Characterization and properties of a new insensitive explosive co-crystal composed of trinitrotoluene and pyrene

A new energetic co-crystal of trinitrotoluene (TNT) and pyrene (PYRN) with a 1:1 molar ratio was prepared by a slow solvent evaporation technique. Co-crystal physicochemical properties have also been examined using optical microscopy, powder X-ray diffraction, single crystal X-ray diffraction, and differential scanning calorimetry. The results of single-crystal X-ray diffraction and non-covalent interaction calculations showed that non-covalent interactions (donor-acceptor π - π interaction) govern the structures of the TNT: PYRN co-crystal. The experimental and theoretical outcomes supported each other in the study. Thermal stability, impact sensitivity, and detonation performance of the co-crystal were investigated. DSC measurement indicates that the co-crystal has a melting point of 167°C and a decomposition temperature of 293°C, indicating outstanding thermal stability. The co-crystal was found to be less impact-sensitive than TNT using the BAM fall hammer instrument. Furthermore, the calculated detonation velocity and detonation pressure of the co-crystal are 5.29 km. s −1 and 8.48 G Pa, respectively. As an outcome, the TNT: PYRN co-crystal may be a promising intermediate energy explosive with low sensitivity and, as such, may be a desirable explosive alternative in the future instead of TNT for low-vulnerability formulations.


Introduction
In both military and civilian uses, energetic materials (EMs), such as explosives, propellants, and pyrotechnics, are frequently utilized because of their large-scale chemical energy storage and rapid release capabilities [1][2][3][4][5][6][7]. Alternative technologies will be unable to accomplish the military and civilian improvements enabled by the widespread obtainability of innovative chemical structures with enhanced detonation and propulsion properties. The science is progressing slowly, however, when it comes to the introduction of novel chemical compounds. Around the world, new EMs are constantly being synthesized, but the majority never nd use because of shortcomings like stability and sensitivity. The fact that the current EMs are reasonably good is arguably the biggest obstacle, especially in light of the considerable effort being made in the eld to create new molecules. The expectation for improvement beyond the state of the art is quite high, both in terms of performance and the requirement for cheap and safe production.
This captures the essence of the synthesis challenge [8][9][10][11]. With the recent introduction of cocrystallization to the EMs community, a different approach to producing better and safe energetics without chemical synthesis has been presented, with the aim of partially resolving this issue [12][13].
Indeed, in comparison to the progression of co-crystallization in the pharmaceutical industry, the application of co-crystals as EMs was slow to develop. A substantial reason for this disparity is that the structures of EMs are mainly de ned by nitro groups. The nitro group is a sort of solitary moiety that, in contrast to polar groups found in pharmaceutical substances, offers less opportunities for two energetic molecules to establish hydrogen bonds [14][15]. Fortunately, recent developments in solid-state supramolecular chemistry have signi cantly improved our comprehension of intermolecular interactions.
A few trailblazing scientists have also demonstrated that energetic co-crystals can be realized based on π ... π stacking and C-H ⋯ nitro interactions [16][17][18][19]. The physical and energetic characteristics of a few different co-crystal examples were compiled and contrasted with their constituents in order to clearly show the advantages of energetic co-crystals over their constituents (Table 1) [5,15,[20][21][22][23][24][25][26]. The formation of these materials depended mainly on donor − acceptor π ... π interactions between the electron-poor ring (energetic) and the electron-rich rings (energetic or non-energetic materials). It should be emphasized in light of these results that energetic co-crystals are best suited for creating powerful and safe explosives.
Based on the literature, structural alterations eventually yield unique features in co-crystals when compared to pure components. Co-crystallization successfully modi es properties crucial to EMs, such as density, packing coe cient, melting point, and decomposition temperature, as well as detonation performance (velocity, pressure, and sensitivity).
TNT, or 2,4,6-trinitrotoluene, was developed in 1863 and is still used as an explosive today [27]. It has poor performance and a detrimental environmental impact, yet it is nevertheless employed due to its low production costs and ability to be melt-cast [10]. At the same time, costly and time-consuming upgrading and puri cation activities are required. As a result, new, environmentally safe, non-toxic, and highly EMs must be developed, require careful handling and storage [11].
Co-crystal formation can alter critical parameters such as density, oxygen balance, melting point, decomposition temperature, crystal packing and void space, molecular electrostatic potentials, impact sensitivity, and detonation performance. To investigate this possibility, the co-crystals formed between PYRN and TNT were chosen for experimental and theoretical study and characterisation. The goal of this study is to present not only achievements in new co-crystal as an EMs, but also new insights into the future design of alternative TNT co-crystal in this area.

Experimental
Caution! All of these EMs were manufactured on a milligram scale, and none exploded or detonated during this experiment. Wear protective clothing such as leather suits and earplugs, as well as safety shields, safety glasses, face shields, and leather gloves.

Characterisation techniques
Single crystal X-ray diffraction of the co-crystal was performed on an Agilent Technologies Super Nova diffractometer equipped with an Oxford Cryo Systems cryoprobe. The crystal was kept at T = 120.0 K during data collection. Using Olex2 [29], the structure was solved with the ShelXT [30] structure solution program, using the Intrinsic Phasing solution method. The model was re ned with version 2017/1 of ShelXL [31] using Least Squares minimisation. Mercury CSD (v.4.1) software was used to aid in the structural study of the co-crystal. [32].
PXRD patterns were obtained using the D2 PHASER Advance instrument using Cu-Kα radiation (λ = 1.54439 Å), and an operating voltage and current 40 kV and 40 Ma, respectively. The data were collected over an angle range of 2θ = 5-50. This technique provided an initial screening of samples for formation of co-crystals.
Optical micrographs of all the crystals were taken using an SK2005A polarization microscope.
DSC was performed on a NETZSCH STA 449 F1 Differential Scanning Calorimeter. Approximately 1.5 mg of salt sample was placed in an aluminum pan and the thermal behavior of the sample was studied under a nitrogen purge (30.0 mL/min) at a heating rate of 10°C/min over a temperature range from 25 to 400°C

Energetic performance and sensitivity
The impact sensitivity was measured using a BAM fall hammer (BFH-12). Impact testing was conducted at the Cavendish Laboratory, Cambridge on the TNT, PYRN, and TNT: PYRN. A sample of 40 mm 3 approximate volume was enclosed in anvil device consisting of two coaxial steel cylinders. A load of 10 kg was dropped on to the sample from heights ranging from 10 cm to 100 cm. The 'one-in-six' test procedure was performed to obtain the limiting impact energy for the tested sample. The method conducted throughout is the 30-trial Bruceton method [33,34]. All detonation parameters such as detonation velocity, pressure and oxygen balance, presented in this work were determined from calculations using the program EXPLO5 v6.03 [

2.4.b. Calculation of heat of formation (HOF)
In this paper, all ab initio calculations were carried out using the program package Gaussian 09 (Revision-D.01) [39] and visualized by GaussView 5.0.9 [40]. In order to calculate the energy of formations and the detonation parameters of compounds, the enthalpies (H) were computed according to the atomization energy method (Eq. (2.1); Table 2) [41,42] based on the complete basis set method (CBS-4M), which is included in the Gaussian program. Structure optimization, the distribution electron density in bonds between atoms, electrostatic potential and electronic density were performed by the DFT-B3LYP method with a 6-31G (d, p) basis set in a gas phase. Energy minimisation of PYRN was calculated with the 6-31 + G (d, p) basis set using the account diffusive function. Single crystal X-ray was used as the input for   [43].  3 Enthalpy values obtained from the CBS-4M calculations. 4 Enthalpy of formation (in gas phase).

2.4.c. Visualization of electrostatic potential (ESP) map
An electrostatic potential map (ESP) is a representation of the electrostatic potential at the molecular surface, displayed through a color scheme. ESP charts are useful tools for visualizing the charge distribution within a molecule. The collected data can provide critical information for analyzing the potential electroactive sub-fragments of the target molecules, which can be used to construct chemical reactivity patterns. The electrostatic potential maps are color-coded, where positive (electron-de cient) regions are depicted in blue, and negative (electron-rich) regions are depicted in red. The green regions are considered to have zero electrostatic potential, while yellow regions indicate intermediate electron density.
To summarize, the potential values represented by different colors can be ranked in descending order as red > orange > yellow > green > blue. For this purpose, the preliminary ndings of the geometry optimization analysis were employed to generate the electrostatic potential (ESP) maps of TNT, PYRN, and TNT: PYRN.

2.4.d. Non-covalent interactions (NCIs)
The NCI study was carried out utilizing a computational methodology known as the reduced density gradient (RDG) technique [44,45]. NCI patterns were investigated utilizing the output data from geometry optimization experiments performed using Gaussian 09 at the B3LYP-D3/def2-TZVP level of theory.

Microscope images
The slow evaporation technique was used to allow methanol to evaporate gradually over a two-week period, resulting in the formation of the target TNT: PYRN in dark orange block-shaped crystals (Fig. 1b) (yield: 87%). The crystal shapes, sizes, and colour differ from those of their co-former (TNT in Fig. 1a).
3.1.2. Determination of the TNT: PYRN crystal structure Table 3   The main non-covalent interaction occurs between the electron-rich and electron-poor aromatic rings, as in between the electron-poor π -system of TNT and the electron-rich π -system of PYRN (Fig. 10a-e). By measuring the excess formation enthalpy of the co-formers, the COSMOtherm [48] program was used to anticipate the likelihood of co-crystal formation while selecting electron-rich aromatic ring compounds. The Turbomole program was used to generate the calculations, which used the BP86 density functional with a TZVP34 basis set (BPTZVP-COSMO level of theory). The general rule is that a pair of materials with a ΔpKa (ΔpKa = pKa [protonated base]-pKa [acid]) more negative than − 1 will form a co-crystal (The pKa-value of PYRN is > 15, while the pKa-value of TNT is 16.99) [49]. When the pKa values of TNT and PYRN are evaluated, it becomes clear that new co-crystal is possible. The predominant non-covalent interaction occurs between electron-rich (PYRN) and electron-poor (TNT) aromatic rings. The slip distances and angles between TNT centroids and PYRN centroids were measured to be 3.65 (6) and 89.9 (1). The values are typical of aromatic face-to-face π-interactions (< 4.00 Å) [50] (Fig. 3).
Both mercury software measurements and non-covalent interaction (NCI) calculations show that the π… π interaction is prominent in the formation of TNT: PYRN co-crystal. When Table 4 is investigated, it is evident that several weak H bonds (C-H...O) are formed. These hydrogen bonds are extremely weak formed due to electrostatic interactions. Based on the NCl calculation results Van der Waals and the steric effect have a considerable effect on co-crystal formation. Molecules display a parallel face to-face π -π interaction, with each side stacked alongside another molecule to form an edge-to-face arrangement.

Powder X-ray Diffraction Analysis (PXRD)
PXRD of the TNT, PYRN, and TNT: PYRN is showed in Fig. 4. TNT: PYRN diffraction patterns are clearly distinct from TNT and PYRN diffraction patterns, con rming the formation of a new compound.

Differential Scanning Calorimetry (DSC)
The melting point and decomposition temperature of EMs are essential to determining their thermal stability. Figure 5a shows a DSC thermogram with two endotherms and one exotherm peak for TNT: PYRN. TNT is thought to have formed from an impurity at the rst endotherm, 62.9°C. The melting point of the TNT: PYRN co-crystal was measured to be 167° C, which is greater than the melting point of TNT (83.5°C). The melting point is determined by the lattice energy, or the total intermolecular interaction that occurs in a crystal cell [19]. The TNT: PYRN co-crystal was formed as a consequence of a π… π stacking interaction with weak hydrogen bonds, resulting in a considerably greater lattice energy than TNT. Intermolecular and non-covalent interactions may cause an increase in melting point. Furthermore, increasing the melting point can be bene cial for developing munitions that withstand deformation before explosion [51]. Co-crystallization can cause signi cant changes in another essential physical property of energetic materials. The mass loss for TNT: PYRN is 96.52%. Hirshfeld surface (HS) studies were carried out in order to acquire a better knowledge of the packing motifs and the contributions of the key intermolecular interactions that in uence the molecular architecture in crystalline compounds. HS mapped over the d norm property in two orientations are shown in Fig. 6a. The surfaces are presented transparently to allow visualization of the molecules within the surfaces. Contacts with distances equal to the sum of the van der Waals radii are depicted in white, while contacts with distances less than and more than van der Waals radii are displayed in red and blue, respectively. The corresponding full two-dimensional ngerprint plots (FPs) and resolved to show O…H, H…H, C…H, O…O, and C…C contacts are displayed in Fig. 7. In structure TNT: PYRN, the bright regions labeled represent H…O/O…H contacts associated to C-H…O hydrogen bonds, which are showed as a pair of symmetrical spikes at de + di ≈ 2.4 Å in the corresponding FP plot. These comprise approximately 40% (O…H) of the entire TNT: PYRN HS. The 32% H…H interactions continue to be the most important contributors to the total Hirshfeld surfaces and can be seen as the ''ridge'' of the 2-D ngerprint plots. The pattern of red and blue triangles on the same region of the shape index surface in Fig. 8b is also characteristic of π ... π stacking which also appear as a distinct pale green area at around d e = d i =1.8 Å [52] in the FP.

Electrostatic potential (ESP) maps
The electrostatic potential (ESP) is a measurable and fundamentally important physical property of compounds because it offers information on charge density distribution and molecular reactivity [53].
According to Fig. 9, positive ESP (red regions) are mostly dispersed on the parent skeleton, whereas negative ESP (blue areas) are concentrated on the borders of the molecules, particularly on the nitrogen and oxygen atoms of nitro groups, owing to their higher electronegativity. As can be seen from Fig. 9, the ESP of the intersection for TNT: PYRN changes from red to light blue and a part of the red region changes to blue, indicating that the positive ESP between molecules decreases and a part of the positive ESP changes into negative ESP. All of the following alterations imply that variations in co-crystal charge distribution cause sensitivity changes.

Non-covalent interaction (NCI) plots
The study of non-covalent interactions (NCI) and π-electron distribution structure provides crucial information in terms of the origin of molecular transformation, crystal packing and stability [54,55] where is the electron density, is the sign of the second largest Hessian eigenvalue.
The RDG maps contain multiple spikes. The spikes can be categorized into three groups: hydrogen bond region, vdW interaction region and steric region, marked by blue, green and red circles (Fig. 10d-e), respectively. More scatter dots in each region indicate higher electron density, implying a bigger contribution to overall interactions. Meanwhile, the strength of intermolecular interaction has a positive correlation with electron density in the corresponding region. The higher the electron density is, the stronger the intermolecular interaction is in each region. Then, turning to the v dW interaction region with ρ ranging from − 0.02 to 0.005 a.u., it is seen that three possesses stronger spikes and more points, indicating notable v dW (Van der Waals) interactions (Fig. 10b, d). From NCI plots of TNT: PYRN, it is clear that there are elliptical green slabs coloured light blue between TNT and the PYRN, which can be identi ed weak hydrogen bond interactions. However, as can be seen from Fig. 10b, c,  Although TNT: PYRN is not denser than TNT, the synthesis of the novel compound signi cantly increased the packing coe cient and voids (Table 6). Additionally, higher compactness implies less free volume in the crystal, which promotes molecular degradation and the formation of hotspots, leading to greater insensitivity [57][58][59]. According to Table 6 and Fig. S1a-b, the calculated free space values are consistent with the impact sensitivity, indicating that it is less sensitive than TNT.  [35]. The detonation parameters were computed at the CJ (Chapman-Jouguet condition) point, which is obtained from the Hugoniot curve of the system by its rst derivative.

Conclusion
A novel TNT: PYRN co-crystal explosive with a 1:1 molar ratio was synthesized by the co-crystallization method and characterised. Considering TNT has an electron poor system, it is proposed that compounds with electron rich groups provide e cient formation opportunities with TNT in co-crystallization. The formation of co-crystal is mainly in uenced by strong intermolecular π − π stacking interactions and weak hydrogen bond (C-H···O) interactions. Hirshfeld surfaces, NCI, and ESP analyses were used to nd the noncovalent interaction and found the link between co-crystal structure and sensitivity.      Electrostatic potential surfaces representing charge density distributions of TNT, PYRN, and TNT: PYRN.