xi’an, China as a high-purity gas.

2.2. Thermogravimetric Analysis

The database of this study was derived from a thermogravimetric analysis (TGA) experiment, and the instrument was purchased from Mettler Toledo Company, OH, USA. The experimental data and analysis were derived by measuring and recording the weight change of substances that were consumed by the reaction. In the experiment, a total of 3.0 ± 0.1 mg sample was placed in a 70 μL alumina crucible, and the reaction was initiated by heating the furnace body of the TGA. The experiment continuously introduced 50 mL min⁻¹ of N₂. The heating rate was set to 300.0 °C at the incremental of 1.0, 2.0, 4.0, 8.0, and 10.0 °C min⁻¹ [21,22].

3. Reaction Mode and Heat Exchange Calculation

3.1. Integrating Activation Energy and Pre-Exponential Factor to Complete the Kinetic Triplet: Determining the Reaction Mechanism

The foundation of characterizing a reaction mode is firmly rooted in the kinetic triplet, which encompasses the activation energy (E_a), the pre-exponential factor (A), and the reaction model function (f(α)). These integral components exhibit interdependence, making their simultaneous determination a prerequisite for portraying a reaction system accurately [23]. Historically, empirical equations, predominantly linear regression, were the go-to methodologies for parameter estimation, especially for E_a. However, such approaches are often inadequate in determining A and f(α), with the latter being particularly elusive [24]. Prominent works by Zaitsau et al. [4] and Liu et al. [5] delved into its decomposition kinetics, elucidating parameters such as E_a and A. Notably, a comprehensive study by Yu et al. [6] combined density functional theory with experimental methods, presenting findings on E_a, A, and f(α); although potential f(α) mechanisms were assessed, a conclusive evaluation remained elusive. Building on this foundation, our investigation primarily focuses on the f(α) reaction mechanism of [EMIM][Tf2N], aiming to fill the knowledge gaps left by previous studies. As research advances, tools like adiabatic calorimetry and exhaustive product analysis will be pivotal in refining our understanding and ensuring the robustness of our findings.

The limitations of linear regression methodologies are evident in their inability to cater to the kinetic triplet’s interconnected nature. In several instances, even when reactions highlighted commendable linearity, the aptness of the chosen mechanism function was not guaranteed [25]. This ambiguity was further exacerbated by the fact that identical datasets could fit multiple mechanism functions, rendering qualitative evaluation of f(α) difficult. In light of these inherent shortcomings, the pivot to non-linear fitting methodologies was not preferred but necessary. These methods, as opposed to their linear counterparts, accommodate the intricate interrelations of the kinetic triplet, thus comprehensively addressing the complex problem of reaction mode analysis [26].

The distinct advantages of non-linear fitting are evident in their adaptability—aptly equipped to deal with numerous reaction types and conditions—which enables a more extensive investigation of chemical systems, providing insights into their intrinsic behavior. Especially in multi-step reactions, as indicated by, Gao, et al., 2019, each segment potentially possesses a unique activation energy, pre-exponential factor, and reaction model function [27]. Non-linear fitting, with its capability to estimate these parameters simultaneously for every step, offers an overarching understanding of the entire reaction process [28].

Furthermore, combining non-linear fitting techniques with modern experimental methodologies, like differential scanning calorimetry and TGA, is a powerful strategy. This combination not only delineates reaction kinetics and thermodynamics but also observes conversion rates, enthalpy shifts, and temperature contours. Such extensive data would refine the reaction model, increasing the precision of kinetic parameter approximations [29,30]. It is pivotal to note that formal reaction models perceive conversion degrees (α) as the system’s state variables:

$$\frac{\mathrm{d}\mathit{\alpha }}{\mathrm{d}\mathit{t}}=\mathit{f}(\mathit{\alpha })k_0{\mathit{e}}^{-\frac{E_{\alpha }}{\mathit{R}\mathit{T}}}$$

(1)

$$\mathit{f}\left(\mathit{\alpha }\right)=(1-\mathit{\alpha }{)}^{\mathit{n}}$$

(2)

The non-interfered reaction system (A→B→C) can be integrated with Equation (2) and restructured as Equation (3):

$$\frac{\mathrm{d}{\alpha }_1}{\mathrm{d}t}=k_1\exp (-\frac{E_{\alpha 1}}{RT})f_1(1-\alpha );\frac{\mathrm{d}{\alpha }_2}{\mathrm{d}t}=k_2\exp (-\frac{E_{\alpha 2}}{RT})f_2(\alpha -{\alpha }_2)$$

(3)

The interaction form: A→B+… or A + B→2B, and the f(α) can be represented as Equation (4):

$$\frac{\mathrm{d}\alpha }{\mathrm{d}t}=k_1\exp (-\frac{E_{\alpha 1}}{RT})(1-\alpha {)}^{n1}+k_2\exp (-\frac{E_{\alpha 2}}{RT}){\alpha }^{n2}(1-\alpha )$$

(4)

where E_a denotes the apparent activation energy; R stands for the gas constant; T is the temperature of the sample; α is the degree of conversion of the reactant; k₀ is the pre-exponential factor; and n is reaction order.

The optimization of the model compares the prediction of the simulated reaction equation with the experimental data to verify the model, which may involve adjusting the model parameters, such as rate constant and activation energy, to improve the consistency between the model prediction and the experimental observation.

3.2. Utilizing the Kinetic Triplet for Estimating Heat Transfer between External Environment and Material

The accumulation of a large volume of material can lead to ineffective cooling due to the heat storage effect, promoting self-accelerating decomposition. In real-world scenarios, both temperature and the quantity of matter play pivotal roles in influencing subsequent thermal hazards. This is particularly evident when inadequate ventilation or cooling leads to partial decomposition of the accumulated substances. The resulting decomposition heat can spur surrounding materials to further react, intensifying the scale of the reaction and leading to significant heat generation. Recognizing this, it becomes evident that temperature control measures are crucial. The time necessary to consume a specific quantity of reactants within a designated temperature range can be determined by considering the temperature-dependent reaction rates. To capture these intricate dynamics, a mathematical model has been crafted to delineate the relationship between α and various reaction conditions, including temperature, pressure, and concentration. This modeling approach involves creating a suite of models that detail how state variables evolve over time and in response to other process variables. Subsequently, integrating the known characteristics—like heat capacity, density, and molecular mass—of the involved species ensures the model genuinely mirrors the system’s behavior across an array of conditions [31,32]:

$$\frac{\mathit{d}\mathit{T}}{\mathit{d}\mathit{t}}=\frac{\lambda }{\rho {\mathit{C}}_{\mathit{p}}}\left(\frac{{\partial }^2\mathit{T}}{\partial {\mathit{x}}^2}+\frac{\mathit{g}}{\mathit{r}}\frac{\partial \mathit{T}}{\partial \mathit{x}}\right)+\frac{-\Delta H_{\mathrm{d}}}{C_{\mathrm{p}}}\frac{\mathrm{d}\alpha }{\mathrm{d}t}$$

(5)

where C_p is the sample heat capacity, λ is the thermal conductivity, ρ is the density of [EMIM][Tf2N], x is the package radius, and g is a geometry factor that varies by the type of packaging.

The initial temperature distribution of the material within the surroundings is evenly distributed with boundary conditions:

$$3\mathrm{rd}\mathrm{kind}:\mathit{\lambda }\frac{\partial \mathit{T}}{\partial \mathit{X}}{|}_{\mathrm{s}}=\mathit{U}\left({\mathit{T}}_{\mathrm{s}}-{\mathit{T}}_{\mathrm{e}}\right)$$

(6)

where X is the outer unit normal on the boundary, subscripts s and e denote the surface and environment, respectively, and U is the heat transfer coefficient.

Validated models were used to analyze the behavior of the system under various conditions, such as TMR_ad [33] in parameters. This can provide valuable insights into the factors affecting the reaction process and help identify potential hazards or areas of process optimization [21]. The decomposition rate is a crucial factor, as it drives the surrounding substances to expand the reaction scale continuously, resulting in significant heat generation. The importance of temperature control measures becomes apparent in such scenarios. The time required to consume a certain quantity of reactants within a specific temperature range can be calculated by considering the temperature-dependent reaction rate variation. The change in substance temperature with the reaction process can be evaluated using the thermal safety software of Thermal Safety Series (TSS) version 3.11 [34].

By examining exothermic reactions, reaction time parameters can be modeled through dependable synthesis and kinetic computations. Kossoy and Akhmetshin have previously estimated TMR_ad values by assessing the thermodynamic properties of reactions. The severity of potential thermal risks is influenced by the pace at which a material produces energy. Building on the work of Wang and colleagues [34], Kinetics Neo version 13 July 1998 of software for thermal analysis, such as those developed by Netzsch Group [17], was employed to evaluate the hazardous parameters associated with reaction time.

4. Results and Discussion

4.1. Evaluate Thermokinetic of [EMIM][Tf2N]

The kinetic triplet, which consists of the reaction mode, pre-exponential factor, and activation energy, plays a crucial role in understanding the intricacies of chemical reactions. By employing non-linear fitting techniques for the analysis of the reaction mode and pre-exponential factor and using the activation energy values determined through the isoconversional method, it is possible to obtain a more accurate representation of the reaction kinetics. This study investigated the reaction kinetic parameters of [EMIM][Tf2N], as shown in Figure 2 and Figure 3. The simulated values were derived by combining the results from prior experiments with the information presented in Section 3.2. The simulation curve was generated by adjusting the parameters within the equation until it closely matched the experimental curve. This close match indicates that the current parameter values, when substituted into the reaction equation, are a close approximation to the actual reaction model.

During our kinetic analysis, we employed various reaction models to best represent the kinetics of the [EMIM][Tf2N] reaction process. Each of these models has unique characteristics and offers different perspectives on the reaction progression. For instance, the nth-order model suggests a direct proportionality between the reaction rate and the concentration of reactants. On the other hand, there are models with a higher degree of complexity, considering factors like autocatalysis or intricate reaction sequences. After implementing these models and comparing their fits with experimental data, we found that the two-stage nth-order model yielded the best fit to our experimental curve. While other models have their distinct merits, none aligned with our data as closely as the two-stage nth-order model did. This suggests that the nth-order model adeptly captured the essential kinetics of the [EMIM][Tf2N] reaction process and provided a reasonable approximation of the reaction kinetics.

However, this finding does not negate the potential applicability and utility of other reaction models in different contexts or for different reactions. The best-fit model largely hinges on the specific characteristics of the reaction under discussion and the accuracy of the experimental data. Therefore, while the two-stage nth-order model fits best in this scenario, other reaction models might be more effective in different circumstances. We intend to further explore and employ various reaction models in future work to deepen our understanding of the kinetics of various chemical reactions.

Our simulation results provided insights into the weight loss rates and the corresponding degradation behavior observed in the TG experiment under different heating rates. A closer numerical analysis of Figure 3 revealed that the reaction of [EMIM][Tf2N] adhered to a two-stage nth-order reaction mechanism. The peak waveforms of both reaction stages were moderate, indicating that the dominant contributions to the overall reaction came from these two stages. Using values derived from the literature as a basis, we ranged the values of the initial simulations within specific intervals. It is worth noting that there were slight variations in the fit values across different heating rates, potentially stemming from errors in different experimental batches. The parameter values derived from this analysis, as presented in

Table 1. Parameter value results of non-linear fitting at different heating rates.

Stage 1
	1.0 °C min−1	2.0 °C min−1	4.0 °C min−1	8.0 °C min−1	10.0 °C min−1
ln (k0)/ln (s−1) (±0.01)	29.74	29.01	30.76	32.25	29.25
Ea/kJ mol−1 (±0.1)	167.9	169.6	183.1	223.7	202.6
n/Dimensionless (±0.01)	1.13	1.44	1.56	1.69	1.44
Stage 2
ln (k0)/ln (s−1) (±0.01)	9.56	8.98	10.43	10.87	10.92
Ea/kJ mol−1 (±0.1)	97.2	97.3	94.6	106.4	102.6
n/Dimensionless (±0.01)	1.03	1.33	1.70	1.30	1.74
The range of parameters in this study Stage 1 ln (k0)/ln (s−1): 29.7423–32.2564 Ea/kJ mol−1: 167.9841–223.7765 n/Dimensionless: 1.1345–1.6920 Stage 2 ln (k0): 8.9837–10.9268 Ea/kJ mol−1: 97.0284–106.4358 n/Dimensionless: 1.0358–1.7468			The range of parameters from Zaitsau et al. [4]: ln (k0)/ln (s−1): 14–16 (Isoconversional approach) Ea/kJ mol−1: 240 ± 15 (at α = 0.2) to 270 ± 15 (Isoconversional approach) n/Dimensionless: No correlation calculation The range of parameters from Liu et al. [5]: ln (k0)/ln (s−1): No correlation calculation Ea/kJ mol−1: 165.2 ± 6.6 and 163.9 ± 6.0 (FWO and Starink methods) 210.4 ± 5.9 (ASTM E698 method) n/Dimensionless: No correlation calculation The range of parameters from [6]: ln (k0)/ln (s−1): 15.4–49.6 (Coats-Redfern method for different f(α)) Ea/kJ mol−1: 25.7–211.4 (Coats-Redfern method for different f(α)) 142.4 (at α = 0.1) to 204.2 (at α = 0.8); (Friedman); 136.3 (at α = 0.1) to 198.7 (at α = 0.8) (KAS); 140.6 (at α = 0.1) to 199.9 (at α = 0.8) (FWO) n/Dimensionless: No correlation calculation

, serve as foundational information for understanding and evaluating the inherent reaction hazards associated with this process.

The application of non-linear fitting techniques to analyze the kinetic triplet’s reaction mode and pre-exponential factor allows for a more accurate representation of the reaction kinetics. Our results suggest that the thermal decomposition of [EMIM][Tf2N] may occur more readily than previously reported [4,5]. The lower activation energy derived from our studies implies that the reaction could start at lower temperatures, which underlines the need for stringent temperature regulation during the processing, handling, and storage of [EMIM][Tf2N]. This fact also emphasizes the necessity of effective temperature regulation systems, particularly in environments where the reaction is carried out or where the substance is stored or transported. The lower activation energy derived from our experiments should be considered when planning safety protocols for using [EMIM][Tf2N] in industrial processes. Further studies could focus on optimizing the safety measures to control potential risks associated with this ionic liquid. In the case of [EMIM][Tf2N], this approach has provided valuable insights into reaction patterns and identified the primary contributions to the overall reaction. These findings have important implications for understanding and managing the reactivity hazards associated with the actual process, ultimately contributing to the development of safer and more efficient chemical processes.

4.2. Evaluating Storage and Transport Hazard Parameters for [EMIM][Tf2N]

Developing a thorough reaction model is crucial for examining and assessing a substance’s heat-releasing behavior and identifying safety aspects linked to reaction speed and duration. In the case of [EMIM][Tf2N], a complete reaction model was devised to account for the thermal interaction between the container holding the substance and the surrounding environment. This model accommodates various storage scenarios, such as the substance stored in a 50 kg drum, and considers the boundary conditions detailed in

Table 2. Physical parameters of packages [32] and [EMIM][Tf2N] [35,36].

	Material	Size/cm	Shell Thickness /mm	Filling Height/cm	Density/g cm–3	Specific Heat Capacity /J g−1 K−1	Thermal Conductivity Coefficient /W m−1 K−1
50 kg Drum	[EMIM][Tf2N]	R × H 23 × 63	7.0	35	0.75	1.7	0.3
Physical parameters for [EMIM][Tf2N].
Surface temperature/°C: 25 (room temperature)
Density/g cm–3: 1.37
Specific heat capacity/J g−1 K−1: 1.53
Thermal conductivity coefficient/W m−1 K−1: 0.165

. This model allowed the analysis of different temperature scenarios for heat transfer in the external environment caused by temperature variances. This is vital for determining the likelihood of thermal runaway in the material and the associated temperature elevation pattern. The model facilitated the investigation of [EMIM][Tf2N]’s decomposition reaction under differing temperature conditions, highlighting the variations in reaction speed and how it influences the consumption of the substance.

The TCL is an essential parameter for evaluating reaction intensity and hazard levels. It denotes the time needed for a specific quantity of the substance to be consumed at a particular ambient temperature. As depicted in Figure 4, when the ambient temperature is under 265.0 °C, more than 70% of the substance engages in the reaction within a day. This insight emphasizes the importance of supervising ambient temperature during [EMIM][Tf2N]’s storage and transportation.

To guarantee the safe management of [EMIM][Tf2N], the impact of heat sources on changes in ambient temperature must be considered. Adequate ventilation and temperature regulation measures should be employed to avert unplanned decomposition reactions, which could result in a faster reaction rate and potential thermal risks. In summary, creating a comprehensive reaction model for [EMIM][Tf2N] has yielded valuable information about the substance’s heat-releasing properties and safety aspects. By analyzing various temperature scenarios and evaluating TCL, the model offers insights into potential dangers associated with storage and transportation, ensuring that proper safety precautions are in place to minimize risks.

The reaction rate significantly impacts the hazard potential of a chemical reaction. As the reaction rate increases, the time taken for the reaction to form decreases, which is unfavorable for controlling process system stability in real-world scenarios and for implementing appropriate measures in emergency response systems. The maximum reaction rate indicates a high amount of heat release and the possibility of fire or explosion accidents. Therefore, the time at which the maximum reaction rate occurs serves as an evaluation basis for determining safety and time-related security measures.

Figure 5 demonstrates the calculation of TMR_ad (time to reach the maximum rate of adiabatic decomposition) across various ambient temperatures. When the temperature surpasses 280 °C, [EMIM][Tf2N]’s TMR_ad is achieved within a day, with this tendency further decreasing as temperatures rise. However, effective monitoring of temperature regulation during operation, storage, and transportation can ensure stability and process protection. Personnel needs to be alert to scenarios where temperature increases due to heat accumulation, even in low-temperature conditions. In such instances, a decomposition reaction may be initiated, leading to elevated overall temperatures and reaction rates, potentially resulting in thermal runaway.

Figure 6 explores the effects of heat accumulation properties on substance storage safety. Temperature is crucial for determining if a substance will experience a runaway reaction due to decomposition in subsequent storage states. When the temperature is close to 270 °C, the curve of the rising temperature of [EMIM][Tf2N] suggests no risk of a runaway reaction. However, if the temperature climbs to 280 °C, a thermal runaway reaction will transpire, and the increased ambient temperature reduces the time needed to achieve a runaway reaction. Furthermore, the storage amount of substances and the container type can impact the thermal runaway mode. The altered temperature curve indicates that a runaway reaction will cause a rapid temperature increase in a brief period.

From a preventive perspective, implementing temperature control measures during routine operations is essential, and disaster mitigation and emergency response plans should be assessed in case of an abrupt runaway reaction. Understanding the impact of reaction rate on hazard development and effectively managing temperature during operation, storage, and transportation can minimize potential risks associated with thermal runaway reactions. Elevated ambient temperatures reduce the time to achieve a runaway reaction, underscoring the significance of temperature control and monitoring in preserving the safety of chemical processes involving substances like [EMIM][Tf2N].

5. Conclusions

The reaction rate heavily influences chemical processes’ safety in ionic liquids like [EMIM][Tf2N]. As demonstrated in our study, when the temperature exceeds 287 °C, the time to reach the maximum reaction rate (TMR) becomes less than one day. This short TMR emphasizes the need for diligent process safety management.

When the temperature exceeds 265 °C, up to 70% of the substance may be consumed, posing a risk for thermal runaway. This finding further highlights the need for appropriate temperature control measures to prevent runaway reactions. It also underscores the importance of having robust emergency response plans in place to manage sudden runaway reactions effectively.

In addition to temperature considerations, this study highlights the influence of different storage container types on thermal safety. For example, a drum container might lose control at 270 °C. This insight should inform decisions about safe storage environments in the context of process safety.

6. Future Prospects

Future evaluation of storage conditions for ionic liquids such as [EMIM][Tf2N] and their storage incompatibilities should be carried out, focusing on long-term behavior (>7 days) using isothermal calorimetry below 100 °C to explore potential degradation and ensure safer industry practices.