Introduction

Homogeneous and heterogeneous solid propellants usually contain various components such as oxidizer, fuel, adhesive, modifier or catalyst, plasticizer, curing agent, stabilizer, and other additives (crosslinker, opacifier, desensitizing agent, etc.). Among them, the oxidizer is the principal component, which could rapidly release mass energy via burning. Cellulosic based energetic polymers such as nitrocellulose (NC), are typical oxidizers in military and civilian applications, which can also serve the basis to develop new energetic materials, such as propellants and explosives (Tarchoun et al. 2020a, b, c, 2021a, b, c). NC is commonly employed in smokeless powders (Chajistamatiou et al. 2016; Botelho et al. 2015), guns (Damse et al. 2009; Shen et al. 2019), blasting gelatin, dynamites, and single- (Jain et al. 2016), double- (Krumlinde et al. 2016; Boulkadid et al. 2019) and triple-base homogeneous propellants (Mendonça-Filho et al. 2019; Liang et al. 2018), in addition to other civilian applications (Yang et al. 2017) due to its high flammability and explosiveness without residual char (Yoon et al. 2016). The thermodynamic properties of oxidizers could appropriately demonstrate the burning performances of the relevant propellants based on them (Singh et al. 2009; Yan et al. 2016). Researchers have shown that the modifiers or catalysts could significantly influence the burning rate and pressure exponents of solid propellants due to their catalytic capacity for the thermal degradation of oxidizers. Therefore, the catalytic effect of catalysts including transition metal oxides (TMOs) and metastable intermolecular composites (MICs) (Mei et al. 2019), on the thermal decomposition process of oxidants has become a research focus (Chaturvedi et al. 2019; Dai et al. 2018).

As representatives of semiconductor materials and catalysts, TMOs can supply abundant electron transfer orbits during redox reactions, thus promoting electron transfer on top of oxidizers (Hu et al. 2020) and changing their decomposition routes and mechanisms. Nanothermites are typical MICs, which contain aluminum (Al) and one or several TMOs that result in enhancing the reaction activity (Stiegman et al. 2009). Based on previous reports, catalysts such as Fe2O3 (Pourmortazavi et al. 2019; Benhammada et al. 2020a), Cr2O3 (Guo et al. 2019), NiO (Wei et al. 2009), and nanothermites Al/Fe2O3 (Pourmortazavi et al. 2014; Li et al. 2013) and Al/CuO (Kim et al. 2019a; Saceleanu et al. 2018) have tremendous potential superiority in the thermal performances and pyrolysis mechanisms of NC. The nonisothermal decomposition kinetics and isoconversional kinetic models of NC with and without catalysts are employed to investigate the corresponding thermal decomposition reactions in the literature. It has been observed that the addition of some catalysts, such as α-Fe2O3, Al/Fe2O3, and Cr2O3, could remarkably decrease the values of apparent activation energy, peak temperature, and the critical temperature for the thermal explosion of NC-based composites. The pyrolysis behaviors of energetic materials could be evaluated more exactly and detailedly by employing thermogravimetry coupled with Fourier transform infrared spectroscopy (TG-FTIR) (Benhammada and Tracheet 2020). Based on the high specificity and sensitivity of TG-FTIR hyphenated techniques, the rupture of O–NO2 bond and secondary self-catalyzed reactions that occur in the coacervate phase are promoted during the decomposition process. Further, the evolved gaseous products during the pyrolysis of NC-based composites are diverse due to the catalytic effect of nano-sized additives compared with that of the pristine NC.

TMOs and MICs are used as active ingredients in solid propellants, which have been displayed to significantly improve the ignition and combustion performances of propellants (Ru et al.2016; Pang et al. 2019). Double-base propellants including nanothermites Al/PbO, Al/CuO and Al/Bi2O3 have been proven to increase the energy release rate of propellants, yet their effects on pressure exponent are not obvious (An et al. 2011). Fe2O3 particles and the corresponding Al/Fe2O3 nanothermites increase the combustion rate to varying degrees and reduce the pressure index of AP/HTPB composite propellant systems contributing to the structural effect of nano Fe2O3 with various sizes and morphologies (Zhao et al. 2014). Compared with blank propellants, the burning rate of AP-CMDB propellant (Hao et al. 2019) has risen by 22.2% with the presence of CuO nanoparticles (NPs). Ultrahigh-burning rate propellants including nano TiO2, and fluorescing propellants including quantum dots (QDs) (Thomas et al. 2018), have been developed to increase the aging stability of propellant formulations by introducing QDs or nano TiO2.

With the deepening of research, more TMOs and MICs are being developed and have tremendous potential superiority in thermal catalytic activity for oxidants, and combustion catalytic performance for solid propellants. In a previous study, we demonstrated the remarkable role of the different morphologies of CuO and corresponding nanothermites on NC decomposition based on nonisothermal decomposition kinetics (Yao et al. 2020). In this study, two novel CuO NPs—gypsophila paniculata- and crab cactus-like NPs—with specific surface areas of 90.14 and 48.93 m2/g were prepared by hydrothermal approach and could significantly improve the thermolysis process of pristine NC. The TG-FTIR hyphenated technique was employed in detecting the free radical or degradation products and deducing the possible catalytic mechanism of catalysts. Depending on the multi isoconversional analysis, the kinetic triplet activation energy (Eα), pre-exponential factor (A), and reaction model of CuO-NC and Al/CuO-NC complexes were obtained through Kissinger–Akahira–Sunose (KAS), Flynn–Wall–Ozawa (FWO), and Friedman methods (Kim et al. 2019b). The reaction models were reconstructed by introducing a modified Šesták–Berggren empirical equation according to the combined kinetic analysis. Moreover, the prepared CuO and Al/CuO nanothermites were applied to AP/HTPB composite propellants to investigate their application in combustion performances.

Experimental

Synthesis of CuO particles

Compared with our previous work (Yao et al. 2020), two novel types of CuO structures, gypsophila paniculata- and crab cactus-like, marked by CuO(G) and CuO(C),were successfully obtained by the hydrothermal methods. Typically, 0.68 g of copric chloride dihydrate (CuCl2·2H2O, 99.99%, Aladdin Biotechnology Co., Ltd., China) and 0.05 g of sodium dihydngen phoshate anhydrous (NaH2PO4, ≥ 99.0%, Aladdin Biotechnology Co., Ltd., China) were added to 20 mL of deionized water with stirring for 30 min to evenly mix the solution. Blue precipitates were obtained by adding 10 mL of sodium hydroxide (NaOH, ≥ 98.0%, Aladdin Biotechnology Co., Ltd., China) solution. The resulting suspension were transferred into a Teflon-lined stainless steel autoclave and then carried out at 120 °C for 2 h. After completing the reaction, the precipitate was washed by deionized water and absolute ethyl alcohol (C2H5OH, ≥ 98.0%, Aladdin Biotechnology Co., Ltd., China) alternately, and then put them into an oven at 60 °C for 10 h. The obtained dried product was CuO(G). Sample CuO(C) was prepared following the above experimental procedure but adjusting the reaction temperature of 80 °C.

Preparation of Al/CuO thermites

Al/CuO thermites were prepared by the sonication dispersion method in N2 atmosphere using n-hexane as effective dispersants. Firstly, CuO particles and Al nanopowders ( ≈ 50 nm in average diameter, Jiaozuo Banlv Nano Material Engineering Co., Ltd., China) were dispersed respectively. Secondly, the CuO dispersion and Al dispersion were ultrasonically mixed in the same vessel and processed at 50 °C for 2 h with ultrasound until n-hexane volatilized completely. Finally, the composite materials were obtained by drying under 60 °C for 8 h in the oven.

Preparation of NC-based composites

Physical mixtures of CuO particles or Al/CuO thermites—as the catalysts—and NC (nitrogen content of 12.6%, Xi’an Modern Chemistry Research Institute, China) were prepared in an agate mortar with a mass ratio of 1:19 at room temperature. The grinding process was maintained for 30 min to obtain the composite materials, including CuO(C)-NC, Al/CuO(C)-NC, CuO(G)-NC and Al/CuO(G)-NC.

Preparation of propellant formulations

A preliminary AP/HTPB composite propellant—as the blank control group—was composed of 14.3 wt% HTPB, 15.3 wt.% Al powders, 18.4 wt% coarse AP, and 52.0 wt% superfine AP. CuO(C), CuO(G), Al/CuO(C), and Al/CuO(G) were added as catalysts into AP/HTPB composite propellant formulations and marked as C1, C2, T1, and T2, respectively. For the experience groups containing CuO particles—C1 and C2—the contents of other ingredients were the same, except the content of CuO, which was 2.0 wt%. For the experience groups containing thermites—T1 and T2—the contents of other ingredients were the same as those of the blank control group, except the contents of Al powders and thermite were 14.5 and 2.8 wt% respectively. AP/HTPB propellants were obtained following the steps (Zhao et al. 2014): all ingredients were completely mixed in a kneading machine to obtain the propellant slurry; afterward, the slurry was poured into a mold under vacuum conditions and solidified at 70 °C for 72 h.

The above fabricating procedures can be clearly illustrated in Scheme 1

Scheme 1
scheme 1

Schematic diagrams showing preparation procedures of catalysts, NC-based composites and AP/HTPB propellants

Full size image

Characterization

X-ray diffraction studies on the crystal structure and purity of as-prepared materials were obtained on D/MAX-3C (Rigaku, Japan) device using Cu Kα1 radiation within 10º–80º. Composition analysis of materials were studied by EDS (INCAIE350, UK). Grain sizes and microstructures of the as-prepared materials were characterized by using SEM (Quanta 400, USA and TESCAN, Czech Republic) and TEM with high-resolution TEM (Carl Zeiss SMT Pte Ltd., Germany). Thermal behavior were analyzed by DSC (Q2000, TA Co.), furthermore, the kinetic triplet were dealt with the mathematic means based on the DSC data. Gases products from the whole pyrolysis process were detected by TG-FTIR technique at Ar atmosphere. The SBET and pore volume were performed on a JW-BK122W static nitrogen adsorption (Beijing JingweiGaobo, China).

Results and discussion

Characterization of CuO and Al/CuO

The crystalline structure and composition of the as-prepared samples were investigated by XRD technology (Fig. 1). The identifiable and broad diffraction peaks centered at about 32.8°, 35.9°, 39.0°, 49.1°, 54.0°, 58.6°, 61.7°, 66.3°, 68.4°, 72.6°, and 75.5° corresponded to the characteristic (110), (11–1), (111), (20–2), (020), (202), (11–3), (31–1), (220), (221), (22–2) planes of CuO (Chen et al. 2020; Liu et al. 2020). CuO displayed weak diffraction peaks at 46.0°, 52.7°, 56.8°, and 75.1°, which could be assigned to the (200), (112), (021), (004) planes. All reflection peaks in Fig. 1a belonged to the monoclinic phase (space group C2/c(15)) with unit cell parameters a = 4.6883 Å, b = 3.4229 Å, c = 5.1319 Å, α = γ = 90°, β = 99.506° of CuO(C) (JCPDS No. 48–1548), which demonstrates its good purity. The signals et al. (111), Al (200), Al (220), and Al (311) corresponded to Al powders. All diffraction peaks in Fig. 1b matched with pure Al (JCPDS No. 65–2869) and CuO(C) (JCPDS No. 48–1548). The obvious characteristic diffraction peaks in Fig. 1c coincided with those given in the JCPDS card 48–1548 of CuO(G). The XRD pattern in Fig. 1d of Al/CuO(G) thermite contained all Al and CuO diffraction peaks, thereby indicating that the crystal structures of Al and CuO were maintained during the sonication dispersion process.

Fig. 1
figure 1

XRD patterns of as-prepared catalysis

Full size image

The surface morphology and composition of the as-prepared CuO NPs were investigated using SEM, TEM, and EDS analyses. The SEM observations Fig 2a,b and high-resolution TEM image (HRTEM) Fig. 2c,d indicated that the as-prepared CuO sample displayed a crab cactus-like structure with a length of approximately 1.6 μm and an average diameter/breadth of approximately 1.0 μm in the middle portion of the structure. The enlargement of the local region (inset i in Fig. 2c) demonstrated each end of CuO with rough surfaces and boundaries. Generally, the crystallinity was poor in these areas, and crystal defects appeared. Figure 2e (inset iii) displays the stacking faults on the surface of CuO(C). CuO 3D hierarchical structure possessed single crystalline nature confirmed from the selected area electron diffraction (SAED) pattern in Fig. 2f. The EDS (Fig. 2g) results show that the product prepared was a pure CuO crystal, and the diffraction peak was obviously broadened, indicating that the crystal grain of the product was small. Determined from SEM (Fig. 2h, i) and TEM images (Fig. 2j, k), it was found that the as-prepared CuO NPs were characterized by their gypsophila paniculata-like structures and inhomogeneous grain size. The TEM micrograph in Fig. 2k suggested that the CuO NPs had rough surfaces due to the adhesion of small CuO spheres with an average particle diameter of 1.5 nm. A representative HRTEM image of CuO NPs is shown in Fig. 2l. It could be inferred that the gypsophila paniculata-like CuO structure was formed by the assembly of small particles. As shown in Fig. 2l, the distinct lattice fringe of 0.232 nm could be indexed to the (111) crystal plane of monoclinic CuO. The fast Fourier transform (FFT) pattern of the as-prepared CuO NPs is depicted in Fig. 2l (inset iv). Figure 2m presents the SAED pattern of CuO polycrystalline, which could be confirmed from the ring pattern diffraction. SEM-based EDS characterization identified that the gypsophila paniculata-like CuO contained Cu and O verified by obvious signal peaks (Fig. 2n).

Fig. 2
figure 2

a, b SEM micrographs and c, d low-magnification TEM images of CuO(C). Inset i is the TEM image of a small part of CuO(C) nanoparticle; e HRTEM image of CuO(C) nanoparticle. FFT pattern (inset, ii) marked “A” is attributed to [111] zone axis of CuO(C) crystal. Dislocation substructures (inset, iii) corresponding to the area “B” are observed; (f) SAED patterns of CuO(C) hierarchical structure; (g) EDS pattern of CuO(C); h, i SEM micrographs and j, k low-magnification TEM images of CuO(G); l HRTEM image of the small part of CuO(G) nanoparticle. FFT pattern (inset, iv) marked “C” correspond to [111] crystal plane of the monoclinic CuO(G); m SAED pattern of CuO(G) crystal; n EDS pattern of CuO(G)

Full size image

SEM micrographs and TEM images in Fig. 3a, c suggested that CuO(C) NPs were tightly bound on the surface of Al NPs. As shown in Fig 3f, h, the gypsophila paniculata-like CuO(G) crystallites attached to the Al NPs uniformly, thereby preventing the aggregation of Al and CuO(G) particles in the thermites. The HRTEM observation and FFT pattern of CuO(C) NP showed that the CuO(C) was a single crystal and grew along the (111) crystal plane (Fig. 3d and inset i). Similarly, the FFT pattern corresponding to the (111) plane of CuO(G) was observed in Fig. 3i, inset ii. Figure 3i, area “C” shows a high-resolution image of CuO(G) containing dislocation tangles/networks on the surface. The microstructure of the Al oxidized layer with a thickness within 0–3.5 nm was confirmed through HRTEM in Fig.3e, j. The chemical compositions of the two types of Al/CuO thermites were measured by the EDS analysis (Fig. 3b, g), which revealed Cu, O, and Al elements in the as-prepared Al/CuO thermite. The Cu and O elements originated from CuO crystallites whereas the Al element was obtained from the Al NPs.

Fig. 3
figure 3

a SEM micrograph, b EDS pattern and c TEM image of Al/CuO(C); d HRTEM image of CuO(C) nanoparticle. FFT pattern (inset, i) marked “A” is attributed to [111] zone axis of CuO(C); e HRTEM image of Al nanoparticle mixed with CuO(C); f SEM micrograph, g EDS pattern and h TEM image of Al/CuO(G); i HRTEM image of CuO(G) nanoparticle. FFT pattern (inset, ii) marked “B” is attributed to [111] zone axis of CuO(G). Dislocation substructures corresponding to the area “C” are observed; (j) HRTEM image of Al nanoparticle mixed with CuO(G)

Full size image

Kinetic parameter, physical model and kinetic compensation effects

DSC curves of NC and CuO(C)-NC, Al/CuO(C)-NC, CuO(G)-NC and Al/CuO(G)-NC composites obtained at a heating rate of 10.0 K min–1 are presented in Fig. S1. Pure NC exhibited a strong exothermic peak at 209.67 °C, and the exothermic peak temperature decreased to 209.63 °C, 208.00 °C, 207.80 °C, and 207.53 °C with the addition of CuO(C), Al/CuO(C), CuO(G), and Al/CuO(G), respectively. Benefiting from the synergistic effect of Al powders and CuO particles, the catalytic activities of Al/CuO thermites were superior to that of pure CuO.

To explore the catalytic effects of CuO(C), Al/CuO(C), CuO(G), and Al/CuO(G) on the pyrolysis process of NC (Fig. S1) and obtain the corresponding kinetic parameters, the DSC curves with linear heating rates (Fig. 4) were analyzed using mathematics methods. From Fig. 4, the temperature increased with the heating rate, which indicated the effect of the heating rate on NC thermolysis, as in the previous reports (Tarchoun et al. 2019; Chai et al. 2019).

Fig. 4
figure 4

Experimental plots obtained for the thermolysis of the NC-based composites in N2 atmosphere with the linear heating rates

Full size image

FWO, KAS, and Friedman isoconversional plots of CuO(C)-NC, Al/CuO(C)-NC, CuO(G) NC and Al/CuO(G)-NC in Fig. 5 were obtained from the simultaneous analysis (Sánchez-Jiménez et al. 2013) for the experimental plots shown in Fig. 4. The FWO isoconversional plot trend generally agreed with the KAS plot; also, it could be initially predicted that the obtained Eα values from the slopes of the straight lines in Fig. 5 seemed very close to each other for the same studied material. The Friedman isoconversional plot shows a slight difference trend due to the employed approximations (Xu et al. 2017) used in the algorithms. From Fig. 5, the lines ran approximately parallel; however, the slopes of the lines corresponding to the high conversion rate (α) were varied. This variation of activation energy might be explained by the autocatalytic reactions of NC and the reactions between the intermediate products during decomposition promoted by the catalysts (CuO and Al/CuO) and the strong oxidizing activity of gas products, such as NO2.

Fig. 5
figure 5

Global kinetic plots of the NC-based composites depending on FWO, KAS and Friedman methods

Full size image

By plotting ln(βi), ln(βi/\({T}_{\alpha ,i}^{2}\)), and ln[βi(dα/dT)α,i] against 1/T for α from 0.025 to 0.975 with a step-size of 0.025, the activation energies could be established by the slope (k) depending on Eα = -kR/1.052 for FWO method, and Eα = -kR for KAS and Friedman methods. Eα and the related confidence intervals (errors bars) evolution with respect to α obtained using the KAS, FWO, and Friedman approaches are shown in Fig. 6. Eα obtained from DSC data almost agreed with calculations obtained using the FWO and KAS methods for the as-prepared NC-based composites. It could be seen that Eα obtained by Friedman method exists some differences from those calculated by the other two model-free methods, which might be due to the accuracy of the mass loss derivative (Mamleev et al. 2004; Hadigheh et al. 2021). Compared with the Eα values of NC (Fig. S2), the Eα values of NC-based composites decreased due to autocatalytic characteristics of NC with the condition of 12.6% nitrogen content and the catalytic activity of catalysts during the initial reaction stage. As the conversion rate increased, the Eα values of the NC-based system with and without catalysts, including CuO(C), Al/CuO(C), and CuO(G), tended to increase, indicating a more complex chemical reaction. However, it revealed that the Eα values of Al/CuO(G)–NC first maintained a certain level and then decreases with the increase in conversion rates. In addition, the plots for the above isoconversional methods exhibited comparable trends, and the obtained Eα values seemed to be close to each other. The results were obviously different from those of CuO(C)-NC, Al/CuO(C)-NC, and CuO(G)-NC composites, which were due to the different catalysis abilities of CuO(C), Al/CuO(C) and CuO(G), as well as the decomposition products. Moreover, as presented in Table 1, the activation energy of Al/CuO(G)-NC decreased most significantly, indicating the best catalytic activity of Al/CuO(G) on the decomposition behavior of NC.

Fig. 6
figure 6

Eα and the associated error evolution with respect to the conversion rate (α) obtained by the three employed isoconversional methods for the NC-based composites

Full size image
Table 1 Kinetic parameters of decomposition reaction for NC-based composites
Full size table

Figure 7 shows the theoretical curves for the kinetic models represented in Table S1 (García-Garrido et al. 2018; Sánchez-Jiménez et al. 2010), and experimental master curves were drawn based on the experimental data from the curves in Fig. 4. The experimental master curves obtained at six heating rates should run similar trends, while cannot be fitted well with any theoretical curve. The reason was that the commonly used models listed in Table S1 were single step reaction models, which were unsuitable for pyrolysis reactions in solid or porous structural solids including the complex multistep pyrolysis reaction of NC. So, the experimental data reconstruction had bad repetitions when the formula assumed as the best fitting kinetic reaction model was chosen from Table S1. That is, even the selected highly fitting model was not a true model of NC pyrolysis (Gao et al. 2019). Therefore, an adjustment function—the modified Šesták–Berggren empirical equation (Eq. (1))—was introduced to reconstruct the reaction model based on the combined kinetic analysis method to be highly compatible with the experimental data. Equation (1) was proved to be fit for every ideal kinetic model in Table S1 by only modifying the proper c, m, and n parameters to span the most common physical models and their possible deviations from ideal conditions (Pérez-Maqueda et al. 2006; Zhongliang et al. 2019).

$$f\left( \alpha \right) \, = c\alpha^{m} \left( {1 - \alpha } \right)^{n}$$
(1)
Fig. 7
figure 7

Comparison between normalized curves of the reaction models for a CuO(C)-NC, b Al/CuO(C)-NC, c CuO(G)-NC and d Al/CuO(G)-NC and theoretical master curves established from the ideal reaction models in Table S1

Full size image

By analyzing the experimental data in Fig. 8 and matching works, the relation curves of the four NC-based composites were obtained from the experiments by which we could obtain rational c, n, and m values and finally realized the reaction models. The reaction model functions of the four NC-based composites could be expressed as quantitative relation formulas (Table 1) (Fig. 8a–d), respectively. The kinetic compensation effect described a linear relationship between Eα and the pre-exponential factor. The better the correlation, the more rational the reaction model. The modified theoretical models were in high accordance with the corresponding experimental data (Fig. 9). From Fig. 9, the slopes and intercepts corresponding to four straight lines varied, indicating that CuO(C), Al/CuO(C), CuO(G), and Al/CuO(G) catalytically influenced on the compensation effect parameters of NC differently. Based on the numerical model reconstruction, the pyrolysis processes of CuO(C)-NC, Al/CuO(C)-NC, CuO(G)-NC and Al/CuO(G)-NC could be controlled mainly by the modified n-order reaction models. A similar integral reaction model has already been reported for pure NC (Gao et al. 2019). In fact, various reaction models and Arrhenius parameters of NC with or without catalysts have been reported (Benhammada and Trache 2021; Benhammada et al. 2020a, b; Tarchoun et al. 2020c), which could be explained by the difference in the nitrogen content of NC (Pourmortazavi et al. 2009), employed analytical techniques, kinetics methods (Chelouche et al. 2019), and catalyst characteristics.

Fig. 8
figure 8

The energy compensation effect for pyrolysis of the NC-based composites calculated via the combined kinetic analysis method

Full size image
Fig. 9
figure 9

TG-TGA curves of aCuO(C)-NC, b Al/CuO(C)-NC, c CuO(G)-NC and d Al/CuO(G)-NC

Full size image

TG-FTIR analysis

The TG-FTIR is an established technique for exploring the pyrolysis process of high energy density materials. Figure 9 shows the TG-TGA curves of NC-based composites under a heating rate of 10 K min–1 within 50–400 °C. A single-stage decomposition was noticed in the TG curves of CuO(C)-NC, Al/CuO(C)-NC, CuO(G)-NC, and Al/CuO(G)-NC composites; the weight losses were 35.73%, 39.37%, 35.94%, and 40.58%, respectively. Compared with the weight loss of NC (68.41%) reported by us (Zhao et al. 2016a), the weight loss of the composites reduced after the addition of the catalyst. The undecomposed remains were attributed to the NC residues, CuO catalyst or products of the aluminothermic reaction. As the TG-TGA curves of CuO(C)-NC in Fig. 9a shown, starting temperature of pyrolysis (Ti), temperature of extrapolated onset (Te), temperature of maximum pyrolysis rate (Tp), extrapolated final temperature (Tc), and terminal temperature (Tf) of the thermal decomposition process were 187.57 °C, 197.87 °C, 208.21 °C, 225.23 °C and 257.18 °C. Figures 9b–d show the representative temperatures (Ti, Te, Tp, Tc, and Tf) for Al/CuO(C)-NC, CuO(G)-NC and Al/CuO(G)-NC, respectively. As in our previous report on NC pyrolysis (Zhao et al. 2016b), the Te and Tp of the composites were lower than these of NC due to the addition of catalysts, which agreed with the DSC results. The pyrolysis behaviors and products of the four NC-based composites should be proposed, and the catalytic mechanism of the catalyst should be revealed.

According to previous studies (Benhammada et al. 2020a, b; Katoh et al. 2010; Trache et al. 2017), including our reports (Zhao et al. 2016c), the pyrolysis process of NC starts from the rupture of the –O–NO2 bond, and the released N2 stands still within the polymeric compound, then reacts with the free radical or thermally degraded products. In addition, we found that Fe2O3 NPs, as catalysts, could promote the breaking of –O–NO2 bonds and keep the gaseous products stagnant within the polymeric skeleton attributed to superior adsorption capacity, thus accelerating autocatalysis.

The catalysts investigated in this study had a wide variance between the Fe2O3 NPs. The TG-FTIR technique was employed to investigate the pyrolysis processes of the four as-prepared NC-based composites. Figures 10 and S3 show IR spectra of gas products at various representative temperatures, including Ti, Te, Tp, Tc, and Tf. To determine whether H2O and CO2 came from the environment or from the thermal decomposition process, the IR spectra at a certain temperature point (Tx) before the initial decomposition temperature was considered for a the comparative study.

Fig. 10
figure 10

IR spectra of the gases evolved from the degradation of the NC-based composites

Full size image

The IR characteristic absorption peaks of CuO-NC at 165.30 °C (Tx) and 187.57 °C (Ti) are shown in Figs. 10a and S3(a). H2O (3605–3741 cm−1), CO2 (2365, 672 cm−1), and NO2 (1592–1636 cm−1) (Li et al. 2012; Wang et al. 2016) could be confirmed. These results were different from the initial stage of the thermal decomposition of NC, which did not decompose to generate NO2 before the initial decomposition temperature. Although NC might slowly decompose as temperature increased, it was enhanced by CuO(C) particles due to their catalysis. The peaks at 3605–3741 cm−1 could be attributed to the O–H bonding vibration of H2O. The bands at 2360 and 670 cm−1 were assigned to the stretching and bending vibration modes of CO2, respectively. By comparing the IR spectra at 165.30 °C (Tx) and 187.57 °C (Ti) in Fig. 10a, the intensities of the discovered H2O, CO2 and NO2 were almost invariant. At 197.87 °C (Te), the absorption peaks of H2O, CO2, NO2, and N2O (2241 cm−1) were easily identifiable (Fig. 10a, Te). When the temperature increased to 208.21 °C, the gaseous products including NO (1763–1966 cm−1), CO (2195 cm−1), N2O (2240 cm−1), HCHO (2815, 1741 cm−1), and HCOOH (1081–1129 cm−1) were detected (Gratien et al. 2007). The bands at 2815 and 1741 cm−1 associated with the C–H and C=O stretching vibration of HCHO due to the cleavage of the −CH2ONO2 groups (Chen et al. 1991). The HCOOH gas products resulted from the secondary self-catalyzed reactions of NC (Shehata et al. 2003). The gas product HCN (2240–2260 cm−1) might be generated by the reaction of HCO• radical with NO (Wei et al. 2010). From Fig. 10 (a, Tp), the strength of these gases was significantly enhanced, which was an essential basis for the faster pyrolysis rate of CuO(C)-NC complex than that of the pristine NC, and showed the catalytic effect of CuO(C) particles. The CuO(C) particles exist rough surfaces and lattice defects (Fig. 2), which could provide adsorption and active sites to absorb the gases produced during the pyrolysis process of NC, thereby promoting a secondary autocatalytic reaction. At 225.23 °C (Tc), the intensities of NO2, N2O, HCHO and CO were weak compared with the plots in Fig. 10(a, Tp and Tc). The intensities of CO2, NO, HCOOH and H2O increase might be due to the remaining thermolysis or the reaction of NO2 and HCHO (Wei et al. 2010) with the catalysis of CuO(C) particles. At the final stage of the decomposition (Fig. 10a, Tf), the peaks of HCOOH, NO and CO disappeared, and the NO2, CO2 and H2O gases were still detected.

Because commercial NC is accompanied by unstable impurities (such as residual acid, cellulose sulfate, nitro-sulfur mixed esters, and various nitro-sugars) during the manufacturing process, these unstable impurities are easily decomposed. The catalysis of decomposition products and the autocatalytic decomposition of NC could promote NC pyrolysis. Although the NC used in the experiment has high purity, NC was also hydrolyzed due to the presence of a catalyst at high temperatures. The hydrolysis products promoted the further decomposition of NC, and the autocatalytic process was repeated. With the Al/CuO(C) catalyst, the gas products H2O, CO2 and NO2 were detected from the IR spectrum of Al/CuO(C)-NC at 165.30 °C (Tx) (Figs. 10b and S3(b)). At the initial decomposition temperature (Ti = 185.29 °C) of Al/CuO(C)-NC, the IR characteristic peaks of H2O, CO2, NO2 and NO could be identified. At 195.67 °C (Te), the intensities of absorption peaks including H2O, CO2, NO2, and NO are slightly increased. It can be mentioned that more gases were released at 209.46 °C (Tp), containing H2O, CO2, NO2, NO, N2O (2200–2300 cm−1), CO, HCN (2240–2260 cm−1), HCHO (2700–2900 cm−1), and HCOOH. The intensities of NO2, N2O, CO2, CO, HCN, and HCOOH peaks increased with temperature. The peak of HCHO was weak, probably due to the reaction between NO2 and HCHO. The decrease in HCHO gas might be due to the addition of Al/CuO(C) catalyst hindering the thermal decomposition of the –CH2O• groups. At 220.37 °C (Tc), the peaks of HCOOH, NO, and CO2 were found to be stronger than the others, and the peaks of H2O, NO2, N2O, CO, HCN, HCHO were detected. Additionally, the distinguishable IR bands of H2O, CO2, NO2 and HCOOH were detected at the final temperature (Tf = 262.68 °C) of the decomposition process of Al/CuO(C)-NC. From the escaped gas products, the catalytic action of Al/CuO(C) was different from that of CuO(C). It can be concluded that the Al/CuO(C) interacted with NC during the catalytic process to generate metal atoms or clusters of “active species” with high catalytic activity, and immediately participated in the catalytic reaction, thereby maximizing the catalytic performance of the catalyst.

For the IR data of CuO(G)-NC at 165.30 °C, the bands in the range of 3557–3739 cm−1 could be attributed to the O–H bond. The obvious IR bands of CO2, NO, and NO2 were presented at 2306–2383 cm−1, 1771–1889 cm−1 and 1571–1639 cm−1, respectively. At 181.02 °C, the peaks of H2O, CO2, NO and NO2 are detected from the IR spectrum of CuO(G)-NC (Figs.10c and S3(c)). At 196.87 °C, more gas phase products such as H2O, CO2, NO, HCHO, HCOOH and NO2 were identified. At 204.90 °C, CuO(G)-NC exhibited characteristic bands in the ranges of 3500–3731 cm−1, 2725–2920 cm−1, 2296–2384 cm−1, 2241 cm−1, 2240–2260 cm−1, 2145–2196 cm−1, 1771–1932 cm−1, 1563 ~ 1634 cm−1 and 1109 cm−1 that corresponded to H2O, HCHO, CO2, N2O, HCN, CO, NO, NO2, and HCOOH, respectively. The bands of N2O, HCN, CO, NO, and NO2 were much stronger at the temperature range from 196.87 °C to 204.90 °C. Otherwise the bands of CO2 and HCOOH gradually weakened after the CuO(G)-NC composites started to decompose. From Fig. 10c, the signals of H2O (3506–3723 cm−1), HCHO (2786–2858 cm−1), CO2 (2300–2383 cm−1), N2O (2241 cm−1), HCN (2236–2254 cm−1), CO (2108–2201 cm−1), NO (1767–1932 cm−1), NO2 (1569–1628 cm−1), and HCOOH (1056–1133 cm−1), corresponding to the IR spectrum of CuO(G)-NC at 222.91 °C, could be observed, whereas the content of NO2 gradually decreased, and the absorbance intensities of gaseous products including HCHO, N2O, HCN, CO, NO, and HCOOH increased with temperature. At 260.37 °C, peaks were revealed in the ranges of 3505–3736 cm−1, 2301–2384 cm−1, 1771–1900 cm−1, 1565–1647 cm−1, and 1068–1133 cm−1 corresponding to H2O, CO2, NO, NO2 and HCOOH, respectively. The thermolysis process of CuO(G)-NC started from the rupture of the –O–NO2 bond and released NO2. With the fracture of the –C–O–C– bond, the skeleton of the NC macromolecule collapsed. Owing to the catalytic action of CuO(G), the escaping NO2 gas was confined in the skeleton and further reacted with the condensed phase, triggering a self-catalytic reaction, thus accelerating the decomposition of NC and producing H2O, HCHO, CO2, N2O, HCN, CO, NO, NO2, and HCOOH.

Figures 10d and S3(d) present the IR spectra of Al/CuO(G)-NC at typical temperatures in the N2 atmosphere. At 165.30 °C, the peaks of H2O, CO2 and NO2 were assigned. At 178.88 °C, the main evolved gases during the Al/CuO(G)-NC thermolysis are H2O, HCHO, CO2, NO, and NO2. The peaks at 197.30 °C were due to the evolution of H2O, CO2, HCHO, N2O, HCN, CO, NO2, NO, and HCOOH gases. By comparing the IR spectra of 197.30 °C and 204.17 °C, the type of the evolved gas was unchanged, but the intensities of the absorption peaks of CO2, N2O, HCN, CO, and NO gases significantly increased. At 225.26 °C, the H2O, HCHO, CO2, N2O, CO, NO, NO2, and HCOOH absorption peaks existed, whereas the intensities of CO2, CO, and HCOOH increased, and the peak of HCHO decreased. At the final stages of the decomposition, the gaseous products of Al/CuO(G)-NC mainly contained H2O, CO2, NO, NO2, and HCOOH. Overall, after adding of Al/CuO(G), the gaseous small molecules remarkably increased, which revealed that the pyrolysis of Al/CuO(G)-NC composite was more complete at the existence of Al/CuO(G).

Based on the TG-FTIR experiment analysis above, the pyrolysis mechanism of the four NC-based composites could be deduced by Scheme S1. During the NC thermolysis, the –O–NO2 bonds first ruptured and released NO2, which was much obvious with the presence of catalysts at high temperatures. The glucopyranosyl ring might further be cleaved to produce H2O, CO2, and NO gases. These gaseous products, especially NO2, could react with the condensed phase, free radical, and other products generated from the polymer degradation. CuO(C), Al/CuO(C), CuO(G), and Al/CuO(G) could accelerate the autocatalytic reaction of NC and produce H2O, CO2, N2O, HCN, CO, NO, NO2, HCHO, and HCOOH.

Catalytic combustion property of AP/HTPB propellants

Figure 11 shows 3D histograms of burning rate versus pressure (Brewster and Mullen 2011) and catalytic efficiency versus pressure of AP/HTPB propellants. From Fig. 11a, the burning rate of the blank group (N0) and experimental groups (C1, C2, T1, and T2) increased with an increase in pressure in the range of 4–15 MPa. At 15 MPa, all formulations (N0, C1, C2, T1, and T2) had the fastest combustion rate and highest catalytic efficiency. For all groups, both the burning rates and catalytic efficiencies within 4–15 MPa followed the order: N0 < C1 < T1 < C2 < T2. Compared with N0, the burning rate of C1, C2, T1, and T2 containing CuO(C), CuO(G), Al/CuO(C), and Al/CuO(G) catalysts at 15 MPa increased by 11%, 17%, 23%, and 43%, respectively. The catalytic efficiencies of C1, C2, T1, and T2 at 15 MPa were 11%, 17%, 23%, and 43% higher than that of N0. The results meant that the catalytic capability of Al/CuO(G) was the best of all catalysts. Thermites had higher catalytic activity than oxides; for the thermites, the catalytic activity was in the order: Al/CuO(C) < Al/CuO(G); for the oxides, the order was CuO(C) < CuO(G) (within 4–15 MPa).

Fig. 11
figure 11

Burning rate (r) a and efficiency b vs pressure (p) 3D histograms of AP/HTPB propellants

Full size image

From the above experimental results, the burning rate exhibited greater sensitivity to the pressure. The ensuing flow oscillations occurring in the interior of a solid rocket motor substantially increase the pressure, which can further influence the instantaneous combustion rate. However, excessive pressure will cause the solid rocket motor to crack. The exponent (n) of propellants was another significant parameter that impacted the chamber pressure. To avoid causing large variations in thrust or combustion chamber in a high pressure environment, the pressure exponent should be as small as possible. In most situations, the n value of composite solid propellant was in the range of 0.2–0.65. Notably, that the n value was also related to the species and composition of the propellant formulations, such as the burning rate modifiers/catalysts. From Fig. 12, the data show that the pressure exponents of C1, C2, T1, and T2 with catalysts were lower than that of N0 in different pressure ranges. The formulations containing CuO(G) possessed lower pressure exponents than the formulations containing CuO(C), especially in the high pressure range. For the formulations including thermite, the pressure exponent of T2 was much smaller than that of T1 within 7–15 MPa.

Fig. 12
figure 12

Pressure exponent (n) of propellant formulations at different pressures

Full size image

Consequently, the burning performances of AP/HTPB propellants only depended on the catalyst; when the kind and composition of a formulation is determined, with the exception of catalyst. T2 exhibited the fastest burning rate, highest catalytic efficiency, and most rational pressure exponent when the Al/CuO(G) catalyst was added. Thus, Al/CuO(G) was the best catalyst. Because the Al powder used in the thermites was the same, the catalysis of the thermite was explained by the CuO particles with different morphologies.

To assess the catalytic combustion of the as-prepared CuO NPs, the SBET, pore volume, and average pore diameter of CuO NPs were carried out. Both CuO(C) and CuO(G) exhibited a type IV physisorption isotherm Fig. 13a in accordance with the International Union of Pure and Applied Chemistry (IUPAC) classification (Sing et al. 1985; Thommes et al. 2015), which was characteristic of mesoporous materials. The specific surface area was evaluated to be 48.93 and 90.14 m2 g−1 for CuO(C) and CuO(G), respectively. Compared with CuO(C), the specific surface area of CuO(G) increased by 41.21 m2 g−1, which suggested that CuO(G) particles exhibited the mesoporous structure and provided more adsorption and active sites. For CuO(C), there was a slight increase in the adsorption capacity values in the high-pressure region (P/Po) of type H3 hysteresis loop, due to nitrogen capillary condensation in a stochastic and interconnected pore network. The pore diameter distribution of CuO(C) indicated the pore size distribution in the mesoporous region with an average pore diameter of ca. 9.1 nm (Fig. 13b). CuO(G) exhibited a type H4 hysteresis loop judging by the IUPAC classification, revealing the existence of narrow pores with the maximal pore size of about ca. 4.3 nm. The total pore volumes of CuO(C) and CuO(G) were 0.066 and 0.046 cm3 g−1, respectively. The results indicated that there were ultrananopores in mesoporous nanomaterials, which are seen in the TEM image (Fig. 2k).

Fig. 13
figure 13

Nitrogen adsorption–desorption isotherms a and pore size distribution curves b of CuO structures

Full size image

Solid-state reactions often occur between crystal lattices or with molecules and may depend on lattice defects. The catalyst with crystal defects and high specific surface area (Yu et al. 2019) will accelerate the bond breaking and the release of small molecular fragments. The specific surface area of CuO(G) was much larger than that of CuO(C). The mesoporous structure, rough surfaces and ultrananopores provided more adsorption positions and active sites, which were favorable for the activation and diffusion of the reactants, thereby enhancing the thermal and combustion catalytic properties of CuO(G).

Conclusion

Structure-catalytic activity relationships of CuO particles and their particle size/shape dependencies on the catalytic thermal degradation of NC as well as the catalytic combustion performance of AP/HTPB propellant were investigated in this study. Gypsophila paniculata- and crab cactus-like CuO structures were successfully obtained by hydrothermal methods and integrated with Al powders to achieve Al/CuO thermite via ultrasonic mixing. The as-prepared CuO and Al/CuO thermites are promising catalysts in reducing the activation energy and the peak temperature of NC thermolysis. With these catalysts, strong catalytic effects occurred, promoting the –O–NO2 bond cleavage during the initial pyrolysis of NC, as well as the condensed-phase decomposition and the reactions between the intermediate products. The initial heat release process of NC follows a certain autocatalytic decomposition reaction, whereas its mechanism model changes to fit well with the n-th order reaction at high- temperature ranges. In addition, the proper f(α) of NC-based composites were determined by model reconstruction with a modified Šesták–Berggren empirical equation based on the combined kinetic analysis method. TG-FTIR analysis certified that the thermolysis mechanism of NC could be changed by mixing different catalysts, which could have resulted from the absorption and desorption procedure of gaseous products generated within the range of different temperatures on catalysts with various adsorption positions and active sites. The burning rate of the formulation containing catalysts Al/CuO(G) increased by 43% compared with the basic formulation without catalyst at 15 MPa due to the large specific surface area (90.14 m2 g−1) for CuO(G) and the addition of Al nanopowders. This study presents two potential catalysts, including CuO particles and their corresponding thermites, by investigating the relationship of microstructure and catalytic activity; and gives a promising way to choose an excellent catalyst for accelerating the pyrolysis process of NC and improving the combustion performance of AP/HTPB composite propellants.