Synthesis and Effects of Two Novel Rare-Earth Energetic Complexes on Thermal Decomposition of Cyclotetramethylene Tetranitramine (HMX)

In order to explore the effect of the energetic complex on the thermal decomposition HMX, two new rare-earth energetic complexes [La(tza)(NO3)2(H2O)4]n (1) and [Ce(tza)(NO3)2(H2O)4]n (2) (Htza = tetrazole-1-acetic acid) were prepared by a solvent evaporation method. The obtained products were structurally characterized by Fourier-transform infrared spectroscopy (FTIR), elemental analysis, powder X-ray diffraction (PXRD), single crystal X-ray diffraction (XRD), and thermogravimetric analysis coupled with differential scanning calorimetry (TG-DSC). In addition, the compatibility of complex 1 with cyclotetramethylene tetranitramine (HMX) was studied by DSC and FTIR, respectively. Structural analysis suggested that complex 1 exhibits an orthorhombic, P 21 21 21 space group, and the La (III) ion was 10-fold coordinated in a distorted double-capped antiprism configuration. Complex 2 featured a one-dimensional, right-handed helical infinite chain. The effect of complexes 1 and 2 on the thermal decomposition of HMX was investigated by DSC, which revealed that complex 1 showed a slightly better effect than 2 on the thermal decomposition of HMX and released more heat. Furthermore, complex 1 had good compatibility with HMX, indicating that it may act as a combustion promoter for nitrate ester plasticized polyether (NEPE) solid propellant.


Introduction
Nitrate ester plasticized polyether (NEPE) propellant is a new type of crosslinked, high-energy, solid propellant which uses polyether polyurethane as a binder, nitrate esters as plasticizers, and is filled with a large amount of the high-energy oxidant, cyclotetramethylene tetranitramine (HMX) [1][2][3]. It also contains a combustion agent, stabilizer, and combustion promoter [4]. Here, HMX is used as the main component of the NEPE solid propellant, and a large amount of HMX increases the pressure index of the propellant, so improving the thermal degradation properties of HMX will significantly improve the combustion performance of NEPE [5][6][7]. Although the combustion promoter accounts for only a small proportion of solid propellants (about 2%-5%), it can greatly improve their combustion performance [8][9][10], so studying the type and performance of combustion promoters is highly important.
Energetic metal complexes are generally assembled with energetic ligands and metal ions, which are synthesized by different coordination techniques to give them certain spatial structures. They also have stable structures and superior thermal stabilities, suggesting they may be used in explosives [11]. On one hand, the energy released by the complexes can provide additional energy during the thermal decomposition of HMX or the combustion of NEPE. The energy can also be used

Materials and Methods
The experimental materials were all analytical grade and commercially available. The purity of all reagents was verified by a Vario ELIII elemental analyzer (Elementar, Hanau, Germany). Samples for FTIR were analyzed on an imported spectrometer. Powder X-ray diffraction (PXRD) patterns were obtained using Cu Kα-ray using monochromatic graphite at room temperature. In this study, the thermal decomposition performance of complexes was evaluated in the range from 40-400 • C under flowing N 2 at 20 mL·min −1 .

X-ray Single-Crystal Diffraction (XRD)
Complexes were analyzed by a SMART-1000 X-ray diffractometer (BRUKER ASX, Karlsruhe, Germany). At 298 (2) K, the sample was scanned by Mo Kα ray (λ = 0.71073 Å) in the form of ω/2θ. All single-crystal structures were solved by direct methods and anisotropically refined using a full-matrix least-squares F 2 method using SHELXTL-97. All non-H atoms were refined in a full-matrix anisotropic thermal parameters approximation. The H atoms of the ligands were obtained using a riding model, while disordered tza − and NO 3 − in complexes were refined by performing split and occupancies refinement.

X-ray Powder Diffraction (PXRD)
Complexes were measured by graphite monochromatic Cu Kα ray with a tube pressure of 40 kV and a tube flow of 100 mA on a Rigaku D/max-rA X-ray diffractometer (Rigaku, Tokyo, Japan).

TG-DSC Thermal Analysis
Samples were analyzed on a Mettler Toledo GC10 TG-DSC (Mettler Toledo, DE, USA). Complex (0.7 mg) was added to a closed platinum crucible and measured under a flowing N 2 atmosphere (20 mL·min −1 ) at a heating rate of 10 K·min −1 .

Effect of the Energetic Complexes on Thermal Degradation of HMX
To evaluate how the energetic complexes facilitated the thermal degradation of HMX, HMX was mixed with the synthetic complexes at a 19:1 mass ratio. The amount of the promoter used in this experiment on the Mettler Toledo DSC823E was similar to the amount of a typical catalyst, which was about 5% (mass ratio).

Compatibility Test between HMX and Complex 1
Since complex 1 had a more pronounced effect on the thermal decomposition of HMX than 2, the compatibility of complex 1 and HMX was then studied by DSC and FTIR [36]. A sample was obtained by mechanically mixing and grinding complex 1 with HMX at 1:1 mass ratio for 5 min to maximize all possible interactions between complex 1 and HMX.  Table S1, and selected bond information is shown in Supplementary Materials Table S2.

Structures of [Ce(tza)(NO 3 ) 2 (H 2 O) 4 ] n (2)
The asymmetric structure of complex 2 includes one Ce(III) ion, one tza − , two NO 3 − , and four coordinated water molecules. The Ce(III) ion is 10-fold coordinated and located in a distorted double-capped antiprism configuration ( Figure 2a). The capping position is occupied by two O atoms (O3 and O6) from two different NO 3 − groups. One square plane of the antiprism is formed by four O atoms (O1, O2, O1W, and O4) from two different tza − groups, one coordinated water, and one NO 3 − group. The other positions are occupied by four O atoms (O2W, O3W, O4W, and O7). The selected bond information is shown in Table S2. As shown in Figure 2b, the tza − anions connect with Ce 3+ cations to form a 1-D infinite chain, which has a left-handed helical configuration along the b direction with adjacent Ce atoms separated by 6.10 Å. Each 1-D helical chain interacts with neighboring ones via O-H . . . N hydrogen bonds between water molecules and tza − anions which leads to the formation of an extended 3-D supramolecular structure (Figure 2c). The corresponding bond information is listed in Table S2. Supplementary Materials of this article include X-ray crystallographic files in CIF reports, crystal data, and structure refinement for the complex, selected bond lengths, and bond angles in Supplementary Materials. CCDC no.18881025 for 1 and 1524937 for 2.

Fourier-Transform Infrared Spectroscopy (FTIR)
The FTIR spectrum of complex 1 has a wide characteristic absorption band from 3000 cm −1 to 3600 cm −1 (Figure 3), which corresponds to the O-H stretching vibration and the formation of hydrogen bonds between crystalline water and coordinated water in [La(tza)(NO 3 ) 2 (H 2 O) 4 ] n . The HOH bending vibration peak is at 1627 cm −1 [37]. The asymmetric stretching absorption vibration of Htza at 1720 cm −1 shifted to around 1600 cm −1 , mainly due to the coordination of a La ion with an O atom in C(O)O − . Due to the symmetrical stretching vibration of C=O, an absorption peak can be seen near 1454 cm −1 . The absorption peaks near 815 cm −1 appeared due to the in-plane bending vibration of -COOH (δ(C=O)). The absorption peak near 1332 cm −1 is the symmetrical vibration peak of -NO 3 [38]. The above analysis is consistent with the synthesized crystal structure. Since an O atom in C(O)O − coordinated with a Ce ion, the stretching vibration of C=O in Htza shifted from 1720 cm −1 to around 1600 cm −1 . The absorption peak near 1440 cm −1 appeared due to the symmetrical stretching vibration of C=O. The stretching vibration peak of -COOH was redshifted, which indicates that the C(O)O − coordinated with Ce ions to form complex 2 [27]. The peak at 1622 cm −1 is due to the bending vibration of HOH [37]. The symmetrical absorption vibration peak of -NO 3 is between 1300 and 1500 cm −1 [38] (Figure 3).

PXRD Diffraction Analysis
The purity of complexes 1 and 2 were analyzed by comparing the simulated diffraction data (the red one) with the experimental powder diffraction data (the black one). The simulated data and the experimental data of complexes 1 and 2 were consistent, respectively, indicating that the two complexes were both pure ( Figure 4). 3.4. TG-DSC Thermal Analysis of Energetic Complexes 1 and 2 TG-DSC was used to investigate the thermal stability and degradation mechanism of complexes 1 and 2 ( Figure 5). The TG curve of complex 1 is divided into three parts. During the first part, the weight loss ratio of coordinated water was 15.2% from 100 to 207 • C, which was similar to the theoretical value of 15.5%. In the second part, the tza − ligands of complex 1 began to decompose, causing the complex framework to collapse between 207 • C to 450 • C. Significant mass loss occurred during this process, and it is inferred that the complex underwent a violent decomposition reaction within this temperature range. The frame of complex 1 collapsed, decomposed into solid products, and released gaseous products and heat. In the third stage, the TG curve was relatively stable from 450 to 600 • C. The whole process contained two distinct peaks in the DSC curve: one endothermic peak at 137.3 • C and one exothermic peak at 277.7 • C. Mass loss occurred due to the further decomposition of products from the second stage as the temperature continued to increase. The remaining 35.8% weight corresponded to the weight content of La 2 O 3 (calcd. 35.3%). The decomposition peak temperature of complex 1 was 277.7 • C with a heat release of 1549 J·g −1 . The TG data of complex 2 shows that initial weight loss occurred from 79.  Figure 5, it can be seen that the exothermic peak temperature of complex 2 is 254.8 • C, with a decomposition heat of 930.1 J·g −1 . The experimental data reveals that both complex 1 and 2 have high decomposition temperatures and release high amounts of energy, indicating that they are thermally stable energetic complexes.

Effect of Energetic Complexes 1 and 2 on Thermal Degradation of HMX
Complexes 1 and 2 were mixed with HMX at a 1:19 mass ratio to investigate how they affected the thermal degradation of HMX using DSC. From the DSC curve of pure HMX, it can be seen that the decomposition process contains two endothermic peaks and one exothermic peak ( Figure 6). One of the endothermic peaks at 194.3 • C was attributed to the crystal transformation of HMX, while the other endothermic peak at 278.2 • C was due to the liquid phase degradation of HMX [39]. The exothermic peak occurred at 282.6 • C, with a heat release of 1198.3 J·g −1 , which is nearly identical to the decomposition process of HMX reported in previous studies [18,40]. Energy can be obtained by integrating the area of Figure 6. By using the following formula, the actual increase or decrease of the heat release of the system can be obtained: where ∆H is the actual added heat release of the system, H t is the total heat release of the mixed system, H n is the heat release of the complex 1 or 2 and H 0 is the heat release per gram of HMX. However, the effects of compound 1 and 2 for HMX thermal decomposition could be explained as follows: (1) In the thermal decomposition processes of HMX with 1 and 2, respectively, fresh metal oxides at the molecule level can be formed ( Figure 5). The in-situ formed oxides with a high specific surface have high density of active sites on the surface and can play an important catalytic role. On the other hand, those fresh metal oxide powders would trap HMX as more adsorption sites formed. Such effects would make HMX less stable and result in the decrease of the decomposition peak temperatures of HMX and HMX mixtures. Furthermore, the formed fresh metal oxides can also adsorb more pyrolysis products of HMX and lead to the release of the adsorption heat [41]. (2) The main gaseous products of HMX are CH 2 O, NO 2 , CO 2 , NO, CO, and N 2 O [42][43][44]. It is noted that the thermal decomposition products of HMX are similar to those of automobile exhaust. Owing to La 2 O 3 and Ce 2 O 3 being able to catalyze the reactions of carbon oxides and nitrogen oxides in automobile exhaust [30,31]. One can conclude that a serious of exothermic reactions between the different gaseous products could be catalyzed with those fresh metal oxide mixtures, including the oxidation reaction of CO, the reaction between CO and NO, and so on. That could lead to the increase of heat release.
Compared with the DSC curve in Figure 6, the position of the first endothermic peak of HMX was nearly unchanged after complex 1 was added, which indicates that the addition of complex 1 had no effect on the crystal transition process of HMX. As an additive, complex 1 not only reduced the exothermic peak temperature of HMX by 2.2 • C, but also increased the heat release of HMX by 129.9 J·g −1 (Table 1). Therefore, the complex could promote the thermal degradation of HMX. When complex 2 was added, the exothermic peak temperature of HMX increased 1.3 • C, and the amount of liberated heat increased by 41.5 J·g −1 . In addition, as an HMX addictive, the decomposition peak temperature reduction and heat release of the La complex (complex 1) is more and higher than that of some transition metal complexes, such as [Ag(tza)] n (3), [Cu(tza) 2 ] n (4) and [Zn(tza) 2 ] n (5) [45], under the same conditions. Table 1. Promoting effect of complexes 1 to 5 on the thermal decomposition of HMX [29,45]. Since the two energetic complexes lowered the exothermic peak temperature of HMX and increased the amount of heat released, each can act as accelerators for the thermal decomposition of HMX. This suggests they may also be excellent combustion promoters of NEPE. Complex 1 exhibits a slightly better effect on the thermal degradation of HMX than 2 and lowers the decomposition temperature of HMX more and releases more heat under the same experimental conditions. Therefore, in this study, the compatibility between HMX and complex 1 was chosen for further study.
3.6. Compatibility between HMX and Complex 1 3.6.1. DSC Analysis As shown in Figure 7, the difference between the exothermic peak temperature (∆T p ) of HMX and the mixture of complex 1 and HMX is 1.1 • C (i.e., less than 2 • C). According to the compatibility evaluation criteria of explosives and contact materials in Table 2, it is known that complex 1 has good compatibility with HMX and is safe for use in any explosive design when the exothermic peak temperature difference (∆T p ) is less than 2 • C [46][47][48].   4 ] n with HMX was further studied using FTIR [35,36]. In Figure 8a, due to the bending vibration and association of O-H from crystalline and coordinated water, complex 1 exhibited a wider absorption peak near 3155 cm −1 . The asymmetric vibrational C=O peaks near 1627 and 1587 cm −1 show that the O atom in C(O)O − was successfully coordinated with the La ion. The absorption peak near 1454 cm −1 appeared due to the stretching vibration of C=N in the tetrazole ring. In addition, the absorption peaks near 1332 cm −1 indicate that the La metal ion was successfully coordinated with NO 3 − [27]. In Figure 8c, the peaks near 3036 cm −1 were due to C-H tensile vibration. The peaks near 1523 cm −1 and 759 cm −1 were due to -NO 2 asymmetric, symmetric, and bending vibrations, respectively. The peaks near 1260 cm −1 were due to the stretching vibration of N-N [49]. In Figure 8b, no new absorption peak appeared, and the position of the peaks was nearly unchanged compared with those of HMX and complex 1.
The absorption spectrum in Figure 8b was obtained by superimposing Figure 8a,c. Therefore, it was concluded that there was no chemical interaction between the pair of components [35]. Consequently, at room temperature, complex 1 does not interact with HMX, showing that the two compounds have good compatibility.

Conclusions
Two novel, thermally-stable rare-earth energetic complexes [La(tza)(NO 3 ) 2 (H 2 O) 4 ] n and [Ce(tza)(NO 3 ) 2 (H 2 O) 4 ] n were prepared and evaluated. The two complexes were shown to promote the thermal degradation of HMX. Complex 1 showed a slightly better effect on the thermal decomposition of HMX than 2, and it was also compatible with HMX. These results show that [La(tza)(NO 3 ) 2 (H 2 O) 4 ] n may be a more efficient combustion promoter for NEPE propellant. This study could provide beneficial ideas for the rational design and preparation of new energetic promoters for solid propellants.

Conflicts of Interest:
The authors declare no conflict of interest.