Thermal hazard assessment and free radical inhibition of decomposition of tert-butyl perbenzoate

Danfeng Zhang; Zhiping Li; Juncheng Jiang; Lei Ni; Zhiquan Chen; Danfeng Zhang; Zhiping Li; Juncheng Jiang; Lei Ni; Zhiquan Chen

doi:10.48130/emst-0024-0029

2025 Volume 5

Article Contents

Abstract

Introduction

Single receiver system

System characteristics

Excitation current based on coupling range

Influence of coupling range

Multi-RX system

System characteristic and optimal driving current

Influence of coupling range

Experimental verification

Rights and permissions

References

Next Previous

ARTICLE Open Access

Thermal hazard assessment and free radical inhibition of decomposition of tert-butyl perbenzoate

College of Safety Science and Engineering, Nanjing Tech University, Nanjing 211816, Jiangsu, China

More Information

Corresponding authors: jcjiang_njtech@163.com; lei_ni@njtech.edu.cn

Received: 09 October 2024
Revised: 08 November 2024
Accepted: 13 November 2024
Published online: 21 January 2025
Emergency Management Science and Technology 5, Article number: e001 (2025) | Cite this article

Abstract

Tert-butyl perbenzoate (TBPB) is a common initiator widely used in polymerization processes, but the peroxide bond in its molecular structure is highly susceptible to breakage, leading to decomposition or even explosion. To explore the thermal behavior of TBPB and to inhibit the thermal hazard of free radicals generated during the reaction process, well-established calorimetric techniques were applied to measure the thermal stability of TBPB. The apparent activation energy of the TBPB decomposition reaction was also calculated using the Kissinger-Akahira-Sunose (KAS), Flynn-Wall-Ozawa (FWO), and Starink kinetic method. The thermal decomposition products of TBPB were determined by Fourier transform infrared spectroscopy (FTIR) experiment, and the qualitative analysis of free radicals generated during the reaction process was conducted by electron paramagnetic resonance spectroscopy (EPR) combined with free radical trapping technology. 2,2,6,6-tetramethylpiperidinooxy (TEMPO), a free radical trapping agent and inhibitor, was selected in this study as the thermal runaway inhibitor of the TBPB thermal decomposition reaction. Its inhibition effects on the corresponding free radicals and the thermal runaway of the TBPB decomposition reaction were verified. It is found that TEMPO can effectively reduce the potential thermal dangers and accident risks of TBPB, which provides a powerful reference for the prevention and management of thermal disasters during the production, storage, and transportation of TBPB.
- Thermal stability,
- Apparent activation energy,
- Thermal decomposition product,
- Free radical,
- Thermal runaway inhibitor

Introduction

In recent years, the widespread use of intelligent mobile terminals has been driving the process of societal intelligence. However, wired charging based on manual connections will severely limit this development^[1]. Inductive power transfer (IPT) has emerged as a promising remedy for charging these intelligent terminals, enabling efficient power transmission from a transmitter (TX) to a receiver (RX). Presently, research focusing on single-TX single-RX IPT systems has achieved considerable maturity. Substantial advancements have been accomplished in coil design, compensation network configuration, output control, efficiency enhancement, and the detection of foreign objects^[2−4]. These developments greatly support the expansion and application of this technology.

Multi-RX charging has emerged as an attractive and challenging area of research. Varied application scenarios present distinct system characteristics, necessitating a focused approach towards system-level design adaptable to diverse power requirements and coupling conditions. In addressing these challenges, extensive discussions have centered around coupler designs, leading to the development of various structures aimed at enhancing and stabilizing the coupling amidst diverse misalignment scenarios^[5,6]. Furthermore, the compensation networks have been recognized as crucial for attaining multiple objectives, including minimizing circulating energy, facilitating soft switching in active circuits, and ensuring output stability amidst load variations^[7]. Notably, beyond passive circuitry, significant progress has been achieved in active circuit designs tailored for multi-RX scenarios^[8,9]. These advancements in coupler and activating circuit designs serve as foundational requisites ensuring the steady-state performance of the system.

A well-performed multi-RX system necessitates an intelligent control methodology to fully exploit its potential benefits. The key control objectives primarily encompass output regulation, power distribution among RXs, and efficiency optimization. For instance, in cases where each RX possesses a specific resonance frequency, employing multiple-frequency excitation allows for independent power control across multiple RXs^[10−12]. The overall efficiency is enhanced in the study by Kim et al.^[13] through the computation of the optimal TX driving current, which heavily relies on the RX-side circuitry and real-time communication. Similarly, a communication-based control strategy has been devised for a system driven by a single Class E inverter, aiding in achieving simultaneous power and efficiency modulation^[14]. Various input modes are formulated in the study of Zhao et al.^[15] to optimize efficiency and cater to diverse power demands. However, these systems predominantly address specific scenarios, failing to comprehensively portray the inherent trade-offs and natural dependencies between achieved performance metrics (such as efficiency and power) and the requisite system complexity (e.g., real-time communication, information sensing, computational burden, and activation of circuits).

This manuscript centers on upholding high efficiency within a single-TX multi-RX system, without knowing real coupling. Employing the single-TX scenario as a case study, the examination delves into the impact of terminal conditions—specifically, the driving current and load—on the resonant tank. This investigation then steers the discourse towards an in-depth analysis of existing approaches aimed at maximizing efficiency at different power. In the absence of real-time coupling knowledge, an optimal current is deduced based on pre-established coupling ranges. As the number of RX grows, the suggested driving logic exhibits considerable scalability and an elevation in system intricacy. The attained efficiency under complete knowledge of all couplings stands as the reference point, only marginally surpassing the efficiency attained through the proposed control strategy.

Single receiver system

System characteristics

The depicted single-TX single-RX IPT system, shown in Fig. 1, encompasses essential components such as the inverter, coupler, compensation unit, rectifier, and load, which collectively elucidate the fundamental control logic. The active circuits, including the inverter, rectifier, and DC/DC converters in both the front-end and post stages, enable input and output modulation for diverse control objectives. Within this system, L_tx and r_tx denote the self-inductance and equivalent series resistance (ESR) of the TX coil, respectively, while L_rx and r_rx represent the self-inductance and ESR of the RX coil. The symbol M signifies the mutual inductance between the coils, while the coefficient k is defined as the coil coupling coefficient: $k=\dfrac{M}{\sqrt{{L}_{tx}{L}_{rx}}}$ .

Figure 1. Configuration of 1TX-1RX system.

DownLoad: Full-Size Img PowerPoint

In Fig. 1, both the TX and RX sides should offer at least two control variables to adjust the output and optimize efficiency indirectly. Typically, these control variables encompass the duty cycle of the DC/DC converter, and the phase shift angle of the inverter, or rectifier. Different combinations of these variables yield various control configurations, as documented in prior studies^[16−18]. Under specific output conditions, such as maintaining a constant voltage (CV), the controllable circuits must satisfy the load's power requirements while ensuring high efficiency throughout the entire system. When dealing with a system incorporating an uncontrollable resonant tank, the diverse configurations and control strategies mentioned earlier can be simplified and clarified through a steady-state analysis of the central resonant tank. This analysis focuses specifically on the integration of compensation and coupler elements and hinges upon terminal conditions—specifically, the input from the TX side and the load output on the RX side.

Despite the existence of various compensations, if assume that losses within the coupler exert a dominant influence on the overall losses, then the efficiency wouldn't be significantly impacted by compensation on the TX side. Consequently, this study is aimed at investigating the passive network depicted in Fig. 2, where TX compensation is disregarded, and a straightforward series compensation is implemented on the RX side. In this setup, C_rx denotes the series compensating capacitor, fulfilling the condition 1/(ωC_rx) = ωL_rx, while i_tx represents the input current to the TX coil.

Figure 2. Equivalent circuit under different control strategies. (a) Optimal load. (b) Optimal excitation.

DownLoad: Full-Size Img PowerPoint

In terms of the load viewpoint, controllable systems can be classified into two distinct types. The first type involves manipulating the RX-side circuitry to attain the optimal load while simultaneously adjusting the TX-side excitation to fulfill the output power requirements, as illustrated in Fig. 2a. In this arrangement, the rectifier and its subsequent circuit components are treated as a load resistance R_rx,opt when aiming for maximum coupler efficiency. The RX's output power during resonance is expressed as:

$P_o=\left(\dfrac{\omega MI_{tx}}{R_{rx,opt}+r_{rx}}\right)^2R_{rx,opt}$

(1)

The optimal load resistance R_rx,opt for efficiency maximization has been well studied and derived as:

$R_{rx,opt}=r_{rx}\sqrt{1+\dfrac{\omega^2M^2}{r_{tx}r_{rx}}}$

(2)

By solving Eqns (1) and (2) concurrently, it gives the root mean square (RMS) value of the necessary input current under the optimal condition, expressed as follows:

$I_{tx,ropt}=\dfrac{\sqrt{P_o}}{\omega M}\sqrt{2r_{rx}+\dfrac{2r_{tx}r_{rx}+\omega^2M^2}{\sqrt{r_{tx}r_{rx}+\omega^2M^2}}\sqrt{\dfrac{r_{rx}}{r_{tx}}}}$

(3)

The efficiency of the coupler is:

$\eta_{c,ropt}=1-\frac{2}{\sqrt{1+\dfrac{\omega^2M^2}{r_{tx}r_{rx}}}+1}$

(4)

which is loading and coupling dependent.

The second type of system aims to provide optimal driving current to maximize efficiency, and its load model is illustrated in Fig. 2b. In this setup, the RX-side circuit maintains the desired output characteristics, such as a constant voltage (CV), while satisfying specific power requirements P_o through RX-side control. Simultaneously, the TX-side circuit adjusts the input excitation to enhance efficiency. Despite the differing control methodologies between these two types, the eventual steady-state effects remain consistent. The primary difference lies in the approaches used to achieve this optimal state, involving distinctions in the activation of circuits, communication protocols, and the requisite information, particularly the coupling.

In Fig. 2b, maximizing efficiency is equivalent to minimizing losses for a target power. The power loss is expressed as:

$P_{\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s}}=P_{\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s},\mathrm{t}\mathrm{x}}+P_{\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s},\mathrm{r}\mathrm{x}}=I_{\mathrm{t}\mathrm{x}}^2r_{\mathrm{t}\mathrm{x}}+I_{\mathrm{r}\mathrm{x}}^2r_{\mathrm{r}\mathrm{x}}$

(5)

The optimal excitation at this point can be determined by solving the equation $\dfrac{d{P}_{\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s}}}{d{I}_{\mathrm{t}\mathrm{x}}}=0$ , yielding the coil's input current excitation I_tx,iopt, which would further determine the corresponding optimum efficiency η_c,iopt. Based on derivation, it would find I_tx,iopt = I_tx,ropt and η_c,iopt = η_c,ropt. The fact that both types of systems share the same excitation in the steady state indicates that, despite different control methods and transient processes, the systems reach identical stable states, i.e., there exists only one global maximum efficiency point. If R_rx,opt $\mathrm \gg$ r_rx, the losses on the TX and RX coils are equal and can be expressed as:

$P_{\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s},\mathrm{t}\mathrm{x}}\approx P_{\mathrm{l}\mathrm{o}\mathrm{s}\mathrm{s},\mathrm{r}\mathrm{x}}=\dfrac{P_o}{\sqrt{1+\dfrac{\omega^2M^2}{r_{\mathrm{t}\mathrm{x}}r_{\mathrm{r}\mathrm{x}}}}}$

(6)

It physically means the optimal efficiency indicates a loss balance condition.

By employing optimal load control as illustrated in Fig. 2a, achieving R_rx,opt demands coupling information for the RX, while the TX relies on receiving feedback (usually via wireless communication) from the RX to regulate power^[19]. Conversely, using the second approach depicted in Fig. 2b allows independent regulation of the RX through a linear controller. In the absence of communication, the TX must resort to either a non-linear tracking algorithm or a sensing-based coupling estimation approach to provide the necessary optimal driving current^[20]. Consequently, a noticeable trade-off emerges involving the communication requirement, controller type, and sensing necessity.

Excitation current based on coupling range

This paper presents a simplified excitation control approach based on the coupling range. The proposed method simplifies the control process by requiring determination of the coupling coefficient range during the coupler design phase. Following implementation, the system operates autonomously, obviating the necessity for real-time coupling estimation or nonlinear tracking algorithms. The core operational principle is demonstrated initially in a straightforward single-RX scenario, subsequently extending its applicability to a more intricate multi-RX scenarios.

In a single-RX system, ensuring sufficient coupling generally requires the RX to be charged within a specified charging area. In such instances, the range of coupling coefficient variations is limited and becomes known upon the fabrication of the coupler. The controllable circuit on the RX side can uphold a constant voltage output (fulfilling the load power requirements), while the TX side employs a real-coupling-independent control. The excitation current is determined by the range of coupling variations.

Figure 3 illustrates the efficiency curve of the coupler, demonstrating variations in input currents while regulating the RX side to achieve the target power. Within the graph, M_min signifies the minimum mutual inductance, M_max signifies the maximum mutual inductance, P_o represents the output power, and the upper and lower boundaries are established by the coupling range. The solid black line represents the system response at maximum coupling, while the solid red line represents the system response at minimum coupling. Within the determined coupling range, the actual response curve of the system is denoted by the dashed blue line. Given offset conditions, the blue line remains constrained within the upper and lower boundaries.

Figure 3. Coupler efficiency at different

$k$ and

${I}_{tx}$ .

DownLoad: Full-Size Img PowerPoint

In Fig. 3, L denotes the optimal efficiency point when the coupling is at its minimum. According to Eqn (3), this point is associated with an optimal input excitation, I_tx,L, given by:

$I_{tx,L}=\dfrac{\sqrt{P_o}}{\omega M_{\mathrm{m}\mathrm{i}\mathrm{n}}}\sqrt{2r_{rx}+\dfrac{2r_{tx}r_{rx}+\omega^2M_{\mathrm{m}\mathrm{i}\mathrm{n}}^2}{\sqrt{r_{tx}r_{rx}+\omega^2M_{\mathrm{m}\mathrm{i}\mathrm{n}}^2}}\sqrt{\dfrac{r_{rx}}{r_{tx}}}}$

(7)

Here, H corresponds to the optimal point when the coupling is at its maximum, associated with an optimal excitation, I_tx,H, given by:

$I_{tx,H}=\dfrac{\sqrt{P_o}}{\omega M_{\mathrm{m}\mathrm{a}\mathrm{x}}}\sqrt{2r_{rx}+\dfrac{2r_{tx}r_{rx}+\omega^2M_{\mathrm{m}\mathrm{a}\mathrm{x}}^2}{\sqrt{r_{tx}r_{rx}+\omega^2M_{\mathrm{m}\mathrm{a}\mathrm{x}}^2}}\sqrt{\dfrac{r_{rx}}{r_{tx}}}}$

(8)

If employing the non-linear perturbation observation method, the system requires continuous tracking of the optimal excitation between I_tx,L and I_tx,H^[21]. Despite the absence of direct communication between the TX and RX sides, disturbances on the TX side can affect the stability of the RX side output. In scenarios without non-linear tracking, estimating the coupling based on TX side information becomes necessary to maintain the coupling-dependent optimal current^[20].

Within the specified coupling range, this paper proposes a strategy that utilizes the average of I_tx,L and I_tx,H as the driving current, denoted as:

$I_{tx,M}\ =\ \dfrac{I_{tx,L}+I_{tx,H}}{2}$

(9)

The median current value, I_tx,M, is influenced by coil parameters, load power, and the coupling range. When the real coupling is M_min, the operating point becomes L' using I_tx,M as the driving current. The efficiency difference between L' and L is Δη_C,L. Conversely, at maximum coupling, the operating point is H', and the efficiency difference between H' and H is Δη_C,H. Δη_C,L and Δη_C,H depict the trade-off incurred by using the real-coupling-independent I_tx,M.

The proposed excitation control strategy eliminates the necessity for real-time tracking of the highest efficiency point and instead maintains operation at a compromise point. At this juncture, the TX side losses stabilize at ${I}_{tx,M}^{2}\cdot {r}_{tx}$ , and different couplings determine varying currents on the RX-side coil. When the current on the RX-side coil reaches a certain value such that the losses at both ends of the coil become equal, i.e., ${I}_{tx,M}^{2}\cdot {r}_{tx}={I}_{rx}^{2}\cdot {r}_{rx}$ , the system attains its actual maximum efficiency point.

Influence of coupling range

To provide a quantitative verification and study the impact of the coupling range, a simulation is employed for further elucidation of the qualitative analysis. The simulation utilizes the following parameters: f = 100 kHz, L_tx = 90 μH, L_rx = 75 μH, r_tx = 0.192 Ω (Q_tx = 250), r_rx = 0.174 Ω (Q_rx = 230), and P_o = 20 W. As the coupling varies, the coupler efficiency changes with the input excitation, as depicted in Fig. 4. Region A delineates the operational area when k ranges within [0.1,0.2]. The simulation results indicate that with an increase in k, the corresponding optimal excitation current decreases.

Figure 4. Efficiency of the 1-RX system under different coupling conditions.

DownLoad: Full-Size Img PowerPoint

If the driving current is controlled to be the optimal value, i.e., I_tx = I_tx,opt, the efficiency is depicted by the black dashed line in Fig. 5, which connects all peak points of Fig. 4. This reference curve serves as the basis for analyzing the influence when I_tx changes to I_tx,M. For instance, considering k $\in$ [0.1, 0.2], the efficiency achieved by I_tx,M—calculated using Eqns (7), (8), and (9)—is represented as case A (yellow solid line) in Fig. 5. It marginally falls below the reference curve, with a maximum drop of 0.5%. Expanding the coupling range to k $\in$ [0.08, 0.2], corresponding to the region between the green and blue solid lines of Fig. 4, yields the efficiency achieved by I_tx,M represented as case B in Fig. 5, exhibiting a maximum drop of 1%. When k $\in$ [0.06, 0.2], as demonstrated in case C, it results in a maximum 2% efficiency drop. These simulation outcomes indicate that adopting I_tx,M based on the coupling range demonstrates efficiency performance close to real-time tracking of the optimal excitation, showcasing favorable system characteristics.

Figure 5. Efficiency when the coupling range changes.

DownLoad: Full-Size Img PowerPoint

Multi-RX system

System characteristic and optimal driving current

Figure 6 illustrates a single-TX n-RX system. L_rx,j and r_rx,j denote the self-inductance and ESR of the j-th RX coil. k_j is the coupling coefficient between the TX and the j-th RX, satisfying: ${k}_{j}={M}_{j}/\sqrt{{L}_{tx}{L}_{rx,j}}$ . C_rx,j is the series compensation capacitor, and R_rx,j is the equivalent AC load resistance. The resonance on the RX side satisfies the condition: ωL_rx,j − 1/(ωC_rx,j) = 0. In scenarios where charging occurs in a flat plane, having multiple RXs positioned over the TX charging area can induce cross-coupling between the RXs. Since the RX coil is enclosed within the device, the minimum coil clearance between adjacent devices becomes a variable during the design phase. Adequate clearance between coils is essential to minimize cross-coupling effects. Consequently, the analysis conducted above assumes a model where the RX cross-coupling is not considered.

Figure 6. Simplified circuit model of one-TX n-RX system.

DownLoad: Full-Size Img PowerPoint

Similar to the single RX scenario depicted in Fig. 2, when employing different control logics in RXs, it becomes essential to individually examine the optimal conditions achieved. In scenarios where the system offers multiple control options, adjusting the load resistance at each RX becomes crucial to maximize coupler efficiency. As outlined in the study by Fu et al.^[22] , the necessary load resistance values and their respective efficiencies are:

$R_{rx,j,opt}=r_{rx,j}\sqrt{1+\displaystyle\sum_{j=1}^n\dfrac{\omega^2M_j^2}{r_{tx}r_{rx,j}}}$

(10)

$\eta_{c,ropt}=1-\dfrac{2}{\sqrt{1+\displaystyle\sum_{j=1}^n\dfrac{\omega^2M_j^2}{r_{tx}r_{rx,j}}}+1}$

(11)

In contrast to the single-RX system, the aforementioned multi-RX system possesses n power outputs and n optimal load resistance values. It offers n + 1 degrees of freedom for control, provided by the TX and RX sides (n on the RX side, 1 on the TX side). To achieve global optimal efficiency, i.e., satisfy Eqn (11), each load should adhere to the optimal load condition, i.e., Eqn (10), through its respective control circuit. In this scenario, the power distribution among the loads is constrained to follow the following proportions:

${P}_{1}:{P}_{2}:\dots :{P}_{n}=\dfrac{{M}_{1}^{2}}{{r}_{rx,1}}:\dfrac{{M}_{2}^{2}}{{r}_{rx,2}}:\dots :\dfrac{{M}_{n}^{2}}{{r}_{rx,n}}$

(12)

There are no degrees of freedom available for power regulation. Therefore, the control logic of Fig. 2a cannot be extended to a multi-RX scenario.

When the RX-side circuit is dedicated to maintaining a constant voltage, the excitation current on the TX side becomes pivotal in enhancing the system's efficiency. This reflects the optimal excitation control employed in a scenario featuring a single TX and a single RX. In this context, maximizing the system's efficiency can be reformulated as a problem centered on minimizing losses while adhering to equality constraints, as articulated below:

$\begin{array}{c}\begin{array}{l}\mathrm{M}\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{m}\mathrm{i}\mathrm{z}\mathrm{e}:\ \ \ I_{tx}^2r_{tx}+\displaystyle\sum_{j=1}^n\left(\dfrac{\omega M_jI_{tx}}{R_{rx,j}+r_{rx,j}}\right)^2r_{rx,j} \\ \mathrm{S}\mathrm{u}\mathrm{b}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}\ \mathrm{t}\mathrm{o}:\ \ \ \left(\dfrac{\omega M_jI_{tx}}{R_{rx,j}+r_{rx,j}}\right)^2R_{rx,j}=P_j,\ \ \ \ j=\mathrm{1,2},...,n\end{array}\end{array}$

(13)

The method for solving such an optimization problem involves the Lagrange multiplier method. Equation (14) represents the Lagrangian, where L is the Lagrange function, and λ_i is the Lagrange multiplier.

$\begin{split}L ={I}_{tx}^{2}{r}_{tx}+\sum _{j=1}^{n}{\left(\dfrac{\omega {M}_{j}{I}_{tx}}{{R}_{rx,j}+{r}_{rx,j}}\right)}^{2}{r}_{rx,j} +\sum _{j=1}^{n}{\lambda }_{j}\left({\left(\dfrac{\omega {M}_{j}{I}_{tx}}{{R}_{rx,j}+{r}_{rx,j}}\right)}^{2}{R}_{rx,j}-{P}_{j}\right)\end{split}$

(14)

If ${R}_{rx,opt}\gg {r}_{rx}$ , by solving $\dfrac{\partial L}{\partial {I}_{tx}}=0$ and $\dfrac{\partial L}{\partial {\lambda }_{j}}=0$ , the optimal input excitation current I_tx, of the system under known coupling coefficients, which maximizes system efficiency, can be obtained as follows:

${I}_{tx,opt}={\left(\dfrac{1}{{r}_{tx}}\sum _{j=1}^{n}\dfrac{{P}_{j}^{2}{r}_{rx,j}}{{\omega }^{2}{M}_{j}^{2}}\right)}^{\frac{1}{4}}$

(15)

Substituting Eqn (15) into the efficiency expression yields:

$\begin{array}{c}{\eta }_{c,iopt}=\dfrac{\sum _{j=1}^{n}{P}_{j}}{\sum _{j=1}^{n}{P}_{j}+2{\left(\sum _{j=1}^{n}\dfrac{{P}_{j}^{2}{r}_{rx,j}}{{\omega }^{2}{M}_{j}^{2}}\right)}^{\frac{1}{2}}}\end{array}$

(16)

At this point, the coupler achieves an efficiency η_c,iopt while facing power constraints on the RX side, rather than the globally optimal efficiency η_c,ropt obtained without power constraints, expresses the system efficiency when operating at the global optimum, representing the theoretical efficiency limit of the system. Therefore, η_c,iopt ≤ η_c,ropt, signifying that under the control strategy of optimal excitation, the efficiency is always lower than the efficiency under optimal load conditions. The losses on the TX and RX sides at this point are given by:

${P}_{loss,tx}={P}_{loss,rx}={\left(\sum _{j=1}^{n}\dfrac{{P}_{j}^{2}{r}_{rx,j}}{{\omega }^{2}{M}_{j}^{2}}\right)}^{\frac{1}{2}}$

(17)

This conclusion indicates that under optimal excitation, the coil conduction losses on the TX side and RX side are equal, consistent with the characteristics of a single-RX system [refer to Eqn (5)].

From the above analysis, it can be concluded that for a general multi-RX system with different RX coils, various couplings, and distinct power requirements, when the coil parameters and output power are determined, the coupler efficiency depends solely on the coupling and the input excitation I_tx. This characteristic is akin to the single-RX case. Despite the increase in the number of outputs, the control freedom of TX remains unchanged. The upper and lower boundaries of the efficiency curve are still determined by the coupling variation range. Therefore, the efficiency curve shown in Fig. 4 is still applicable to a multi-RX system. The red solid line represents the efficiency curve when the coupling between the TX and all RXs is minimized (i.e., M_j = M_min). At this point, the current at the efficiency peak L is given by:

$I_{tx,L}=\left(\dfrac{1}{r_{tx}}\sum_{j=1}^n\dfrac{P_j^2r_{rx,j}}{\omega^2M_{\mathrm{m}\mathrm{i}\mathrm{n}}^2}\right)^{\frac{1}{4}}$

(18)

The black solid line represents the efficiency curve when the coupling between the TX and all RXs is maximized (i.e., M_j = M_max). The current at the corresponding efficiency peak is given by the following:

$I_{tx,H}=\left(\dfrac{1}{r_{tx}}\sum_{j=1}^n\dfrac{P_j^2r_{rx,j}}{\omega^2M_{\mathrm{m}\mathrm{a}\mathrm{x}}^2}\right)^{\frac{1}{4}}$

(19)

It is possible to adopt the average current, which mirrors the approach used in the single-RX system, i.e.,

$I_{tx,M}=\dfrac{I_{tx,L}+I_{tx,H}}{2}$

(20)

which is also influenced by coil parameters, load power, and the coupling range. Given the known coupling range, the input excitation is determined without the need for real-time coupling information. When using the proposed excitation, the TX coil loss is ${P}_{loss,tx}={I}_{tx,M}^{2}{R}_{tx}$ , while the RX-side losses vary with the coupling. When the coupling changes to the condition given by Eqn (17), where the sum of RX-side losses equals the TX side losses, the system reaches its optimal efficiency state, calculated as Eqn (16).

In a single-RX system, the TX side must obtain specific coupling and load information through communication or detection means and then calculate the optimal excitation. Similarly, in a multi-RXs system, if coupling estimation or similar methods are employed, the difficulty of estimating information and the circuit cost will increase dramatically due to the addition of RX devices. Therefore, the proposed driving strategy would show great scalability and is particularly advantageous in multi-RX systems.

Influence of coupling range

Taking the example of a dual-RX system to study the influence of coupling, Table 1 shows the simulation parameters. The coupling range is k₁ $\in$ [0.1, 0.34] for RX1, and k₂ $\in$ [0.15, 0.45] for RX2. In Fig. 7, when meeting the power demand, the coupler efficiency changes with input excitation, and different colors represent different coupling conditions. The blue line represents the coil efficiency change curve when the coupling is strongest (k₁ = 0.2, k₂ = 0.25), the orange curve represents an intermediate state (k₁ = 0.34, k₂ = 0.45), and the black solid line is used to represent the weakest coupling (k₁ = 0.1, k₂ = 0.15). Similar to Fig. 4, stronger coupling leads to a smaller optimal excitation current and higher efficiency. Additionally, the power level and coil parameters also influence the shape and position of the efficiency curve.

Table 1. Parameters of 1TX-2RX IPT system.

Symbol	Value	Symbol	Value	Symbol	Value
f	100 kHz	L_tx	91.3 μH	r_tx & r_rx	0.195 Ω
L_rx,1	22.3 μH	r_rx,1	0.059 Ω	k₁	[0.1, 0.34]
L_rx,2	26.4 μH	r_rx,2	0.074 Ω	k₂	[0.15, 0.45]
P₁	15 W	P₂	10 W

| Show Table

DownLoad: CSV

Figure 7. Coupler efficiency of a 2-RX system under different coupling conditions.

DownLoad: Full-Size Img PowerPoint

Analogous to the single-RX system, for certain spatial offsets in practical applications, real-coupling-independent I_tx,M is calculated by simultaneously solving Eqns (18), (19), and (20). Therefore, when I_tx,M and I_tx,opt are used, the efficiency difference can explain the high-efficiency operation.

The coupling range denoted by region C of Fig. 7 means k₁ $\in$ [0.2, 0.34] and k₂ $\in$ [0.25, 0.45]. It is defined as Case C, and the corresponding simulation results are shown in Fig. 8a. The left graph illustrates the coupler efficiency η_c,M under excitation I_tx,M, while the right graph displays the efficiency difference Δη_c between the maximum coil efficiency η_c,iopt (using I_tx,opt) and η_c,M (using I_tx,M). Within the coupling range, the efficiency difference is less than 0.2%. As shown in Fig. 8b, with a further expansion of the coupling range in Case D (k₁ $\in$ [0.1, 0.34], k₂ $\in$ [0.15, 0.45]), the maximum efficiency drop is 0.8%. From these examples, it is evident that adopting the proposed current would maintain relatively high coupler efficiency without dramatically increasing the requirement of real-time coupling information.

Figure 8. Efficiency comparison. (a) Case C: η_s,M (left), Δη_c (right); (b) Δη_c (left), Δη_c (right).

DownLoad: Full-Size Img PowerPoint

Experimental verification

An experimental platform was implemented using a single-TX dual-RX system to validate the theoretical conclusion. As illustrated in Fig. 9 the setup includes a DC source, a full-bridge inverter, a large square TX coil, two small RX coils (rectangular RX1 and square RX2), rectifiers, and loads. The TX side employs an LCC compensation structure, while the RX sides utilize series compensation. This compensation is convenient to use the DC input voltage for the TX coil current control due to the clamping effect. The coil parameters are provided in Table 1, and the compensation parameters are as follows: L_t = 32 μH, C_t = 110 nF, C_tx = 57 nF, C_rx1 = 227 nF, C_rx2 = 132.8 nF. The charging powers for the two loads are P₁ = 10 W and P₂ = 15 W.

Figure 9. Experimental setup.

DownLoad: Full-Size Img PowerPoint

When two RXs are placed within a specific charging area in Fig. 9 (i.e., not moving outside the TX), the coupling variation ranges are measured as k₁ $\in$ [0.26, 0.34] and k₂ $\in$ [0.31, 0.45]. Using the driving current Eqn (20), with an effective value of I_tx,M = 1.151 A, a corresponding DC voltage of V_dc = 23 V is provided. Note that the coil current is clamped by the input DC voltage in LCC compensation. Within the mentioned coupling variation range and power demands, the experimental waveforms of inverter output voltage v_in, input current i_in, and TX coil input current i_tx are depicted in Fig. 10. When employing I_tx,M, v_in, and i_tx remain constant, while i_in exhibits slight fluctuations under coupling variation.

Figure 10. Experimental waveform.

DownLoad: Full-Size Img PowerPoint

The efficiency analysis in the preceding section exclusively concentrates on the coupler itself, but the efficiency is also influenced by compensations, inverters, and rectifiers. And the characteristic traits of the efficiency curve primarily stem from the attributes of the coupler. Past studies have emphasized that a precise analysis of system efficiency necessitates the consideration of all loss models. However, when optimizing efficiency, the control mechanisms often resort to simplified analyses to deduce the terminal requisites for the resonant tank. Hence, in the ensuing analysis, a comprehensive system-level evaluation could be undertaken to yield results akin to those illustrated in Figs 7 and 8.

Similar to the AC simulation in Fig. 7, the influence of the coupling could be studied in a complete DC to DC system. The overall efficiency η_s is measured by sweeping the input current as shown in Fig. 11. An optimal current exists, and it is still valid to have a conclusion that a stronger coupling leads to a lower optimal current.

Figure 11. Efficiency under different coupling conditions.

DownLoad: Full-Size Img PowerPoint

When employing I_tx,M, as illustrated in Fig. 10, and moving two RX coils to change their coupling, the input and output power were measured using a power analyzer. The system efficiency η_s,M is then obtained in Fig. 12a, which changes from 89.5% to 92.6%. Simultaneously, the optimal efficiency values η_c,sopt (obtained by sweeping the input current) under different couplings are measured and compared with η_s,M. The difference is illustrated in Fig. 12b. At higher coupling, η_s,M is closer to η_s,opt, differing by only 0.4%. As the coupling weakens, the efficiency difference gradually increases, reaching a maximum of 1.7%. The proposed control can maintain high-efficiency operations without the need for coupling estimation.

Figure 12. (a) System efficiency η_s,M. (b) Efficiency drop:

$\Delta$ η_s = η_s,opt − η_s,M.

DownLoad: Full-Size Img PowerPoint

The similar trends in simulation and experiment indicate that the differences in efficiency are mainly attributed to the losses in the active circuits (inverter and rectifier) and conductor losses. The AC analysis in the previous sections would be used to guide the current control in a complete DC-to-DC system. The experimental results demonstrate that the proposed excitation based on the coupling range maintains a certain level of high efficiency, with stable input and simple control strategies.

Conclusions

This paper proposes a simple driving logic to dramatically improve the scalability of a general single-TX multi-RX IPT system. The high-efficiency operation is achieved without knowing the real coupling and does not suffer from the coupling variation issue. Using the one-RX example, the coupler analysis unified the understanding of optimal conditions, which serves as the reference to justify the benefit of the proposed excitation. The same control logic is then extended to a general multi-RX case based on the derivation of optimal conditions. The claimed benefit is finally verified in a complete two-RX system.

Author contributions

The authors confirm contribution to the paper as follows: theoretical derivation, experimental validation, and manuscript writing: Wang X, He R; manuscript writing: Lin Z; conceptualization, funding acquisition, and supervision: Fu M. All authors reviewed the results and approved the final version of the manuscript.

Data availability

The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.

Acknowledgments

This work was supported by National Natural Science Foundation of China (Grant No. 52477013).

Conflict of interest

The authors declare that they have no conflict of interest.

Rights and permissions
Copyright: © 2025 by the author(s). Published by Maximum Academic Press on behalf of Nanjing Tech University. This article is an open access article distributed under Creative Commons Attribution License (CC BY 4.0), visit https://creativecommons.org/licenses/by/4.0/.

References

[1]	Zhou HL, Jiang JC, Huang AC. 2023. Thermal hazard assessment of tert-butyl perbenzoate using advanced calorimetric techniques and thermokinetic methods. Journal of Loss Prevention in the Process Industries 85:105166 doi: 10.1016/j.jlp.2023.105166 CrossRef Google Scholar
[2]	Zhou HL, Jiang JC, Huang AC. 2024. Calorimetric and kinetic evaluation of thermal stability and process safety in tert-butyl peroxy-2-ethylhexanoate and tert-butyl peroxybenzoate. Process Safety and Environmental Protection 185:602−13 doi: 10.1016/j.psep.2024.03.045 CrossRef Google Scholar
[3]	Huang AC, Huang CF, Xing ZX, Jiang JC, Shu CM. 2019. Thermal hazard assessment of the thermal stability of acne cosmeceutical therapy using advanced calorimetry technology. Process Safety and Environmental Protection 131:197−204 doi: 10.1016/j.psep.2019.09.016 CrossRef Google Scholar
[4]	Milas NA, Surgenor DM. 1946. Studies in organic peroxides; t-amyl hydroperoxide and di-t-amyl peroxide. Journal of the American Chemical Society 68:643 doi: 10.1021/ja01208a034 CrossRef Google Scholar
[5]	Wei W, Zhang C, Xu Y, Wan X. 2011. Synthesis of tert-butyl peresters from aldehydes by Bu4NI-catalyzed metal-free oxidation and its combination with the Kharasch–Sosnovsky reaction. Chemical Communications 47:10827−29 doi: 10.1039/c1cc14602e CrossRef Google Scholar
[6]	Zhang H, Dong DQ, Hao SH, Wang ZL. 2016. Bu4NI-catalyzed construction of tert-butyl peresters from alcohols. RSC Advances 6:8465−68 doi: 10.1039/C5RA27500H CrossRef Google Scholar
[7]	Chen X, Li Y, Wu M, Guo H, Jiang L, et al. 2016. An efficient method for the preparation of tert-butyl esters from benzyl cyanide and tert-butyl hydroperoxide under the metal free condition. RSC Advances 6:102023−27 doi: 10.1039/C6RA20966A CrossRef Google Scholar
[8]	Hashemi H, Saberi D, Poorsadeghi S, Niknam K. 2017. Temperature-controlled solvent-free selective synthesis of tert-butyl peresters or acids from benzyl cyanides in the presence of the TBHP/Cu(OAc)₂ system. RSC Advances 7:7619−22 doi: 10.1039/C6RA27921J CrossRef Google Scholar
[9]	Singha R, Shit P. 2020. Sunlight assisted solvent free synthesis of tert-butylperesters. Synthetic Communications 50:2698−703 doi: 10.1080/00397911.2020.1783560 CrossRef Google Scholar
[10]	Tseng JM, Lin YF. 2011. Evaluation of a tert-butyl peroxybenzoate runaway reaction by five kinetic models. Industrial & Engineering Chemistry Research 50:4783−87 doi: 10.1021/ie100640t CrossRef Google Scholar
[11]	Lv JY, Wei S, Chen WH, Chen GF, Chen LP, et al. 2012. Thermal kinetic analysis of tert-butyl peroxybenzoate under dynamic and adiabatic conditions. Advanced Materials Research 550−553:2782−85 doi: 10.4028/www.scientific.net/AMR.550-553.2782 CrossRef Google Scholar
[12]	Wei TT, Qian XM, Yuan MQ. 2015. Thermal hazard analysis for tert-butyl peroxybenzoate contaminated by acid or alkali. CIESC Journal 66(10):3931−39 doi: 10.11949/j.issn.0438-1157.20141395 CrossRef Google Scholar
[13]	Jiang JC, Li L, Jiang JJ, Wang Y, Lo SM, et al. 2019. Effect of ionic liquids on the thermal decomposition process of tert-butyl peroxybenzoate (TBPB) by DSC. Thermochimica Acta 671:127−33 doi: 10.1016/j.tca.2018.11.017 CrossRef Google Scholar
[14]	Moane S, Raftery DP, Smyth MR, Leonard RG. 1999. Decomposition of peroxides by transition metal ions in anaerobic adhesive cure chemistry. International Journal of Adhesion and Adhesives 19:49−57 doi: 10.1016/S0143-7496(98)00056-6 CrossRef Google Scholar
[15]	Lv J, Chen W, Chen L, Tian Y, Yan J. 2014. Thermal risk evaluation on decomposition processes for four organic peroxides. Thermochimica Acta 589:11−18 doi: 10.1016/j.tca.2014.05.013 CrossRef Google Scholar
[16]	Dobbs AP, Jones P, Penny MJ, Rigby SE. 2009. Light-fluorous TEMPO: reagent, spin trap and stable free radical. Tetrahedron 65:5271−77 doi: 10.1016/j.tet.2009.04.078 CrossRef Google Scholar
[17]	Barton DHR, Le Gloahec VN, Smith J. 1998. Study of a new reaction: trapping of peroxyl radicals by TEMPO. Tetrahedron Letters 39:7483−86 doi: 10.1016/S0040-4039(98)01628-1 CrossRef Google Scholar
[18]	Makino K, Hagiwara T, Murakami A. 1991. A mini review: Fundamental aspects of spin trapping with DMPO. International Journal of Radiation Applications and Instrumentation. Part C: Radiation Physics and Chemistry 37:657−65 doi: 10.1016/1359-0197(91)90164-W CrossRef Google Scholar
[19]	Buettner GR. 1993. The spin trapping of superoxide and hydroxyl free radicals with DMPO (5,5-dimethylpyrroline-N-oxide): more about iron. Free Radical Research Communications 19:s79−s87 doi: 10.3109/10715769309056s79 CrossRef Google Scholar
[20]	Villamena FA, Locigno EJ, Rockenbauer A, Hadad CM, Zweier JL. 2006. Theoretical and experimental studies of the spin trapping of inorganic radicals by 5,5-dimethyl-1-pyrroline N-oxide (DMPO). 1. carbon dioxide radical anion. The Journal of Physical Chemistry A 110:13253−58 doi: 10.1021/jp064892m CrossRef Google Scholar
[21]	Yao H, Jiang J, Li B, Ni L, Ni Y, et al. 2022. Investigation of pyrolysis kinetics, mechanism and thermal stability of tert-butyl peroxy-2-ethyl hexanoate. Process Safety and Environmental Protection 160:734−48 doi: 10.1016/j.psep.2022.02.059 CrossRef Google Scholar
[22]	Tang Y, Li ZP, Zhou HL, Miao CF, Jiang JC, et al. 2023. Thermal stability assessment of nitrocellulose by using multiple calorimetric techniques and advanced thermokinetics. Journal of Thermal Analysis and Calorimetry 148:5029−38 doi: 10.1007/s10973-022-11754-1 CrossRef Google Scholar
[23]	Huang AC, Li ZP, Liu YC, Tang Y, Huang CF, et al. 2021. Essential hazard and process safety assessment of para-toluene sulfonic acid through calorimetry and advanced thermokinetics. Journal of Loss Prevention in the Process Industries 72:104558 doi: 10.1016/j.jlp.2021.104558 CrossRef Google Scholar
[24]	Liu YC, Zhou HL, Tang Y, Li Y, Zhai J, et al. 2023. Thermal hazard assessment by TGA, DSC, and ARC experimental and simulated thermokinetic approaches for trinitrophloroglucinol. Journal of Thermal Analysis and Calorimetry 148:5039−49 doi: 10.1007/s10973-022-11649-1 CrossRef Google Scholar
[25]	Li ZP, Huang AC, Tang Y, Zhou HL, Liu YC, et al. 2022. Thermokinetic prediction and safety evaluation for toluene sulfonation process and product using calorimetric technology. Journal of Thermal Analysis and Calorimetry 147:12177−86 doi: 10.1007/s10973-022-11384-7 CrossRef Google Scholar
[26]	Alavi SE, Ali Abdoli M, Khorasheh F, Bayandori Moghaddam A. 2020. Non-isothermal pyrolysis of used lubricating oil and the catalytic effect of carbon-based nanomaterials on the process performance. Journal of Thermal Analysis and Calorimetry 139:1025−36 doi: 10.1007/s10973-019-08436-w CrossRef Google Scholar
[27]	Wu ZH, Huang AC, Tang Y, Yang YP, Liu YC, et al. 2021. Thermal effect and mechanism analysis of flame-retardant modified polymer electrolyte for lithium-Ion battery. Polymers 13:1675 doi: 10.3390/polym13111675 CrossRef Google Scholar
[28]	Zhou HL, Jiang JC, Huang AC, Tang Y, Zhang Y, et al. 2022. Calorimetric evaluation of thermal stability and runaway hazard based on thermokinetic parameters of O, O–dimethyl phosphoramidothioate. Journal of Loss Prevention in the Process Industries 75:104697 doi: 10.1016/j.jlp.2021.104697 CrossRef Google Scholar
[29]	Li X, Yao H, Lu X, Chen C, Cao Y, et al. 2020. Effect of pyrogallol on the ring-opening polymerization and curing kinetics of a fully bio-based benzoxazine. Thermochimica Acta 694:178787 doi: 10.1016/j.tca.2020.178787 CrossRef Google Scholar
[30]	Cao CR, Liu SH, Huang AC, Lee MH, Ho SP, et al. 2018. Application of thermal ignition theory of di(2, 4-dichlorobenzoyl) peroxide by kinetic-based curve fitting. Journal of Thermal Analysis and Calorimetry 133:753−61 doi: 10.1007/s10973-018-7002-8 CrossRef Google Scholar
[31]	Wei R, Huang S, Wang Z, Wang C, Zhou T, et al. 2018. Effect of plasticizer dibutyl phthalate on the thermal decomposition of nitrocellulose. Journal of Thermal Analysis and Calorimetry 134:953−69 doi: 10.1007/s10973-018-7521-3 CrossRef Google Scholar
[32]	Li ZP, Jiang JC, Huang AC, Tang Y, Miao CF, et al. 2021. Thermal hazard evaluation on spontaneous combustion characteristics of nitrocellulose solution under different atmospheric conditions. Scientific Reports 11:24053 doi: 10.1038/s41598-021-03579-z CrossRef Google Scholar
[33]	Huang AC, Huang CF, Tang Y, Xing ZX, Jiang JC. 2021. Evaluation of multiple reactions in dilute benzoyl peroxide concentrations with additives using calorimetric technology. Journal of Loss Prevention in the Process Industries 69:104373 doi: 10.1016/j.jlp.2020.104373 CrossRef Google Scholar
[34]	Vyazovkin S, Burnham AK, Criado JM, Pérez-Maqueda LA, Popescu C, et al. 2011. ICTAC Kinetics Committee recommendations for performing kinetic computations on thermal analysis data. Thermochimica Acta 520:1−19 doi: 10.1016/j.tca.2011.03.034 CrossRef Google Scholar
[35]	Shen S, Jiang J, Zhang W, Ni L, Shu CM. 2018. Process safety evaluation of the synthesis of tert-butyl peracetate. Journal of Loss Prevention in the Process Industries 54:153−62 doi: 10.1016/j.jlp.2018.03.009 CrossRef Google Scholar
[36]	Zhang H, Jiang J, Fei M, Ni L, Hang Y. 2022. Thermal hazard characteristics and essential mechanism study of 1-hydroxybenzotriazole: Thermodynamic study combined DFT simulation. Process Safety and Environmental Protection 168:713−22 doi: 10.1016/j.psep.2022.10.043 CrossRef Google Scholar
[37]	Zhang H, Jiang JC, Yan TY, Ni L, Liu SH. 2023. Thermal hazard risk and decomposition mechanism identification of 1-Hexyl-2,3-dimethylimidazolium nitrate: combined thermal analysis experiment and DFT emulation. Process Safety and Environmental Protection 172:38−47 doi: 10.1016/j.psep.2023.01.065 CrossRef Google Scholar

About this article

Cite this article

Zhang D, Li Z, Jiang J, Ni L, Chen Z. 2025. Thermal hazard assessment and free radical inhibition of decomposition of tert-butyl perbenzoate. Emergency Management Science and Technology 5: e001 doi: 10.48130/emst-0024-0029

Zhang D, Li Z, Jiang J, Ni L, Chen Z. 2025. Thermal hazard assessment and free radical inhibition of decomposition of tert-butyl perbenzoate. Emergency Management Science and Technology 5: e001 doi: 10.48130/emst-0024-0029

Figures(12) / Tables(4)

Download PDF

Article Metrics

Article views(881) PDF downloads(299)

Parameters	Kinetic methods			Average values
Parameters	KAS	FWO	Starink	Average values
$\overline E_{\rm a}$ (kJ/mol)	96.80	98.72	97.14	97.55
$\overline R^2$	0.9636	0.9681	0.9641	0.9653

Absorption peaks	Free radical attribution	g_iso	A_N (Gauss)	A_H (Gauss)
1	Alkoxy 1	2.0081	13.63	8.91
2	Alkoxy 2	2.0072	13.63	9.12
3	Alkyl	2.0075	14.06	20.49
4	The oxide of DMPO	2.0073	13.95	–
−: Not applicable.

Samples	T_o (°C)	T_p (°C)	ΔH (J/g)
TBPB	100.93	165.60	865.87
TBPB/TEMPO	128.65	162.49	585.00

Samples	T_oad (°C)	T_fad (°C)	ΔT_ad (°C)	ΔT_ad^* (°C)	P_oad (MPa)	P_fad (MPa)	ΔP_ad (MPa)	ΔH_ad (J/g)
TBPB	84.93	230.76	145.83	193.17	0.20	1.81	1.61	1989.65
TBPB/TEMPO	92.03	204.73	112.70	149.25	0.28	1.46	1.18	1537.29

Contents

Policies

Services

Partnerships

Thermal hazard assessment and free radical inhibition of decomposition of tert-butyl perbenzoate

Abstract

Introduction

Single receiver system

System characteristics

Excitation current based on coupling range

Influence of coupling range

Multi-RX system

System characteristic and optimal driving current

Influence of coupling range

Experimental verification

Conclusions

Author contributions

Data availability

Acknowledgments

Conflict of interest

Rights and permissions

References

About this article

Cite this article

Article Metrics

Access History

Other Articles By Authors