Thermal and combustion behaviors of aluminum/manganese dioxide/fluoroelastomer terpolymer nanothermite

Fluoroelastomer has received increasing attention for energetic materials application due to its high fluorine contents. Different contents (0 wt%–30 wt%) of poly(VDF‐ter‐HFP‐ter‐TFE) terpolymer are added into Al/MnO2 nanothermite. The peak exothermic temperature of thermite reaction for Al/MnO2 system is about 554°C with 1070 Jg−1 heat release. After adding terpolymer, it mainly exists in the gap among Al nanoparticles and MnO2 nanorods, and react with Al and MnO2 at the range of 350 to 540°C before the occurrence of the thermite reaction. 10 wt% terpolymer has relatively little effect on the thermite reaction, but for the samples with higher terpolymer content, more nanothermite components react with terpolymer at an early stage. Ignition and combustion performance show terpolymer can reduce ignition current threshold by up to 9.82% and increase combustion duration time at least several times. The potential reasons for the above results are analyzed. This work can shed light on the application of fluoroelastomer in energetic‐materials.

necessary to control the energy release rate, duration time of combustion and energy density. In nature, fluorine is the most electronegative element in the periodic table of elements with the electronegativity of 3.98 on Pauling scale. 12,13 Thus, recently, the combination of nanothermites and fluorine polymers has aroused wide concern from researchers, which could help to control and enhance the thermal reaction and long-term storage stability. [14][15][16] Fluorine polymers have been already widely used in many applications, such as membrane, tubing, fibers, filler, and coating materials, which contain fluorine atoms in the molecular chains. [17][18][19] The most famous fluorine polymer is polytetrafluoroethylene (PTFE) because of the highest fluorine content, about 76%. 20 Tang and his co-author had paid much attention to the mechanical properties and impact-induced characteristics of PTFE/Al/CuO reactive materials. 21,22 The flame and energy release rate greatly improved due to the addition of Al/CuO thermite. Recently, comparative research reported about three kinds of PTFE-based thermite materials, PTFE/Al/Fe 2 O 3 , PTFE/Al/MnO 2 , and PTFE/Al/MoO 3 . 23 The results showed that the formula of PTFE/Al/MnO 2 released the most energy during compression, and the detailed thermal reaction processes of PTFE/Al/MnO 2 were also reported. 24 But the components were mainly at the micron-scale with slow energy release. In fact, because of the poor solubility and agglomeration of PTFE, it is hard to obtain uniform well-mixed samples, especially for nanothermites, which is also the main drawback for further application in the energetic materials field. 25 Thus, poly(vinylidene fluoride-ter-hexafluoropropylene-ter-tetrafluoroethylene) (poly(VDF-ter-HFP-ter-TFE)) terpolymer with both high fluorine content and good solubility can be used as a substitute for PTFE in nanoenergetic materials application. 25 However, few researches focus on the application of poly(VDF-ter-HFP-ter-TFE) on nanothermites.
Therefore, in this article, poly(VDF-ter-HFP-ter-TFE) terpolymer is selected as an energetic additive and binder, and the Al/MnO 2 /poly(VDF-ter-HFP-ter-TFE) nanothermite samples are prepared by using electrospray method. For energetic additive, the addition of terpolymer can help to regulate the rate of energy release; for binder, the role of terpolymer can significantly improve the integrity for further industrial application. Then, the morphologies, thermal properties and combustion performance are also investigated. Furthermore, the mechanism of terpolymer on nanothermite system is also discussed and analyzed, which will provide practical guidance for fluorine-containing nanothermite research and application.

Materials
All chemicals can be used directly without any further purification any more. Al nanoparticles (Al NPs) (∼100 nm diameter) are purchased from Shanghai Nai-ou Nano Technology Co., LTD, and the thickness of the alumina shell on the surface of Al NPs is about 5 nm, so the active aluminum content is about 64.5%. 26 As for the detailed processes, please see the References 27,28. Poly(VDF-ter-HFP-ter-TFE) terpolymer is selected from Shanghai 3F New Chemical Materials Co., LTD, with about 67.5% fluorine content. As for solvents and dispersants, ethyl alcohol, deionized water, and dimethylformamide (DMF) are purchased from Nanjing Chemical Reagent Co., LTD.

Sample preparation
The ratio of each component is listed in Table 1. In a typical preparation experiment, the mixture of Al NPs and MnO 2 NRs is dispersed in ethyl alcohol (about 5 ml) by using ultrasonic method. The oxide of Al NPs is considered for the design of the ratio of Al NPs and MnO 2 NRs. In the meanwhile, the corresponding poly(VDF-ter-HFP-ter-TFE) terpolymer component is dissolved by DMF (about 3 ml). Then, the terpolymer -DMF solution is poured into the mixture of Al NPs and MnO 2 NRs as the precursor (about 8 ml) for further electrospray method. The diagram of electrospray experiment is shown in Figure 1. The precursor is put into the syringe with a 0.43 mm flat nozzle. The distance between the end of the

Characteristics analysis
The structures and morphologies of samples prepared from electrospray method are observed by using field emission scanning electron microscopy (FE-SEM) (HITACHI High Technologies Corporation, S-4800II Japan). The secondary electron images are captured at 15 kV accelerating voltage. As for thermal analysis, thermogravimetric-differential scanning calorimetry (TG-DSC) (NETZSCH STA 449C Germany) is introduced to study the thermal properties of samples. The heating rate is 15 • C min −1 from room temperature to 800 • C at argon atmosphere. The mass of sample in an 80 μL corundum crucible is about 3 mg. Finally, the x-ray diffraction (XRD) (Bruker, D8 Advance, Germany) analysis is used to figure out the phase in residues after thermal analysis.

Onset combustion test
To verify and record the characteristics of samples combustion, the heating wire ignition combustion system is designed and made, as shown in Figure 2. The diameter of the heating wire is about 0.1 mm. The DC is employed as the power supply. Ignition and combustion processes are recorded by the high-speed camera (FASTCAM SA-Z Japan) with a shutter speed of 50 us per picture. In each ignition and combustion test, the mass weight of sample is about 10 mg (±1 mg) directly touching the heating wire. Then, the current value on the heating wire increases slowly and gradually. At the moment of ignition, the corresponding current value is recorded, which is set as the ignition current threshold, and in the meanwhile the shutter of the high-speed camera is also pressed. A time period of 200 ms before the shutter time can be also recorded which ensures that we can fully document the process from the ignition moment to its extinction. To protect the camera from the burning sparks, a transparent tempered glass is placed between the samples and the camera. Figure 3 represents the SEM images of Al/MnO 2 and Al/MnO 2 / terpolymer nanothermite samples. Figure 3A is the Al/MnO 2 nanothermite without the addition of terpolymer. There are two main structures, nanoparticles and nanorods. The nanoparticles represent the fuel, Al NPs, while the nanorods are the oxidizer, MnO 2 NRs. The diameter of Al NPs is in the range of 50-100 nm. The length of rods is about several micrometers and the diameter of the rods is merely about 20-50 nm. The Al NPs can directly contact the oxidizer. Besides, the agglomeration phenomenon is not very clear. Figure 3B-D are the Al/MnO 2 /terpolymer nanothermite with different contents of P(VDF-ter-HFP-ter-TFE) terpolymer, from 10 to 30 wt% respectively. In Figure 3B, with 10 wt% terpolymer addition, some terpolymer binder can be found among the nanothermite components. In fact, the existence of terpolymer is not pretty obvious due to its content in nanothermite system is not high. So, a fair amount of fuel still can contact the direct oxidizer. When the mass fraction of terpolymer reach 20 and 30 wt%, the existence of terpolymer can be more easily found in Figure 3C,D. As a kind of binder, terpolymer mainly exists among the components, which can bond the nanothermite components together to improve the integrity of the materials.

Simultaneous thermal analysis
We carried out TG-DSC simultaneous thermal analysis tests for Al/MnO 2 and Al/MnO 2 /terpolymer nanothermite samples, as shown in Figure 3. The red curves mean the DSC curves while the black curves represent the TG curves within the temperature range from 100 to 800 • C. Figure 4A illustrates that the TG-DSC results of Al/MnO 2 nanothermite samples. Below 350 • C, the total mass continuously loses with the rise of temperature, but there is no obvious endothermic or exothermic DSC signal, which is caused by the evaporation of residual solvents, such as free and structural water, ethyl alcohol. 29 The mass is reduced by about 3.0%. Then, in the range of 500-600 • C, the main exothermic DSC signal appears, indicating that the thermite exothermic reaction between Al NPs and MnO 2 NRs occurs. At the same time, no obvious mass changes from TG curve at such a temperature range. During thermite reaction, the O element transfers from oxidizer, MnO 2 NRs, to Al NPs. Thus, no mass change happens. The peak temperature of thermite reaction of Al/MnO 2 nanothermite sample is about 554 • C with the 1070 Jg −1 heat release.  Figure 4B shows the TG-DSC results of Al/MnO 2 /10 wt%-P(VDF-ter-HFP-ter-TFE) terpolymer nanothermite sample. At about 472 • C, the minimum total mass appears, about 89.1%. Considering the evaporation of the solvent, more than 80% of terpolymer decomposes to the gaseous products, leading to mass loss. However, D'Orazio and his co-authors 30 have reported the thermal decomposition property of P(VDF-ter-HFP-ter-TFE) terpolymer, and the results show that all terpolymer will convert the gaseous products before 500 • C. Meanwhile, combined with previous reports, 24,30 a few parts of terpolymer react with nanothermite components directly, which can explain those exothermic DSC signals (signal b-1, b-2, and b-3), implying reactions among Al NPs, MnO 2 NRs, terpolymer degradation products and/or the terpolymer matrix. Then, as the temperature increases gradually, the exothermic DSC signal (b-4) occurs, indicating the thermite reaction between remaining Al NPs and MnO 2 NRs, similar to the exothermic signal in Figure 4A. By contrast, the peak temperature point (b-4) is delayed by 26 • C since some sections of thermite components react with P(VDF-ter-HFP-ter-TFE) terpolymer before.
Next, the content of P(VDF-ter-HFP-ter-TFE) terpolymer reaches 20 wt% in Figure 4C. At about 483 • C, the minimum total mass appears, about 86.2%, implying that there is a 13.8% mass loss. Similarly, it is caused by the evaporation of the solvent and thermal decomposition of terpolymer. It is inferred that near 50% of terpolymer directly reacts with the Al NPs and MnO 2 NRs rather than decomposition with gaseous products release, leading to the three exothermic DSC signals (c-1, c-2, and c-3) accordingly. 24,30,31 Compared to Figure 4B, those three exothermic DSC signals (c-1, c-2, and c-3) are bigger and more obvious, especially signal c-2. However, the more proportion of terpolymer reacts with nanothermite components, the less amount of nanothermite components remains for thermite reaction, which results in the corresponding smaller exothermic DSC signal (c-4) for thermite reaction with less heat release.
Finally, the formulation with 30 wt% P(VDF-ter-HFP-ter-TFE) terpolymer content is shown in Figure 4D. At about 491 • C, the minimum total mass appears, about 83.9%. Same to the above reason for mass loss, after estimating, more than 50% of terpolymer directly reacts with Al NPs and/or MnO 2 NRs, which leads to the big exothermic DSC signal (d-2). Similarly, much proportion terpolymer reaction with nanothermite components will affect the further thermite reaction and the corresponding exothermic DSC signal (d-4).
Besides, from Figure 4B-D, it can be found that not only the terpolymer itself will influence the thermal reaction process of Al/MnO 2 nanothermite system, but the mass fraction of P(VDF-ter-HFP-ter-TFE) terpolymer also affect the thermal properties and processes. Specifically, from the above TG-DSC results, as the content of terpolymer is no more than 10 wt%, the main exothermic reaction is still thermite reaction between Al NPs and MnO 2 NRs. In contrast, if the mass fraction of terpolymer is more than 20 wt%, the thermite reaction will be greatly decreased and the main exothermic signal will be replaced by the reaction between terpolymer and nanothermite components.

Residues XRD analysis
From TG-DSC results, it can be found that the addition and content of P(VDF-ter-HFP-ter-TFE) terpolymer can significantly change the thermal processes with the rise of temperature. To further understand the reaction products from different nanothermite samples, we collected the residues in crucibles after TG-DSC tests, and used XRD analysis to study the phase of residues, as shown in Figure 5. Without the addition of terpolymer, the reaction products are mainly Mn 3 O 4 , MnO and Al 2 O 3 caused by thermite reaction between Al NPs and MnO 2 NRs (green curve in Figure 5, AMT-0-residues). Then, the blue curve presents the XRD results of Al/MnO 2 /10 wt%-terpolymer nanothermite reaction products. The reaction products have mainly changed to galaxite (MnAl 2 O 4 ) with some sections of MnO and aluminum manganese (Al 2 Mn 3 ). The red and black curves show the residues of Al/MnO 2 /20 wt%-terpolymer and Al/MnO 2 /30 wt%-terpolymer nanothermite samples, respectively, and their reaction products are almost the same, containing Al 2 O 3 , manganese carbide (Mn 7 C 3 ), Al 2 Mn 3 , MnO, AlF 3 , C and possible remaining Al, which are greatly different from Al/MnO 2 /10 wt%-terpolymer nanothermite sample as well as Al/MnO 2 nanothermite sample. During the thermal reaction, both Al and MnO 2 can react with fluoropolymer, which could greatly break the equilibrium of original chemical reactions, leading to a little bit of possible remaining Al. We guess that the different content of P(VDF-ter-HFP-ter-TFE) terpolymer influences the detailed thermal processes during the TG-DSC tests, and then the different thermal processes lead to the different reaction products finally.

F I G U R E 5 XRD results of residues after TG-DSC tests
The different content of terpolymer can distinctly change the reaction products. More than 80% terpolymer decomposes before the thermite reaction. Namely, a few terpolymer and/or terpolymer degradation products involves into thermite reaction. So, the phases of reaction products are mainly MnAl 2 O 4 , MnO and Al 2 Mn 3 rather than AlF 3 . In contrast, when the content of terpolymer increases, more and more terpolymers react with components, leading to the AlF 3 occurrence from residues analysis.

Ignition and combustion
Ignition and combustion performance represent the practical characteristics for energetic materials future applications.
In this work, we carried out the ignition and combustion by using fast heating wire test as well as a high-speed camera. The detailed ignition and combustion performance for each sample is shown in Figure 6. The first picture of firelight is defined as the beginning time, set as 0 us. The DC power supplier is selected, and the current value for ignition moment is recorded as the ignition current threshold. In Figure 6A, when the current rise to 1.192 A, the Al/MnO 2 nanothermite sample without any additive successfully ignites. The speed of flame growth is very rapid, and at about 350 us, the size of flame reaches the maximum. Then, the flame fades away gradually with clear flying sparks (Supplementary movie 1).
In Figure 6B, the sample with 10 wt%-P(VDF-ter-HFP-ter-TFE) terpolymer additive ignites at about 1.137 A current value. The speed of flame growth is still relatively fast, and at about 1 ms, the size of flame reaches the maximum. However, compared to the Al/MnO 2 nanothermite sample in Figure 6A, the speed of flame growth and the size of maximum flame are decreased clearly. The addition of terpolymer can help reduce the number of flying sparks to concentrate the flame. Besides, the duration of combustion is longer than that of Al/MnO 2 nanothermite sample (Supplementary movie 2).
Furthermore, when the contents of terpolymer reach 20 and 30 wt%, the current value at ignition moment are 1.083 and 1.075 A, respectively. As for Al/MnO 2 /20 wt%-terpolymer sample, in Figure 6C (Supplementary movie 3) the flame grows pretty slow, and the maximum flame appears at about 7 ms. After the maximum flame moment, the flame seems not to fade away gradually but further forms a certain burning flame shape, which might be caused by the unevenly distributed terpolymer, leading to the self-sustaining combustion phenomenon. 32,33 Such phenomenon is also found in the last sample, in Figure 6D (Supplementary movie 4). After the maximum flame moment at about 8-9 ms, a small but clear bulge appears at the top of the flame.
When the contents of terpolymer are below or equal to 10 wt%, the phenomenon of self-sustaining combustion is not obvious, and the shapes of flame are a mainly concentrated sphere. In contrast, once the terpolymer content is over 20 wt%, the self-sustaining combustion phenomenon and different shapes of flame can be captured. Besides, the addition of P(VDF-ter-HFP-ter-TFE) terpolymer can reduce the ignition current threshold from the current value at the ignition moment, but it has a great negative influence on flame growth, leading to the long combustion duration time.
To further understand the difference between ignition and combustion phenomenon, the potential reasons are analyzed and discussed combined with the above morphology, thermal process and residues analysis, as shown in Figure 7. The red arrow presents the direction of flame growth and spread, and its length and width collectively mean the speed of flame growth and spread. From the above results of TG-DSC test, the thermite reaction occurs before the melting point of Al NPs, implying that thermite reaction belongs to solid-phase reaction.
As for Al/MnO 2 nanothermite ( Figure 7A), without any addition of terpolymer, the ignition point should be near the heating wire and between the Al NPs and MnO 2 NRs, and then theoretically the flame travels equally fast in each direction.
In Figure 7B, the 10 wt% content of P(VDF-ter-HFP-ter-TFE) terpolymer exists in the gaps among components, decreasing the heat spread and direct contact of nanothermite components. It can be the reason for the delay of flame growth. But the content of terpolymer is not high, so the negative effect on flame growth is not terrible, either. Each coin has two sides. Based on the ignition test, the addition of terpolymer can reduce the ignition current threshold from the current value at the ignition moment. The terpolymer is very sensitive to the temperature increment. On one hand, high temperature can lead to the process of thermal decomposition with heat release. On the other hand, the decomposition of terpolymer can release the fluoride. The fluoride can directly react with nanothermite components at an early stage. Besides, since fluorine is the most electronegative element, it can deal with the alumina shell covered on the Al NPs, leading to the increasing active aluminum directly reacting with oxidizer. Figure 7C,D shows the combustion diagram of Al/MnO 2 /20 wt%-terpolymer and Al/MnO 2 /30 wt%-terpolymer nanothermite, respectively. Comparatively, the negative effect on flame growth becomes increasingly great with the increasing content of terpolymer. Terpolymers even totally cover the nanothermite components in some parts, which leads to the clear self-sustaining combustion phenomenon. At the same time, much more heat will release at an early stage due to the more content of terpolymer decomposition as well as reaction with components from TG-DSC results, resulting in a lower ignition current threshold accordingly.

CONCLUSION
In this article, the investigation of P(VDF-ter-HFP-ter-TFE) terpolymer additive on Al/MnO 2 nanothermite system was presented, and the effect of terpolymer mass fraction on energetic materials was also explained. The terpolymer can stick nanothermite components together to enhance the integrality of materials. The thermal properties can be significantly influenced by the contents of P(VDF-ter-HFP-ter-TFE) terpolymer. Without any terpolymer additive, the thermite reaction between Al and MnO 2 occurred at about 554 • C. For the 10 wt% mass fraction of terpolymer, three exothermic signals appeared before the main thermite reaction, indicating the thermal reactions among Al nanoparticles, MnO 2 nanorods, terpolymer degradation products and/or the terpolymer matrix. However, those three thermal reactions became the main exothermic reaction instead of the thermite reaction when the content of terpolymer was over 20 wt%, and the residues analysis verified the difference. The main reaction products of Al/MnO 2 nanothermite were Mn 3 O 4 , MnO, and Al 2 O 3 . For Al/MnO 2 /10 wt%-terpolymer, the residues were MnAl 2 O 3 and Al 2 Mn 3 . When the content of terpolymer was more than 20 wt%, such Mn 7 C 3 and AlF 3 were found in the residues. The ignition and combustion processes were also different accordingly. The terpolymer could reduce the ignition current threshold but also decrease the speed of flame growth. The terpolymer could hinder the direct contact of nanothermite components and influence the flame spread. For ignition current threshold and duration time of self-sustaining combustion, we suggest that the 20 wt% or 30 wt% addition is the best. For the reaction rate, the 0 and 10 wt% additions are better. This work revealed the thermal and combustion performance of Al/MnO 2 /P(VDF-ter-HFP-ter-TFE) terpolymer nanothermites, which could provide a reference for application of fluoroelastomer in energetic materials.