Special catalytic effects of intermediate-water for rapid shock initiation of β-HMX

Abstract

Quantum-based multiscale calculations were carried out to reveal the rapid shock initiation mechanism of β-HMX (octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine). Tracking the dissociation process of HMX, we discovered a special catalytic action of intermediate-water. During early decomposition of HMX, water and its derivative (·OH) acted as oxidizer, dominated the carbon oxidation reaction, and promoted the oxygen-transport from nitrogen to carbon. Water monomer and its small polymers efficiently transferred proton and hydroxyl moieties between reaction centers, also contributing to the acceleration of the carbon oxidation reaction. The carbon oxidation significantly reduced the dissociation energy barrier of C–N bonds, causing fast and deep cracking of HMX. This water catalytic mechanism may contribute to the illustration of the intrinsic difference in reaction properties between sensitive and insensitive explosives, and shed light on understanding the rapid shock initiation process.

---

1. Introduction

High performance explosives have been widely applied in modern military and commercial fields. Most of the them always contained both the fuel and oxidizer as separate chemical groups within the same molecule,^1,2 enabling them to rapidly decompose and release huge internal energy on timescales of a few nanoseconds to microseconds.^3,4 The traditional explosives could be typically classified into sensitive and insensitive species according to their impact sensitivities.⁵ For sensitive and insensitive explosives, they had distinctly different initial reaction mechanisms and detonation properties, because of different chemical components. The insensitive explosives possessed slower reaction and energy release process, with a wilder reaction zone of several millimeters.⁴ While the sensitive explosives had higher reaction velocity, with a shorter reaction zone of few micrometers.⁶ Despite extensive efforts were devoted to explore the relevant chemical processes,^7–21 the fundamental reaction characteristics were not clearly understood yet. The previous experimental results revealed that, for insensitive explosives, such as TATB (1,3,5-triamino-2,4,6-trinitrobenzene, a carbon-rich explosive), a lot of solid-state diamond or graphite structures were generated during explosion.²² The slow diffusion and coagulation of carbon atoms resulted in a low reaction rate and long reaction zone.^23,24 Recently, Manaa and Reed²⁵ discovered that high concentrative nitrogen-rich heterocyclic clusters were formed during shock decomposition of TATB, which obviously impeded the further reactions to final gas products. It was another mechanism for the slow reactivity of the carbon-rich explosives. As to sensitive explosives, such as HMX (octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine), it seemed that the rapid decomposition and less carbon clustering resulted in a short reaction zone,^26–28 because HMX had a smaller negative oxygen balance comparing with the TATB. However, the key factors responsible for this fast reaction process were not sufficiently illustrated.

In this study, our aim was to discover the intrinsic mechanism of fast decomposition of sensitive explosives under shock loading. HMX (C₄H₈N₈O₈), as a typical specimen of sensitive explosive, had an extraordinary detonation velocity and pressure.^29,30 The early chemical reaction mechanism and main products evolutions had been widely studied by experiments and theories.^{11–18,31–39} Three primary initial reaction models were proposed for HMX thermal decomposition: (I) cleavage of N-nitro bond to produce NO₂,^12,32,37 (II) HONO elimination by electrophilic adsorption of alkyl H and nitro O,³¹ (III) breaking of molecular ring.³⁶ Kuklja and Sharia^16,17,33 calculated the activation energies for these reactions using density functional theory, and illustrated the promoting effects of surface and defect. Long and Chen⁶ obtained 507 kinds of intermediates and 9677 kinds of chemical reactions for HMX thermal decomposition based on molecular dynamics (MD) simulation, and confirmed the main products of N₂, H₂O, CO₂ under 42 GPa up to 50 ps. Goddard III and coworkers³⁷ simulated the decomposition of HMX at different initial densities for 30 ps. They found that NO₂ was firstly produced by cleavage of N-nitro bond. NO, H₂O and N₂ were formed shortly after the dissociation of NO₂. For the shock decomposition of HMX, the molecular ring breaking,^13,15 nitro O¹³ and alkyl H^11,14 direct elimination were identified as the main reaction patterns. He and Chen¹⁵ found that the initial reaction sites closely depended on the angles between chemical bonds and shock direction. Ge et al.¹⁴ explored the early decomposition of HMX at the different shocks of 8, 10 and 11 km s⁻¹ and suggested that the cleavage of N-nitro bond was suppressed at high pressure. Although the earliest reaction sites were studied deeply both in thermal and shock conditions, more details of the reaction mechanisms from first chemical bond breaking to fast decomposition were not sufficiently investigated. More importantly, the complex reaction performance of early intermediate products was not clearly understood yet. A clearly understanding of such reaction properties would not only open the new insight for the fast shock initiation, but also help to uncover the intrinsic difference for the reactivity between sensitive and insensitive explosives.

The quantum-based multiscale molecular dynamics (MD), were performed for β-HMX to reveal its shock decomposition process. Tracking the main small molecules reaction actions, we identified a special catalytic mechanism of intermediate-water (water derived from the early decomposition of HMX) for the rapid decomposition of β-HMX. We first analyzed the microstructure properties of water to more clearly understand its catalytic effects. Intermediate-water and its derivative (·OH) directly participated carbon oxidation reaction and transferred oxygen to carbon. Their monomer and small polymers displayed high efficient transfer capability for proton and hydroxyl and accelerated the carbon oxidation process. All the catalytic behaviors of intermediate-water obviously reduced the breaking energy barrier of C–N bonds within HMX molecules, causing a rapid and deep cracking of HMX.

2. Computational details

The shock decomposition of β-HMX was simulated, using quantum-based MD method in conjunction with multiscale shock technique (MSST)⁴⁰ implemented in the CP2K code. MSST is based on the Navier–Stokes equations, combining molecular dynamics and the one dimension (1D) Euler equations to model the propagation of the shock wave for compressible flow. The Hugoniot relation and Rayleigh line derive from the conservation equations of mass, momentum, and energy across the shock front. The computational cell of the MSST method follows a Lagrangian point through the shock wave, enabling the simulation of the shock wave with really fewer atoms and significantly lower computational cost. The electronic structure calculations were based on the self-consistent-charge density-functional tight-binding method (SCC-DFTB).⁴¹ This method is based on second-order expansion of the Kohn–Sham total energy in density-functional theory (DFT) with respect to charge density fluctuations, which allows one to describe the total energies, atomic forces, and charge transfer in a self-consistent manner. It was applied to study of nitromethane⁴² and TATB²⁵ under pressure, which gave the great consistent results with the other density functional calculations. Besides, the method was also successfully used to describe the lattice parameters and reaction properties of HMX.¹⁴

For illustrating the catalytic effect of intermediate-water on HMX decomposition, the energy barriers of the some key reactions were calculated using density functional theory (DFT) with Gaussian03 (ref. 43) program package. The B3LYP^44,45 functional combined with 6-31G(d) basis set were applied to the geometrical structure optimization for reactants, products and transition states (TS). The electron spin was considered for all open-shell species. The initial TS structures were obtained by reaction potential energy surface (PES) scanning.⁴⁶ Furthermore, harmonic vibrational frequency calculations were performed to identify the various stationary points to be either a minimum or a transition state.⁴⁷ Intrinsic reaction coordinate (IRC)^48,49 calculations were carried out to make sure that each TS connected its reactions and products. The energies discussed below were relative Gibbs free energies (ΔG_298K). And ZPE (zero-point energy) corrected electronic energies (ΔE_0K) were also provided for reference.

Condensed phase β-HMX possessed typical monoclinic molecular crystal structure, with the crystallographic constants of the unit cell a = 6.54 Å, b = 11.05 Å, c = 8.70 Å, and α = 90°, β = 124.3°, γ = 90°.⁵⁰ In this study, a 5 × 2 × 2 supercell structure was constructed. The cell optimization was carried out to obtain the reasonable original configuration. Seeing in Fig. 1, the supercell structure was cut along the (1 0 0) direction, with the high activity N–NO₂ groups exposed on surface. The surface of the model was employed to parallel to the y–z plane. An extra 5 Å vacuum was placed on top of it to eliminate the interaction between molecular layers. The lattice parameters of this supercell were a = 21.253 Å, b = 17.512 Å, c = 34.143 Å, and α = β = γ = 90°. To accelerate the reaction process, two-molecule vacancy was set in center of the model. Meanwhile, the models with six-molecule vacancy and without molecular vacancy were also used for comparison. Firstly, those structures were equilibrated under 300 K for 3 ps using MD–NVT method. And then, uniaxial compression of the shock wave was loaded along the x-axis with MSST method. The fictitious cell mass of 7 × 10⁷ au was employed for MSST calculations. The shock speed of 9 km s⁻¹ was selected, considering the experimental detonation velocity of HMX.²⁹ The initial temperature and pressure were set to 300 K and 0 GPa, respectively. The MD–MSST calculation was run for 25 ps to reveal the complete decomposition process of HMX. For all MD calculations, the target accuracy for the SCF convergence was 10⁻⁶ au, and the time step was 0.1 fs. To identify the stable molecular species, we implemented a post-processing procedure based on bond-length criteria.^6,25

Fig. 1 (a) Molecular structure of β-HMX (C₄H₈N₈O₈), (b) slab model of 5 × 2 × 2 supercell model of β-HMX with (1 0 0) surface (two-molecule vacancy labeled with green ring).

3. Results and discussion

Firstly, the main chemical components involved in decomposition of β-HMX were analyzed within our calculation time-scale. We detected several typical gas products, such as HCN, CO, CO₂, N₂, N₂O, NO₂, NO, H₂O, and some molecular fragments. The population evolutions of these gas products were displayed in Fig. 2 and S1 in ESI.† All the gas products were also observed in the experimental and other theoretical studies on HMX thermal dissociation.^37,51 Although the molecular vacancy could efficiently enhance the dynamical response and promote the earliest reactions,¹⁵ it was really too small to form the ‘hot spot’ and further to affect the detonation initiation like the other simulation results.^52,53 In Fig. S1,† the more fast reaction process was observed with the larger vacancy in condensed phase TATB, but the main products distribution and their population change trends were strictly consistent with each other. For simplification, we just took the two-molecule vacancy model as an example to carefully analyze the HMX decomposition process. Under shock loading, β-HMX molecules were rapidly depleted within 0.6 ps (seeing the inserted panel in Fig. 2). Most molecular rings cracked to form the chainlike molecule residues. The first gas product was NO₂, which produced by homolytic cleavage of N–NO₂ bond. In the following reactions, it was quickly consumed as oxidant. This was consistent with the conclusion drawn from the HMX thermal decomposition theoretical calculation.³⁷ H₂O and NO were produced at almost the same time by the electrophilic adsorption between nitro O and alkyl H. Some deoxygenation N–nitro groups directly converted into N₂O and N₂. The populations of these initial products gradually increased to reach their maximums along with the reaction progresses, for example, at ∼2 ps for water and ∼6 ps for N₂ (seeing Fig. 2). Then, these intermediates were gradually consumed at the following reaction process. According to our MD trajectory analysis, the nitrogen oxides were mainly depleted with secondary reactions to donate the oxygen atoms. Some of initial product N₂ and carbon oxides participated the C–N fragment reorganization. Unexpectedly, we observed an intriguing performance of H₂O at early decomposition stage. Water molecules not only involved in the carbon oxidation process, but also displayed efficient transport capability for hydroxyl and proton.

Fig. 2 Evolutions of main components involved in decomposition of β-HMX.

During the shock compression process, the reaction system reached extremely high temperature and pressure (seeing Fig. S2 in ESI†), which were more than 4000 K and 70 GPa, respectively. Water molecules were rapidly produced under that condition, with the maximum quasi-density (the ratio of mass and system volume) of ∼0.31 g cm⁻³ at 2.0 ps. It was noteworthy that this reaction temperature and pressure were much higher than the critical point of water (T = 647 K, P = 22.1 MPa),^54,55 and the quasi-density of intermediate-water was also close to the critical density of water (0.32 g cm⁻³).⁵⁶ Thus, we could reasonably think that the intermediate-water might have the similar microscopic properties^57–60 and reaction activities^61–63 as the supercritical water (SCW). In fact, before ∼6 ps, the intermediate-water molecules preferred to form the small polymers (such as structure 1–3 in Fig. S3 in ESI†), which were the typical polymeric structures reported in SCW systems.^56,60 And they were proposed to have effective transfer capability for proton.⁶³ However, up to 25 ps, most of water molecules and early small polymers, even some of hydroxyl and proton, were locked within the final water clusters (seeing structure (6) in Fig. S3 in ESI†), diminishing their original reaction activity. We ascribed the water aggregation to extreme high pressure under shock compression, because the similar phenomenon was never detected in the thermal decomposition.^37,64

Oxygen-transport mechanism

It was a widely accepted opinion that, for the traditional energetic materials, such as PETN, TATB and HMX, the explosion process was mainly the rapid oxidation reaction of C–N backbone.^2,51,65 Because the oxidizer and fuels were separately distributed in the explosive molecules, this process could be viewed as the diffusion of oxygen from storage sites (nitro) to fuel sites, such as carbon to form CO and CO₂, or hydrogen to form ·OH and H₂O. Wu and coworkers⁶⁵ suggested that water molecules could efficiently transport oxygen between the reaction centers during dissociation of PETN at Chapman–Jouguet (CJ) state. Here, we further investigated the more details of oxygen-transport process for shock decomposition of HMX. The different mechanisms at different reaction stages were detected.

For the shock decomposition of β-HMX, we first explored the evolutions of the chemical bonds of oxygen, to reveal the sketch of oxygen-transport process. In Fig. 3a, the population of O–N bonds decreased dramatically with reaction proceeding, while that of O–H bonds increased correspondingly. Although the number of O–C bonds also increased, which was obviously lower than that of O–H bonds during the initial 2 ps. It indicated that oxygen atoms were most and first transported from nitrogen to hydrogen. After that, the population of O–H bonds reached their maximum and almost kept constant during the following reaction processes. Whereas oxygen was sequentially transported from nitrogen to carbon, resulting in the number of O–C bonds to continue to increase. Up to ∼3 ps, the O–C and O–N bonds were also reached their dynamical equilibriums, with the number of 160 and 60, respectively. For more clearly understanding about this process, we reversely tracked the oxygen source of the new formed C–O bonds at different reaction periods. It was surprised to find that the earliest C–O bonds were mainly produced by direct N–O–C interaction, while more than 60% new C–O bonds derived from the interactions between carbon and water molecules after 2 ps (seeing Fig. 3b). It denoted that the oxygen was transferred from water molecules to carbon atom. The transport behavior became more significant with water population increase, and even dominated the further carbon oxidation process. Therefore, during early shock decomposition of β-HMX, oxygen atoms were mostly transferred from nitro groups to H atoms to form large number of hydroxyls and H₂O, and then, transported to C atoms through the hydroxyl and H₂O. In essence, intermediate-water acted as an inducer and carrier for the oxygen diffusion between N and C atoms. It was a special catalyst during this period. In comparing with the similar catalytic phenomenon observed in PETN decomposition,⁶⁵ the oxygen-transport behavior of intermediate-water mainly occurred within the first 3 ps. So, we just distinguished this phenomenon as an “early carbon oxidation”. After 3 ps, although most of the new C–O bonds still derived from H₂O, almost an equal amount of oxygen atoms were transferred back to H₂O from the earlier C-oxide components (seeing Fig. 3c). There was no net oxygen transferred from water to carbon. The intermediate-water just promoted the oxygen diffusion between different C-species. We found that this process efficiently promoted the formation of carbon oxides (CO, CO₂), causing a fast increase for their concentrations as showed in Fig. 2.

Fig. 3 Oxygen-transport during β-HMX decomposition process: (a) number of different bonds of O atom, (b) C–O bonds formed by oxidizing of water molecule, (c) oxygen-transport between water and C-species.

A detailed schematic of the oxygen-transport was represented in Fig. 4, which illustrated overall decomposition process of β-HMX to final gas products, like CO and CO₂. Fig. 4a showed the main reaction pathways for the formation of initial water. The nitro O preferred to adsorb H atom to form nitrite group by interacting with vicinity alkyl hydrogen, which finally converted into water molecules with other hydrogen atom. Water also could be produced by adsorption of initial carbonyl oxygen and hydrogen atom. The oxidation of hydrogen displayed obviously advantage than that of carbon. Under explosion condition, the early formed water molecule was not a final stable gas product. Their dissociation occurred much faster than that in ambient condition with different mechanisms. Wu and coworkers⁶⁵ proposed that unimolecular dissociation (H₂O → OH + H) was the most frequent event during PETN decomposition at CJ state. While Schwegler et al.⁶⁶ observed that the bimolecular mechanism (2H₂O → H₃O⁺ + OH⁻) was the main dissociation way in pure water system even at 2000 K. It seemed that the bimolecular dissociation mechanism was favored at lower temperature and higher water concentration. In this study, the reversible reactions of H₂O ↔ ·OH were widely observed since water molecules produced, supplying the active center (·OH) to the reaction system directly. In contrast, water molecules mostly dissociated by interacting with surrounding electrophilic groups to eliminate the H atom, which was neither the unimolecular nor the bimolecular dissociation mechanism as mentioned above. This discrepancy was probably due to extreme high temperature and pressure of the explosion system, with a considerably high activity and close spatial location. By tracking thousands of key reactions within first 6 ps, we found that H₂O and ·OH apparently participated the oxygen transport process. In Fig. 4b, three dominated reaction pathways of oxygen-transport were proposed according to MD trajectory analysis: I, carbon hydroxylation; II, C-carbonyl carboxylation; III, nitrogen hydroxylation. In pathway I, H₂O/·OH interacted with C atom to form C-hydroxyl group. It was followed by two different branch reactions. For the first one, the C–N bond adjacent with hydroxyl directly ruptured to produce the terminal carbonyl (–C–O) (pathway I-1), which further converted into CO. Also, it could interacted with hydroxyl radical to form carboxyl structure (pathway I-3). For the second reaction path, C-hydroxyl reacted with terminal carbonyl to produce ether structure, eliminating H atom and transferring O atom to terminal carbonyl to generate the carboxyl group (pathway I-2). The carboxyl group was also produced by direct adsorption of water (·OH) O atom on carbonyl C as represented in pathway II. We noted that most of the oxidation reactions always produced the typical carboxyl structures labeled with red box in Fig. 4b. It seemed that the carboxylation of the C-carbonyl was the favorable reaction route for the oxygen-transport. H₂O/·OH preferred to transfer oxygen to the partly oxidized carbon atoms. Fig. 4c displayed a typical example of that reaction process. Water O atom firstly adsorbed on the carbonyl C atom to form the analogue carboxyl structure. The water H atom interacted with the other O-species around it, such as hydroxyl and other carbonyl, to be eliminated. The corresponding C–N bond ruptured to generate the terminal carboxyl radical (labeled with red dotted ring). Further dissociation of terminal carboxyl group produced CO₂ molecule. This process confirmed that intermediate-water directly participated the carbon oxidation reaction during β-HMX decomposition.

Fig. 4 Schematic for oxygen-transport during early decomposition of β-HMX, (a) initial formation of water, (b) most favorable pathway of oxygen-transport from water to carbon, (c) a typical example for oxygen-transport form water to carbon. (R represented different C–N fragments).

Besides, water molecule also interacted with the N atom in C–N backbones, to form N-hydroxyl structure (seeing pathway III in Fig. 4b). Like the pathway I-2, N-hydroxyl group could interact with terminal carbonyl, transferring the hydroxyl O to carbonyl C atom and producing carboxyl group (pathway III-1). Or it directly dissociated to form NO. During these oxygen-transport processes, water and ·OH actually acted as an oxidizer and donated oxygen to carbon atom. The initial small molecules, like CO, CO₂ and NO, could not exist stably yet. Some of them reacted with other active radicals to return the oxygen recycle process, or combined into new C–N fragments. For instance, CO and CO₂ preferred to interact with ·OH to form the carbonate groups (labeled with green ring in Fig. 4b), which was the important intermediate for carbon oxides transition.

Proton and hydroxyl transport of intermediate-water

Besides the oxygen transfer actions, the intermediate-water also displayed high performance transport ability for proton and hydroxyl. The details of these processes were typically represented in Fig. 5a. The proton transfer in aqueous solution was widely reported, which was really important for the proton-triggered chemical reactions.⁶³ During β-HMX explosion, protons were frequently produced and transferred among the electrophilic species around them, causing fast dehydrogenation of alkyl carbon. Water molecule and its small polymers acted as the perfect carrier, efficiently inducing and promoting this process. Concretely, the alkyl H atom was first attracted by water molecule to form the hydronium radical (H₃O), and transferred rapidly to another water molecule or the other electrophilic species via the Grotthuss exchange mechanism,⁶⁷ contributing to accelerate chemical reactions. For example, most hydronium radical collided with hydroxyl radical to form new water molecule (seeing Fig. 5b). Some of them also adsorbed on adjacent N atom. More importantly, the alkyl H atom was abstracted and transferred to the nitro O to form nitrite structure (seeing Fig. 5c), which further converted into the high reactive hydroxyl radical. This process substantially supplemented oxygen source into the water and ·OH oxidation system. The proton transfer processes promoted dehydrogenation of alkyl carbon and accelerated the oxygen diffusion from N to H, which all were beneficial to the following carbon oxidation reaction.

Fig. 5 Proton and hydroxyl transport of intermediate-water in β-HMX decomposition, (a) schematic for proton and hydroxyl transport, (b and c) proton transport, (d) hydroxyl transport.

·OH was one of the main oxidants for the supercritical water oxidation process.⁶² It was also identified as the primary free radical carrier involved in hydrocarbon explosion,⁴⁶ promoting the chain transfer and branching reactions. The hydroxyl transport was a reverse process for proton transport (seeing Fig. 5a), which was already observed in the gas explosion.⁶⁸ In our study, hydroxyl radical combined with adjacent water molecule to form hydration-hydroxyl (H–O–H⋯OH). The water H atom was attracted by hydroxyl O to produce a new ·OH and H₂O. The hydroxyl radical was transferred according to this schema till it collided with C atom to form C-hydroxyl groups (seeing Fig. 5d). This process efficiently increased the collision frequency between hydroxyl radical and carbon, promoting the oxygen diffusion from H to C and accelerating the carbon oxidation reaction.

Reaction potential energy curve for C–N bond

Both the oxidizing and transfer actions of intermediate-water all caused rapid carbon oxidation process. The effects of this catalytic behaviors were further studied by comparing the energy barrier of C–N bond breaking. We calculated the reaction potential energy profiles of typical C–N chains within HMX molecule using density functional theory (DFT). In Fig. 6, the breaking of alkyl C–N bond required 17.47 kcal mol⁻¹ activation free energy (R1–TS1). When the alkyl C was oxidized into carbonyl structure (R2), the fission of C–N bond only needed to overcome a free energy barrier of 13.71 kcal mol⁻¹ (R2–TS2). It was considerably lower than that of alkyl C–N (R1) condition. When a hydroxyl adsorbed on alkyl C, the cleavage of C–N bond just required an activation free energy of 9.56 kcal mol⁻¹ (R3–TS3). Obviously, the reaction energy barrier of C–N bond rupture was significantly reduced by carbon oxidation. The hydroxyl-C and carbonyl-C structures had the better reaction activity for the C–N bond breaking than alkyl C. Thus, the carbon oxidation process could efficiently promote the C–N bond dissociation, causing fast terminalization of C-hydroxyl and C-carbonyl species, which further converted into small carbon oxides. It was consistent with the MD trajectory analysis above.

Fig. 6 Potential energy profiles for C–N bond breaking (the R, TS and P represented reactant, transition state and product, respectively).

We detected that the carbon oxidation effects of intermediate-water were mainly based on the form of single water molecules or their derivative (·OH). Their transport actions for proton and hydroxyl were more significant among the small water polymeric groups. Therefore, the catalytic behaviors of intermediate-water just concentrated on the early decomposition of β-HMX molecules, when it randomly distributed within reaction system. These catalytic effects caused fast oxidation of carbon atom, accelerating the decomposition of HMX. However, after the aggregation of water molecules, most of them were bounded within the clusters by quasi hydrogen-bonding net and their reaction activities were diminished. They would not possess significant promoting effect on further reaction of explosion any more.

4. Conclusions

The shock initiation mechanism of β-HMX was systematically investigated using quantum-based molecular dynamics (MD) calculation in conjunction with multiscale shock technique (MSST). The rapid decomposition process of HMX was analyzed to identify the catalytic mechanism of intermediate-water. (I) Oxygen-transport, the nitro O was mostly transferred to alkyl H to form H₂O and ·OH. H₂O (·OH) acted as oxidizer, directly participated carbon oxidation reactions, promoting terminalization of C-oxide species. (II) Proton-transport, intermediate-water efficiently induced and enhanced the hydrogen atom transfer process, accelerating the dehydrogenation of alkyl carbon and promoting oxygen diffusion between nitro and H. (III) Hydroxyl-transport. ·OH radical was transferred by water molecule, efficiently enhancing collision frequency of ·OH and carbon atoms and accelerating the oxygen diffusion from H to C. All the special catalytic behaviors of intermediate-water, resulting in fast oxidation of carbon, which significantly decreased further dissociation energy barrier of β-HMX. Thus, this catalytic mechanism of intermediate-water caused fast and deep cracking of HMX, which was probably responsible for rapid shock initiation process of sensitive explosives.

Acknowledgements

This work was supported by the Science Challenging Program, National Natural Science Foundation of China (No. 11572053), Development Foundation of China Academy of Engineering Physics (No. 2014A0101004).

References

1. W. G. Proud, North Atlantic Treaty Organization, 2013, STO-EN-AVT-214, 203(211–220).
2. H. Gao and J. M. Shreeve, Chem. Rev., 2011, 111, 7377–7436.
3. C. L. Mader, Numerical Modeling of Detonations, University of California: Berkeley, CA, 1979.
4. C. M. Tarver, AIP Conf. Proc., 2006, 845, 1026–1029.
5. R. Wang, H. Xu, Y. Guo, R. Sa and J. n. M. Shreeve, J. Am. Chem. Soc., 2010, 132, 11904–11905.
6. Y. Long and J. Chen, J. Phys. Chem. A, 2015, 119, 4073–4082.
7. C. Ye, Q. An, W. A. Goddard III, T. Cheng, S. V. Zybin and X. Ju, J. Phys. Chem. C, 2015, 119, 2290–2296.
8. E. J. Reed, M. Riad Manaa, L. E. Fried, K. R. Glaesemann and J. D. Joannopoulos, Nat. Phys., 2008, 4, 72–76.
9. Y. Q. Guo, M. Greenfield and E. R. Bernstein, J. Chem. Phys., 2005, 122, 224310.
10. O. Sharia and M. M. Kuklja, AIP Conf. Proc., 2012, 1426, 1223–1226.
11. N.-N. Ge, Y.-K. Wei, Z.-F. Song, X.-R. Chen, G.-F. Ji, F. Zhao and D.-Q. Wei, J. Phys. Chem. B, 2014, 118, 8691–8699.
12. M. R. Manaa, L. E. Fried, C. F. Melius, M. Elstner and T. Frauenheim, J. Phys. Chem. A, 2002, 106, 9024–9029.
13. W. Zhu, H. Huang, H. Huang and H. Xiao, J. Chem. Phys., 2012, 136, 044516.
14. N.-N. Ge, Y.-K. Wei, G.-F. Ji, X.-R. Chen, F. Zhao and D.-Q. Wei, J. Phys. Chem. B, 2012, 116, 13696–13704.
15. Z.-H. He, J. Chen, G.-F. Ji, L.-M. Liu, W.-J. Zhu and Q. Wu, J. Phys. Chem. B, 2015, 119, 10673–10681.
16. O. Sharia and M. M. Kuklja, J. Phys. Chem. B, 2011, 115, 12677–12686.
17. O. Sharia and M. M. Kuklja, J. Am. Chem. Soc., 2012, 134, 11815–11820.
18. O. Sharia, R. Tsyshevsky and M. M. Kuklja, J. Phys. Chem. Lett., 2013, 4, 730–734.
19. M. M. Kuklja, E. V. Stefanovich and A. B. Kunz, J. Chem. Phys., 2000, 112, 3417–3423.
20. V. V. Chaban, E. E. Fileti and O. V. Prezhdo, J. Phys. Chem. Lett., 2015, 6, 913–917.
21. L.-M. Liu, R. Car, A. Selloni, D. M. Dabbs, I. A. Aksay and R. A. Yetter, J. Am. Chem. Soc., 2012, 134, 19011–19016.
22. C. M. Tarver, J. Phys. Chem. A, 1997, 101, 4845–4851.
23. M. S. Shaw and J. D. Johnson, J. Appl. Phys., 1987, 62, 2080–2085.
24. J. A. Viecelli and F. H. Ree, J. Appl. Phys., 1999, 86, 237–248.
25. M. R. Manaa, E. J. Reed, L. E. Fried and N. Goldman, J. Am. Chem. Soc., 2009, 131, 5483–5487.
26. A. Strachan, A. C. T. van Duin, D. Chakraborty, S. Dasgupta and W. A. Goddard, Phys. Rev. Lett., 2003, 91, 098301.
27. Q. Wu, G. Xiong, W. Zhu and H. Xiao, Phys. Chem. Chem. Phys., 2015, 17, 22823–22831.
28. L. Zhang, S. V. Zybin, A. C. T. Van Duin and W. A. Goddard, J. Energ. Mater., 2010, 28, 92–127.
29. C. R. Pulham, D. I. A. Millar, I. D. H. Oswald and W. G. Marshall, High-Pressure Crystallography: From Fundamental Phenomena to Technological Applications-NATO Science for Peace and Security Series B: Physics and Biophysics, Springer, New York, 2010.
30. D. D. Dlott, Annu. Rev. Phys. Chem., 2011, 62, 575–597.
31. D. Chakraborty, R. P. Muller, S. Dasgupta and W. A. Goddard, J. Phys. Chem. A, 2001, 105, 1302–1314.
32. R. Behrens and S. Bulusu, J. Phys. Chem., 1991, 95, 5838–5845.
33. O. Sharia and M. M. Kuklja, J. Phys. Chem. C, 2012, 116, 11077–11081.
34. T. Zhou, H. Song, Y. Liu and F. Huang, Phys. Chem. Chem. Phys., 2014, 16, 13914–13931.
35. M. M. Kuklja, R. V. Tsyshevsky and O. Sharia, J. Am. Chem. Soc., 2014, 136, 13289–13302.
36. K. Prasad, R. A. Yetter and M. D. Smooke, Combust. Flame, 1998, 115, 406–416.
37. L. Zhang, S. V. Zybin, A. C. T. van Duin, S. Dasgupta, W. A. Goddard and E. M. Kober, J. Phys. Chem. A, 2009, 113, 10619–10640.
38. Y. Wen, C. Zhang, X. Xue and X. Long, Phys. Chem. Chem. Phys., 2015, 17, 12013–12022.
39. Q. Peng, Rahul, G. Wang, G.-R. Liu and S. De, Phys. Chem. Chem. Phys., 2014, 16, 19972–19983.
40. E. J. Reed, L. E. Fried and J. D. Joannopoulos, Phys. Rev. Lett., 2003, 90, 235503.
41. M. Elstner, D. Porezag, G. Jungnickel, J. Elsner, M. Haugk, T. Frauenheim, S. Suhai and G. Seifert, Phys. Rev. B: Condens. Matter Mater. Phys., 1998, 58, 7260–7268.
42. D. Margetis, E. Kaxiras, M. Elstner, T. Frauenheim and M. R. Manaa, J. Chem. Phys., 2002, 117, 788–799.
43. M. J. Frisch, Gaussian 03, revision C.02, Gaussian, Inc., Pittsburgh, PA, 2004.
44. C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter Mater. Phys., 1988, 37, 785–789.
45. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
46. Z. He, X.-B. Li, L.-M. Liu and W. Zhu, Fuel, 2014, 124, 85–90.
47. J. Zhang, Z. He, W. Li and Y. Han, RSC Adv., 2012, 2, 4814–4821.
48. C. Gonzalez and H. B. Schlegel, J. Chem. Phys., 1989, 90, 2154–2161.
49. C. Gonzalez and H. B. Schlegel, J. Phys. Chem., 1990, 94, 5523–5527.
50. H. H. Cady, A. C. Larson and D. T. Cromer, Acta Crystallogr., 1963, 16, 617–623.
51. R. Behrens, J. Phys. Chem., 1990, 94, 6706–6718.
52. M. A. Wood, M. J. Cherukara, E. M. Kober and A. Strachan, J. Phys. Chem. C, 2015, 119, 22008–22015.
53. T.-R. Shan, R. R. Wixom and A. P. Thompson, Phys. Rev. B: Condens. Matter Mater. Phys., 2016, 94, 054308.
54. S. Wang, Y. Guo, L. Wang, Y. Wang, D. Xu and H. Ma, Fuel Process. Technol., 2011, 92, 291–297.
55. A. A. Vostrikov, D. Y. Dubov and S. A. Psarov, Tech. Phys. Lett., 2001, 27, 847–849.
56. M. Boero, K. Terakura, T. Ikeshoji, C. C. Liew and M. Parrinello, Phys. Rev. Lett., 2000, 85, 3245–3248.
57. V. Marques Leite dos Santos, F. G. Brady Moreira and R. L. Longo, Chem. Phys. Lett., 2004, 390, 157–161.
58. S. Krishtal, M. Kiselev, Y. Puhovski, T. Kerdcharoen, S. Hannongbua and K. Heinzinger, Zeitschrift für Naturforschung A, 2001, 56, 579–584.
59. P. E. Savage, Chem. Rev., 1999, 99, 603–622.
60. T. Tassaing, P. A. Garrain, D. Bégué and I. Baraille, J. Chem. Phys., 2010, 133, 034103.
61. M. Boero, T. Ikeshoji, C. C. Liew, K. Terakura and M. Parrinello, J. Am. Chem. Soc., 2004, 126, 6280–6286.
62. J. Feng, S. N. V. K. Aki, J. E. Chateauneuf and J. F. Brennecke, J. Am. Chem. Soc., 2002, 124, 6304–6311.
63. M. Boero, T. Ikeshoji and K. Terakura, ChemPhysChem, 2005, 6, 1775–1779.
64. T.-T. Zhou and F.-L. Huang, J. Phys. Chem. B, 2011, 115, 278–287.
65. C. J. Wu, L. E. Fried, L. H. Yang, N. Goldman and S. Bastea, Nat. Chem., 2009, 1, 57–62.
66. E. Schwegler, G. Galli, F. Gygi and R. Q. Hood, Phys. Rev. Lett., 2001, 87, 265501.
67. D. Marx, ChemPhysChem, 2006, 7, 1848–1870.
68. Z.-H. He, X.-B. Li, W.-J. Zhu, L.-M. Liu and G.-F. Ji, RSC Adv., 2014, 4, 35048–35054.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra21384g

---