Introduction

The development of energetic materials is of great significance from the military to civil aspects [1, 2]. As the energy component of explosive and propellant, the performances of the energetic compounds have great influence on the operational efficiency and survivability of missile weapons [3,4,5,6]. Therefore, it is of great interest to synthesize energetic materials with both high energy and low sensitivity [7, 8]. The synthesized 5, 5′-bistetrazole-1, 1′-diolate (TKX-50) has attracted extensive attention not only owing to its high theoretical density (1.918 g cm−3), detonation velocity (9679 M s−1) and detonation pressure (42.4 MPa), but also to its low friction and impact sensitivity even lower than that of the traditional explosives octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine (HMX) and hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) [9,10,11,12,13]. TKX-50 used in solid propellant can not only improve the energetic performance, but also avoid the potential risks caused by the introduction of highly sensitive energetic compounds [9].

Additionally, TKX-50 burns faster than HMX and approaches the burning rate of 2,4,6,8,10,12-hexanitro-2,4,6,8,10,12-hexaazaisowurtzitane (CL-20), which owns excellent application prospect in the field of solid propellants [10]. The thermal decomposition property of energetic component has a critical influence on the combustion performance of solid propellant [11,12,13]. Differential scanning calorimetry (DSC) and thermogravimetric (TG) methods are usually used methods for the investigation of thermal decomposition properties of energetic materials [14]. Besides, the kinetic parameters including apparent activation energies and Arrhenius parameters calculated using DSC or TG-DTG data play the important role on the understanding of thermal decomposition of energetic component [15]. Trache et al. [16]. studied the thermal decomposition performance and kinetic parameters using iso-conversional models, which better reflect the decomposition process than traditional liner methods.

The thermal decomposition performance of pristine TKX-50 was studied, and the kinetic parameters of TKX-50 were calculated using iso-conversional Friedman and Vyazovkin methods [17]. Additionally, the combustion catalyst plays an essential role in the thermal decomposition and combustion performance regulation of solid propellant although its tiny addition amounts [18]. However, most of the published studies involved only experimental and theoretical studies of pristine TKX-50 [17, 18]. It is vital to study the catalytic effects of combustion catalysts (transition metal oxide) on the thermal decomposition of TKX-50 [19,20,21,22]. In our preliminary study, the excellent catalytic activity of iron for thermal decomposition of energetic ionic salts including TKX-50 were illustrated [23, 24].

Therefore, three kinds of bimetallic iron oxides (NiFe2O4, ZnFe2O4 and CoFe2O4) and iron oxide were prepared and used for thermal decomposition of TKX-50 in the present work. The morphology, size and composition of the bimetallic iron oxides were characterized using SEM, TEM, XRD, FTIR, BET and XPS instruments. The catalytic action of different iron containing catalysts for TKX-50 thermal decomposition was characterized by means of DSC and TG-DTG. The kinetic parameters for thermal decomposition of TKX-50 were calculated using multi-kinetics methods including traditional, iso-conversional and iteration methods. Based on the above studies, the catalytic performance of various iron containing catalysts for TKX-50 thermal decomposition was analyzed (see Fig. 1).

Fig. 1
figure 1

Illustration of the MFe2O4 fabrication and the catalytic mechanism for TKX-50 thermal decomposition

Full size image

Experimental

Materials

All the chemicals used were of analytical grade and used without further purification. Ferric nitrate hydrate (Fe(NO3)3·9H2O), ferric trichloride hexahydrate (FeCl3·6H2O), nickel chloride hexahydrate (NiCl2·6H2O), zinc chloride (ZnCl2), cobalt chloride hexahydrate (CoCl2·6H2O), sodium acetate and polyethylene glycol (Mn = 4000) were purchased from Aladdin Inc. Ethylene glycol (Sinopharm Chemical Reagent Co., Ltd.), ethanol (EA) (Xi’an Chemical Reagent Factory) and distilled water were used for the sample preparation and treatment. Ammonia (FuYu Chemical Co., Ltd. of TianJin, 25%) was used to adjust pH value. TKX-50 with purity higher than 99.5% was obtained from Xi’an Modern Chemistry Research Institute.

Preparation of MFe2O4 (M = Ni, Zn and Co)

Certain amounts of FeCl3·6H2O (5 mmol) was dissolved and mixed with NiCl2·6H2O, ZnCl2 and CoCl2·6H2O (2.5 mmol) in 60 mL ethylene glycol under magnetic agitation, respectively. Then, sodium acetate (3.6 g) and polyethylene glycol (1.0 g) were added to the above three solutions and mixed homogeneous. The reactants were transformed into 100 mL Teflon-sealed autoclave and heated at 180 °C for 24 h. After cooling, the products were washed several times by distilled water and absolute alcohol and cured at 60 °C in a vacuum oven over night. The obtained powder was ground for further characterization. The yields of the NiFe2O4, CoFe2O4 and ZnFe2O4 are 76%, 77% and 79%, respectively. Besides, the preparation method of Fe2O3 refers to our previous published literature [24].

Characterization

The morphology and size of Fe2O3 and MFe2O4 (M = Ni, Zn and Co) were characterized by scanning electron microscope (SEM, Quanta600, Quantachrome, America) and transmission electron microscopy (TEM, Tecnai G2 F20, FEI, America). The structure and composition were analyzed using X-ray diffraction (XRD, Empyrean, PANalytical, Netherlands), Fourier transform infrared spectroscopy (FTIR, Tensor 27, Bruker, Germany) and X-ray photoelectron spectroscopy (XPS, NEXSA, Thermo scientific, Britain) instruments. XRD was collected with Cu Kα source in the measurement angle range of 2θ = 5°–90° with a scan rate of 8° min−1. The specific surface area of Fe2O3 and MFe2O4 (M = Ni, Zn and Co) samples was characterized using BET (JW-BK112, China). The catalytic action of Fe2O3 and MFe2O4 samples on thermal decomposition of TKX-50 (the mass ratio of the catalysts in the TKX-50-based composites is 1/11) were studied by differential scanning calorimeter (DSC, 200 F3, NETZSCH, Germany) with a heating rate of 5, 10, 15 and 20 °C min−1, respectively (N2 flow rate of 50 mL min−1, sample mass of 0.5 ± 0.2 mg). TG-DTG (Netzsch STA 449C with TASC 414/4 controller, Germany) with a heating rate of 10 °C min−1 (Ar flow rate of 50 mL min−1, sample mass of 1.0 ± 0.2 mg) was studied.

Kinetics calculation

The traditional Kissinger (Eq. 1), Ozawa (Eq. 2), Flynn–Wall–Ozawa (FWO, Eq. 3) [25] and nonlinear iso-conversional Kissinger (Eq. 4) and Ozawa (Eq. 5) iterative methods were employed to obtain the kinetic parameters of TKX-50, respectively [24].

$$\mathrm{I}\mathrm{n}\left(\frac{\beta }{{T}^{2}}\right)=\mathrm{I}\mathrm{n}\frac{AR}{E}-\frac{E}{RT}$$
(1)
$$Lg\left(\beta \right)=C-\frac{0.4567E}{RT}$$
(2)
$$\mathrm{lg}\beta =\mathrm{lg}\left(\frac{AE}{RF\left(\alpha \right)}\right)-2.315-0.4567\frac{E}{RT}$$
(3)
$$\mathrm{I}\mathrm{n}\left(\frac{\upbeta }{{Q}_{4}\left(u\right){T}^{2}}\right)=\mathrm{I}\mathrm{n}\left(\frac{AR}{EG(\alpha )}\right)-\frac{E}{RT}$$
(4)
$$\mathrm{I}\mathrm{n}\left(\frac{\upbeta }{H(u)}\right)=\mathrm{I}\mathrm{n}\frac{0.00484AE}{RG(\alpha )}-\frac{1.0516E}{RT}$$
(5)

Here, E is activation energy (J mol−1), T is temperature (K), β is heating rate (K min−1), R is universal gas constant (8.314 J mol−1 K−1), A is pre-exponential factor (s−1), and the functions in the formula (4) and (5) are expressed as Eqs. (6)–(10).

$${Q}_{4}\left(u\right)=\frac{{u}^{4}+18{u}^{3}+86{u}^{2}+96u}{{u}^{4}+20{u}^{3}+120{u}^{2}+240u+120}$$
(6)
$$H\left(u\right)=\frac{{e}^{-u}{Q}_{4}(u)/{u}^{2}}{0.00484{\mathrm{e}}^{-1.0516u}}$$
(7)
$$u=\frac{E}{RT}$$
(8)
$$G\left(\alpha \right)=\frac{AE}{\beta R}0.00484{\mathrm{e}}^{-1.0516{\text{u}}}H(u)$$
(9)

Compared with the traditional methods, the activation energy (iteration calculation until Ei − Ei−1 < 0.1 kJ mol−1) calculated by the nonlinear iso-conversional iteration methods is more reasonable.

Results and discussion

Morphology and composition characterization

Morphology and size

The morphology and size of MFe2O4 (M = Ni, Zn and Co) samples were characterized using SEM and TEM, and the results are shown in Fig. 2. SEM images showed that the MFe2O4 (M = Ni, Zn and Co) samples exhibit spherical morphology with a uniform size distribution. As can be seen from Fig. 2a–c, NiFe2O4, ZnFe2O4 and CoFe2O4 samples present rough surface with particle size about 70 nm, 360 nm and 220 nm, respectively.

Fig. 2
figure 2

SEM (a–c), TEM (d–f) and HRTEM (g–i) images of CoFe2O4 (a, d, g), NiFe2O4 (b, e, h) and ZnFe2O4 (c, f, i) and Fe2O3 samples

Full size image

TEM results (Fig. 2d–f) indicated that the MFe2O4 (M = Ni, Zn and Co) were composed of agglomerated hollow particles, which are consistent with the rough surface shown in SEM images. The HRTEM images of the NiFe2O4, ZnFe2O4 and CoFe2O4 (Fig. 2g–i) showed clear lattice fringes with the inter-planar distance of 0.250, 0.247 and 0.248 nm, respectively, which can be indexed to the (311) plane of the cubic structure. In addition, the BET results indicated that the specific surface area of NiFe2O4, ZnFe2O4, CoFe2O4 and Fe2O3 samples are 105.1, 34.68, 42.07 and 21.26 m2 g−1, respectively. The BET results are consistent well with the results of SEM and TEM, that is, smaller particle size possess lager specific surface area.

Composition and structure

The crystalline structure and composition of the as-synthesized MFe2O4 (M = Ni, Zn and Co) and Fe2O3 samples were characterized by XRD, and the patterns are presented in Fig. 3. The XRD patterns of CoFe2O4 at 18.2°, 30.0°, 35.4°, 37.0°, 43.0°, 53.4°, 56.9°, 62.5° and 74.0° can be indexed to (111), (220), (311), (222), (400), (422), (511), (440) and (533) of cobalt iron oxide (JCPDS No. 22-1086). The diffraction peaks shown in NiFe2O4 and ZnFe2O4 patterns correspond well with the crystal planes of (111), (220), (311), (222), (400), (422), (511), (440) and (533) planes of trevorite (JCPDS No. 54-0964) and franklinite (JCPDS No. 22-1012), respectively. The XRD patterns demonstrate the successful fabrication of MFe2O4 (M = Ni, Zn and Co) via the solvothermal process. As regards the other impurity phases, no diffraction peaks were observed, which is also indicative of the high purity of the synthesized MFe2O4 (M = Ni, Zn and Co). Besides, the average crystallite size of MFe2O4 (M = Ni, Zn and Co) was calculated using Scherrer’s equation, and the average crystallite size of NiFe2O4, ZnFe2O4 and CoFe2O4 samples are 13.17, 26.16 and 18.14 nm, respectively.

Fig. 3
figure 3

XRD results of NiFe2O4, ZnFe2O4, CoFe2O4 and Fe2O3 samples

Full size image

FTIR measurement was also used to investigate the composition of Fe2O3 and MFe2O4 (M = Ni, Zn and Co), and the corresponding spectra are shown in Fig. 4. The absorption peaks of MFe2O4 appear around 3420 cm−1 and 1625 cm−1, which can be derived from the O–H stretching vibration and bending vibration of adsorbed water molecules, respectively. For Fe2O3 samples, strong peaks of Fe–O at 564 and 477 cm−1 appeared in the Fe2O3 sample, which are consisted well with the previous study [23]. Moreover, for MFe2O4 (M = Ni, Zn and Co) samples, two strong absorption peaks at lower frequency (around 550 and 415 cm−1) can be assigned to the stretching vibrations of the M–O bonds in tetrahedral positions and the Fe–O bonds in octahedral positions, respectively [26]. Additionally, peaks around 1400, 1050 and 800 cm−1 appear in MFe2O4 samples, which can be assigned to the deformation vibrations of C–OH, stretching vibrations of C–O and the bending vibration of C–H bonds, respectively [26]. These functional groups are derived from the polyethylene glycol, which is used as surfactant for dispersion.

Fig. 4
figure 4

FTIR spectra of NiFe2O4, ZnFe2O4, CoFe2O4 and Fe2O3 samples

Full size image

XPS was carried out to study the elemental composition and chemical state of MFe2O4 (M = Ni, Zn and Co) samples, and the results are shown in Fig. 5. The wide scan XPS confirmed the existence of C, O and Fe elements in all MFe2O4 samples. Besides, Ni, Zn and Co elements have also been confirmed to exist separately in NiFe2O4, ZnFe2O4 and CoFe2O4 samples, and no other hetero-elements are detected. The peaks around 711, 719.0 and 725 eV in CoFe2O4, NiFe2O4 and ZnFe2O4 samples are related to Fe 2p3/2, satellite peak and Fe 2p1/2, providing clear evidence for the existence of Fe3+ [27]. For CoFe2O4 sample, two peaks located at 780.9and 796.6 eV can be assigned to the spectra of Co 2p3/2 and Co 2p1/2 of Co2+ [28]. Peaks at 856.0 and 873.8 eV are related to the Ni 2p3/2 and Ni 2p1/2 spectra of Ni2+ in NiFe2O4 sample [29]. For ZnFe2O4 sample, two peaks at 1021.8 and 1044.8 eV can be assigned to the spectra of Zn 2p3/2 and Zn 2p1/2 of Zn2+, respectively [30]. The appearance of Co2+, Ni2+ and Zn2+ peaks separation in CoFe2O4, NiFe2O4 and ZnFe2O4 also confirmed the successful fabrication of MFe2O4 (M = Ni, Zn and Co) samples.

Fig. 5
figure 5

XPS results of MFe2O4 (M = Ni, Zn and Co) samples. a Fe 2p spectrum of MFe2O4 (M = Ni, Zn and Co) samples, b Zn 2p spectrum of ZnFe2O4, c Co 2p spectrum of CoFe2O4, d Ni 2p spectrum of NiFe2O4

Full size image

Catalytic performance characterization

DSC analysis

DSC curves of TKX-50 before and after mixed with different catalysts are shown in Fig. 6, and the corresponding decomposition peak temperatures and heat released are listed in Table S1. The results indicated that the low thermal decomposition peak temperature (TLDP) and THDP of TKX-50 reduced obviously after being mixed with NiFe2O4, ZnFe2O4 and CoFe2O4 samples. The TLDP of TKX-50 mixed with NiFe2O4, ZnFe2O4 and CoFe2O4 samples are 210.6, 201.9 and 200.7 °C, respectively, and decreased by 29.3, 38.0 and 39.2 °C in comparison with the TLDP of pristine TKX-50 (239.9 °C) at 10 °C min−1. Additionally, the THDP of TKX-50 mixed with NiFe2O4, ZnFe2O4 and CoFe2O4 samples are decreased by 39.9, 47.0 and 51.2 °C in comparison with the THDP of pristine TKX-50 (268.0 °C) at 10 °C min−1. The advanced peak temperature of TKX-50 showed the excellent catalytic performance of bimetallic iron oxides. Besides, the MFe2O4 (M = Zn and Co) samples have better effects for the reduction of decomposition temperature than Fe2O3, while the NiFe2O4 has similar effect to Fe2O3 sample. The enhancement of the heat generated during TKX-50 pyrolysis also confirmed the catalytic activity of bimetallic iron oxide for TKX-50 thermal decomposition.

Fig. 6
figure 6

DSC and TG-DTG curves of TKX-50 before and after being mixed with different catalysts at 10 °C min−1. a DSC curves, b TG curves, c DTG curves

Full size image

TG-DTG analysis

The TG and the corresponding DTG curves of TKX-50, before and after mixing with Fe2O3 and MFe2O4 (M = Ni, Zn and Co), at the heating rate of 10 K min−1 are shown in Fig. 6. The decomposition process of pristine TKX-50 can be divided into two stages [31, 32]. The TG results indicated that the mass loss of pristine TKX-50 after the first decomposition process is 81.80%, and the total mass loss until 450 °C is 95.20%. After mixed with combustion catalysts, the residual mass of TKX-50 is shown in Fig. 6b. As shown in Fig. 6b, the less residue mass of TKX-50 confirms the excellent catalytic of ZnFe2O4 and CoFe2O4 samples. Additionally, the DTG peak temperatures decreased to the maximum degree indicating the excellent catalytic activity of ZnFe2O4 and CoFe2O4 for TKX-50 thermal decomposition. The results of TG-DTG and DSC are consistent, that is, CoFe2O4 and ZnFe2O4 have better catalytic effects than NiFe2O4 and Fe2O3 samples for TKX-50 thermal decomposition.

Kinetic analysis

Traditional liner Kissinger and Ozawa methods

The apparent activation energy (Ea), pre-exponential factor (LgA) and correlation coefficent (r) for the first and high thermal decomposition process of pristine TKX-50 and TKX-50 mixed with Fe2O3, ZnFe2O4, CoFe2O4 and NiFe2O4 were evaluated by traditional Kissinger and Ozawa methods. The kinetics data were calculated using thermal decomposition peak temperature (Tp, shown in Table S1 and Fig. S1) at different heating rates (5, 10, 15 and 20 °C min−1), and the corresponding kinetic parameters are shown in Table S2. The results indicated that the Ea of the low temperature decomposition process of TKX-50+Fe2O3, TKX-50+ZnFe2O4, TKX-50+CoFe2O4 and TKX-50+NiFe2O4 samples are 168.70, 141.94 and 147.41, 156.66 kJ mol−1 respectively, which are decreased by 6.45, 33.21, 28.04 and 18.79 kJ mol−1 in comparison with the Ea of pristine TKX-50. Among different catalysts, the activation energies of TKX-50 decrease to the maximum degree demonstrating the excellent catalytic performance of ZnFe2O4, while the activation energies of pristine TKX-50 calculated using traditional Kissinger and Ozawa methods could not explained the high temperature decomposition process owing to the high error bars and low linearly dependent coefficient (r). Thus, the iso-conversional and iteration methods were used to calculate the Arrhenius parameters of TKX-50 before and after mixed catalysts.

Iso-conversional methods

The kinetic parameters of TKX-50 calculated using iso-conversional Flynn–Wall–Ozawa method are shown in Fig. 7 and Table 1. The results showed that the activation energy of TKX-50 thermal decomposition reduced obviously after the addition of catalysts. Besides, bimetallic iron oxides own better catalytic effects for the reduction of activation energies of TKX-50, which can be attributed to the synergistic effects between different metals. Besides, the activation energies (conversion degree between 0.3 and 0.8) of TKX-50, before and after mixing with catalysts, are calculated and shown in Table 2. As shown in Table 2, the activation energies of TKX-50 decreased to the max degree after the addition of ZnFe2O4, indicating the excellent catalytic performance of ZnFe2O4.

Fig. 7
figure 7

Activation energy of TKX-50 thermal decomposition obtained from DSC data by iso-conversional Flynn–Wall–Ozawa method

Full size image
Table 1 Kinetic parameters calculated using iso-conversional Flynn–Wall–Ozawa method
Full size table
Table 2 Kinetic parameters of TKX-50 obtained using the DSC data
Full size table

Iteration Kissinger and Ozawa methods

Besides, the apparent activation energies were also calculated using iso-conversional iteration Kissinger and Ozawa methods, and the calculated results are shown in Fig. 8 (the corresponding Arrhenius parameters are shown in Table S2 and S3). The activation energies calculated by iteration Kissinger method are in good agreement with the results of iteration Ozawa method. The average activation energies and the error bars of TKX-50 were calculated using iso-conversional methods, and the results are shown in Table 2. As shown in Table 2, the activation energy calculated by multi-kinetic methods is very close and in good agreement with the previous studies, which showed the reliability of the results. In bimetallic iron oxide, ZnFe2O4 is an efficient combustion catalyst for solid propellant containing TKX-50, which has excellent effect of reducing activation energy and decomposition peak temperature of TKX-50.

Fig. 8
figure 8

Activation energy of TKX-50 thermal decomposition obtained from DSC data by nonlinear iso-conversional. a Iteration Kissinger method; b iteration Ozawa method

Full size image

Mechanism analysis

The reduction of the decomposition peak temperature and activation energies confirmed the catalytic effects of MFe2O4 (M = Ni, Zn and Co) for TKX-50 thermal decomposition. The catalytic effect of MFe2O4 (M = Ni, Zn and Co) for thermal decomposition of TKX-50 can be attributed to the reduction of activation energies of TKX-50. Among various catalysts used, CoFe2O4 has the best catalytic activity for thermal decomposition of TKX-50 which can significantly reduce the decomposition peak temperature and activation energy of TKX-50. The excellent catalytic activities of MFe2O4 (M = Ni, Zn, Co) are closely related to its electronic structures, grain size and particle size [33]. The excellent catalytic activity of ZnFe2O4 can be also attributed to the synergistic interaction between Co and Fe, which is beneficial for the thermal decomposition of TKX-50 [34].

Conclusions

The MFe2O4 (M = Ni, Zn and Co), which exhibit hollow structure with particle size about 70, 360 and 220 nm, were successfully fabricated via the facile one-pot solvothermal method. DSC and TG-DTG results indicated that the bimetallic iron oxides could effectively promote the thermal decomposition of TKX-50. Besides, the kinetic parameters of TKX-50 were calculated using multi-methods. The ZnFe2O4 and CoFe2O4 possess the better catalytic activity for TKX-50 thermal decomposition compared with NiFe2O4 and Fe2O3. After being mixed with ZnFe2O4 and CoFe2O4, the high thermal decomposition peak temperatures of TKX-50 were reduced by 47.0 and 51.2 °C in comparison with the pristine TKX-50 at 10 °C min−1. Besides, the activation energies of TKX-50 (α = 0.3–0.8) were decreased by 33.52 and 25.62 in comparison with the pristine TKX-50. Combined with the results of DSC and kinetic parameters, ZnFe2O4 owns excellent catalytic activity on the thermal decomposition of TKX-50, which can act as an efficient combustion catalyst of solid propellant containing TKX-50. The excellent catalytic performance of ZnFe2O4 can be attributed to the synergistic effects between Fe and Co, which are beneficial for the fracture of TKX-50.