3.1 Characterizations of FGO with different fluorine content
The results of elements content of fluorinated graphene were shown in Fig. 3. It can be seen that with the increase of fluorinating agent, the C/O and the C/F ratio both decreases.
The FTIR spectrum of fluorinated graphene was shown in Fig. 4. There was an obvious C-F bond stretching vibration peak (υC−F) around 1217 cm− 1[34],the C-C bond stretching vibration peak (υC−C) of carbon skeleton around 1050 cm− 1 and the hydroxyl bending vibration peak (βO−H) about 1415 cm− 1 were obvious, but the hydroxyl stretching vibration peak (υO−H) about 3440 cm− 1, the carbon-oxygen double bond stretching vibration peak (υC=O) in carboxyl group or ester group around 1718 cm− 1 and the epoxy group bond stretching vibration peak (υC−O−C) near 1220 cm− 1 were almost invisible [35–37]. For the three samples of FGO, the peak intensity of the C-F bond stretching vibration were 50.24%, 47.04% and 24.26% respectively, which proves that the degree of fluorination gradually deepens with the increase of the amount of fluorinating agent.
By fitting the C1s peak of XPS spectrum (see supplementary information S3), the types of functional groups connected to the carbon skeleton and the proportion of each functional group can be analyzed. The binding energy of non covalent C-F bond was 288.9 eV, covalent C-F bond was 290.3 eV and covalent C-F2 was 291.2 eV [38], The binding energies of sp2 C, sp3 C, C-OH, C-O-C, C = O and HO-C = O were 284.5 eV, 285.3 eV, 286.3 eV, 287.1 eV, 288.2 eV and 289.4 eV respectively [39, 40]. The content of each functional group was shown in Fig. 5. It can be seen that after fluorination, the content of oxygen-containing functional groups was gradually decreasing, and the content of fluorine-containing functional groups was increasing instead, especially C-F2 bond. The increase of F content makes more and more C atoms hang two F atoms at the same time.
3.2 Thermal decomposition of FGO under argon atmosphere
FGO contains a large number of fluorine-containing functional groups, and the C-F bond was relatively stable, but under sufficient thermal stimulation, it will produce fluorocarbon fragments. The higher the fluorine content, the higher the gas-phase product volume Vg and pressure P. The fluorine-containing gas had a certain oxidizing ability, and could promote oxidation. FGO-3 with the highest fluorine content was selected as the research object. As shown in Fig. 6, GO began to decompose at about 200°C[8], however, FGO-3 began to decompose at 400°C. Combined with the mass spectrum in Fig. 7, FGO-3 had severe weight loss between 400 ℃ and 900 ℃, which was divided into two stages. At 481 ℃ ~ 661 ℃ stage, the weight loss was about 76.83%, and the heat release was 4668 J·g− 1, mainly producing fragments of F ions, HF, CxFy, CO, and CO2 compounds. The second stage from 661 ℃ to 880 ℃ was relatively gentle, the weight loss was about 14.00%, and the heat release was 2628 J·g− 1, and a small amount of C2F, F, CO2 compound fragments produced. After reaction, FGO-3 only a small amount of solids remained. Compared with GO, FGO-3 was more stable, the decomposition temperature was higher, the heat generated during decomposition was much higher than that of GO, and in addition to carbon oxides, more fluorocarbons were generated during decomposition.
3.3 Thermal decomposition of B/KNO3/FGO
The thermal analysis curves of the B/KNO3 with different contents of FGO-3 were shown in Fig. 8. After adding FGO-3, the DSC curves were similar to B/KNO3 (S0), with two endothermic peaks and one two-step exothermic peak. While adding FGO-3, the first step of the exothermic peak temperature was delayed and the second step peak temperature was advanced. With the increase of FGO-3 content, above-mentioned phenomenon was more obvious. By integrating the area of curve peak, the heat release of samples S0 and S1 ~ S3 was 4470 J·g− 1, 4496 J·g− 1, 4544 J·g− 1, 4848 J·g− 1 respectively, indicating that the addition of fluorinated graphene will increase the heat release of composite energy materials, and with the increase of the content of fluorinated graphene, the heat release became larger.
Analyzing the mass spectrometry data of the sample S3 under air atmosphere with heating rate of 5 ℃·min− 1, as shown in Fig. 9. the first stage mainly produced BO2−, NO2, NO−, N2O, C, F−, BF2+, HBO2, H3O+ ion fragments and fluorocarbon fragments. the second stage mainly produced NO2, NO−, N2O, BO2− ion fragments. Combined with Fig. 8 and Fig. 9, it was deduced that delay of the first step exothermic peak temperature of B/KNO3/FGO was arisen from FGO need absorb some energy before decomposition. The degree of postponement of 1st decomposition temperature was proportional to the content of fluorinated graphene. This meant that B/KNO3/FGO showed higher safety than B/KNO3 under the stimulation of low external energy. The reason for the advance of the 2nd exothermic peak temperature is that FGO begins to decompose above 400 ℃, and generated fluorocarbon fragments can promote the reaction. Similarly, with the deepening of the degree of fluorination, the promotion of the 2nd decomposition peak became more obviously. It was indicated that when the external energy stimulation was large enough, the reaction temperature of B/KNO3 advanced and the heat release increased. In conclude, FGO was added into B/KNO3 to make the whole materials had “intelligent” response behavior to different thermal stimulus modes.
3.4 Moisture resistance
The contact angle was used to characterize the ability of the liquid to wet the solid surface and was related to the surface tension of liquid-solid contact surface[41]. The contact angles of the liquid drop on the surface of the solid sample can be directly observed. Pressed samples into discs, and then placed them on the slide, droped deionized water on the discs to observe the change of the contact angle. The contact angles of B/KNO3/FGO versus water were shown in Fig. 10. The contact angles were 103.65°, 106.25°, and 111.75° respectively. The contact angle becomes larger with the increase of FGO content. It shows that the addition of hydrophobic fluorinated graphene into energetic materials had a significant effect on surface properties. FGO contained hydrophobic groups C-F, and hybrid it into the composites will reduce its hygroscopicity. It can be seen that the addition of hydrophobic FGO can improve the hygroscopic property of the B/KNO3, which was beneficial to maintain the dryness during storage.
In order to test the moisture resistance ability, the hygroscopicity of S0 ~ S3 samples were measured. The empty bottle was weighed first. Bottles containing 3 ± 0.0005 g samples (two samples of S0, S1, S2, S3, with the bottle caps all open) were placed in an oven at 60 ± 2°C for 2 h, took it out and sealed in desiccator I to cool to room temperature and weighed, recorded the mass of the sample bottles as m1, and calculated the samples mass m. The relative humidity of the potassium nitrate saturated solution at different temperatures was different, and it was necessary to ensure that the temperature was constant during the moisture absorption process. Then, placed the desiccator II in an oven, preheated at 30 ± 1°C for 8 h, placed the empty weighing bottle and sample bottles in the desiccator II, opened the weighing bottles, kept them warm for 24 h, after the bottle caps were closed, them were placed in the dryer I for cooling and weighing. Repeated these steps several times until the mass difference of weighing bottle for two consecutive times was less than 0.0002 g. The mass of the weighing bottle with the samples was m2, and the increment of the empty weighing bottle was m3. The incremental score was calculated by formula (1):
In the formula, W was the mass increment fraction of the sample after moisture absorption, %. m2 was the mass of the sample and bottle after moisture absorption, g. m1 was the mass of the sample and bottle before moisture absorption, g. m3 was the weight gain of empty bottle after moisture absorption, g. m was the mass of the sample, g.
The results of hygroscopicity were shown in Table 2. The hygroscopic increment of B/KNO3 was 9.54%. After adding fluorinated graphene, the hygroscopic increment decreased. It was proved that hydrophobic FGO was benefit to improve the hygroscopicity of B/KNO3.
Table 2
Hygroscopic increment of energetic composites
sample number | m1/g | m/g | m2/g | W/% | /% |
S0 | 38.6020 | 2.9993 | 38.8884 | 9.5422 | 9.54 |
39.0721 | 2.9995 | 39.3586 | 9.5449 |
S1 | 39.1896 | 2.9996 | 39.4490 | 8.6412 | 8.64 |
38.7049 | 2.9991 | 38.9643 | 8.6426 |
S2 | 39.2130 | 2.9997 | 39.3958 | 6.0873 | 6.09 |
39.4472 | 2.9998 | 39.6301 | 6.0904 |
S3 | 39.8514 | 2.9997 | 40.0154 | 5.4605 | 5.46 |
38.9778 | 2.9996 | 39.1419 | 5.4641 |
3.5 Thermal safety
The response of energetic materials to thermal stimuli was closely related to their thermal conductivity. Heat transferred from hot spot to adjacent, raising their temperature to ignition temperature. An excellent thermal conductivity was essential for stable propagation of combustion. If specific heat capacity and density of the sample were known, the thermal conductivity at corresponding temperature can be obtained through formula (2)[42].
In there, λ was thermal conductivity, W·m− 1·K− 1. α was thermal diffusivity, m2·s− 1. cp was specific heat capacity, J·kg− 1·K− 1. ρ was density, kg·m− 3. The thermal diffusivity can be measured by Laser Flash method. The thermal diffusivity was calculated according to the Fourier[43, 44] heat transfer equation as Eq. (3).
In the formula, α was thermal diffusivity, L was thickness of the sample, and t1/2 was half of heating time.
The samples were pressed into discs(Φ12.7 mm) with a thickness of more than 2 mm. Thermal diffusivity of the samples at 25°C were measured by a flash laser thermal conductivity meter, and the average value was obtained after multiple measurements. The specific heat capacity at the same temperature was measured by differential scanning calorimeter (TA DSC 25). So we got the thermal conductivity λ by formula (2). The results were shown in Table 3.
Table 3
Thermal conductivity of energetic composites
sample number | ρ/(kg·m− 3) | cp/(J·kg− 1·K− 1) | α/(×10− 7 m2·s− 1) | λ/(W·m− 1·K− 1) |
S0 | 1873 | 841.0 | 2.59 | 0.408 |
S1 | 1837 | 933.4 | 3.19 | 0.547 |
S2 | 1817 | 1108.8 | 2.98 | 0.600 |
S3 | 1798 | 1242.3 | 2.73 | 0.610 |
The specific heat capacity cp can measure the ability of materials to absorb and dissipate heat[45]. The specific heat capacity of B/KNO3 addition with FGO was improved, indicating that the ability of energetic composite to transfer heat to surrounding was improved. For materials with high thermal diffusivity, the internal temperature was easier to equalize[46]. The high thermal diffusivity made energetic materials difficult to form hot spots. On the contrary, it was conductive to improving thermal safety. The thermal diffusivity α increases after adding FGO, thus improving thermal safety.
The flame sensitivity reflected the ignition probability of energetic materials under the action of flame. The flame sensitivity was evaluated by critical ignition distance H50, which was defined as the distance between samples and flame when the ignition probability was 50% ignition probability of the sample. The lower the H50 value was, the higher the thermal safety was [47]. As shown in Fig. 11, the test was carried out in a protective box with a vertically movable tray. The samples were placed on the tray and kept a readable distance from the ignition source of the test device. When the ignition source ignites, the samples were exposed to hot temperature field.
In order to study the change law of thermal safety of B/KNO3 after adding FGO, the critical ignition distance H50 of samples were measured by HGY-1 flame sensitivity meter. 20.0 ± 5.0 mg composites were weighed and pressed into test samples at 58.8 MPa. Placed the test samples on the vertically movable tray, and read the distance between the test samples and the center of the flame through the ruler. The top ignition device ignited black powder column to burn, observed the combusting condition of the sample and recorded, then obtained the critical ignition distance H50 and standard error.
The results of the critical ignition distance were shown in Fig. 12. The critical ignition distance H50 of the B/KNO3 was 39.4 ± 0.5 cm. Adding FGO, H50 all decreased, and with increase of FGO content, the H50 decreased obviously. Due to the low thermal conductivity of B/KNO3, the part of column close to flame was in high temperature filed. The ability to dissipate heat to the environment was weak, and it was easy to form a hot spot and be ignited at a small energy. While adding FGO, heat dissipation was improved, so heat was easily conducted to adjacent area. The bulk temperature was lower than ignition point, B/KNO3/FGO was less likely to burn.
It was deduced that thermal conductivity of the composite energetic materials had an important impact on the flame sensitivity[48]. In Table 3, addition of FGO enhanced the thermal conductivity of B/KNO3, which could reduce the flame sensitivity, and improve the thermal safety under lower energy stimulation.
3.6 Energy output
The pressure-time curve (P-t curve) was a common method to characterize the output performance of energetic materials[49]. The experiment was carried out using a closed bomb as shown in Fig. 13. The vessel’s volume was 330 mL. A metal wire was buried in the samples to ignite, and a pressure sensor was installed on the side to record the pressure change.
The pressure-time curves of the B/KNO3 were shown in Fig. 14(a). After the samples ignited, pressure value raised sharply. The pressure rise time was 0.0537s. As the temperature in the vessel declined, the pressure value gradually decreased, It can be seen from Fig. 14(b) that the pressure rise times of S1 ~ S3 samples were 0.0422s, 0.0485s, and 0.0387s respectively. With the increment of FGO content, the pressure peak increased.
As shown in the Table 4, we calculated the peak pressure rise rate. The peak pressure raised more rapidly after adding FGO,. This was because the oxidation reaction of FGO can also generate a certain amount of gas, thus enhancing the total exothermic effect of B/KNO3 system and accelerating the reaction speed.
Table 4
data of pressure-time experiment
sample number | Pmax/KPa | percent change in Pmax/% | (dP⁄dt)max/(KPa·s− 1) |
S0 | 468.08 | — — | 8716 |
S1 | 515.09 | 110.04 | 12218 |
S2 | 533.45 | 113.97 | 11050 |
S3 | 557.89 | 119.19 | 14397 |
Combustion heat was an important parameter to measure the energy released by energetic materials. The results were shown in Table 5. The introduction of a small amount of FGO would increase the combustion heat value. This was because the decomposition of FGO can produce fluorine ions and fluorocarbons with high activity, which can promote the reaction. At the same time, the heat released from the generation of boron fluoride was more than that from the generation of boron oxide[50].
Table 5
combustion heat value of B/KNO3 systems
sample number | S0 | S1 | S2 | S3 |
heat of combustion / (J·g− 1) | 8258 | 8317 | 8326 | 8395 |
3.7 Combustion performance
The B/KNO3 was a typical laser ignition agent. The laser experiment was fast and efficiently, and the reaction activity was judged by observing the ignition delay. The sample was ignited by optical fiber, which was inserted into a tube whose size was ϕ 5 mm×100 mm charged with 1.750 ± 0.050 g. High speed photography was used to record the entire combustion process[51], layout was shown in Fig. 15.
Combustion time was recorded by high speed camera and linear combustion velocity was calculated by formula (4). In there, the length of charge was h and the burning time was t. Ignition delay represented the time difference between the laser signal and combustion luminous signal.
The combustion process of four samples were shown in Fig. 16. Trigger laser times as starting point. The ignition delay time of S0 ~ S3 can be calculated to be 46 ms, 39 ms, 34 ms, and 19 ms respectively. Ignition delay time shortened with the increase of FGO content. Combustion time of S0 ~ S3 are 428 ms, 1064 ms, 1421 ms, and 2653 ms respectively. The combustion time of B/KNO3 was short and the reaction was violent. Because B/KNO3 reaction would produce solid boron oxides, accompanied by splashing during combustion. The samples containing FGO had longer combustion time, and produced more spatters.
The combustion time and linear burning velocity of the energetic materials were obtained by formula (4), as shown in Fig. 17. The linear combustion velocity of B/KNO3 was 0.233 m·s− 1, and the linear combustion velocity decreased sharply after adding FGO. FGO would improve the thermal conductivity. Heat was easy to spread into environment, which will be quickly to reach the combustion temperature. The expansion of gas products took away heat, which made the temperature of S1 ~ S3 samples lower than that of S0 sample, but the reaction time was prolonged, and combustion trends to self-sustaining. The combustion speed reduced, so that the generated heat has enough time to be transferred to unfired area. At the same time, the flame lasted for longer time, which was conducive to ignite.
3.8 “Intelligent” response to thermal stimulus
Excellent energetic materials needed to meet the requirements of high energy and high safety simultaneous. However high energy and high safety were contrary. It was particularly important to improve thermal safety of materials as much as possible on the premise of maintaining high energy. Energetic materials were well known as high-risk chemical products, it was very important to ensure safe disposal likely transportation, storage and maintenance etc. For this reason, the upper and lower limits of the sensitivity of energetic materials were proposed. The upper limit of sensitivity indicated that fire probability of energetic materials was 100% if stimulating energy was higher than this value, which can be used for product reliability design.The lower limit of sensitivity indicated that fire probability of energetic materials is 0% if stimulating energy was lower than this value, which was used to ensure safe handling. Therefore, reduced the range between the upper and lower sensitivity limits, it meant that we narrow the uncertainty space and have more control. As shown in Fig. 18, B/KNO3 adding with FGO-3 showed higher thermal safety under low-energy thermal stimulis, and faster and higher energy release under high-energy thermal stimulis, which reflected "smart" response of FGO to thermal input. After FGO-3 added to B/KNO3, the uncertain ignition probability range was compressed, which had a wider space for safe disposal in practical application.