Skip to main content
NCBI home page
ACS Omega logoLink to ACS Omega
. 2020 Sep 1;5(36):23193–23200. doi: 10.1021/acsomega.0c02968

Thermal Decomposition Mechanism of 1,3,5,7-Tetranitro-1,3,5,7-tetrazocane Accelerated by Nano-Aluminum Hydride (AlH3): ReaxFF-Lg Molecular Dynamics Simulation

Ying Zhao †, Zheng Mei †, Feng-Qi Zhao ‡, Si-Yu Xu ‡, Xue-Hai Ju †,*
PMCID: PMC7495736  PMID: 32954170

Abstract

graphic file with name ao0c02968_0018.jpg

ReaxFF-low-gradient reactive force field with CHONAl parameters is used to simulate thermal decomposition of 1,3,5,7-tetranitro-1,3,5,7-tetrazocane (HMX) and AlH3 composite. Perfect AlH3 and surface-passivated AlH3 particles were constructed to mix with HMX. The simulation results indicate HMX is adsorbed on the surface of particles to form O–Al and N–Al bonds. The decomposition of HMX and AlH3 composite is an exothermic reaction without energy barrier, but the decomposition of pure HMX needs to overcome the energy barrier of 133.57 kcal/mol. Active nano-AlH3 causes HMX to decompose rapidly at low temperature, and the primary decomposition pathway is the rupture of N–O and C–N bonds. Adiabatic simulation shows that the energy release and temperature increase of HMX/AlH3 is much larger than those of the HMX system. Surface-passivated AlH3 particles only affect the initial decomposition rate of HMX. In HMX and AlH3 composites, the strong attraction of Al in AlH3 to O and the activation of the intermediate reaction by H2 cause HMX to decompose rapidly. The final decomposition products of pure HMX are H2O, N2, and CO2, and those of HMX/AlH3 are H2O, N2, and Al-containing clusters dominated by C–Al. The final gas production shows that the specific impulse of HMX/AlH3 is larger than that of HMX.

1. Introduction

Energetic materials (EMs) consisting of CHON elements are a class of significant substances containing a large amount of chemical energy. They can release huge amounts of energy after decomposition, which make them widely used in military and aerospace fields.1 Nevertheless, with the improvement of the requirement of comprehensive properties of EMs, conventional EMs cannot gradually meet the requirements of high energy density materials. As a result, a growing number of researchers are working to synthesize EMs with better properties. Alternatively, the addition of high active metal additives is an effective way to improve the properties of EMs.2 Research shows that EMs with high active metal additives have higher detonation and combustion properties.3−5 Commonly used high active metal additives include aluminum, magnesium, boron, and so forth. However, some studies have found that metal hydrides can replace metals as better additives to mix with EMs.6,7 Compared with metal power, metal hydrides possess high hydrogen storage capacity and higher combustion heat. Their decomposition can release large amounts of hydrogen, which allows them to release more energy and decline the average molecular weight of the gas in the combustion process.8,9 Common metal hydrides include MgH2, CaH2, AlH3, and so forth. Among these metal hydrides, the application of AlH3 is considered important for the development of EMs because of its high hydrogen storage capacity (10.1%).10,11 At present, AlH3 has been successfully applied in solid propellants as an active additive. Compared with traditional propellants, the performance of propellants containing AlH3 is greatly improved. For instance, Deluca et al. found that adding AlH3 into propellants could increase the specific impulse.12 Moreover, in the application of active additive, the size of the particles also has an important effect on its properties. Some investigations have found that the reactivity of nano-Al powder obviously increases as the size decreases.13−15 Hence, it is important to study the influence of micron or even nanometer AlH3 on explosives. However, because of the hazard of EMs and the instability of AlH3, it is difficult to explore the mechanism of action of AlH3 on explosives by experiments. Therefore, it is necessary to study the application of nano-AlH3 in EMs theoretically. Reactive molecular dynamics (RMD) simulation is an effective method to investigate the reaction mechanism between nano-AlH3 and EMs.

RMD can favorably describe the performance of large-scale systems and study the chemical reactions that may exist. It can simulate the free rupture and generation of chemical bonds between atoms, making it possible to analyze the chemical reaction process in the system. Besides, compared with quantum chemical methods, RMD has excellent simulation speed. RMD is primarily simulated using the ReaxFF force field.16 Furthermore, in order to describe the long-range interaction of atoms, ReaxFF-lg that uses a low-gradient (lg) model to correct long-range effects is developed based on ReaxFF.17 At present, ReaxFF-lg is being successfully applied to the study the thermal decomposition of EMs.18−21

1,3,5,7-Tetranitro-1,3,5,7-tetrazocane (HMX) is an EM with high detonation and thermal stability. It has extensive application in the field of propellants and explosives because of its outstanding energy characteristics. So far, researchers have conducted many theoretical and experimental studies on the decomposition mechanism of HMX.22−25 However, there has been only limited studies on the effect of AlH3 on the thermal decomposition of HMX. Studying the interaction mechanism between these two significant components of the propellant is important for their application.

In this paper, the ReaxFF-lg reaction force field with CHONAl parameters was used to simulate the thermal decomposition of HMX and HMX/AlH3 systems. The decomposition pathways, intermediates, products, evolution of AlH3 clusters, and diffusion coefficients of each atom were analyzed through the C++ program developed by our group.

2. Computational Methods and Details

A ReaxFF-lg reactive force field containing CHONAl parameters was used to perform the RMD of HMX and HMX/AlH3 systems. It has been proved that this force field is suitable for studying the thermal decomposition of high EM systems containing CHONAl.26 LAMMPS, the large-scale atomic molecular massively parallel simulator, was used to perform all simulations.27 ReaxFF-lg uses the following equation to determine the energy and then the force on each atom

2.

where Ebond is the bond energy; Elp is the lone pair energy; Eover and Eunder are over- and under-coordination corrections, respectively; Eangle is the bond angle energy; Edouble and Etriple are double- and triple-bond corrections; Etors is the dihedral angle energy; Econj is the conjugation energy; EHbond is the hydrogen bond energy; EvdWaals is the van der Waals energy; and ECoulomb is the electrostatic energy.

In this work, HMX supercell was derived from the extended original cell of β-HMX (P21̅n). The orthogonal HMX(001) supercell was built through Materials Studio 7.0 tools.28 The spherical AlH3 nanoparticle with a diameter of 4.0 nm was obtained by cutting the supercell obtained from the extended α-AlH3 crystal. Moreover, in practical applications, in order to keep them reactive, the nano-AlH3 particles are usually covered with an oxide layer. However, the diameter of the nano-AlH3 model constructed in this work is only 4 nm, which is far smaller than the size of the nano-AlH3 particles in practice (about 100 nm). Therefore, in order to maintain the reactivity of aluminum hydride, only core–shell aluminum hydride particles coated with 0.1 nm oxide layer were established. First, MS tool was used to cut aluminum oxide balls with a diameter of 4.8 nm. A self-made C++ program was used to embed the 4 nm aluminum hydride into aluminum oxide with internal vacuum to form aluminum hydride nanoparticles coated with 0.1 nm aluminum oxide on the surface. The constructed core–shell aluminum hydride particles were fully relaxed under the NVT ensemble of 30 ps. The result is the core–shell aluminum hydride particles (AlH3O) shown in Figure 1. In order to embed aluminum hydride particles in the center of the box, the orthogonal HMX(001) supercell was built, and then the AlH3 and AlH3O particles were embedded into the HMX(001) supercell to build HMX/AlH3 and HMX/AlH3O composite systems. The thickness of the vacuum layer between both particles and HMX molecules is 0.2 nm. Figure 1 presents the model of the HMX/AlH3 composite. Molecular dynamics simulation was performed to relax the HMX, HMX/AlH3, and HMX/AlH3O systems in the isobaric–isothermal (NPT) ensemble up to 20 ps (300 K). The properties before and after the relaxation of the two systems are listed in Table 1; here in, aluminum and hydrogen in the aluminum hydride particle accounted for 19.74 and 23.35% of HMX/AlH3 and HMX/AlH3O, respectively. In practical application, the additive proportion of aluminum hydride can reach 10–30% in weight. After that, a canonical ensemble (NVT) was used to simulate the thermal decomposition of both the models. In the NVT simulation process, the initial temperature was set as 300 K and then increased to 3500 K within 100 ps, followed by 100 ps constant temperature simulation at 3500 K. The main reason is that 3500 K is more in line with the temperature after the explosive explosion. On the other hand, the decomposition of the system is more complete at 3500 K. The time step was 0.1 fs. A Nose–Hoover barostat and a thermostat were used to control the pressure and temperature during the NPT and NVT simulations. The charge distribution method is the QEq method, and the tolerance value is set as 1 × 10–6. The coupling parameter is 10. Moreover, 30 ps NVE was used to simulate the ignition reasonably for HMX and HMX/AlH3 in highly confined spaces almost without mass and heat transfer. In the species analysis, the bond-length cutoff is 1.2 times the standard single bond length. When the distance between atoms is larger than the standard bond length, the bond is considered ruptured.

Figure 1.

Figure 1

Open in a new tab

Construction process of the composite model. C, H, O, and N atoms are shown in gray, white, red, and blue, respectively. AlH3 and AlH3O particles after embedding are shown in yellow and green, respectively.

Table 1. Properties of HMX and HMX/AlH3 Systems.

      density (g/cm3)
model composition mass fraction of AlH3 particles (%) initial relaxed
HMX 420 HMX   1.96 2.04
HMX/AlH3 411 HMX + Al997H3000 19.74 1.50 1.94
HMX/AlH3 455 HMX + Al1589H3006O842 23.35 1.65 1.90
Open in a new tab

3. Results and Discussion

3.1. Applicability of the Calculation Method

Radial distribution function can provide information about the probability of finding an atom in a spherical shell at a distance r from an arbitrary atom. According to the distribution of atomic probability at r, we can understand the interaction distance between atomic pairs in the crystal, that is, the length of the chemical bond between the two atoms. Hence, the 3 × 3 × 3 of AlH3, 2 × 2 × 2 of α-Al2O3, 3 × 3 × 3 of Al4C3, 3 × 3 × 3 of AlN, 3 × 3 × 3 of Al, and 3 × 3 × 3 of HMX supercells were established to relax under a 30 ps NPT ensemble; here in, all crystal protocells come from the crystallography open database. The final 5 ps structure of the supercell was used to analyze the radial distribution function between atomic pairs. In the radial distribution function, the first peak tends to represent the nearest pair of atoms, that is, the length of the bond in the crystal. Figure 2 presents the radial distribution function between atomic pairs in the HMX crystal. In the HMX supercell, the first peak of rC,H is 1.075–1.175 Å, and the experimental of the HMX crystal shows that the C–H bond length range is 1.093–1.0981 Å. The first peak of rC,N is 1.465–1.525 Å, and the C–N bond length in the experimental is 1.436–1.480 Å. The first peak of rN,O is 1.275–1.315 Å, and the N–O bond length in the experimental is 1.252–1.260 Å. The first peak of rN,N is 1.485–1.525 Å, and the N–N bond length in the experimental is 1.374–1.525 Å. To sum up, it can be observed that there is little difference between the experimental value and the calculated value. The results show that the method is suitable for HMX system simulation.

Figure 2.

Figure 2

Open in a new tab

Radial distribution function between atomic pairs in the HMX crystal.

Figure 3 presents the calculated radial distribution function between atomic pairs of AlH3, α-Al2O3, Al4C3, AlN, and Al crystals. In α-Al2O3, the first peak between O and Al occurs at 1.815–2.095 Å. Based on the experimental crystal structure of α-Al2O3, the minimum and maximum values of O–Al bond length are 1.843 and 1.982 Å, respectively. The first peak of rO,Al is 1.815–2.025 Å. In AlH3, the first peak of rH,Al is 1.825–1.915 Å, and the minimum and maximum values of the H–Al bond length in Al4C3 are 1.769 and 1.784 Å. In Al4C3, the first peak of rC,Al occurs at 1.875–2.225 Å, and the minimum and maximum values of the C–Al bond length in the Al4C3 crystal are 1.898 and 2.223 Å. In AlN, the first peak of rN,Al is 1.895–2.025 Å, and the minimum and maximum values of the C–Al bond length in the Al4C3 crystal are 1.893 and 1.894 Å. In Al, the largest peak of rAl,Al is 2.855 Å, and the length of the Al–Al bond in the Al crystal is 2.852 Å. Hence, the calculated value is close to the experimental value. This further confirms the accuracy of the method used in the simulation in describing the interaction with Al atoms.

Figure 3.

Figure 3

Open in a new tab

Calculated radial distribution function between atomic pairs of AlH3, α-Al2O3, Al4C3, AlN, and Al crystals. r is the atomic distance.

3.2. Adsorption of HMX on the Surface of AlH3 Particles in NPT Simulation

In the relaxation process of HMX/AlH3 and HMX/AlH3O systems, both AlH3 particle surfaces strongly attract HMX molecules, causing some HMX molecules to absorb on the AlH3 particle surface. Figure 4 shows that the AlH3 particles of HMX were adsorbed on the surface. Figure 5 shows the variation of bond numbers of HMX/AlH3 and HMX/AlH3O systems during 20 ps NPT simulation. It can be observed that the HMX molecules are adsorbed on the surface of both AlH3 particles, and a large number of Al–O bonds and a handful of N–Al bonds are produced. The phenomenon suggests that the strong attraction between AlH3 and HMX comes from the strong attraction of Al to the O atom and the medium attraction to the N atom. In addition, a part of the N–O bonds of the HMX molecule rupture in both systems, and then the dissociative O atoms are attracted to Al to form the Al–O–Al and HO–Al bonds. It is interesting to note that the N–NO2 bond as the decomposition trigger bond in HMX is not ruptured. Figure 4 shows the reduction proportion in the HMX quantity and released energy during NPT simulation. It can be seen from Figure 6 that the adsorption reaction between HMX and aluminum hydride particles results in the decomposition of HMX, and the adsorption process was a spontaneous exothermic reaction. This spontaneous reaction is caused by the strong attraction of Al to the O atoms in HMX. Compared with HMX/AlH3, the increasing number of O–Al and N–Al bonds and the decreasing number of N–O bonds in the HMX/AlH3O system are smaller. This indicates that the presence of oxide layer, to a certain extent, inhibits the reactivity between aluminum hydride particles and HMX in the adsorption stage. Nevertheless, the HMX molecule reacts with both the particles depending on the attraction of Al to oxygen. This can be judged from the large increase of O–Al bonds between the two systems during the adsorption phase. In addition, the adsorption reaction stops after 15 ps because of the stable passivation layer formed by the adsorption of HMX on the surface of aluminum hydride. These results indicate that HMX/AlH3 composites can exist stably under certain conditions.

Figure 4.

Figure 4

Open in a new tab

Sectional views of HMX molecules on the surface of AlH3 particles after 20 ps NPT simulation. C, H, O, N, and Al atoms are shown in gray, white, red, blue, and pink, respectively.

Figure 5.

Figure 5

Open in a new tab

Variation of bond numbers of HMX/AlH3 and HMX/AlH3O systems during 20 ps NPT (300 K) simulation.

Figure 6.

Figure 6

Open in a new tab

Variation of HMX and PE in NPT (300 K) simulation. HMX loss refers to the ratio of the number of HMX to the initial number; PE released refers to the ratio of the PE released by the system to the initial PE.

3.3. Heating and Adiabatic Simulations

Figures 7 and 8 present the evolution of HMX molecules and the corresponding potential energy (PE) from 300 to 3500 K within the NVT ensemble. In the PE calculation, all PEs are referenced by the PE values of the relaxed system and divided by the number of HMX molecules in the system. It can be observed that the decomposition of pure HMX begins at about 1100 K. Hereafter, the HMX decomposition rate accelerates rapidly with the increase of temperature. Nevertheless, in the HMX/AlH3 system, the attraction of AlH3 can lead to the decomposition of some HMX molecules in the relaxation process, and HMX/AlH3 starts to decompose after 5 ps heating. The variation of PE shows that the decomposition of HMX/AlH3 is an exothermic reaction without energy barrier. However, there is an energy barrier of 133.57 kcal/mol for the decomposition of pure HMX system. Moreover, it can be noted that HMX/AlH3 shows much higher reaction heat than HMX from Figure 7.

Figure 7.

Figure 7

Open in a new tab

Evolution of HMX molecules during the heating from 300 to 3500 K. System was relaxed within 20 ps NPT (300 K) simulation before NVT simulation.

Figure 8.

Figure 8

Open in a new tab

Evolution of PE from 300 to 3500 K.

In order to explore the decomposition path of these three systems, the evolution of the number of bonds is calculated and illustrated in Figure 9. In terms of the tendency of bonds to decrease in Figure 9 (HMX), the scission of N–N bonds and C–H bond are the main decomposition pathways of pure HMX. It is consistent with other experimental and theoretical studies.29−31 However, the decomposition pathway in HMX/AlH3 and HMX/AlH3O is different from that in pure HMX. It can be noted that the rate of decrease of N–O bonds is significantly higher than that of other bonds from Figure 9 (HMX/AlH3, HMX/AlH3O). The results demonstrated that the scission of N–O bonds is the dominating pathway of HMX in HMX/AlH3 and HMX/AlH3O. Moreover, in HMX/AlH3 and HMX/AlH3O systems, the decreasing rate of C–N bonds becomes relatively quick after the heating up of the system, indicating that AlH3 particles cause the C–N bond to break easily. Nevertheless, in the same simulation time, the reduction of N–O and C–N bonds in the HMX/AlH3O system is much smaller than that in the HMX/AlH3 system. The results show that the oxide layer inhibited the reactivity of aluminum hydride.

Figure 9.

Figure 9

Open in a new tab

Variation of bonds of pure HMX, HMX/AlH3, and HMX/AlH3O.

The NVE simulations can describe the ignition reasonably for EMs in highly confined spaces with less mass and heat transfer. Hence, the adiabatic processes of HMX and HMX/AlH3 systems are simulated by the NVE ensemble. The evolution is preceded by preheating of 2 ps at 1500 K within the NVT ensemble. It is worthwhile to point out that the 1500 K preheating ignites the EM system. Figure 10 presents the evolution of HMX molecules, PE, and temperature within the NVE ensemble. These profiles show clearly that the decomposition rate determines the PE release rate and the temperature increase rate of the adiabatic system. The rate of overall decomposition, energy release, and temperature increase of AlH3/HMX with a highly active AlH3 content are much larger than those of the HMX system. Moreover, Figure 8 clearly exhibits the remarkably low rate of energy release and temperature rise within the 5–15 ps. This is consistent with the fact that HMX decomposition in NVT simulations requires a large number of barriers.

Figure 10.

Figure 10

Open in a new tab

Evolution of HMX molecules, PE, and temperature with time in the NVE ensemble. The evolution is preceded by preheating of 2 ps at 1500 K.

3.4. Evolution of Al-Containing Clusters

The evolution of the components of Al-containing clusters was analyzed to study the role of Al in the decomposition process of HMX/AlH3. Figure S1 shows the evolution curve of the X–Al bond of HMX/AlH3O system. Figure 11 reveals the evolution of X–Al (X presents C, H, O, or N) bond in Al-containing clusters. It can be noted that even though there is a difference in the number of X–Al bonds, it depends on the composition of the system, but the overall evolution trend of the two is consistent. Therefore, the evolution of aluminum-containing clusters in the HMX/Al system is discussed in this section. This is also appropriate for the HMX/AlH3O system. As can be seen from Figure 11, the number of H–Al bonds decreases sharply as the temperature increases. The Al atoms are constantly oxidized by the O atoms in HMX, causing the number of O–Al bonds to increase rapidly to the peak at 90 ps; soon afterward, the amount of O–Al bond decreases gradually. The number of C–Al bonds increases slowly before 60 ps but increases to a peak in a rapid rate from 60 to 130 ps. After that, the quantity of C–Al bonds remains constant. The number of N–Al bonds increases to a peak at 110 ps and then decreases gradually. The above statements indicate that the composition and quantity of the Al-containing clusters vary in the reactions. The O atoms in HMX are strongly attracted to the Al atoms of AlH3. Al atoms are gradually oxidized to form a mass of Al–O bonds. Besides, a handful of C and N atoms in HMX also contact with Al-containing clusters to form a fraction of C–Al and N–Al bonds. With the temperature increasing, Al atoms are oxidized to saturation, forming Al-containing clusters dominated by Al–O bonds. After that, C, as the reducing agent, makes Al–O convert into Al–C cluster. Hence, the number of O–Al bonds decreases, while the number of C–Al bonds increases. After the Al–O cluster is replaced basically by the C–Al cluster, the excess C in the system begins to react with the N–Al cluster, causing the quantity of N–Al bonds to decrease. As a result, Al-containing clusters, which are dominated by C–Al products, were formed after 150 ps.

Figure 11.

Figure 11

Open in a new tab

Evolution of X–Al (X presents C, H, O, or N) bond of Al-containing clusters in the NVT simulation.

Figure 12 presents the configurations of Al-containing clusters with HMX molecules in the NVT simulation. Herein, OVITO program is used to analyze the variation of Al-containing clusters, and the cutoff distance of the cluster is set to 3.5 Å.32 As can be seen from Figure 12, in the process of reaction between HMX molecules and AlH3 particles, CHO atoms in the HMX molecule gradually enter the inside of AlH3 particles to oxidize Al atoms with the increase of temperature, while the Al in AlH3 spreads outward. This result causes holes to form inside the AlH3 particles. Figure 13 presents the variation of the number of Al-containing clusters. It can be observed that the number of Al-containing clusters increases dramatically in 0–30 ps because the structure of AlH3 is destroyed as temperatures rise and oxygen atoms penetrate. After that, Al-containing clusters are gradually oxidized to form alumina. Despite the increase of temperature, it is difficult to make a significant change in the number of Al-containing clusters within 30–80 ps. However, the replacement reaction mentioned above and the high temperature (>2220 K) after 60 ps cause a sharp increase in the number of Al clusters. The substitution of C and the high temperature destroy the structure of alumina, causing the aluminum clusters to melt, disperse, and agglomerate (Figure 12; 100, 110, and 120 ps). When the temperature reaches 3500 K, the C–Al clusters fuse at high temperature (Figure 10; 150 ps). As a result, the number of Al-containing clusters decline rapidly. The reaction of C and N–Al clusters makes the number of Al-containing clusters increase slowly until the Al-containing clusters dominated by C–Al bond are formed.

Figure 12.

Figure 12

Open in a new tab

Configurations of Al-containing clusters with HMX molecules in the NVT simulation. C, H, O, N, and Al atoms are shown in gray, white, red, blue, and pink, respectively.

Figure 13.

Figure 13

Open in a new tab

Variation of the number of Al-containing clusters in the NVT simulation. Cutoff distance of the neighbor cluster model is 3.5 Å.

3.5. Stable Products and Major Intermediates

Figure 14 illustrates the evolution of stable products in HMX and HMX/AlH3 systems. Figure S2 shows the evolution of stable gaseous products of the HMX/AlH3O system. It can be noted that the evolution of gaseous products is basically consistent in HMX/AlH3 and HMX/AlH3O systems. Hence, only HMX and HMX/AlH3 systems are discussed. Consistent with previous experimental and theoretical studies, the final stable products of HMX system thermal decomposition are N2, H2O, and CO2,28 of which N2 is the largest amount, while CO2 is the least amount. These gas products start to form after 50 ps (at 1200 K), and accumulate rapidly to a peak. Comparatively, the amount of stable products in HMX/AlH3 is different from that in the HMX system. In addition to the formation of Al-containing clusters, it is observed that H2O is the most abundant product in HMX/AlH3, and its amount is much larger than that in the HMX system. However, the amount of N2 is reduced, and no CO2 is produced. Moreover, in the HMX/AlH3 system, N2 and H2O can be produced rapidly at low temperatures. As temperatures rise, these gases are produced at a high rate before 100 ps,;after that, they achieve the maximum (150 ps) yields in a relatively slow rate. The productions of H2O and N2 after 100 ps are related to the decrease of N–Al and O–Al clusters discussed previously in 3.1 (Figure 10). The formation of H2O indicates that H2 also reacts with O–Al clusters. Consequently, the peaks of product numbers are delayed in the HMX/AlH3 system in comparison with pure HMX. Meanwhile, the results also show that the combustion time in HMX/AlH3 has been extended.

Figure 14.

Figure 14

Open in a new tab

Evolution of products in NVT simulation.

These stable gas productions have important influences on the performance of the propellant. Although the amounts of N2 and CO2 in the HMX/AlH3 system are reduced, the amount of H2O increases significantly, and the total gas yield of HMX/AlH3 is larger than that of the HMX system. Furthermore, HMX/AlH3 system reduces emission of carbon dioxide; it is favorable for the protection of the environment. Figure S3 presents the evolution of pressures under the NVT ensemble. It can be noted that the rate of pressure increase and the final stability value in the HMX/AlH3 system are larger than those in the HMX system.

Figure 15 presents the evolutions of the main intermediates during the decomposition process of HMX and HMX/AlH3. Figure S4 reveals the evolution of intermediates during the decomposition of the HMX/AlH3O system. In the initial decomposition of HMX, NO2 and HNO2 are produced first as important intermediates. The results show that the scission of N–NO2 and C–H bonds are the main pathways for the initial decomposition of HMX. Moreover, it is observed obviously that these intermediates are mainly composed of O-containing free radicals and unstable products containing O atoms in pure HMX, indicating that O atom plays an important role in the process of HMX decomposition. In HMX/AlH3 and HMX/AlH3O, H2 as the most important intermediate is produced first due to the decomposition of AlH3 at low temperature. Although there are O-containing intermediates in the decomposition of HMX/AlH3 and HMX/AlH3O, their amounts are significantly reduced. Among these O-containing intermediates, OH is the most abundant and the first to be produced in HMX/AlH3 and HMX/AlH3O, which is different from that in HMX. The results indicate that the decomposition pathway of HMX in HMX/AlH3 and HMX/AlH3O systems is different from that of pure HMX. Considering the evolution of Al–X discussed above, it can be known that the strong attraction of AlH3 particles to O atoms causes HMX to decompose prematurely. Meanwhile, the binding effect of Al on CNO atoms makes these atoms difficult to interact with each other to form O-containing intermediates, and the active H2 reacts easily with O to form OH, causing OH to become the most intermediate in the decomposition process of HMX/AlH3 and HMX/AlH3O.

Figure 15.

Figure 15

Open in a new tab

Evolution of intermediates in the decomposition process of HMX and HMX/AlH3.

3.6. Diffusion Analysis

The diffusion analysis can quantitatively describe the atomic interactions in the decomposition process. Hence, the diffusion coefficients of all atoms in these two systems were calculated. Diffusion coefficient is calculated by Einstein’s diffusion equation, which is mean square displacement (MSD) divided by time and then divided by 6 (three-dimensional diffusion). The MSD is the average of displacement of each type of atoms, as ⟨|r(t) – r(0)|2⟩. r(0) is the initial coordination of an atom. In this work, the MSD is computed every 1 picosecond to calculate the current diffusion coefficient. Figure 16 presents the average diffusion coefficient of atoms in HMX and HNM/AlH3 systems. The magnitude of diffusion coefficient is related to the relative atomic mass. Therefore, the diffusion coefficient of H atom is the largest in both systems. Nevertheless, although O has a larger relative atomic mass than C and N, the diffusion coefficients of C and N are less than that of O. The phenomenon indicates that there are other factors influencing the diffusion of atoms in the decomposition process of HMX. In combination with the intermediates in HMX decomposition discussed earlier, O atoms are involved in more intense chemical reactions. Meanwhile, it can be observed that the diffusions of C, N, and O in HMX are larger than those in HMX/AlH3. This is because some of the CNO atoms are attracted to Al, making them difficult to spread. However, the diffusion of H in HMX/AlH3 is larger than that in HMX. This is because H2 produced by decomposition of AlH3 participates in violent chemical reactions. However, C is easily deposited on aluminum particles, and its diffusion coefficient decreases most obviously. The result is consistent with the phenomenon that the final Al-containing cluster is dominated by C–Al.

Figure 16.

Figure 16

Open in a new tab

Average diffusion coefficient of different atoms in NVT simulation.

4. Conclusions

The ReaxFF-lg reactive field was used to simulate the thermal decomposition process of HMX and aluminum hydride nanocomposite system. The simulation results show that aluminum hydride changes the initial decomposition mechanism of the HMX molecule. HMX was first adsorbed on the surface of hydrogenated aluminum particles and then decomposed by a new way of N–O bond and C–N bond. The strong attraction of aluminum to oxygen and the medium attraction to nitrogen are the main reasons for the change of HMX initial decomposition mechanism. The increase of O–Al and N–Al bonds in the evolution of aluminum-containing clusters confirms this view. During the oxidation of aluminum hydride particles, a large number of oxygen atoms and a small number of N atoms permeate into the interior of aluminum hydride particles. The aluminum atoms of aluminum hydride spread out. The internal structure of aluminum hydride is destroyed, forming hydrogen accumulation cavity. As the temperature increases, hydrogen breaks through the outer oxide layer and participates in the intermediate reaction. Because of the ability of H2 to compete for oxygen atoms, the H2O yield of HMX/AlH3 system increases and the CO2 yield decreases. The adiabatic simulation indicates that energy release and temperature increase of AlH3/HMX are much larger than those of the HMX system.

Acknowledgments

Y.Z. gratefully thanks the Postgraduate Innovation Project of Jiangsu Province for the partial financial support.

Supporting Information Available

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.0c02968.

  • Evolution of X–Al (X presents C, H, O, or N) bond of Al-containing clusters in HMX/AlH3O; evolution of products of HMX/AlH3O in NVT simulation; evolution of system pressure during NVT simulation; and evolution of intermediates in the decomposition process of HMX/AlH3O (PDF)

The authors declare no competing financial interest.

Supplementary Material

ao0c02968_si_001.pdf (255.4KB, pdf)

References

  1. Chandrasekaran N.; Oommen C.; Kumar V. R. S.; Lukin A. N.; Abrukov V. S.; Anufrieva D. A. Prediction of Detonation Velocity and N–O Composition of High Energy C–H–N–O Explosives by Means of Artificial Neural Networks. Propellants, Explos., Pyrotech. 2019, 44, 579–587. 10.1002/prep.201800325. [DOI] [Google Scholar]
  2. Gottfried J. L.; Bukowski E. J. Laser-shocked energetic materials with metal additives: evaluation of chemistry and detonation performance. Appl. Opt. 2017, 56, B47–B57. 10.1364/ao.56.000b47. [DOI] [PubMed] [Google Scholar]
  3. Zhi J.; Shu-Fen L.; Feng-Qi Z.; Zi-Ru L.; Cui-Mei Y.; Yang L.; Shang-Wen L. Research on the combustion properties of propellants with low content of nano metal powders. Propellants, Explos., Pyrotech. 2006, 31, 139. 10.1002/prep.200600021. [DOI] [Google Scholar]
  4. Babuk V. A.; Dolotkazin I. N.; Sviridov V. V. Simulation of agglomerate dispersion in combustion of aluminized solid propellants. Combust., Explos. Shock Wave 2003, 39, 195–203. 10.1023/a:1022917201681. [DOI] [Google Scholar]
  5. Maggi F.; Dossi S.; Paravan C.; DeLuca L. T.; Liljedahl M. Activated aluminum powders for space propulsion. Powder Technol. 2015, 270, 46–52. 10.1016/j.powtec.2014.09.048. [DOI] [Google Scholar]
  6. Liu J.; Yang W.; Liu Y.; Liu X. Mechanism and characteristics of thermal action of HMX explosive mixture containing high-efficiency fuel. Sci. China: Technol. Sci. 2019, 62, 578–586. 10.1007/s11431-018-9464-8. [DOI] [Google Scholar]
  7. Xue B.; Lin M.; Ma H.; Wang Y.; Shen Z. Energy Performance and Aging of RDX-based TiH2, MgH2 Explosive Composites. Propellants, Explos., Pyrotech. 2018, 43, 671–678. 10.1002/prep.201700269. [DOI] [Google Scholar]
  8. Lawrence A. R.; Laktas J. M.; Place G. J.; Jelliss P. A.; Buckner S. W.; Sippel T. R. Organically-Capped, Nanoscale Alkali Metal Hydride and Aluminum Particles as Solid Propellant Additives. J. Propul. Power 2019, 35, 736–746. 10.2514/1.b37409. [DOI] [Google Scholar]
  9. Elmas U.; Bedir F.; Kayfeci M. Computational analysis of hydrogen storage capacity using process parameters for three different metal hydride materials. Int. J. Hydrogen Energy 2018, 43, 10741–10754. 10.1016/j.ijhydene.2017.12.057. [DOI] [Google Scholar]
  10. Bazyn T.; Eyer R.; Krier H.; Glumac N. Combustion characteristics of aluminum hydride at elevated pressure and temperature. J. Propul. Power 2004, 20, 427–431. 10.2514/1.3160. [DOI] [Google Scholar]
  11. Il’in A. P.; Bychin N. V.; Gromov A. Products of combustion of aluminum hydride in air. Combust., Explos. Shock Waves 2001, 37, 490–491. 10.1023/a:1017973601611. [DOI] [Google Scholar]
  12. Wang L.; Rawal A.; Aguey-Zinsou K.-F. Hydrogen storage properties of nanoconfined aluminium hydride (AlH3). Chem. Eng. Sci. 2019, 194, 64–70. 10.1016/j.ces.2018.02.014. [DOI] [Google Scholar]
  13. Deluca L. T.; Galfetti L.; Severini F.; Rossettini L.; Meda L.; Marra G.; D’Andrea B.; Weiser V.; Calabro M.; Vorozhtsov A. B.; Glazunov A. A.; Pavlovets G. J. Physical and ballistic characterization of AlH3-based space propellants. Aerosp. Sci. Technol. 2007, 11, 18–25. 10.1016/j.ast.2006.08.010. [DOI] [Google Scholar]
  14. Weiser V.; Kelzenberg S.; Eisenreich N. Influence of the metal particle size on the ignition of energetic materials. Propellants, Explos., Pyrotech. 2001, 26, 284–289. . [DOI] [Google Scholar]
  15. Li C.-F.; Mei Z.; Zhao F.-Q.; Xu S.-Y.; Ju X.-H. Molecular dynamic simulation for thermal decomposition of RDX with nano-AlH3 particles. Phys. Chem. Chem. Phys. 2018, 20, 14192–14199. 10.1039/c8cp01621f. [DOI] [PubMed] [Google Scholar]
  16. van Duin A. C. T.; Dasgupta S.; Lorant F.; Goddard W. A. ReaxFF: A reactive force field for hydrocarbons. J. Phys. Chem. A 2001, 105, 9396–9409. 10.1021/jp004368u. [DOI] [Google Scholar]
  17. Liu L.; Liu Y.; Zybin S. V.; Sun H.; Goddard W. A. ReaxFF-lg: correction of the ReaxFF reactive force field for London dispersion, with applications to the equations of state for energetic materials. J. Phys. Chem. A 2011, 115, 11016–11022. 10.1021/jp201599t. [DOI] [PubMed] [Google Scholar]
  18. Zhang Y.; Liu H.; Yang Z.; Li Q.; He Y. Shock-induced hot spot formation and spalling in 1,3,5-trinitroperhydro-1,3,5-triazine containing a cube void. ACS Omega 2019, 4, 8031–8038. 10.1021/acsomega.9b00589. [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Wen Y.; Zhang C.; Xue X.; Long X. Cluster evolution during the early stages of heating explosives and its relationship to sensitivity: a comparative study of TATB, beta-HMX and PETN by molecular reactive force field simulations. Phys. Chem. Chem. Phys. 2015, 17, 12013–12022. 10.1039/c5cp00006h. [DOI] [PubMed] [Google Scholar]
  20. Hai L.; Qi-Kai L.; Yuan-Hang H. Pyrolysis of CL20-TNT cocrystal from ReaxFF/lg reactive molecular dynamics simulations. Acta Phys. Sin. 2013, 62, 208202. 10.7498/aps.62.208202. [DOI] [Google Scholar]
  21. Zhai P.; Shi C.; Zhao S.; Mei Z.; Pan Y. Molecular dynamics simulations of a cyclotetramethylene tetra-nitramine/hydrazine 5,5 ’-bitetrazole-1,1 ’-diolate cocrysta. RSC Adv. 2019, 9, 19390–19396. 10.1039/c9ra02966d. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Patidar L.; Khichar M.; Thynell S. T. Identification of initial decomposition reactions in liquid-phase HMX using quantum mechanics calculations. Combust 2018, 188, 170–179. 10.1016/j.combustflame.2017.09.042. [DOI] [Google Scholar]
  23. Liu H.; Li Y.; Li Y.; Ma Z.; Zhou Z.; Li J.; He Y. Study on the Initial Decomposition Mechanism of Energetic Co-Crystal 2, 4, 6, 8, 10, 12-Hexanitro-2, 4, 6, 8, 10, 12-Hexaazaiso-Wurtzitane (CL-20)/1, 3, 5, 7-Tetranitro-1, 3, 5, 7-Tetrazacy-Clooctane (HMX) under a Steady Shock Wave. Acta Phys.-Chim. Sin. 2019, 35, 858–867. 10.3866/pku.whxb201812011. [DOI] [Google Scholar]
  24. Manelis G. B.; Nazin G. M.; Rubtsov Y. I.; Strunin A. A.. Thermal Decomposition and Combustion of Explosives and Propellants; Taylor & Francis: New York, 2003. [Google Scholar]
  25. Kuklja M.Quantum/Chemical Modeling of Energetic Materials. Quantum Chemistry, Energetic Materials; Elsevier, 2014; Vol. 69, p 71. [Google Scholar]
  26. Mei Z.; An Q.; Zhao F.-Q.; Xu S.-Y.; Ju X.-H. Reactive molecular dynamics simulation of thermal decomposition for nano-aluminized explosives. Phys. Chem. Chem. Phys. 2018, 20, 29341–29350. 10.1039/c8cp05006f. [DOI] [PubMed] [Google Scholar]
  27. Plimpton S. Fast parallel algorithms for short-range molecular dynamics. J. Comput. Phys. 1995, 117, 1–19. 10.1006/jcph.1995.1039. [DOI] [Google Scholar]
  28. Acceryls, Inc. Material Studio 7.0; Acceryls Inc.: San Diego, 2013.
  29. Ye C.-C.; An Q.; Zhang W.-Q.; Goddard W. A. Initial Decomposition of HMX Energetic Material from Quantum Molecular Dynamics and the Molecular Structure Transition of beta-HMX to delta-HMX. J. Phys. Chem. C 2019, 123, 9231–9236. 10.1021/acs.jpcc.9b01169. [DOI] [Google Scholar]
  30. Long Y.; Chen J. Systematic Study of the Reaction Kinetics for HMX. J. Phys. Chem. A 2015, 119, 4073–4082. 10.1021/jp509144v. [DOI] [PubMed] [Google Scholar]
  31. Eisenreich N.; Kessler A.; Koleczko A.; Weiser V. On the kinetics of AlH3 decomposition and the subsequent Al oxidation. Int. J. Hydrogen Energy 2014, 39, 6286–6294. 10.1016/j.ijhydene.2014.01.167. [DOI] [Google Scholar]
  32. Stukowski A. Visualization and analysis of atomistic simulation data with OVITO — the open visualization tool. Mater. Sci. Eng. 2010, 18, 015012. 10.1088/0965-0393/18/1/015012. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ao0c02968_si_001.pdf (255.4KB, pdf)

Articles from ACS Omega are provided here courtesy of American Chemical Society

RESOURCES